1
0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-29 23:12:55 +01:00
Commit Graph

96056 Commits

Author SHA1 Message Date
Juergen Ributzka
d4aac28519 Fix the buildbot
llvm-svn: 191133
2013-09-21 05:15:01 +00:00
Juergen Ributzka
32cca125e1 [X86] Emulate AVX 256bit MIN/MAX support by splitting the vector.
In AVX 256bit vectors are valid vectors and therefore the Type Legalizer doesn't
split the VSELECT and SETCC nodes. AVX only supports MIN/MAX on 128bit vectors
and this fix enables vector splitting for this special case in the X86 DAG
Combiner.

This fix is related to PR16695, PR17002, and <rdar://problem/14594431>.

llvm-svn: 191131
2013-09-21 04:55:22 +00:00
Juergen Ributzka
67e5289ff2 SelectionDAG: Teach the legalizer to split SETCC if VSELECT needs splitting too.
The Type Legalizer recognizes that VSELECT needs to be split, because the type
is to wide for the given target. The same does not always apply to SETCC,
because less space is required to encode the result of a comparison. As a result
VSELECT is split and SETCC is unrolled into scalar comparisons.

This commit fixes the issue by checking for VSELECT-SETCC patterns in the DAG
Combiner. If a matching pattern is found, then the result mask of SETCC is
promoted to the expected vector mask for the given target. This mask has usually
te same size as the VSELECT return type (except for Intel KNL). Now the type
legalizer will split both VSELECT and SETCC.

This allows the following X86 DAG Combine code to sucessfully detect the MIN/MAX
pattern. This fixes PR16695, PR17002, and <rdar://problem/14594431>.

llvm-svn: 191130
2013-09-21 04:55:18 +00:00
NAKAMURA Takumi
81025ea333 Initialize BSSSection explicitly in InitMachOMCObjectFileInfo() to appease msvc.
This can revert r191087.

llvm-svn: 191128
2013-09-21 02:34:45 +00:00
Reed Kotler
c8c27affb8 Set .reorder for the stub so that gas takes care of delay slot processing.
llvm-svn: 191125
2013-09-21 01:37:52 +00:00
Arnold Schwaighofer
c1f8473eb6 Reapply "SLPVectorizer: Handle more horizontal reductions (disabled)""
Reapply r191108 with a fix for a memory corruption error I introduced.  Of
course, we can't reference the scalars that we replace by vectorizing and then
call their eraseFromParent method. I only 'needed' the scalars to get the
DebugLoc. Just store the DebugLoc before actually vectorizing instead. As a nice
side effect, this also simplifies the interface between BoUpSLP and the
HorizontalReduction class to returning a value pointer (the vectorized tree
root).

radar://14607682

llvm-svn: 191123
2013-09-21 01:06:00 +00:00
Nadav Rotem
a10aa3ffa1 LoopVectorizer: Only allow vectorization of intrinsics. We can't know for sure that the functions 'abs' or 'round' are the functions from libm.
rdar://15012650

llvm-svn: 191122
2013-09-21 00:27:05 +00:00
Arnold Schwaighofer
c81e140238 Revert "SLPVectorizer: Handle more horizontal reductions (disabled)"
This reverts commit r191108.

The horizontal.ll test case fails under libgmalloc. Thanks Shuxin for pointing
this out to me.

llvm-svn: 191121
2013-09-21 00:06:20 +00:00
Eric Christopher
a092932bd3 Move emission of the debug string table to early in the debug
info finalization to greatly reduce the number of fixups that the
assembler has to handle in order to improve compile time.

llvm-svn: 191119
2013-09-20 23:22:52 +00:00
Shuxin Yang
640fbb44a5 Resurrect r191017 " GVN proceeds in the presence of dead code" plus a fix to PR17307 & 17308.
The problem of r191017 is that when GVN fabricate a val-number for a dead instruction (in order
to make following expr-PRE happy), it forget to fabricate a leader-table entry for it as well.

llvm-svn: 191118
2013-09-20 23:12:57 +00:00
Jim Grosbach
cf9257a3a6 MC: Tidy up.
Clean up some simple code quality issues. Bring internal naming
conventions up to current standard, fix inconsistent formatting, and
tidy up a couple of odd contructs.

llvm-svn: 191117
2013-09-20 23:08:21 +00:00
Eric Christopher
6ef1d748cb Migrate addGlobalName to the .cpp file as an intermediate step
to further work.

llvm-svn: 191113
2013-09-20 22:20:55 +00:00
Benjamin Kramer
89ff6bf9f0 InstCombine: Remove unused argument. No functionality change.
llvm-svn: 191112
2013-09-20 22:12:42 +00:00
Andrew Kaylor
6360b74df8 Fix some more MCJIT PIC test XFAILs (for i386)
Patch by Dimitry Andric

llvm-svn: 191111
2013-09-20 22:12:03 +00:00
Akira Hatanaka
449ced08a3 [mips] MUL should clobber HI0 and LO0.
I cannot think of a test case that reliably triggers this bug.

llvm-svn: 191109
2013-09-20 21:22:28 +00:00
Arnold Schwaighofer
db78859615 SLPVectorizer: Handle more horizontal reductions (disabled)
Match reductions starting at binary operation feeding into a phi. The code
handles trees like

 r += v1 + v2 + v3 ...

and

 r += v1
 r += v2
 ...

and

 r *= v1 + v2 + ...

We currently only handle associative operations (add, fadd fast).

The code can now also handle reductions feeding into stores.

 a[i] = v1 + v2 + v3 + ...

The code is currently disabled behind the flag "-slp-vectorize-hor".  The cost
model for most architectures is not there yet.

I found one opportunity of a horizontal reduction feeding a phi in TSVC
(LoopRerolling-flt) and there are several opportunities where reductions feed
into stores.

radar://14607682

llvm-svn: 191108
2013-09-20 21:18:20 +00:00
Joerg Sonnenberger
ed165ebb1f Delete empty files.
llvm-svn: 191105
2013-09-20 20:40:22 +00:00
Joerg Sonnenberger
d4339cb110 Revert r191017, it results in segmentation faults in Qt.
llvm-svn: 191104
2013-09-20 20:33:57 +00:00
Tareq A. Siraj
da412cbe5f Fixed typo in CreateProcessTrailingSlash test
--gtest_filter was filtering an invalid name for the test.

Differential Revision: http://llvm-reviews.chandlerc.com/D1733

llvm-svn: 191100
2013-09-20 18:21:59 +00:00
Benjamin Kramer
7ea950a209 InstCombine: Canonicalize (gep i8* X, -(ptrtoint Y)) to (sub (ptrtoint X), (ptrtoint Y))
The GEP pattern is what SCEV expander emits for "ugly geps". The latter is what
you get for pointer subtraction in C code. The rest of instcombine already
knows how to deal with that so just canonicalize on that.

llvm-svn: 191090
2013-09-20 14:38:44 +00:00
NAKAMURA Takumi
98762f0455 llvm/tools/Makefile: Suppress building llvm-lto on cygming, for now, probably due to LTO.dll.
llvm-svn: 191088
2013-09-20 13:12:24 +00:00
NAKAMURA Takumi
69102d22a5 llvm/test: Mark 3 tests as XFAIL:msvc.
llvm-svn: 191087
2013-09-20 12:57:34 +00:00
NAKAMURA Takumi
5f50cb98e9 test/tools/lto/lit.local.cfg: Suppress llvm-lto test on Cygwin and Win32, for now.
llvm-svn: 191086
2013-09-20 12:57:25 +00:00
Benjamin Kramer
0d760b7cc7 Move x86-dependent tests into the right directory.
llvm-svn: 191082
2013-09-20 10:56:40 +00:00
Anders Waldenborg
70a0f09515 Revert "llvm-c: Add LLVMGetPointerToFunction"
This reverts r191030

llvm-svn: 191075
2013-09-20 07:00:36 +00:00
Craig Topper
6670549a2c Lift alignment restrictions on load/store folding of VEXTRACTI128/VINSERTI128.
llvm-svn: 191073
2013-09-20 05:37:49 +00:00
Andrew Trick
439918c874 Allow subtarget selection of the default MachineScheduler and document the interface.
The global registry is used to allow command line override of the
scheduler selection, but does not work well as the normal selection
API. For example, the same LLVM process should be able to target
multiple targets or subtargets.

llvm-svn: 191071
2013-09-20 05:14:41 +00:00
Richard Smith
6dfe60433d Revert r191062; the build break was also fixed in a different (incompatible) way in r191060.
llvm-svn: 191065
2013-09-20 01:24:10 +00:00
Richard Smith
255a969b21 Unbreak Clang build after r191050: don't pass a StringRef to snprintf.
llvm-svn: 191062
2013-09-20 00:38:18 +00:00
David Blaikie
8dd7f9c9a7 DebugInfo: GDBIndexEntry*String conversion functions now return const char* for easy llvm::formating
This was previously invoking UB by passing a user-defined type to
format. Thanks to Jordan Rose for pointing this out.

llvm-svn: 191060
2013-09-20 00:33:15 +00:00
David Blaikie
7e8457a199 Add braces to suppress Clang's dangling-else warning.
These violations were introduced in r191049

llvm-svn: 191059
2013-09-20 00:33:11 +00:00
David Blaikie
954130581b DebugInfo: constrain gnu pubnames test further
Ensures that the pubnames entries actually refer to the intended
entities. This test could be more flexible if there was a way to do
multiline FileCheck matches with captures (in that way the test wouldn't
need to have hardcoded offset values and would thus be resilient to
changes in the layout of the DIEs in this CU).

llvm-svn: 191055
2013-09-19 23:43:46 +00:00
Richard Mitton
f43e48210e Fixed warning
llvm-svn: 191053
2013-09-19 23:21:07 +00:00
Richard Mitton
fe7ffd01f6 Added support for generate DWARF .debug_aranges sections automatically.
llvm-svn: 191052
2013-09-19 23:21:01 +00:00
Andrew Trick
6fd8b52af9 Rename ConvergingScheduler to GenericScheduler.
This was an experimental scheduler a year ago. It's now used by
several subtargets, both in-order and out-of-order, and it
is about to be enabled by default for x86 and armv7. It will be the
new GenericScheduler for subtargets that don't provide their own
SchedulingStrategy.

llvm-svn: 191051
2013-09-19 23:10:59 +00:00
David Blaikie
f60a2aeec3 DebugInfo: llvm-dwarfdump support for gnu_pubnames section
llvm-svn: 191050
2013-09-19 23:01:29 +00:00
Kai Nacke
64a3fccd60 PR16726: extend rol/ror matching
C-like languages promote types like unsigned short to unsigned int before
performing an arithmetic operation. Currently the rotate matcher in the
DAGCombiner does not consider this situation.

This commit extends the DAGCombiner in the way that the pattern

(or (shl ([az]ext x), (*ext y)), (srl ([az]ext x), (*ext (sub 32, y))))

is folded into

([az]ext (rotl x, y))

The matching is restricted to aext and zext because in this cases the upper
bits are either undefined or known. Test case is included.

This fixes PR16726.

llvm-svn: 191049
2013-09-19 23:00:28 +00:00
Kai Nacke
21c0476931 Revert PR16726: extend rol/ror matching
There is a buildbot failure. Need to investigate this.

llvm-svn: 191048
2013-09-19 22:53:36 +00:00
Kai Nacke
fe02753846 PR16726: extend rol/ror matching
C-like languages promote types like unsigned short to unsigned int before
performing an arithmetic operation. Currently the rotate matcher in the
DAGCombiner does not consider this situation.

This commit extends the DAGCombiner in the way that the pattern

(or (shl ([az]ext x), (*ext y)), (srl ([az]ext x), (*ext (sub 32, y))))

is folded into

([az]ext (rotl x, y))

The matching is restricted to aext and zext because in this cases the upper
bits are either undefined or known. Test case is included.

This fixes PR16726.

llvm-svn: 191045
2013-09-19 22:36:39 +00:00
David Blaikie
7d5a020241 DebugInfo: Improve IR annotation comments for GNU pubthings.
llvm-svn: 191043
2013-09-19 22:19:37 +00:00
Peter Collingbourne
2a807b9eca Fix LTO handling of module-level assembly (PR14152).
Patch by Tom Roeder!

llvm-svn: 191042
2013-09-19 22:15:52 +00:00
Bill Wendling
01837a249d Add testcase to make sure we don't generate too many jumps for a une compare.
<rdar://problem/7859988>

llvm-svn: 191040
2013-09-19 21:58:20 +00:00
Benjamin Kramer
824049fa6e Unbreak C++03 build.
llvm-svn: 191039
2013-09-19 21:30:00 +00:00
Shuxin Yang
44b8e62121 [Fast-math] Disable "(C1/X)*C2 => (C1*C2)/X" if C1/X has multiple uses.
If "C1/X" were having multiple uses, the only benefit of this
transformation is to potentially shorten critical path. But it is at the
cost of instroducing additional div.

  The additional div may or may not incur cost depending on how div is
implemented. If it is implemented using Newton–Raphson iteration, it dosen't
seem to incur any cost (FIXME). However, if the div blocks the entire
pipeline, that sounds to be pretty expensive. Let CodeGen to take care 
this transformation.

  This patch sees 6% on a benchmark.

rdar://15032743

llvm-svn: 191037
2013-09-19 21:13:46 +00:00
Benjamin Kramer
b681a45715 InstCombine: Don't allow turning vector-of-pointer loads into vector-of-integer.
The code below can't handle any pointers. PR17293.

llvm-svn: 191036
2013-09-19 20:59:04 +00:00
David Blaikie
a8dfbd9e12 Unshift the GDB index/GNU pubnames constants modified in r191025
Based on code review feedback from Eric Christopher, unshifting these
constants as they can appear in the gdb_index itself, shifted a further
24 bits. This means that keeping them preshifted is a bit inflexible, so
let's not do that.

Given the motivation, wrap up some nicer enums, more type safety, and
some utility functions.

llvm-svn: 191035
2013-09-19 20:40:26 +00:00
Anders Waldenborg
c34c43d63b llvm-c: Add LLVMGetPointerToFunction
Differential Revision: http://llvm-reviews.chandlerc.com/D1715

llvm-svn: 191030
2013-09-19 19:55:06 +00:00
Anders Waldenborg
ce324aaf69 llvm-c: Make LLVMGetFirstTarget a proper prototype
This avoids warnings when included in a application that
uses -Wstrict-prototypes.

Differential Revision: http://llvm-reviews.chandlerc.com/D1713

llvm-svn: 191029
2013-09-19 19:43:55 +00:00
Eric Christopher
61eb11007b Remove extraneous space, the asm printing infrastructure adds a space
in normally.

llvm-svn: 191026
2013-09-19 18:41:40 +00:00
David Blaikie
d4db76886f DebugInfo: Simplify gnu_pubnames index computation.
Names open to bikeshedding. Could switch back to the constants being
unshifted, but this way seems a bit easier to work with.

llvm-svn: 191025
2013-09-19 18:39:59 +00:00