[formatv][FmtAlign] Use fill count of type size_t instead of uint32_t #78459
Conversation
FmtAlign::fill() accepts a uint32_t variable while the usages operate on size_t values. On some platform / compiler combinations, this ends up being a narrowing conversion. Fix this by changing the function's signature. Co-authored-by: Orest Chura <[email protected]>
|
@llvm/pr-subscribers-llvm-support Author: Andrei Golubev (andrey-golubev) ChangesFmtAlign::fill() accepts a uint32_t variable while the usages operate on size_t values. On some platform / compiler combinations, this ends up being a narrowing conversion. Fix this by changing the function's signature. Full diff: https://github.com/llvm/llvm-project/pull/78459.diff 1 Files Affected:
diff --git a/llvm/include/llvm/Support/FormatCommon.h b/llvm/include/llvm/Support/FormatCommon.h
index 3c119d12529aeed..24a40c325e13148 100644
--- a/llvm/include/llvm/Support/FormatCommon.h
+++ b/llvm/include/llvm/Support/FormatCommon.h
@@ -66,8 +66,8 @@ struct FmtAlign {
}
private:
- void fill(llvm::raw_ostream &S, uint32_t Count) {
- for (uint32_t I = 0; I < Count; ++I)
+ void fill(llvm::raw_ostream &S, size_t Count) {
+ for (size_t I = 0; I < Count; ++I)
S << Fill;
}
};
|
|
@zturner0826 @labath please take a look (you seem to be the ones who worked on this before)! |
|
also, I guess @inglorion could have some experience with the surroundings. |
|
For the record, can you please post the compiler message that you see before this PR? (As well as compiler version) |
aha. this would be rather complicated since I'm not the original author (also, this patch was submitted ~1 year ago if not earlier). to the best of my knowledge, it fails on some MSVC version on x86. i can try to ping whoever brought it up in our downstream originally, maybe they still have something of the kind around. |
yeah, no luck. I found the original patch in our fork but the issues were fixed in bulk for MSVC 2019 so there's no special error for this :| from our internal chats I guess it is roughly "conversion from 'size_t' to 'uint32_t' requires a narrowing conversion". |
|
No worries, thanks for investigating! |
FmtAlign::fill() accepts a uint32_t variable while the usages operate on size_t values. On some platform / compiler combinations, this ends up being a narrowing conversion. Fix this by changing the function's signature.