On Fri, 2020-08-14 at 17:24 -0700, Nick Desaulniers wrote:
> LLVM implemented a recent "libcall optimization" that lowers calls to
> `sprintf(dest, "%s", str)` where the return value is used to
> `stpcpy(dest, str) - dest`. This generally avoids the machinery involved
> in parsing format strings.
[]
> diff --git a/include/linux/string.h b/include/linux/string.h
[]
> @@ -31,6 +31,9 @@ size_t strlcpy(char *, const char *, size_t);
> #ifndef __HAVE_ARCH_STRSCPY
> ssize_t strscpy(char *, const char *, size_t);
> #endif
> +#ifndef __HAVE_ARCH_STPCPY
> +extern char *stpcpy(char *__restrict, const char *__restrict__);
Why use two different forms for __restrict and __restrict__?
Any real reason to use __restrict__ at all?
It's used nowhere else in the kernel.
$ git grep -w -P '__restrict_{0,2}'
scripts/genksyms/keywords.c: // According to rth, c99 defines "_Bool", __restrict", __restrict__", "restrict". KAO
scripts/genksyms/keywords.c: { "__restrict__", RESTRICT_KEYW },
