1
0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-31 07:52:55 +01:00
Commit Graph

81651 Commits

Author SHA1 Message Date
Jim Grosbach
5da9e32405 Tidy up. Test formatting.
llvm-svn: 154839
2012-04-16 18:03:14 +00:00
Akira Hatanaka
0f31530336 Do not add offset in applyFixup. This has already been accounted for in Value.
llvm-svn: 154838
2012-04-16 18:00:19 +00:00
Tobias Grosser
68d502462e Describe Polly as a 'Polyhedral Optimizer'
llvm-svn: 154834
2012-04-16 17:18:49 +00:00
Tobias Grosser
f33f842695 Add Polly news to the release notes
llvm-svn: 154833
2012-04-16 17:17:00 +00:00
Jim Grosbach
d961988871 ARM two-operand aliases for VRHADD instructions.
rdar://11252521

llvm-svn: 154832
2012-04-16 17:14:11 +00:00
Jim Grosbach
33a32d0d92 Tidy up. Testcase formatting.
llvm-svn: 154831
2012-04-16 17:14:07 +00:00
Hal Finkel
8f9dfd5a0c Link to the autovectorization EuroLLVM slides in the release notes.
llvm-svn: 154830
2012-04-16 17:06:49 +00:00
Sirish Pande
051c2d4395 Hexagon V5 (Floating Point) Support.
llvm-svn: 154829
2012-04-16 17:05:06 +00:00
Bill Wendling
a16a178883 Move to X86 directory because this fails on non-X86 platforms.
llvm-svn: 154825
2012-04-16 16:38:48 +00:00
Duncan Sands
f61d49df40 Make it possible to indicate relaxed floating point requirements at the IR level
through the use of 'fpmath' metadata.  Currently this only provides a 'fpaccuracy'
value, which may be a number in ULPs or the keyword 'fast', however the intent is
that this will be extended with additional information about NaN's, infinities
etc later.  No optimizations have been hooked up to this so far.

llvm-svn: 154822
2012-04-16 16:28:59 +00:00
Chandler Carruth
728acc9bd9 Flip the new block-placement pass to be on by default.
This is mostly to test the waters. I'd like to get results from FNT
build bots and other bots running on non-x86 platforms.

This feature has been pretty heavily tested over the last few months by
me, and it fixes several of the execution time regressions caused by the
inlining work by preventing inlining decisions from radically impacting
block layout.

I've seen very large improvements in yacr2 and ackermann benchmarks,
along with the expected noise across all of the benchmark suite whenever
code layout changes. I've analyzed all of the regressions and fixed
them, or found them to be impossible to fix. See my email to llvmdev for
more details.

I'd like for this to be in 3.1 as it complements the inliner changes,
but if any failures are showing up or anyone has concerns, it is just
a flag flip and so can be easily turned off.

I'm switching it on tonight to try and get at least one run through
various folks' performance suites in case SPEC or something else has
serious issues with it. I'll watch bots and revert if anything shows up.

llvm-svn: 154816
2012-04-16 13:49:17 +00:00
Chandler Carruth
20d3a870c6 Remove an overly brittle test. This test will no longer be interesting
once we start changing the block layout, so just nuke it. If anyone has
ideas about how to craft a code layout agnostic form of the test please
let me know.

llvm-svn: 154815
2012-04-16 13:49:09 +00:00
Duncan Sands
e0ea0f7eb6 Remove the methods for attaching metadata to instructions/retrieving metadata
from instructions.  Chandler doesn't like them being here.

llvm-svn: 154813
2012-04-16 13:44:35 +00:00
Chandler Carruth
fbb6219d5b Add a somewhat hacky heuristic to do something different from whole-loop
rotation. When there is a loop backedge which is an unconditional
branch, we will end up with a branch somewhere no matter what. Try
placing this backedge in a fallthrough position above the loop header as
that will definitely remove at least one branch from the loop iteration,
where whole loop rotation may not.

I haven't seen any benchmarks where this is important but loop-blocks.ll
tests for it, and so this will be covered when I flip the default.

llvm-svn: 154812
2012-04-16 13:33:36 +00:00
Duncan Sands
80725e26de Add convenience methods to MDBuilder for attaching metadata to instructions,
and retrieving it from instructions.  I don't have a use for this but is seems
logical for it to exist.  While there, remove some 'const' markings from methods
which are in fact 'const' in practice, but aren't logically 'const'.

llvm-svn: 154811
2012-04-16 13:21:26 +00:00
Hal Finkel
5e614e7520 Fix style violation in BBVectorize (pointed out by Bill Wendling)
llvm-svn: 154810
2012-04-16 12:39:17 +00:00
Richard Barton
9e62efdf5f Add -disassemble support for -show-inst and -show-encode capability llvm-mc. Also refactor so all MC paraphernalia are created once for all uses as much as possible.
The test change is to account for the fact that the default disassembler behaviour has changed with regards to specifying the assembly syntax to use.

llvm-svn: 154809
2012-04-16 11:32:10 +00:00
Rafael Espindola
3c1722e285 Remove lto_codegen_set_whole_program_optimization. It is a work in progress,
so we don't want it to show up in the stable 3.1 interface.

While at it, add a comment about why LTOCodeGenerator manually creates the
internalize pass.

llvm-svn: 154807
2012-04-16 10:58:38 +00:00
Chandler Carruth
33b200ad13 Tweak the loop rotation logic to check whether the loop is naturally
laid out in a form with a fallthrough into the header and a fallthrough
out of the bottom. In that case, leave the loop alone because any
rotation will introduce unnecessary branches. If either side looks like
it will require an explicit branch, then the rotation won't add any, do
it to ensure the branch occurs outside of the loop (if possible) and
maximize the benefit of the fallthrough in the bottom.

llvm-svn: 154806
2012-04-16 09:31:23 +00:00
Benjamin Kramer
a72a6005f8 Reapply 'Add reverseColor to raw_ostream'.
To be used in printing unprintable source in clang diagnostics.
Patch by Seth Cantrell, with a minor fix for mingw by me.

llvm-svn: 154805
2012-04-16 08:56:50 +00:00
Eli Bendersky
b9e9796e6f Documentation fixes to LLVMBuild.html [PR 11563]
llvm-svn: 154804
2012-04-16 08:42:55 +00:00
Argyrios Kyrtzidis
4950ffbb4f Revert r154800 which breaks windows builders.
llvm-svn: 154802
2012-04-16 07:59:39 +00:00
Craig Topper
db4fcf7088 Replace vpermd/vpermps intrinic patterns with custom lowering to target specific nodes.
llvm-svn: 154801
2012-04-16 07:13:00 +00:00
Argyrios Kyrtzidis
3d576f296a Add reverseColor to raw_ostream.
To be used in printing unprintable source in clang diagnostics.
Patch by Seth Cantrell!

llvm-svn: 154800
2012-04-16 07:07:38 +00:00
Craig Topper
a986fc78e2 Change type profile for vpermv back to using operand type for the mask argument to match intrinsic behavior. Add a bitcast to the lowering code to convert mask from v8i32 to v8f32 for vpermps.
llvm-svn: 154798
2012-04-16 06:43:40 +00:00
Craig Topper
129dccdc84 Flip the arguments when converting vpermd/vpermps intrinsics into instructions. The intrinsic has the mask as the last operand, but the instruction has it as the second.
llvm-svn: 154797
2012-04-16 06:26:15 +00:00
Bill Wendling
3514c015bd Add credit and release notes for r150307. By Kai Nacke.
llvm-svn: 154796
2012-04-16 05:24:52 +00:00
Bill Wendling
66282c6d7f Add a Fixme.
llvm-svn: 154793
2012-04-16 04:23:52 +00:00
Sebastian Pop
d60bf3baf0 add configure flag --with-default-sysroot
llvm-svn: 154791
2012-04-16 04:11:45 +00:00
Hal Finkel
4ac041693e Say something about -vectorize in the release notes.
llvm-svn: 154788
2012-04-16 03:49:43 +00:00
Hal Finkel
4f7adc1f50 Simplify checking for pointer types in BBVectorize (this change was suggested by Duncan).
llvm-svn: 154787
2012-04-16 03:49:42 +00:00
Hal Finkel
457fbe481c Remove dead SD nodes after the combining pass. Fixes PR12201.
llvm-svn: 154786
2012-04-16 03:33:22 +00:00
Chandler Carruth
fc5ab5d388 Rewrite how machine block placement handles loop rotation.
This is a complex change that resulted from a great deal of
experimentation with several different benchmarks. The one which proved
the most useful is included as a test case, but I don't know that it
captures all of the relevant changes, as I didn't have specific
regression tests for each, they were more the result of reasoning about
what the old algorithm would possibly do wrong. I'm also failing at the
moment to craft more targeted regression tests for these changes, if
anyone has ideas, it would be welcome.

The first big thing broken with the old algorithm is the idea that we
can take a basic block which has a loop-exiting successor and a looping
successor and use the looping successor as the layout top in order to
get that particular block to be the bottom of the loop after layout.
This happens to work in many cases, but not in all.

The second big thing broken was that we didn't try to select the exit
which fell into the nearest enclosing loop (to which we exit at all). As
a consequence, even if the rotation worked perfectly, it would result in
one of two bad layouts. Either the bottom of the loop would get
fallthrough, skipping across a nearer enclosing loop and thereby making
it discontiguous, or it would be forced to take an explicit jump over
the nearest enclosing loop to earch its successor. The point of the
rotation is to get fallthrough, so we need it to fallthrough to the
nearest loop it can.

The fix to the first issue is to actually layout the loop from the loop
header, and then rotate the loop such that the correct exiting edge can
be a fallthrough edge. This is actually much easier than I anticipated
because we can handle all the hard parts of finding a viable rotation
before we do the layout. We just store that, and then rotate after
layout is finished. No inner loops get split across the post-rotation
backedge because we check for them when selecting the rotation.

That fix exposed a latent problem with our exitting block selection --
we should allow the backedge to point into the middle of some inner-loop
chain as there is no real penalty to it, the whole point is that it
*won't* be a fallthrough edge. This may have blocked the rotation at all
in some cases, I have no idea and no test case as I've never seen it in
practice, it was just noticed by inspection.

Finally, all of these fixes, and studying the loops they produce,
highlighted another problem: in rotating loops like this, we sometimes
fail to align the destination of these backwards jumping edges. Fix this
by actually walking the backwards edges rather than relying on loopinfo.

This fixes regressions on heapsort if block placement is enabled as well
as lots of other cases where the previous logic would introduce an
abundance of unnecessary branches into the execution.

llvm-svn: 154783
2012-04-16 01:12:56 +00:00
Craig Topper
1b15347812 Merge vpermps/vpermd and vpermpd/vpermq SD nodes.
llvm-svn: 154782
2012-04-16 00:41:45 +00:00
Craig Topper
c217784dc3 Fix SDTypeProfile for vpermps. The mask operand should be v8i32.
llvm-svn: 154781
2012-04-16 00:12:20 +00:00
Craig Topper
e274a2cc61 Spacing fixes and 80 column fixes. Use 0 instead of 0x80 for undef indices in vpermps/vpermd. Hardware only looks at lower 3-bits.
llvm-svn: 154780
2012-04-15 23:48:57 +00:00
Craig Topper
788250eec1 Remove AVX2 vpermq and vpermpd intrinsics. These can now be handled with normal shuffle vectors.
llvm-svn: 154778
2012-04-15 22:43:31 +00:00
Craig Topper
a6f7e1a202 Make member variables of AsmToken private. Remove unnecessary forward declarations. Remove an unnecessary include.
llvm-svn: 154775
2012-04-15 22:00:22 +00:00
Jakub Staszak
64c3ee0cea Fix class name.
llvm-svn: 154773
2012-04-15 20:22:36 +00:00
Nadav Rotem
7a2a2ae678 Do not convert between fp128 <-> ppc_fp128 since there is no legal cast conversion between the two.
Patch by nobled <nobled@dreamwidth.org>

llvm-svn: 154772
2012-04-15 20:17:14 +00:00
Jakub Staszak
bde8ec16d4 Fix filename and register numbers.
llvm-svn: 154771
2012-04-15 20:13:47 +00:00
Nadav Rotem
2a4e2ef10c Fix PR12529. The Vxx family of instructions are only supported by AVX.
Use non-vex instructions for SSE4.

llvm-svn: 154770
2012-04-15 19:36:44 +00:00
Duncan Sands
f6cbb0b2cb Add the MDBuilder helper class for conveniently creating metadata.
llvm-svn: 154766
2012-04-15 18:03:49 +00:00
Benjamin Kramer
d4a8bf07d5 Wire up support for diagnostic ranges in the ARMAsmParser.
As an example, attach range info to the "invalid instruction" message:

$ clang -arch arm -c asm.c
asm.c:2:11: error: invalid instruction
  __asm__("foo r0");
          ^
<inline asm>:1:2: note: instantiated into assembly here
        foo r0
        ^~~

llvm-svn: 154765
2012-04-15 17:04:27 +00:00
Nadav Rotem
b8710ee43f When emulating vselect using OR/AND/XOR make sure to bitcast the result back to the original type.
llvm-svn: 154764
2012-04-15 15:08:09 +00:00
Elena Demikhovsky
92fb3e613e Added VPERM optimization for AVX2 shuffles
llvm-svn: 154761
2012-04-15 11:18:59 +00:00
NAKAMURA Takumi
0133680b3d HexagonCopyToCombine.cpp: Silence two warnings, -Wunused-variable, with -Asserts.
llvm-svn: 154759
2012-04-15 05:33:43 +00:00
NAKAMURA Takumi
ddf2dc407e Target/Hexagon: Tweak to fix msvc build.
llvm-svn: 154758
2012-04-15 05:09:09 +00:00
Anshuman Dasgupta
68108fede2 Remove trailing whitespace.
llvm-svn: 154755
2012-04-14 20:59:13 +00:00
Anshuman Dasgupta
9c02269a1c Add VLIW packetizer to ReleaseNotes.html and CREDITS.TXT. Committing patch
by Sundeep Kushwaha.

llvm-svn: 154754
2012-04-14 20:57:13 +00:00