Hongtao Yu [Tue, 15 Sep 2020 23:12:12 +0000 (16:12 -0700)]
[CSSPGO] Turn on Profi by default
As titled.
Reviewed By: wenlei, wlei
Differential Revision: https://reviews.llvm.org/D115011
Mehdi Amini [Thu, 2 Dec 2021 17:06:18 +0000 (17:06 +0000)]
Using make_unique instead of `new` (NFC)
Fix a clang-tidy warning.
Geoffrey Martin-Noble [Fri, 3 Dec 2021 01:33:19 +0000 (17:33 -0800)]
[Bazel] Set the right default for LLVM_WINDOWS_PREFER_FORWARD_SLASH on Windows
This cmake configure option was added in
df0ba47c36f6bd0865e3286853b76d37e037c2d7, and was ported to
Bazel in
7d323dc7738e3152c4bd54a23ac27554bfbbf583.
However, the setting chosen in Bazel seems accidental, not necessarily
intentional.
LLVM_WINDOWS_PREFER_FORWARD_SLASH has no effect on Unix, and on
Windows, setting it to 0 is the default, which gets the same behaviour
as before. Setting it to 1 enables new experimental behaviours
(which is enabled by default on MinGW targets only).
As I don't see any explicit intent to opt in to the new experimental
behaviour, I believe the current configuration in bazel was a
mistake.
Differential Revision: https://reviews.llvm.org/D114065
Keith Smiley [Thu, 2 Dec 2021 19:10:41 +0000 (11:10 -0800)]
[clang][Darwin] Remove old lld implementation handling
This now assumes that for the darwin driver any lld is the "new" macho
lld implementation.
Differential Revision: https://reviews.llvm.org/D114974
Reid Kleckner [Fri, 3 Dec 2021 00:25:28 +0000 (16:25 -0800)]
[bazel] Update static analyzer unit test deps for clangTesting
Mingming Liu [Fri, 3 Dec 2021 00:21:11 +0000 (00:21 +0000)]
Run update_test_checks.py on test cases.
In this way, each instruction has a line, and diffs will be more clear.
Differential Revision: https://reviews.llvm.org/D115006
Reid Kleckner [Fri, 3 Dec 2021 00:15:38 +0000 (16:15 -0800)]
[Bazel] Add LLVM_ENABLE_CURL to Bazel llvm-config.h.cmake for
e0b259f2
Daniil Fukalov [Fri, 29 Oct 2021 15:29:57 +0000 (18:29 +0300)]
[CostModel][AMDGPU] Fix instructions costs estimation for vector types.
1. Fixed vector instructions costs estimations incosistency - removed different
logic for "not simple types" since it biases costs for these types.
2. Fixed legalization penalty for vectors too big for the target: changed from
overwrite default legalization cost value estimation to added penalty.
3. Fixed few typos in tests.
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D114893
Greg Clayton [Thu, 2 Dec 2021 23:47:15 +0000 (15:47 -0800)]
Include extra input contents on this test so we can see why lldb-arm-ubuntu buildbot is failing.
Only lldb-arm-ubuntu is failing after https://reviews.llvm.org/D114288 and there isn't enough input context to see why this is failing. It works on x86_64 linux just fine.
Mogball [Thu, 2 Dec 2021 23:42:47 +0000 (23:42 +0000)]
[mlir][ods] update attr/type def format docs
Vy Nguyen [Thu, 2 Dec 2021 20:22:43 +0000 (15:22 -0500)]
[clang-tidy][objc] Finds and fixes improper usages of XCTAssertEquals and XCTAssertNotEquals.
Using XCTAssertEqual on NSString* objects is almost always wrong.
Unfortunately, we have seen a lot of tests doing this and reyling on pointer equality for strings with the same values (which happens to work sometimes - depending on the linker, but this assumption is not guaranteed by the language)
These fixes would make tests less brittle.
Differential Revision: https://reviews.llvm.org/D114975
Steven Wan [Thu, 2 Dec 2021 23:29:43 +0000 (18:29 -0500)]
[analyzer]Skip unstable CSA tests failing on several platforms
Clang static analyzer uses bitwidth to infer the integer value type, that is, any 32-bit integer is considered of type `int`, and any 64-bit integer is considered of type `long`. This isn't always true, for instance, in ILP32 (e.g., 32-bit AIX), 32-bit could be `long`, and in LP64 (e.g., 64-bit wasm64), 64-bit could be `long long`.
Reviewed By: steakhal
Differential Revision: https://reviews.llvm.org/D114454
George Koehler [Thu, 2 Dec 2021 23:28:37 +0000 (15:28 -0800)]
[ELF][PPC32] Make R_PPC32_PLTREL retain .got
PLT usage needs the first 12 bytes of the .got section. We need to keep .got and
DT_GOT_PPC even if .got/_GLOBAL_OFFSET_TABLE_ are not referenced (large PIC code
may only reference .got2), which is the case in OpenBSD's ld.so, leading
to a misleading error, "unsupported insecure BSS PLT object".
Fix this by adding R_PPC32_PLTREL to the list of hasGotOffRel.
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D114982
Matt Arsenault [Thu, 2 Dec 2021 21:31:56 +0000 (16:31 -0500)]
AMDGPU: Sanitized functions require implicit arguments
Do not infer no-amdgpu-implicitarg-ptr for sanitized functions. If a
function is explicitly marked amdgpu-no-implicitarg-ptr and
sanitize_address, infer that it is required.
Ulysse Beaugnon [Thu, 2 Dec 2021 22:49:02 +0000 (23:49 +0100)]
[MLIR] Use a shared uniquer for affine maps and integer sets.
Affine maps and integer sets previously relied on a single lock for creating unique instances. In a multi-threaded setting, this lock becomes a contention point. This commit updates AffineMap and IntegerSet to use StorageUniquer instead. StorageUniquer internally uses sharded locks and thread-local caches to reduce contention. It is already used for affine expressions, types and attributes. On my local machine, this gives me a 5X speedup for an application that manipulates a lot of affine maps and integer sets.
This commit also removes the integer set uniquer threshold. The threshold was used to avoid adding integer sets with a lot of constraints to the hash_map containing unique instances, but the constraints and the integer set were still allocated in the same allocator and never freed, thus not saving any space expect for the hash-map entry.
Reviewed By: rriddle
Differential Revision: https://reviews.llvm.org/D114942
Vitaly Buka [Thu, 2 Dec 2021 20:44:38 +0000 (12:44 -0800)]
[NFC][sanitizer] Remove SetSoftRssLimitExceededCallback
According comments on D44404, something like that was the goal.
Reviewed By: morehouse, kstoimenov
Differential Revision: https://reviews.llvm.org/D114991
Vitaly Buka [Thu, 2 Dec 2021 21:27:11 +0000 (13:27 -0800)]
[NFC][sanitizer] Reorder functions in cpp file
David Blaikie [Thu, 2 Dec 2021 20:16:10 +0000 (12:16 -0800)]
libcxx pretty printers: remove non-lazy_string fallback
This has been supported on gdb for something like ~10 years, so doesn't
seem necessary to carry a fallback.
Differential Revision: https://reviews.llvm.org/D114986
Paul Robinson [Thu, 2 Dec 2021 22:28:07 +0000 (14:28 -0800)]
[TLI checker] Follow good practice with -COUNT directives
FileCheck's -COUNT suffix doesn't fail if there are more matches
than you asked for, so it's good practice to put a -NOT after.
Yitzhak Mandelbaum [Thu, 2 Dec 2021 19:52:47 +0000 (19:52 +0000)]
[clang-tidy] Allow disabling support for NOLINTBEGIN/NOLINTEND blocks.
This patch parameterizes the clang-tidy diagnostic consumer with a boolean that
controls whether to honor NOLINTBEGIN/NOLINTEND blocks. The current support for
scanning these blocks is very costly -- O(n*m) in the size of files (n) and
number of diagnostics found (m), with a large constant factor. So, the patch
allows clients to disable it.
Future patches should make the feature more efficient, but this will mitigate in
the interim.
Differential Revision: https://reviews.llvm.org/D114981
Groverkss [Thu, 2 Dec 2021 22:14:36 +0000 (03:44 +0530)]
[MLIR][FlatAffineConstraints] Remove duplicate divisions while merging local ids
This patch implements detecting duplicate local identifiers by extracting their
division representation while merging local identifiers.
For example, given the FACs A, B:
```
A: (x, y)[s0] : (exists d0 = [x / 4], d1 = [y / 4]: d0 <= s0, d1 <= s0, x + y >= 2)
B: (x, y)[s0] : (exists d0 = [x / 4], d1 = [y / 4]: d0 <= s0, d1 <= s0, x + y >= 5)
```
The intersection of A and B without this patch would lead to the following FAC:
```
(x, y)[s0] : (exists d0 = [x / 4], d1 = [y / 4], d2 = [x / 4], d3 = [x / 4]: d0 <= s0, d1 <= s0, d2 <= s0, d3 <= s0, x + y >= 2, x + y >= 5)
```
after this patch, merging of local ids will detect that `d0 = d2` and `d1 = d3`,
and the intersection of these two FACs will be (after removing duplicate constraints):
```
(x, y)[s0] : (exists d0 = [x / 4], d1 = [y / 4] : d0 <= s0, d1 <= s0, x + y >= 2, x + y >= 5)
```
This reduces the number of constraints by 2 (constraints) + 4 (2 constraints for each extra division) for this case.
This is used to reduce the output size representation of operations like
PresburgerSet::subtract, PresburgerSet::intersect which require merging local
variables.
Reviewed By: arjunp, bondhugula
Differential Revision: https://reviews.llvm.org/D112867
Groverkss [Thu, 2 Dec 2021 22:08:00 +0000 (03:38 +0530)]
Revert changes that should have been sent as a patch
Revert changes that were meant to be sent as a single commit with
summary for the differential review, but were accidently sent directly.
This reverts commit
3bc5353fc6f291d6ac12c256600b5b05d7de8f74.
Alexey Bataev [Tue, 29 Jun 2021 20:56:00 +0000 (13:56 -0700)]
[OPENMP]Fix PR50733: unexpected final value of list-item in linear clause in loop construct.
Currently the last value of linear is calculated as var = init + num_iters * step.
Replaced it with var = var_priv, i.e. original variable gets the value
of the last private copy.
Differential Revision: https://reviews.llvm.org/D105151
Amy Kwan [Thu, 2 Dec 2021 21:24:06 +0000 (15:24 -0600)]
[PowerPC] Fix load/store selection infrastructure when load/store intrinsics are used on P10.
The load/store infrastructure previously made an incorrect assumption that
whenever it is used with a load/store intrinsic on Power10 - those intrinsics
would automatically be the lxvp/stxvp intrinsics introduced in Power10.
However, this is obviously not the case as there are multiple instances of
pre-P10 intrinsics that use the refactored load/store implementation.
This patch corrects this assumption, and produces the expected intrinsic on pre-P10.
Differential Revision: https://reviews.llvm.org/D114978
Groverkss [Thu, 2 Dec 2021 21:39:19 +0000 (03:09 +0530)]
Address bondhugula's comments
Groverkss [Thu, 18 Nov 2021 17:04:36 +0000 (22:34 +0530)]
Addressed comments
Groverkss [Thu, 18 Nov 2021 15:34:00 +0000 (21:04 +0530)]
Fix doc comment for mergeLocalIds.
Groverkss [Thu, 18 Nov 2021 14:51:46 +0000 (20:21 +0530)]
Address more comments.
Groverkss [Fri, 12 Nov 2021 22:57:37 +0000 (04:27 +0530)]
Addressed more comments
Groverkss [Thu, 11 Nov 2021 17:12:04 +0000 (22:42 +0530)]
Fix asserts as suggested by Arjun
Groverkss [Wed, 10 Nov 2021 10:18:48 +0000 (15:48 +0530)]
Fix clang-format errors
Groverkss [Mon, 8 Nov 2021 22:02:40 +0000 (03:32 +0530)]
Update docs
Groverkss [Mon, 8 Nov 2021 21:43:43 +0000 (03:13 +0530)]
Update tests for mergeLocalIds
Groverkss [Mon, 8 Nov 2021 21:43:25 +0000 (03:13 +0530)]
Address arjun's comments
Groverkss [Sun, 7 Nov 2021 05:15:20 +0000 (10:45 +0530)]
Move division representation to a common function
Groverkss [Sat, 6 Nov 2021 10:45:51 +0000 (16:15 +0530)]
Update mergeLocalIds docs
Groverkss [Sun, 31 Oct 2021 21:34:08 +0000 (03:04 +0530)]
Update docs for mergeLocalIds
Groverkss [Sat, 30 Oct 2021 12:12:47 +0000 (17:42 +0530)]
Implement division merging
Jake Egan [Thu, 2 Dec 2021 21:39:52 +0000 (16:39 -0500)]
[LTO] Specify triple to address unknown binary format assertion
On AIX, this test generates an XCOFF file and hits "unknown binary format" assertion in llvm-nm. This patch specifies the triple to mitigate this issue.
Reviewed By: steven_wu
Differential Revision: https://reviews.llvm.org/D114683
Alexey Bataev [Wed, 24 Nov 2021 19:21:10 +0000 (11:21 -0800)]
[OPENMP]Fix PR51327: Range based for loop not working if range's type is a template.
Need to postpone anlysis of the ranged for loops till the actual
instantiation to avoid erroneous emission of error messages.
Differential Revision: https://reviews.llvm.org/D114560
Mogball [Thu, 2 Dec 2021 21:25:04 +0000 (21:25 +0000)]
[mlir][ods] fix defgen on empty files
Philip Reames [Thu, 2 Dec 2021 20:56:09 +0000 (12:56 -0800)]
[funcattrs] Infer writeonly argument attribute
This change extends the current logic for inferring readonly and readnone argument attributes to also infer writeonly.
This change is deliberately minimal; there's a couple of areas for follow up.
* I left out all call handling and thus any benefit from the SCC walk. When examining the test changes, I realized the existing code is imprecise, and am going to fix that in it's own revision before adding in the writeonly handling. (Mostly because updating the tests is hard when I, the human, can't figure out whether the result is correct.)
* I left out handling for storing a value (as opposed to storing to a pointer). This should benefit readonly/readnone as well, and applies to a bunch of other instructions. Seemed worth having as a separate review.
Differential Revision: https://reviews.llvm.org/D114963
Alexey Bataev [Fri, 26 Nov 2021 15:36:25 +0000 (07:36 -0800)]
[OPENMP]Fix error emission for dependent expressions in iterators for depend clauses.
Need to postpone analysis for addressable lvalue in a depend clause with
iterators, otherwise the incorrect error message is emitted.
Differential Revision: https://reviews.llvm.org/D114653
Ron Lieberman [Thu, 2 Dec 2021 20:58:19 +0000 (20:58 +0000)]
Restric xfail on openmp/libomptarget/test/mapping/reduction_implicit_map.cpp to amdgcn-amd-amdhsa
Vitaly Buka [Thu, 2 Dec 2021 20:47:47 +0000 (12:47 -0800)]
[NFC][sanitizer] Rename RssLimitExceeded -> IsRssLimitExceeded
Nico Weber [Thu, 2 Dec 2021 20:50:56 +0000 (15:50 -0500)]
[gn build] (manually) port
e0b259f22c003ffe9
Ron Lieberman [Thu, 2 Dec 2021 20:32:06 +0000 (20:32 +0000)]
xfail: libomptarget reduction_implicit_map.cpp after reapply of Start calling setTargetAttributes
Noah Shutty [Thu, 2 Dec 2021 19:33:04 +0000 (19:33 +0000)]
[llvm] [Support] Add CURL HTTP Client.
Provides an implementation of `HTTPClient` that wraps libcurl.
Reviewed By: dblaikie
Differential Revision: https://reviews.llvm.org/D112753
Ron Lieberman [Thu, 2 Dec 2021 20:29:30 +0000 (20:29 +0000)]
Revert "xfail: reduction_implicit_map.cpp after reapply of Start calling setTargetAttributes"
This reverts commit
8b646f212e9ccc8e0fbf1a84603d0a40bfca1338.
Kirill Stoimenov [Thu, 2 Dec 2021 20:19:29 +0000 (20:19 +0000)]
[ASan] Fixed asan_mapping.h format.
Reviewed By: kstoimenov
Differential Revision: https://reviews.llvm.org/D114987
Ron Lieberman [Thu, 2 Dec 2021 20:22:11 +0000 (20:22 +0000)]
xfail: reduction_implicit_map.cpp after reapply of Start calling setTargetAttributes
Nikolas Klauser [Thu, 2 Dec 2021 13:12:51 +0000 (14:12 +0100)]
[libc++][NFC] Update namespace comments in include/
update the namspace comments in include/
Reviewed By: ldionne, #libc
Spies: smeenai, libcxx-commits
Differential Revision: https://reviews.llvm.org/D114947
spupyrev [Thu, 2 Dec 2021 19:59:33 +0000 (11:59 -0800)]
profi - a flow-based profile inference algorithm: Part III (out of 3)
This is a continuation of D109860 and D109903.
An important challenge for profile inference is caused by the fact that the
sample profile is collected on a fully optimized binary, while the block and
edge frequencies are consumed on an early stage of the compilation that operates
with a non-optimized IR. As a result, some of the basic blocks may not have
associated sample counts, and it is up to the algorithm to deduce missing
frequencies. The problem is illustrated in the figure where three basic
blocks are not present in the optimized binary and hence, receive no samples
during profiling.
We found that it is beneficial to treat all such blocks equally. Otherwise the
compiler may decide that some blocks are “cold” and apply undesirable
optimizations (e.g., hot-cold splitting) regressing the performance. Therefore,
we want to distribute the counts evenly along the blocks with missing samples.
This is achieved by a post-processing step that identifies "dangling" subgraphs
consisting of basic blocks with no sampled counts; once the subgraphs are
found, we rebalance the flow so as every branch probability is 50:50 within the
subgraphs.
Our experiments indicate up to 1% performance win using the optimization on
some binaries and a significant improvement in the quality of profile counts
(when compared to ground-truth instrumentation-based counts)
{
F19093045}
Reviewed By: hoy
Differential Revision: https://reviews.llvm.org/D109980
Nico Weber [Thu, 2 Dec 2021 20:00:40 +0000 (15:00 -0500)]
[gn build] (manually) port
9e3552523ebd (no more old mach-o lld)
Fangrui Song [Thu, 2 Dec 2021 19:58:25 +0000 (11:58 -0800)]
[ELF] Hint -z nostart-stop-gc for __start_ undefined references
Make users aware what to do with ld.lld 13.0.0 / GNU ld<2015-10 --gc-sections
behavior.
Differential Revision: https://reviews.llvm.org/D114830
Matt Arsenault [Mon, 29 Nov 2021 20:47:42 +0000 (15:47 -0500)]
Reapply "OpenMP: Start calling setTargetAttributes for generated kernels"
This reverts commit
25eb7fa01d7ebbe67648ea03841cda55b4239ab2.
Previous buildbot failures appear to have been a fluke from a dirty
build.
Vitaly Buka [Thu, 2 Dec 2021 02:41:54 +0000 (18:41 -0800)]
[NFC][sanitizer] Use more bytes of sanitizer_stack_store_test pointers
Leonard Chan [Thu, 2 Dec 2021 19:45:41 +0000 (11:45 -0800)]
[compiler-rt] Fix incorrect variable names used
Vitaly Buka [Thu, 2 Dec 2021 08:41:41 +0000 (00:41 -0800)]
[sanitizer] Start background thread once
Reviewed By: morehouse
Differential Revision: https://reviews.llvm.org/D114933
Keith Smiley [Thu, 2 Dec 2021 19:24:08 +0000 (11:24 -0800)]
[Bazel] Remove old macho lld port
This code and cmake was removed in https://reviews.llvm.org/D114842
Differential Revision: https://reviews.llvm.org/D114976
Vitaly Buka [Thu, 2 Dec 2021 08:53:02 +0000 (00:53 -0800)]
[asan] Remove confusing workaround
The goal is to identify the bot and try to fix it.
SetSoftRssLimitExceededCallback is AsanInitInternal as I assume
that only MaybeStartBackgroudThread needs to be delayed to constructors.
Later I want to move MaybeStartBackgroudThread call into sanitizer_common.
If it needs to be reverted please provide to more info, like bot, or details about setup.
Reviewed By: kstoimenov
Differential Revision: https://reviews.llvm.org/D114934
David Blaikie [Fri, 5 Nov 2021 05:45:55 +0000 (22:45 -0700)]
Simplify the libcxx std::string_view gdb pretty printer
Seems better to rely on the existing formatting, makes the output
smaller/simpler - this is consistent with libstdc++'s std::string_view
pretty printing too.
Differential Revision: https://reviews.llvm.org/D113244
Reid Kleckner [Thu, 2 Dec 2021 19:30:26 +0000 (11:30 -0800)]
[Bazel] Remove old MachO LLD from the Bazel build
Updates Bazel files for
9e3552523ebd3385487e01e3e7af37b8c0efaf57
Keith Smiley [Wed, 1 Dec 2021 06:01:07 +0000 (22:01 -0800)]
[lld-macho] Remove old macho darwin lld
During the llvm round table it was generally agreed that the newer macho
lld implementation is feature complete enough to replace the old
implementation entirely. This will reduce confusion for new users who
aren't aware of the history.
Differential Revision: https://reviews.llvm.org/D114842
spupyrev [Thu, 2 Dec 2021 18:19:32 +0000 (10:19 -0800)]
profi - a flow-based profile inference algorithm: Part II (out of 3)
This is a continuation of D109860.
Traditional flow-based algorithms cannot guarantee that the resulting edge
frequencies correspond to a *connected* flow in the control-flow graph. For
example, for an instance in the attached figure, a flow-based (or any other)
inference algorithm may produce an output in which the hot loop is disconnected
from the entry block (refer to the rightmost graph in the figure). Furthermore,
creating a connected minimum-cost maximum flow is a computationally NP-hard
problem. Hence, we apply a post-processing adjustments to the computed flow
by connecting all isolated flow components ("islands").
This feature helps to keep all blocks with sample counts connected and results
in significant performance wins for some binaries.
{
F19077343}
Reviewed By: hoy
Differential Revision: https://reviews.llvm.org/D109903
Aart Bik [Tue, 30 Nov 2021 23:34:23 +0000 (15:34 -0800)]
[mlir][bufferization] fixed typo in to_memref doc
Reviewed By: pifon2a
Differential Revision: https://reviews.llvm.org/D114824
Alexey Bataev [Thu, 2 Dec 2021 12:22:55 +0000 (04:22 -0800)]
[SLP]Fix reused extracts cost.
If the extractelement instruction is used multiple times in the
different tree entries (either vectorized, or gathered), need to
compensate the scalar cost of such instructions. They are completely
removed if all users are part of the tree but we need to compensate the
cost only once for each instruction.
Differential Revision: https://reviews.llvm.org/D114958
Vitaly Buka [Mon, 22 Nov 2021 05:54:08 +0000 (21:54 -0800)]
[sanitizer] Add delta compression stack depot
Compress by factor 4x, takes about 10ms per 8 MiB block.
Depends on D114498.
Reviewed By: morehouse
Differential Revision: https://reviews.llvm.org/D114503
Vitaly Buka [Wed, 24 Nov 2021 00:31:26 +0000 (16:31 -0800)]
[sanitizer] Add compress_stack_depot flag
Depends on D114494.
Reviewed By: morehouse
Differential Revision: https://reviews.llvm.org/D114495
Reid Kleckner [Thu, 2 Dec 2021 18:19:54 +0000 (10:19 -0800)]
[ELF] Fix driver.test after
8c3641d0 when cwd is readonly
Vitaly Buka [Thu, 2 Dec 2021 07:52:31 +0000 (23:52 -0800)]
[sanitizer] DEFINE_REAL_PTHREAD_FUNCTIONS for hwasan, lsan, msan
It should be NFC, as they already intercept pthread_create.
This will let us to fix BackgroundThread for these sanitizerts.
In in followup patches I will fix MaybeStartBackgroudThread for them
and corresponding tests.
Reviewed By: kstoimenov
Differential Revision: https://reviews.llvm.org/D114935
Stella Stamenova [Thu, 2 Dec 2021 17:53:28 +0000 (09:53 -0800)]
[lldb] Skip two lldb tests on Windows because they are flaky
These tests work fine with VS2017, but become more flaky with VS2019 and the buildbot is about to get upgraded.
Differential Revision: https://reviews.llvm.org/D114907
Mircea Trofin [Thu, 2 Dec 2021 17:50:09 +0000 (09:50 -0800)]
[bazel][mlgo] Remove the mlgo-related build excludes
They aren't needed anymore, we handle conditional compilation in those
files.
Reviewed By: GMNGeoffrey
Differential Revision: https://reviews.llvm.org/D114970
Florian Hahn [Thu, 2 Dec 2021 17:38:07 +0000 (17:38 +0000)]
[BasicAA] Add tests for strcat/strncat/strcpy.
Florian Hahn [Thu, 2 Dec 2021 17:37:59 +0000 (17:37 +0000)]
[DSE] Read after strcpy test.
Joseph Huber [Thu, 2 Dec 2021 17:28:21 +0000 (12:28 -0500)]
[OpenMP] Remove the new runtime default for AMDGPU
The new runtime is currently broken for AMD offloading. This patch makes
the default the old runtime only for the AMD target.
Reviewed By: ronlieb
Differential Revision: https://reviews.llvm.org/D114965
Kazu Hirata [Thu, 2 Dec 2021 17:27:47 +0000 (09:27 -0800)]
[llvm] Use range-based for loops (NFC)
Kazu Hirata [Thu, 2 Dec 2021 17:20:17 +0000 (09:20 -0800)]
[llvm] Fix "unused variable" warnings
Alexey Bataev [Wed, 1 Dec 2021 18:01:56 +0000 (10:01 -0800)]
[SLP]Outline and fix code for finding common insertelement vectors.
Need to outline the code for finding common vectors in insertelement
instructions into a separate function for future patches. It also
improves the process by adding some extra checks for early exit and
fixes a bug where it always finds the match because of erroneous compare
of the same values.
Differential Revision: https://reviews.llvm.org/D114909
David Green [Thu, 2 Dec 2021 17:10:26 +0000 (17:10 +0000)]
[ARM] Introduce i8neg and i8pos addressing modes
Some instructions with i8 immediate ranges can only hold negative values
(like t2LDRHi8), only hold positive values (like t2STRT) or hold +/-
depending on the U bit (like the pre/post inc instructions. e.g
t2LDRH_POST). This patch splits the AddrModeT2_i8 into AddrModeT2_i8,
AddrModeT2_i8pos and AddrModeT2_i8neg to make this clear.
This allows us to get the offset ranges of t2LDRHi8 correct in the
load/store optimizer, fixing issues where we could end up creating
instructions with positive offsets (which may then be encoded as ldrht).
Differential Revision: https://reviews.llvm.org/D114638
Nico Weber [Thu, 2 Dec 2021 14:12:50 +0000 (09:12 -0500)]
[clang-cl] Define _MSVC_LANG for -std=c++2b
This matches the value that msvc v19.29 VS16.11 uses for
_MSVC_LANG with /std:c++latest.
Differential Revision: https://reviews.llvm.org/D114952
Paul Robinson [Thu, 2 Dec 2021 16:34:16 +0000 (08:34 -0800)]
Reapply "[TLI checker] Add more tests"
This reverts commit
8cd61aac0030b8add686a98b8902ea49ec9c1deb.
I had missed one place in a test that needed updating; it passed on my
dirty build tree but not on a clean one.
Original commit message:
D114478 identified testing gaps; this patch fills them.
Differential Revision: https://reviews.llvm.org/D114913
Dmitry Vyukov [Thu, 2 Dec 2021 12:35:04 +0000 (13:35 +0100)]
tsan: tolerate munmap with invalid arguments
We call UnmapShadow before the actual munmap, at that point we don't yet
know if the provided address/size are sane. We can't call UnmapShadow
after the actual munmap becuase at that point the memory range can
already be reused for something else, so we can't rely on the munmap
return value to understand is the values are sane.
While calling munmap with insane values (non-canonical address, negative
size, etc) is an error, the kernel won't crash. We must also try to not
crash as the failure mode is very confusing (paging fault inside of the
runtime on some derived shadow address).
Such invalid arguments are observed on Chromium tests:
https://bugs.chromium.org/p/chromium/issues/detail?id=1275581
Reviewed By: melver
Differential Revision: https://reviews.llvm.org/D114944
Alexey Bataev [Mon, 15 Nov 2021 17:18:27 +0000 (09:18 -0800)]
[SLP]Improve registering and merging of compatible shuffles.
If several shuffle instructions are emitted, some of them might
same/compatible (less defined) with the previously emitted ones. Such
shuffles can be removed safely, improving the total cost of the
vectorized code.
Differential Revision: https://reviews.llvm.org/D114087
Dmitry Vyukov [Thu, 2 Dec 2021 14:15:14 +0000 (15:15 +0100)]
tsan: fix false positives in dynamic libs with static tls
The added test demonstrates loading a dynamic library with static TLS.
Such static TLS is a hack that allows a dynamic library to have faster TLS,
but it can be loaded only iff all threads happened to allocate some excess
of static TLS space for whatever reason. If it's not the case loading fails with:
dlopen: cannot load any more object with static TLS
We used to produce a false positive because dlopen will write into TLS
of all existing threads to initialize/zero TLS region for the loaded library.
And this appears to be racing with initialization of TLS in the thread
since we model a write into the whole static TLS region (we don't what part
of it is currently unused):
WARNING: ThreadSanitizer: data race (pid=2317365)
Write of size 1 at 0x7f1fa9bfcdd7 by main thread:
0 memset
1 init_one_static_tls
2 __pthread_init_static_tls
[[ this is where main calls dlopen ]]
3 main
Previous write of size 8 at 0x7f1fa9bfcdd0 by thread T1:
0 __tsan_tls_initialization
Fix this by ignoring accesses during dlopen.
Reviewed By: melver
Differential Revision: https://reviews.llvm.org/D114953
Sam Clegg [Thu, 2 Dec 2021 02:10:37 +0000 (18:10 -0800)]
[lld][WebAssembly] Fix for debug relocations against undefined function symbols
This is very similar to https://reviews.llvm.org/D103557 but applies to
symbols which are undefined at link time rather than compile time.
We already have code that handles symbols which were defined at link
time but dead stripped by `--gc-sections` (See
`test/wasm/debug-removed-fn.ll`). In that case the symbols are not live
(!isLive()). However, we can also have live symbols (which are
references by the program) but which are undefined at link time and are
imported by the linker.
In the test case here the symbol `undef` is used but is not defined
in the program but is imported by the linker due to the
`--import-undefined` flag.
Fixes: https://github.com/emscripten-core/emscripten/issues/15528
Differential Revision: https://reviews.llvm.org/D114921
Paul Robinson [Thu, 2 Dec 2021 16:30:47 +0000 (08:30 -0800)]
Revert "[TLI checker] Add more tests"
This reverts commit
2778554971dada8ef7df9ee6954c52a753d90c22.
Some bots are failing on the updated tests.
David Greene [Wed, 6 Oct 2021 19:10:08 +0000 (12:10 -0700)]
[clang] Do not duplicate "EnableSplitLTOUnit" module flag
If clang's output is set to bitcode and LTO is enabled, clang would
unconditionally add the flag to the module. Unfortunately, if the input were a
bitcode or IR file and had the flag set, this would result in two copies of the
flag, which is illegal IR. Guard the setting of the flag by checking whether it
already exists. This follows existing practice for the related "ThinLTO" module
flag.
Differential Revision: https://reviews.llvm.org/D112177
Paul Robinson [Wed, 1 Dec 2021 22:56:00 +0000 (14:56 -0800)]
[TLI checker] Add more tests
D114478 identified testing gaps; this patch fills them.
Differential Revision: https://reviews.llvm.org/D114913
Joseph Huber [Wed, 1 Dec 2021 18:14:40 +0000 (13:14 -0500)]
[OpenMP] Make the new device runtime the default
This patch changes the `-fopenmp-target-new-runtime` option which controls if
the new or old device runtime is used to be true by default. Disabling this to
use the old runtime now requires using `-fno-openmp-target-new-runtime`.
Reviewed By: JonChesterfield, tianshilei1992, gregrodgers, ronlieb
Differential Revision: https://reviews.llvm.org/D114890
Sanjay Patel [Thu, 2 Dec 2021 15:30:06 +0000 (10:30 -0500)]
[InstCombine] add tests for icmp with mul op; NFC
Christian Kühnel [Thu, 2 Dec 2021 13:43:13 +0000 (13:43 +0000)]
[clangd] cleanup of header guard names
Renaming header guards to match the LLVM convention.
This patch was created by automatically applying the fixes from
clang-tidy.
I've removed the [NFC] tag from the title, as we're adding header guards in some files and thus might trigger behavior changes.
Differential Revision: https://reviews.llvm.org/D113896
Florian Hahn [Thu, 2 Dec 2021 15:41:31 +0000 (15:41 +0000)]
[Clang] Fix LTO pipeline test after
770a50b28c00211f9a.
Florian Hahn [Thu, 2 Dec 2021 14:50:14 +0000 (14:50 +0000)]
[AnnotationRemarks] Support generating annotation remarks with -O0.
This matches the legacy pass manager behavior. If remarks are not
enabled the pass is effectively a no-op.
Yitzhak Mandelbaum [Thu, 2 Dec 2021 14:41:00 +0000 (14:41 +0000)]
[clang-tidy] Fix build broken by commit
6a9487df73e917c4faf5e060f2bb33c6ade3f967 (D113148)
Alexey Bataev [Thu, 2 Dec 2021 14:29:27 +0000 (06:29 -0800)]
[SLP][NFC]Add a test for extractelements with many uses vectorization, NFC.
David Stuttard [Tue, 30 Nov 2021 09:29:14 +0000 (09:29 +0000)]
[AMDGPU] Add support for in-order bvh in waitcnt pass
bvh should be handled separately from vmem and vmem with sampler instructions
for waitcnt handling.
Differential Revision: https://reviews.llvm.org/D114794
David Stuttard [Tue, 30 Nov 2021 09:29:14 +0000 (09:29 +0000)]
[AMDGPU] Test for in-order waitcnt insertion for bvh instructions
In-order bvh instructions don't require a waitcnt as order is
guaranteed.
However, waitcnt IS required for other image instruction types vs
bvh.
Pre-commit test for new functionality in https://reviews.llvm.org/D114794
Differential Revision: https://reviews.llvm.org/D114792
Simon Moll [Thu, 2 Dec 2021 12:59:24 +0000 (13:59 +0100)]
[VE][NFC] Cleanup redundant namespace wrapper
Florian Hahn [Thu, 2 Dec 2021 14:18:04 +0000 (14:18 +0000)]
[MemoryLocation] Support strncpy in getForArgument.
The size argument of strncpy can be used as bound for the size of
its pointer arguments.
strncpy is guaranteed to write N bytes and reads up to N bytes.
Reviewed By: xbolva00
Differential Revision: https://reviews.llvm.org/D114871