Rui Ueyama [Fri, 17 Mar 2017 23:28:41 +0000 (23:28 +0000)]
Define Config::isLE and Config::wordsize.
isLE() return true if the target is little-endian.
wordsize() returns 8 for 64-bit and 4 for 32-bit.
llvm-svn: 298167
Sanjay Patel [Fri, 17 Mar 2017 23:04:18 +0000 (23:04 +0000)]
[x86] regenerate checks; NFC
llvm-svn: 298166
Duncan P. N. Exon Smith [Fri, 17 Mar 2017 22:55:13 +0000 (22:55 +0000)]
Modules: Cache PCMs in memory and avoid a use-after-free
Clang's internal build system for implicit modules uses lock files to
ensure that after a process writes a PCM it will read the same one back
in (without contention from other -cc1 commands). Since PCMs are read
from disk repeatedly while invalidating, building, and importing, the
lock is not released quickly. Furthermore, the LockFileManager is not
robust in every environment. Other -cc1 commands can stall until
timeout (after about eight minutes).
This commit changes the lock file from being necessary for correctness
to a (possibly dubious) performance hack. The remaining benefit is to
reduce duplicate work in competing -cc1 commands which depend on the
same module. Follow-up commits will change the internal build system to
continue after a timeout, and reduce the timeout. Perhaps we should
reconsider blocking at all.
This also fixes a use-after-free, when one part of a compilation
validates a PCM and starts using it, and another tries to swap out the
PCM for something new.
The PCMCache is a new type called MemoryBufferCache, which saves memory
buffers based on their filename. Its ownership is shared by the
CompilerInstance and ModuleManager.
- The ModuleManager stores PCMs there that it loads from disk, never
touching the disk if the cache is hot.
- When modules fail to validate, they're removed from the cache.
- When a CompilerInstance is spawned to build a new module, each
already-loaded PCM is assumed to be valid, and is frozen to avoid
the use-after-free.
- Any newly-built module is written directly to the cache to avoid the
round-trip to the filesystem, making lock files unnecessary for
correctness.
Original patch by Manman Ren; most testcases by Adrian Prantl!
llvm-svn: 298165
Sanjay Patel [Fri, 17 Mar 2017 22:47:21 +0000 (22:47 +0000)]
[x86] regenerate checks; NFC
llvm-svn: 298164
Evgeniy Stepanov [Fri, 17 Mar 2017 22:31:13 +0000 (22:31 +0000)]
Fix docs-llvm-html build.
llvm-svn: 298163
Jessica Paquette [Fri, 17 Mar 2017 22:26:55 +0000 (22:26 +0000)]
[Outliner] Add outliner for AArch64
This commit adds the necessary target hooks for outlining in AArch64. It also
refactors the switch statement used in `getMemOpBaseRegImmOfsWidth` into a
more general function, `getMemOpInfo`. This allows the outliner to share that
code without copying and pasting it.
The AArch64 outliner can be run using -mllvm -enable-machine-outliner, as with
the X86-64 outliner.
The test for this pass verifies that the outliner does, in fact outline
functions, fixes up the stack accesses properly, and can correctly generate a
tail call. In the future, this test should be replaced with a MIR test, so that
we can properly test immediate offset overflows in fixed-up instructions.
llvm-svn: 298162
Eli Friedman [Fri, 17 Mar 2017 22:19:52 +0000 (22:19 +0000)]
[SCEV] Use const Loop *L instead of Loop *L. NFC
Use const pointer in the trip count and trip multiple calculations.
Patch by Huihui Zhang <huihuiz@codeaurora.org>
llvm-svn: 298161
Davide Italiano [Fri, 17 Mar 2017 22:19:20 +0000 (22:19 +0000)]
[Sema] Unbreak GCC -Werror build (enum compare).
llvm-svn: 298160
Evgeniy Stepanov [Fri, 17 Mar 2017 22:17:38 +0000 (22:17 +0000)]
[asan] Fix dead stripping of globals on Linux (compiler-rt)
Runtime support for the new instrumentation of globals based on !associated, and a bunch of tests.
Differential Revision: https://reviews.llvm.org/D30120
llvm-svn: 298159
Evgeniy Stepanov [Fri, 17 Mar 2017 22:17:29 +0000 (22:17 +0000)]
[asan] Fix dead stripping of globals on Linux.
Use a combination of !associated, comdat, @llvm.compiler.used and
custom sections to allow dead stripping of globals and their asan
metadata. Sometimes.
Currently this works on LLD, which supports SHF_LINK_ORDER with
sh_link pointing to the associated section.
This also works on BFD, which seems to treat comdats as
all-or-nothing with respect to linker GC. There is a weird quirk
where the "first" global in each link is never GC-ed because of the
section symbols.
At this moment it does not work on Gold (as in the globals are never
stripped).
Differential Revision: https://reviews.llvm.org/D30121
llvm-svn: 298158
Evgeniy Stepanov [Fri, 17 Mar 2017 22:17:24 +0000 (22:17 +0000)]
Add !associated metadata.
This is an ELF-specific thing that adds SHF_LINK_ORDER to the global's section
pointing to the metadata argument's section. The effect of that is a reverse dependency
between sections for the linker GC.
!associated does not change the behavior of global-dce. The global
may also need to be added to llvm.compiler.used.
Since SHF_LINK_ORDER is per-section, !associated effectively enables
fdata-sections for the affected globals, the same as comdats do.
Differential Revision: https://reviews.llvm.org/D29104
llvm-svn: 298157
Eli Friedman [Fri, 17 Mar 2017 22:15:50 +0000 (22:15 +0000)]
[SelectionDAG] Remove redundant stores more aggressively.
Handle TokenFactors more aggressively in
SDValue::reachesChainWithoutSideEffects. This isn't really a
very effective change anymore because of other changes to
chain handling, but it's a cheap check, and the expanded
comments are still useful.
It might be possible to loosen the hasOneUse() requirement with a
deeper analysis, but a naive implementation of that check would be
expensive.
Differential Revision: https://reviews.llvm.org/D29845
llvm-svn: 298156
Peter Collingbourne [Fri, 17 Mar 2017 22:06:36 +0000 (22:06 +0000)]
ELF: Change check(Expected<T>, const Twine &) to call toString instead of converting to an error code.
llvm-svn: 298155
Evgeniy Stepanov [Fri, 17 Mar 2017 22:04:52 +0000 (22:04 +0000)]
[ELF] Restore GC handling of LINK_ORDER, C-named sections.
__start_xxx symbol keeps section xxx alive only if it is not
SHF_LINK_ORDER. Such sections can be used for user metadata, when
__start_xxx is used to iterate over section contents at runtime, and
the liveness is determined solely by the linked (associated) section.
This was earlier implemented in r294592, and broken in r296723.
Differential Revision: https://reviews.llvm.org/D30964
llvm-svn: 298154
Peter Collingbourne [Fri, 17 Mar 2017 21:49:09 +0000 (21:49 +0000)]
LTO: Work around libstdc++ version mismatch bug, see D31063 review thread.
llvm-svn: 298127
Matthias Gehre [Fri, 17 Mar 2017 21:41:20 +0000 (21:41 +0000)]
Implement DR 373 "Lookup on namespace qualified name in using-directive"
Summary:
3.4.6 [basic.lookup.udir] paragraph 1:
In a using-directive or namespace-alias-definition, during the lookup for a namespace-name or for a name in a nested-name-specifier, only namespace names are considered.
Reviewers: rsmith, aaron.ballman
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D30848
llvm-svn: 298126
Matt Arsenault [Fri, 17 Mar 2017 21:36:28 +0000 (21:36 +0000)]
AMDGPU: Fix broken condition in hazard recognizer
Fixes bug 32248.
llvm-svn: 298125
Bob Haarman [Fri, 17 Mar 2017 21:32:49 +0000 (21:32 +0000)]
recommend using llvm-ar when finding undefined references and empty archives
Summary:
When we perform LTO builds with a version of ar that does not
understand LLVM bitcode objects, we end up with undefined references,
because our archive files do not list the bitcode symbols in their
indices. The error messages do not make it clear what the real problem
is. This change adds a note that points out the likely problem and
solution. It is similar in spirit to r282633, but aims to avoid false
positives by only triggering when we see both undefined references and
archives without symbols in their indices.
Fixes PR32281.
Reviewers: davide, ruiu, tejohnson
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D31011
llvm-svn: 298124
Tim Hammerquist [Fri, 17 Mar 2017 21:00:35 +0000 (21:00 +0000)]
allow for specification of compiler/lldb executables basename
llvm-svn: 298123
Vassil Vassilev [Fri, 17 Mar 2017 20:58:08 +0000 (20:58 +0000)]
[coverity] Fix uninit variable.
Patch by John Harvey!
llvm-svn: 298122
Matt Arsenault [Fri, 17 Mar 2017 20:52:21 +0000 (20:52 +0000)]
AMDGPU: Fix handling of constant phi input loop conditions
If the loop condition was an i1 phi with a constantexpr input, this
would add a loop intrinsic fed by a phi dependent on a call to
if.break in the same block. Insert the call in the loop header.
llvm-svn: 298121
Rong Xu [Fri, 17 Mar 2017 20:51:44 +0000 (20:51 +0000)]
[PGO] Change the internal options description. nfc.
llvm-svn: 298120
Matt Arsenault [Fri, 17 Mar 2017 20:41:45 +0000 (20:41 +0000)]
AMDGPU: Cleanup control flow intrinsics
Move backend internal intrinsics along with the rest of the
normal intrinsics, and use the Intrinsic::getDeclaration
API instead of manually constructing the type list.
It's surprising this was working before. fdiv.fast had
the wrong number of parameters. The control flow intrinsic
declaration attributes were not being applied, and
their types were inconsistent. The actual IR use types
did not match the declaration, and were closer to the
types used for the patterns. The brcond lowering
was changing the types, so introduce new nodes for those.
llvm-svn: 298119
Sanjay Patel [Fri, 17 Mar 2017 20:29:40 +0000 (20:29 +0000)]
[x86] clean up setcc with negated operand transform and add missing test; NFCI
llvm-svn: 298118
Vedant Kumar [Fri, 17 Mar 2017 20:27:33 +0000 (20:27 +0000)]
[ubsan] Add e2e test for -fsanitize=nullability
llvm-svn: 298117
Reid Kleckner [Fri, 17 Mar 2017 20:25:49 +0000 (20:25 +0000)]
[X86] Emit fewer instructions to allocate >16GB stack frames
Summary:
Use this code pattern when RAX is live, instead of emitting up to 2
billion adjustments:
pushq %rax
movabsq +-$Offset+-8, %rax
addq %rsp, %rax
xchg %rax, (%rsp)
movq (%rsp), %rsp
Try to clean this code up a bit while I'm here. In particular, hoist the
logic that handles the entire adjustment with `movabsq $imm, %rax` out
of the loop.
This negates the offset in the prologue and uses ADD because X86 only
has a two operand subtract which always subtracts from the destination
register, which can no longer be RSP.
Fixes PR31962
Reviewers: majnemer, sdardis
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D30052
llvm-svn: 298116
Rong Xu [Fri, 17 Mar 2017 20:05:13 +0000 (20:05 +0000)]
[PGO] Add omitted test cases.
llvm-svn: 298115
Jun Bum Lim [Fri, 17 Mar 2017 19:05:21 +0000 (19:05 +0000)]
[CodeGenPrep]Restructure promoting Ext to form ExtLoad
Summary:
Instead of just looking for a load which is mergable with Ext to form ExtLoad, trying to promote Exts as long as the cost is acceptable. This change is not a NFC as it continue promoting Exts even after finding a load during promotions; the change in arm64-codegen-prepare-extload.ll described in 2.b might show the case.
This change was motivated from D26524. Based on this change, I will move the transformation performed in aarch64-type-promotion into CGP.
Reviewers: jmolloy, qcolombet, mcrosier, javed.absar
Reviewed By: qcolombet
Subscribers: rengolin, llvm-commits, aemerson
Differential Revision: https://reviews.llvm.org/D27853
llvm-svn: 298114
Rong Xu [Fri, 17 Mar 2017 18:41:33 +0000 (18:41 +0000)]
Resubmit r295469 [PGO] Suspend SIGKILL for PR_SET_PDEATHSIG in profile-write
And also r295364 [PGO] remove unintended debug trace. NFC
I removed the test case: it's hard to write synchronized test b/w processes
in this framework. I will revisit the test-case later.
llvm-svn: 298113
Tim Hammerquist [Fri, 17 Mar 2017 18:10:58 +0000 (18:10 +0000)]
executables should be validated before spawning subprocesses
dotest.py script doesn't validate executables passed on the command line
before spawning dozens of subprocesses, all of which fail silently,
leaving an empty results file.
We should validate the lldb and compiler executables on
configuration, aborting when given invalid paths, to prevent numerous,
cryptic, and spurious failures.
<rdar://problem/
31117272>
llvm-svn: 298111
Rong Xu [Fri, 17 Mar 2017 18:07:26 +0000 (18:07 +0000)]
[PGO] Value profile for size of memory intrinsic calls
This patch annotates the valuesites profile to memory intrinsics.
Differential Revision: http://reviews.llvm.org/D31002
llvm-svn: 298110
Vedant Kumar [Fri, 17 Mar 2017 17:53:26 +0000 (17:53 +0000)]
[Bitcode] Add compatibility test for the 4.0 release
Fork off compatibility.ll for the 4.0 release. The *.bc file in this
commit was produced using a Release build of the release_40 branch.
llvm-svn: 298109
Simon Pilgrim [Fri, 17 Mar 2017 17:45:36 +0000 (17:45 +0000)]
[SelectionDAG] Add SelectionDAG.computeKnownBits test support for ISD::ABS
llvm-svn: 298108
Sanjay Patel [Fri, 17 Mar 2017 17:27:31 +0000 (17:27 +0000)]
[x86] avoid adc/sbb assert when both sides of add are zexted (PR32316)
As noted in the comment, we might want to account for this case,
but I didn't look at what that would mean for the asm.
I'm also not sure why this only reproduces with avx512, but I'm
putting a conservative fix in for now to avoid the crash.
Also, if both sides of an add are zexted, shouldn't we shrink that add?
https://bugs.llvm.org/show_bug.cgi?id=32316
llvm-svn: 298107
Reid Kleckner [Fri, 17 Mar 2017 17:24:03 +0000 (17:24 +0000)]
Fix wasm build after arg_begin iterator type change
llvm-svn: 298106
Reid Kleckner [Fri, 17 Mar 2017 17:16:39 +0000 (17:16 +0000)]
Store Arguments in a flat array instead of an iplist
This saves two pointers from Argument and eliminates some extra
allocations.
Arguments cannot be inserted or removed from a Function because that
would require changing its Type, which LLVM does not allow. Instead,
passes that change prototypes, like DeadArgElim, create a new Function
and copy over argument names and attributes. The primary benefit of
iplist is O(1) random insertion and removal. We just don't need that for
arguments, so don't use it.
Reviewed By: chandlerc
Subscribers: dlj, inglorion, llvm-commits
Differential Revision: https://reviews.llvm.org/D31058
llvm-svn: 298105
Stanislav Mekhanoshin [Fri, 17 Mar 2017 17:13:41 +0000 (17:13 +0000)]
Only unswitch loops with uniform conditions
Loop unswitching can be extremely harmful for a SIMT target. In case
if hoisted condition is not uniform a SIMT machine will execute both
clones of a loop sequentially. Therefor LoopUnswitch checks if the
condition is non-divergent.
Since DivergenceAnalysis adds an expensive PostDominatorTree analysis
not needed for non-SIMT targets a new option is added to avoid unneded
analysis initialization. The method getAnalysisUsage is called when
TargetTransformInfo is not yet available and we cannot use it here.
For that reason a new field DivergentTarget is added to PassManagerBuilder
to control the behavior and set this field from a target.
Differential Revision: https://reviews.llvm.org/D30796
llvm-svn: 298104
Simon Pilgrim [Fri, 17 Mar 2017 16:58:15 +0000 (16:58 +0000)]
[X86] Add SelectionDAG.computeKnownBits test showing inability to handle ISD::ABS
We have to be careful as abs(INT_MIN) == INT_MIN.
llvm-svn: 298103
George Rimar [Fri, 17 Mar 2017 16:50:07 +0000 (16:50 +0000)]
[ELF] - Move template instantiations to the end of file. NFC.
To be consistent with other code, addresses post
commit review comments.
llvm-svn: 298102
Alexander Kornienko [Fri, 17 Mar 2017 16:40:34 +0000 (16:40 +0000)]
[Clang-tidy] Fix for misc-noexcept-move-constructor false triggers on defaulted declarations
Summary:
There is no need for triggering warning when noexcept specifier in move constructor or move-assignment operator is neither evaluated nor uninstantiated.
This fixes bug reported here: bugs.llvm.org/show_bug.cgi?id=24712
Reviewers: alexfh
Reviewed By: alexfh
Subscribers: JonasToth, JDevlieghere, cfe-commits
Tags: #clang-tools-extra
Patch by Marek Jenda!
Differential Revision: https://reviews.llvm.org/D31049
llvm-svn: 298101
Zachary Turner [Fri, 17 Mar 2017 16:33:37 +0000 (16:33 +0000)]
CMake requires normalized paths when appending.
Patch by Hugh Bellamy
Differential Revision: https://reviews.llvm.org/D30927
llvm-svn: 298100
Zachary Turner [Fri, 17 Mar 2017 16:32:43 +0000 (16:32 +0000)]
Fix some signed/unsigned comparison warnings.
Patch by Hugh Bellamy
Differential Revision: https://reviews.llvm.org/D30926
llvm-svn: 298099
Zachary Turner [Fri, 17 Mar 2017 16:24:34 +0000 (16:24 +0000)]
[clang-cl] Fix cross-compilation with MSVC 2017.
clang-cl works best when the user runs vcvarsall to set up
an environment before running, but even this is not enough
on VC 2017 when cross compiling (e.g. using an x64 toolchain
to target x86, or vice versa).
The reason is that although clang-cl itself will have a
valid environment, it will shell out to other tools (such
as link.exe) which may not. Generally we solve this through
adding the appropriate linker flags, but this is not enough
in VC 2017.
The cross-linker and the regular linker both link against
some common DLLs, but these DLLs live in the binary directory
of the native linker. When setting up a cross-compilation
environment through vcvarsall, it will add *both* directories
to %PATH%, so that when cl shells out to any of the associated
tools, those tools will be able to find all of the dependencies
that it links against. If you don't do this, link.exe will
fail to run because the loader won't be able to find all of
the required DLLs that it links against.
To solve this we teach the driver how to spawn a process with
an explicitly specified environment. Then we modify the
PATH before shelling out to subtools and run with the modified
PATH.
Patch by Hamza Sood
Differential Revision: https://reviews.llvm.org/D30991
llvm-svn: 298098
David Green [Fri, 17 Mar 2017 15:38:49 +0000 (15:38 +0000)]
Test commit.
llvm-svn: 298097
Rafael Espindola [Fri, 17 Mar 2017 14:55:36 +0000 (14:55 +0000)]
Handle & and | of non abs values.
Handling & in particular is probably important because of its use in
aligning addresses.
llvm-svn: 298096
Siddharth Bhat [Fri, 17 Mar 2017 14:52:19 +0000 (14:52 +0000)]
[ScheduleOptimiser] fix typos in top comment [NFC]
coice -> choice
Transations -> Transactions
llvm-svn: 298095
Rafael Espindola [Fri, 17 Mar 2017 14:51:07 +0000 (14:51 +0000)]
Refuse to add two non absolute symbols.
Since there is no way to produce the correct answer at runtime, it is
probably better to just err.
llvm-svn: 298094
Simon Atanasyan [Fri, 17 Mar 2017 14:27:55 +0000 (14:27 +0000)]
[ELF][MIPS] Always collect .reginfo, .MIPS.options, and .MIPS.abiflags input sections
Do not take in account the `Live` flag while collecting .reginfo, .MIPS.options,
and .MIPS.abiflags input sections to produce corresponding output sections.
These sections have information purpose and should be always produced per
ABI requirements.
llvm-svn: 298093
Chad Rosier [Fri, 17 Mar 2017 14:19:55 +0000 (14:19 +0000)]
[AArch64] Use alias analysis in the load/store optimization pass.
This allows the optimization to rearrange loads and stores more aggressively.
Differential Revision: http://reviews.llvm.org/D30903
llvm-svn: 298092
George Rimar [Fri, 17 Mar 2017 14:12:51 +0000 (14:12 +0000)]
[ELF] - Detemplate SymbolBody::getGotOffset(). NFC.
llvm-svn: 298091
Eric Liu [Fri, 17 Mar 2017 14:05:39 +0000 (14:05 +0000)]
[change-namespace] do not rename specialized template parameters.
Reviewers: hokein
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D31076
llvm-svn: 298090
Michael Kruse [Fri, 17 Mar 2017 13:56:53 +0000 (13:56 +0000)]
Revert "Remove references to AssumptionCache. NFC."
The AssumptionCache removal of r289756 has been reverted in
r290086/r290087. A different solution has been implemented in r291671
which keeps the AssumptionCache. We can therefore use it again in Polly.
This reverts r289791.
llvm-svn: 298089
Rafael Espindola [Fri, 17 Mar 2017 13:45:36 +0000 (13:45 +0000)]
Support non abs values in the rhs of +.
llvm-svn: 298088
George Rimar [Fri, 17 Mar 2017 13:43:24 +0000 (13:43 +0000)]
[ELF] - Detemplate findSection() global static method. NFC.
llvm-svn: 298087
George Rimar [Fri, 17 Mar 2017 13:31:07 +0000 (13:31 +0000)]
[ELF] - Recommit r298078 "[ELF] - Simplify logic of creating "COMMON" section."
With fix of next warning:
Writer.cpp:361:3: warning: suggest parentheses around assignment used as truth value [-Wparentheses]
Original commit message:
Patch reuses BssSection section to simplify creation of
COMMON section.
Differential revision: https://reviews.llvm.org/D30690
llvm-svn: 298086
Siddharth Bhat [Fri, 17 Mar 2017 13:26:10 +0000 (13:26 +0000)]
[DependenceInfo] Remove idempotent union: must-writes with may-writes [NFC]
Since may-writes are always a superset of the must-writes, there is no
point in taking a union of one with the other.
llvm-svn: 298085
George Rimar [Fri, 17 Mar 2017 13:21:22 +0000 (13:21 +0000)]
[ELF] - Revert r298078 and 298082
It segfaults.
llvm-svn: 298084
Rafael Espindola [Fri, 17 Mar 2017 13:19:15 +0000 (13:19 +0000)]
Add a test that now passes.
llvm-svn: 298083
George Rimar [Fri, 17 Mar 2017 13:15:35 +0000 (13:15 +0000)]
[ELF] - Fix warning.
Writer.cpp:361:3: warning: suggest parentheses around assignment used as truth value [-Wparentheses]
llvm-svn: 298082
Michael Kruse [Fri, 17 Mar 2017 13:10:05 +0000 (13:10 +0000)]
[ScopInfo/PruneUnprofitable] Move default profitability check.
In the previous default ScopInfo applied the profitability heuristic for
scalar accesses (-polly-unprofitable-scalar-accs=true) and the
-polly-prune-unprofitable was disabled by default
(-polly-enable-prune-unprofitable=false) as that pruning was already done.
This changes switches the defaults to -polly-unprofitable-scalar-accs=true
-polly-enable-prune-unprofitable=false such that the scalar access
heuristic check is done by the pass. This allows passes between ScopInfo
and PruneUnprofitable to optimize away scalar accesses.
Without enabling such intermediate passes, there is no change in
behaviour of profitability checks in a PassManagerBuilder built
pass chain, but it allows us to cover this configuration with the
buildbots.
Suggested-by: Tobias Grosser <tobias@grosser.es>
llvm-svn: 298081
Michael Kruse [Fri, 17 Mar 2017 13:09:52 +0000 (13:09 +0000)]
[PruneUnprofitable] Add -polly-prune-unprofitable pass.
ScopInfo's normal profitability heuristic considers SCoPs where all
statements have scalar writes as not profitably optimizable and
invalidate the SCoP in that case. However, -polly-delicm and
-polly-simplify may be able to remove some of the scalar writes such
that the flag -polly-unprofitable-scalar-accs=false allows disabling
that part of the heuristic.
In cases where DeLICM (or other passes after ScopInfo) are not
successful in removing scalar writes, the SCoP is still not profitably
optimizable. The schedule optimizer would again try computing another
schedule, resulting in slower compilation.
The -polly-prune-unprofitable pass applies the profitability heuristic
again before the schedule optimizer Polly can still bail out even with
-polly-unprofitable-scalar-accs=false.
Differential Revision: https://reviews.llvm.org/D31033
llvm-svn: 298080
Rafael Espindola [Fri, 17 Mar 2017 13:05:04 +0000 (13:05 +0000)]
Change our linker script expr representation.
This fixes pr32031 by representing the expressions results as a
SectionBase and offset. This allows us to use an input section
directly instead of getting lost trying to compute an offset in an
outputsection when not all the information is available yet.
This also creates a struct to represent the *value* of and expression,
allowing the expression itself to be a simple typedef. I think this is
easier to read and will make it easier to extend the expression
computation to handle more complicated cases.
llvm-svn: 298079
George Rimar [Fri, 17 Mar 2017 13:04:06 +0000 (13:04 +0000)]
[ELF] - Simplify logic of creating "COMMON" section.
Patch reuses BssSection section to simplify creation of
COMMON section.
Differential revision: https://reviews.llvm.org/D30690
llvm-svn: 298078
Tobias Grosser [Fri, 17 Mar 2017 13:00:53 +0000 (13:00 +0000)]
[ScopInfo] Add option to not add parameter bounds to context [NFC]
For experiments it is sometimes helpful to provide parameter bound information
to polly and to not use these parameter bounds for simplification.
Add a new option "-polly-ignore-parameter-bounds" which does precisely this.
llvm-svn: 298077
Rafael Espindola [Fri, 17 Mar 2017 13:00:45 +0000 (13:00 +0000)]
Remove unnecessary (). NFC.
llvm-svn: 298076
Siddharth Bhat [Fri, 17 Mar 2017 12:59:01 +0000 (12:59 +0000)]
[DependenceInfo] Replace use of deprecated isl_dim_n_out [NFC]
Change isl_dim_n_out to isl_map_dim(*, isl_dim_out)
llvm-svn: 298075
Siddharth Bhat [Fri, 17 Mar 2017 12:31:28 +0000 (12:31 +0000)]
[DependenceInfo] Track may-writes and build flow information in
Dependences::calculateDependences.
This ensures that we handle may-writes correctly when building
dependence information. Also add a test case checking correctness of
may-write information. Not handling it before was an oversight.
Differential Revision: https://reviews.llvm.org/D31075
llvm-svn: 298074
Tobias Grosser [Fri, 17 Mar 2017 12:26:58 +0000 (12:26 +0000)]
[ScopInfo] Do not take inbounds assumptions [NFC]
For experiments it is sometimes helpful to not take any inbounds assumptions.
Add a new option "-polly-ignore-inbounds" which does precisely this.
llvm-svn: 298073
George Rimar [Fri, 17 Mar 2017 12:07:44 +0000 (12:07 +0000)]
[ELF] - Detemplate DynamicReloc class. NFC.
llvm-svn: 298072
George Rimar [Fri, 17 Mar 2017 11:56:54 +0000 (11:56 +0000)]
[ELF] - Detemplate SymbolBody::getVA and SymbolBody::getPltVA. NFC.
llvm-svn: 298071
George Rimar [Fri, 17 Mar 2017 11:41:56 +0000 (11:41 +0000)]
[ELF] - Fix linux build bot.
Error was:
Relocations.cpp:792:38: error: expected primary-expression before ‘>’ token
In<ELFT>::Iplt->addEntry<ELFT>(Body);
llvm-svn: 298070
Pavel Labath [Fri, 17 Mar 2017 11:33:57 +0000 (11:33 +0000)]
Fix FreeBSD build broken by r298066
llvm-svn: 298069
Oliver Stannard [Fri, 17 Mar 2017 11:10:17 +0000 (11:10 +0000)]
[Asm] Don't list '@<type>' in diag when '@' is a comment
This fixes https://bugs.llvm.org//show_bug.cgi?id=31280
Differential revision: https://reviews.llvm.org/D31026
llvm-svn: 298067
Pavel Labath [Fri, 17 Mar 2017 11:08:40 +0000 (11:08 +0000)]
Move GetAuxvData from Host to relevant process plugins
Summary:
GetAuxvData was causing dependencies from host to target and linux
process modules. It also does not fit netbsd use case, as there we can
only read the auxiliary vector with ptrace, which is better done in the
process plugin, with the other ptrace calls.
I resolve these issues by moving the freebsd and linux versions into the
relevant process plugins. In case of linux, this required adding an
interface in NativeProcessProtocol. The empty definitions on other
platforms can simply be removed.
To get the code compiling I had to add ProcessGdbRemote -> ProcessLinux
dependency, which was not caught before because we depended on it
transitively.
Reviewers: zturner, emaste
Subscribers: srhines, mgorny, lldb-commits
Differential Revision: https://reviews.llvm.org/D31031
llvm-svn: 298066
George Rimar [Fri, 17 Mar 2017 11:01:57 +0000 (11:01 +0000)]
[ELF] - Detemplate PltSection. NFC.
Alternative approach can be remove templated method
either, like D31028 do.
llvm-svn: 298065
George Rimar [Fri, 17 Mar 2017 10:32:10 +0000 (10:32 +0000)]
[ELF] - Fixed cache.ll testcase.
It failed for me under windows,
touch reported invalid date format.
Seems related to 12h/24h format, though
I don't understant why
197001010200 does not work,
but
197001010300 works, I see no logic here.
Changed to
197001011200.
llvm-svn: 298064
Pavel Labath [Fri, 17 Mar 2017 10:30:42 +0000 (10:30 +0000)]
One more attempt to fix FreeBSD
It seems sysctl.h is not self-contained, as I get missing symbols in the
header itself now. I am going to include all files that the file I moved
this from included, and hope that is enough.
llvm-svn: 298063
George Rimar [Fri, 17 Mar 2017 10:14:53 +0000 (10:14 +0000)]
[ELF] - Recommit "[ELF] - Make Bss and BssRelRo sections to be synthetic (#3)."
Was fixed, details on review page.
Original commit message:
That removes CopyRelSection class completely, making
Bss/BssRelRo to be just regular synthetics.
This is splitted from D30541 and polished.
Difference from D30541 that all logic of SharedSymbol
converting to DefinedRegular was removed for now and
probably will be posted as separate patch.
Differential revision: https://reviews.llvm.org/D30892
llvm-svn: 298062
Pavel Labath [Fri, 17 Mar 2017 10:09:56 +0000 (10:09 +0000)]
Speculative build fix for FreeBSD
broken by r298058.
llvm-svn: 298061
Alexander Kornienko [Fri, 17 Mar 2017 10:05:49 +0000 (10:05 +0000)]
[clang-tidy] Added a test with a different format.
llvm-svn: 298060
Alexander Kornienko [Fri, 17 Mar 2017 09:58:30 +0000 (09:58 +0000)]
[clang-tidy] readability-misleading-indentation: fix chained if
Summary:
Fixed erroneously flagging of chained if statements when styled like this:
```
if (cond) {
}
else if (cond) {
}
else {
}
```
Reviewers: xazax.hun, alexfh
Reviewed By: xazax.hun, alexfh
Subscribers: JDevlieghere, cfe-commits
Patch by Florian Gross!
Differential Revision: https://reviews.llvm.org/D30841
llvm-svn: 298059
Pavel Labath [Fri, 17 Mar 2017 09:51:23 +0000 (09:51 +0000)]
Remove HostThreadLinux/Free/NetBSD
Summary:
These classes existed only because of the GetName() static function,
which can be moved to a more natural place anyway. I move the linux
version to NativeProcessLinux (and get rid of ProcFileReader), the
freebsd version to ProcessFreeBSD (and fix a bug where it was using the
current process ID, instead of the inferior pid), and remove the NetBSD
version (which was probably incorrect anyway, as it assumes the current
process instead of the inferior.
I also add an llgs test to that verifies thread names are read
correctly.
Reviewers: zturner, krytarowski, emaste
Subscribers: lldb-commits, mgorny
Differential Revision: https://reviews.llvm.org/D30981
llvm-svn: 298058
Alexander Kornienko [Fri, 17 Mar 2017 09:47:05 +0000 (09:47 +0000)]
[clang-tidy] Verify some conditions in a matcher instead of check(). NFC
llvm-svn: 298057
Andre Vieira [Fri, 17 Mar 2017 09:37:10 +0000 (09:37 +0000)]
[ARM] Fix triple format in test branch disassemble test
Fixing triple format in the tests added for the branch label fix for Thumb
Targets. Also recommitting previously approved patch, see
https://reviews.llvm.org/D30943.
Reviewed by: samparker
Differential Revision: https://reviews.llvm.org/D30987
llvm-svn: 298056
Andre Vieira [Fri, 17 Mar 2017 09:37:02 +0000 (09:37 +0000)]
[ARM] Change tests after fixing branch label for Thumb targets
Patch to fix lld tests after my llvm-objdump patch to fix branch labels for
Thumb targets.
Reviewed by: emaste
Differential Revision: https://reviews.llvm.org/D30986
llvm-svn: 298055
Tobias Grosser [Fri, 17 Mar 2017 09:02:53 +0000 (09:02 +0000)]
[ScopInfo] Do not try to eliminate parameter dimensions that do not exist
In subsequent changes we will make Polly a little bit more lazy in adding
parameter dimensions to different sets. As a result, not all parameters will
always be part of the parameter space. This change ensures that we do not use
the '-1' returned when a parameter dimension cannot be found, but instead
just do not try to eliminate the anyhow non-existing dimension.
llvm-svn: 298054
Tobias Grosser [Fri, 17 Mar 2017 09:02:50 +0000 (09:02 +0000)]
[ScopInfo] Do not expand getDomains() to full parameter space.
Since several years, isl can perform most operations on sets with differing
parameter spaces, by expanding the parameter space on demand relying using
named isl ids to distinguish different parameter dimensions.
By not always expanding to full dimensionality the set remain smaller and can
likely be operated on faster. This change by itself did not yet result in
measurable performance benefits, but it is a step into the right direction
needed to ensure that subsequent changes indeed can work with lower-dimensional
sets and these sets do not get blown up by accident when later intersected with
the domain context.
llvm-svn: 298053
Alexander Kornienko [Fri, 17 Mar 2017 08:40:07 +0000 (08:40 +0000)]
[clang-tidy] Ignore deleted members in google-explicit-constructor.
This fixes http://llvm.org/PR32221.
llvm-svn: 298052
Craig Topper [Fri, 17 Mar 2017 07:37:31 +0000 (07:37 +0000)]
[AVX-512] Make VEX encoded FMA instructions available when AVX512 is enabled regardless of whether +fma was added on the command line.
We weren't able to handle isel of the 128/256-bit FMA instructions when AVX512F was enabled but VLX and FMA weren't.
I didn't mask FeatureAVX512 imply FeatureFMA as I wasn't sure I wanted disabling FMA to also disable AVX512. Instead we just can't prevent FMA instructions if AVX512 is enabled.
Another option would be to promote 128/256-bit to 512-bit, do the operation and extract it. But that requires a lot of extra isel patterns. Since no CPUs exist that support AVX512, but not FMA just using the VEX instructions seems better.
llvm-svn: 298051
Craig Topper [Fri, 17 Mar 2017 07:37:27 +0000 (07:37 +0000)]
[X86] Remove unused predicate. NFC
llvm-svn: 298050
Jonas Paulsson [Fri, 17 Mar 2017 07:11:46 +0000 (07:11 +0000)]
[SystemZ] New CodeGen tests for vector compare / select.
New SystemZ tests for the improved codegen of vector compare and select,
including cases with a logical combination of two compares.
Review: Ulrich Weigand.
https://reviews.llvm.org/D29489
llvm-svn: 298049
Jonas Paulsson [Fri, 17 Mar 2017 07:11:42 +0000 (07:11 +0000)]
Add SystemZ to utils/update_llc_test_checks.py.
Extend script for auto-generating CHECK lines so that it works for SystemZ.
This is a pre-commit for the new tests resulting from
https://reviews.llvm.org/D29489
llvm-svn: 298048
Jonas Paulsson [Fri, 17 Mar 2017 06:47:08 +0000 (06:47 +0000)]
[SystemZ] Add use of super-reg in splitMove()
If one of the subregs of the 128 bit reg is undefined when splitMove() splits
a store into two instructions, a use of an undefined physical register
results.
To remedy this, an implicit use of the super register is added onto both new
instructions, along with propagated kill and undef flags.
This was discovered with llvm-stress, and that test case is attached as
test/CodeGen/SystemZ/splitMove_undefReg_mverifier.ll
Thanks to Matthias Braun for helping with a nice explanation.
Review: Ulrich Weigand
llvm-svn: 298047
Craig Topper [Fri, 17 Mar 2017 06:10:37 +0000 (06:10 +0000)]
[AVX-512] Give priority to EVEX encoded scalar FMA instructions when we have FMA, AVX512 and no VLX.
We were giving priority if VLX was enabled.
llvm-svn: 298046
Craig Topper [Fri, 17 Mar 2017 06:00:01 +0000 (06:00 +0000)]
[X86] Use update_llc_test_checks.py to regenerate a test and add command lines to demonstrate that we don't pick EVEX encoded instruction when AVX512 and FMA3 are both enabled.
This bug only exists on the scalar llvm.fma instrinsics. Looks like we don't test the llvm.fma intrinsics very thoroughly. In fact I don't see any tests for the vector versions.
llvm-svn: 298045
Craig Topper [Fri, 17 Mar 2017 05:59:57 +0000 (05:59 +0000)]
[X86] Use update_llc_test_checks.py to regenerate a test.
llvm-svn: 298044
Craig Topper [Fri, 17 Mar 2017 05:59:54 +0000 (05:59 +0000)]
[X86] Cleanup the AddedComplexity values on move immediate instructions. NFC
This makes the values a little more consistent between similar instruction and reduces the values some. This results in better grouping in the isel table saving a few bytes.
llvm-svn: 298043
Craig Topper [Fri, 17 Mar 2017 05:59:25 +0000 (05:59 +0000)]
[AVX-512] Change the input type for some load intrinsics to take void type like the spec (and the test cases say).
llvm-svn: 298042
Craig Topper [Fri, 17 Mar 2017 05:59:22 +0000 (05:59 +0000)]
[AVX-512] Fix test cases that were using the builtins directly without typecasts instead of the intrinsic header.
llvm-svn: 298041
Craig Topper [Fri, 17 Mar 2017 05:14:37 +0000 (05:14 +0000)]
[AVX-512] Add missing typecasts and parentheses to _mm512_mask_i64gather_ps. My macro cleanup script I used on the others last year must have missed it.
llvm-svn: 298040