[LTP] [PATCH] syscalls/statvfs01: Correcly zero terminate the strings

[Martin Doucha]

Hi,

On 18. 01. 23 14:36, Cyril Hrubis wrote:
> Fixes random failures caused by the fact that the stack is not
> guaranteed to be zeroed.
> 
> Fixes: e305ac4a305f ("statvfs01: Convert to new LTP API")
> Signed-off-by: Cyril Hrubis <chrubis@suse.cz>
> CC: Petr Vorel <pvorel@suse.cz>
> CC: Richard Palethorpe <rpalethorpe@suse.com>
> CC: Avinesh Kumar <akumar@suse.de>
> ---
>   testcases/kernel/syscalls/statvfs/statvfs01.c | 3 +++
>   1 file changed, 3 insertions(+)
> 
> diff --git a/testcases/kernel/syscalls/statvfs/statvfs01.c b/testcases/kernel/syscalls/statvfs/statvfs01.c
> index dd14d6a0e..f17dc4cfa 100644
> --- a/testcases/kernel/syscalls/statvfs/statvfs01.c
> +++ b/testcases/kernel/syscalls/statvfs/statvfs01.c
> @@ -38,6 +38,9 @@ static void run(void)
>   	memset(valid_fname, 'a', valid_len);
>   	memset(toolong_fname, 'b', valid_len + 1);
>   
> +	valid_fname[valid_len] = 0;
> +	toolong_fname[valid_len+1] = 0;

Is there a possibility that valid_len could be equal to PATH_MAX-1? I 
think toolong_fname[] size should be bumped to at least PATH_MAX+1.

> +
>   	TST_EXP_FD(creat(valid_fname, 0444));
>   	SAFE_CLOSE(TST_RET);
>   

-- 
Martin Doucha   mdoucha@suse.cz
QA Engineer for Software Maintenance
SUSE LINUX, s.r.o.
CORSO IIa
Krizikova 148/34
186 00 Prague 8
Czech Republic