projects
/
oota-llvm.git
/ history
commit
grep
author
committer
pickaxe
?
search:
re
summary
|
shortlog
|
log
|
commit
|
commitdiff
|
tree
first ⋅ prev ⋅
next
Allow 256-bit shuffles to still be split even if only half of the shuffle comes from...
[oota-llvm.git]
/
lib
/
Target
/
X86
/
X86ISelLowering.cpp
2012-05-21
Craig Topper
Allow 256-bit shuffles to still be split even if only...
blob
|
commitdiff
|
raw
2012-05-19
Nadav Rotem
On Haswell, perfer storing YMM registers using a single...
blob
|
commitdiff
|
raw
|
diff to current
2012-05-19
Nadav Rotem
Add support for additional in-reg vbroadcast patterns
blob
|
commitdiff
|
raw
|
diff to current
2012-05-18
Craig Topper
Simplify code a bit. No functional change intended.
blob
|
commitdiff
|
raw
|
diff to current
2012-05-18
Craig Topper
Simplify handling of v16i8 shuffles and fix a missed...
blob
|
commitdiff
|
raw
|
diff to current
2012-05-11
Hans Wennborg
Implement initial-exec TLS model for 32-bit PIC x86
blob
|
commitdiff
|
raw
|
diff to current
2012-05-10
Nadav Rotem
Fix merge-typo and cleanup
blob
|
commitdiff
|
raw
|
diff to current
2012-05-10
Nadav Rotem
AVX2: Add an additional broadcast idiom.
blob
|
commitdiff
|
raw
|
diff to current
2012-05-10
Nadav Rotem
Generate AVX/AVX2 shuffles even when there is a memory...
blob
|
commitdiff
|
raw
|
diff to current
2012-05-07
Chad Rosier
Fix a regression from r147481. This combine should...
blob
|
commitdiff
|
raw
|
diff to current
2012-05-07
Manman Ren
X86: optimization for -(x != 0)
blob
|
commitdiff
|
raw
|
diff to current
2012-05-06
Craig Topper
Use MVT instead of EVT as the argument to all the shuff...
blob
|
commitdiff
|
raw
|
diff to current
2012-05-06
Craig Topper
Add VPERMQ/VPERMPD to the list of target specific shuff...
blob
|
commitdiff
|
raw
|
diff to current
2012-05-05
Benjamin Kramer
Add a new target hook "predictableSelectIsExpensive".
blob
|
commitdiff
|
raw
|
diff to current
2012-05-04
Craig Topper
Fix some loops to match coding standards. No functional...
blob
|
commitdiff
|
raw
|
diff to current
2012-05-04
Craig Topper
Fix up some spacing. No functional change.
blob
|
commitdiff
|
raw
|
diff to current
2012-05-04
Craig Topper
Simplify broadcast lowering code. No functional change...
blob
|
commitdiff
|
raw
|
diff to current
2012-05-04
Craig Topper
Allow v16i16 and v32i8 shuffles to be rewritten as...
blob
|
commitdiff
|
raw
|
diff to current
2012-05-04
Craig Topper
Simplify shuffle narrowing code a bit. No functional...
blob
|
commitdiff
|
raw
|
diff to current
2012-05-03
Craig Topper
Use 'unsigned' instead of 'int' in a few places dealing...
blob
|
commitdiff
|
raw
|
diff to current
2012-05-03
Craig Topper
Fix 256-bit vpshuflw and vpshufhw immediate encoding...
blob
|
commitdiff
|
raw
|
diff to current
2012-05-02
Preston Gurd
For Intel Atom, use ILP scheduling always, instead...
blob
|
commitdiff
|
raw
|
diff to current
2012-05-02
Manman Ren
Revert r155853
blob
|
commitdiff
|
raw
|
diff to current
2012-05-02
Craig Topper
Add support for selecting AVX2 vpshuflw and vpshufhw...
blob
|
commitdiff
|
raw
|
diff to current
2012-05-01
Manman Ren
X86: optimization for max-like struct
blob
|
commitdiff
|
raw
|
diff to current
2012-04-30
Manman Ren
X86: optimization for -(x != 0)
blob
|
commitdiff
|
raw
|
diff to current
2012-04-30
Chad Rosier
Tidy up. No functional change intended.
blob
|
commitdiff
|
raw
|
diff to current
2012-04-30
Craig Topper
No need to normalize index before calling Extract128Bit...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-29
Jakub Staszak
Remove unneeded casts. No functionality change.
blob
|
commitdiff
|
raw
|
diff to current
2012-04-29
Craig Topper
Simplify code a bit. No functional change intended.
blob
|
commitdiff
|
raw
|
diff to current
2012-04-27
Craig Topper
Use 'unsigned' instead of 'int' in several places when...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-27
Chad Rosier
Add x86-specific DAG combine to simplify:
blob
|
commitdiff
|
raw
|
diff to current
2012-04-27
Craig Topper
Tidy up spacing.
blob
|
commitdiff
|
raw
|
diff to current
2012-04-27
Benjamin Kramer
X86: Don't emit conditional floating point moves on...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-25
Craig Topper
Use vector_shuffles instead of target specific unpack...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-24
Nadav Rotem
AVX2: The BLENDPW instruction selects between vectors...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-24
Craig Topper
Remove dangling spaces. Fix some other formatting.
blob
|
commitdiff
|
raw
|
diff to current
2012-04-24
Craig Topper
Simplify code a bit and make it compile better. Remove...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-23
Nadav Rotem
Optimize the vector UINT_TO_FP, SINT_TO_FP and FP_TO_SI...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-23
Craig Topper
Use MVT instead of EVT through all of LowerVECTOR_SHUFF...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-23
Craig Topper
Make getZeroVector and getOnesVector more alike as...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-23
Craig Topper
Tidy up by removing some 'else' after 'return'
blob
|
commitdiff
|
raw
|
diff to current
2012-04-23
Craig Topper
Tidy up spacing in LowerVECTOR_SHUFFLEtoBlend. Remove...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-23
Craig Topper
Add a couple llvm_unreachables.
blob
|
commitdiff
|
raw
|
diff to current
2012-04-23
Craig Topper
Remove some tab characers.
blob
|
commitdiff
|
raw
|
diff to current
2012-04-23
Craig Topper
Remove some 'else' after 'return'. No functional change.
blob
|
commitdiff
|
raw
|
diff to current
2012-04-22
Craig Topper
Make Extract128BitVector and Insert128BitVector take...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-22
Craig Topper
Convert getNode(UNDEF) to getUNDEF.
blob
|
commitdiff
|
raw
|
diff to current
2012-04-22
Craig Topper
Make calls to getVectorShuffle more consistent. Use...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-22
Craig Topper
Tidy up. 80 columns and argument alignment.
blob
|
commitdiff
|
raw
|
diff to current
2012-04-22
Craig Topper
Simplify code by converting multiple places that were...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-22
Elena Demikhovsky
ZERO_EXTEND/SIGN_EXTEND/TRUNCATE optimization for AVX2
blob
|
commitdiff
|
raw
|
diff to current
2012-04-21
Craig Topper
Make some fixed arrays const. Use array_lengthof in...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-21
Craig Topper
Tidy up. 80 columns and some other spacing issues.
blob
|
commitdiff
|
raw
|
diff to current
2012-04-20
Craig Topper
Convert some uses of XXXRegisterClass to &XXXRegClass...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-18
Craig Topper
Remove AVX vpermil intrinsics. I removed their uses...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-17
Craig Topper
Don't decode vperm2i128 or vperm2f128 into a shuffle...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-16
Richard Smith
Fix incorrect atomics codegen introduced in r154705...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-16
Craig Topper
Replace vpermd/vpermps intrinic patterns with custom...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-16
Craig Topper
Change type profile for vpermv back to using operand...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-16
Craig Topper
Merge vpermps/vpermd and vpermpd/vpermq SD nodes.
blob
|
commitdiff
|
raw
|
diff to current
2012-04-15
Craig Topper
Spacing fixes and 80 column fixes. Use 0 instead of...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-15
Elena Demikhovsky
Added VPERM optimization for AVX2 shuffles
blob
|
commitdiff
|
raw
|
diff to current
2012-04-13
Richard Smith
Fix X86 codegen for 'atomicrmw nand' to generate *x...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-11
Nadav Rotem
remove unused argument
blob
|
commitdiff
|
raw
|
diff to current
2012-04-11
Nadav Rotem
Reapply 154396 after fixing a test.
blob
|
commitdiff
|
raw
|
diff to current
2012-04-10
Chad Rosier
Whitespace.
blob
|
commitdiff
|
raw
|
diff to current
2012-04-10
Chad Rosier
Revert r154396, which looks to be the real culprit...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-10
Eric Christopher
Temporarily revert this patch to see if it brings the...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-10
David Blaikie
Remove unused variable.
blob
|
commitdiff
|
raw
|
diff to current
2012-04-10
Nadav Rotem
Modify the code that lowers shuffles to blends from...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-10
Evan Cheng
Fix a long standing tail call optimization bug. When...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-09
Nadav Rotem
Lower some x86 shuffle sequences to the vblend family...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-09
Nadav Rotem
Fix a bug in the lowering of broadcasts: ConstantPools...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-08
Chandler Carruth
Move the TLSModel information into the TargetMachine...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-08
Nadav Rotem
AVX2: Build splat vectors by broadcasting a scalar...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-06
Benjamin Kramer
Fix narrowing conversion.
blob
|
commitdiff
|
raw
|
diff to current
2012-04-06
Craig Topper
Allow 256-bit shuffles to be split if a 128-bit lane...
blob
|
commitdiff
|
raw
|
diff to current
2012-04-04
Rafael Espindola
Always compute all the bits in ComputeMaskedBits.
blob
|
commitdiff
|
raw
|
diff to current
2012-04-01
Nadav Rotem
This commit contains a few changes that had to go in...
blob
|
commitdiff
|
raw
|
diff to current
2012-03-21
Craig Topper
Spacing fixes and using 'unsigned' instead of 'int...
blob
|
commitdiff
|
raw
|
diff to current
2012-03-20
Craig Topper
Remove code that prevented lowering shuffles if they...
blob
|
commitdiff
|
raw
|
diff to current
2012-03-20
Craig Topper
Factor out target shuffle mask decoding from getShuffle...
blob
|
commitdiff
|
raw
|
diff to current
2012-03-18
Craig Topper
isCommutedMOVLMask should only look at 128-bit vectors...
blob
|
commitdiff
|
raw
|
diff to current
2012-03-11
Craig Topper
Convert more static tables of registers used by calling...
blob
|
commitdiff
|
raw
|
diff to current
2012-03-05
Chad Rosier
Address Evan's comments for r151877.
blob
|
commitdiff
|
raw
|
diff to current
2012-03-02
Chad Rosier
Prevent obscure and incorrect tail-call optimization.
blob
|
commitdiff
|
raw
|
diff to current
2012-02-28
Evan Cheng
Re-commit r151623 with fix. Only issue special no-retur...
blob
|
commitdiff
|
raw
|
diff to current
2012-02-28
Daniel Dunbar
Revert r151623 "Some ARM implementaions, e.g. A-series...
blob
|
commitdiff
|
raw
|
diff to current
2012-02-28
Evan Cheng
Some ARM implementaions, e.g. A-series, does return...
blob
|
commitdiff
|
raw
|
diff to current
2012-02-25
NAKAMURA Takumi
Target/X86: Fix assertion failures and warnings caused...
blob
|
commitdiff
|
raw
|
diff to current
2012-02-24
Michael J. Spencer
Add WIN_FTOL_* psudo-instructions to model the unique...
blob
|
commitdiff
|
raw
|
diff to current
2012-02-22
Craig Topper
Make all pointers to TargetRegisterClass const since...
blob
|
commitdiff
|
raw
|
diff to current
2012-02-19
Craig Topper
Remove some unneeded includes and fix ordering in X86IS...
blob
|
commitdiff
|
raw
|
diff to current
2012-02-19
Craig Topper
Unify all shuffle mask checking functions take a mask...
blob
|
commitdiff
|
raw
|
diff to current
2012-02-19
Craig Topper
Make a bunch of X86ISelLowering shuffle functions stati...
blob
|
commitdiff
|
raw
|
diff to current
2012-02-16
Jakob Stoklund Olesen
Use the same CALL instructions for Windows as for every...
blob
|
commitdiff
|
raw
|
diff to current
2012-02-16
Jakob Stoklund Olesen
Enable register mask operands for x86 calls.
blob
|
commitdiff
|
raw
|
diff to current
2012-02-13
Craig Topper
Update CanXFormVExtractWithShuffleIntoLoad to ensure...
blob
|
commitdiff
|
raw
|
diff to current
2012-02-11
Anton Korobeynikov
Add support for implicit TLS model used with MS VC...
blob
|
commitdiff
|
raw
|
diff to current
next