Jim Grosbach [Mon, 9 Nov 2009 00:11:35 +0000 (00:11 +0000)]
Use Unified Assembly Syntax for the ARM backend.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86494
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Sun, 8 Nov 2009 22:36:55 +0000 (22:36 +0000)]
Fix an issue where the ordering of blocks within a function could lead to different constraint
graphs being produced. The cause was that we were incorrectly marking sigma instructions as
processed after handling the sigma-specific constraints for them, potentially neglecting to
process them as normal instructions as well.
Unfortunately, the testcase that inspired this still doesn't work because of a bug in the solver,
which is next on the list to debug.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86486
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sun, 8 Nov 2009 21:51:53 +0000 (21:51 +0000)]
Add a 'zkill' script, which is more-or-less a fancy (although not necessarily
very robust) version of killall. Because I like making shiny new wheels out of
spare parts.
For use by buildbots when people insist on making cc1 infinite loop. :)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86484
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 8 Nov 2009 21:20:06 +0000 (21:20 +0000)]
Teach an instcombine to not pull trunc instructions through PHI nodes
when both the source and dest are illegal types, since it would cause
the phi to grow (for example, we shouldn't transform test14b's phi to
a phi on i320). This fixes an infinite loop on i686 bootstrap with
phi slicing turned on, so turn it back on.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86483
91177308-0d34-0410-b5e6-
96231b3b80d8
Duncan Sands [Sun, 8 Nov 2009 20:55:48 +0000 (20:55 +0000)]
Revert commit 81144, and add a comment. It caused bugpoint timeouts
not to work any more on linux.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86481
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 8 Nov 2009 19:23:30 +0000 (19:23 +0000)]
reapply r8644[3-5] with only the scary part
(SliceUpIllegalIntegerPHI) disabled.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86480
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sun, 8 Nov 2009 17:52:47 +0000 (17:52 +0000)]
Speculatively revert r8644[3-5], they seem to be leading to infinite loops in
llvm-gcc bootstrap.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86478
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Sun, 8 Nov 2009 15:33:12 +0000 (15:33 +0000)]
Add and-not (bic) patterns. Based heavily on patch by Brian Lucas!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86471
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Sun, 8 Nov 2009 15:32:44 +0000 (15:32 +0000)]
Move OR patterns upper to all logical stuff. No functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86470
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Sun, 8 Nov 2009 15:32:28 +0000 (15:32 +0000)]
Some nice peephole patterns. Based on patch by Brian Lucas!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86469
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Sun, 8 Nov 2009 15:32:11 +0000 (15:32 +0000)]
Print tab before operand of jcc
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86468
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Sun, 8 Nov 2009 14:27:38 +0000 (14:27 +0000)]
Fix invalid operand updates & implement post-inc memory operands
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86466
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Sun, 8 Nov 2009 12:58:40 +0000 (12:58 +0000)]
Throw an error when stack realignment stuff fails instead of silent
code miscompilation
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86463
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Sun, 8 Nov 2009 12:14:54 +0000 (12:14 +0000)]
It is invalid to infer the value type from the result #0 of the node
since the instruction might use the other result of different type.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86462
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sun, 8 Nov 2009 09:46:57 +0000 (09:46 +0000)]
Remove ByteswapSCANFResults, it is dead.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86458
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sun, 8 Nov 2009 09:34:14 +0000 (09:34 +0000)]
NNT: Remove DejaGNU test from NewNightlyTest reports, this aspect of testing is
handled by buildbots now.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86454
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sun, 8 Nov 2009 09:29:52 +0000 (09:29 +0000)]
Two small fixes for site.exp for cmake.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86453
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sun, 8 Nov 2009 09:08:00 +0000 (09:08 +0000)]
Derive the right paths to use during testing instead of passing it in via make.
Also, fix a few other details of the cmake test target and rename it to
'check'. CMake tests now work for the most part, but there are a handful of
failures left due to missing site.exp bits.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86452
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sun, 8 Nov 2009 09:07:51 +0000 (09:07 +0000)]
Switch to using 'lit.site.cfg.in' for the site config template for Unit tests,
and generate it for CMake builds as well.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86451
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sun, 8 Nov 2009 09:07:42 +0000 (09:07 +0000)]
Cleanup some unused RUN lines.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86450
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sun, 8 Nov 2009 09:07:33 +0000 (09:07 +0000)]
lit: Hardcode whence seek value, os.SEEK_END isn't always available.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86449
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sun, 8 Nov 2009 09:07:26 +0000 (09:07 +0000)]
lit: Warn when a test suite contains no tests.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86448
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sun, 8 Nov 2009 09:07:13 +0000 (09:07 +0000)]
lit: Drop require_and_and support.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86447
91177308-0d34-0410-b5e6-
96231b3b80d8
Lang Hames [Sun, 8 Nov 2009 08:49:59 +0000 (08:49 +0000)]
Moved some ManagedStatics out of the SlotIndexes header.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86446
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 8 Nov 2009 08:36:40 +0000 (08:36 +0000)]
another more interesting test.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86445
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 8 Nov 2009 08:30:58 +0000 (08:30 +0000)]
feature test for the new transformation in r86443
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86444
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 8 Nov 2009 08:21:13 +0000 (08:21 +0000)]
teach a couple of instcombine transformations involving PHIs to
not turn a PHI in a legal type into a PHI of an illegal type, and
add a new optimization that breaks up insane integer PHI nodes into
small pieces (PR3451).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86443
91177308-0d34-0410-b5e6-
96231b3b80d8
Nick Lewycky [Sun, 8 Nov 2009 05:45:04 +0000 (05:45 +0000)]
We don't need to byteswap, the interpreter assumes the program is running
native anyways. This fixes a crash using %d and similar in a scanf statement.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86440
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sun, 8 Nov 2009 03:43:06 +0000 (03:43 +0000)]
lit: Workaround a Win32/subprocess bug when appending.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86437
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sun, 8 Nov 2009 03:35:19 +0000 (03:35 +0000)]
lit: Preserve the PATHEXT variable when running subcommands, this is important on Win32
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86436
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 8 Nov 2009 02:32:01 +0000 (02:32 +0000)]
Make TargetData::getStringRepresentation spit out native integer types,
this gives llvm-gcc generated modules the right data.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86435
91177308-0d34-0410-b5e6-
96231b3b80d8
Nick Lewycky [Sun, 8 Nov 2009 02:23:15 +0000 (02:23 +0000)]
Remove test. Execution tests are slow and generally not worth it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86434
91177308-0d34-0410-b5e6-
96231b3b80d8
Nick Lewycky [Sun, 8 Nov 2009 01:04:45 +0000 (01:04 +0000)]
Fix run line.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86429
91177308-0d34-0410-b5e6-
96231b3b80d8
Nick Lewycky [Sun, 8 Nov 2009 00:45:29 +0000 (00:45 +0000)]
Fix the interpreter to not crash due to zeroext/signext
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86428
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sun, 8 Nov 2009 00:34:22 +0000 (00:34 +0000)]
Prevent warning spew about -fPIC when using CMake generated Xcode project files.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86427
91177308-0d34-0410-b5e6-
96231b3b80d8
Jim Grosbach [Sun, 8 Nov 2009 00:27:19 +0000 (00:27 +0000)]
Use aligned load/store instructions for spilling Q registers when we know the stack slot is 128 bit aligned
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86425
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Sun, 8 Nov 2009 00:15:23 +0000 (00:15 +0000)]
Refactor code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86423
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sat, 7 Nov 2009 23:52:27 +0000 (23:52 +0000)]
Fix CMake reporting of target triple.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86419
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sat, 7 Nov 2009 23:52:20 +0000 (23:52 +0000)]
Stop running get_target_triple more than we need to.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86418
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sat, 7 Nov 2009 23:51:55 +0000 (23:51 +0000)]
Fix MSVC warning ( | with bool and unsigned int).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86417
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Sat, 7 Nov 2009 23:21:30 +0000 (23:21 +0000)]
Fix class -> struct tag.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86416
91177308-0d34-0410-b5e6-
96231b3b80d8
Nate Begeman [Sat, 7 Nov 2009 23:17:15 +0000 (23:17 +0000)]
x86 vector shuffle cleanup/fixes:
1. rename the movhp patfrag to movlhps, since thats what it actually matches
2. eliminate the bogus movhps load and store patterns, they were incorrect. The load transforms are already handled (correctly) by shufps/unpack.
3. revert a recent test change to its correct form.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86415
91177308-0d34-0410-b5e6-
96231b3b80d8
Jim Grosbach [Sat, 7 Nov 2009 22:00:39 +0000 (22:00 +0000)]
80-column cleanup of file header comments
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86408
91177308-0d34-0410-b5e6-
96231b3b80d8
Jim Grosbach [Sat, 7 Nov 2009 21:25:39 +0000 (21:25 +0000)]
Support alignment specifier for NEON vld/vst instructions
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86404
91177308-0d34-0410-b5e6-
96231b3b80d8
Nick Lewycky [Sat, 7 Nov 2009 21:10:15 +0000 (21:10 +0000)]
Improve tail call elimination to handle the switch statement.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86403
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Sat, 7 Nov 2009 19:40:04 +0000 (19:40 +0000)]
t2ldrpci_pic can be used for blockaddress as well.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86400
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 7 Nov 2009 19:13:17 +0000 (19:13 +0000)]
temporarily remove these tests, as they are breaking in the buildbot,
Eric, please investigate.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86399
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 7 Nov 2009 19:11:46 +0000 (19:11 +0000)]
make instcombine only rewrite a chain of computation
(eliminating some extends) if the new type of the
computation is legal or if both the source and dest
are illegal. This prevents instcombine from changing big
chains of computation into i64 on 32-bit targets for
example.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86398
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 7 Nov 2009 19:07:32 +0000 (19:07 +0000)]
indicate what the native integer types for the target are.
Please verify.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86397
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 7 Nov 2009 18:53:00 +0000 (18:53 +0000)]
all targets should be required to declare legal integer types. My plan to
make it optional doesn't work out. If you don't want to specify this, don't
specify a TD string at all.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86394
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 7 Nov 2009 18:03:32 +0000 (18:03 +0000)]
remove empty files.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86392
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 7 Nov 2009 17:59:32 +0000 (17:59 +0000)]
Revert r86359, it is breaking the self host on the
llvm-gcc-i386-darwin9 build bot.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86391
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Sat, 7 Nov 2009 17:15:25 +0000 (17:15 +0000)]
First try of the post-inc operands handling... Not fully worked, though :(
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86386
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Sat, 7 Nov 2009 17:15:06 +0000 (17:15 +0000)]
Add some dummy support for post-incremented loads
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86385
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Sat, 7 Nov 2009 17:14:39 +0000 (17:14 +0000)]
Add 8 bit libcalls and make use of them for msp430
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86384
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Sat, 7 Nov 2009 17:13:57 +0000 (17:13 +0000)]
Add few pseudo-source-values
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86383
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Sat, 7 Nov 2009 17:13:35 +0000 (17:13 +0000)]
Initial support for addrmode handling. Tests by Brian Lucas!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86382
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Sat, 7 Nov 2009 17:12:58 +0000 (17:12 +0000)]
Some preliminary variable asmprinting
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86381
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Sat, 7 Nov 2009 17:12:38 +0000 (17:12 +0000)]
Use '.L' for global private prefix (as mspgcc)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86380
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Sat, 7 Nov 2009 17:12:21 +0000 (17:12 +0000)]
Drop old asmprinter stuff
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86379
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Sat, 7 Nov 2009 15:20:32 +0000 (15:20 +0000)]
It turns out that the testcase in question uncovered subreg-handling bug.
Add assert in asmprinter to catch such cases and xfail the tests.
PR is to be filled.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86375
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 7 Nov 2009 09:35:34 +0000 (09:35 +0000)]
add the ability for TargetData to return information about legal integer
datatypes on a given CPU. This is intended to allow instcombine and other
transformations to avoid converting big sequences of operations to an
inconvenient width, and will help clean up after SRoA. See also "Adding
legal integer sizes to TargetData" on Feb 1, 2009 on llvmdev, and PR3451.
Comments welcome.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86370
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 7 Nov 2009 09:23:04 +0000 (09:23 +0000)]
more cleanup.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86369
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 7 Nov 2009 09:20:54 +0000 (09:20 +0000)]
add some missing #includes
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86367
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 7 Nov 2009 09:13:23 +0000 (09:13 +0000)]
rewrite TargetData to use StringRef/raw_ostream instead of thrashing std::strings.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86366
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 7 Nov 2009 09:07:01 +0000 (09:07 +0000)]
prune #include / layering violation
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86365
91177308-0d34-0410-b5e6-
96231b3b80d8
Jeffrey Yasskin [Sat, 7 Nov 2009 08:51:52 +0000 (08:51 +0000)]
Make the need-stub variables accurate and consistent. In the case of
MachineRelocations, "stub" always refers to a far-call stub or a
load-a-faraway-global stub, so this patch adds "Far" to the term. (Other stubs
are used for lazy compilation and dlsym address replacement.) The variable was
also inconsistent between the positive and negative sense, and the positive
sense ("NeedStub") was more demanding than is accurate (since a nearby-enough
function can be called directly even if the platform often requires a stub).
Since the negative sense causes double-negatives, I switched to
"MayNeedFarStub" globally.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86363
91177308-0d34-0410-b5e6-
96231b3b80d8
Eric Christopher [Sat, 7 Nov 2009 08:45:53 +0000 (08:45 +0000)]
Fix a couple of shuffle patterns to use movhlps instead
of movhps as the constraint. Changes optimizations so
update testcases as appropriate as well.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86360
91177308-0d34-0410-b5e6-
96231b3b80d8
Nick Lewycky [Sat, 7 Nov 2009 08:34:40 +0000 (08:34 +0000)]
Teach dead store elimination that certain intrinsics write to memory just like
a store.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86359
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 7 Nov 2009 08:31:52 +0000 (08:31 +0000)]
remove the win32 tree, it's stale and confusing.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86358
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 7 Nov 2009 08:05:03 +0000 (08:05 +0000)]
reapply 86289, 86278, 86270, 86267, 86266 & 86264 plus a fix
(making pred factoring only happen if threading is guaranteed
to be successful).
This now survives an X86-64 bootstrap of llvm-gcc.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86355
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 7 Nov 2009 07:50:34 +0000 (07:50 +0000)]
Fix PR5421 by APInt'izing switch lowering.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86354
91177308-0d34-0410-b5e6-
96231b3b80d8
Nick Lewycky [Sat, 7 Nov 2009 07:42:38 +0000 (07:42 +0000)]
Oops, FunctionContainsEscapingAllocas is really used to mean two different
things. Back out part of r86349 for a moment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86353
91177308-0d34-0410-b5e6-
96231b3b80d8
Nick Lewycky [Sat, 7 Nov 2009 07:10:01 +0000 (07:10 +0000)]
Dust off tail recursion elimination. Fix a fixme by applying CaptureTracking
and add a .ll to demo the new capability.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86349
91177308-0d34-0410-b5e6-
96231b3b80d8
Mikhail Glushenkov [Sat, 7 Nov 2009 06:33:58 +0000 (06:33 +0000)]
llvmc: Add a '-time' option.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86348
91177308-0d34-0410-b5e6-
96231b3b80d8
Mikhail Glushenkov [Sat, 7 Nov 2009 06:33:12 +0000 (06:33 +0000)]
Trailing whitespace.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86347
91177308-0d34-0410-b5e6-
96231b3b80d8
Mikhail Glushenkov [Sat, 7 Nov 2009 06:33:01 +0000 (06:33 +0000)]
80-col violation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86346
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 7 Nov 2009 06:19:20 +0000 (06:19 +0000)]
merge cmp1 into cmp0 and filecheckize.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86345
91177308-0d34-0410-b5e6-
96231b3b80d8
Lang Hames [Sat, 7 Nov 2009 05:50:28 +0000 (05:50 +0000)]
Update some globals to use ManagedStatic.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86342
91177308-0d34-0410-b5e6-
96231b3b80d8
Mon P Wang [Sat, 7 Nov 2009 04:46:25 +0000 (04:46 +0000)]
Fix memoizing of CvtRndSatSDNode
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86340
91177308-0d34-0410-b5e6-
96231b3b80d8
Mon P Wang [Sat, 7 Nov 2009 04:07:33 +0000 (04:07 +0000)]
Fixed Overload table bug noticed by Jakob
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86332
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Sat, 7 Nov 2009 04:07:30 +0000 (04:07 +0000)]
Missed this.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86331
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Sat, 7 Nov 2009 04:04:34 +0000 (04:04 +0000)]
Refactor code. Fix a potential missing check. Teach isIdentical() about tLDRpci_pic.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86330
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Sat, 7 Nov 2009 03:52:02 +0000 (03:52 +0000)]
- Add TargetInstrInfo::isIdentical(). It's similar to MachineInstr::isIdentical
except it doesn't care if the definitions' virtual registers differ. This is
used by machine LICM and other MI passes to perform CSE.
- Teach Thumb2InstrInfo::isIdentical() to check two t2LDRpci_pic are identical.
Since pc relative constantpool entries are always different, this requires it
it check if the values can actually the same.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86328
91177308-0d34-0410-b5e6-
96231b3b80d8
Ted Kremenek [Sat, 7 Nov 2009 03:26:59 +0000 (03:26 +0000)]
Update CMake file.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86325
91177308-0d34-0410-b5e6-
96231b3b80d8
Kenneth Uildriks [Sat, 7 Nov 2009 02:11:54 +0000 (02:11 +0000)]
Add code to check at SelectionDAGISel::LowerArguments time to see if return values can be lowered to registers. Coming soon, code to perform sret-demotion if return values cannot be lowered to registers
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86324
91177308-0d34-0410-b5e6-
96231b3b80d8
Jakob Stoklund Olesen [Sat, 7 Nov 2009 01:58:40 +0000 (01:58 +0000)]
Fix inverted conflict test in -early-coalesce.
A non-identity copy cannot be coalesced when the phi join destination register
is live at the copy site.
Also verify the condition that the PHI join source register is only used in
the PHI join. Otherwise the coalescing is invalid.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86322
91177308-0d34-0410-b5e6-
96231b3b80d8
Devang Patel [Sat, 7 Nov 2009 01:32:59 +0000 (01:32 +0000)]
Revert following patches to fix llvmgcc bootstrap.
86289, 86278, 86270, 86267, 86266 & 86264
Chris, please take a look.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86321
91177308-0d34-0410-b5e6-
96231b3b80d8
Johnny Chen [Sat, 7 Nov 2009 00:54:36 +0000 (00:54 +0000)]
My previous patch (r84124) for setting the encoding bits 4 and 7 of DPSoRegFrm
was wrong and too aggressive in the sense that DPSoRegFrm includes both constant
shifts (with Inst{4} = 0) and register controlled shifts (with Inst{4} = 1 and
Inst{7} = 0). The 'rr' fragment of the multiclass definitions actually means
register/register with no shift, see A8-11.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86319
91177308-0d34-0410-b5e6-
96231b3b80d8
Victor Hernandez [Sat, 7 Nov 2009 00:41:19 +0000 (00:41 +0000)]
- new SROA mallocs should have the mallocs running-or'ed, not the malloc's bitcast
- fix ProcessInternalGlobal() debug output
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86317
91177308-0d34-0410-b5e6-
96231b3b80d8
Victor Hernandez [Sat, 7 Nov 2009 00:36:50 +0000 (00:36 +0000)]
Fit in 80 columns
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86316
91177308-0d34-0410-b5e6-
96231b3b80d8
Jeffrey Yasskin [Sat, 7 Nov 2009 00:26:47 +0000 (00:26 +0000)]
Avoid "ambiguous 'else'" warning from gcc.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86314
91177308-0d34-0410-b5e6-
96231b3b80d8
Victor Hernandez [Sat, 7 Nov 2009 00:16:28 +0000 (00:16 +0000)]
Re-commit r86077 now that r86290 fixes the 179.art and 175.vpr ARM regressions.
Here is the original commit message:
This commit updates malloc optimizations to operate on malloc calls that have constant int size arguments.
Update CreateMalloc so that its callers specify the size to allocate:
MallocInst-autoupgrade users use non-TargetData-computed allocation sizes.
Optimization uses use TargetData to compute the allocation size.
Now that malloc calls can have constant sizes, update isArrayMallocHelper() to use TargetData to determine the size of the malloced type and the size of malloced arrays.
Extend getMallocType() to support malloc calls that have non-bitcast uses.
Update OptimizeGlobalAddressOfMalloc() to optimize malloc calls that have non-bitcast uses. The bitcast use of a malloc call has to be treated specially here because the uses of the bitcast need to be replaced and the bitcast needs to be erased (just like the malloc call) for OptimizeGlobalAddressOfMalloc() to work correctly.
Update PerformHeapAllocSRoA() to optimize malloc calls that have non-bitcast uses. The bitcast use of the malloc is not handled specially here because ReplaceUsesOfMallocWithGlobal replaces through the bitcast use.
Update OptimizeOnceStoredGlobal() to not care about the malloc calls' bitcast use.
Update all globalopt malloc tests to not rely on autoupgraded-MallocInsts, but instead use explicit malloc calls with correct allocation sizes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86311
91177308-0d34-0410-b5e6-
96231b3b80d8
Jim Grosbach [Sat, 7 Nov 2009 00:13:30 +0000 (00:13 +0000)]
80-columns
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86310
91177308-0d34-0410-b5e6-
96231b3b80d8
Jeffrey Yasskin [Sat, 7 Nov 2009 00:00:10 +0000 (00:00 +0000)]
Give the JITResolver a direct pointer to its JITEmitter, and use that instead
of going through the global TheJIT variable. This makes it easier to use
features of JITEmitter that aren't in JITCodeEmitter for fixing PR5201.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86305
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Fri, 6 Nov 2009 23:52:48 +0000 (23:52 +0000)]
- Add pseudo instructions tLDRpci_pic and t2LDRpci_pic which does a pc-relative
load of a GV from constantpool and then add pc. It allows the code sequence to
be rematerializable so it would be hoisted by machine licm.
- Add a late pass to break these pseudo instructions into a number of real
instructions. Also move the code in Thumb2 IT pass that breaks up t2MOVi32imm
to this pass. This is done before post regalloc scheduling to allow the
scheduler to proper schedule these instructions. It also allow them to be
if-converted and shrunk by later passes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86304
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Fri, 6 Nov 2009 23:45:15 +0000 (23:45 +0000)]
Honour subreg machine operands during asmprinting
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86303
91177308-0d34-0410-b5e6-
96231b3b80d8
Bob Wilson [Fri, 6 Nov 2009 23:33:28 +0000 (23:33 +0000)]
Print VMOV (immediate) operands as hexadecimal values. Apple's assembler
will not accept negative values for these. LLVM's default operand printing
sign extends values, so that valid unsigned values appear as negative
immediates. Print all VMOV immediate operands as hex values to resolve this.
Radar
7372576.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86301
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Fri, 6 Nov 2009 23:19:58 +0000 (23:19 +0000)]
Fix a bug where we'd call SplitBlockPredecessors with a pred in the
set only once even if it has multiple edges to BB.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86299
91177308-0d34-0410-b5e6-
96231b3b80d8
Bob Wilson [Fri, 6 Nov 2009 23:06:42 +0000 (23:06 +0000)]
Fix a broken test.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86298
91177308-0d34-0410-b5e6-
96231b3b80d8