Commit Graph

110508 Commits

Author SHA1 Message Date
Rafael Espindola
cc0061b299 This reverts commit r223306 and r223277.
The code is using uninitialized memory and failing on linux.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223315 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 23:29:34 +00:00
Kostya Serebryany
0fd28b40a4 [msan] allow -fsanitize-coverage=N together with -fsanitize=memory, llvm part
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223312 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 23:28:26 +00:00
Jacques Pienaar
a78929d201 Test commit.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223310 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 23:21:02 +00:00
Rafael Espindola
a63b95fd14 Make llvm-ar a real build target and install it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223309 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 23:17:24 +00:00
Chris Matthews
0baca9564c Add a test-case for lit xunit output
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223307 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 23:01:35 +00:00
Kevin Enderby
15d4003c46 Move tests for llvm-objdump for universal files to X86 directory to fix build bots.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223306 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 23:00:16 +00:00
Chris Matthews
06a1a5f256 Don't allow test-suite names with . in them in xml output
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223305 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 22:59:15 +00:00
Rafael Espindola
55d5ea3be3 Split the set of identified struct types into opaque and non-opaque ones.
The non-opaque part can be structurally uniqued. To keep this to just
a hash lookup, we don't try to unique cyclic types.

Also change the type mapping algorithm to be optimistic about a type
not being recursive and only create a new type when proven to be wrong.
This is not as strong as trying to speculate that we can keep the source
type, but is simpler (no speculation to revert) and more powerfull
than what we had before (we don't copy non-recursive types at least).

I initially wrote this to try to replace the name based type merging.
It is not strong enough to replace it, but is is a useful addition.

With this patch the number of named struct types is a clang lto bootstrap goes
from 49674 to 15986.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223278 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 22:36:37 +00:00
Kevin Enderby
5b9d42b68c Add support to llvm-objdump for Mach-O universal files and archives with -macho.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223277 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 22:29:40 +00:00
Sanjay Patel
7e4c9bda0a fix typos, grammar, formatting; NFC
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223276 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 22:28:05 +00:00
Philip Reames
07caf6b709 Strength Verifier checks around the types involved in a statepoint
Add checks that the types in a gc.statepoint sequence match the wrapper callee and that relocating a pointer doesn't change it's type.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223275 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 22:23:24 +00:00
Matthias Braun
704923b2ba correct spelling, NFC
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223274 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 22:10:39 +00:00
Matthias Braun
b0ec6c21b7 [SimplifyLibCalls] Improve double->float shrinking to consider constants
This allows cases like float x; fmin(1.0, x); to be optimized to fminf(1.0f, x);

rdar://19049359

Differential Revision: http://reviews.llvm.org/D6496

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223270 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 21:46:33 +00:00
Matthias Braun
9d362ec2a4 [SimplifyLibCalls] Enable double to float shrinking for copysign
rdar://19049359

Differential Revision: http://reviews.llvm.org/D6495

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223269 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 21:46:29 +00:00
Colin LeMahieu
ec42b05190 [Hexagon] Converting member InstrDesc to static variable.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223268 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 21:40:25 +00:00
Nico Weber
e3f12719fd Fix grammar-o.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223265 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 20:58:23 +00:00
Colin LeMahieu
ca1a325ec7 [Hexagon] Converting subclass members to an implicit operand.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223264 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 20:23:22 +00:00
David Blaikie
687f237b53 range-for some things
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223263 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 19:58:49 +00:00
David Blaikie
4496906657 Simplify CodeGenRegBank::inferMatchingSuperRegClass & its caller by passing an iterator rather than index
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223262 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 19:58:48 +00:00
David Blaikie
7a16f34aa6 Simplify ownership of RegClasses by using list<CodeGenRegisterClass> instead of vector<CodeGenRegisterClass*>
This complicates a few algorithms due to not having random access, but
not by a huge degree I don't think (open to debate/design
discussion/etc).

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223261 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 19:58:45 +00:00
David Blaikie
89036eb44c Range-for some stuff related to RegClasses, and comment cases where range-for isn't suitable.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223260 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 19:58:41 +00:00
Philip Reames
93ce526d0a Make the Verifier more strict about gc.statepoints
The recently added documentation for statepoints claimed that we checked the parameters of the various intrinsics for validity.  This patch adds the code to actually do so.  I also removed a couple of redundant checks for conditions which are checked elsewhere in the Verifier and simplified the logic using the helper functions from Statepoint.h.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223259 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 19:53:15 +00:00
Will Schmidt
ad304153f4 Add TableGen info for Power8.
This is based on the Power7 version, with units added and renamed to match P8.

Differential Revision: http://reviews.llvm.org/D6358




git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223257 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 18:46:30 +00:00
Roman Divacky
788af07d5d Change the name to be in style.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223255 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 18:39:44 +00:00
Matt Arsenault
99eff50c00 Fix sphinx error from Statepoints.rst
It was complaining it wasn't included in any toctree

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223254 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 18:35:11 +00:00
Tom Stellard
f41e1d1a45 R600/SI: Move SIInsertWaits into AMDGPUPassConfig::addPreSched2()
This pass needs to be run after PrologEpilogInserter, because
that pass may inserter spill code which reads or writes memory.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223253 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 18:27:08 +00:00
Tom Stellard
8a8dc6a29d R600/SI: Don't run SI passes on R600 subtargets
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223252 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 18:27:05 +00:00
Tim Northover
34c1d6673e AArch64: fix wrong-endian parameter passing.
The blocked arguments code didn't take account of the hacks needed to support
it.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223247 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 17:49:26 +00:00
Colin LeMahieu
cf17914f89 [NFC] Fixing pendantic warning extra semicolons.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223246 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 17:36:39 +00:00
Colin LeMahieu
d54ac604b3 [Hexagon] [NFC] Moving function implementations out of header. Clang-formatting files.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223245 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 17:35:39 +00:00
Nick Lewycky
0f87df2033 Fix test to use the right metadata node (reapply r223239 plus a fix) and also to use the correct path to the GCNO file.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223244 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 17:32:44 +00:00
Colin LeMahieu
e6c2d47e82 [Hexagon] [NFC] Renaming *packetStart to *packetBegin
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223243 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 17:31:43 +00:00
Alexander Potapenko
22b6c01fc8 Revert r223239, which broke some bots.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223240 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 16:03:08 +00:00
Alexander Potapenko
2afd191abd Fix the metadata number used by llvm.gcov to match the number of the inserted metadata node.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223239 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 15:15:58 +00:00
Aaron Ballman
42fcf3d61a Silencing several "multiple copy constructors" warnings from MSVC; NFC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223238 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 14:44:16 +00:00
Aaron Ballman
13e20a9dac Silencing a 32-bit implicit conversion warning in MSVC; NFC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223237 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 14:39:58 +00:00
Evgeniy Stepanov
78d5d1ea1b msan] Add compile-time checks for missing origins.
This change makes MemorySanitizer instrumentation a bit more strict
about instructions that have no origin id assigned to them.

This would have caught the bug that was fixed in r222918.

This is re-commit of r222997, reverted in r223211, with 3 more
missing origins added.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223236 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 14:15:53 +00:00
Erik Eckstein
10e28ca6b1 InstCombine: simplify signed range checks
Try to convert two compares of a signed range check into a single unsigned compare.
Examples:
(icmp sge x, 0) & (icmp slt x, n) --> icmp ult x, n
(icmp slt x, 0) | (icmp sgt x, n) --> icmp ugt x, n




git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223224 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 10:39:15 +00:00
Hal Finkel
cf988bca67 [PowerPC] Print all inline-asm consts as signed numbers
Almost all immediates in PowerPC assembly (both 32-bit and 64-bit) are signed
numbers, and it is important that we print them as such. To make sure that
happens, we change PPCTargetLowering::LowerAsmOperandForConstraint so that it
does all intermediate checks on a signed-extended int64_t value, and then
creates the resulting target constant using MVT::i64. This will ensure that all
negative values are printed as negative values (mirroring what is done in other
backends to achieve the same sign-extension effect).

This came up in the context of inline assembly like this:
  "add%I2   %0,%0,%2", ..., "Ir"(-1ll)
where we used to print:
  addi   3,3,4294967295
and gcc would print:
  addi   3,3,-1
and gas accepts both forms, but our builtin assembler (correctly) does not. Now
we print -1 like gcc does.

While here, I replaced a bunch of custom integer checks with isInt<16> and
friends from MathExtras.h.

Thanks to Paul Hargrove for the bug report.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223220 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 09:37:50 +00:00
Charlie Turner
10cae8e352 Emit ABI_FP_rounding attribute.
LLVM understands a -enable-sign-dependent-rounding-fp-math codegen option. When
the user has specified this option, the Tag_ABI_FP_rounding attribute should be
emitted with value 1. This option currently does not appear to disable
transformations and optimizations that assume default floating point rounding
behavior, AFAICT, but the intention should be recorded in the build attributes,
regardless of what the compiler actually does with the intention.

Change-Id: If838578df3dc652b6f2796b8d152545674bcb30e

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223218 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 08:12:26 +00:00
Charlie Turner
78f9ab5f7c Add tests for default value of Tag_ABI_FP_rounding.
Change-Id: I051866d073fc6ce87ce3e693a3762da6d81f4393

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223217 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 07:59:50 +00:00
Benjamin Poulain
a941057b6c Fix a typo in the documentation of LTO
Fix defininitions->definitions.

Reviewed by David Blaikie.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223216 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 07:32:36 +00:00
Rafael Espindola
eee41dbb65 Ask the module for its the identified types.
When lazy reading a module, the types used in a function will not be visible to
a TypeFinder until the body is read.

This patch fixes that by asking the module for its identified struct types.
If a materializer is present, the module asks it. If not, it uses a TypeFinder.

This fixes pr21374.

I will be the first to say that this is ugly, but it was the best I could find.

Some of the options I looked at:

* Asking the LLVMContext. This could be made to work for gold, but not currently
  for ld64. ld64 will load multiple modules into a single context before merging
  them. This causes us to see types from future merges. Unfortunately,
  MappedTypes is not just a cache when it comes to opaque types. Once the
  mapping has been made, we have to remember it for as long as the key may
  be used. This would mean moving MappedTypes to the Linker class and having
  to drop the Linker::LinkModules static methods, which are visible from C.

* Adding an option to ignore function bodies in the TypeFinder. This would
  fix the PR by picking the worst result. It would work, but unfortunately
  we are currently quite dependent on the upfront type merging. I will
  try to reduce our dependency, but it is not clear that we will be able
  to get rid of it for now.

The only clean solution I could think of is making the Module own the types.
This would have other advantages, but it is a much bigger change. I will
propose it, but it is nice to have this fixed while that is discussed.

With the gold plugin, this patch takes the number of types in the LTO clang
binary from 52817 to 49669.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223215 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 07:18:23 +00:00
Duncan P. N. Exon Smith
59c9cb75d2 ADT: Rename argument in emplace_back_impl
Rename a functor argument in r223201 from `emplace` to `construct` to
reduce confusion.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223212 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 05:53:24 +00:00
Nick Lewycky
3c6860c8a1 Revert r222997. The newly added compile-time checks are finding missing origins, testcase is being reduced and a PR will be posted shortly.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223211 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 05:47:00 +00:00
Duncan P. N. Exon Smith
5ed8a767a0 LoopVectorize: Remove unnecessary RAUW
Remove an unnecessary `MDNode::replaceAllUsesWith()`.  In the preceding
line, `TheLoop->setLoopID()` visits all backedges and sets the new loop
ID.  This sufficiently updates the loop metadata.

Metadata RAUW is going away as part of PR21532.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223210 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 05:41:20 +00:00
Matt Arsenault
a4ec31ed52 R600/SI: Fix SIFixSGPRCopies for copies to physical registers
This shows up when operands required to be passed in VCC are copied
to.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223208 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 05:22:39 +00:00
Matt Arsenault
5c2d295d86 R600/SI: Remove incorrect assertion
This can be a COPY to a physical register, such as VCC

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223207 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 05:22:38 +00:00
Matt Arsenault
ec0a7cd15a R600/SI: Remove i1 pseudo VALU ops
Select i1 logical ops directly to 64-bit SALU instructions.
Vector i1 values are always really in SGPRs, with each
bit for each item in the wave. This saves about 4 instructions
when and/or/xoring any condition, and also helps write conditions
that need to be passed in vcc.

This should work correctly now that the SGPR live range
fixing pass works. More work is needed to eliminate the VReg_1
pseudo regclass and possibly the entire SILowerI1Copies pass.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223206 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 05:22:35 +00:00
Matt Arsenault
bd5f9f45d1 R600/SI: Fix suspicious indexing
The loop is over the operands of an instruction, and checks the
register with the sub reg index of the dest register. This probably
meant to be checking the sub reg index of the same operand.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223205 91177308-0d34-0410-b5e6-96231b3b80d8
2014-12-03 05:22:32 +00:00