Miscellaneous fixes proposed for 3.5.1

David Majnemer david.majnemer at gmail.com
Wed Nov 26 15:11:17 PST 2014


to this list, I would add:
r222856 Revert "Added inst combine transforms for single bit tests from
Chris's note"

On Tue, Nov 25, 2014 at 11:50 PM, David Majnemer <david.majnemer at gmail.com>
wrote:

> Hi, I think the following list of commit revisions should be merged into
> 3.5.1:
>
> r214385 InstCombine: Correctly propagate NSW/NUW for x-(-A) -> x+A
> r215818 InstCombine: Fix a potential bug in 0 - (X sdiv C)  -> (X sdiv -C)
> r216261 InstCombine: Don't unconditionally preserve 'nsw' when shrinking
> constants
> r216273 InstCombine: Don't unconditionally preserve 'nuw' when shrinking
> constants
> r216911 LICM: Don't crash when an instruction is used by an unreachable BB
> r219566 InstCombine: mul to shl shouldn't preserve nsw
> r219567 InstCombine, InstSimplify: (%X /s C1) /s C2 isn't always 0 when
> C1 * C2 overflow
> r219568 InstCombine: Don't fold (X <<s log(INT_MIN)) /s INT_MIN to X
> r219634 InstCombine: Don't miscompile (x lshr C1) udiv C2
> r219639 InstCombine: Fix miscompile in X % -Y -> X % Y transform
> r219713 InstCombine: Don't miscompile X % ((Pow2 << A) >>u B)
> r220664 InstCombine: Fix a combine assuming that icmp operands were
> integers
> r221069 InstCombine: Don't assume that m_ZExt matches an Instruction
> r221187 InstCombine: Remove infinite loop caused by FoldOpIntoPhi
> r221318 Analysis: Make isSafeToSpeculativelyExecute fire less for divides
> r221513 SCCP: overdefined calls cannot become constant
> r222040 InstCombine: Fix infinite loop caused by visitFPTrunc
> r222093 ScalarEvolution: HowFarToZero was wrongly using signed division
> r222104 ScalarEvolution: Introduce SCEVSDivision and SCEVUDivision
> r222126 ScalarEvolution: Construct SCEVDivision's Derived type instead of
> itself
> r222265 InstCombine: Fix another infinite loop caused by visitFPTrunc
>
> --
> David Majnemer
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20141126/f59427ae/attachment.html>


More information about the llvm-commits mailing list