Re: [PATCH v5] x86: bitops: fix build regression

[Brian Gerst]

On Fri, May 8, 2020 at 2:32 PM Nick Desaulniers <ndesaulniers@xxxxxxxxxx> wrote:
>
> This is easily reproducible via CC=clang+CONFIG_STAGING=y+CONFIG_VT6656=m.
>
> It turns out that if your config tickles __builtin_constant_p via
> differences in choices to inline or not, these statements produce
> invalid assembly:
>
> $ cat foo.c
> long a(long b, long c) {
>   asm("orb\t%1, %0" : "+q"(c): "r"(b));
>   return c;
> }
> $ gcc foo.c
> foo.c: Assembler messages:
> foo.c:2: Error: `%rax' not allowed with `orb'
>
> Use the `%b` "x86 Operand Modifier" to instead force register allocation
> to select a lower-8-bit GPR operand.
>
> The "q" constraint only has meaning on -m32 otherwise is treated as
> "r". Not all GPRs have low-8-bit aliases for -m32.
>
> Cc: Jesse Brandeburg <jesse.brandeburg@xxxxxxxxx>
> Link: https://github.com/ClangBuiltLinux/linux/issues/961
> Link: https://lore.kernel.org/lkml/20200504193524.GA221287@xxxxxxxxxx/
> Link: https://gcc.gnu.org/onlinedocs/gcc/Extended-Asm.html#x86Operandmodifiers
> Fixes: 1651e700664b4 ("x86: Fix bitops.h warning with a moved cast")
> Reported-by: Sedat Dilek <sedat.dilek@xxxxxxxxx>
> Reported-by: kernelci.org bot <bot@xxxxxxxxxxxx>
> Suggested-by: Andy Shevchenko <andriy.shevchenko@xxxxxxxxx>
> Suggested-by: Brian Gerst <brgerst@xxxxxxxxx>
> Suggested-by: H. Peter Anvin <hpa@xxxxxxxxx>
> Suggested-by: Ilie Halip <ilie.halip@xxxxxxxxx>
> Signed-off-by: Nick Desaulniers <ndesaulniers@xxxxxxxxxx>

Reviewed-By: Brian Gerst <brgerst@xxxxxxxxx>