Re: [PATCH] selftests/filesystems:fix build error in overlayfs

From: Günther Noack
Date: Mon Jan 15 2024 - 09:26:45 EST


Hello!

On Fri, Jan 12, 2024 at 03:40:59PM +0800, Hu Yadi wrote:
> One build issue comes up due to both mount.h included dev_in_maps.c
>
> In file included from dev_in_maps.c:10:
> /usr/include/sys/mount.h:35:3: error: expected identifier before numeric constant
> 35 | MS_RDONLY = 1, /* Mount read-only. */
> | ^~~~~~~~~
> In file included from dev_in_maps.c:13:
>
> Remove one of them to solve conflict, another error comes up:
>
> dev_in_maps.c:170:6: error: implicit declaration of function ‘mount’ [-Werror=implicit-function-declaration]
> 170 | if (mount(NULL, "/", NULL, MS_SLAVE | MS_REC, NULL) == -1) {
> | ^~~~~
> cc1: all warnings being treated as errors
>
> and then , add sys_mount definition to solve it
> After both above, dev_in_maps.c can be built correctly on my mache(gcc 102,glibc-2.32,kernel-5.10)

This is apparently the same error as in
https://lore.kernel.org/all/11cdac1e-e96c-405f-63e8-35b0e2926337@xxxxxxx/

I'm getting the impression that we are fixing the issue at the wrong layer here?
After all, the mount() syscall is supposed to be used with <sys/mount.h>
according to the mount(2) man page? It feels a bit like cheating to resort to
sys_mount() instead...?

Do you have any deeper thoughts on what could be the underlying issue here?
With my newer GCC toolchains, I have been unable to reproduce this.

Thanks,
—Günther