On Mon, Jan 22, 2018 at 12:49:12PM +0100, Martin Kletzander wrote:
On Thu, Jan 18, 2018 at 10:16:55AM +0100, Ján Tomko wrote:
> After the latest CPU additions, the build fails with clang:
> cputest.c:905:1: error: stack frame size of 26136 bytes
> in function 'mymain' [-Werror,-Wframe-larger-than=]
>
> Raise the relaxed limit which is used for tests.
> ---
> m4/virt-compile-warnings.m4 | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> Pushed as a build breaker fix
>
> diff --git a/m4/virt-compile-warnings.m4 b/m4/virt-compile-warnings.m4
> index f18a08a8f..b9c974842 100644
> --- a/m4/virt-compile-warnings.m4
> +++ b/m4/virt-compile-warnings.m4
> @@ -200,7 +200,7 @@ AC_DEFUN([LIBVIRT_COMPILE_WARNINGS],[
> # but using 1024 bytes sized buffers (mostly for virStrerror)
> # stops us from going down further
> gl_WARN_ADD([-Wframe-larger-than=4096], [STRICT_FRAME_LIMIT_CFLAGS])
> - gl_WARN_ADD([-Wframe-larger-than=25600], [RELAXED_FRAME_LIMIT_CFLAGS])
> + gl_WARN_ADD([-Wframe-larger-than=32768], [RELAXED_FRAME_LIMIT_CFLAGS])
>
Remind me again why don't we do -Wno-frame-larger-than (or something to that
effect) for tests? Is it just because "We should fix it at some point"? I
can't really recall the reasoning behind that (and if it is still valid) even
though I already asked for it.
I don't think there's a strong reason, given the way we currently write
tests with huge amounts of stack variables.
For -Wframe-larger-than to be useful, we'd need to move all the big data
blobs to be static, global variables.
Regards,
Daniel
--
|:
https://berrange.com -o-
https://www.flickr.com/photos/dberrange :|
|:
https://libvirt.org -o-
https://fstop138.berrange.com :|
|:
https://entangle-photo.org -o-
https://www.instagram.com/dberrange :|