llvm-project/llvm/lib/ExecutionEngine/Orc/ObjectLinkingLayer.cpp

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

834 lines
28 KiB
C++
Raw Normal View History

Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
//===------- ObjectLinkingLayer.cpp - JITLink backed ORC ObjectLayer ------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
#include "llvm/ExecutionEngine/Orc/ObjectLinkingLayer.h"
#include "llvm/ADT/Optional.h"
#include "llvm/ExecutionEngine/JITLink/EHFrameSupport.h"
[Orc] Add JITLink debug support plugin for ELF x86-64 Add a new ObjectLinkingLayer plugin `DebugObjectManagerPlugin` and infrastructure to handle creation of `DebugObject`s as well as their registration in OrcTargetProcess. The current implementation only covers ELF on x86-64, but the infrastructure is not limited to that. The journey starts with a new `LinkGraph` / `JITLinkContext` pair being created for a `MaterializationResponsibility` in ORC's `ObjectLinkingLayer`. It sends a `notifyMaterializing()` notification, which is forwarded to all registered plugins. The `DebugObjectManagerPlugin` aims to create a `DebugObject` form the provided target triple and object buffer. (Future implementations might create `DebugObject`s from a `LinkGraph` in other ways.) On success it will track it as the pending `DebugObject` for the `MaterializationResponsibility`. This patch only implements the `ELFDebugObject` for `x86-64` targets. It follows the RuntimeDyld approach for debug object setup: it captures a copy of the input object, parses all section headers and prepares to patch their load-address fields with their final addresses in target memory. It instructs the plugin to report the section load-addresses once they are available. The plugin overrides `modifyPassConfig()` and installs a JITLink post-allocation pass to capture them. Once JITLink emitted the finalized executable, the plugin emits and registers the `DebugObject`. For emission it requests a new `JITLinkMemoryManager::Allocation` with a single read-only segment, copies the object with patched section load-addresses over to working memory and triggers finalization to target memory. For registration, it notifies the `DebugObjectRegistrar` provided in the constructor and stores the previously pending`DebugObject` as registered for the corresponding MaterializationResponsibility. The `DebugObjectRegistrar` registers the `DebugObject` with the target process. `llvm-jitlink` uses the `TPCDebugObjectRegistrar`, which calls `llvm_orc_registerJITLoaderGDBWrapper()` in the target process via `TargetProcessControl` to emit a `jit_code_entry` compatible with the GDB JIT interface [1]. So far the implementation only supports registration and no removal. It appears to me that it wouldn't raise any new design questions, so I left this as an addition for the near future. [1] https://sourceware.org/gdb/current/onlinedocs/gdb/JIT-Interface.html Reviewed By: lhames Differential Revision: https://reviews.llvm.org/D97335
2021-03-02 19:37:48 +08:00
#include "llvm/ExecutionEngine/Orc/DebugObjectManagerPlugin.h"
#include "llvm/Support/MemoryBuffer.h"
#include <string>
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
#include <vector>
#define DEBUG_TYPE "orc"
using namespace llvm;
using namespace llvm::jitlink;
using namespace llvm::orc;
namespace {
class LinkGraphMaterializationUnit : public MaterializationUnit {
public:
static std::unique_ptr<LinkGraphMaterializationUnit>
Create(ObjectLinkingLayer &ObjLinkingLayer, std::unique_ptr<LinkGraph> G) {
auto LGI = scanLinkGraph(ObjLinkingLayer.getExecutionSession(), *G);
return std::unique_ptr<LinkGraphMaterializationUnit>(
new LinkGraphMaterializationUnit(ObjLinkingLayer, std::move(G),
std::move(LGI)));
}
StringRef getName() const override { return G->getName(); }
void materialize(std::unique_ptr<MaterializationResponsibility> MR) override {
ObjLinkingLayer.emit(std::move(MR), std::move(G));
}
private:
static Interface scanLinkGraph(ExecutionSession &ES, LinkGraph &G) {
Interface LGI;
for (auto *Sym : G.defined_symbols()) {
// Skip local symbols.
if (Sym->getScope() == Scope::Local)
continue;
assert(Sym->hasName() && "Anonymous non-local symbol?");
JITSymbolFlags Flags;
if (Sym->getScope() == Scope::Default)
Flags |= JITSymbolFlags::Exported;
if (Sym->isCallable())
Flags |= JITSymbolFlags::Callable;
LGI.SymbolFlags[ES.intern(Sym->getName())] = Flags;
}
if ((G.getTargetTriple().isOSBinFormatMachO() && hasMachOInitSection(G)) ||
(G.getTargetTriple().isOSBinFormatELF() && hasELFInitSection(G)))
LGI.InitSymbol = makeInitSymbol(ES, G);
return LGI;
}
static bool hasMachOInitSection(LinkGraph &G) {
for (auto &Sec : G.sections())
if (Sec.getName() == "__DATA,__obj_selrefs" ||
Sec.getName() == "__DATA,__objc_classlist" ||
Sec.getName() == "__TEXT,__swift5_protos" ||
Sec.getName() == "__TEXT,__swift5_proto" ||
Sec.getName() == "__TEXT,__swift5_types" ||
Sec.getName() == "__DATA,__mod_init_func")
return true;
return false;
}
static bool hasELFInitSection(LinkGraph &G) {
for (auto &Sec : G.sections())
if (Sec.getName() == ".init_array")
return true;
return false;
}
static SymbolStringPtr makeInitSymbol(ExecutionSession &ES, LinkGraph &G) {
std::string InitSymString;
raw_string_ostream(InitSymString)
<< "$." << G.getName() << ".__inits" << Counter++;
return ES.intern(InitSymString);
}
LinkGraphMaterializationUnit(ObjectLinkingLayer &ObjLinkingLayer,
std::unique_ptr<LinkGraph> G, Interface LGI)
: MaterializationUnit(std::move(LGI)), ObjLinkingLayer(ObjLinkingLayer),
G(std::move(G)) {}
void discard(const JITDylib &JD, const SymbolStringPtr &Name) override {
for (auto *Sym : G->defined_symbols())
if (Sym->getName() == *Name) {
assert(Sym->getLinkage() == Linkage::Weak &&
"Discarding non-weak definition");
G->makeExternal(*Sym);
break;
}
}
ObjectLinkingLayer &ObjLinkingLayer;
std::unique_ptr<LinkGraph> G;
static std::atomic<uint64_t> Counter;
};
std::atomic<uint64_t> LinkGraphMaterializationUnit::Counter{0};
} // end anonymous namespace
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
namespace llvm {
namespace orc {
class ObjectLinkingLayerJITLinkContext final : public JITLinkContext {
public:
ObjectLinkingLayerJITLinkContext(
ObjectLinkingLayer &Layer,
std::unique_ptr<MaterializationResponsibility> MR,
std::unique_ptr<MemoryBuffer> ObjBuffer)
: JITLinkContext(&MR->getTargetJITDylib()), Layer(Layer),
MR(std::move(MR)), ObjBuffer(std::move(ObjBuffer)) {}
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
~ObjectLinkingLayerJITLinkContext() {
// If there is an object buffer return function then use it to
// return ownership of the buffer.
if (Layer.ReturnObjectBuffer && ObjBuffer)
Layer.ReturnObjectBuffer(std::move(ObjBuffer));
}
JITLinkMemoryManager &getMemoryManager() override { return Layer.MemMgr; }
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
[Orc] Add JITLink debug support plugin for ELF x86-64 Add a new ObjectLinkingLayer plugin `DebugObjectManagerPlugin` and infrastructure to handle creation of `DebugObject`s as well as their registration in OrcTargetProcess. The current implementation only covers ELF on x86-64, but the infrastructure is not limited to that. The journey starts with a new `LinkGraph` / `JITLinkContext` pair being created for a `MaterializationResponsibility` in ORC's `ObjectLinkingLayer`. It sends a `notifyMaterializing()` notification, which is forwarded to all registered plugins. The `DebugObjectManagerPlugin` aims to create a `DebugObject` form the provided target triple and object buffer. (Future implementations might create `DebugObject`s from a `LinkGraph` in other ways.) On success it will track it as the pending `DebugObject` for the `MaterializationResponsibility`. This patch only implements the `ELFDebugObject` for `x86-64` targets. It follows the RuntimeDyld approach for debug object setup: it captures a copy of the input object, parses all section headers and prepares to patch their load-address fields with their final addresses in target memory. It instructs the plugin to report the section load-addresses once they are available. The plugin overrides `modifyPassConfig()` and installs a JITLink post-allocation pass to capture them. Once JITLink emitted the finalized executable, the plugin emits and registers the `DebugObject`. For emission it requests a new `JITLinkMemoryManager::Allocation` with a single read-only segment, copies the object with patched section load-addresses over to working memory and triggers finalization to target memory. For registration, it notifies the `DebugObjectRegistrar` provided in the constructor and stores the previously pending`DebugObject` as registered for the corresponding MaterializationResponsibility. The `DebugObjectRegistrar` registers the `DebugObject` with the target process. `llvm-jitlink` uses the `TPCDebugObjectRegistrar`, which calls `llvm_orc_registerJITLoaderGDBWrapper()` in the target process via `TargetProcessControl` to emit a `jit_code_entry` compatible with the GDB JIT interface [1]. So far the implementation only supports registration and no removal. It appears to me that it wouldn't raise any new design questions, so I left this as an addition for the near future. [1] https://sourceware.org/gdb/current/onlinedocs/gdb/JIT-Interface.html Reviewed By: lhames Differential Revision: https://reviews.llvm.org/D97335
2021-03-02 19:37:48 +08:00
void notifyMaterializing(LinkGraph &G) {
for (auto &P : Layer.Plugins)
P->notifyMaterializing(*MR, G, *this,
ObjBuffer ? ObjBuffer->getMemBufferRef()
: MemoryBufferRef());
[Orc] Add JITLink debug support plugin for ELF x86-64 Add a new ObjectLinkingLayer plugin `DebugObjectManagerPlugin` and infrastructure to handle creation of `DebugObject`s as well as their registration in OrcTargetProcess. The current implementation only covers ELF on x86-64, but the infrastructure is not limited to that. The journey starts with a new `LinkGraph` / `JITLinkContext` pair being created for a `MaterializationResponsibility` in ORC's `ObjectLinkingLayer`. It sends a `notifyMaterializing()` notification, which is forwarded to all registered plugins. The `DebugObjectManagerPlugin` aims to create a `DebugObject` form the provided target triple and object buffer. (Future implementations might create `DebugObject`s from a `LinkGraph` in other ways.) On success it will track it as the pending `DebugObject` for the `MaterializationResponsibility`. This patch only implements the `ELFDebugObject` for `x86-64` targets. It follows the RuntimeDyld approach for debug object setup: it captures a copy of the input object, parses all section headers and prepares to patch their load-address fields with their final addresses in target memory. It instructs the plugin to report the section load-addresses once they are available. The plugin overrides `modifyPassConfig()` and installs a JITLink post-allocation pass to capture them. Once JITLink emitted the finalized executable, the plugin emits and registers the `DebugObject`. For emission it requests a new `JITLinkMemoryManager::Allocation` with a single read-only segment, copies the object with patched section load-addresses over to working memory and triggers finalization to target memory. For registration, it notifies the `DebugObjectRegistrar` provided in the constructor and stores the previously pending`DebugObject` as registered for the corresponding MaterializationResponsibility. The `DebugObjectRegistrar` registers the `DebugObject` with the target process. `llvm-jitlink` uses the `TPCDebugObjectRegistrar`, which calls `llvm_orc_registerJITLoaderGDBWrapper()` in the target process via `TargetProcessControl` to emit a `jit_code_entry` compatible with the GDB JIT interface [1]. So far the implementation only supports registration and no removal. It appears to me that it wouldn't raise any new design questions, so I left this as an addition for the near future. [1] https://sourceware.org/gdb/current/onlinedocs/gdb/JIT-Interface.html Reviewed By: lhames Differential Revision: https://reviews.llvm.org/D97335
2021-03-02 19:37:48 +08:00
}
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
void notifyFailed(Error Err) override {
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
for (auto &P : Layer.Plugins)
Err = joinErrors(std::move(Err), P->notifyFailed(*MR));
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
Layer.getExecutionSession().reportError(std::move(Err));
MR->failMaterialization();
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
}
[ORC][JITLink] Add support for weak references, and improve handling of static libraries. This patch substantially updates ORCv2's lookup API in order to support weak references, and to better support static archives. Key changes: -- Each symbol being looked for is now associated with a SymbolLookupFlags value. If the associated value is SymbolLookupFlags::RequiredSymbol then the symbol must be defined in one of the JITDylibs being searched (or be able to be generated in one of these JITDylibs via an attached definition generator) or the lookup will fail with an error. If the associated value is SymbolLookupFlags::WeaklyReferencedSymbol then the symbol is permitted to be undefined, in which case it will simply not appear in the resulting SymbolMap if the rest of the lookup succeeds. Since lookup now requires these flags for each symbol, the lookup method now takes an instance of a new SymbolLookupSet type rather than a SymbolNameSet. SymbolLookupSet is a vector-backed set of (name, flags) pairs. Clients are responsible for ensuring that the set property (i.e. unique elements) holds, though this is usually simple and SymbolLookupSet provides convenience methods to support this. -- Lookups now have an associated LookupKind value, which is either LookupKind::Static or LookupKind::DLSym. Definition generators can inspect the lookup kind when determining whether or not to generate new definitions. The StaticLibraryDefinitionGenerator is updated to only pull in new objects from the archive if the lookup kind is Static. This allows lookup to be re-used to emulate dlsym for JIT'd symbols without pulling in new objects from archives (which would not happen in a normal dlsym call). -- JITLink is updated to allow externals to be assigned weak linkage, and weak externals now use the SymbolLookupFlags::WeaklyReferencedSymbol value for lookups. Unresolved weak references will be assigned the default value of zero. Since this patch was modifying the lookup API anyway, it alo replaces all of the "MatchNonExported" boolean arguments with a "JITDylibLookupFlags" enum for readability. If a JITDylib's associated value is JITDylibLookupFlags::MatchExportedSymbolsOnly then the lookup will only match against exported (non-hidden) symbols in that JITDylib. If a JITDylib's associated value is JITDylibLookupFlags::MatchAllSymbols then the lookup will match against any symbol defined in the JITDylib.
2019-11-26 13:57:27 +08:00
void lookup(const LookupMap &Symbols,
std::unique_ptr<JITLinkAsyncLookupContinuation> LC) override {
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
JITDylibSearchOrder LinkOrder;
MR->getTargetJITDylib().withLinkOrderDo(
[&](const JITDylibSearchOrder &LO) { LinkOrder = LO; });
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
auto &ES = Layer.getExecutionSession();
[ORC][JITLink] Add support for weak references, and improve handling of static libraries. This patch substantially updates ORCv2's lookup API in order to support weak references, and to better support static archives. Key changes: -- Each symbol being looked for is now associated with a SymbolLookupFlags value. If the associated value is SymbolLookupFlags::RequiredSymbol then the symbol must be defined in one of the JITDylibs being searched (or be able to be generated in one of these JITDylibs via an attached definition generator) or the lookup will fail with an error. If the associated value is SymbolLookupFlags::WeaklyReferencedSymbol then the symbol is permitted to be undefined, in which case it will simply not appear in the resulting SymbolMap if the rest of the lookup succeeds. Since lookup now requires these flags for each symbol, the lookup method now takes an instance of a new SymbolLookupSet type rather than a SymbolNameSet. SymbolLookupSet is a vector-backed set of (name, flags) pairs. Clients are responsible for ensuring that the set property (i.e. unique elements) holds, though this is usually simple and SymbolLookupSet provides convenience methods to support this. -- Lookups now have an associated LookupKind value, which is either LookupKind::Static or LookupKind::DLSym. Definition generators can inspect the lookup kind when determining whether or not to generate new definitions. The StaticLibraryDefinitionGenerator is updated to only pull in new objects from the archive if the lookup kind is Static. This allows lookup to be re-used to emulate dlsym for JIT'd symbols without pulling in new objects from archives (which would not happen in a normal dlsym call). -- JITLink is updated to allow externals to be assigned weak linkage, and weak externals now use the SymbolLookupFlags::WeaklyReferencedSymbol value for lookups. Unresolved weak references will be assigned the default value of zero. Since this patch was modifying the lookup API anyway, it alo replaces all of the "MatchNonExported" boolean arguments with a "JITDylibLookupFlags" enum for readability. If a JITDylib's associated value is JITDylibLookupFlags::MatchExportedSymbolsOnly then the lookup will only match against exported (non-hidden) symbols in that JITDylib. If a JITDylib's associated value is JITDylibLookupFlags::MatchAllSymbols then the lookup will match against any symbol defined in the JITDylib.
2019-11-26 13:57:27 +08:00
SymbolLookupSet LookupSet;
for (auto &KV : Symbols) {
orc::SymbolLookupFlags LookupFlags;
switch (KV.second) {
case jitlink::SymbolLookupFlags::RequiredSymbol:
LookupFlags = orc::SymbolLookupFlags::RequiredSymbol;
break;
case jitlink::SymbolLookupFlags::WeaklyReferencedSymbol:
LookupFlags = orc::SymbolLookupFlags::WeaklyReferencedSymbol;
break;
}
LookupSet.add(ES.intern(KV.first), LookupFlags);
}
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
// OnResolve -- De-intern the symbols and pass the result to the linker.
auto OnResolve = [LookupContinuation =
std::move(LC)](Expected<SymbolMap> Result) mutable {
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
if (!Result)
LookupContinuation->run(Result.takeError());
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
else {
AsyncLookupResult LR;
for (auto &KV : *Result)
LR[*KV.first] = KV.second;
LookupContinuation->run(std::move(LR));
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
}
};
for (auto &KV : InternalNamedSymbolDeps) {
SymbolDependenceMap InternalDeps;
InternalDeps[&MR->getTargetJITDylib()] = std::move(KV.second);
MR->addDependencies(KV.first, InternalDeps);
}
ES.lookup(LookupKind::Static, LinkOrder, std::move(LookupSet),
[ORC][JITLink] Add support for weak references, and improve handling of static libraries. This patch substantially updates ORCv2's lookup API in order to support weak references, and to better support static archives. Key changes: -- Each symbol being looked for is now associated with a SymbolLookupFlags value. If the associated value is SymbolLookupFlags::RequiredSymbol then the symbol must be defined in one of the JITDylibs being searched (or be able to be generated in one of these JITDylibs via an attached definition generator) or the lookup will fail with an error. If the associated value is SymbolLookupFlags::WeaklyReferencedSymbol then the symbol is permitted to be undefined, in which case it will simply not appear in the resulting SymbolMap if the rest of the lookup succeeds. Since lookup now requires these flags for each symbol, the lookup method now takes an instance of a new SymbolLookupSet type rather than a SymbolNameSet. SymbolLookupSet is a vector-backed set of (name, flags) pairs. Clients are responsible for ensuring that the set property (i.e. unique elements) holds, though this is usually simple and SymbolLookupSet provides convenience methods to support this. -- Lookups now have an associated LookupKind value, which is either LookupKind::Static or LookupKind::DLSym. Definition generators can inspect the lookup kind when determining whether or not to generate new definitions. The StaticLibraryDefinitionGenerator is updated to only pull in new objects from the archive if the lookup kind is Static. This allows lookup to be re-used to emulate dlsym for JIT'd symbols without pulling in new objects from archives (which would not happen in a normal dlsym call). -- JITLink is updated to allow externals to be assigned weak linkage, and weak externals now use the SymbolLookupFlags::WeaklyReferencedSymbol value for lookups. Unresolved weak references will be assigned the default value of zero. Since this patch was modifying the lookup API anyway, it alo replaces all of the "MatchNonExported" boolean arguments with a "JITDylibLookupFlags" enum for readability. If a JITDylib's associated value is JITDylibLookupFlags::MatchExportedSymbolsOnly then the lookup will only match against exported (non-hidden) symbols in that JITDylib. If a JITDylib's associated value is JITDylibLookupFlags::MatchAllSymbols then the lookup will match against any symbol defined in the JITDylib.
2019-11-26 13:57:27 +08:00
SymbolState::Resolved, std::move(OnResolve),
[this](const SymbolDependenceMap &Deps) {
registerDependencies(Deps);
});
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
}
Error notifyResolved(LinkGraph &G) override {
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
auto &ES = Layer.getExecutionSession();
SymbolFlagsMap ExtraSymbolsToClaim;
bool AutoClaim = Layer.AutoClaimObjectSymbols;
SymbolMap InternedResult;
for (auto *Sym : G.defined_symbols())
if (Sym->hasName() && Sym->getScope() != Scope::Local) {
auto InternedName = ES.intern(Sym->getName());
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
JITSymbolFlags Flags;
if (Sym->isCallable())
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
Flags |= JITSymbolFlags::Callable;
if (Sym->getScope() == Scope::Default)
Flags |= JITSymbolFlags::Exported;
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
InternedResult[InternedName] =
JITEvaluatedSymbol(Sym->getAddress().getValue(), Flags);
if (AutoClaim && !MR->getSymbols().count(InternedName)) {
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
assert(!ExtraSymbolsToClaim.count(InternedName) &&
"Duplicate symbol to claim?");
ExtraSymbolsToClaim[InternedName] = Flags;
}
}
for (auto *Sym : G.absolute_symbols())
if (Sym->hasName()) {
auto InternedName = ES.intern(Sym->getName());
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
JITSymbolFlags Flags;
Flags |= JITSymbolFlags::Absolute;
if (Sym->isCallable())
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
Flags |= JITSymbolFlags::Callable;
if (Sym->getLinkage() == Linkage::Weak)
Flags |= JITSymbolFlags::Weak;
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
InternedResult[InternedName] =
JITEvaluatedSymbol(Sym->getAddress().getValue(), Flags);
if (AutoClaim && !MR->getSymbols().count(InternedName)) {
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
assert(!ExtraSymbolsToClaim.count(InternedName) &&
"Duplicate symbol to claim?");
ExtraSymbolsToClaim[InternedName] = Flags;
}
}
if (!ExtraSymbolsToClaim.empty())
if (auto Err = MR->defineMaterializing(ExtraSymbolsToClaim))
return Err;
[ORC] Add generic initializer/deinitializer support. Initializers and deinitializers are used to implement C++ static constructors and destructors, runtime registration for some languages (e.g. with the Objective-C runtime for Objective-C/C++ code) and other tasks that would typically be performed when a shared-object/dylib is loaded or unloaded by a statically compiled program. MCJIT and ORC have historically provided limited support for discovering and running initializers/deinitializers by scanning the llvm.global_ctors and llvm.global_dtors variables and recording the functions to be run. This approach suffers from several drawbacks: (1) It only works for IR inputs, not for object files (including cached JIT'd objects). (2) It only works for initializers described by llvm.global_ctors and llvm.global_dtors, however not all initializers are described in this way (Objective-C, for example, describes initializers via specially named metadata sections). (3) To make the initializer/deinitializer functions described by llvm.global_ctors and llvm.global_dtors searchable they must be promoted to extern linkage, polluting the JIT symbol table (extra care must be taken to ensure this promotion does not result in symbol name clashes). This patch introduces several interdependent changes to ORCv2 to support the construction of new initialization schemes, and includes an implementation of a backwards-compatible llvm.global_ctor/llvm.global_dtor scanning scheme, and a MachO specific scheme that handles Objective-C runtime registration (if the Objective-C runtime is available) enabling execution of LLVM IR compiled from Objective-C and Swift. The major changes included in this patch are: (1) The MaterializationUnit and MaterializationResponsibility classes are extended to describe an optional "initializer" symbol for the module (see the getInitializerSymbol method on each class). The presence or absence of this symbol indicates whether the module contains any initializers or deinitializers. The initializer symbol otherwise behaves like any other: searching for it triggers materialization. (2) A new Platform interface is introduced in llvm/ExecutionEngine/Orc/Core.h which provides the following callback interface: - Error setupJITDylib(JITDylib &JD): Can be used to install standard symbols in JITDylibs upon creation. E.g. __dso_handle. - Error notifyAdding(JITDylib &JD, const MaterializationUnit &MU): Generally used to record initializer symbols. - Error notifyRemoving(JITDylib &JD, VModuleKey K): Used to notify a platform that a module is being removed. Platform implementations can use these callbacks to track outstanding initializers and implement a platform-specific approach for executing them. For example, the MachOPlatform installs a plugin in the JIT linker to scan for both __mod_inits sections (for C++ static constructors) and ObjC metadata sections. If discovered, these are processed in the usual platform order: Objective-C registration is carried out first, then static initializers are executed, ensuring that calls to Objective-C from static initializers will be safe. This patch updates LLJIT to use the new scheme for initialization. Two LLJIT::PlatformSupport classes are implemented: A GenericIR platform and a MachO platform. The GenericIR platform implements a modified version of the previous llvm.global-ctor scraping scheme to provide support for Windows and Linux. LLJIT's MachO platform uses the MachOPlatform class to provide MachO specific initialization as described above. Reviewers: sgraenitz, dblaikie Subscribers: mgorny, hiraditya, mgrang, ributzka, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D74300
2019-12-16 18:50:40 +08:00
{
// Check that InternedResult matches up with MR->getSymbols(), overriding
// flags if requested.
// This guards against faulty transformations / compilers / object caches.
// First check that there aren't any missing symbols.
size_t NumMaterializationSideEffectsOnlySymbols = 0;
SymbolNameVector ExtraSymbols;
SymbolNameVector MissingSymbols;
for (auto &KV : MR->getSymbols()) {
auto I = InternedResult.find(KV.first);
// If this is a materialization-side-effects only symbol then bump
// the counter and make sure it's *not* defined, otherwise make
// sure that it is defined.
if (KV.second.hasMaterializationSideEffectsOnly()) {
++NumMaterializationSideEffectsOnlySymbols;
if (I != InternedResult.end())
ExtraSymbols.push_back(KV.first);
continue;
} else if (I == InternedResult.end())
MissingSymbols.push_back(KV.first);
else if (Layer.OverrideObjectFlags)
I->second.setFlags(KV.second);
}
// If there were missing symbols then report the error.
if (!MissingSymbols.empty())
return make_error<MissingSymbolDefinitions>(
Layer.getExecutionSession().getSymbolStringPool(), G.getName(),
std::move(MissingSymbols));
// If there are more definitions than expected, add them to the
// ExtraSymbols vector.
if (InternedResult.size() >
MR->getSymbols().size() - NumMaterializationSideEffectsOnlySymbols) {
for (auto &KV : InternedResult)
if (!MR->getSymbols().count(KV.first))
ExtraSymbols.push_back(KV.first);
}
// If there were extra definitions then report the error.
if (!ExtraSymbols.empty())
return make_error<UnexpectedSymbolDefinitions>(
Layer.getExecutionSession().getSymbolStringPool(), G.getName(),
std::move(ExtraSymbols));
}
if (auto Err = MR->notifyResolved(InternedResult))
return Err;
Layer.notifyLoaded(*MR);
return Error::success();
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
}
void notifyFinalized(JITLinkMemoryManager::FinalizedAlloc A) override {
if (auto Err = Layer.notifyEmitted(*MR, std::move(A))) {
Layer.getExecutionSession().reportError(std::move(Err));
MR->failMaterialization();
return;
}
if (auto Err = MR->notifyEmitted()) {
Layer.getExecutionSession().reportError(std::move(Err));
MR->failMaterialization();
}
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
}
LinkGraphPassFunction getMarkLivePass(const Triple &TT) const override {
return [this](LinkGraph &G) { return markResponsibilitySymbolsLive(G); };
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
}
Error modifyPassConfig(LinkGraph &LG, PassConfiguration &Config) override {
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
// Add passes to mark duplicate defs as should-discard, and to walk the
// link graph to build the symbol dependence graph.
Config.PrePrunePasses.push_back([this](LinkGraph &G) {
return claimOrExternalizeWeakAndCommonSymbols(G);
});
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
Layer.modifyPassConfig(*MR, LG, Config);
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
Config.PostPrunePasses.push_back(
[this](LinkGraph &G) { return computeNamedSymbolDependencies(G); });
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
return Error::success();
}
private:
[ORC] Improve computeLocalDeps / computeNamedSymbolDependencies performance. The computeNamedSymbolDependencies and computeLocalDeps methods on ObjectLinkingLayerJITLinkContext are responsible for computing, for each symbol in the current MaterializationResponsibility, the set of non-locally-scoped symbols that are depended on. To calculate this we have to consider the effect of chains of dependence through locally scoped symbols in the LinkGraph. E.g. .text .globl foo foo: callq bar ## foo depneds on external 'bar' movq Ltmp1(%rip), %rcx ## foo depends on locally scoped 'Ltmp1' addl (%rcx), %eax retq .data Ltmp1: .quad x ## Ltmp1 depends on external 'x' In this example symbol 'foo' depends directly on 'bar', and indirectly on 'x' via 'Ltmp1', which is locally scoped. Performance of the existing implementations appears to have been mediocre: Based on flame graphs posted by @drmeister (in #jit on the LLVM discord server) the computeLocalDeps function was taking up a substantial amount of time when starting up Clasp (https://github.com/clasp-developers/clasp). This commit attempts to address the performance problems in three ways: 1. Using jitlink::Blocks instead of jitlink::Symbols as the nodes of the dependencies-introduced-by-locally-scoped-symbols graph. Using either Blocks or Symbols as nodes provides the same information, but since there may be more than one locally scoped symbol per block the block-based version of the dependence graph should always be a subgraph of the Symbol-based version, and so faster to operate on. 2. Improved worklist management. The older version of computeLocalDeps used a fixed worklist containing all nodes, and iterated over this list propagating dependencies until no further changes were required. The worklist was not sorted into a useful order before the loop started. The new version uses a variable work-stack, visiting nodes in DFS order and only adding nodes when there is meaningful work to do on them. Compared to the old version the new version avoids revisiting nodes which haven't changed, and I suspect it converges more quickly (due to the DFS ordering). 3. Laziness and caching. Mappings of... jitlink::Symbol* -> Interned Name (as SymbolStringPtr) jitlink::Block* -> Immediate dependencies (as SymbolNameSet) jitlink::Block* -> Transitive dependencies (as SymbolNameSet) are all built lazily and cached while running computeNamedSymbolDependencies. According to @drmeister these changes reduced Clasp startup time in his test setup (averaged over a handful of starts) from 4.8 to 2.8 seconds (with ORC/JITLink linking ~11,000 object files in that time), which seems like enough to justify switching to the new algorithm in the absence of any other perf numbers.
2021-07-08 13:29:47 +08:00
// Symbol name dependencies:
// Internal: Defined in this graph.
// External: Defined externally.
struct BlockSymbolDependencies {
[ORC] Add generic initializer/deinitializer support. Initializers and deinitializers are used to implement C++ static constructors and destructors, runtime registration for some languages (e.g. with the Objective-C runtime for Objective-C/C++ code) and other tasks that would typically be performed when a shared-object/dylib is loaded or unloaded by a statically compiled program. MCJIT and ORC have historically provided limited support for discovering and running initializers/deinitializers by scanning the llvm.global_ctors and llvm.global_dtors variables and recording the functions to be run. This approach suffers from several drawbacks: (1) It only works for IR inputs, not for object files (including cached JIT'd objects). (2) It only works for initializers described by llvm.global_ctors and llvm.global_dtors, however not all initializers are described in this way (Objective-C, for example, describes initializers via specially named metadata sections). (3) To make the initializer/deinitializer functions described by llvm.global_ctors and llvm.global_dtors searchable they must be promoted to extern linkage, polluting the JIT symbol table (extra care must be taken to ensure this promotion does not result in symbol name clashes). This patch introduces several interdependent changes to ORCv2 to support the construction of new initialization schemes, and includes an implementation of a backwards-compatible llvm.global_ctor/llvm.global_dtor scanning scheme, and a MachO specific scheme that handles Objective-C runtime registration (if the Objective-C runtime is available) enabling execution of LLVM IR compiled from Objective-C and Swift. The major changes included in this patch are: (1) The MaterializationUnit and MaterializationResponsibility classes are extended to describe an optional "initializer" symbol for the module (see the getInitializerSymbol method on each class). The presence or absence of this symbol indicates whether the module contains any initializers or deinitializers. The initializer symbol otherwise behaves like any other: searching for it triggers materialization. (2) A new Platform interface is introduced in llvm/ExecutionEngine/Orc/Core.h which provides the following callback interface: - Error setupJITDylib(JITDylib &JD): Can be used to install standard symbols in JITDylibs upon creation. E.g. __dso_handle. - Error notifyAdding(JITDylib &JD, const MaterializationUnit &MU): Generally used to record initializer symbols. - Error notifyRemoving(JITDylib &JD, VModuleKey K): Used to notify a platform that a module is being removed. Platform implementations can use these callbacks to track outstanding initializers and implement a platform-specific approach for executing them. For example, the MachOPlatform installs a plugin in the JIT linker to scan for both __mod_inits sections (for C++ static constructors) and ObjC metadata sections. If discovered, these are processed in the usual platform order: Objective-C registration is carried out first, then static initializers are executed, ensuring that calls to Objective-C from static initializers will be safe. This patch updates LLJIT to use the new scheme for initialization. Two LLJIT::PlatformSupport classes are implemented: A GenericIR platform and a MachO platform. The GenericIR platform implements a modified version of the previous llvm.global-ctor scraping scheme to provide support for Windows and Linux. LLJIT's MachO platform uses the MachOPlatform class to provide MachO specific initialization as described above. Reviewers: sgraenitz, dblaikie Subscribers: mgorny, hiraditya, mgrang, ributzka, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D74300
2019-12-16 18:50:40 +08:00
SymbolNameSet Internal, External;
};
[ORC] Improve computeLocalDeps / computeNamedSymbolDependencies performance. The computeNamedSymbolDependencies and computeLocalDeps methods on ObjectLinkingLayerJITLinkContext are responsible for computing, for each symbol in the current MaterializationResponsibility, the set of non-locally-scoped symbols that are depended on. To calculate this we have to consider the effect of chains of dependence through locally scoped symbols in the LinkGraph. E.g. .text .globl foo foo: callq bar ## foo depneds on external 'bar' movq Ltmp1(%rip), %rcx ## foo depends on locally scoped 'Ltmp1' addl (%rcx), %eax retq .data Ltmp1: .quad x ## Ltmp1 depends on external 'x' In this example symbol 'foo' depends directly on 'bar', and indirectly on 'x' via 'Ltmp1', which is locally scoped. Performance of the existing implementations appears to have been mediocre: Based on flame graphs posted by @drmeister (in #jit on the LLVM discord server) the computeLocalDeps function was taking up a substantial amount of time when starting up Clasp (https://github.com/clasp-developers/clasp). This commit attempts to address the performance problems in three ways: 1. Using jitlink::Blocks instead of jitlink::Symbols as the nodes of the dependencies-introduced-by-locally-scoped-symbols graph. Using either Blocks or Symbols as nodes provides the same information, but since there may be more than one locally scoped symbol per block the block-based version of the dependence graph should always be a subgraph of the Symbol-based version, and so faster to operate on. 2. Improved worklist management. The older version of computeLocalDeps used a fixed worklist containing all nodes, and iterated over this list propagating dependencies until no further changes were required. The worklist was not sorted into a useful order before the loop started. The new version uses a variable work-stack, visiting nodes in DFS order and only adding nodes when there is meaningful work to do on them. Compared to the old version the new version avoids revisiting nodes which haven't changed, and I suspect it converges more quickly (due to the DFS ordering). 3. Laziness and caching. Mappings of... jitlink::Symbol* -> Interned Name (as SymbolStringPtr) jitlink::Block* -> Immediate dependencies (as SymbolNameSet) jitlink::Block* -> Transitive dependencies (as SymbolNameSet) are all built lazily and cached while running computeNamedSymbolDependencies. According to @drmeister these changes reduced Clasp startup time in his test setup (averaged over a handful of starts) from 4.8 to 2.8 seconds (with ORC/JITLink linking ~11,000 object files in that time), which seems like enough to justify switching to the new algorithm in the absence of any other perf numbers.
2021-07-08 13:29:47 +08:00
// Lazily populated map of blocks to BlockSymbolDependencies values.
class BlockDependenciesMap {
public:
BlockDependenciesMap(ExecutionSession &ES,
DenseMap<const Block *, DenseSet<Block *>> BlockDeps)
: ES(ES), BlockDeps(std::move(BlockDeps)) {}
const BlockSymbolDependencies &operator[](const Block &B) {
// Check the cache first.
auto I = BlockTransitiveDepsCache.find(&B);
if (I != BlockTransitiveDepsCache.end())
return I->second;
// No value. Populate the cache.
BlockSymbolDependencies BTDCacheVal;
auto BDI = BlockDeps.find(&B);
assert(BDI != BlockDeps.end() && "No block dependencies");
for (auto *BDep : BDI->second) {
auto &BID = getBlockImmediateDeps(*BDep);
for (auto &ExternalDep : BID.External)
BTDCacheVal.External.insert(ExternalDep);
for (auto &InternalDep : BID.Internal)
BTDCacheVal.Internal.insert(InternalDep);
}
return BlockTransitiveDepsCache
.insert(std::make_pair(&B, std::move(BTDCacheVal)))
.first->second;
}
SymbolStringPtr &getInternedName(Symbol &Sym) {
auto I = NameCache.find(&Sym);
if (I != NameCache.end())
return I->second;
return NameCache.insert(std::make_pair(&Sym, ES.intern(Sym.getName())))
.first->second;
}
private:
BlockSymbolDependencies &getBlockImmediateDeps(Block &B) {
// Check the cache first.
auto I = BlockImmediateDepsCache.find(&B);
if (I != BlockImmediateDepsCache.end())
return I->second;
BlockSymbolDependencies BIDCacheVal;
for (auto &E : B.edges()) {
auto &Tgt = E.getTarget();
if (Tgt.getScope() != Scope::Local) {
if (Tgt.isExternal())
BIDCacheVal.External.insert(getInternedName(Tgt));
else
BIDCacheVal.Internal.insert(getInternedName(Tgt));
}
}
return BlockImmediateDepsCache
.insert(std::make_pair(&B, std::move(BIDCacheVal)))
.first->second;
}
ExecutionSession &ES;
DenseMap<const Block *, DenseSet<Block *>> BlockDeps;
DenseMap<const Symbol *, SymbolStringPtr> NameCache;
DenseMap<const Block *, BlockSymbolDependencies> BlockImmediateDepsCache;
DenseMap<const Block *, BlockSymbolDependencies> BlockTransitiveDepsCache;
};
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
Error claimOrExternalizeWeakAndCommonSymbols(LinkGraph &G) {
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
auto &ES = Layer.getExecutionSession();
SymbolFlagsMap NewSymbolsToClaim;
std::vector<std::pair<SymbolStringPtr, Symbol *>> NameToSym;
auto ProcessSymbol = [&](Symbol *Sym) {
if (Sym->hasName() && Sym->getLinkage() == Linkage::Weak &&
Sym->getScope() != Scope::Local) {
auto Name = ES.intern(Sym->getName());
if (!MR->getSymbols().count(ES.intern(Sym->getName()))) {
JITSymbolFlags SF = JITSymbolFlags::Weak;
if (Sym->getScope() == Scope::Default)
SF |= JITSymbolFlags::Exported;
NewSymbolsToClaim[Name] = SF;
NameToSym.push_back(std::make_pair(std::move(Name), Sym));
}
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
}
};
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
for (auto *Sym : G.defined_symbols())
ProcessSymbol(Sym);
for (auto *Sym : G.absolute_symbols())
ProcessSymbol(Sym);
// Attempt to claim all weak defs that we're not already responsible for.
// This cannot fail -- any clashes will just result in rejection of our
// claim, at which point we'll externalize that symbol.
cantFail(MR->defineMaterializing(std::move(NewSymbolsToClaim)));
for (auto &KV : NameToSym)
if (!MR->getSymbols().count(KV.first))
G.makeExternal(*KV.second);
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
return Error::success();
}
Error markResponsibilitySymbolsLive(LinkGraph &G) const {
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
auto &ES = Layer.getExecutionSession();
for (auto *Sym : G.defined_symbols())
if (Sym->hasName() && MR->getSymbols().count(ES.intern(Sym->getName())))
Sym->setLive(true);
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
return Error::success();
}
Error computeNamedSymbolDependencies(LinkGraph &G) {
auto &ES = MR->getTargetJITDylib().getExecutionSession();
[ORC] Improve computeLocalDeps / computeNamedSymbolDependencies performance. The computeNamedSymbolDependencies and computeLocalDeps methods on ObjectLinkingLayerJITLinkContext are responsible for computing, for each symbol in the current MaterializationResponsibility, the set of non-locally-scoped symbols that are depended on. To calculate this we have to consider the effect of chains of dependence through locally scoped symbols in the LinkGraph. E.g. .text .globl foo foo: callq bar ## foo depneds on external 'bar' movq Ltmp1(%rip), %rcx ## foo depends on locally scoped 'Ltmp1' addl (%rcx), %eax retq .data Ltmp1: .quad x ## Ltmp1 depends on external 'x' In this example symbol 'foo' depends directly on 'bar', and indirectly on 'x' via 'Ltmp1', which is locally scoped. Performance of the existing implementations appears to have been mediocre: Based on flame graphs posted by @drmeister (in #jit on the LLVM discord server) the computeLocalDeps function was taking up a substantial amount of time when starting up Clasp (https://github.com/clasp-developers/clasp). This commit attempts to address the performance problems in three ways: 1. Using jitlink::Blocks instead of jitlink::Symbols as the nodes of the dependencies-introduced-by-locally-scoped-symbols graph. Using either Blocks or Symbols as nodes provides the same information, but since there may be more than one locally scoped symbol per block the block-based version of the dependence graph should always be a subgraph of the Symbol-based version, and so faster to operate on. 2. Improved worklist management. The older version of computeLocalDeps used a fixed worklist containing all nodes, and iterated over this list propagating dependencies until no further changes were required. The worklist was not sorted into a useful order before the loop started. The new version uses a variable work-stack, visiting nodes in DFS order and only adding nodes when there is meaningful work to do on them. Compared to the old version the new version avoids revisiting nodes which haven't changed, and I suspect it converges more quickly (due to the DFS ordering). 3. Laziness and caching. Mappings of... jitlink::Symbol* -> Interned Name (as SymbolStringPtr) jitlink::Block* -> Immediate dependencies (as SymbolNameSet) jitlink::Block* -> Transitive dependencies (as SymbolNameSet) are all built lazily and cached while running computeNamedSymbolDependencies. According to @drmeister these changes reduced Clasp startup time in his test setup (averaged over a handful of starts) from 4.8 to 2.8 seconds (with ORC/JITLink linking ~11,000 object files in that time), which seems like enough to justify switching to the new algorithm in the absence of any other perf numbers.
2021-07-08 13:29:47 +08:00
auto BlockDeps = computeBlockNonLocalDeps(G);
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
[ORC] Add generic initializer/deinitializer support. Initializers and deinitializers are used to implement C++ static constructors and destructors, runtime registration for some languages (e.g. with the Objective-C runtime for Objective-C/C++ code) and other tasks that would typically be performed when a shared-object/dylib is loaded or unloaded by a statically compiled program. MCJIT and ORC have historically provided limited support for discovering and running initializers/deinitializers by scanning the llvm.global_ctors and llvm.global_dtors variables and recording the functions to be run. This approach suffers from several drawbacks: (1) It only works for IR inputs, not for object files (including cached JIT'd objects). (2) It only works for initializers described by llvm.global_ctors and llvm.global_dtors, however not all initializers are described in this way (Objective-C, for example, describes initializers via specially named metadata sections). (3) To make the initializer/deinitializer functions described by llvm.global_ctors and llvm.global_dtors searchable they must be promoted to extern linkage, polluting the JIT symbol table (extra care must be taken to ensure this promotion does not result in symbol name clashes). This patch introduces several interdependent changes to ORCv2 to support the construction of new initialization schemes, and includes an implementation of a backwards-compatible llvm.global_ctor/llvm.global_dtor scanning scheme, and a MachO specific scheme that handles Objective-C runtime registration (if the Objective-C runtime is available) enabling execution of LLVM IR compiled from Objective-C and Swift. The major changes included in this patch are: (1) The MaterializationUnit and MaterializationResponsibility classes are extended to describe an optional "initializer" symbol for the module (see the getInitializerSymbol method on each class). The presence or absence of this symbol indicates whether the module contains any initializers or deinitializers. The initializer symbol otherwise behaves like any other: searching for it triggers materialization. (2) A new Platform interface is introduced in llvm/ExecutionEngine/Orc/Core.h which provides the following callback interface: - Error setupJITDylib(JITDylib &JD): Can be used to install standard symbols in JITDylibs upon creation. E.g. __dso_handle. - Error notifyAdding(JITDylib &JD, const MaterializationUnit &MU): Generally used to record initializer symbols. - Error notifyRemoving(JITDylib &JD, VModuleKey K): Used to notify a platform that a module is being removed. Platform implementations can use these callbacks to track outstanding initializers and implement a platform-specific approach for executing them. For example, the MachOPlatform installs a plugin in the JIT linker to scan for both __mod_inits sections (for C++ static constructors) and ObjC metadata sections. If discovered, these are processed in the usual platform order: Objective-C registration is carried out first, then static initializers are executed, ensuring that calls to Objective-C from static initializers will be safe. This patch updates LLJIT to use the new scheme for initialization. Two LLJIT::PlatformSupport classes are implemented: A GenericIR platform and a MachO platform. The GenericIR platform implements a modified version of the previous llvm.global-ctor scraping scheme to provide support for Windows and Linux. LLJIT's MachO platform uses the MachOPlatform class to provide MachO specific initialization as described above. Reviewers: sgraenitz, dblaikie Subscribers: mgorny, hiraditya, mgrang, ributzka, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D74300
2019-12-16 18:50:40 +08:00
// Compute dependencies for symbols defined in the JITLink graph.
for (auto *Sym : G.defined_symbols()) {
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
// Skip local symbols: we do not track dependencies for these.
if (Sym->getScope() == Scope::Local)
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
continue;
assert(Sym->hasName() &&
"Defined non-local jitlink::Symbol should have a name");
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
[ORC] Improve computeLocalDeps / computeNamedSymbolDependencies performance. The computeNamedSymbolDependencies and computeLocalDeps methods on ObjectLinkingLayerJITLinkContext are responsible for computing, for each symbol in the current MaterializationResponsibility, the set of non-locally-scoped symbols that are depended on. To calculate this we have to consider the effect of chains of dependence through locally scoped symbols in the LinkGraph. E.g. .text .globl foo foo: callq bar ## foo depneds on external 'bar' movq Ltmp1(%rip), %rcx ## foo depends on locally scoped 'Ltmp1' addl (%rcx), %eax retq .data Ltmp1: .quad x ## Ltmp1 depends on external 'x' In this example symbol 'foo' depends directly on 'bar', and indirectly on 'x' via 'Ltmp1', which is locally scoped. Performance of the existing implementations appears to have been mediocre: Based on flame graphs posted by @drmeister (in #jit on the LLVM discord server) the computeLocalDeps function was taking up a substantial amount of time when starting up Clasp (https://github.com/clasp-developers/clasp). This commit attempts to address the performance problems in three ways: 1. Using jitlink::Blocks instead of jitlink::Symbols as the nodes of the dependencies-introduced-by-locally-scoped-symbols graph. Using either Blocks or Symbols as nodes provides the same information, but since there may be more than one locally scoped symbol per block the block-based version of the dependence graph should always be a subgraph of the Symbol-based version, and so faster to operate on. 2. Improved worklist management. The older version of computeLocalDeps used a fixed worklist containing all nodes, and iterated over this list propagating dependencies until no further changes were required. The worklist was not sorted into a useful order before the loop started. The new version uses a variable work-stack, visiting nodes in DFS order and only adding nodes when there is meaningful work to do on them. Compared to the old version the new version avoids revisiting nodes which haven't changed, and I suspect it converges more quickly (due to the DFS ordering). 3. Laziness and caching. Mappings of... jitlink::Symbol* -> Interned Name (as SymbolStringPtr) jitlink::Block* -> Immediate dependencies (as SymbolNameSet) jitlink::Block* -> Transitive dependencies (as SymbolNameSet) are all built lazily and cached while running computeNamedSymbolDependencies. According to @drmeister these changes reduced Clasp startup time in his test setup (averaged over a handful of starts) from 4.8 to 2.8 seconds (with ORC/JITLink linking ~11,000 object files in that time), which seems like enough to justify switching to the new algorithm in the absence of any other perf numbers.
2021-07-08 13:29:47 +08:00
auto &SymDeps = BlockDeps[Sym->getBlock()];
if (SymDeps.External.empty() && SymDeps.Internal.empty())
continue;
auto SymName = ES.intern(Sym->getName());
[ORC] Improve computeLocalDeps / computeNamedSymbolDependencies performance. The computeNamedSymbolDependencies and computeLocalDeps methods on ObjectLinkingLayerJITLinkContext are responsible for computing, for each symbol in the current MaterializationResponsibility, the set of non-locally-scoped symbols that are depended on. To calculate this we have to consider the effect of chains of dependence through locally scoped symbols in the LinkGraph. E.g. .text .globl foo foo: callq bar ## foo depneds on external 'bar' movq Ltmp1(%rip), %rcx ## foo depends on locally scoped 'Ltmp1' addl (%rcx), %eax retq .data Ltmp1: .quad x ## Ltmp1 depends on external 'x' In this example symbol 'foo' depends directly on 'bar', and indirectly on 'x' via 'Ltmp1', which is locally scoped. Performance of the existing implementations appears to have been mediocre: Based on flame graphs posted by @drmeister (in #jit on the LLVM discord server) the computeLocalDeps function was taking up a substantial amount of time when starting up Clasp (https://github.com/clasp-developers/clasp). This commit attempts to address the performance problems in three ways: 1. Using jitlink::Blocks instead of jitlink::Symbols as the nodes of the dependencies-introduced-by-locally-scoped-symbols graph. Using either Blocks or Symbols as nodes provides the same information, but since there may be more than one locally scoped symbol per block the block-based version of the dependence graph should always be a subgraph of the Symbol-based version, and so faster to operate on. 2. Improved worklist management. The older version of computeLocalDeps used a fixed worklist containing all nodes, and iterated over this list propagating dependencies until no further changes were required. The worklist was not sorted into a useful order before the loop started. The new version uses a variable work-stack, visiting nodes in DFS order and only adding nodes when there is meaningful work to do on them. Compared to the old version the new version avoids revisiting nodes which haven't changed, and I suspect it converges more quickly (due to the DFS ordering). 3. Laziness and caching. Mappings of... jitlink::Symbol* -> Interned Name (as SymbolStringPtr) jitlink::Block* -> Immediate dependencies (as SymbolNameSet) jitlink::Block* -> Transitive dependencies (as SymbolNameSet) are all built lazily and cached while running computeNamedSymbolDependencies. According to @drmeister these changes reduced Clasp startup time in his test setup (averaged over a handful of starts) from 4.8 to 2.8 seconds (with ORC/JITLink linking ~11,000 object files in that time), which seems like enough to justify switching to the new algorithm in the absence of any other perf numbers.
2021-07-08 13:29:47 +08:00
if (!SymDeps.External.empty())
ExternalNamedSymbolDeps[SymName] = SymDeps.External;
if (!SymDeps.Internal.empty())
InternalNamedSymbolDeps[SymName] = SymDeps.Internal;
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
}
[ORC] Add generic initializer/deinitializer support. Initializers and deinitializers are used to implement C++ static constructors and destructors, runtime registration for some languages (e.g. with the Objective-C runtime for Objective-C/C++ code) and other tasks that would typically be performed when a shared-object/dylib is loaded or unloaded by a statically compiled program. MCJIT and ORC have historically provided limited support for discovering and running initializers/deinitializers by scanning the llvm.global_ctors and llvm.global_dtors variables and recording the functions to be run. This approach suffers from several drawbacks: (1) It only works for IR inputs, not for object files (including cached JIT'd objects). (2) It only works for initializers described by llvm.global_ctors and llvm.global_dtors, however not all initializers are described in this way (Objective-C, for example, describes initializers via specially named metadata sections). (3) To make the initializer/deinitializer functions described by llvm.global_ctors and llvm.global_dtors searchable they must be promoted to extern linkage, polluting the JIT symbol table (extra care must be taken to ensure this promotion does not result in symbol name clashes). This patch introduces several interdependent changes to ORCv2 to support the construction of new initialization schemes, and includes an implementation of a backwards-compatible llvm.global_ctor/llvm.global_dtor scanning scheme, and a MachO specific scheme that handles Objective-C runtime registration (if the Objective-C runtime is available) enabling execution of LLVM IR compiled from Objective-C and Swift. The major changes included in this patch are: (1) The MaterializationUnit and MaterializationResponsibility classes are extended to describe an optional "initializer" symbol for the module (see the getInitializerSymbol method on each class). The presence or absence of this symbol indicates whether the module contains any initializers or deinitializers. The initializer symbol otherwise behaves like any other: searching for it triggers materialization. (2) A new Platform interface is introduced in llvm/ExecutionEngine/Orc/Core.h which provides the following callback interface: - Error setupJITDylib(JITDylib &JD): Can be used to install standard symbols in JITDylibs upon creation. E.g. __dso_handle. - Error notifyAdding(JITDylib &JD, const MaterializationUnit &MU): Generally used to record initializer symbols. - Error notifyRemoving(JITDylib &JD, VModuleKey K): Used to notify a platform that a module is being removed. Platform implementations can use these callbacks to track outstanding initializers and implement a platform-specific approach for executing them. For example, the MachOPlatform installs a plugin in the JIT linker to scan for both __mod_inits sections (for C++ static constructors) and ObjC metadata sections. If discovered, these are processed in the usual platform order: Objective-C registration is carried out first, then static initializers are executed, ensuring that calls to Objective-C from static initializers will be safe. This patch updates LLJIT to use the new scheme for initialization. Two LLJIT::PlatformSupport classes are implemented: A GenericIR platform and a MachO platform. The GenericIR platform implements a modified version of the previous llvm.global-ctor scraping scheme to provide support for Windows and Linux. LLJIT's MachO platform uses the MachOPlatform class to provide MachO specific initialization as described above. Reviewers: sgraenitz, dblaikie Subscribers: mgorny, hiraditya, mgrang, ributzka, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D74300
2019-12-16 18:50:40 +08:00
for (auto &P : Layer.Plugins) {
[ORC] Improve computeLocalDeps / computeNamedSymbolDependencies performance. The computeNamedSymbolDependencies and computeLocalDeps methods on ObjectLinkingLayerJITLinkContext are responsible for computing, for each symbol in the current MaterializationResponsibility, the set of non-locally-scoped symbols that are depended on. To calculate this we have to consider the effect of chains of dependence through locally scoped symbols in the LinkGraph. E.g. .text .globl foo foo: callq bar ## foo depneds on external 'bar' movq Ltmp1(%rip), %rcx ## foo depends on locally scoped 'Ltmp1' addl (%rcx), %eax retq .data Ltmp1: .quad x ## Ltmp1 depends on external 'x' In this example symbol 'foo' depends directly on 'bar', and indirectly on 'x' via 'Ltmp1', which is locally scoped. Performance of the existing implementations appears to have been mediocre: Based on flame graphs posted by @drmeister (in #jit on the LLVM discord server) the computeLocalDeps function was taking up a substantial amount of time when starting up Clasp (https://github.com/clasp-developers/clasp). This commit attempts to address the performance problems in three ways: 1. Using jitlink::Blocks instead of jitlink::Symbols as the nodes of the dependencies-introduced-by-locally-scoped-symbols graph. Using either Blocks or Symbols as nodes provides the same information, but since there may be more than one locally scoped symbol per block the block-based version of the dependence graph should always be a subgraph of the Symbol-based version, and so faster to operate on. 2. Improved worklist management. The older version of computeLocalDeps used a fixed worklist containing all nodes, and iterated over this list propagating dependencies until no further changes were required. The worklist was not sorted into a useful order before the loop started. The new version uses a variable work-stack, visiting nodes in DFS order and only adding nodes when there is meaningful work to do on them. Compared to the old version the new version avoids revisiting nodes which haven't changed, and I suspect it converges more quickly (due to the DFS ordering). 3. Laziness and caching. Mappings of... jitlink::Symbol* -> Interned Name (as SymbolStringPtr) jitlink::Block* -> Immediate dependencies (as SymbolNameSet) jitlink::Block* -> Transitive dependencies (as SymbolNameSet) are all built lazily and cached while running computeNamedSymbolDependencies. According to @drmeister these changes reduced Clasp startup time in his test setup (averaged over a handful of starts) from 4.8 to 2.8 seconds (with ORC/JITLink linking ~11,000 object files in that time), which seems like enough to justify switching to the new algorithm in the absence of any other perf numbers.
2021-07-08 13:29:47 +08:00
auto SynthDeps = P->getSyntheticSymbolDependencies(*MR);
if (SynthDeps.empty())
[ORC] Add generic initializer/deinitializer support. Initializers and deinitializers are used to implement C++ static constructors and destructors, runtime registration for some languages (e.g. with the Objective-C runtime for Objective-C/C++ code) and other tasks that would typically be performed when a shared-object/dylib is loaded or unloaded by a statically compiled program. MCJIT and ORC have historically provided limited support for discovering and running initializers/deinitializers by scanning the llvm.global_ctors and llvm.global_dtors variables and recording the functions to be run. This approach suffers from several drawbacks: (1) It only works for IR inputs, not for object files (including cached JIT'd objects). (2) It only works for initializers described by llvm.global_ctors and llvm.global_dtors, however not all initializers are described in this way (Objective-C, for example, describes initializers via specially named metadata sections). (3) To make the initializer/deinitializer functions described by llvm.global_ctors and llvm.global_dtors searchable they must be promoted to extern linkage, polluting the JIT symbol table (extra care must be taken to ensure this promotion does not result in symbol name clashes). This patch introduces several interdependent changes to ORCv2 to support the construction of new initialization schemes, and includes an implementation of a backwards-compatible llvm.global_ctor/llvm.global_dtor scanning scheme, and a MachO specific scheme that handles Objective-C runtime registration (if the Objective-C runtime is available) enabling execution of LLVM IR compiled from Objective-C and Swift. The major changes included in this patch are: (1) The MaterializationUnit and MaterializationResponsibility classes are extended to describe an optional "initializer" symbol for the module (see the getInitializerSymbol method on each class). The presence or absence of this symbol indicates whether the module contains any initializers or deinitializers. The initializer symbol otherwise behaves like any other: searching for it triggers materialization. (2) A new Platform interface is introduced in llvm/ExecutionEngine/Orc/Core.h which provides the following callback interface: - Error setupJITDylib(JITDylib &JD): Can be used to install standard symbols in JITDylibs upon creation. E.g. __dso_handle. - Error notifyAdding(JITDylib &JD, const MaterializationUnit &MU): Generally used to record initializer symbols. - Error notifyRemoving(JITDylib &JD, VModuleKey K): Used to notify a platform that a module is being removed. Platform implementations can use these callbacks to track outstanding initializers and implement a platform-specific approach for executing them. For example, the MachOPlatform installs a plugin in the JIT linker to scan for both __mod_inits sections (for C++ static constructors) and ObjC metadata sections. If discovered, these are processed in the usual platform order: Objective-C registration is carried out first, then static initializers are executed, ensuring that calls to Objective-C from static initializers will be safe. This patch updates LLJIT to use the new scheme for initialization. Two LLJIT::PlatformSupport classes are implemented: A GenericIR platform and a MachO platform. The GenericIR platform implements a modified version of the previous llvm.global-ctor scraping scheme to provide support for Windows and Linux. LLJIT's MachO platform uses the MachOPlatform class to provide MachO specific initialization as described above. Reviewers: sgraenitz, dblaikie Subscribers: mgorny, hiraditya, mgrang, ributzka, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D74300
2019-12-16 18:50:40 +08:00
continue;
[ORC] Improve computeLocalDeps / computeNamedSymbolDependencies performance. The computeNamedSymbolDependencies and computeLocalDeps methods on ObjectLinkingLayerJITLinkContext are responsible for computing, for each symbol in the current MaterializationResponsibility, the set of non-locally-scoped symbols that are depended on. To calculate this we have to consider the effect of chains of dependence through locally scoped symbols in the LinkGraph. E.g. .text .globl foo foo: callq bar ## foo depneds on external 'bar' movq Ltmp1(%rip), %rcx ## foo depends on locally scoped 'Ltmp1' addl (%rcx), %eax retq .data Ltmp1: .quad x ## Ltmp1 depends on external 'x' In this example symbol 'foo' depends directly on 'bar', and indirectly on 'x' via 'Ltmp1', which is locally scoped. Performance of the existing implementations appears to have been mediocre: Based on flame graphs posted by @drmeister (in #jit on the LLVM discord server) the computeLocalDeps function was taking up a substantial amount of time when starting up Clasp (https://github.com/clasp-developers/clasp). This commit attempts to address the performance problems in three ways: 1. Using jitlink::Blocks instead of jitlink::Symbols as the nodes of the dependencies-introduced-by-locally-scoped-symbols graph. Using either Blocks or Symbols as nodes provides the same information, but since there may be more than one locally scoped symbol per block the block-based version of the dependence graph should always be a subgraph of the Symbol-based version, and so faster to operate on. 2. Improved worklist management. The older version of computeLocalDeps used a fixed worklist containing all nodes, and iterated over this list propagating dependencies until no further changes were required. The worklist was not sorted into a useful order before the loop started. The new version uses a variable work-stack, visiting nodes in DFS order and only adding nodes when there is meaningful work to do on them. Compared to the old version the new version avoids revisiting nodes which haven't changed, and I suspect it converges more quickly (due to the DFS ordering). 3. Laziness and caching. Mappings of... jitlink::Symbol* -> Interned Name (as SymbolStringPtr) jitlink::Block* -> Immediate dependencies (as SymbolNameSet) jitlink::Block* -> Transitive dependencies (as SymbolNameSet) are all built lazily and cached while running computeNamedSymbolDependencies. According to @drmeister these changes reduced Clasp startup time in his test setup (averaged over a handful of starts) from 4.8 to 2.8 seconds (with ORC/JITLink linking ~11,000 object files in that time), which seems like enough to justify switching to the new algorithm in the absence of any other perf numbers.
2021-07-08 13:29:47 +08:00
DenseSet<Block *> BlockVisited;
for (auto &KV : SynthDeps) {
[ORC] Add generic initializer/deinitializer support. Initializers and deinitializers are used to implement C++ static constructors and destructors, runtime registration for some languages (e.g. with the Objective-C runtime for Objective-C/C++ code) and other tasks that would typically be performed when a shared-object/dylib is loaded or unloaded by a statically compiled program. MCJIT and ORC have historically provided limited support for discovering and running initializers/deinitializers by scanning the llvm.global_ctors and llvm.global_dtors variables and recording the functions to be run. This approach suffers from several drawbacks: (1) It only works for IR inputs, not for object files (including cached JIT'd objects). (2) It only works for initializers described by llvm.global_ctors and llvm.global_dtors, however not all initializers are described in this way (Objective-C, for example, describes initializers via specially named metadata sections). (3) To make the initializer/deinitializer functions described by llvm.global_ctors and llvm.global_dtors searchable they must be promoted to extern linkage, polluting the JIT symbol table (extra care must be taken to ensure this promotion does not result in symbol name clashes). This patch introduces several interdependent changes to ORCv2 to support the construction of new initialization schemes, and includes an implementation of a backwards-compatible llvm.global_ctor/llvm.global_dtor scanning scheme, and a MachO specific scheme that handles Objective-C runtime registration (if the Objective-C runtime is available) enabling execution of LLVM IR compiled from Objective-C and Swift. The major changes included in this patch are: (1) The MaterializationUnit and MaterializationResponsibility classes are extended to describe an optional "initializer" symbol for the module (see the getInitializerSymbol method on each class). The presence or absence of this symbol indicates whether the module contains any initializers or deinitializers. The initializer symbol otherwise behaves like any other: searching for it triggers materialization. (2) A new Platform interface is introduced in llvm/ExecutionEngine/Orc/Core.h which provides the following callback interface: - Error setupJITDylib(JITDylib &JD): Can be used to install standard symbols in JITDylibs upon creation. E.g. __dso_handle. - Error notifyAdding(JITDylib &JD, const MaterializationUnit &MU): Generally used to record initializer symbols. - Error notifyRemoving(JITDylib &JD, VModuleKey K): Used to notify a platform that a module is being removed. Platform implementations can use these callbacks to track outstanding initializers and implement a platform-specific approach for executing them. For example, the MachOPlatform installs a plugin in the JIT linker to scan for both __mod_inits sections (for C++ static constructors) and ObjC metadata sections. If discovered, these are processed in the usual platform order: Objective-C registration is carried out first, then static initializers are executed, ensuring that calls to Objective-C from static initializers will be safe. This patch updates LLJIT to use the new scheme for initialization. Two LLJIT::PlatformSupport classes are implemented: A GenericIR platform and a MachO platform. The GenericIR platform implements a modified version of the previous llvm.global-ctor scraping scheme to provide support for Windows and Linux. LLJIT's MachO platform uses the MachOPlatform class to provide MachO specific initialization as described above. Reviewers: sgraenitz, dblaikie Subscribers: mgorny, hiraditya, mgrang, ributzka, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D74300
2019-12-16 18:50:40 +08:00
auto &Name = KV.first;
[ORC] Improve computeLocalDeps / computeNamedSymbolDependencies performance. The computeNamedSymbolDependencies and computeLocalDeps methods on ObjectLinkingLayerJITLinkContext are responsible for computing, for each symbol in the current MaterializationResponsibility, the set of non-locally-scoped symbols that are depended on. To calculate this we have to consider the effect of chains of dependence through locally scoped symbols in the LinkGraph. E.g. .text .globl foo foo: callq bar ## foo depneds on external 'bar' movq Ltmp1(%rip), %rcx ## foo depends on locally scoped 'Ltmp1' addl (%rcx), %eax retq .data Ltmp1: .quad x ## Ltmp1 depends on external 'x' In this example symbol 'foo' depends directly on 'bar', and indirectly on 'x' via 'Ltmp1', which is locally scoped. Performance of the existing implementations appears to have been mediocre: Based on flame graphs posted by @drmeister (in #jit on the LLVM discord server) the computeLocalDeps function was taking up a substantial amount of time when starting up Clasp (https://github.com/clasp-developers/clasp). This commit attempts to address the performance problems in three ways: 1. Using jitlink::Blocks instead of jitlink::Symbols as the nodes of the dependencies-introduced-by-locally-scoped-symbols graph. Using either Blocks or Symbols as nodes provides the same information, but since there may be more than one locally scoped symbol per block the block-based version of the dependence graph should always be a subgraph of the Symbol-based version, and so faster to operate on. 2. Improved worklist management. The older version of computeLocalDeps used a fixed worklist containing all nodes, and iterated over this list propagating dependencies until no further changes were required. The worklist was not sorted into a useful order before the loop started. The new version uses a variable work-stack, visiting nodes in DFS order and only adding nodes when there is meaningful work to do on them. Compared to the old version the new version avoids revisiting nodes which haven't changed, and I suspect it converges more quickly (due to the DFS ordering). 3. Laziness and caching. Mappings of... jitlink::Symbol* -> Interned Name (as SymbolStringPtr) jitlink::Block* -> Immediate dependencies (as SymbolNameSet) jitlink::Block* -> Transitive dependencies (as SymbolNameSet) are all built lazily and cached while running computeNamedSymbolDependencies. According to @drmeister these changes reduced Clasp startup time in his test setup (averaged over a handful of starts) from 4.8 to 2.8 seconds (with ORC/JITLink linking ~11,000 object files in that time), which seems like enough to justify switching to the new algorithm in the absence of any other perf numbers.
2021-07-08 13:29:47 +08:00
auto &DepsForName = KV.second;
for (auto *Sym : DepsForName) {
if (Sym->getScope() == Scope::Local) {
auto &BDeps = BlockDeps[Sym->getBlock()];
for (auto &S : BDeps.Internal)
InternalNamedSymbolDeps[Name].insert(S);
for (auto &S : BDeps.External)
ExternalNamedSymbolDeps[Name].insert(S);
} else {
if (Sym->isExternal())
ExternalNamedSymbolDeps[Name].insert(
BlockDeps.getInternedName(*Sym));
else
InternalNamedSymbolDeps[Name].insert(
BlockDeps.getInternedName(*Sym));
}
[ORC] Add generic initializer/deinitializer support. Initializers and deinitializers are used to implement C++ static constructors and destructors, runtime registration for some languages (e.g. with the Objective-C runtime for Objective-C/C++ code) and other tasks that would typically be performed when a shared-object/dylib is loaded or unloaded by a statically compiled program. MCJIT and ORC have historically provided limited support for discovering and running initializers/deinitializers by scanning the llvm.global_ctors and llvm.global_dtors variables and recording the functions to be run. This approach suffers from several drawbacks: (1) It only works for IR inputs, not for object files (including cached JIT'd objects). (2) It only works for initializers described by llvm.global_ctors and llvm.global_dtors, however not all initializers are described in this way (Objective-C, for example, describes initializers via specially named metadata sections). (3) To make the initializer/deinitializer functions described by llvm.global_ctors and llvm.global_dtors searchable they must be promoted to extern linkage, polluting the JIT symbol table (extra care must be taken to ensure this promotion does not result in symbol name clashes). This patch introduces several interdependent changes to ORCv2 to support the construction of new initialization schemes, and includes an implementation of a backwards-compatible llvm.global_ctor/llvm.global_dtor scanning scheme, and a MachO specific scheme that handles Objective-C runtime registration (if the Objective-C runtime is available) enabling execution of LLVM IR compiled from Objective-C and Swift. The major changes included in this patch are: (1) The MaterializationUnit and MaterializationResponsibility classes are extended to describe an optional "initializer" symbol for the module (see the getInitializerSymbol method on each class). The presence or absence of this symbol indicates whether the module contains any initializers or deinitializers. The initializer symbol otherwise behaves like any other: searching for it triggers materialization. (2) A new Platform interface is introduced in llvm/ExecutionEngine/Orc/Core.h which provides the following callback interface: - Error setupJITDylib(JITDylib &JD): Can be used to install standard symbols in JITDylibs upon creation. E.g. __dso_handle. - Error notifyAdding(JITDylib &JD, const MaterializationUnit &MU): Generally used to record initializer symbols. - Error notifyRemoving(JITDylib &JD, VModuleKey K): Used to notify a platform that a module is being removed. Platform implementations can use these callbacks to track outstanding initializers and implement a platform-specific approach for executing them. For example, the MachOPlatform installs a plugin in the JIT linker to scan for both __mod_inits sections (for C++ static constructors) and ObjC metadata sections. If discovered, these are processed in the usual platform order: Objective-C registration is carried out first, then static initializers are executed, ensuring that calls to Objective-C from static initializers will be safe. This patch updates LLJIT to use the new scheme for initialization. Two LLJIT::PlatformSupport classes are implemented: A GenericIR platform and a MachO platform. The GenericIR platform implements a modified version of the previous llvm.global-ctor scraping scheme to provide support for Windows and Linux. LLJIT's MachO platform uses the MachOPlatform class to provide MachO specific initialization as described above. Reviewers: sgraenitz, dblaikie Subscribers: mgorny, hiraditya, mgrang, ributzka, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D74300
2019-12-16 18:50:40 +08:00
}
}
}
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
return Error::success();
}
[ORC] Improve computeLocalDeps / computeNamedSymbolDependencies performance. The computeNamedSymbolDependencies and computeLocalDeps methods on ObjectLinkingLayerJITLinkContext are responsible for computing, for each symbol in the current MaterializationResponsibility, the set of non-locally-scoped symbols that are depended on. To calculate this we have to consider the effect of chains of dependence through locally scoped symbols in the LinkGraph. E.g. .text .globl foo foo: callq bar ## foo depneds on external 'bar' movq Ltmp1(%rip), %rcx ## foo depends on locally scoped 'Ltmp1' addl (%rcx), %eax retq .data Ltmp1: .quad x ## Ltmp1 depends on external 'x' In this example symbol 'foo' depends directly on 'bar', and indirectly on 'x' via 'Ltmp1', which is locally scoped. Performance of the existing implementations appears to have been mediocre: Based on flame graphs posted by @drmeister (in #jit on the LLVM discord server) the computeLocalDeps function was taking up a substantial amount of time when starting up Clasp (https://github.com/clasp-developers/clasp). This commit attempts to address the performance problems in three ways: 1. Using jitlink::Blocks instead of jitlink::Symbols as the nodes of the dependencies-introduced-by-locally-scoped-symbols graph. Using either Blocks or Symbols as nodes provides the same information, but since there may be more than one locally scoped symbol per block the block-based version of the dependence graph should always be a subgraph of the Symbol-based version, and so faster to operate on. 2. Improved worklist management. The older version of computeLocalDeps used a fixed worklist containing all nodes, and iterated over this list propagating dependencies until no further changes were required. The worklist was not sorted into a useful order before the loop started. The new version uses a variable work-stack, visiting nodes in DFS order and only adding nodes when there is meaningful work to do on them. Compared to the old version the new version avoids revisiting nodes which haven't changed, and I suspect it converges more quickly (due to the DFS ordering). 3. Laziness and caching. Mappings of... jitlink::Symbol* -> Interned Name (as SymbolStringPtr) jitlink::Block* -> Immediate dependencies (as SymbolNameSet) jitlink::Block* -> Transitive dependencies (as SymbolNameSet) are all built lazily and cached while running computeNamedSymbolDependencies. According to @drmeister these changes reduced Clasp startup time in his test setup (averaged over a handful of starts) from 4.8 to 2.8 seconds (with ORC/JITLink linking ~11,000 object files in that time), which seems like enough to justify switching to the new algorithm in the absence of any other perf numbers.
2021-07-08 13:29:47 +08:00
BlockDependenciesMap computeBlockNonLocalDeps(LinkGraph &G) {
// First calculate the reachable-via-non-local-symbol blocks for each block.
struct BlockInfo {
DenseSet<Block *> Dependencies;
DenseSet<Block *> Dependants;
bool DependenciesChanged = true;
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
};
[ORC] Improve computeLocalDeps / computeNamedSymbolDependencies performance. The computeNamedSymbolDependencies and computeLocalDeps methods on ObjectLinkingLayerJITLinkContext are responsible for computing, for each symbol in the current MaterializationResponsibility, the set of non-locally-scoped symbols that are depended on. To calculate this we have to consider the effect of chains of dependence through locally scoped symbols in the LinkGraph. E.g. .text .globl foo foo: callq bar ## foo depneds on external 'bar' movq Ltmp1(%rip), %rcx ## foo depends on locally scoped 'Ltmp1' addl (%rcx), %eax retq .data Ltmp1: .quad x ## Ltmp1 depends on external 'x' In this example symbol 'foo' depends directly on 'bar', and indirectly on 'x' via 'Ltmp1', which is locally scoped. Performance of the existing implementations appears to have been mediocre: Based on flame graphs posted by @drmeister (in #jit on the LLVM discord server) the computeLocalDeps function was taking up a substantial amount of time when starting up Clasp (https://github.com/clasp-developers/clasp). This commit attempts to address the performance problems in three ways: 1. Using jitlink::Blocks instead of jitlink::Symbols as the nodes of the dependencies-introduced-by-locally-scoped-symbols graph. Using either Blocks or Symbols as nodes provides the same information, but since there may be more than one locally scoped symbol per block the block-based version of the dependence graph should always be a subgraph of the Symbol-based version, and so faster to operate on. 2. Improved worklist management. The older version of computeLocalDeps used a fixed worklist containing all nodes, and iterated over this list propagating dependencies until no further changes were required. The worklist was not sorted into a useful order before the loop started. The new version uses a variable work-stack, visiting nodes in DFS order and only adding nodes when there is meaningful work to do on them. Compared to the old version the new version avoids revisiting nodes which haven't changed, and I suspect it converges more quickly (due to the DFS ordering). 3. Laziness and caching. Mappings of... jitlink::Symbol* -> Interned Name (as SymbolStringPtr) jitlink::Block* -> Immediate dependencies (as SymbolNameSet) jitlink::Block* -> Transitive dependencies (as SymbolNameSet) are all built lazily and cached while running computeNamedSymbolDependencies. According to @drmeister these changes reduced Clasp startup time in his test setup (averaged over a handful of starts) from 4.8 to 2.8 seconds (with ORC/JITLink linking ~11,000 object files in that time), which seems like enough to justify switching to the new algorithm in the absence of any other perf numbers.
2021-07-08 13:29:47 +08:00
DenseMap<Block *, BlockInfo> BlockInfos;
SmallVector<Block *> WorkList;
// Pre-allocate map entries. This prevents any iterator/reference
// invalidation in the next loop.
for (auto *B : G.blocks())
(void)BlockInfos[B];
// Build initial worklist, record block dependencies/dependants and
// non-local symbol dependencies.
for (auto *B : G.blocks()) {
auto &BI = BlockInfos[B];
for (auto &E : B->edges()) {
if (E.getTarget().getScope() == Scope::Local) {
auto &TgtB = E.getTarget().getBlock();
if (&TgtB != B) {
BI.Dependencies.insert(&TgtB);
BlockInfos[&TgtB].Dependants.insert(B);
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
}
}
}
[ORC] Improve computeLocalDeps / computeNamedSymbolDependencies performance. The computeNamedSymbolDependencies and computeLocalDeps methods on ObjectLinkingLayerJITLinkContext are responsible for computing, for each symbol in the current MaterializationResponsibility, the set of non-locally-scoped symbols that are depended on. To calculate this we have to consider the effect of chains of dependence through locally scoped symbols in the LinkGraph. E.g. .text .globl foo foo: callq bar ## foo depneds on external 'bar' movq Ltmp1(%rip), %rcx ## foo depends on locally scoped 'Ltmp1' addl (%rcx), %eax retq .data Ltmp1: .quad x ## Ltmp1 depends on external 'x' In this example symbol 'foo' depends directly on 'bar', and indirectly on 'x' via 'Ltmp1', which is locally scoped. Performance of the existing implementations appears to have been mediocre: Based on flame graphs posted by @drmeister (in #jit on the LLVM discord server) the computeLocalDeps function was taking up a substantial amount of time when starting up Clasp (https://github.com/clasp-developers/clasp). This commit attempts to address the performance problems in three ways: 1. Using jitlink::Blocks instead of jitlink::Symbols as the nodes of the dependencies-introduced-by-locally-scoped-symbols graph. Using either Blocks or Symbols as nodes provides the same information, but since there may be more than one locally scoped symbol per block the block-based version of the dependence graph should always be a subgraph of the Symbol-based version, and so faster to operate on. 2. Improved worklist management. The older version of computeLocalDeps used a fixed worklist containing all nodes, and iterated over this list propagating dependencies until no further changes were required. The worklist was not sorted into a useful order before the loop started. The new version uses a variable work-stack, visiting nodes in DFS order and only adding nodes when there is meaningful work to do on them. Compared to the old version the new version avoids revisiting nodes which haven't changed, and I suspect it converges more quickly (due to the DFS ordering). 3. Laziness and caching. Mappings of... jitlink::Symbol* -> Interned Name (as SymbolStringPtr) jitlink::Block* -> Immediate dependencies (as SymbolNameSet) jitlink::Block* -> Transitive dependencies (as SymbolNameSet) are all built lazily and cached while running computeNamedSymbolDependencies. According to @drmeister these changes reduced Clasp startup time in his test setup (averaged over a handful of starts) from 4.8 to 2.8 seconds (with ORC/JITLink linking ~11,000 object files in that time), which seems like enough to justify switching to the new algorithm in the absence of any other perf numbers.
2021-07-08 13:29:47 +08:00
// If this node has both dependants and dependencies then add it to the
// worklist to propagate the dependencies to the dependants.
if (!BI.Dependants.empty() && !BI.Dependencies.empty())
WorkList.push_back(B);
}
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
[ORC] Improve computeLocalDeps / computeNamedSymbolDependencies performance. The computeNamedSymbolDependencies and computeLocalDeps methods on ObjectLinkingLayerJITLinkContext are responsible for computing, for each symbol in the current MaterializationResponsibility, the set of non-locally-scoped symbols that are depended on. To calculate this we have to consider the effect of chains of dependence through locally scoped symbols in the LinkGraph. E.g. .text .globl foo foo: callq bar ## foo depneds on external 'bar' movq Ltmp1(%rip), %rcx ## foo depends on locally scoped 'Ltmp1' addl (%rcx), %eax retq .data Ltmp1: .quad x ## Ltmp1 depends on external 'x' In this example symbol 'foo' depends directly on 'bar', and indirectly on 'x' via 'Ltmp1', which is locally scoped. Performance of the existing implementations appears to have been mediocre: Based on flame graphs posted by @drmeister (in #jit on the LLVM discord server) the computeLocalDeps function was taking up a substantial amount of time when starting up Clasp (https://github.com/clasp-developers/clasp). This commit attempts to address the performance problems in three ways: 1. Using jitlink::Blocks instead of jitlink::Symbols as the nodes of the dependencies-introduced-by-locally-scoped-symbols graph. Using either Blocks or Symbols as nodes provides the same information, but since there may be more than one locally scoped symbol per block the block-based version of the dependence graph should always be a subgraph of the Symbol-based version, and so faster to operate on. 2. Improved worklist management. The older version of computeLocalDeps used a fixed worklist containing all nodes, and iterated over this list propagating dependencies until no further changes were required. The worklist was not sorted into a useful order before the loop started. The new version uses a variable work-stack, visiting nodes in DFS order and only adding nodes when there is meaningful work to do on them. Compared to the old version the new version avoids revisiting nodes which haven't changed, and I suspect it converges more quickly (due to the DFS ordering). 3. Laziness and caching. Mappings of... jitlink::Symbol* -> Interned Name (as SymbolStringPtr) jitlink::Block* -> Immediate dependencies (as SymbolNameSet) jitlink::Block* -> Transitive dependencies (as SymbolNameSet) are all built lazily and cached while running computeNamedSymbolDependencies. According to @drmeister these changes reduced Clasp startup time in his test setup (averaged over a handful of starts) from 4.8 to 2.8 seconds (with ORC/JITLink linking ~11,000 object files in that time), which seems like enough to justify switching to the new algorithm in the absence of any other perf numbers.
2021-07-08 13:29:47 +08:00
// Propagate block-level dependencies through the block-dependence graph.
while (!WorkList.empty()) {
2021-09-20 04:44:23 +08:00
auto *B = WorkList.pop_back_val();
[ORC] Improve computeLocalDeps / computeNamedSymbolDependencies performance. The computeNamedSymbolDependencies and computeLocalDeps methods on ObjectLinkingLayerJITLinkContext are responsible for computing, for each symbol in the current MaterializationResponsibility, the set of non-locally-scoped symbols that are depended on. To calculate this we have to consider the effect of chains of dependence through locally scoped symbols in the LinkGraph. E.g. .text .globl foo foo: callq bar ## foo depneds on external 'bar' movq Ltmp1(%rip), %rcx ## foo depends on locally scoped 'Ltmp1' addl (%rcx), %eax retq .data Ltmp1: .quad x ## Ltmp1 depends on external 'x' In this example symbol 'foo' depends directly on 'bar', and indirectly on 'x' via 'Ltmp1', which is locally scoped. Performance of the existing implementations appears to have been mediocre: Based on flame graphs posted by @drmeister (in #jit on the LLVM discord server) the computeLocalDeps function was taking up a substantial amount of time when starting up Clasp (https://github.com/clasp-developers/clasp). This commit attempts to address the performance problems in three ways: 1. Using jitlink::Blocks instead of jitlink::Symbols as the nodes of the dependencies-introduced-by-locally-scoped-symbols graph. Using either Blocks or Symbols as nodes provides the same information, but since there may be more than one locally scoped symbol per block the block-based version of the dependence graph should always be a subgraph of the Symbol-based version, and so faster to operate on. 2. Improved worklist management. The older version of computeLocalDeps used a fixed worklist containing all nodes, and iterated over this list propagating dependencies until no further changes were required. The worklist was not sorted into a useful order before the loop started. The new version uses a variable work-stack, visiting nodes in DFS order and only adding nodes when there is meaningful work to do on them. Compared to the old version the new version avoids revisiting nodes which haven't changed, and I suspect it converges more quickly (due to the DFS ordering). 3. Laziness and caching. Mappings of... jitlink::Symbol* -> Interned Name (as SymbolStringPtr) jitlink::Block* -> Immediate dependencies (as SymbolNameSet) jitlink::Block* -> Transitive dependencies (as SymbolNameSet) are all built lazily and cached while running computeNamedSymbolDependencies. According to @drmeister these changes reduced Clasp startup time in his test setup (averaged over a handful of starts) from 4.8 to 2.8 seconds (with ORC/JITLink linking ~11,000 object files in that time), which seems like enough to justify switching to the new algorithm in the absence of any other perf numbers.
2021-07-08 13:29:47 +08:00
auto &BI = BlockInfos[B];
assert(BI.DependenciesChanged &&
"Block in worklist has unchanged dependencies");
BI.DependenciesChanged = false;
for (auto *Dependant : BI.Dependants) {
auto &DependantBI = BlockInfos[Dependant];
for (auto *Dependency : BI.Dependencies) {
if (Dependant != Dependency &&
DependantBI.Dependencies.insert(Dependency).second)
if (!DependantBI.DependenciesChanged) {
DependantBI.DependenciesChanged = true;
WorkList.push_back(Dependant);
}
}
[ORC] Add generic initializer/deinitializer support. Initializers and deinitializers are used to implement C++ static constructors and destructors, runtime registration for some languages (e.g. with the Objective-C runtime for Objective-C/C++ code) and other tasks that would typically be performed when a shared-object/dylib is loaded or unloaded by a statically compiled program. MCJIT and ORC have historically provided limited support for discovering and running initializers/deinitializers by scanning the llvm.global_ctors and llvm.global_dtors variables and recording the functions to be run. This approach suffers from several drawbacks: (1) It only works for IR inputs, not for object files (including cached JIT'd objects). (2) It only works for initializers described by llvm.global_ctors and llvm.global_dtors, however not all initializers are described in this way (Objective-C, for example, describes initializers via specially named metadata sections). (3) To make the initializer/deinitializer functions described by llvm.global_ctors and llvm.global_dtors searchable they must be promoted to extern linkage, polluting the JIT symbol table (extra care must be taken to ensure this promotion does not result in symbol name clashes). This patch introduces several interdependent changes to ORCv2 to support the construction of new initialization schemes, and includes an implementation of a backwards-compatible llvm.global_ctor/llvm.global_dtor scanning scheme, and a MachO specific scheme that handles Objective-C runtime registration (if the Objective-C runtime is available) enabling execution of LLVM IR compiled from Objective-C and Swift. The major changes included in this patch are: (1) The MaterializationUnit and MaterializationResponsibility classes are extended to describe an optional "initializer" symbol for the module (see the getInitializerSymbol method on each class). The presence or absence of this symbol indicates whether the module contains any initializers or deinitializers. The initializer symbol otherwise behaves like any other: searching for it triggers materialization. (2) A new Platform interface is introduced in llvm/ExecutionEngine/Orc/Core.h which provides the following callback interface: - Error setupJITDylib(JITDylib &JD): Can be used to install standard symbols in JITDylibs upon creation. E.g. __dso_handle. - Error notifyAdding(JITDylib &JD, const MaterializationUnit &MU): Generally used to record initializer symbols. - Error notifyRemoving(JITDylib &JD, VModuleKey K): Used to notify a platform that a module is being removed. Platform implementations can use these callbacks to track outstanding initializers and implement a platform-specific approach for executing them. For example, the MachOPlatform installs a plugin in the JIT linker to scan for both __mod_inits sections (for C++ static constructors) and ObjC metadata sections. If discovered, these are processed in the usual platform order: Objective-C registration is carried out first, then static initializers are executed, ensuring that calls to Objective-C from static initializers will be safe. This patch updates LLJIT to use the new scheme for initialization. Two LLJIT::PlatformSupport classes are implemented: A GenericIR platform and a MachO platform. The GenericIR platform implements a modified version of the previous llvm.global-ctor scraping scheme to provide support for Windows and Linux. LLJIT's MachO platform uses the MachOPlatform class to provide MachO specific initialization as described above. Reviewers: sgraenitz, dblaikie Subscribers: mgorny, hiraditya, mgrang, ributzka, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D74300
2019-12-16 18:50:40 +08:00
}
}
[ORC] Improve computeLocalDeps / computeNamedSymbolDependencies performance. The computeNamedSymbolDependencies and computeLocalDeps methods on ObjectLinkingLayerJITLinkContext are responsible for computing, for each symbol in the current MaterializationResponsibility, the set of non-locally-scoped symbols that are depended on. To calculate this we have to consider the effect of chains of dependence through locally scoped symbols in the LinkGraph. E.g. .text .globl foo foo: callq bar ## foo depneds on external 'bar' movq Ltmp1(%rip), %rcx ## foo depends on locally scoped 'Ltmp1' addl (%rcx), %eax retq .data Ltmp1: .quad x ## Ltmp1 depends on external 'x' In this example symbol 'foo' depends directly on 'bar', and indirectly on 'x' via 'Ltmp1', which is locally scoped. Performance of the existing implementations appears to have been mediocre: Based on flame graphs posted by @drmeister (in #jit on the LLVM discord server) the computeLocalDeps function was taking up a substantial amount of time when starting up Clasp (https://github.com/clasp-developers/clasp). This commit attempts to address the performance problems in three ways: 1. Using jitlink::Blocks instead of jitlink::Symbols as the nodes of the dependencies-introduced-by-locally-scoped-symbols graph. Using either Blocks or Symbols as nodes provides the same information, but since there may be more than one locally scoped symbol per block the block-based version of the dependence graph should always be a subgraph of the Symbol-based version, and so faster to operate on. 2. Improved worklist management. The older version of computeLocalDeps used a fixed worklist containing all nodes, and iterated over this list propagating dependencies until no further changes were required. The worklist was not sorted into a useful order before the loop started. The new version uses a variable work-stack, visiting nodes in DFS order and only adding nodes when there is meaningful work to do on them. Compared to the old version the new version avoids revisiting nodes which haven't changed, and I suspect it converges more quickly (due to the DFS ordering). 3. Laziness and caching. Mappings of... jitlink::Symbol* -> Interned Name (as SymbolStringPtr) jitlink::Block* -> Immediate dependencies (as SymbolNameSet) jitlink::Block* -> Transitive dependencies (as SymbolNameSet) are all built lazily and cached while running computeNamedSymbolDependencies. According to @drmeister these changes reduced Clasp startup time in his test setup (averaged over a handful of starts) from 4.8 to 2.8 seconds (with ORC/JITLink linking ~11,000 object files in that time), which seems like enough to justify switching to the new algorithm in the absence of any other perf numbers.
2021-07-08 13:29:47 +08:00
DenseMap<const Block *, DenseSet<Block *>> BlockDeps;
for (auto &KV : BlockInfos)
BlockDeps[KV.first] = std::move(KV.second.Dependencies);
return BlockDependenciesMap(Layer.getExecutionSession(),
std::move(BlockDeps));
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
}
void registerDependencies(const SymbolDependenceMap &QueryDeps) {
for (auto &NamedDepsEntry : ExternalNamedSymbolDeps) {
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
auto &Name = NamedDepsEntry.first;
auto &NameDeps = NamedDepsEntry.second;
SymbolDependenceMap SymbolDeps;
for (const auto &QueryDepsEntry : QueryDeps) {
JITDylib &SourceJD = *QueryDepsEntry.first;
const SymbolNameSet &Symbols = QueryDepsEntry.second;
auto &DepsForJD = SymbolDeps[&SourceJD];
for (const auto &S : Symbols)
if (NameDeps.count(S))
DepsForJD.insert(S);
if (DepsForJD.empty())
SymbolDeps.erase(&SourceJD);
}
MR->addDependencies(Name, SymbolDeps);
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
}
}
ObjectLinkingLayer &Layer;
std::unique_ptr<MaterializationResponsibility> MR;
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
std::unique_ptr<MemoryBuffer> ObjBuffer;
DenseMap<SymbolStringPtr, SymbolNameSet> ExternalNamedSymbolDeps;
DenseMap<SymbolStringPtr, SymbolNameSet> InternalNamedSymbolDeps;
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
};
ObjectLinkingLayer::Plugin::~Plugin() {}
char ObjectLinkingLayer::ID;
using BaseT = RTTIExtends<ObjectLinkingLayer, ObjectLayer>;
ObjectLinkingLayer::ObjectLinkingLayer(ExecutionSession &ES)
: BaseT(ES), MemMgr(ES.getExecutorProcessControl().getMemMgr()) {
ES.registerResourceManager(*this);
}
ObjectLinkingLayer::ObjectLinkingLayer(ExecutionSession &ES,
JITLinkMemoryManager &MemMgr)
: BaseT(ES), MemMgr(MemMgr) {
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
ES.registerResourceManager(*this);
}
ObjectLinkingLayer::ObjectLinkingLayer(
ExecutionSession &ES, std::unique_ptr<JITLinkMemoryManager> MemMgr)
: BaseT(ES), MemMgr(*MemMgr), MemMgrOwnership(std::move(MemMgr)) {
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
ES.registerResourceManager(*this);
}
ObjectLinkingLayer::~ObjectLinkingLayer() {
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
assert(Allocs.empty() && "Layer destroyed with resources still attached");
getExecutionSession().deregisterResourceManager(*this);
}
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
Error ObjectLinkingLayer::add(ResourceTrackerSP RT,
std::unique_ptr<LinkGraph> G) {
auto &JD = RT->getJITDylib();
return JD.define(LinkGraphMaterializationUnit::Create(*this, std::move(G)),
std::move(RT));
}
void ObjectLinkingLayer::emit(std::unique_ptr<MaterializationResponsibility> R,
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
std::unique_ptr<MemoryBuffer> O) {
assert(O && "Object must not be null");
[Orc] Add JITLink debug support plugin for ELF x86-64 Add a new ObjectLinkingLayer plugin `DebugObjectManagerPlugin` and infrastructure to handle creation of `DebugObject`s as well as their registration in OrcTargetProcess. The current implementation only covers ELF on x86-64, but the infrastructure is not limited to that. The journey starts with a new `LinkGraph` / `JITLinkContext` pair being created for a `MaterializationResponsibility` in ORC's `ObjectLinkingLayer`. It sends a `notifyMaterializing()` notification, which is forwarded to all registered plugins. The `DebugObjectManagerPlugin` aims to create a `DebugObject` form the provided target triple and object buffer. (Future implementations might create `DebugObject`s from a `LinkGraph` in other ways.) On success it will track it as the pending `DebugObject` for the `MaterializationResponsibility`. This patch only implements the `ELFDebugObject` for `x86-64` targets. It follows the RuntimeDyld approach for debug object setup: it captures a copy of the input object, parses all section headers and prepares to patch their load-address fields with their final addresses in target memory. It instructs the plugin to report the section load-addresses once they are available. The plugin overrides `modifyPassConfig()` and installs a JITLink post-allocation pass to capture them. Once JITLink emitted the finalized executable, the plugin emits and registers the `DebugObject`. For emission it requests a new `JITLinkMemoryManager::Allocation` with a single read-only segment, copies the object with patched section load-addresses over to working memory and triggers finalization to target memory. For registration, it notifies the `DebugObjectRegistrar` provided in the constructor and stores the previously pending`DebugObject` as registered for the corresponding MaterializationResponsibility. The `DebugObjectRegistrar` registers the `DebugObject` with the target process. `llvm-jitlink` uses the `TPCDebugObjectRegistrar`, which calls `llvm_orc_registerJITLoaderGDBWrapper()` in the target process via `TargetProcessControl` to emit a `jit_code_entry` compatible with the GDB JIT interface [1]. So far the implementation only supports registration and no removal. It appears to me that it wouldn't raise any new design questions, so I left this as an addition for the near future. [1] https://sourceware.org/gdb/current/onlinedocs/gdb/JIT-Interface.html Reviewed By: lhames Differential Revision: https://reviews.llvm.org/D97335
2021-03-02 19:37:48 +08:00
MemoryBufferRef ObjBuffer = O->getMemBufferRef();
auto Ctx = std::make_unique<ObjectLinkingLayerJITLinkContext>(
*this, std::move(R), std::move(O));
[Orc] Add JITLink debug support plugin for ELF x86-64 Add a new ObjectLinkingLayer plugin `DebugObjectManagerPlugin` and infrastructure to handle creation of `DebugObject`s as well as their registration in OrcTargetProcess. The current implementation only covers ELF on x86-64, but the infrastructure is not limited to that. The journey starts with a new `LinkGraph` / `JITLinkContext` pair being created for a `MaterializationResponsibility` in ORC's `ObjectLinkingLayer`. It sends a `notifyMaterializing()` notification, which is forwarded to all registered plugins. The `DebugObjectManagerPlugin` aims to create a `DebugObject` form the provided target triple and object buffer. (Future implementations might create `DebugObject`s from a `LinkGraph` in other ways.) On success it will track it as the pending `DebugObject` for the `MaterializationResponsibility`. This patch only implements the `ELFDebugObject` for `x86-64` targets. It follows the RuntimeDyld approach for debug object setup: it captures a copy of the input object, parses all section headers and prepares to patch their load-address fields with their final addresses in target memory. It instructs the plugin to report the section load-addresses once they are available. The plugin overrides `modifyPassConfig()` and installs a JITLink post-allocation pass to capture them. Once JITLink emitted the finalized executable, the plugin emits and registers the `DebugObject`. For emission it requests a new `JITLinkMemoryManager::Allocation` with a single read-only segment, copies the object with patched section load-addresses over to working memory and triggers finalization to target memory. For registration, it notifies the `DebugObjectRegistrar` provided in the constructor and stores the previously pending`DebugObject` as registered for the corresponding MaterializationResponsibility. The `DebugObjectRegistrar` registers the `DebugObject` with the target process. `llvm-jitlink` uses the `TPCDebugObjectRegistrar`, which calls `llvm_orc_registerJITLoaderGDBWrapper()` in the target process via `TargetProcessControl` to emit a `jit_code_entry` compatible with the GDB JIT interface [1]. So far the implementation only supports registration and no removal. It appears to me that it wouldn't raise any new design questions, so I left this as an addition for the near future. [1] https://sourceware.org/gdb/current/onlinedocs/gdb/JIT-Interface.html Reviewed By: lhames Differential Revision: https://reviews.llvm.org/D97335
2021-03-02 19:37:48 +08:00
if (auto G = createLinkGraphFromObject(ObjBuffer)) {
Ctx->notifyMaterializing(**G);
link(std::move(*G), std::move(Ctx));
[Orc] Add JITLink debug support plugin for ELF x86-64 Add a new ObjectLinkingLayer plugin `DebugObjectManagerPlugin` and infrastructure to handle creation of `DebugObject`s as well as their registration in OrcTargetProcess. The current implementation only covers ELF on x86-64, but the infrastructure is not limited to that. The journey starts with a new `LinkGraph` / `JITLinkContext` pair being created for a `MaterializationResponsibility` in ORC's `ObjectLinkingLayer`. It sends a `notifyMaterializing()` notification, which is forwarded to all registered plugins. The `DebugObjectManagerPlugin` aims to create a `DebugObject` form the provided target triple and object buffer. (Future implementations might create `DebugObject`s from a `LinkGraph` in other ways.) On success it will track it as the pending `DebugObject` for the `MaterializationResponsibility`. This patch only implements the `ELFDebugObject` for `x86-64` targets. It follows the RuntimeDyld approach for debug object setup: it captures a copy of the input object, parses all section headers and prepares to patch their load-address fields with their final addresses in target memory. It instructs the plugin to report the section load-addresses once they are available. The plugin overrides `modifyPassConfig()` and installs a JITLink post-allocation pass to capture them. Once JITLink emitted the finalized executable, the plugin emits and registers the `DebugObject`. For emission it requests a new `JITLinkMemoryManager::Allocation` with a single read-only segment, copies the object with patched section load-addresses over to working memory and triggers finalization to target memory. For registration, it notifies the `DebugObjectRegistrar` provided in the constructor and stores the previously pending`DebugObject` as registered for the corresponding MaterializationResponsibility. The `DebugObjectRegistrar` registers the `DebugObject` with the target process. `llvm-jitlink` uses the `TPCDebugObjectRegistrar`, which calls `llvm_orc_registerJITLoaderGDBWrapper()` in the target process via `TargetProcessControl` to emit a `jit_code_entry` compatible with the GDB JIT interface [1]. So far the implementation only supports registration and no removal. It appears to me that it wouldn't raise any new design questions, so I left this as an addition for the near future. [1] https://sourceware.org/gdb/current/onlinedocs/gdb/JIT-Interface.html Reviewed By: lhames Differential Revision: https://reviews.llvm.org/D97335
2021-03-02 19:37:48 +08:00
} else {
Ctx->notifyFailed(G.takeError());
[Orc] Add JITLink debug support plugin for ELF x86-64 Add a new ObjectLinkingLayer plugin `DebugObjectManagerPlugin` and infrastructure to handle creation of `DebugObject`s as well as their registration in OrcTargetProcess. The current implementation only covers ELF on x86-64, but the infrastructure is not limited to that. The journey starts with a new `LinkGraph` / `JITLinkContext` pair being created for a `MaterializationResponsibility` in ORC's `ObjectLinkingLayer`. It sends a `notifyMaterializing()` notification, which is forwarded to all registered plugins. The `DebugObjectManagerPlugin` aims to create a `DebugObject` form the provided target triple and object buffer. (Future implementations might create `DebugObject`s from a `LinkGraph` in other ways.) On success it will track it as the pending `DebugObject` for the `MaterializationResponsibility`. This patch only implements the `ELFDebugObject` for `x86-64` targets. It follows the RuntimeDyld approach for debug object setup: it captures a copy of the input object, parses all section headers and prepares to patch their load-address fields with their final addresses in target memory. It instructs the plugin to report the section load-addresses once they are available. The plugin overrides `modifyPassConfig()` and installs a JITLink post-allocation pass to capture them. Once JITLink emitted the finalized executable, the plugin emits and registers the `DebugObject`. For emission it requests a new `JITLinkMemoryManager::Allocation` with a single read-only segment, copies the object with patched section load-addresses over to working memory and triggers finalization to target memory. For registration, it notifies the `DebugObjectRegistrar` provided in the constructor and stores the previously pending`DebugObject` as registered for the corresponding MaterializationResponsibility. The `DebugObjectRegistrar` registers the `DebugObject` with the target process. `llvm-jitlink` uses the `TPCDebugObjectRegistrar`, which calls `llvm_orc_registerJITLoaderGDBWrapper()` in the target process via `TargetProcessControl` to emit a `jit_code_entry` compatible with the GDB JIT interface [1]. So far the implementation only supports registration and no removal. It appears to me that it wouldn't raise any new design questions, so I left this as an addition for the near future. [1] https://sourceware.org/gdb/current/onlinedocs/gdb/JIT-Interface.html Reviewed By: lhames Differential Revision: https://reviews.llvm.org/D97335
2021-03-02 19:37:48 +08:00
}
}
void ObjectLinkingLayer::emit(std::unique_ptr<MaterializationResponsibility> R,
std::unique_ptr<LinkGraph> G) {
[Orc] Add JITLink debug support plugin for ELF x86-64 Add a new ObjectLinkingLayer plugin `DebugObjectManagerPlugin` and infrastructure to handle creation of `DebugObject`s as well as their registration in OrcTargetProcess. The current implementation only covers ELF on x86-64, but the infrastructure is not limited to that. The journey starts with a new `LinkGraph` / `JITLinkContext` pair being created for a `MaterializationResponsibility` in ORC's `ObjectLinkingLayer`. It sends a `notifyMaterializing()` notification, which is forwarded to all registered plugins. The `DebugObjectManagerPlugin` aims to create a `DebugObject` form the provided target triple and object buffer. (Future implementations might create `DebugObject`s from a `LinkGraph` in other ways.) On success it will track it as the pending `DebugObject` for the `MaterializationResponsibility`. This patch only implements the `ELFDebugObject` for `x86-64` targets. It follows the RuntimeDyld approach for debug object setup: it captures a copy of the input object, parses all section headers and prepares to patch their load-address fields with their final addresses in target memory. It instructs the plugin to report the section load-addresses once they are available. The plugin overrides `modifyPassConfig()` and installs a JITLink post-allocation pass to capture them. Once JITLink emitted the finalized executable, the plugin emits and registers the `DebugObject`. For emission it requests a new `JITLinkMemoryManager::Allocation` with a single read-only segment, copies the object with patched section load-addresses over to working memory and triggers finalization to target memory. For registration, it notifies the `DebugObjectRegistrar` provided in the constructor and stores the previously pending`DebugObject` as registered for the corresponding MaterializationResponsibility. The `DebugObjectRegistrar` registers the `DebugObject` with the target process. `llvm-jitlink` uses the `TPCDebugObjectRegistrar`, which calls `llvm_orc_registerJITLoaderGDBWrapper()` in the target process via `TargetProcessControl` to emit a `jit_code_entry` compatible with the GDB JIT interface [1]. So far the implementation only supports registration and no removal. It appears to me that it wouldn't raise any new design questions, so I left this as an addition for the near future. [1] https://sourceware.org/gdb/current/onlinedocs/gdb/JIT-Interface.html Reviewed By: lhames Differential Revision: https://reviews.llvm.org/D97335
2021-03-02 19:37:48 +08:00
auto Ctx = std::make_unique<ObjectLinkingLayerJITLinkContext>(
*this, std::move(R), nullptr);
Ctx->notifyMaterializing(*G);
link(std::move(G), std::move(Ctx));
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
}
void ObjectLinkingLayer::modifyPassConfig(MaterializationResponsibility &MR,
LinkGraph &G,
PassConfiguration &PassConfig) {
for (auto &P : Plugins)
P->modifyPassConfig(MR, G, PassConfig);
}
void ObjectLinkingLayer::notifyLoaded(MaterializationResponsibility &MR) {
for (auto &P : Plugins)
P->notifyLoaded(MR);
}
Error ObjectLinkingLayer::notifyEmitted(MaterializationResponsibility &MR,
FinalizedAlloc FA) {
Error Err = Error::success();
for (auto &P : Plugins)
Err = joinErrors(std::move(Err), P->notifyEmitted(MR));
if (Err)
return Err;
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
return MR.withResourceKeyDo(
[&](ResourceKey K) { Allocs[K].push_back(std::move(FA)); });
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
}
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
Error ObjectLinkingLayer::handleRemoveResources(ResourceKey K) {
{
Error Err = Error::success();
for (auto &P : Plugins)
Err = joinErrors(std::move(Err), P->notifyRemovingResources(K));
if (Err)
return Err;
}
std::vector<FinalizedAlloc> AllocsToRemove;
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
getExecutionSession().runSessionLocked([&] {
auto I = Allocs.find(K);
if (I != Allocs.end()) {
std::swap(AllocsToRemove, I->second);
Allocs.erase(I);
}
});
if (AllocsToRemove.empty())
return Error::success();
return MemMgr.deallocate(std::move(AllocsToRemove));
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
}
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
void ObjectLinkingLayer::handleTransferResources(ResourceKey DstKey,
ResourceKey SrcKey) {
auto I = Allocs.find(SrcKey);
if (I != Allocs.end()) {
auto &SrcAllocs = I->second;
auto &DstAllocs = Allocs[DstKey];
DstAllocs.reserve(DstAllocs.size() + SrcAllocs.size());
for (auto &Alloc : SrcAllocs)
DstAllocs.push_back(std::move(Alloc));
// Erase SrcKey entry using value rather than iterator I: I may have been
// invalidated when we looked up DstKey.
Allocs.erase(SrcKey);
}
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
for (auto &P : Plugins)
P->notifyTransferringResources(DstKey, SrcKey);
}
EHFrameRegistrationPlugin::EHFrameRegistrationPlugin(
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
ExecutionSession &ES, std::unique_ptr<EHFrameRegistrar> Registrar)
: ES(ES), Registrar(std::move(Registrar)) {}
void EHFrameRegistrationPlugin::modifyPassConfig(
MaterializationResponsibility &MR, LinkGraph &G,
PassConfiguration &PassConfig) {
PassConfig.PostFixupPasses.push_back(createEHFrameRecorderPass(
G.getTargetTriple(), [this, &MR](ExecutorAddr Addr, size_t Size) {
if (Addr) {
std::lock_guard<std::mutex> Lock(EHFramePluginMutex);
assert(!InProcessLinks.count(&MR) &&
"Link for MR already being tracked?");
InProcessLinks[&MR] = {Addr, Size};
}
}));
}
Error EHFrameRegistrationPlugin::notifyEmitted(
MaterializationResponsibility &MR) {
ExecutorAddrRange EmittedRange;
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
{
std::lock_guard<std::mutex> Lock(EHFramePluginMutex);
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
auto EHFrameRangeItr = InProcessLinks.find(&MR);
if (EHFrameRangeItr == InProcessLinks.end())
return Error::success();
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
EmittedRange = EHFrameRangeItr->second;
assert(EmittedRange.Start && "eh-frame addr to register can not be null");
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
InProcessLinks.erase(EHFrameRangeItr);
}
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
if (auto Err = MR.withResourceKeyDo(
[&](ResourceKey K) { EHFrameRanges[K].push_back(EmittedRange); }))
return Err;
return Registrar->registerEHFrames(EmittedRange);
}
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
Error EHFrameRegistrationPlugin::notifyFailed(
MaterializationResponsibility &MR) {
std::lock_guard<std::mutex> Lock(EHFramePluginMutex);
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
InProcessLinks.erase(&MR);
return Error::success();
}
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
Error EHFrameRegistrationPlugin::notifyRemovingResources(ResourceKey K) {
std::vector<ExecutorAddrRange> RangesToRemove;
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
ES.runSessionLocked([&] {
auto I = EHFrameRanges.find(K);
if (I != EHFrameRanges.end()) {
RangesToRemove = std::move(I->second);
EHFrameRanges.erase(I);
}
});
Error Err = Error::success();
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
while (!RangesToRemove.empty()) {
auto RangeToRemove = RangesToRemove.back();
RangesToRemove.pop_back();
assert(RangeToRemove.Start && "Untracked eh-frame range must not be null");
Err = joinErrors(std::move(Err),
Registrar->deregisterEHFrames(RangeToRemove));
}
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
return Err;
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
}
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
void EHFrameRegistrationPlugin::notifyTransferringResources(
ResourceKey DstKey, ResourceKey SrcKey) {
auto SI = EHFrameRanges.find(SrcKey);
if (SI == EHFrameRanges.end())
return;
auto DI = EHFrameRanges.find(DstKey);
if (DI != EHFrameRanges.end()) {
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
auto &SrcRanges = SI->second;
auto &DstRanges = DI->second;
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
DstRanges.reserve(DstRanges.size() + SrcRanges.size());
for (auto &SrcRange : SrcRanges)
DstRanges.push_back(std::move(SrcRange));
EHFrameRanges.erase(SI);
} else {
// We need to move SrcKey's ranges over without invalidating the SI
// iterator.
auto Tmp = std::move(SI->second);
EHFrameRanges.erase(SI);
EHFrameRanges[DstKey] = std::move(Tmp);
[ORC] Add support for resource tracking/removal (removable code). This patch introduces new APIs to support resource tracking and removal in Orc. It is intended as a thread-safe generalization of the removeModule concept from OrcV1. Clients can now create ResourceTracker objects (using JITDylib::createResourceTracker) to track resources for each MaterializationUnit (code, data, aliases, absolute symbols, etc.) added to the JIT. Every MaterializationUnit will be associated with a ResourceTracker, and ResourceTrackers can be re-used for multiple MaterializationUnits. Each JITDylib has a default ResourceTracker that will be used for MaterializationUnits added to that JITDylib if no ResourceTracker is explicitly specified. Two operations can be performed on ResourceTrackers: transferTo and remove. The transferTo operation transfers tracking of the resources to a different ResourceTracker object, allowing ResourceTrackers to be merged to reduce administrative overhead (the source tracker is invalidated in the process). The remove operation removes all resources associated with a ResourceTracker, including any symbols defined by MaterializationUnits associated with the tracker, and also invalidates the tracker. These operations are thread safe, and should work regardless of the the state of the MaterializationUnits. In the case of resource transfer any existing resources associated with the source tracker will be transferred to the destination tracker, and all future resources for those units will be automatically associated with the destination tracker. In the case of resource removal all already-allocated resources will be deallocated, any if any program representations associated with the tracker have not been compiled yet they will be destroyed. If any program representations are currently being compiled then they will be prevented from completing: their MaterializationResponsibility will return errors on any attempt to update the JIT state. Clients (usually Layer writers) wishing to track resources can implement the ResourceManager API to receive notifications when ResourceTrackers are transferred or removed. The MaterializationResponsibility::withResourceKeyDo method can be used to create associations between the key for a ResourceTracker and an allocated resource in a thread-safe way. RTDyldObjectLinkingLayer and ObjectLinkingLayer are updated to use the ResourceManager API to enable tracking and removal of memory allocated by the JIT linker. The new JITDylib::clear method can be used to trigger removal of every ResourceTracker associated with the JITDylib (note that this will only remove resources for the JITDylib, it does not run static destructors). This patch includes unit tests showing basic usage. A follow-up patch will update the Kaleidoscope and BuildingAJIT tutorial series to OrcV2 and will use this API to release code associated with anonymous expressions.
2020-09-12 00:50:41 +08:00
}
}
Initial implementation of JITLink - A replacement for RuntimeDyld. Summary: JITLink is a jit-linker that performs the same high-level task as RuntimeDyld: it parses relocatable object files and makes their contents runnable in a target process. JITLink aims to improve on RuntimeDyld in several ways: (1) A clear design intended to maximize code-sharing while minimizing coupling. RuntimeDyld has been developed in an ad-hoc fashion for a number of years and this had led to intermingling of code for multiple architectures (e.g. in RuntimeDyldELF::processRelocationRef) in a way that makes the code more difficult to read, reason about, extend. JITLink is designed to isolate format and architecture specific code, while still sharing generic code. (2) Support for native code models. RuntimeDyld required the use of large code models (where calls to external functions are made indirectly via registers) for many of platforms due to its restrictive model for stub generation (one "stub" per symbol). JITLink allows arbitrary mutation of the atom graph, allowing both GOT and PLT atoms to be added naturally. (3) Native support for asynchronous linking. JITLink uses asynchronous calls for symbol resolution and finalization: these callbacks are passed a continuation function that they must call to complete the linker's work. This allows for cleaner interoperation with the new concurrent ORC JIT APIs, while still being easily implementable in synchronous style if asynchrony is not needed. To maximise sharing, the design has a hierarchy of common code: (1) Generic atom-graph data structure and algorithms (e.g. dead stripping and | memory allocation) that are intended to be shared by all architectures. | + -- (2) Shared per-format code that utilizes (1), e.g. Generic MachO to | atom-graph parsing. | + -- (3) Architecture specific code that uses (1) and (2). E.g. JITLinkerMachO_x86_64, which adds x86-64 specific relocation support to (2) to build and patch up the atom graph. To support asynchronous symbol resolution and finalization, the callbacks for these operations take continuations as arguments: using JITLinkAsyncLookupContinuation = std::function<void(Expected<AsyncLookupResult> LR)>; using JITLinkAsyncLookupFunction = std::function<void(const DenseSet<StringRef> &Symbols, JITLinkAsyncLookupContinuation LookupContinuation)>; using FinalizeContinuation = std::function<void(Error)>; virtual void finalizeAsync(FinalizeContinuation OnFinalize); In addition to its headline features, JITLink also makes other improvements: - Dead stripping support: symbols that are not used (e.g. redundant ODR definitions) are discarded, and take up no memory in the target process (In contrast, RuntimeDyld supported pointer equality for weak definitions, but the redundant definitions stayed resident in memory). - Improved exception handling support. JITLink provides a much more extensive eh-frame parser than RuntimeDyld, and is able to correctly fix up many eh-frame sections that RuntimeDyld currently (silently) fails on. - More extensive validation and error handling throughout. This initial patch supports linking MachO/x86-64 only. Work on support for other architectures and formats will happen in-tree. Differential Revision: https://reviews.llvm.org/D58704 llvm-svn: 358818
2019-04-21 01:10:34 +08:00
} // End namespace orc.
} // End namespace llvm.