OSDN Git Service
Chad Rosier [Fri, 12 Sep 2014 17:55:16 +0000 (17:55 +0000)]
FileCheckize. NFC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217698
91177308-0d34-0410-b5e6-
96231b3b80d8
JF Bastien [Fri, 12 Sep 2014 17:54:17 +0000 (17:54 +0000)]
Add support for le64.
Summary:
le64 is a generic little-endian 64-bit processor, mimicking le32.
Depends on D5318.
Test Plan: make check-all
Reviewers: dschuff
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D5319
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217697
91177308-0d34-0410-b5e6-
96231b3b80d8
Chad Rosier [Fri, 12 Sep 2014 17:40:39 +0000 (17:40 +0000)]
[AArch64] Enable post-RA MI scheduler.
Phabricator Revision: http://reviews.llvm.org/D5278
Patch by Sanjin Sijaric!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217693
91177308-0d34-0410-b5e6-
96231b3b80d8
James Molloy [Fri, 12 Sep 2014 16:55:32 +0000 (16:55 +0000)]
[A57FPLoadBalancing] Remove support for vector types
Vector MUL/MLAs have tied operands, which gives us extra constraints
that we currently can't handle. Instead of silently doing the wrong
thing, remove support to be readded later properly.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217690
91177308-0d34-0410-b5e6-
96231b3b80d8
James Molloy [Fri, 12 Sep 2014 16:55:26 +0000 (16:55 +0000)]
[A57FPLoadBalancing] Ignore <def>s when checking if a chain may be killed.
Defs are seen before uses, so a def without the kill flag doesn't necessarily
mean that the register is not killed on that instruction. It may be killed
in a later use operand.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217689
91177308-0d34-0410-b5e6-
96231b3b80d8
Jordan Rose [Fri, 12 Sep 2014 16:46:05 +0000 (16:46 +0000)]
[lit] Parse all strings as UTF-8 rather than ASCII.
As far as I can tell UTF-8 has been supported since the beginning of Python's
codec support, and it's the de facto standard for text these days, at least
for primarily-English text. This allows us to put Unicode into lit RUN lines.
rdar://problem/
18311663
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217688
91177308-0d34-0410-b5e6-
96231b3b80d8
NAKAMURA Takumi [Fri, 12 Sep 2014 15:12:32 +0000 (15:12 +0000)]
Move sys::fs::AccessMode out of @brief in the function. [-Wdocumentation]
FIXME: Annotate sys::fs::AccessMode.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217685
91177308-0d34-0410-b5e6-
96231b3b80d8
NAKAMURA Takumi [Fri, 12 Sep 2014 15:12:21 +0000 (15:12 +0000)]
sys::fs::access(): Fix @param [-Wdocumentation]
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217684
91177308-0d34-0410-b5e6-
96231b3b80d8
NAKAMURA Takumi [Fri, 12 Sep 2014 15:10:56 +0000 (15:10 +0000)]
llvm/test/CodeGen/X86/vec_ctbits.ll: Add explicit -mtriple=x86_64-unknown. It was incompatible to Win32 x64.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217683
91177308-0d34-0410-b5e6-
96231b3b80d8
James Molloy [Fri, 12 Sep 2014 14:35:17 +0000 (14:35 +0000)]
[A57LoadBalancing] unique_ptr-ify.
Thanks to David Blakie for the in-depth review!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217682
91177308-0d34-0410-b5e6-
96231b3b80d8
Zoran Jovanovic [Fri, 12 Sep 2014 14:29:54 +0000 (14:29 +0000)]
[mips][microMIPS] Implement JRADDIUSP instruction
Differential Revision: http://reviews.llvm.org/D5046
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217681
91177308-0d34-0410-b5e6-
96231b3b80d8
Bill Schmidt [Fri, 12 Sep 2014 14:26:36 +0000 (14:26 +0000)]
Address comments on r217622
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217680
91177308-0d34-0410-b5e6-
96231b3b80d8
Zoran Jovanovic [Fri, 12 Sep 2014 13:51:58 +0000 (13:51 +0000)]
[mips][microMIPS] Implement BGEZALS and BLTZALS instructions
Differential Revision: http://reviews.llvm.org/D5004
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217678
91177308-0d34-0410-b5e6-
96231b3b80d8
Zoran Jovanovic [Fri, 12 Sep 2014 13:43:41 +0000 (13:43 +0000)]
[mips][microMIPS] Implement JALS and JALRS instructions.
Differential Revision: http://reviews.llvm.org/D5003
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217676
91177308-0d34-0410-b5e6-
96231b3b80d8
Zoran Jovanovic [Fri, 12 Sep 2014 13:33:33 +0000 (13:33 +0000)]
[mips][microMIPS] Implement TLBP, TLBR, TLBWI and TLBWR instructions
Differential Revision: http://reviews.llvm.org/D5211
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217675
91177308-0d34-0410-b5e6-
96231b3b80d8
James Molloy [Fri, 12 Sep 2014 13:29:40 +0000 (13:29 +0000)]
[ARM] Teach the cost model that cross-class copies are costly.
Cross-class copies being expensive is actually a trait of the microarchitecture, but as I haven't yet seen an example of a microarchitecture where they're cheap it seems best to just enable this by default, covering the non-mcpu build case.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217674
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Fri, 12 Sep 2014 12:50:27 +0000 (12:50 +0000)]
Legalizer: Use the scalar bit width when promoting bit counting instrs on
vectors.
e.g. when promoting ctlz from <2 x i32> to <2 x i64> we have to fixup
the result by 32 bits, not 64. PR20917.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217671
91177308-0d34-0410-b5e6-
96231b3b80d8
Patrik Hagglund [Fri, 12 Sep 2014 12:32:08 +0000 (12:32 +0000)]
Fix gcc -Wpedantic.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217669
91177308-0d34-0410-b5e6-
96231b3b80d8
Evgeniy Stepanov [Fri, 12 Sep 2014 11:08:59 +0000 (11:08 +0000)]
Add CMake check for libatomic.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217666
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Fri, 12 Sep 2014 08:56:53 +0000 (08:56 +0000)]
Fix an ODR violation consisting of two 'struct Query' in the global namespace.
Put them in their own anonymous namespaces. Found by GCC's new -Wodr (PR20915).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217662
91177308-0d34-0410-b5e6-
96231b3b80d8
Joerg Sonnenberger [Fri, 12 Sep 2014 08:07:31 +0000 (08:07 +0000)]
Add Tom Stellard's role as 3.5 release manager.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217659
91177308-0d34-0410-b5e6-
96231b3b80d8
Justin Bogner [Fri, 12 Sep 2014 06:52:44 +0000 (06:52 +0000)]
llvm-cov: Move FunctionCoverageMapping into CoverageMapping.h (NFC)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217657
91177308-0d34-0410-b5e6-
96231b3b80d8
Craig Topper [Fri, 12 Sep 2014 05:17:20 +0000 (05:17 +0000)]
Remove a temporary variable and just construct a unique_ptr directly using make_unique.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217655
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Thu, 11 Sep 2014 23:42:10 +0000 (23:42 +0000)]
Object: Small cleanup in COFFObjectFile::getSymbol
NFC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217648
91177308-0d34-0410-b5e6-
96231b3b80d8
Justin Bogner [Thu, 11 Sep 2014 23:20:48 +0000 (23:20 +0000)]
Revert "llvm-cov: Remove an overly system specific test"
This fixes a call to sys::fs::equivalent that should've been to
CodeCoverageTool::equivalentFiles, which lets us restore the test of
r217476 that was removed in r217478.
This reverts r217478, but the test works this time.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217646
91177308-0d34-0410-b5e6-
96231b3b80d8
Lang Hames [Thu, 11 Sep 2014 23:09:22 +0000 (23:09 +0000)]
[MCJIT] Improve the "stub not found" diagnostic in RuntimeDyldChecker.
A "stub found found" diagnostic is emitted when RuntimeDyldChecker's stub lookup
logic fails to find the requested stub. The obvious reason for the failure is
that no such stub has been created, but it can also fail for internal symbols if
the symbol offset is not computed correctly (E.g. due to a mangled relocation
addend). This patch adds a comment about the latter case so that it's not
overlooked.
Inspired by confusion experienced during test case construction for r217635.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217643
91177308-0d34-0410-b5e6-
96231b3b80d8
Rui Ueyama [Thu, 11 Sep 2014 22:55:25 +0000 (22:55 +0000)]
[Support][Endian] Overload += and -=
This patch is to overload operator+= and operator-= for
{u}{little}{big}{16,32,64}_t.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217637
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Thu, 11 Sep 2014 22:51:37 +0000 (22:51 +0000)]
R600/SI: Fix off by 1 error in used register count
The register numbers start at 0, so if only 1 register
was used, this was reported as 0.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217636
91177308-0d34-0410-b5e6-
96231b3b80d8
Lang Hames [Thu, 11 Sep 2014 22:43:36 +0000 (22:43 +0000)]
[MCJIT] Make sure we test ARM BR24 relocations with both internal and external
symbols.
Previously we have only been testing these relocations with external symbols.
<rdar://problem/
18308413>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217635
91177308-0d34-0410-b5e6-
96231b3b80d8
Rui Ueyama [Thu, 11 Sep 2014 22:34:32 +0000 (22:34 +0000)]
Support: Use llvm::COFF::BigObjMagic
Use llvm::COFF::BigObjMagic insetad of the string literal.
Also checks the version number.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217633
91177308-0d34-0410-b5e6-
96231b3b80d8
Rui Ueyama [Thu, 11 Sep 2014 21:46:33 +0000 (21:46 +0000)]
Support: Delete {aligned_,}{u,}{little,big}8_t
The byte has no endianness, so these types don't make sense.
uint8_t should be used instead.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217631
91177308-0d34-0410-b5e6-
96231b3b80d8
Juergen Ributzka [Thu, 11 Sep 2014 21:32:32 +0000 (21:32 +0000)]
[C API] Make the 'lower switch' pass available via the C API.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217630
91177308-0d34-0410-b5e6-
96231b3b80d8
Quentin Colombet [Thu, 11 Sep 2014 21:22:14 +0000 (21:22 +0000)]
[CodeGenPrepare] Teach the addressing mode matcher how to promote zext.
I.e., teach it about 'sext (zext a to ty) to ty2' => zext a to ty2.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217629
91177308-0d34-0410-b5e6-
96231b3b80d8
David Blaikie [Thu, 11 Sep 2014 21:12:48 +0000 (21:12 +0000)]
Remove the unused string section symbol parameter from DwarfFile::emitStrings
And since it /looked/ like the DwarfStrSectionSym was unused, I tried
removing it - but then it turned out that DwarfStringPool was
reconstructing the same label (and expecting it to have already been
emitted) and uses that.
So I kept it around, but wanted to pass it in to users - since it seemed
a bit silly for DwarfStringPool to have it passed in and returned but
itself have no use for it. The only two users don't handle strings in
both .dwo and .o files so they only ever need the one symbol - no need
to keep it (and have an unused symbol) in the DwarfStringPool used for
fission/.dwo.
Refactor a bunch of accelerator table usage to remove duplication so I
didn't have to touch 4-5 callers.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217628
91177308-0d34-0410-b5e6-
96231b3b80d8
Rui Ueyama [Thu, 11 Sep 2014 21:09:57 +0000 (21:09 +0000)]
Support: improve identify_magic to recognize COFF bigobj
identify_magic recognized a COFF bigobj as an import library file.
This patch fixes that.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217627
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Thu, 11 Sep 2014 20:30:02 +0000 (20:30 +0000)]
Misc cleanups to the FileSytem api.
The main difference is the removal of
std::error_code exists(const Twine &path, bool &result);
It was an horribly redundant interface since a file not existing is also a valid
error_code. Now we have an access function that returns just an error_code. This
is the only function that has to be implemented for Unix and Windows. The
functions can_write, exists and can_execute an now just wrappers.
One still has to be very careful using these function to avoid introducing
race conditions (Time of check to time of use).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217625
91177308-0d34-0410-b5e6-
96231b3b80d8
Bill Schmidt [Thu, 11 Sep 2014 20:13:52 +0000 (20:13 +0000)]
Add missing colon to RUN line...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217623
91177308-0d34-0410-b5e6-
96231b3b80d8
Bill Schmidt [Thu, 11 Sep 2014 20:10:03 +0000 (20:10 +0000)]
[PATCH, PowerPC] Accept 'U' and 'X' constraints in inline asm
Inline asm may specify 'U' and 'X' constraints to print a 'u' for an
update-form memory reference, or an 'x' for an indexed-form memory
reference. However, these are really only useful in GCC internal code
generation. In inline asm the operand of the memory constraint is
typically just a register containing the address, so 'U' and 'X' make
no sense.
This patch quietly accepts 'U' and 'X' in inline asm patterns, but
otherwise does nothing. If we ever unexpectedly see a non-register,
we'll assert and sort it out afterwards.
I've added a new test for these constraints; the test case should be
used for other asm-constraints changes down the road.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217622
91177308-0d34-0410-b5e6-
96231b3b80d8
Lang Hames [Thu, 11 Sep 2014 19:21:14 +0000 (19:21 +0000)]
[MCJIT] Add support for ARM HALF_DIFF relocations to MCJIT.
Fixes <rdar://problem/
18297804>.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217620
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Thu, 11 Sep 2014 19:11:02 +0000 (19:11 +0000)]
Use simpler version of sys::fs::exists. NFC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217618
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Thu, 11 Sep 2014 18:44:26 +0000 (18:44 +0000)]
Use the simpler sys::fs:;exists. NFC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217617
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Thu, 11 Sep 2014 17:50:20 +0000 (17:50 +0000)]
Add triple to test to fix bots
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217612
91177308-0d34-0410-b5e6-
96231b3b80d8
Brad Smith [Thu, 11 Sep 2014 17:40:51 +0000 (17:40 +0000)]
Provide an implementation of getNoopForMachoTarget for SPARC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217611
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Thu, 11 Sep 2014 17:34:19 +0000 (17:34 +0000)]
Add DAG combine for shl + add of constants.
Do
(shl (add x, c1), c2) -> (add (shl x, c2), c1 << c2)
This is already done for multiplies, but since multiplies
by powers of two are turned into shifts, we also need
to handle it here.
This might want checks for isLegalAddImmediate to avoid
transforming an add of a legal immediate with one that isn't.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217610
91177308-0d34-0410-b5e6-
96231b3b80d8
Lang Hames [Thu, 11 Sep 2014 17:27:01 +0000 (17:27 +0000)]
[MCJIT] Take the relocation addend into account when applying ARM MachO VANILLA
and BR24 relocations.
<rdar://problem/
18296496>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217605
91177308-0d34-0410-b5e6-
96231b3b80d8
Ismail Pazarbasi [Thu, 11 Sep 2014 17:19:54 +0000 (17:19 +0000)]
Added missing LLVM_NOEXCEPT to the definition of _obj2yaml_error_category::name
LLVM_NOEXCEPT was added in r210591.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217603
91177308-0d34-0410-b5e6-
96231b3b80d8
Adam Nemet [Thu, 11 Sep 2014 16:51:10 +0000 (16:51 +0000)]
[AVX512] Fix miscompile for unpack
r189189 implemented AVX512 unpack by essentially performing a 256-bit unpack
between the low and the high 256 bits of src1 into the low part of the
destination and another unpack of the low and high 256 bits of src2 into the
high part of the destination.
I don't think that's how unpack works. AVX512 unpack simply has more 128-bit
lanes but other than it works the same way as AVX. So in each 128-bit lane,
we're always interleaving certain parts of both operands rather different
parts of one of the operands.
E.g. for this:
__v16sf a = { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 };
__v16sf b = { 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31 };
__v16sf c = __builtin_shufflevector(a, b, 0, 8, 1, 9, 4, 12, 5, 13, 16,
24, 17, 25, 20, 28, 21, 29);
we generated punpcklps (notice how the elements of a and b are not interleaved
in the shuffle). In turn, c was set to this:
0 16 1 17 4 20 5 21 8 24 9 25 12 28 13 29
Obviously this should have just returned the mask vector of the shuffle
vector.
I mostly reverted this change and made sure the original AVX code worked
for 512-bit vectors as well.
Also updated the tests because they matched the logic from the code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217602
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Thu, 11 Sep 2014 16:08:44 +0000 (16:08 +0000)]
Add triple and remove hashes to account for buildbot differences in comment strings.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217601
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Thu, 11 Sep 2014 15:58:39 +0000 (15:58 +0000)]
Move constant-sized bitvector to the stack.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217600
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Thu, 11 Sep 2014 15:45:27 +0000 (15:45 +0000)]
Combine fmul vector FP constants when unsafe math is allowed.
This is an extension of the change made with r215820:
http://llvm.org/viewvc/llvm-project?view=revision&revision=215820
That patch allowed combining of splatted vector FP constants that are multiplied.
This patch allows combining non-uniform vector FP constants too by relaxing the
check on the type of vector. Also, canonicalize a vector fmul in the
same way that we already do for scalars - if only one operand of the fmul is a
constant, make it operand 1. Otherwise, we miss potential folds.
This fold is also done by -instcombine, but it's possible that extra
fmuls may have been generated during lowering.
Differential Revision: http://reviews.llvm.org/D5254
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217599
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Thu, 11 Sep 2014 15:41:01 +0000 (15:41 +0000)]
typo
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217597
91177308-0d34-0410-b5e6-
96231b3b80d8
Aaron Watry [Thu, 11 Sep 2014 15:02:54 +0000 (15:02 +0000)]
R600: Add cmpxchg instruction for evergreen
Refactored the R600_LDS_1A2D class a bit to get it to actually work.
It seemed to be previously unused and broken.
We also have to disable the conversion to the noret variant for now in
R600ISelLowering because the getLDSNoRetOp method only handles 1A1D LDS ops.
Someone can feel free to modify the AMDGPU::getLDSNoRetOp method to
work for more than 1A1D variants of LDS operations. It's being left as a
future TODO for now.
Signed-off-by: Aaron Watry <awatry at gmail.com>
Reviewed-by: Matt Arsenault <matthew.arsenault@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217596
91177308-0d34-0410-b5e6-
96231b3b80d8
Aaron Watry [Thu, 11 Sep 2014 15:02:52 +0000 (15:02 +0000)]
R600: Test local atomics for evergreen
Now that the operations are all implemented, we can test this sub-arch here.
Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Matt Arsenault <matthew.arsenault@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217595
91177308-0d34-0410-b5e6-
96231b3b80d8
Aaron Watry [Thu, 11 Sep 2014 15:02:49 +0000 (15:02 +0000)]
R600: Add LDS_WRXCHG[_RET] instructions for Evergreen.
Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Matt Arsenault <matthew.arsenault@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217594
91177308-0d34-0410-b5e6-
96231b3b80d8
Aaron Watry [Thu, 11 Sep 2014 15:02:47 +0000 (15:02 +0000)]
R600: Add LDS_MIN_[U]INT[_RET] instructions for Evergreen
Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Matt Arsenault <matthew.arsenault@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217593
91177308-0d34-0410-b5e6-
96231b3b80d8
Aaron Watry [Thu, 11 Sep 2014 15:02:46 +0000 (15:02 +0000)]
R600: Add LDS_XOR[_RET] instructions for Evergreen
Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Matt Arsenault <matthew.arsenault@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217592
91177308-0d34-0410-b5e6-
96231b3b80d8
Aaron Watry [Thu, 11 Sep 2014 15:02:44 +0000 (15:02 +0000)]
R600: Add LDS_OR[_RET] instructions for Evergreen
Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Matt Arsenault <matthew.arsenault@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217591
91177308-0d34-0410-b5e6-
96231b3b80d8
Aaron Watry [Thu, 11 Sep 2014 15:02:43 +0000 (15:02 +0000)]
R600: Add LDS_AND[_RET] instructions for Evergreen
Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Matt Arsenault <matthew.arsenault@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217590
91177308-0d34-0410-b5e6-
96231b3b80d8
Aaron Watry [Thu, 11 Sep 2014 15:02:41 +0000 (15:02 +0000)]
R600: Add LDS_MAX_[U]INT[_RET] instructions for Evergreen
This was only present for SI before.
Cayman may still be missing, but I am unable to test that currently.
v2: Don't create atomicrmw max tests in separate file
Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Matt Arsenault <matthew.arsenault@amd.com>
CC: Tom Stellard <thomas.stellard@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217589
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Thu, 11 Sep 2014 12:02:59 +0000 (12:02 +0000)]
[docs] Mention character array constants in docs/LangRef.rst
Summary:
They were used in the 'Module Structure' example but weren't otherwise
documented.
Credit to Reed Kotler for noticing.
Reviewers: hans
Reviewed By: hans
Subscribers: hans, llvm-commits
Differential Revision: http://reviews.llvm.org/D5191
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217583
91177308-0d34-0410-b5e6-
96231b3b80d8
Tilmann Scheller [Thu, 11 Sep 2014 10:45:50 +0000 (10:45 +0000)]
[ARM] Add Thumb-2 code size optimization regression test for LSR (register).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217582
91177308-0d34-0410-b5e6-
96231b3b80d8
Tilmann Scheller [Thu, 11 Sep 2014 10:42:17 +0000 (10:42 +0000)]
[ARM] Add Thumb-2 code size optimization regression test for LSR (immediate).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217581
91177308-0d34-0410-b5e6-
96231b3b80d8
Arnaud A. de Grandmaison [Thu, 11 Sep 2014 10:39:52 +0000 (10:39 +0000)]
[AArch64] Reenable the PBQP test now that the leak issue has been fixed.
David Blaikie's commits r217563 & r217564, which added shared_ptr to the
CostPool have fixed some memory leak issues exposed by the PBQP with
coalescing constraints.
The sanitizer bot was failing because of those leaks. Now that the leaks
are gone, we can reenable the aarch64/pbqp test.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217580
91177308-0d34-0410-b5e6-
96231b3b80d8
Tilmann Scheller [Thu, 11 Sep 2014 10:33:39 +0000 (10:33 +0000)]
[ARM] Add Thumb-2 code size optimization regression test for LSL (register).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217579
91177308-0d34-0410-b5e6-
96231b3b80d8
Tilmann Scheller [Thu, 11 Sep 2014 10:29:42 +0000 (10:29 +0000)]
[ARM] Add Thumb2 code size optimization regression test for LSL (immediate).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217576
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Thu, 11 Sep 2014 10:21:25 +0000 (10:21 +0000)]
[x86] Fixup r217565 which baked in an assumption about the function
name that breaks on some platforms. This part of the test just doesn't
matter...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217575
91177308-0d34-0410-b5e6-
96231b3b80d8
Hal Finkel [Thu, 11 Sep 2014 08:40:17 +0000 (08:40 +0000)]
[AlignmentFromAssumptions] Don't crash just because the target is 32-bit
We used to crash processing any relevant @llvm.assume on a 32-bit target
(because we'd ask SE to subtract expressions of differing types). I've copied
our 'simple.ll' test, but with the data layout from arm-linux-gnueabihf to get
some meaningful test coverage here.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217574
91177308-0d34-0410-b5e6-
96231b3b80d8
David Xu [Thu, 11 Sep 2014 05:10:28 +0000 (05:10 +0000)]
Build correct vector filled with undef nodes
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217570
91177308-0d34-0410-b5e6-
96231b3b80d8
Justin Bogner [Thu, 11 Sep 2014 01:47:38 +0000 (01:47 +0000)]
utils: Teach lldbDataFormatters how to format ArrayRefs
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217567
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Thu, 11 Sep 2014 00:13:35 +0000 (00:13 +0000)]
[x86] FileCheck-ize this test.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217565
91177308-0d34-0410-b5e6-
96231b3b80d8
David Blaikie [Thu, 11 Sep 2014 00:08:54 +0000 (00:08 +0000)]
Use CostPool::PoolRef typedef some more
Cleanup to 217563 suggested by Lang Hames in post-commit review.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217564
91177308-0d34-0410-b5e6-
96231b3b80d8
David Blaikie [Wed, 10 Sep 2014 23:54:45 +0000 (23:54 +0000)]
shared_ptrify ownershp of PoolEntries in PBQP's CostPool
Leveraging both intrusive shared_ptr-ing (std::enable_shared_from_this)
and shared_ptr<T>-owning-U (to allow external users to hold
std::shared_ptr<CostT> while keeping the underlying PoolEntry alive).
The intrusiveness could be removed if we had a weak_set that implicitly
removed items from the set when their underlying data went away.
This /might/ fix an existing memory leak reported by LeakSanitizer in
r217504.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217563
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Wed, 10 Sep 2014 23:26:19 +0000 (23:26 +0000)]
R600/SI: Fix losing chain when fixing reg class of loads.
The lost chain resulting in earlier side effecting nodes
being deleted.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217561
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Wed, 10 Sep 2014 23:26:16 +0000 (23:26 +0000)]
R600/SI: Report offset in correct units for st64 DS instructions
Need to convert the 64 element offset into bytes, not just the element
size like the normal case instructions.
Noticed by inspection. This can't be hit now because
st64 instructions aren't emitted during instruction selection,
and the post-RA scheduler isn't enabled.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217560
91177308-0d34-0410-b5e6-
96231b3b80d8
Peter Collingbourne [Wed, 10 Sep 2014 22:20:49 +0000 (22:20 +0000)]
Add LLVMgold target to test dependencies.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217557
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Wed, 10 Sep 2014 21:44:27 +0000 (21:44 +0000)]
R600: Custom lower frem
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217553
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Wed, 10 Sep 2014 21:27:43 +0000 (21:27 +0000)]
Add doInitialization/doFinalization to DataLayoutPass.
With this a DataLayoutPass can be reused for multiple modules.
Once we have doInitialization/doFinalization, it doesn't seem necessary to pass
a Module to the constructor.
Overall this change seems in line with the idea of making DataLayout a required
part of Module. With it the only way of having a DataLayout used is to add it
to the Module.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217548
91177308-0d34-0410-b5e6-
96231b3b80d8
Hal Finkel [Wed, 10 Sep 2014 21:06:11 +0000 (21:06 +0000)]
Enable use of __builtin_assume_aligned when self-hosting
Clang/LLVM trunk now have support for __builtin_assume_aligned, turn this &&
into an || so we can use it ourselves.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217545
91177308-0d34-0410-b5e6-
96231b3b80d8
Hal Finkel [Wed, 10 Sep 2014 21:05:52 +0000 (21:05 +0000)]
[AlignmentFromAssumptions] Don't divide by zero for unknown starting alignment
The routine that determines an alignment given some SCEV returns zero if the
answer is unknown. In a case where we could determine the increment of an
AddRec but not the starting alignment, we would compute the integer modulus by
zero (which is illegal and traps). Prevent this by returning early if either
the start or increment alignment is unknown (zero).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217544
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Liew [Wed, 10 Sep 2014 20:43:03 +0000 (20:43 +0000)]
[sphinx cleanup]
Fix sphinx warning introduced by r217537
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217541
91177308-0d34-0410-b5e6-
96231b3b80d8
Gerolf Hoflehner [Wed, 10 Sep 2014 20:31:57 +0000 (20:31 +0000)]
[AArch64] Revert r216141 for cyclone
The increase of the interleave factor to 4 has side-effects
like performance losses eg. due to reminder loops being executed
more frequently and may increase code size. It requires more
analysis and careful heuristic tuning. Expect double digit gains
in small benchmarks like lowercase.c and losses in puzzle.c.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217540
91177308-0d34-0410-b5e6-
96231b3b80d8
Gerolf Hoflehner [Wed, 10 Sep 2014 20:24:03 +0000 (20:24 +0000)]
[PassManager] Adding Hidden attribute to EnableMLSM option
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217539
91177308-0d34-0410-b5e6-
96231b3b80d8
Gerolf Hoflehner [Wed, 10 Sep 2014 19:55:29 +0000 (19:55 +0000)]
[MergedLoadStoreMotion] Move pass enabling option to PassManagerBuilder
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217538
91177308-0d34-0410-b5e6-
96231b3b80d8
Nico Weber [Wed, 10 Sep 2014 19:50:55 +0000 (19:50 +0000)]
Fix docs reference to inexistent class.
Patch sent via telegraph by TNorthover. Thanks!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217537
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Wed, 10 Sep 2014 19:39:41 +0000 (19:39 +0000)]
Remember to eraseFromParent after replaceAllUsesWith.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217536
91177308-0d34-0410-b5e6-
96231b3b80d8
Adrian Prantl [Wed, 10 Sep 2014 18:52:29 +0000 (18:52 +0000)]
Cleanup: Use the appropriate API for accessing the DIVariable of a
DBG_VALUE intrinsic.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217533
91177308-0d34-0410-b5e6-
96231b3b80d8
Arnaud A. de Grandmaison [Wed, 10 Sep 2014 18:40:18 +0000 (18:40 +0000)]
[AArch64] Temporarily desactivate the PBQP test, while I investigate some leaks in the allocator
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217531
91177308-0d34-0410-b5e6-
96231b3b80d8
Alexey Samsonov [Wed, 10 Sep 2014 18:00:17 +0000 (18:00 +0000)]
Make CallingConv::ID an alias of "unsigned".
Summary:
Make CallingConv::ID a plain unsigned instead of enum with a
fixed set of valus. LLVM IR allows arbitraty calling conventions (you are
free to write
cc12345), and loading them as enum is an undefined
behavior. This was reported by UBSan.
Test Plan: llvm regression test suite
Reviewers: nicholas
Reviewed By: nicholas
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D5248
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217529
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Wed, 10 Sep 2014 17:58:16 +0000 (17:58 +0000)]
Rename getMaximumUnrollFactor -> getMaxInterleaveFactor; also rename option names controlling this variable.
"Unroll" is not the appropriate name for this variable. Clang already uses
the term "interleave" in pragmas and metadata for this.
Differential Revision: http://reviews.llvm.org/D5066
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217528
91177308-0d34-0410-b5e6-
96231b3b80d8
Gerolf Hoflehner [Wed, 10 Sep 2014 17:54:50 +0000 (17:54 +0000)]
Removed misleading comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217527
91177308-0d34-0410-b5e6-
96231b3b80d8
Gerolf Hoflehner [Wed, 10 Sep 2014 17:52:27 +0000 (17:52 +0000)]
Added missing blank
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217526
91177308-0d34-0410-b5e6-
96231b3b80d8
Hans Wennborg [Wed, 10 Sep 2014 17:05:08 +0000 (17:05 +0000)]
LangRef: @baz should be @bar in the COMDAT example
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217520
91177308-0d34-0410-b5e6-
96231b3b80d8
Arnaud A. de Grandmaison [Wed, 10 Sep 2014 17:03:25 +0000 (17:03 +0000)]
[AArch64] Address Chad's post commit review comments for r217504 (PBQP experimental support)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217518
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Wed, 10 Sep 2014 16:58:40 +0000 (16:58 +0000)]
typo
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217516
91177308-0d34-0410-b5e6-
96231b3b80d8
Frederic Riss [Wed, 10 Sep 2014 16:03:14 +0000 (16:03 +0000)]
Fix comments of createReplaceableForwardDecl() and createForwardDecl().
Noticed while trying to understand how the merge of forward decalred types
and defintions work.
Reviewers: echristo, dblaikie, aprantl
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D5291
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217514
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Wed, 10 Sep 2014 15:50:08 +0000 (15:50 +0000)]
Replace a few virtual with override.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217513
91177308-0d34-0410-b5e6-
96231b3b80d8
Arnaud A. de Grandmaison [Wed, 10 Sep 2014 14:24:02 +0000 (14:24 +0000)]
[AArch64] Pacify lld buildbot complaining about an unused static function in release build.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217505
91177308-0d34-0410-b5e6-
96231b3b80d8
Arnaud A. de Grandmaison [Wed, 10 Sep 2014 14:06:10 +0000 (14:06 +0000)]
[AArch64] Add experimental PBQP support
This adds target specific support for using the PBQP register allocator on the
AArch64, for the A57 cpu.
By default, the PBQP allocator is not used, unless explicitely required
on the command line with "-aarch64-pbqp".
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217504
91177308-0d34-0410-b5e6-
96231b3b80d8
Asiri Rathnayake [Wed, 10 Sep 2014 13:54:38 +0000 (13:54 +0000)]
[AArch 64] Use a constant pool load for weak symbol references when
using static relocation model and small code model.
Summary: currently we generate GOT based relocations for weak symbol
references regardless of the underlying relocation model. This should
be change so that in static relocation model we use a constant pool
load instead.
Patch from: Keith Walker
Reviewers: Renato Golin, Tim Northover
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217503
91177308-0d34-0410-b5e6-
96231b3b80d8
Sid Manning [Wed, 10 Sep 2014 13:09:25 +0000 (13:09 +0000)]
Add missing HWEncoding to base register class.
This change gives tblgen the information needed to fill in the
HexagonRegEncodingTable.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@217500
91177308-0d34-0410-b5e6-
96231b3b80d8