[PATCH] D94869: [LV] Fix crash when computing max VF too early

[Cullen Rhodes via Phabricator via llvm-commits]

c-rhodes created this revision.
c-rhodes added a reviewer: iajbar.
Herald added a subscriber: hiraditya.
c-rhodes requested review of this revision.
Herald added a project: LLVM.

D90687 <https://reviews.llvm.org/D90687> introduced a crash:

  llvm::LoopVectorizationCostModel::computeMaxVF(llvm::ElementCount, unsigned int): Assertion `WideningDecisions.empty() && Uniforms.empty() && Scalars.empty() && "No decisions should have been taken at this point"' failed.

when compiling the following C code:

  typedef struct {
  char a;
  } b;
  
  b *c;
  int d, e;
  
  int f() {
    int g = 0;
    for (; d; d++) {
      e = 0;
      for (; e < c[d].a; e++)
        g++;
    }
    return g;
  }

with:

  clang -Os -target hexagon -mhvx -fvectorize -mv67 testcase.c -S -o -

This occurred since prior to D90687 <https://reviews.llvm.org/D90687> computeFeasibleMaxVF would only be
called in computeMaxVF when a scalar epilogue was allowed, but now it's
always called. This causes the assert above since computeFeasibleMaxVF
collects all viable VFs larger than the default MaxVF, and for each VF
calculates the register usage which results in analysis being done the
assert above guards against. This can occur in computeFeasibleMaxVF if
TTI.shouldMaximizeVectorBandwidth and this target hook is implemented in
the hexagon backend to always return true.

Reported by @iajbar.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D94869

Files:
  llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
  llvm/test/Transforms/LoopVectorize/Hexagon/maximum-vf-crash.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D94869.317183.patch
Type: text/x-patch
Size: 4310 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20210116/d89fa7d7/attachment.bin>