Commit Graph

51365 Commits

Author SHA1 Message Date
Chandler Carruth
51901d85f7 Prevent rotating the blocks of a loop (and thus getting a backedge to be
fallthrough) in cases where we might fail to rotate an exit to an outer
loop onto the end of the loop chain.

Having *some* rotation, but not performing this rotation, is the primary
fix of thep performance regression with -enable-block-placement for
Olden/em3d (a whopping 30% regression). Still working on reducing the
test case that actually exercises this and the new rotation strategy out
of this code, but I want to check if this regresses other test cases
first as that may indicate it isn't the correct fix.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145195 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-27 20:18:00 +00:00
Chandler Carruth
fac1305da1 Take two on rotating the block ordering of loops. My previous attempt
was centered around the premise of laying out a loop in a chain, and
then rotating that chain. This is good for preserving contiguous layout,
but bad for actually making sane rotations. In order to keep it safe,
I had to essentially make it impossible to rotate deeply nested loops.
The information needed to correctly reason about a deeply nested loop is
actually available -- *before* we layout the loop. We know the inner
loops are already fused into chains, etc. We lose information the moment
we actually lay out the loop.

The solution was the other alternative for this algorithm I discussed
with Benjamin and some others: rather than rotating the loop
after-the-fact, try to pick a profitable starting block for the loop's
layout, and then use our existing layout logic. I was worried about the
complexity of this "pick" step, but it turns out such complexity is
needed to handle all the important cases I keep teasing out of benchmarks.

This is, I'm afraid, a bit of a work-in-progress. It is still
misbehaving on some likely important cases I'm investigating in Olden.
It also isn't really tested. I'm going to try to craft some interesting
nested-loop test cases, but it's likely to be extremely time consuming
and I don't want to go there until I'm sure I'm testing the correct
behavior. Sadly I can't come up with a way of getting simple, fine
grained test cases for this logic. We need complex loop structures to
even trigger much of it.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145183 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-27 13:34:33 +00:00
Chandler Carruth
790552c20f Revert r145180 as it is causing test failures on all the bots.
Original commit message:
Fixed ObjectFile functions:
- getSymbolOffset() renamed as getSymbolFileOffset()
- getSymbolFileOffset(), getSymbolAddress(), getRelocationAddress() returns same result for ELFObjectFile, MachOObjectFile and COFFObjectFile.
- added getRelocationOffset()
- fixed MachOObjectFile::getSymbolSize()
- fixed MachOObjectFile::getSymbolSection()
- fixed MachOObjectFile::getSymbolOffset() for symbols without section data.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145182 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-27 10:37:47 +00:00
Chandler Carruth
7096692fd9 Fix an impressive type-o / spell-o Duncan noticed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145181 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-27 10:32:16 +00:00
Danil Malyshev
9b24738bd1 Fixed ObjectFile functions:
- getSymbolOffset() renamed as getSymbolFileOffset()
- getSymbolFileOffset(), getSymbolAddress(), getRelocationAddress() returns same result for ELFObjectFile, MachOObjectFile and COFFObjectFile.
- added getRelocationOffset()
- fixed MachOObjectFile::getSymbolSize()
- fixed MachOObjectFile::getSymbolSection()
- fixed MachOObjectFile::getSymbolOffset() for symbols without section data.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145180 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-27 10:12:52 +00:00
Chandler Carruth
2eb5a744b1 Rework a bit of the implementation of loop block rotation to not rely so
heavily on AnalyzeBranch. That routine doesn't behave as we want given
that rotation occurs mid-way through re-ordering the function. Instead
merely check that there are not unanalyzable branching constructs
present, and then reason about the CFG via successor lists. This
actually simplifies my mental model for all of this as well.

The concrete result is that we now will rotate more loop chains. I've
added a test case from Olden highlighting the effect. There is still
a bit more to do here though in order to regain all of the performance
in Olden.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145179 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-27 09:22:53 +00:00
Chris Lattner
b5dd9de724 Eli managed to kill off llvm.membarrier in llvm 3.0 also, this means
that mainline needs no autoupgrade logic for intrinsics yet, woohoo!


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145178 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-27 08:42:07 +00:00
Chris Lattner
875882a184 The llvm.atomic intrinsics *were* removed in LLVM 3.0 (in r141333), remove the
autoupgrade logic for 2.9 and before.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145176 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-27 08:18:55 +00:00
Chris Lattner
3211c6e31b remove autoupgrade support for old forms of llvm.prefetch and the old
trampoline forms.  Both of these were correct in LLVM 3.0, and we don't
need to support LLVM 2.9 and earlier in mainline.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145174 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-27 07:42:04 +00:00
Chris Lattner
fbe910e7f4 remove asmparsing and documentation support for "volatile load", which was only produced by LLVM 2.9 and earlier. LLVM 3.0 and later prefers "load volatile".
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145172 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-27 06:56:53 +00:00
Chris Lattner
a7e6f74631 remove autoupgrade support for really old-style debug info intrinsics.
I think this is the last of autoupgrade that can be removed in 3.1.
Can the atomic upgrade stuff also go?


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145169 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-27 06:18:33 +00:00
Chris Lattner
663aebf8d6 remove some old autoupgrade logic
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145167 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-27 06:10:54 +00:00
Chris Lattner
9d5f6ccd26 remove autoupgrade support for LLVM 2.9 exception stuff. Mainline supports
LLVM 3.0 and later.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145165 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-27 05:56:16 +00:00
Chris Lattner
dbc88494ee remove support for reading llvm 2.9 .bc files. LLVM 3.1 is only compatible back to 3.0
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145164 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-27 05:48:27 +00:00
Wesley Peck
4c729f1152 Add several new instructions supported by the latest MicroBlaze.
These instructions are not generated by the backend yet, this will come in a later commit.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145161 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-27 05:16:58 +00:00
Wesley Peck
a0aa45c759 Optimize comparison against 0 in conditional instructions.
Fix a couple of 80-column violations.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145159 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-27 01:36:20 +00:00
Chandler Carruth
2e38cf961d Introduce a loop block rotation optimization to the new block placement
pass. This is designed to achieve one of the important optimizations
that the old code placement pass did, but more simply.

This is a somewhat rough and *very* conservative version of the
transform. We could get a lot fancier here if there are profitable cases
to do so. In particular, this only looks for a single pattern, it
insists that the loop backedge being rotated away is the last backedge
in the chain, and it doesn't provide any means of doing better in-loop
placement due to the rotation. However, it appears that it will handle
the important loops I am finding in the LLVM test suite.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145158 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-27 00:38:03 +00:00
Benjamin Kramer
0861f5793a Move code into anonymous namespaces.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145154 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-26 23:01:57 +00:00
Craig Topper
38034c568c Merge 128-bit and 256-bit X86ISD node types for VPERMILPS and VPERMILPD. Simplify some shuffle lowering code since V1 can never be UNDEF due to canonalizing that occurs when shuffle nodes are created.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145153 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-26 22:55:48 +00:00
Wesley Peck
a18f08318a Rename a couple of options and fix some simple typos.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145152 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-26 21:50:38 +00:00
Craig Topper
06cb680779 Collapse X86ISD node types for PUNPCKH*, PUNPCKL*, UNPCKLP*, and UNPCKHP* to not be type specific. Now we just have integer high and low and floating point high and low. Pattern matching will choose the correct instruction based on the vector type.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145148 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-26 20:47:44 +00:00
Eli Friedman
4455142a95 Fix APFloat::convert so that it handles narrowing conversions correctly; it
was returning incorrect values in rare cases, and incorrectly marking
exact conversions as inexact in some more common cases. Fixes PR11406, and a
missed optimization in test/CodeGen/X86/fp-stack-O0.ll.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145141 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-26 03:38:02 +00:00
Bruno Cardoso Lopes
1b9b377975 This patch contains support for encoding FMA4 instructions and
tablegen patterns for scalar FMA4 operations and intrinsic. Also
add tests for vfmaddsd.

Patch by Jan Sjodin

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145133 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-25 19:33:42 +00:00
NAKAMURA Takumi
70aaf37c11 ARMLoadStoreOptimizer.cpp: Fix MSVC(Debug) build.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145129 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-25 09:19:57 +00:00
Craig Topper
705f2431a0 Remove 256-bit specific node types for UNPCKHPS/D and instead use the 128-bit versions and let the operand type disinquish. Also fix the load form of the v8i32 patterns for these to realize that the load would be promoted to v4i64.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145126 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-24 22:57:10 +00:00
Craig Topper
f475a55bd4 Remove AVX2 specific X86ISD node types for PUNPCKH/L and instead just reuse the 128-bit versions and let the vector type distinguish.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145125 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-24 22:20:08 +00:00
Benjamin Kramer
8a42633273 Devirtualize Pass::getPassID, overriding it isn't useful and it gets called a lot.
While at it pull the trivial ctor in line.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145124 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-24 21:14:11 +00:00
Benjamin Kramer
b3ff49e923 Make ConstantRange::truncate a bit more efficient.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145122 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-24 17:24:33 +00:00
Benjamin Kramer
9d399b1fc2 X86: alias cqo to cqto.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145121 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-24 12:02:46 +00:00
Chandler Carruth
4aae4f9007 Fix a silly use-after-free issue. A much earlier version of this code
need lots of fanciness around retaining a reference to a Chain's slot in
the BlockToChain map, but that's all gone now. We can just go directly
to allocating the new chain (which will update the mapping for us) and
using it.

Somewhat gross mechanically generated test case replicates the issue
Duncan spotted when actually testing this out.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145120 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-24 11:23:15 +00:00
Chandler Carruth
a2deea1dcf When adding blocks to the list of those which no longer have any CFG
conflicts, we should only be adding the first block of the chain to the
list, lest we try to merge into the middle of that chain. Most of the
places we were doing this we already happened to be looking at the first
block, but there is no reason to assume that, and in some cases it was
clearly wrong.

I've added a couple of tests here. One already worked, but I like having
an explicit test for it. The other is reduced from a test case Duncan
reduced for me and used to crash. Now it is handled correctly.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145119 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-24 08:46:04 +00:00
Akira Hatanaka
421455f1ea This patch makes the following changes necessary for MIPS' direct code emission.
- lower unaligned loads/stores.
- encode the size operand of instructions INS and EXT.
- emit relocation information needed for JAL (jump-and-link).  


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145113 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-23 22:19:28 +00:00
Akira Hatanaka
84bfc2f090 This patch addresses gp relative fixups/relocations for jump tables.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145112 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-23 22:18:04 +00:00
Richard Smith
e73db4e2a7 Correctly byte-swap APInts with bit-widths greater than 64.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145111 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-23 21:33:37 +00:00
Benjamin Kramer
1db1c239fe Validate the return type when checking if a function is malloc.
Fixes PR11426. Not sure if a test case with a "wrong" malloc would be useful.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145106 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-23 17:58:47 +00:00
Duncan Sands
a8f5cd3539 Fix a crash in which a multiplication was being reported as being both negative
and positive: positive, because it could be directly computed to be positive;
negative, because the nsw flags means it is either negative or undefined (the
multiplication always overflowed).


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145104 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-23 16:26:47 +00:00
Benjamin Kramer
f238f50aaf X86: Use btq for bit tests if the immediate can't be encoded in 32 bits.
Before:
	movabsq	$4294967296, %rax       ## encoding: [0x48,0xb8,0x00,0x00,0x00,0x00,0x01,0x00,0x00,0x00]
	testq	%rax, %rdi              ## encoding: [0x48,0x85,0xf8]
	jne	LBB0_2                  ## encoding: [0x75,A]

After:
	btq	$32, %rdi               ## encoding: [0x48,0x0f,0xba,0xe7,0x20]
	jb	LBB0_2                  ## encoding: [0x72,A]

btq is usually slower than testq because it doesn't fuse with the jump, but here we're better off
saving one register and a giant movabsq.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145103 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-23 13:54:17 +00:00
Chandler Carruth
598894ff25 Relax an invariant that block placement was trying to assert a bit
further. This invariant just wasn't going to work in the face of
unanalyzable branches; we need to be resillient to the phenomenon of
chains poking into a loop and poking out of a loop. In fact, we already
were, we just needed to not assert on it.

This was found during a bootstrap with block placement turned on.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145100 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-23 10:35:36 +00:00
Elena Demikhovsky
52a35a89e6 I added several lines in X86 code generator that allow to choose
VSHUFPS/VSHUFPD instructions while lowering VECTOR_SHUFFLE node. I check a commuted VSHUFP mask.

The patch was reviewed by Bruno.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145099 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-23 10:23:16 +00:00
Chandler Carruth
521fc5bcd7 Handle the case of a no-return invoke correctly. It actually still has
successors, they just are all landing pad successors. We handle this the
same way as no successors. Comments attached for the next person to wade
through here and another lovely test case courtesy of Benjamin Kramer's
bugpoint reduction.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145098 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-23 08:23:54 +00:00
Bob Wilson
23d66a58b7 Enable stack protectors for all arrays, not just char arrays. rdar://5875909
Patch by Bill Wendling.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145097 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-23 07:13:56 +00:00
Jakob Stoklund Olesen
7f5e43f61d Fix PR11422.
This was a bug in keeping track of the available domains when merging
domain values.

The wrong domain mask caused ExecutionDepsFix to try to move VANDPSYrr
to the integer domain which is only available in AVX2.

Also add an assertion to catch future attempts at emitting AVX2
instructions.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145096 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-23 04:03:08 +00:00
Chandler Carruth
47fb954f74 Fix a crash in block placement due to an inner loop that happened to be
reversed in the function's original ordering, and we happened to
encounter it while handling an outer unnatural CFG structure.

Thanks to the test case reduced from GCC's source by Benjamin Kramer.
This may also fix a crasher in gzip that Duncan reduced for me, but
I haven't yet gotten to testing that one.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145094 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-23 03:03:21 +00:00
Kostya Serebryany
d2703dec27 [asan] do not instrument threadlocal globals, this is buggy
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145092 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-23 02:10:54 +00:00
Hal Finkel
768c65f677 add basic PPC register-pressure feedback; adjust the vaarg test to match the new register-allocation pattern
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145065 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-22 16:21:04 +00:00
Craig Topper
796c193768 More fixes to the X86InstComments for shuffle instructions. In particular add AVX flavors of many instructions and fix the destination operand for some of the existing AVX entries.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145063 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-22 14:27:57 +00:00
Chandler Carruth
3b7b209bf8 Fix a devilish miscompile exposed by block placement. The
updateTerminator code didn't correctly handle EH terminators in one very
specific case. AnalyzeBranch would find no terminator instruction, and
so the fallback in updateTerminator is to assume fallthrough. This is
correct, but the destination of the fallthrough was assumed to be the
first successor.

This is *almost always* true, but in certain cases the loop
transformations will cause the landing pad to be the first successor!
Instead of this brittle logic, actually look through the successors for
a non-landing-pad accessor, and to assert if more than one is found.

This will hopefully fix some (if not all) of the self host miscompiles
with block placement. Thanks to Benjamin Kramer for reporting, Nick
Lewycky for an initial stab at a reduction, and Duncan for endless
advice on EH (which I know nothing about) as well as reviewing the
actual fix.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145062 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-22 13:13:16 +00:00
Benjamin Kramer
5745fbce16 Add configure checking for pread(2) and use it to save a syscall when reading files.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145061 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-22 12:31:53 +00:00
Chandler Carruth
f264568bae Fix an obvious omission in the SelectionDAGBuilder where we were
dropping weights on the floor for invokes. This was impeding my writing
further test cases for invoke when interacting with probabilities and
block placement.

No test case as there doesn't appear to be a way to test this stuff. =/
Suggestions for a test case of course welcome. I hope to be able to add
test cases that indirectly cover this eventually by adding probabilities
to the exceptional edge and reordering blocks as a result.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145060 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-22 11:37:46 +00:00
Benjamin Kramer
a673e8352a Turn error recovery into an assert.
This was put in because in a certain version of DragonFlyBSD stat(2) lied about the
size of some files. This was fixed a long time ago so we can remove the workaround.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145059 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-22 11:37:11 +00:00