[LTP] [PATCH] fanotify: the mount fs_type should be given

[LTP] [PATCH] fanotify: the mount fs_type should be given

[Li Wang]

This commmit 28507e514c(safe_mount: Do not try mount() syscall for FUSE fs)
involves FUSE fs check in safe_mount(), so we should give the "fs_type" when
calling that in case the system kill our program.

  cmdline="fanotify06"
  contacts=""
  analysis=exit
  <<<test_output>>>
  tst_test.c:980: INFO: Timeout per run is 0h 10m 00s
  tst_test.c:1025: BROK: Test killed by SIGSEGV!

Signed-off-by: Li Wang <liwang@redhat.com>
---
 testcases/kernel/syscalls/fanotify/fanotify06.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/testcases/kernel/syscalls/fanotify/fanotify06.c b/testcases/kernel/syscalls/fanotify/fanotify06.c
index e63e457..8cbe1ad 100644
--- a/testcases/kernel/syscalls/fanotify/fanotify06.c
+++ b/testcases/kernel/syscalls/fanotify/fanotify06.c
@@ -221,7 +221,7 @@ void test01(void)
 static void setup(void)
 {
 	SAFE_MKDIR(MOUNT_NAME, 0755);
-	SAFE_MOUNT(MOUNT_NAME, MOUNT_NAME, NULL, MS_BIND, NULL);
+	SAFE_MOUNT(MOUNT_NAME, MOUNT_NAME, "none", MS_BIND, NULL);
 	mount_created = 1;
 	SAFE_CHDIR(MOUNT_NAME);
 
-- 
2.9.3

[LTP] [PATCH] fanotify: the mount fs_type should be given

[Jan Stancek]

----- Original Message -----
> This commmit 28507e514c(safe_mount: Do not try mount() syscall for FUSE fs)
> involves FUSE fs check in safe_mount(), so we should give the "fs_type" when
> calling that in case the system kill our program.

Shouldn't we rather change is_fuse() to return 0 if fs_type is NULL?

> 
>   cmdline="fanotify06"
>   contacts=""
>   analysis=exit
>   <<<test_output>>>
>   tst_test.c:980: INFO: Timeout per run is 0h 10m 00s
>   tst_test.c:1025: BROK: Test killed by SIGSEGV!
> 
> Signed-off-by: Li Wang <liwang@redhat.com>
> ---
>  testcases/kernel/syscalls/fanotify/fanotify06.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/testcases/kernel/syscalls/fanotify/fanotify06.c
> b/testcases/kernel/syscalls/fanotify/fanotify06.c
> index e63e457..8cbe1ad 100644
> --- a/testcases/kernel/syscalls/fanotify/fanotify06.c
> +++ b/testcases/kernel/syscalls/fanotify/fanotify06.c
> @@ -221,7 +221,7 @@ void test01(void)
>  static void setup(void)
>  {
>  	SAFE_MKDIR(MOUNT_NAME, 0755);
> -	SAFE_MOUNT(MOUNT_NAME, MOUNT_NAME, NULL, MS_BIND, NULL);
> +	SAFE_MOUNT(MOUNT_NAME, MOUNT_NAME, "none", MS_BIND, NULL);
>  	mount_created = 1;
>  	SAFE_CHDIR(MOUNT_NAME);
>  
> --
> 2.9.3
> 
> 

[LTP] [PATCH] fanotify: the mount fs_type should be given

[Cyril Hrubis]

Hi!
> > This commmit 28507e514c(safe_mount: Do not try mount() syscall for FUSE fs)
> > involves FUSE fs check in safe_mount(), so we should give the "fs_type" when
> > calling that in case the system kill our program.
> 
> Shouldn't we rather change is_fuse() to return 0 if fs_type is NULL?

Agreed.

-- 
Cyril Hrubis
chrubis@suse.cz

[LTP] [PATCH] fanotify: the mount fs_type should be given

[Li Wang]

On Tue, Jan 23, 2018 at 4:43 PM, Cyril Hrubis <chrubis@suse.cz> wrote:

> Hi!
> > > This commmit 28507e514c(safe_mount: Do not try mount() syscall for
> FUSE fs)
> > > involves FUSE fs check in safe_mount(), so we should give the
> "fs_type" when
> > > calling that in case the system kill our program.
> >
> > Shouldn't we rather change is_fuse() to return 0 if fs_type is NULL?
>

​Ok, that sounds better, let me re-write this.

Thanks!​



>
> Agreed.
>
> --
> Cyril Hrubis
> chrubis@suse.cz
>



-- 
Li Wang
liwang@redhat.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20180123/5196ec7f/attachment.html>