Petar Jovanovic [Wed, 15 Nov 2017 15:24:04 +0000 (15:24 +0000)]
[mips] Improve genConstMult() to work with arbitrary precision
APInt is now used instead of uint64_t in function genConstMult() allowing
multiplication optimizations with constants of arbitrary length.
Patch by Milos Stojanovic.
Differential Revision: https://reviews.llvm.org/D38130
llvm-svn: 318296
Igor Laevsky [Wed, 15 Nov 2017 15:07:37 +0000 (15:07 +0000)]
[llvm-opt-fuzzer] Add opt fuzzer to the test-depends list.
This should help with the buildbot failures after rL318293.
llvm-svn: 318295
Igor Laevsky [Wed, 15 Nov 2017 13:35:42 +0000 (13:35 +0000)]
[llvm-opt-fuzzer] Only run tests for the x86 target.
This fixes build bot failures after rL318293.
llvm-svn: 318294
Igor Laevsky [Wed, 15 Nov 2017 12:36:57 +0000 (12:36 +0000)]
[llvm-opt-fuzzer] NFC. Add sanity tests.
llvm-svn: 318293
Benjamin Kramer [Wed, 15 Nov 2017 12:20:41 +0000 (12:20 +0000)]
[libclang] Fix cursors for in-class initializer of field declarations
Fixes PR33745.
Patch by Nikolai Kosjar!
Differential Revision: https://reviews.llvm.org/D40027
llvm-svn: 318292
Momchil Velikov [Wed, 15 Nov 2017 12:02:55 +0000 (12:02 +0000)]
[ARM] Split Arm jump table branch into i12 and rs suffixed versions
This is a refactoring/cleanup of Arm `addrmode2` operand class. The patch
removes it completely.
Differential Revision: https://reviews.llvm.org/D39832
llvm-svn: 318291
Alexey Bader [Wed, 15 Nov 2017 11:38:17 +0000 (11:38 +0000)]
[OpenCL] Fix code generation of function-scope constant samplers.
Summary:
Constant samplers are handled as static variables and clang's code generation
library, which leads to llvm::unreachable. We bypass emitting sampler variable
as static since it's translated to a function call later.
Reviewers: yaxunl, Anastasia
Reviewed By: yaxunl, Anastasia
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D34342
llvm-svn: 318290
Jonas Devlieghere [Wed, 15 Nov 2017 10:57:05 +0000 (10:57 +0000)]
[DebugInfo] Fix potential CU mismatch for SubprogramScopeDIEs.
In constructAbstractSubprogramScopeDIE there can be a potential mismatch
between `this` and the CU of ContextDIE when a scope is shared between
two DISubprograms belonging to a different CU. In that case, `this` is
the CU that was specified in the IR, but the CU of ContextDIE is that of
the first subprogram that was emitted. This patch fixes the mismatch by
looking up the CU of ContextDIE, and switching to use that.
This fixes PR35212 (https://bugs.llvm.org/show_bug.cgi?id=35212)
Patch by Philip Craig!
Differential revision: https://reviews.llvm.org/D39981
llvm-svn: 318289
Ilya Biryukov [Wed, 15 Nov 2017 10:50:43 +0000 (10:50 +0000)]
Workaround CodeGen/WebAssembly/cfg-stackify.ll failure after r318202
By disabling the introduced optimization.
llvm-svn: 318288
Sam McCall [Wed, 15 Nov 2017 09:16:29 +0000 (09:16 +0000)]
[clangd] Support returning a limited number of completion results.
Summary:
All results are scored, we only process CodeCompletionStrings for the winners.
We now return CompletionList rather than CompletionItem[] (both are valid).
sortText is now based on CodeCompletionResult::orderedName (mostly the same).
This is the first clangd-only completion option, so plumbing changed.
It requires a small clangd patch (exposing CodeCompletionResult::orderedName).
(This can't usefully be enabled yet: we don't support server-side filtering)
Reviewers: ilya-biryukov
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D39852
llvm-svn: 318287
Sam McCall [Wed, 15 Nov 2017 09:15:06 +0000 (09:15 +0000)]
[clang] Expose orderedString from CodeCompletionResult. NFC
llvm-svn: 318286
Martin Storsjo [Wed, 15 Nov 2017 08:18:28 +0000 (08:18 +0000)]
[MinGW] Ignore the --build-id, --pie-executable and --disable-auto-image-base options
LLD already writes a build id if debug info is enabled.
Some projects set --pie-executable to avoid GNU ld bugs about stripping
base relocations from an executable when they actually are used.
Since -fixed:no is the default (and we don't support setting the
-fixed option via the MinGW frontend), we don't need to handle this.
--disable-auto-image-base is ignored just like --enable-auto-image-base
(as we ignore from before).
Differential Revision: https://reviews.llvm.org/D40031
llvm-svn: 318285
Martin Storsjo [Wed, 15 Nov 2017 08:18:25 +0000 (08:18 +0000)]
[COFF] Always include the size of the string table size field
Even if we don't actually write any string table contents, the
4 byte size for the string table will always be written. Make
sure we accommodate for this in the file size. Since this size
is aligned up, this would seldom be an issue in practice.
Differential Revision: https://reviews.llvm.org/D39891
llvm-svn: 318284
Martin Storsjo [Wed, 15 Nov 2017 08:18:20 +0000 (08:18 +0000)]
[MinGW] Implement the --[no-]gc-sections and --icf options
GNU ld doesn't seem to support --icf at all, but this was suggested
in D39885, and GNU gold seems to support it.
Differential Revision: https://reviews.llvm.org/D40019
llvm-svn: 318283
Martin Storsjo [Wed, 15 Nov 2017 08:18:15 +0000 (08:18 +0000)]
[MinGW] Implement support for the --image-base option
Differential Revision: https://reviews.llvm.org/D40018
llvm-svn: 318282
Martin Storsjo [Wed, 15 Nov 2017 08:18:11 +0000 (08:18 +0000)]
[MinGW] Add support for --dynamicbase, ignore --nxcompat, --tsaware and --high-entropy-va
All of these are disabled by default in GNU ld, but enabled by default
in lld.
Disable dynamicbase by default since it potentially could cause
compatibility issues, but just ignore the others since the lld
default should be fine for most concievable cases.
Differential Revision: https://reviews.llvm.org/D40017
llvm-svn: 318281
Martin Storsjo [Wed, 15 Nov 2017 08:18:06 +0000 (08:18 +0000)]
[MinGW] Handle --large-address-aware
In GNU ld, this option is only available on i386, not on x86_64
(where it's enabled by default with no option to disable it either).
Differential Revision: https://reviews.llvm.org/D40015
llvm-svn: 318280
Mikael Holmen [Wed, 15 Nov 2017 07:46:48 +0000 (07:46 +0000)]
[Lint] Don't warn about passing alloca'd value to tail call if using byval
Summary:
This fixes PR35241.
When using byval, the data is effectively copied as part of the call
anyway, so the pointer returned by the alloca will not be leaked to the
callee and thus there is no reason to issue a warning.
Reviewers: rnk
Reviewed By: rnk
Subscribers: Ka-Ka, llvm-commits
Differential Revision: https://reviews.llvm.org/D40009
llvm-svn: 318279
Craig Topper [Wed, 15 Nov 2017 07:46:43 +0000 (07:46 +0000)]
[X86] Redefine the 128-bit version of VPGATHERQD and VGATHERQPS to use a VK2 mask instead of a VK4 mask.
This allows us to remove extra extend creation during lowering and more accurately reflects the semantics of the instruction.
While there add an extra output VT to X86 masked gather node to better match the isel pattern predicate. Currently we're exploiting the fact that the isel table doesn't count how many output results a node actually has if the result type of any can be inferred from the first result and the type constraints defined in tablegen. I think we might ultimately want to lower all MGATHER/MSCATTER to an X86ISD node with the extra mask result and stop relying on this hole in the isel checking.
llvm-svn: 318278
Billy Robert O'Neal III [Wed, 15 Nov 2017 07:45:07 +0000 (07:45 +0000)]
Tolerate even more [[nodiscard]] in the STL. Reviewed as https://reviews.llvm.org/D39080
llvm-svn: 318277
Billy Robert O'Neal III [Wed, 15 Nov 2017 07:40:37 +0000 (07:40 +0000)]
Tolerate [[nodiscard]] annotations in the STL. Reviewed as https://reviews.llvm.org/D39033
llvm-svn: 318276
NAKAMURA Takumi [Wed, 15 Nov 2017 07:34:35 +0000 (07:34 +0000)]
GISelWorkList.h: Fix -fmodules build in rL318210.
llvm-svn: 318275
NAKAMURA Takumi [Wed, 15 Nov 2017 06:53:45 +0000 (06:53 +0000)]
ASTMatchers.h: Avoid warnings due to "@throw". [-Wdocumentation]
llvm-svn: 318274
NAKAMURA Takumi [Wed, 15 Nov 2017 06:46:58 +0000 (06:46 +0000)]
Fix llvm/test/Transforms/LoopRotate/pr35210.ll in rL318237, it uses debug options.
llvm-svn: 318273
Fangrui Song [Wed, 15 Nov 2017 06:17:32 +0000 (06:17 +0000)]
NFC Remove default argument of DataLayout::getPointerABIAlignment
Differential Revision: https://reviews.llvm.org/D40005
llvm-svn: 318272
Craig Topper [Wed, 15 Nov 2017 06:02:43 +0000 (06:02 +0000)]
[X86] Add getHostCPUName support for the Gemini Lake model number which also uses Goldmont.
llvm-svn: 318271
Craig Topper [Wed, 15 Nov 2017 06:02:42 +0000 (06:02 +0000)]
[X86] Add getHostCPUName support for cannonlake.
This adds an explicit model number check and fallback path to the unknown family 6 detection.
llvm-svn: 318270
Marshall Clow [Wed, 15 Nov 2017 05:51:26 +0000 (05:51 +0000)]
First part of P0600 - '[[nodiscard] in the standard library'. Mark the 'empty()' methods of all the containers as nodiscard. If you're calling empty() w/o looking at the result, you probably meanto to call 'clear()'. c++2a only
llvm-svn: 318269
Marshall Clow [Wed, 15 Nov 2017 05:25:36 +0000 (05:25 +0000)]
Still more missing tests - this time for the unordered containers
llvm-svn: 318268
Craig Topper [Wed, 15 Nov 2017 05:23:02 +0000 (05:23 +0000)]
[InstCombine] Simplify binops that are only used by a select and are fed by a select with the same condition.
Summary:
This patch optimizes a binop sandwiched between 2 selects with the same condition. Since we know its only used by the select we can propagate the appropriate input value from the earlier select.
As I'm writing this I realize I may need to avoid doing this for division in case the select was protecting a divide by zero?
Reviewers: spatel, majnemer
Reviewed By: majnemer
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D39999
llvm-svn: 318267
Hiroshi Inoue [Wed, 15 Nov 2017 04:23:26 +0000 (04:23 +0000)]
[PowerPC] fix up in redundant compare elimination
This patch fixes a potential problem in my previous commit (https://reviews.llvm.org/rL312514) by introducing an additional check.
llvm-svn: 318266
Jan Vesely [Wed, 15 Nov 2017 04:10:39 +0000 (04:10 +0000)]
math: Implement minmag
Reviewer: Aaron Watry
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
llvm-svn: 318265
Jan Vesely [Wed, 15 Nov 2017 04:10:37 +0000 (04:10 +0000)]
math: Implement maxmag
Reviewer: Aaron Watry
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
llvm-svn: 318264
Rafael Espindola [Wed, 15 Nov 2017 04:00:54 +0000 (04:00 +0000)]
Make the test more strict. NFC.
llvm-svn: 318263
Jason Molenda [Wed, 15 Nov 2017 03:41:47 +0000 (03:41 +0000)]
Roll back r318260 because it is causing the windows bot to
break. The alignas(__uint128_t) is not recognized with MSVC
it looks like. Zachary, is there a similar type on windows?
I suppose I can go with alignas(16) here but I'd prefer to
specify the type alignment that I want & let the ABI dictate
how much padding is required.
llvm-svn: 318262
Dean Michael Berris [Wed, 15 Nov 2017 03:35:42 +0000 (03:35 +0000)]
[XRay][compiler-rt][x86_64] Align the stack before and after calling handlers
Summary:
This change fixes the XRay trampolines aside from the __xray_CustomEvent
trampoline to align the stack to 16-byte boundaries before calling the
handler. Before this change we've not been explicitly aligning the stack
to 16-byte boundaries, which makes it dangerous when calling handlers
that leave the stack in a state that isn't strictly 16-byte aligned
after calling the handlers.
We add a test that makes sure we can handle these cases appropriately
after the changes, and prevents us from regressing the state moving
forward.
Fixes http://llvm.org/PR35294.
Reviewers: pelikan, pcc
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D40004
llvm-svn: 318261
Jason Molenda [Wed, 15 Nov 2017 03:18:24 +0000 (03:18 +0000)]
Two small fixes to handle arm64 fpu register contexts in
a Mach-O file load command correctly, patch by Ryan
Mansfield.
<rdar://problem/
35468499>
llvm-svn: 318260
Marshall Clow [Wed, 15 Nov 2017 03:08:39 +0000 (03:08 +0000)]
Named the macro wrong in the test.
llvm-svn: 318259
Richard Smith [Wed, 15 Nov 2017 03:03:56 +0000 (03:03 +0000)]
PR35214: don't crash if we see an array of unknown bound added to an empty but invalid designator.
llvm-svn: 318258
Vedant Kumar [Wed, 15 Nov 2017 02:58:45 +0000 (02:58 +0000)]
[docs] Document a way to simplify names in bugpoint output
llvm-svn: 318257
Marshall Clow [Wed, 15 Nov 2017 02:31:14 +0000 (02:31 +0000)]
More missing tests - array<>::size() and array<>::max_size()
llvm-svn: 318256
Eric Fiselier [Wed, 15 Nov 2017 01:37:11 +0000 (01:37 +0000)]
Update Appveyor LLVM install link to be current and non-broken
llvm-svn: 318247
Matt Arsenault [Wed, 15 Nov 2017 01:34:06 +0000 (01:34 +0000)]
AMDGPU: Add separate definitions for DS insts without m0 use
llvm-svn: 318246
Richard Smith [Wed, 15 Nov 2017 01:33:46 +0000 (01:33 +0000)]
[modules] Fix crash in complex class merging scenario.
When we merge together class definitions, we can end up with the canonical
declaration of a field not being the one that was lexically within the
canonical definition of the class. Additionally, when we merge class
definitions via update records (eg, for a template specialization whose
declaration is instantiated in one module and whose definition is instantiated
in multiple others), we can end up with the list of lexical contents for the
class not including a particular declaration of a field whose lexical parent is
that class definition. In the worst case, we have a field whose canonical
declaration's lexical parent has no fields, and in that case this attempt to
number the fields by walking the fields in the declaration of the class that
contained one of the canonical fields will fail.
Instead, when numbering fields in a class, do the obvious thing: walk the
fields in the definition.
I'm still trying to reduce a testcase; the setup that leads to the above
scenario seems to be quite fragile.
llvm-svn: 318245
Marshall Clow [Wed, 15 Nov 2017 01:33:33 +0000 (01:33 +0000)]
Added tests for xxx.size() and xxx.empty() for all the sequence containers
llvm-svn: 318244
Craig Topper [Wed, 15 Nov 2017 01:01:50 +0000 (01:01 +0000)]
[X86] Correct the spelling of pentiumpro in X86TargetParser.def
Thanks to Erich Keane for spotting this.
llvm-svn: 318243
Matt Arsenault [Wed, 15 Nov 2017 00:45:43 +0000 (00:45 +0000)]
AMDGPU: Don't use MUBUF vaddr if address may overflow
Effectively revert r263964. Before we would not
allow this if vaddr was not known to be positive.
llvm-svn: 318240
Hans Wennborg [Wed, 15 Nov 2017 00:38:13 +0000 (00:38 +0000)]
Revert r318193 "[SLPVectorizer] Failure to beneficially vectorize 'copyable' elements in integer binary ops."
It crashes building sqlite; see reply on the llvm-commits thread.
> [SLPVectorizer] Failure to beneficially vectorize 'copyable' elements in integer binary ops.
>
> Patch tries to improve vectorization of the following code:
>
> void add1(int * __restrict dst, const int * __restrict src) {
> *dst++ = *src++;
> *dst++ = *src++ + 1;
> *dst++ = *src++ + 2;
> *dst++ = *src++ + 3;
> }
> Allows to vectorize even if the very first operation is not a binary add, but just a load.
>
> Fixed issues related to previous commit.
>
> Reviewers: spatel, mzolotukhin, mkuper, hfinkel, RKSimon, filcab, ABataev
>
> Reviewed By: ABataev, RKSimon
>
> Subscribers: llvm-commits, RKSimon
>
> Differential Revision: https://reviews.llvm.org/D28907
llvm-svn: 318239
Mitch Phillips [Wed, 15 Nov 2017 00:35:26 +0000 (00:35 +0000)]
[cfi-verify] Validate there are no register clobbers between CFI-check and instruction execution.
Summary:
This patch adds another failure mode for `validateCFIProtection(..)`, wherein any register that affects the indirect control flow instruction is clobbered to between the CFI-check and the instruction's execution.
Also includes a modification to make MCInstrDesc::hasDefOfPhysReg public.
Reviewers: vlad.tsyrklevich
Reviewed By: vlad.tsyrklevich
Subscribers: llvm-commits, pcc, kcc
Differential Revision: https://reviews.llvm.org/D39820
llvm-svn: 318238
Craig Topper [Wed, 15 Nov 2017 00:22:42 +0000 (00:22 +0000)]
[LoopRotate] processLoop should return true even if it just simplified the loop latch without making any other changes
Simplifying a loop latch changes the IR and we need to make sure the pass manager knows to invalidate analysis passes if that happened.
PR35210 discovered a case where we failed to invalidate the post dominator tree after this simplification because we no changes other than simplifying the loop latch.
Fixes PR35210.
Differential Revision: https://reviews.llvm.org/D40035
llvm-svn: 318237
Evgeniy Stepanov [Wed, 15 Nov 2017 00:11:51 +0000 (00:11 +0000)]
[asan] Prevent rematerialization of &__asan_shadow.
Summary:
In the mode when ASan shadow base is computed as the address of an
external global (__asan_shadow, currently on android/arm32 only),
regalloc prefers to rematerialize this value to save register spills.
Even in -Os. On arm32 it is rather expensive (2 loads + 1 constant
pool entry).
This changes adds an inline asm in the function prologue to suppress
this behavior. It reduces AsanTest binary size by 7%.
Reviewers: pcc, vitalybuka
Subscribers: aemerson, kristof.beyls, hiraditya, llvm-commits
Differential Revision: https://reviews.llvm.org/D40048
llvm-svn: 318235
Erich Keane [Wed, 15 Nov 2017 00:11:24 +0000 (00:11 +0000)]
Simplify CpuIs code to use include from LLVM
LLVM exposes a file in the backend (X86TargetParser.def) that
contains information about the correct list of CpuIs values.
This patch removes 2 of the copied and pasted versions of this
list from clang and instead includes the data from the .def file.
Differential Revision: https://reviews.llvm.org/D40054
llvm-svn: 318234
Vedant Kumar [Tue, 14 Nov 2017 23:57:58 +0000 (23:57 +0000)]
[profile] Update InstrProfData.inc to sync with llvm
llvm-svn: 318230
Vedant Kumar [Tue, 14 Nov 2017 23:56:53 +0000 (23:56 +0000)]
[PGO] Detect more structural changes with the stable hash
Lifting from Bob Wilson's notes: The hash value that we compute and
store in PGO profile data to detect out-of-date profiles does not
include enough information. This means that many significant changes to
the source will not cause compiler warnings about the profile being out
of date, and worse, we may continue to use the outdated profile data to
make bad optimization decisions. There is some tension here because
some source changes won't affect PGO and we don't want to invalidate the
profile unnecessarily.
This patch adds a new hashing scheme which is more sensitive to loop
nesting, conditions, and out-of-order control flow. Here are examples
which show snippets which get the same hash under the current scheme,
and different hashes under the new scheme:
Loop Nesting Example
--------------------
// Snippet 1
while (foo()) {
while (bar()) {}
}
// Snippet 2
while (foo()) {}
while (bar()) {}
Condition Example
-----------------
// Snippet 1
if (foo())
bar();
baz();
// Snippet 2
if (foo())
bar();
else
baz();
Out-of-order Control Flow Example
---------------------------------
// Snippet 1
while (foo()) {
if (bar()) {}
baz();
}
// Snippet 2
while (foo()) {
if (bar())
continue;
baz();
}
In each of these cases, it's useful to differentiate between the
snippets because swapping their profiles gives bad optimization hints.
The new hashing scheme considers some logical operators in an effort to
detect more changes in conditions. This isn't a perfect scheme. E.g, it
does not produce the same hash for these equivalent snippets:
// Snippet 1
bool c = !a || b;
if (d && e) {}
// Snippet 2
bool f = d && e;
bool c = !a || b;
if (f) {}
This would require an expensive data flow analysis. Short of that, the
new hashing scheme looks reasonably complete, based on a scan over the
statements we place counters on.
Profiles which use the old version of the PGO hash remain valid and can
be used without issue (there are tests in tree which check this).
rdar://
17068282
Differential Revision: https://reviews.llvm.org/D39446
llvm-svn: 318229
Vedant Kumar [Tue, 14 Nov 2017 23:56:48 +0000 (23:56 +0000)]
[PGO] Bump the indexed profile format version
Differential Revision: https://reviews.llvm.org/D39447
llvm-svn: 318228
Petr Hosek [Tue, 14 Nov 2017 23:56:05 +0000 (23:56 +0000)]
[CMake][runtimes] Don't process common options in runtimes build
This is no longer needed for any of the runtimes build and it breaks
in case we don't have the working compiler yet, e.g. when building
a compiler that uses compiler-rt and libc++ as a default runtime,
because these common options check whether these are available.
Differential Revision: https://reviews.llvm.org/D39932
llvm-svn: 318227
Craig Topper [Tue, 14 Nov 2017 23:54:28 +0000 (23:54 +0000)]
[X86] Fix the parameter order in the default implementation of X86_VENDOR macro in X86TargetParser.def
The default implementation doesn't do anything so the order doesn't matter, but good for cleanliness.
llvm-svn: 318226
Petr Hosek [Tue, 14 Nov 2017 23:47:20 +0000 (23:47 +0000)]
[CMake][runtimes] Set compiler as working even for default target
Even when building builtins and runtimes for the default target
we shouldn't assume that the just built compiler is already useable.
When the compiler uses compiler-rt and libc++ as the default runtime
and C++ library, it won't be usable until we finish building runtimes.
Differential Revision: https://reviews.llvm.org/D39715
llvm-svn: 318224
Matt Arsenault [Tue, 14 Nov 2017 23:46:42 +0000 (23:46 +0000)]
AMDGPU: Handle or in multi-use shl ptr combine
llvm-svn: 318223
Eugene Zelenko [Tue, 14 Nov 2017 23:35:42 +0000 (23:35 +0000)]
[AST] Fix some Clang-tidy modernize and Include What You Use warnings; other minor fixes (NFC).
llvm-svn: 318221
Hans Wennborg [Tue, 14 Nov 2017 23:30:28 +0000 (23:30 +0000)]
Fix switch-lower-peel-top-case.ll isel pass is not registered error
The test was doing -stop-after=isel, but that pass is actually the
AMDGPUDAGToDAGISel pass, which might not be built when targeting x86_64.
This changes the test to -stop-after=expand-isel-pseudos instead.
Follow-up to r318202.
llvm-svn: 318220
Jason Molenda [Tue, 14 Nov 2017 23:15:35 +0000 (23:15 +0000)]
Update xcode project file to track ArchSpec.cpp
move and LibCxxBitset.cpp addition.
llvm-svn: 318218
Davide Italiano [Tue, 14 Nov 2017 23:13:38 +0000 (23:13 +0000)]
[EntryExitInstrumenter] Placate GCC, the semicolon is redundant. NFCI.
llvm-svn: 318217
Eugene Zelenko [Tue, 14 Nov 2017 23:13:32 +0000 (23:13 +0000)]
[AST] Fix some Clang-tidy modernize and Include What You Use warnings; other minor fixes (NFC).
llvm-svn: 318216
Alex Lorenz [Tue, 14 Nov 2017 23:10:50 +0000 (23:10 +0000)]
[refactor][selection] canonicalize decl ref callee to the call expr
We would like to extract the full call when just the callee function is
selected
llvm-svn: 318215
Hans Wennborg [Tue, 14 Nov 2017 23:10:04 +0000 (23:10 +0000)]
Try to fix the instrument-functions tests
On e.g. PPC the return value and argument were marked 'signext'. This
makes the test expectations a bit more flexible.
Follow-up to r318199.
llvm-svn: 318214
Tim Renouf [Tue, 14 Nov 2017 23:05:36 +0000 (23:05 +0000)]
[AMDGPU] updated PAL metadata record keys
Summary: The ABI changed before specification was finalized.
Reviewers: kzhuravl, dstuttard
Subscribers: wdng, nhaehnle, yaxunl, t-tye, llvm-commits
Differential Revision: https://reviews.llvm.org/D39807
llvm-svn: 318213
Sanjay Patel [Tue, 14 Nov 2017 23:03:56 +0000 (23:03 +0000)]
[Reassociate] use dyn_cast instead of isa+cast; NFCI
llvm-svn: 318212
Mitch Phillips [Tue, 14 Nov 2017 22:43:13 +0000 (22:43 +0000)]
[cfi-verify] Add DOT graph printing for GraphResult objects.
Allows users to view GraphResult objects in a DOT directed-graph format. This feature can be turned on through the --print-graphs flag.
Also enabled pretty-printing of instructions in output. Together these features make analysis of unprotected CF instructions much easier by providing a visual control flow graph.
Reviewers: pcc
Subscribers: llvm-commits, kcc, vlad.tsyrklevich
Differential Revision: https://reviews.llvm.org/D39819
llvm-svn: 318211
Aditya Nandakumar [Tue, 14 Nov 2017 22:42:19 +0000 (22:42 +0000)]
[GISel]: Rework legalization algorithm for better elimination of
artifacts along with DCE
Legalization Artifacts are all those insts that are there to make the
type system happy. Currently, the target needs to say all combinations
of extends and truncs are legal and there's no way of verifying that
post legalization, we only have *truly* legal instructions. This patch
changes roughly the legalization algorithm to process all illegal insts
at one go, and then process all truncs/extends that were added to
satisfy the type constraints separately trying to combine trivial cases
until they converge. This has the added benefit that, the target
legalizerinfo can only say which truncs and extends are okay and the
artifact combiner would combine away other exts and truncs.
Updated legalization algorithm to roughly the following pseudo code.
WorkList Insts, Artifacts;
collect_all_insts_and_artifacts(Insts, Artifacts);
do {
for (Inst in Insts)
legalizeInstrStep(Inst, Insts, Artifacts);
for (Artifact in Artifacts)
tryCombineArtifact(Artifact, Insts, Artifacts);
} while(!Insts.empty());
Also, wrote a simple wrapper equivalent to SetVector, except for
erasing, it avoids moving all elements over by one and instead just
nulls them out.
llvm-svn: 318210
Hans Wennborg [Tue, 14 Nov 2017 22:32:49 +0000 (22:32 +0000)]
CMake: Turn LLVM_ENABLE_LIBXML2 into a tri-state option
In addition to the current ON and OFF options, this adds the FORCE_ON
option, which causes a configuration error if libxml2 cannot be used.
Differential revision: https://reviews.llvm.org/D40050
llvm-svn: 318209
Marshall Clow [Tue, 14 Nov 2017 22:26:50 +0000 (22:26 +0000)]
Add two new macros: _LIBCPP_NODISCARD_AFTER_CXX17 and _LIBCPP_CONSTEXPR_AFTER_CXX17, along with a way to turn off the NODISCARD one: _LIBCPP_DISABLE_NODISCARD_AFTER_CXX17. No one is using these yet, but we will be ... soon
llvm-svn: 318208
Simon Dardis [Tue, 14 Nov 2017 22:26:42 +0000 (22:26 +0000)]
Reland "[mips][mt][6/7] Add support for mftr, mttr instructions."
This adjusts the tests to hopfully pacify the
llvm-clang-x86_64-expensive-checks-win buildbot.
Unlike many other instructions, these instructions have aliases which
take coprocessor registers, gpr register, accumulator (and dsp accumulator)
registers, floating point registers, floating point control registers and
coprocessor 2 data and control operands.
For the moment, these aliases are treated as pseudo instructions which are
expanded into the underlying instruction. As a result, disassembling these
instructions shows the underlying instruction and not the alias.
Reviewers: slthakur, atanasyan
Differential Revision: https://reviews.llvm.org/D35253
llvm-svn: 318207
Rong Xu [Tue, 14 Nov 2017 22:08:37 +0000 (22:08 +0000)]
[CodeGen] Fix the test case added in r318202
Add the -mtriple option to filter some platforms.
llvm-svn: 318206
Alex Lorenz [Tue, 14 Nov 2017 22:06:55 +0000 (22:06 +0000)]
[refactor][selection] canonicalize member expr callee to the full
member call expression
We would like to extract the full call when just the callee is selected.
llvm-svn: 318205
Jan Vesely [Tue, 14 Nov 2017 21:55:41 +0000 (21:55 +0000)]
native_powr: Switch implementation to native_exp2 and native_log2
v2: don't use assume
check only for x<0, the other conditions are handled transparently
v3: don't check inputs at all, nan propagation works as expected
Reviewer: Jeroen Ketema
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
llvm-svn: 318204
Reid Kleckner [Tue, 14 Nov 2017 21:49:06 +0000 (21:49 +0000)]
Make salvageDebugInfo of casts work for dbg.declare and dbg.addr
Summary:
Instcombine (and probably other passes) sometimes want to change the
type of an alloca. To do this, they generally create a new alloca with
the desired type, create a bitcast to make the new pointer type match
the old pointer type, replace all uses with the cast, and then simplify
the casts. We already knew how to salvage dbg.value instructions when
removing casts, but we can extend it to cover dbg.addr and dbg.declare.
Fixes a debug info quality issue uncovered in Chromium in
http://crbug.com/784609
Reviewers: aprantl, vsk
Subscribers: hiraditya, llvm-commits
Differential Revision: https://reviews.llvm.org/D40042
llvm-svn: 318203
Rong Xu [Tue, 14 Nov 2017 21:44:09 +0000 (21:44 +0000)]
[CodeGen] Peel off the dominant case in switch statement in lowering
This patch peels off the top case in switch statement into a branch if the
probability exceeds a threshold. This will help the branch prediction and
avoids the extra compares when lowering into chain of branches.
Differential Revision: http://reviews.llvm.org/D39262
llvm-svn: 318202
Richard Smith [Tue, 14 Nov 2017 21:26:46 +0000 (21:26 +0000)]
Fix unused variable warning.
llvm-svn: 318201
Hans Wennborg [Tue, 14 Nov 2017 21:13:27 +0000 (21:13 +0000)]
Switch -mcount and -finstrument-functions to emit EnterExitInstrumenter attributes
This updates -mcount to use the new attribute names (LLVM r318195), and
switches over -finstrument-functions to also use these attributes rather
than inserting instrumentation in the frontend.
It also adds a new flag, -finstrument-functions-after-inlining, which
makes the cygprofile instrumentation get inserted after inlining rather
than before.
Differential Revision: https://reviews.llvm.org/D39331
llvm-svn: 318199
Hans Wennborg [Tue, 14 Nov 2017 21:09:45 +0000 (21:09 +0000)]
Rename CountingFunctionInserter and use for both mcount and cygprofile calls, before and after inlining
Clang implements the -finstrument-functions flag inherited from GCC, which
inserts calls to __cyg_profile_func_{enter,exit} on function entry and exit.
This is useful for getting a trace of how the functions in a program are
executed. Normally, the calls remain even if a function is inlined into another
function, but it is useful to be able to turn this off for users who are
interested in a lower-level trace, i.e. one that reflects what functions are
called post-inlining. (We use this to generate link order files for Chromium.)
LLVM already has a pass for inserting similar instrumentation calls to
mcount(), which it does after inlining. This patch renames and extends that
pass to handle calls both to mcount and the cygprofile functions, before and/or
after inlining as controlled by function attributes.
Differential Revision: https://reviews.llvm.org/D39287
llvm-svn: 318195
Alexey Bataev [Tue, 14 Nov 2017 21:01:01 +0000 (21:01 +0000)]
[OPENMP] Fix DSA analysis for threadprivates after deserialization.
If threadprivate vaible is deserialized, it is not marked as
threadprivate in DSAStack.
llvm-svn: 318194
Dinar Temirbulatov [Tue, 14 Nov 2017 20:55:08 +0000 (20:55 +0000)]
[SLPVectorizer] Failure to beneficially vectorize 'copyable' elements in integer binary ops.
Patch tries to improve vectorization of the following code:
void add1(int * __restrict dst, const int * __restrict src) {
*dst++ = *src++;
*dst++ = *src++ + 1;
*dst++ = *src++ + 2;
*dst++ = *src++ + 3;
}
Allows to vectorize even if the very first operation is not a binary add, but just a load.
Fixed issues related to previous commit.
Reviewers: spatel, mzolotukhin, mkuper, hfinkel, RKSimon, filcab, ABataev
Reviewed By: ABataev, RKSimon
Subscribers: llvm-commits, RKSimon
Differential Revision: https://reviews.llvm.org/D28907
llvm-svn: 318193
Jake Ehrlich [Tue, 14 Nov 2017 20:36:04 +0000 (20:36 +0000)]
[llvm-objcopy] Improve command line option help messages
I was being inconsistent with the way I was capitalizing help messages
for command line options. Additionally --remove-section wasn't using
value_desc even though it benefited from it.
Differential Revision: https://reviews.llvm.org/D39978
llvm-svn: 318190
Matt Arsenault [Tue, 14 Nov 2017 20:33:14 +0000 (20:33 +0000)]
AMDGPU: Error on stack size overflow
llvm-svn: 318189
Ulrich Weigand [Tue, 14 Nov 2017 20:00:34 +0000 (20:00 +0000)]
[SystemZ] Do not crash when selecting an OR of two constants
In rare cases, common code will attempt to select an OR of two
constants. This confuses the logic in splitLargeImmediate,
causing an internal error during isel. Fixed by simply leaving
this case to common code to handle.
This fixes PR34859.
llvm-svn: 318187
Evandro Menezes [Tue, 14 Nov 2017 19:59:43 +0000 (19:59 +0000)]
[AArch64] Adjust the cost model for Exynos M1 and M2
Fix the modeling of loads and stores of registers pairs.
llvm-svn: 318186
Martin Storsjo [Tue, 14 Nov 2017 19:58:36 +0000 (19:58 +0000)]
[llvm-strings] Add support for the -a/--all options
They don't actually change nay behaviour, as llvm-strings currently
checks the whole object without looking at individual sections anyway.
This allows using llvm-strings in a context that explicitly passes
the -a option.
Differential Revision: https://reviews.llvm.org/D40020
llvm-svn: 318185
Martin Storsjo [Tue, 14 Nov 2017 19:57:59 +0000 (19:57 +0000)]
[ARM, AArch64] Fix an assert message, Darwin isn't the only target supporting TLS. NFC.
llvm-svn: 318184
Hiroshi Yamauchi [Tue, 14 Nov 2017 19:48:59 +0000 (19:48 +0000)]
Simplify irreducible loop metadata test code.
Summary:
Shorten the irreducible loop metadata test code by removing insignificant
instructions.
Reviewers: davidxl
Reviewed By: davidxl
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D40043
llvm-svn: 318182
Easwaran Raman [Tue, 14 Nov 2017 19:31:51 +0000 (19:31 +0000)]
[CodeGenPrepare] Disable div bypass when working set size is huge.
Summary:
Bypass of slow divs based on operand values is currently disabled for
-Os. Do the same when profile summary is available and the working set
size of the application is huge. This is similar to how loop peeling is
guarded by hasHugeWorkingSetSize. In the div bypass case, the generated
extra code (and the extra branch) tendss to outweigh the benefits of the
bypass. This results in noticeable performance improvement on an
internal application.
Reviewers: davidxl
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D39992
llvm-svn: 318179
Ulrich Weigand [Tue, 14 Nov 2017 19:20:46 +0000 (19:20 +0000)]
[SystemZ] Fix invalid codegen using RISBMux on out-of-range bits
Before using the 32-bit RISBMux set of instructions we need to
verify that the input bits are actually within range of the 32-bit
instruction. This fixer PR35289.
llvm-svn: 318177
Alex Bradbury [Tue, 14 Nov 2017 19:16:08 +0000 (19:16 +0000)]
Set hasSideEffects=0 for TargetOpcode::{CFI_INSTRUCTION,EH_LABEL,GC_LABEL,ANNOTATION_LABEL}
D37065 (committed as rL317674) explicitly set hasSideEffects for all
TargetOpcode::* instructions where it was inferred previously. This is a
follow-up to that patch, setting hasSideEffects=0 for CFI_INSTRUCTION,
EH_LABEL, GC_LABEL and ANNOTATION_LABEL. All LLVM tests pass after this
change.
This patch also modifies MachineInstr::isLabel returns true for a
TargetOpcode::ANNOTATION_LABEL, which ensures that an annotation label won't
be incorrectly considered safe to move.
Differential Revision: https://reviews.llvm.org/D39941
llvm-svn: 318174
Artem Belevich [Tue, 14 Nov 2017 19:14:00 +0000 (19:14 +0000)]
Mark intrinsics operating on the whole warp as IntrInaccessibleMemOnly
It's needed to model the fact that they do access data from other threads in a
warp and thus can't be CSE'd.
llvm-svn: 318173
Simon Dardis [Tue, 14 Nov 2017 19:11:45 +0000 (19:11 +0000)]
[mips] Simplify test for 5.0.1 (NFC)
Simplify testing that an emergency spill slot is used when MSA
is used so that it can be included in the 5.0.1 release.
llvm-svn: 318172
Adam Nemet [Tue, 14 Nov 2017 19:00:08 +0000 (19:00 +0000)]
Adjust test after r318159
llvm-svn: 318170
Alex Lorenz [Tue, 14 Nov 2017 18:59:01 +0000 (18:59 +0000)]
[refactor][extract] avoid extracting expressions from types in functions
llvm-svn: 318169
Jake Ehrlich [Tue, 14 Nov 2017 18:50:24 +0000 (18:50 +0000)]
[llvm-objcopy] Add -strip-non-alloc option to remove all non-allocated sections
This change adds a new flag not present in GNU objcopy that we call
--strip-non-alloc.
Differential Revision: https://reviews.llvm.org/D39926
llvm-svn: 318168
Yaxun Liu [Tue, 14 Nov 2017 18:46:52 +0000 (18:46 +0000)]
CodeGen: Fix TargetLowering::LowerCallTo for sret value type
TargetLowering::LowerCallTo assumes that sret value type corresponds to a
pointer in default address space, which is incorrect, since sret value type
should correspond to a pointer in alloca address space, which may not
be the default address space. This causes assertion for amdgcn target
in amdgiz environment.
This patch fixes that.
Differential Revision: https://reviews.llvm.org/D39996
llvm-svn: 318167
Jake Ehrlich [Tue, 14 Nov 2017 18:41:47 +0000 (18:41 +0000)]
[llvm-objcopy] Support the rest of the ELF formats
We haven't been supporting anything but ELF64LE since the start. Luckily
this was always accounted for and the change is pretty trivial. B35281
requests this change for ELF32LE. This change adds support for ELF32LE,
ELF64BE, and ELF32BE with all supported features that already existed
for ELF64LE.
Differential Revision: https://reviews.llvm.org/D39977
llvm-svn: 318166