Refactor this so it's similar to the existing integer comparison code.
Also add some missing 64-bit testcases to select-fcmp.mir.
Refactoring to prep for improving selection for G_FCMP-related conditional
branches etc.
Differential Revision: https://reviews.llvm.org/D88614
This is a follow-up from https://reviews.llvm.org/D61717. Where Richard
described the issue with compiling arm_neon.h under
-flax-vector-conversions=none. It looks like the example reproducer does
actually work but what was missing was a test entry for that target.
Differential Revision: https://reviews.llvm.org/D88546
Currently, the parser used to tokenize the TSAN_OPTIONS in libomp uses
only spaces as separators, even though TSAN in compiler-rt supports
other separators like ':' or ','.
CTest uses ':' to separate sanitizer options by default.
The documentation for other sanitizers mentions ':' as separator,
but TSAN only lists spaces, which is probably where this mismatch originated.
Patch provided by upsj
Differential Revision: https://reviews.llvm.org/D87144
Patch IEEEFloat::isSignificandAllZeros and IEEEFloat::isSignificandAllOnes to behave correctly in the case that the size of the significand is a multiple of the width of the integerParts making up the significand.
The patch to IEEEFloat::isSignificandAllOnes fixes bug 34579, and the patch to IEEE:Float:isSignificandAllZeros fixes the unit test "APFloatTest.x87Next" I added here. I have included both in this diff since the changes are very similar.
Patch by Andrew Briand
This patch adds outer product instructions for MMA, including related infrastructure, and their tests.
Depends on D84968.
Reviewed By: #powerpc, bsaleil, amyk
Differential Revision: https://reviews.llvm.org/D88043
This patch updates the expected results for the GOMP interface patches: D87267, D87269, and D87271.
The taskwait-depend test is changed to really use taskwait-depend and copied to an task_if0-depend test.
To pass the tests, the handling of the return address was fixed.
Differential Revision: https://reviews.llvm.org/D87680
OpenMP 5.1 defines omp_get_initial_device to return the same value as omp_get_num_devices.
Since this change is also 5.0 compliant, no versioning is needed.
Differential Revision: https://reviews.llvm.org/D88149
Represent FINAL subroutines in the symbol table entries of
derived types. Enforce constraints. Update tests that have
inadvertent violations or modified messages. Added a test.
The specific procedure distinguishability checking code for generics
was used to enforce distinguishability of FINAL procedures.
(Also cleaned up some confusion and redundancy noticed in the
type compatibility infrastructure while digging into that area.)
Differential revision: https://reviews.llvm.org/D88613
Stored Error objects have to be checked, even if they are success
values.
This reverts commit 8d250ac3cd.
Relands commit 49b3459930655d879b2dc190ff8fe11c38a8be5f..
Original commit message:
-----------------------------------------
This makes type merging much faster (-24% on chrome.dll) when multiple
threads are available, but it slightly increases the time to link (+10%)
when /threads:1 is passed. With only one more thread, the new type
merging is faster (-11%). The output PDB should be identical to what it
was before this change.
To give an idea, here is the /time output placed side by side:
BEFORE | AFTER
Input File Reading: 956 ms | 968 ms
Code Layout: 258 ms | 190 ms
Commit Output File: 6 ms | 7 ms
PDB Emission (Cumulative): 6691 ms | 4253 ms
Add Objects: 4341 ms | 2927 ms
Type Merging: 2814 ms | 1269 ms -55%!
Symbol Merging: 1509 ms | 1645 ms
Publics Stream Layout: 111 ms | 112 ms
TPI Stream Layout: 764 ms | 26 ms trivial
Commit to Disk: 1322 ms | 1036 ms -300ms
----------------------------------------- --------
Total Link Time: 8416 ms 5882 ms -30% overall
The main source of the additional overhead in the single-threaded case
is the need to iterate all .debug$T sections up front to check which
type records should go in the IPI stream. See fillIsItemIndexFromDebugT.
With changes to the .debug$H section, we could pre-calculate this info
and eliminate the need to do this walk up front. That should restore
single-threaded performance back to what it was before this change.
This change will cause LLD to be much more parallel than it used to, and
for users who do multiple links in parallel, it could regress
performance. However, when the user is only doing one link, it's a huge
improvement. In the future, we can use NT worker threads to avoid
oversaturating the machine with work, but for now, this is such an
improvement for the single-link use case that I think we should land
this as is.
Algorithm
----------
Before this change, we essentially used a
DenseMap<GloballyHashedType, TypeIndex> to check if a type has already
been seen, and if it hasn't been seen, insert it now and use the next
available type index for it in the destination type stream. DenseMap
does not support concurrent insertion, and even if it did, the linker
must be deterministic: it cannot produce different PDBs by using
different numbers of threads. The output type stream must be in the same
order regardless of the order of hash table insertions.
In order to create a hash table that supports concurrent insertion, the
table cells must be small enough that they can be updated atomically.
The algorithm I used for updating the table using linear probing is
described in this paper, "Concurrent Hash Tables: Fast and General(?)!":
https://dl.acm.org/doi/10.1145/3309206
The GHashCell in this change is essentially a pair of 32-bit integer
indices: <sourceIndex, typeIndex>. The sourceIndex is the index of the
TpiSource object, and it represents an input type stream. The typeIndex
is the index of the type in the stream. Together, we have something like
a ragged 2D array of ghashes, which can be looked up as:
tpiSources[tpiSrcIndex]->ghashes[typeIndex]
By using these side tables, we can omit the key data from the hash
table, and keep the table cell small. There is a cost to this: resolving
hash table collisions requires many more loads than simply looking at
the key in the same cache line as the insertion position. However, most
supported platforms should have a 64-bit CAS operation to update the
cell atomically.
To make the result of concurrent insertion deterministic, the cell
payloads must have a priority function. Defining one is pretty
straightforward: compare the two 32-bit numbers as a combined 64-bit
number. This means that types coming from inputs earlier on the command
line have a higher priority and are more likely to appear earlier in the
final PDB type stream than types from an input appearing later on the
link line.
After table insertion, the non-empty cells in the table can be copied
out of the main table and sorted by priority to determine the ordering
of the final type index stream. At this point, item and type records
must be separated, either by sorting or by splitting into two arrays,
and I chose sorting. This is why the GHashCell must contain the isItem
bit.
Once the final PDB TPI stream ordering is known, we need to compute a
mapping from source type index to PDB type index. To avoid starting over
from scratch and looking up every type again by its ghash, we save the
insertion position of every hash table insertion during the first
insertion phase. Because the table does not support rehashing, the
insertion position is stable. Using the array of insertion positions
indexed by source type index, we can replace the source type indices in
the ghash table cells with the PDB type indices.
Once the table cells have been updated to contain PDB type indices, the
mapping for each type source can be computed in parallel. Simply iterate
the list of cell positions and replace them with the PDB type index,
since the insertion positions are no longer needed.
Once we have a source to destination type index mapping for every type
source, there are no more data dependencies. We know which type records
are "unique" (not duplicates), and what their final type indices will
be. We can do the remapping in parallel, and accumulate type sizes and
type hashes in parallel by type source.
Lastly, TPI stream layout must be done serially. Accumulate all the type
records, sizes, and hashes, and add them to the PDB.
Differential Revision: https://reviews.llvm.org/D87805
The pattern is structured similar to other patterns like
LinalgTilingPattern. The fusion patterns takes options that allows you
to fuse with producers of multiple operands at once.
- The pattern fuses only at the level that is known to be legal, i.e
if a reduction loop in the consumer is tiled, then fusion should
happen "before" this loop. Some refactoring of the fusion code is
needed to fuse only where it is legal.
- Since the fusion on buffers uses the LinalgDependenceGraph that is
not mutable in place the fusion pattern keeps the original
operations in the IR, but are tagged with a marker that can be later
used to find the original operations.
This change also fixes an issue with tiling and
distribution/interchange where if the tile size of a loop were 0 it
wasnt account for in these.
Differential Revision: https://reviews.llvm.org/D88435
This makes type merging much faster (-24% on chrome.dll) when multiple
threads are available, but it slightly increases the time to link (+10%)
when /threads:1 is passed. With only one more thread, the new type
merging is faster (-11%). The output PDB should be identical to what it
was before this change.
To give an idea, here is the /time output placed side by side:
BEFORE | AFTER
Input File Reading: 956 ms | 968 ms
Code Layout: 258 ms | 190 ms
Commit Output File: 6 ms | 7 ms
PDB Emission (Cumulative): 6691 ms | 4253 ms
Add Objects: 4341 ms | 2927 ms
Type Merging: 2814 ms | 1269 ms -55%!
Symbol Merging: 1509 ms | 1645 ms
Publics Stream Layout: 111 ms | 112 ms
TPI Stream Layout: 764 ms | 26 ms trivial
Commit to Disk: 1322 ms | 1036 ms -300ms
----------------------------------------- --------
Total Link Time: 8416 ms 5882 ms -30% overall
The main source of the additional overhead in the single-threaded case
is the need to iterate all .debug$T sections up front to check which
type records should go in the IPI stream. See fillIsItemIndexFromDebugT.
With changes to the .debug$H section, we could pre-calculate this info
and eliminate the need to do this walk up front. That should restore
single-threaded performance back to what it was before this change.
This change will cause LLD to be much more parallel than it used to, and
for users who do multiple links in parallel, it could regress
performance. However, when the user is only doing one link, it's a huge
improvement. In the future, we can use NT worker threads to avoid
oversaturating the machine with work, but for now, this is such an
improvement for the single-link use case that I think we should land
this as is.
Algorithm
----------
Before this change, we essentially used a
DenseMap<GloballyHashedType, TypeIndex> to check if a type has already
been seen, and if it hasn't been seen, insert it now and use the next
available type index for it in the destination type stream. DenseMap
does not support concurrent insertion, and even if it did, the linker
must be deterministic: it cannot produce different PDBs by using
different numbers of threads. The output type stream must be in the same
order regardless of the order of hash table insertions.
In order to create a hash table that supports concurrent insertion, the
table cells must be small enough that they can be updated atomically.
The algorithm I used for updating the table using linear probing is
described in this paper, "Concurrent Hash Tables: Fast and General(?)!":
https://dl.acm.org/doi/10.1145/3309206
The GHashCell in this change is essentially a pair of 32-bit integer
indices: <sourceIndex, typeIndex>. The sourceIndex is the index of the
TpiSource object, and it represents an input type stream. The typeIndex
is the index of the type in the stream. Together, we have something like
a ragged 2D array of ghashes, which can be looked up as:
tpiSources[tpiSrcIndex]->ghashes[typeIndex]
By using these side tables, we can omit the key data from the hash
table, and keep the table cell small. There is a cost to this: resolving
hash table collisions requires many more loads than simply looking at
the key in the same cache line as the insertion position. However, most
supported platforms should have a 64-bit CAS operation to update the
cell atomically.
To make the result of concurrent insertion deterministic, the cell
payloads must have a priority function. Defining one is pretty
straightforward: compare the two 32-bit numbers as a combined 64-bit
number. This means that types coming from inputs earlier on the command
line have a higher priority and are more likely to appear earlier in the
final PDB type stream than types from an input appearing later on the
link line.
After table insertion, the non-empty cells in the table can be copied
out of the main table and sorted by priority to determine the ordering
of the final type index stream. At this point, item and type records
must be separated, either by sorting or by splitting into two arrays,
and I chose sorting. This is why the GHashCell must contain the isItem
bit.
Once the final PDB TPI stream ordering is known, we need to compute a
mapping from source type index to PDB type index. To avoid starting over
from scratch and looking up every type again by its ghash, we save the
insertion position of every hash table insertion during the first
insertion phase. Because the table does not support rehashing, the
insertion position is stable. Using the array of insertion positions
indexed by source type index, we can replace the source type indices in
the ghash table cells with the PDB type indices.
Once the table cells have been updated to contain PDB type indices, the
mapping for each type source can be computed in parallel. Simply iterate
the list of cell positions and replace them with the PDB type index,
since the insertion positions are no longer needed.
Once we have a source to destination type index mapping for every type
source, there are no more data dependencies. We know which type records
are "unique" (not duplicates), and what their final type indices will
be. We can do the remapping in parallel, and accumulate type sizes and
type hashes in parallel by type source.
Lastly, TPI stream layout must be done serially. Accumulate all the type
records, sizes, and hashes, and add them to the PDB.
Differential Revision: https://reviews.llvm.org/D87805
These types have to distinguish list-directed I/O from formatted I/O,
and the subscript incrementation call was in the formatted branch
of the if() rather than after the if().
Differential revision: https://reviews.llvm.org/D88606
`Posix/no_asan_gen_globals.c` currently `FAIL`s on Solaris:
$ nm no_asan_gen_globals.c.tmp.exe | grep ___asan_gen_
0809696a r .L___asan_gen_.1
0809a4cd r .L___asan_gen_.2
080908e2 r .L___asan_gen_.4
0809a4cd r .L___asan_gen_.5
0809a529 r .L___asan_gen_.7
0809a4cd r .L___asan_gen_.8
As detailed in Bug 47607, there are two factors here:
- `clang` plays games by emitting some local labels into the symbol
table. When instead one uses `-fno-integrated-as` to have `gas` create
the object files, they don't land in the objects in the first place.
- Unlike GNU `ld`, the Solaris `ld` doesn't support support
`-X`/`--discard-locals` but instead relies on the assembler to follow its
specification and not emit local labels.
Therefore this patch `XFAIL`s the test on Solaris.
Tested on `amd64-pc-solaris2.11` and `x86_64-pc-linux-gnu`.
Differential Revision: https://reviews.llvm.org/D88218
This will be further canonicalized to a compare involving 0
which will enable the use of test instructions. Either using
cmovg for signed for cmovne for unsigned.
Fixes more case for PR47049
The patch adds a new TargetMachine member "registerPassBuilderCallbacks" for targets to add passes to the pass pipeline using the New Pass Manager (similar to adjustPassManager for the Legacy Pass Manager).
Reviewed By: aeubanks
Differential Revision: https://reviews.llvm.org/D88138
This is the first of several steps to support distributing large vectors. This
adds instructions extract_map and insert_map that allow us to do incremental
lowering. Right now the transformation only apply to simple pointwise operation
with a vector size matching the multiplicity of the IDs used to distribute the
vector.
This can be used to distribute large vectors to loops or SPMD.
Differential Revision: https://reviews.llvm.org/D88341
Failing tests on Arm due to the tests automatically populating
incomatible pointer width architectures. Reverting until the tests are
updated. Failing tests:
OpenMP/distribute_parallel_for_num_threads_codegen.cpp
OpenMP/distribute_parallel_for_if_codegen.cpp
OpenMP/distribute_parallel_for_simd_if_codegen.cpp
OpenMP/distribute_parallel_for_simd_num_threads_codegen.cpp
OpenMP/target_teams_distribute_parallel_for_if_codegen.cpp
OpenMP/target_teams_distribute_parallel_for_simd_if_codegen.cpp
OpenMP/teams_distribute_parallel_for_if_codegen.cpp
OpenMP/teams_distribute_parallel_for_simd_if_codegen.cpp
This reverts commit 90eaedda9b.
Failing tests on Arm due to the tests automatically populating
incomatible pointer width architectures. Reverting until the tests are
updated. Failing tests:
OpenMP/distribute_parallel_for_num_threads_codegen.cpp
OpenMP/distribute_parallel_for_if_codegen.cpp
OpenMP/distribute_parallel_for_simd_if_codegen.cpp
OpenMP/distribute_parallel_for_simd_num_threads_codegen.cpp
OpenMP/target_teams_distribute_parallel_for_if_codegen.cpp
OpenMP/target_teams_distribute_parallel_for_simd_if_codegen.cpp
OpenMP/teams_distribute_parallel_for_if_codegen.cpp
OpenMP/teams_distribute_parallel_for_simd_if_codegen.cpp
This reverts commit 9d2378b591.
Also, use the StructuredData::Dump method to print the StructuredData if there
is no plugin, rather than just returning an error.
Differential Revision: https://reviews.llvm.org/D88266
Per the DAP spec for SetBreakpoints [1], the way to clear breakpoints is: `To clear all breakpoint for a source, specify an empty array.`
However, leaving the breakpoints field unset is also a well formed request (note the `breakpoints?:` in the `SetBreakpointsArguments` definition). If it's unset, we have a couple choices:
1. Crash (current behavior)
2. Clear breakpoints
3. Return an error response that the breakpoints field is missing.
I propose we do (2) instead of (1), and treat an unset breakpoints field the same as an empty breakpoints field.
[1] https://microsoft.github.io/debug-adapter-protocol/specification#Requests_SetBreakpoints
Reviewed By: wallace, labath
Differential Revision: https://reviews.llvm.org/D88513
When running in an ipv6-only environment where `AF_INET` sockets are not available, many lldb tests (mostly gdb remote tests) fail because things like `127.0.0.1` don't work there.
Use `localhost` instead of `127.0.0.1` whenever possible, or include a fallback of creating `AF_INET6` sockets when `AF_INET` fails.
Reviewed By: labath
Differential Revision: https://reviews.llvm.org/D87333
This is part of the Propeller framework to do post link code layout optimizations. Please see the RFC here: https://groups.google.com/forum/#!msg/llvm-dev/ef3mKzAdJ7U/1shV64BYBAAJ and the detailed RFC doc here: https://github.com/google/llvm-propeller/blob/plo-dev/Propeller_RFC.pdf
This patch provides exception support for basic block sections by splitting the call-site table into call-site ranges corresponding to different basic block sections. Still all landing pads must reside in the same basic block section (which is guaranteed by the the core basic block section patch D73674 (ExceptionSection) ). Each call-site table will refer to the landing pad fragment by explicitly specifying @LPstart (which is omitted in the normal non-basic-block section case). All these call-site tables will share their action and type tables.
The C++ ABI somehow assumes that no landing pads point directly to LPStart (which works in the normal case since the function begin is never a landing pad), and uses LP.offset = 0 to specify no landing pad. In the case of basic block section where one section contains all the landing pads, the landing pad offset relative to LPStart could actually be zero. Thus, we avoid zero-offset landing pads by inserting a **nop** operation as the first non-CFI instruction in the exception section.
**Background on Exception Handling in C++ ABI**
https://github.com/itanium-cxx-abi/cxx-abi/blob/master/exceptions.pdf
Compiler emits an exception table for every function. When an exception is thrown, the stack unwinding library queries the unwind table (which includes the start and end of each function) to locate the exception table for that function.
The exception table includes a call site table for the function, which is used to guide the exception handling runtime to take the appropriate action upon an exception. Each call site record in this table is structured as follows:
| CallSite | --> Position of the call site (relative to the function entry)
| CallSite length | --> Length of the call site.
| Landing Pad | --> Position of the landing pad (relative to the landing pad fragment’s begin label)
| Action record offset | --> Position of the first action record
The call site records partition a function into different pieces and describe what action must be taken for each callsite. The callsite fields are relative to the start of the function (as captured in the unwind table).
The landing pad entry is a reference into the function and corresponds roughly to the catch block of a try/catch statement. When execution resumes at a landing pad, it receives an exception structure and a selector value corresponding to the type of the exception thrown, and executes similar to a switch-case statement. The landing pad field is relative to the beginning of the procedure fragment which includes all the landing pads (@LPStart). The C++ ABI requires all landing pads to be in the same fragment. Nonetheless, without basic block sections, @LPStart is the same as the function @Start (found in the unwind table) and can be omitted.
The action record offset is an index into the action table which includes information about which exception types are caught.
**C++ Exceptions with Basic Block Sections**
Basic block sections break the contiguity of a function fragment. Therefore, call sites must be specified relative to the beginning of the basic block section. Furthermore, the unwinding library should be able to find the corresponding callsites for each section. To do so, the .cfi_lsda directive for a section must point to the range of call-sites for that section.
This patch introduces a new **CallSiteRange** structure which specifies the range of call-sites which correspond to every section:
`struct CallSiteRange {
// Symbol marking the beginning of the precedure fragment.
MCSymbol *FragmentBeginLabel = nullptr;
// Symbol marking the end of the procedure fragment.
MCSymbol *FragmentEndLabel = nullptr;
// LSDA symbol for this call-site range.
MCSymbol *ExceptionLabel = nullptr;
// Index of the first call-site entry in the call-site table which
// belongs to this range.
size_t CallSiteBeginIdx = 0;
// Index just after the last call-site entry in the call-site table which
// belongs to this range.
size_t CallSiteEndIdx = 0;
// Whether this is the call-site range containing all the landing pads.
bool IsLPRange = false;
};`
With N basic-block-sections, the call-site table is partitioned into N call-site ranges.
Conceptually, we emit the call-site ranges for sections sequentially in the exception table as if each section has its own exception table. In the example below, two sections result in the two call site ranges (denoted by LSDA1 and LSDA2) placed next to each other. However, their call-sites will refer to records in the shared Action Table. We also emit the header fields (@LPStart and CallSite Table Length) for each call site range in order to place the call site ranges in separate LSDAs. We note that with -basic-block-sections, The CallSiteTableLength will not actually represent the length of the call site table, but rather the reference to the action table. Since the only purpose of this field is to locate the action table, correctness is guaranteed.
Finally, every call site range has one @LPStart pointer so the landing pads of each section must all reside in one section (not necessarily the same section). To make this easier, we decide to place all landing pads of the function in one section (hence the `IsLPRange` field in CallSiteRange).
| @LPStart | ---> Landing pad fragment ( LSDA1 points here)
| CallSite Table Length | ---> Used to find the action table.
| CallSites |
| … |
| … |
| @LPStart | ---> Landing pad fragment ( LSDA2 points here)
| CallSite Table Length |
| CallSites |
| … |
| … |
…
…
| Action Table |
| Types Table |
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D73739
since that is the normal behaviour of other compilers on the platform.
Reviewed By: hubert.reinterpretcast
Differential Revision: https://reviews.llvm.org/D88500
Summary:
Replace the OpenMP Runtime Library functions used in CGOpenMPRuntimeGPU
for OpenMP device code generation with ones in OMPKinds.def and use
OMPIRBuilder for generating runtime calls. This allows us to consolidate
more OpenMP code generation into the OMPIRBuilder. This patch also
invalidates specifying target architectures with conflicting pointer
sizes.
Reviewers: jdoerfert
Subscribers: aaron.ballman cfe-commits guansong llvm-commits sstefan1 yaxunl
Tags: #OpenMP #Clang #LLVM
Differential Revision: https://reviews.llvm.org/D88430
Summary:
This patch adds an error to Clang that detects if OpenMP offloading is used
between two architectures with incompatible pointer sizes. This ensures that
the data mapping can be done correctly and solves an issue in code generation
generating the wrong size pointer.
Reviewer: jdoerfert
Subscribers:
Tags: #OpenMP #Clang
Differential Revision:
We previously took a shortcut and said that weak variables never have
constant initializers (because those initializers are never correct to
use outside the variable). We now say that weak variables can have
constant initializers, but are never usable in constant expressions.
The current half vector was enforcing an assert expecting
"(LHS is half vector) == (RHS is half vector)"
for comma.
Reviewed By: ahatanak, fhahn
Differential Revision: https://reviews.llvm.org/D88265
This goes with the APFloat change proposed in
D88238.
This is copied from the MIPS-specific test in
builtin-nan-legacy.c to verify that the normal
behavior is correct on other targets without the
complication of an inverted quiet bit.
When pairing loads, we should check if in between the two loads the
base register has been modified. If that is the case then avoid pairing
them because the second load actually loads from a different address.
Reviewed By: fhahn
Differential Revision: https://reviews.llvm.org/D86956
`Posix/unpoison-alternate-stack.cpp` currently `FAIL`s on Solaris/i386.
Some of the problems are generic:
- `clang` warns compiling the testcase:
compiler-rt/test/asan/TestCases/Posix/unpoison-alternate-stack.cpp:83:7: warning: nested designators are a C99 extension [-Wc99-designator]
.sa_sigaction = signalHandler,
^~~~~~~~~~~~~
compiler-rt/test/asan/TestCases/Posix/unpoison-alternate-stack.cpp:84:7: warning: ISO C++ requires field designators to be specified in declaration order; field '_funcptr' will be initialized after field 'sa_flags' [-Wreorder-init-list]
.sa_flags = SA_SIGINFO | SA_NODEFER | SA_ONSTACK,
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
and some more instances. This can all easily be avoided by initializing
each field separately.
- The test `SEGV`s in `__asan_memcpy`. The default Solaris/i386 stack size
is only 4 kB, while `__asan_memcpy` tries to allocate either 5436
(32-bit) or 10688 bytes (64-bit) on the stack. This patch avoids this by
requiring at least 16 kB stack size.
- Even without `-fsanitize=address` I get an assertion failure:
Assertion failed: !isOnSignalStack(), file compiler-rt/test/asan/TestCases/Posix/unpoison-alternate-stack.cpp, line 117
The fundamental problem with this testcase is that `longjmp` from a
signal handler is highly unportable; XPG7 strongly warns against it and
it is thus unspecified which stack is used when `longjmp`ing from a
signal handler running on an alternative stack.
So I'm `XFAIL`ing this testcase on Solaris.
Tested on `amd64-pc-solaris2.11` and `x86_64-pc-linux-gnu`.
Differential Revision: https://reviews.llvm.org/D88501
Move some of the flags previously in Options, as well as the
UseMemoryTagging flag previously in the primary allocator, into an
atomic variable so that it can be updated while other threads are
running. Relaxed accesses are used because we only have the requirement
that the other threads see the new value eventually.
The code is set up so that the variable is generally loaded once per
allocation function call with the exception of some rarely used code
such as error handlers. The flag bits can generally stay in a register
during the execution of the allocation function which means that they
can be branched on with minimal overhead (e.g. TBZ on aarch64).
Differential Revision: https://reviews.llvm.org/D88523
This is mostly for the benefit of the LBR latency mode.
Right now, it performs no checking. If this is run on non-supported hardware, it will produce all zeroes for latency.
Differential Revision: https://reviews.llvm.org/D85254
Switch to a dummy op in the test dialect so we can remove the -allow-unregistred-dialect
on ops.mlir and invalid.mlir. Change after comment on D88272.
Reviewed By: mehdi_amini
Differential Revision: https://reviews.llvm.org/D88587
This patch achieves two things:
1. It breaks up the `join_blocks` interface between the SDA to the DA to
return two separate sets for divergent loops exits and divergent,
disjoint path joins.
2. It updates the SDA algorithm to run in O(n) time and improves the
precision on divergent loop exits.
This fixes `https://bugs.llvm.org/show_bug.cgi?id=46372` (by virtue of
the improved `join_blocks` interface) and revealed an imprecise expected
result in the `Analysis/DivergenceAnalysis/AMDGPU/hidden_loopdiverge.ll`
test.
Reviewed By: sameerds
Differential Revision: https://reviews.llvm.org/D84413
All the state of VRAI is allocator-wide, so we can avoid creating it
every time we need it. In addition, the normalization function is
allocator-specific. In a next change, we can simplify that design in
favor of just having it as a virtual member.
Differential Revision: https://reviews.llvm.org/D88499
As mentioned on PR47191, if we're bswap'ing some bytes and the 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.
Jessica Paquette