diff options
| author | Philip Reames <listmail@philipreames.com> | 2017-10-30 23:59:51 +0000 |
|---|---|---|
| committer | Philip Reames <listmail@philipreames.com> | 2017-10-30 23:59:51 +0000 |
| commit | 9c3cbeea3943dcfa096634eb94c8acc5ff8bf640 (patch) | |
| tree | 5b22227e9913884476f782da3708c7414c95e999 /llvm/lib/CodeGen/CodeGenPrepare.cpp | |
| parent | c38ba6697fc8b9f5cb577d27c07dc9ff46ae6b17 (diff) | |
| download | bcm5719-llvm-9c3cbeea3943dcfa096634eb94c8acc5ff8bf640.tar.gz bcm5719-llvm-9c3cbeea3943dcfa096634eb94c8acc5ff8bf640.zip | |
[CGP] Fix crash on i96 bit multiply
Issue found by llvm-isel-fuzzer on OSS fuzz, https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=3725
If anyone actually cares about > 64 bit arithmetic, there's a lot more to do in this area. There's a bunch of obviously wrong code in the same function. I don't have the time to fix all of them and am just using this to understand what the workflow for fixing fuzzer cases might look like.
llvm-svn: 316967
Diffstat (limited to 'llvm/lib/CodeGen/CodeGenPrepare.cpp')
| -rw-r--r-- | llvm/lib/CodeGen/CodeGenPrepare.cpp | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/llvm/lib/CodeGen/CodeGenPrepare.cpp b/llvm/lib/CodeGen/CodeGenPrepare.cpp index 0a417e34084..13bdad71252 100644 --- a/llvm/lib/CodeGen/CodeGenPrepare.cpp +++ b/llvm/lib/CodeGen/CodeGenPrepare.cpp @@ -4032,7 +4032,7 @@ bool AddressingModeMatcher::matchOperationAddr(User *AddrInst, unsigned Opcode, case Instruction::Shl: { // Can only handle X*C and X << C. ConstantInt *RHS = dyn_cast<ConstantInt>(AddrInst->getOperand(1)); - if (!RHS) + if (!RHS || RHS->getBitWidth() > 64) return false; int64_t Scale = RHS->getSExtValue(); if (Opcode == Instruction::Shl) |
