aboutsummaryrefslogtreecommitdiff
path: root/test/Transforms
AgeCommit message (Collapse)Author
2008-10-17remove v[if]cmp functionality from the 2.4 release. These instructions will ↵Tanya Lattner
be removed or substantially changes in mainline after the 2.4 release and we don't want to have to support these for backwards compatibility purposes. Patch by Chris. git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_24@57732 91177308-0d34-0410-b5e6-96231b3b80d8
2008-10-17Merge from mainline.Tanya Lattner
Fix PR2697 by rewriting the '(X / pos) op neg' logic. This also changes a couple other cases for clarity, but shouldn't affect correctness. git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_24@57698 91177308-0d34-0410-b5e6-96231b3b80d8
2008-10-06Fix typo, fix PR 2865.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57221 91177308-0d34-0410-b5e6-96231b3b80d8
2008-10-06Allow scalarrepl to treat an all-zero GEP just as bitcast.Matthijs Kooijman
This includes not marking a GEP involving a vector as unsafe, but only when it has all zero indices. This allows scalarrepl to work in a few more cases. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57177 91177308-0d34-0410-b5e6-96231b3b80d8
2008-10-05rewrite bswap matching to be more general, allowing arbitraryChris Lattner
shifting and masking inside a bswap expr. This allows it to handle the cases from PR2842, which involve the intermediate 'or' expressions being shifted, not just the input value. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57095 91177308-0d34-0410-b5e6-96231b3b80d8
2008-10-04Ignore loads from and stores to local memory (i.e. allocas)Duncan Sands
when deciding whether to mark a function readnone/readonly. Since the pass is currently run before SROA, this may be quite helpful. Requested by Chris on IRC. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57050 91177308-0d34-0410-b5e6-96231b3b80d8
2008-10-04Allow the construction of SCEVs with SCEVCouldNotCompute operands, byNick Lewycky
implementing folding. Fixes PR2857. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57049 91177308-0d34-0410-b5e6-96231b3b80d8
2008-10-03Nick Lewycky's patch.Devang Patel
While hosting instruction check PHI node. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57025 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-30Fix misoptimization of: xor i1 (icmp eq (X, C1), icmp s[lg]t (X, C2))Nick Lewycky
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56834 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-29Support inreg, zext and sext as return value attributes.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56801 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-29Add a testcase showing that scalarrepl supports first class structs.Matthijs Kooijman
I originally made this script to show that scalarrepl didn't support them, but it turned out it does. Better to still add the testcase then. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56781 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-26Implement function notes as function attributes. Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56716 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-23Rationalize the names of passes that print information:Duncan Sands
-callgraph => print-callgraph -callscc => print-callgraph-sccs -cfgscc => print-cfg-sccs -externalfnconstants => print-externalfnconstants -print => print-function -print-alias-sets (no change) -print-callgraph => dot-callgraph -print-cfg => dot-cfg -print-cfg-only => dot-cfg-only -print-dom-info (no change) -printm => print-module -printusedtypes => print-used-types git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56487 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-19Add test for improvement of readonly to readnone,Duncan Sands
and non-demotion of readnone to readonly. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56344 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-19Turn on these tests!Duncan Sands
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56343 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-19Add a new pass AddReadAttrs which works out which functionsDuncan Sands
can get the readnone/readonly attributes, and gives them it. The plan is to remove markmodref (which did the same thing by querying GlobalsModRef) and delete the analogous functionality from GlobalsModRef. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56341 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-19Test the callgraph directly for the missing edge.Duncan Sands
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56338 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-18splitLoop does not handle split condition EQ.Devang Patel
Fixes PR 2805 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56321 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-17Do not hoist instruction above branch condition. The instruction may use ↵Devang Patel
branch condition. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56286 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-17Do not ignore iv uses outside the loop.Devang Patel
This one slipped through cracks very well. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56284 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-16Improve instcombine's handling of integer min and max in two ways:Dan Gohman
- Recognize expressions like "x > -1 ? x : 0" as min/max and turn them into expressions like "x < 0 ? 0 : x", which is easily recognizable as a min/max operation. - Refrain from folding expression like "y/2 < 1" to "y < 2" when the comparison is being used as part of a min or max idiom, like "y/2 < 1 ? 1 : y/2". In that case, the division has another use, so folding doesn't eliminate it, and obfuscates the min/max, making it harder to recognize as a min/max operation. These benefit ScalarEvolution, CodeGen, and anything else that wants to recognize integer min and max. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56246 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-11On 64-bit targets, change 32-bit getelementptr indices to be 64-bitDan Gohman
getelementptr indices, inserting an explicit cast if necessary. This helps expose the sign-extension operation to other optimizations. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56133 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-11Fix a vectorshuffle instcombine bug introduced by r55995.Dan Gohman
Patch by Nicolas Capens! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56129 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-10Fix an icmp+sdiv optimization to check for and handle an overflowDan Gohman
condition. This fixes PR2740. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56076 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-09Remove. Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56018 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-09if loop induction variable is always sign or zero extended then Devang Patel
extend the type of induction variable. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56017 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-09fix overflow check.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56011 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-09Resolve aliases, when possibleAnton Korobeynikov
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56001 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-09Make SimplifyDemandedVectorElts simplify vectors with multipleDan Gohman
users, and teach it about shufflevector instructions. Also, fix a subtle bug in SimplifyDemandedVectorElts' insertelement code. This is a patch that was originally written by Eli Friedman, with some fixes and cleanup by me. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55995 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-09Fix simplifycfg crash in handing block merge.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55971 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-08xfailDevang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55914 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-08Update the callgraph correctly in ArgumentPromotion.Duncan Sands
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55895 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-06When PruneEH turned an invoke into an ordinaryDuncan Sands
call (thus changing the call site) it didn't inform the callgraph about this. But the call site does matter - as shown by the testcase, the callgraph become invalid after the inliner ran (with an edge between two functions simply missing), resulting in wrong deductions by GlobalsModRef. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55872 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-03Don't crash when trying to constant fold a vector with some elements that can'tNick Lewycky
be folded. Instead, fail to fold the entire vector. We could also return a vector with some elements folded and some not. If anyone thinks that's a better approach, please speak up! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55689 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-03Check iteration count.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55680 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-02If all IV uses are extending integer IV then change the type of IV itself, ↵Devang Patel
if possible. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55674 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-02respect inline=never and inline=always notes.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55673 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-26If IV is used in a int-to-float cast inside the loop then try to eliminate ↵Devang Patel
the cast operation. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55374 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-23Fix PR2423 by checking all indices for out of range access, not only Chris Lattner
indices that start with an array subscript. x->field[10000] is just as bad as (*X)[14][10000]. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55226 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-21Revert r54876 r54877 r54906 and r54907. Evan found that these caused a 20%Nick Lewycky
slowdown in bzip2. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55113 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-18XFAIL this test for now.Bill Wendling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@54929 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-17Consider the case where xor by -1 and xor by 128 have been combined already toNick Lewycky
produce an xor by 127. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@54906 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-17Didn't mean to change this.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@54904 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-17Fix a (u)comiss intrinsic lowering bug. It was using anyext which can return ↵Evan Cheng
junk in higher bits. Patch by Nate Begeman. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@54903 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-17Xor'ing both sides of icmp by sign-bit is equivalent to swapping signedness ofNick Lewycky
the predicate. Also, make this optz'n apply in more cases where it's safe to do so. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@54876 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-16Remove GCSE and LoadVN from the testsuite.Owen Anderson
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@54832 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-15Reapply 54786. Add overflow and number of mantissa bits checks.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@54821 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-15Revert 54786. It's not checking for overflows, etc.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@54813 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-14If IV is used in a int-to-float cast inside the loop then try to eliminate ↵Devang Patel
the cast opeation. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@54786 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-13Fix a bogus srem rule - a negative value srem'd by a power-of-2Dan Gohman
can have a non-negative result; for example, -16%16 is 0. Also, clarify the related comments. This fixes PR2670. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@54767 91177308-0d34-0410-b5e6-96231b3b80d8
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2026-02-10 20:35:40 +0000