Commit Graph

368216 Commits

Author SHA1 Message Date
Jianzhou Zhao 045a620c45 Release the shadow memory used by the mmap range at munmap
When an application does a lot of pairs of mmap and munmap, if we did
not release shadoe memory used by mmap addresses, this would increase
memory usage.

Reviewed-by: morehouse

Differential Revision: https://reviews.llvm.org/D88686
2020-10-02 20:17:22 +00:00
Tim Keith 11622d0fed [flang][NFC] Fix mis-matched struct/class declarations
The template `ListDirectedStatementState` was declared as a struct and then as a class.
Fix it so they match.

Differential Revision: https://reviews.llvm.org/D88711
2020-10-02 13:08:51 -07:00
Jonas Devlieghere 128e999d63 [lldb] Add a "design" section to the documentation.
Create a "Design" section for the LLDB documentation. The goal is to
have design documents that describe how the LLDB internals work.

Currently similar pages  are mixed together under the "Development". The
existing pages describing the architecture, the reproducers, the
structured data plugins, and the SB API could be housed here. I hope
we'd see more pages being added here in the future.

Differential revision: https://reviews.llvm.org/D88516
2020-10-02 13:07:31 -07:00
Nikita Popov 94704ed008 [MemCpyOpt] Add helper to erase instructions (NFC)
Next to erasing the instruction, we also always want to remove
it from MSSA and MD. Use a common function to do so.

This is a refactoring split out from D26739.
2020-10-02 21:52:10 +02:00
Nikita Popov 87b63c1726 [MemCpyOpt] Avoid double invalidation (NFCI)
The removal of the cpy instruction is left to the caller of
performCallSlotOptzn(), including the invalidation of MD. Both
call-sites already do this.

Also handle incrementation of NumMemCpyInstr consistently at the
call-site. One of the call-site was already doing this, which
ended up incrementing the statistic twice.

This fix was part of D26739.
2020-10-02 21:50:46 +02:00
Douglas Yung 24f406e6d3 Relax newly added opcode alias check to check only for a number instead of a specific opcode. 2020-10-02 12:18:56 -07:00
ergawy 0c8f9b8099 [MLIR][SPIRV] Add initial support for OpSpecConstantComposite.
This commit adds support to SPIR-V's composite specialization constants.
These are specialization constants which are composed of other spec
constants (whehter scalar or composite), regular constatns, or undef
values.

This commit adds support for parsing, printing, verification, and
(De)serialization.

A few TODOs are still in order:
- Supporting more types of constituents; currently, only scalar spec constatns are supported.
- Extending `spv._reference_of` to support composite spec constatns.

Reviewed By: antiagainst

Differential Revision: https://reviews.llvm.org/D88568
2020-10-02 15:18:16 -04:00
Louis Dionne 870827f652 [libc++] NFCI: Remove the _LIBCPP_DEBUG_MODE helper macro
It was used inconsistently and the name was pretty confusing, so we might
as well use `#if _LIBCPP_DEBUG_LEVEL == 2` consistently everywhere.
2020-10-02 15:11:23 -04:00
Louis Dionne 31e820378b [libc++] NFCI: Simplify macro definitions for the debug mode
The debug mode always had three possibilities:
- _LIBCPP_DEBUG is undefined => no assertions
- _LIBCPP_DEBUG == 0         => some assertions
- _LIBCPP_DEBUG == 1         => some assertions + iterator checks

This was documented that way, however the code did not make this clear
at all. The discrepancy between _LIBCPP_DEBUG and _LIBCPP_DEBUG_LEVEL
was especially confusing. I reworked how the various macros are defined
without changing anything else to make the code clearer.
2020-10-02 15:11:23 -04:00
Evgenii Stepanov 66cf68ed46 [docs] Update ControlFlowIntegrity.rst.
Expand the list of targets that support cfi-icall.
Add ThinLTO everywhere LTO is mentioned. AFAIK all CFI features are
supported with ThinLTO.

Differential Revision: https://reviews.llvm.org/D87717
2020-10-02 12:01:05 -07:00
Fangrui Song 322519ee12 [llc] Initialize TargetOptions after Triple is available
Some targets have different defaults. This patch defers initialization of `TargetOptions` so that a future patch can pass `TargetOptions` to `InitTargetOptionsFromCodeGenFlags`

Reviewed By: jasonliu

Differential Revision: https://reviews.llvm.org/D88748
2020-10-02 11:43:40 -07:00
Amara Emerson 1e020b2a17 Update legalizer-info-validation.mir test to test all opcodes.
The test doesn't fail if we add opcodes to the end of the opcodes definition
list, so we were missing some.
2020-10-02 11:32:54 -07:00
Louis Dionne 1a92de0064 [libc++] NFCI: Remove _LIBCPP_EXTERN_TEMPLATE2
This seems to have been added a long time ago as a temporary help
for debugging some <regex> issue, but it's really the same as
_LIBCPP_EXTERN_TEMPLATE.
2020-10-02 14:31:43 -04:00
Nikita Popov 84feca6a84 [MemCpyOpt] Add tests from D40802 (NFC)
Even though that patch didn't stick, we should retain the test
coverage.
2020-10-02 20:28:38 +02:00
Arthur Eubanks 7468afe9ca [DAE] MarkLive in MarkValue(MaybeLive) if any use is live
While looping through all args or all return values, we may mark a use
of a later iteration as live. Previously when we got to that later value
it would ignore that and continue adding to Uses instead of marking it
live. For example, when looping through arg#0 and arg#1,
MarkValue(arg#0, Live) may cause some use of arg#1 to be live, but
MarkValue(arg#1, MaybeLive) will not notice that and continue adding
into Uses.

Now MarkValue(RA, MaybeLive) will MarkLive(RA) if any use is live.

Fixes PR47444.

Reviewed By: rnk

Differential Revision: https://reviews.llvm.org/D88529
2020-10-02 10:55:08 -07:00
Peter Collingbourne a8938f3da3 scudo: Simplify AtomicOptions::setFillContentsMode. NFCI.
Differential Revision: https://reviews.llvm.org/D88747
2020-10-02 10:52:41 -07:00
Arthur Eubanks eb55735073 Reland [AlwaysInliner] Update BFI when inlining
Reviewed By: davidxl

Differential Revision: https://reviews.llvm.org/D88324
2020-10-02 10:46:57 -07:00
Arthur Eubanks 354ba1cb80 [gn build] Don't define CINDEX_EXPORTS
This causes
../../clang/include\clang-c/Platform.h(23,11): warning: 'CINDEX_EXPORTS' macro redefined [-Wmacro-redefined]
  #define CINDEX_EXPORTS
2020-10-02 10:39:49 -07:00
Arthur Eubanks 9b8c0b8b46 Revert "[AlwaysInliner] Update BFI when inlining"
This reverts commit b1bf24667f.
2020-10-02 10:34:51 -07:00
Arthur Eubanks b1bf24667f [AlwaysInliner] Update BFI when inlining
Reviewed By: davidxl

Differential Revision: https://reviews.llvm.org/D88324
2020-10-02 10:26:34 -07:00
Simon Pilgrim 0364721e3e Revert rG3d14a1e982ad27 - "[InstCombine] recognizeBSwapOrBitReverseIdiom - support for 'partial' bswap patterns (PR47191)"
This reverts commit 3d14a1e982.

This is breaking on some 2stage clang buildbots
2020-10-02 18:17:14 +01:00
Thomas Raoux d1c8e179d8 [mlir][vector] Add canonicalization patterns for extractMap/insertMap
Add basic canonicalization patterns for the extractMap/insertMap to allow them
to be folded into Transfer ops.
Also mark transferRead as memory read so that it can be removed by dead code.

Differential Revision: https://reviews.llvm.org/D88622
2020-10-02 10:13:11 -07:00
Simon Pilgrim d0dd7cadbd [InstCombine] Add trunc(bswap(trunc/zext(x))) vector tests 2020-10-02 18:05:16 +01:00
Louis Dionne aac2de1b1a [libc++] Remove unnecessary usage of <iostream> in the test suite
Tests should strive to be as minimal as possible, since it makes them
relevant on platforms where <iostream> does not work.
2020-10-02 13:00:34 -04:00
Jonas Devlieghere 07c112574a [lldb] Fix bug in fallback logic for finding the resource directory.
Both of the if-clauses modify the raw_path variable and only one of them
was resetting the variable for the fallback. Avoid future bugs like that
by always resetting the variable.

Differential revision: https://reviews.llvm.org/D88704
2020-10-02 09:56:01 -07:00
Florian Hahn 0867a9e85a [VPlan] Use isa<> instead of directly checking VPRecipeID (NFC).
getVPRecipeID is intended to be only used in `classof` helpers. Instead
of checking it directly, use isa<> with the correct recipe type.
2020-10-02 17:47:35 +01:00
Nikita Popov 64c54c5459 [MemCpyOpt] Regnerate test checks (NFC) 2020-10-02 18:42:13 +02:00
zhanghb97 2fc0d4a8e8 [mlir] Add Float Attribute, Integer Attribute and Bool Attribute subclasses to python bindings.
Based on PyAttribute and PyConcreteAttribute classes, this patch implements the bindings of Float Attribute, Integer Attribute and Bool Attribute subclasses.
This patch also defines the `mlirFloatAttrDoubleGetChecked` C API which is bound with the `FloatAttr.get_typed` python method.

Differential Revision: https://reviews.llvm.org/D88531
2020-10-03 00:32:51 +08:00
Stephen Neuendorffer 34d12c15f7 [MLIR] Better message for FuncOp type mismatch
Previously the actual types were not shown, which makes the message
difficult to grok in the context of long lowering chains.  Also, it
appears that there were no actual tests for this.

Differential Revision: https://reviews.llvm.org/D88318
2020-10-02 09:31:44 -07:00
Sanjay Patel 33fa3dbce9 [CostModel] move default handling after switch; NFC
We will need to add intrinsics to the switch (such as
the ones that are currently in the switch above this
one) that deal with special cases and then break to
the default handling.
2020-10-02 12:26:49 -04:00
Stella Stamenova 432e4e56d3 Revert "[WebAssembly] Emulate v128.const efficiently"
This reverts commit 542523a61a.
2020-10-02 09:26:21 -07:00
Simon Pilgrim 3d14a1e982 [InstCombine] recognizeBSwapOrBitReverseIdiom - support for 'partial' bswap patterns (PR47191)
If we're bswap'ing some bytes and zero'ing the remainder we can perform this as a bswap+mask which helps us match 'partial' bswaps as a first step towards folding into a more complex bswap pattern.

Differential Revision: https://reviews.llvm.org/D88578
2020-10-02 17:25:12 +01:00
Simon Pilgrim 0347f3ea72 TruncInstCombine.cpp - fix header include ordering to fix llvm-include-order clang-tidy warning. NFCI. 2020-10-02 17:25:12 +01:00
Simon Pilgrim 5e8e89d814 TruncInstCombine.cpp - use auto * to fix llvm-qualified-auto clang-tidy warning. NFCI. 2020-10-02 17:25:11 +01:00
Vinay Madhusudan f192594956 [AArch64] Generate dot for v16i8 sum reduction to i32
Convert VECREDUCE_ADD( EXTEND(v16i8_type) ) to VECREDUCE_ADD( DOTv16i8(v16i8_type) ) whenever the result type is i32. This gains in one of the SPECCPU 2017 benchmark.

This partially solves the bug: https://bugs.llvm.org/show_bug.cgi?id=46888
Meta ticket: https://bugs.llvm.org/show_bug.cgi?id=46929

Differential Revision: https://reviews.llvm.org/D88577
2020-10-02 17:11:02 +01:00
Utkarsh Saxena db2a646c5f [clangd] Add bencmark for measuring latency of DecisionForest model.
Differential Revision: https://reviews.llvm.org/D88590
2020-10-02 18:04:31 +02:00
Diego Caballero a611f9a5c6 [mlir] Fix call op conversion in bare-ptr calling convention
We hit an llvm_unreachable related to unranked memrefs for call ops
with scalar types. Removing the llvm_unreachable since the conversion
should gracefully bail out in the presence of unranked memrefs. Adding
tests to verify that.

Reviewed By: ftynse

Differential Revision: https://reviews.llvm.org/D88709
2020-10-02 08:48:21 -07:00
Nicolas Vasilache 86b14d0969 [mlir] Attempt to appease gcc-5 const char* -> StringLiteral conversion issu 2020-10-02 10:53:48 -04:00
serge-sans-paille f2c6bfa350 Fix interaction between stack alignment and inline-asm stack clash protection
As reported in https://github.com/rust-lang/rust/issues/70143 alignment is not
taken into account when doing the probing. Fix that by adjusting the first probe
if the stack align is small, or by extending the dynamic probing if the
alignment is large.

Differential Revision: https://reviews.llvm.org/D84419
2020-10-02 16:51:49 +02:00
Denis Antrushin 7b19cd06d7 [Statepoints][ISEL] visitGCRelocate: chain to current DAG root.
This is similar to D87251, but for CopyFromRegs nodes.
Even for local statepoint uses we generate CopyToRegs/CopyFromRegs
nodes.  When generating CopyFromRegs in visitGCRelocate, we must chain
to current DAG root, not EntryNode, to ensure proper ordering of copy
w.r.t. statepoint node producing result for it.

Reviewed By: reames

Differential Revision: https://reviews.llvm.org/D88639
2020-10-02 21:41:22 +07:00
Yaxun (Sam) Liu c87c017a4c Fix failure in test hip-macros.hip
requires amdgpu-registered-target.
2020-10-02 10:33:32 -04:00
Kamil Rytarowski 2a9ce60de9 [compiler-rt] [netbsd] Improve the portability of ThreadSelfTlsTcb
Use __lwp_gettcb_fast() and __lwp_getprivate_fast(), as _lwp_getprivate()
can be a biased pointer and invalid for use in this function on all CPUs.
2020-10-02 16:32:58 +02:00
LLVM GN Syncbot d9e3972080 [gn build] Port 0c1bb4f885 2020-10-02 14:24:01 +00:00
Paul C. Anagnostopoulos 0c1bb4f885 [TableGen] New backend to print detailed records.
Pertinent lints are fixed.
2020-10-02 10:22:13 -04:00
Yaxun (Sam) Liu 36501b180a Emit predefined macro for wavefront size for amdgcn
Also fix the issue of multiple -m[no-]wavefrontsize64
options to make the last one wins.

Differential Revision: https://reviews.llvm.org/D88370
2020-10-02 10:17:21 -04:00
Haojian Wu 0f0cbcc4b1 [clangd] Extend the rename API.
several changes:
- return a structure result in rename API;
- prepareRename now returns more information (main-file occurrences);
- remove the duplicated detecting-touch-identifier code in prepareRename (which is implemented in rename API);

Differential Revision: https://reviews.llvm.org/D88634
2020-10-02 16:03:44 +02:00
Simon Pilgrim fa59135bf1 [Analysis] Drop local maxAPInt/minAPInt helpers. NFCI.
Use standard APIntOps::smax/smin helpers instead.
2020-10-02 14:56:12 +01:00
Alexandre Ganea fe1f0a1a19 [LLD] Fix /time formatting for very long runs. NFC. 2020-10-02 09:53:43 -04:00
Alexandre Ganea 55b97a6d2a [LLD][COFF] Add more type record information to /summary
This adds the following two new lines to /summary:

      21351 Input OBJ files (expanded from all cmd-line inputs)
         61 PDB type server dependencies
         38 Precomp OBJ dependencies
 1420669231 Input type records         <<<<
78665073382 Input type records bytes   <<<<
    8801393 Merged TPI records
    3177158 Merged IPI records
      59194 Output PDB strings
   71576766 Global symbol records
   25416935 Module symbol records
    2103431 Public symbol records

Differential Revision: https://reviews.llvm.org/D88703
2020-10-02 09:36:11 -04:00
Louis Dionne c7d4aa711a [libc++] Move the weak symbols list to libc++abi
Those symbols are exported from libc++abi in the first place, so it
makes more sense to have them there.
2020-10-02 09:22:23 -04:00