[llvm-commits] PATCH: wide strings alignment fix in clang

[NAKAMURA Takumi]

Good evening, Sundeep!

Thanks to submit your patches, though, it was wrong mailing list :p
You should re-submit to cfe-commits, please.

It seems good functionally to me.

...Takumi


2011/8/2 Sundeep <sundeepk at codeaurora.org>:
> Hi All,
>
> I have been working on a bug in clang dealing with alignment of wide
> strings. Clang is aligning wide string literals to 1 byte boundary. However,
> LLVM treats wide strings as 4 byte aligned and generates memory operations
> accessing 4 bytes at a time. This works fine on architectures that allow
> unaligned access. But for architectures which don't allow unaligned access,
> this results in an exception and segfault.
>
> I have attached the patch along with a simple test case to reproduce the
> problem.
>
> patch.llvm - clang fix to handle wide string literal alignment correctly.
>
> patch.test - patch to a test case problem in llvm/test regression directory.
>
> wstr_main.c - declares a string "ab" and a wide string L"AZ" and calls foo.
> String "ab" causes string literal "AZ" to start from an offset of 3 ('a',
> 'b', '\0').
>
> wstr_copy.c - attempts to copy a single wide char from L"AZ" into a global
> wide string array. This results in unaligned load exception since it tries
> to load 4 bytes from an address which is 3 byte aligned.
>
> makefile    - makefile to reproduce the problem.
>
> failures.txt - failure report and explanation from running llvm/test and
> projects/test-suite.
>
> This is my first attempt to commit into LLVM repository. I followed
> guidelines from "LLVM Developer Policy". Please let me know if I missed
> something or if you guys need any additional information.
>
> Thanks,
> Sundeep