[llvm-bugs] [Bug 49141] New: opt -indvars -loop-load-elim fails with old PM

via llvm-bugs llvm-bugs at lists.llvm.org
Thu Feb 11 01:15:57 PST 2021


            Bug ID: 49141
           Summary: opt -indvars -loop-load-elim fails with old PM
           Product: libraries
           Version: trunk
          Hardware: PC
                OS: Linux
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: Scalar Optimizations
          Assignee: unassignedbugs at nondot.org
          Reporter: mikael.holmen at ericsson.com
                CC: llvm-bugs at lists.llvm.org

Created attachment 24513
  --> https://bugs.llvm.org/attachment.cgi?id=24513&action=edit
bbi-52668.ll reproducer

Reproduce with
 opt -enable-new-pm=0 -S -o /dev/null bbi-52668.ll -indvars -loop-load-elim


opt: ../include/llvm/IR/ValueHandle.h:490: ValueTy
*llvm::PoisoningVH<llvm::BasicBlock>::getValPtr() const [ValueTy =
llvm::BasicBlock]: Assertion `!Poisoned && "Accessed a poisoned value handle!"'
PLEASE submit a bug report to https://bugs.llvm.org/ and include the crash

This starts happening with commit 664e1da485d2:

    [LoopLoadElim] Make sure all loops are in simplify form. PR48150

    LoopLoadElim may end up expanding an AddRec from a loop
    which is not the current loop. This loop may not be in simplify
    form. We figure it out after the no-return point, so cannot bail
    in this case.

    AddRec requires simplify form to expand. The only way to ensure
    this does not crash is to simplify all loops beforehand.

    The issue only exists in new PM. Old PM requests LoopSimplify
    required pass and it simplifies all loops before the opt begins.

    Differential Revision: https://reviews.llvm.org/D91525
    Reviewed By: asbirlea, aeubanks

You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20210211/57d5dcdf/attachment.html>

More information about the llvm-bugs mailing list