On Thu, Jan 28, 2016 at 02:31:58PM +1100, Michael Ellerman wrote:
>
> Looking at GCC history it looks like the fix is in 4.9.0 and anything later.
Good. But 4.8.5 has a buggy -mprofile-kernel, and there will be no 4.8.6, Bad.
> But a version check doesn't work with patched distro/vendor toolchains. So we
> probably need some sort of runtime check.
Agreed.
/bin/echo -e '#include <linux/compiler.h>\nnotrace int func() { return 0; }' |
gcc -D__KERNEL__ -Iinclude -p -mprofile-kernel -x c -O2 - -S -o - | grep mcount
should be empty. If it yields "bl _mcount" your compiler is buggy.
I haven't looked at the kernel's "autoconf" yet, but it's probably capable
of testing this.
Torsten
--
To unsubscribe from this list: send the line "unsubscribe live-patching" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
