summaryrefslogtreecommitdiffstats
path: root/llvm/test/CodeGen/X86/sse-minmax.ll
Commit message (Expand)AuthorAgeFilesLines
* [X86] Fix execution domain for cmpss/sd instructions.Craig Topper2017-02-261-102/+102
* [DAGCombine] Recognise any_extend_vector_inreg and truncation style shuffle m...Simon Pilgrim2017-02-171-4/+0
* [X86] Fix printing of blendvpd/blendvps/pblendvb to include the implicit %xmm...Craig Topper2017-02-051-8/+8
* Regenerate test with update_llc_test_checks.pySimon Pilgrim2017-01-131-65/+1
* [X86][SSE] Fix domains for VZEXT_LOAD type instructionsSimon Pilgrim2016-12-151-38/+38
* [X86][SSE] Add initial support for combining (V)PMOVZX with shuffles.Simon Pilgrim2016-11-281-6/+2
* [x86] add fake scalar FP logic instructions to ReplaceableInstrs to save some...Sanjay Patel2016-11-161-56/+56
* [x86] auto-generate better checks; NFCSanjay Patel2016-11-151-720/+1112
* fix typo; NFCSanjay Patel2015-12-051-1/+1
* [X86][SSE] Provide execution domains for scalar floating point operationsSimon Pilgrim2015-04-211-12/+12
* [X86] Also create+widen FMIN/FMAX nodes for v2f32.Ahmed Bougacha2015-01-121-0/+36
* [X86] Make SSE min/max testcases more explicit. NFC.Ahmed Bougacha2015-01-121-32/+84
* Improve logic that decides if its profitable to commute when some of the virt...Craig Topper2014-11-051-32/+16
* X86: Turn fp selects into mask operations.Benjamin Kramer2013-08-041-24/+31
* Disambiguate function names in some CodeGen tests. (Some tests were using fun...Stephen Lin2013-07-181-8/+8
* Mass update to CodeGen tests to use CHECK-LABEL for labels corresponding to f...Stephen Lin2013-07-141-172/+172
* When unsafe math is used, we can use commutative FMAX and FMIN. In some casesNadav Rotem2012-08-191-48/+32
* X86 Peephole: fold loads to the source register operand if possible.Manman Ren2012-08-021-20/+40
* X86 Peephole: fold loads to the source register operand if possible.Manman Ren2012-08-021-43/+23
* Revert r160920 and r160919 due to dragonegg and clang selfhost failureManman Ren2012-07-291-23/+43
* X86 Peephole: fold loads to the source register operand if possible.Manman Ren2012-07-281-3/+3
* X86 Peephole: fold loads to the source register operand if possible.Manman Ren2012-07-281-40/+20
* Remove the "-promote-elements" flag. This flag is now enabled by default.Nadav Rotem2012-06-041-3/+3
* Remove -join-physregs from the test suite.Jakob Stoklund Olesen2012-05-171-144/+141
* Add mcpu to tests to prevent them from using AVX instructions on Sandy Bridge...Craig Topper2012-04-271-3/+3
* Make X86::FsFLD0SS / FsFLD0SD real pseudo-instructions.Jakob Stoklund Olesen2011-11-291-40/+40
* Synthesize x86 max/min instructions also for vectors (i.e. produceDuncan Sands2011-09-171-3/+35
* Prepare remaining tests for -join-physreg going away.Jakob Stoklund Olesen2011-05-041-3/+6
* Split -enable-finite-only-fp-math to two options:Evan Cheng2010-07-151-2/+2
* -enable-unsafe-fp-math should not imply -enable-finite-only-fp-math.Evan Cheng2010-07-131-1/+1
* Fix a few testsJakob Stoklund Olesen2010-07-091-60/+60
* When forming SSE min and max nodes for UGE and ULE comparisons, it'sDan Gohman2010-02-241-13/+429
* Actually enable the -enable-unsafe-fp-math tests.Dan Gohman2010-02-221-144/+144
* Remove the logic for reasoning about NaNs from the code that formsDan Gohman2010-02-221-44/+168
* Recognize SSE min and max opportunities in even more cases.Dan Gohman2009-09-211-1/+320
* Eliminate more uses of llvm-as and llvm-dis.Dan Gohman2009-09-081-1/+1
* Recognize more opportunities to use SSE min and max instructions,Dan Gohman2009-09-031-0/+73
OpenPOWER on IntegriCloud