All of lore.kernel.org
 help / color / mirror / Atom feed
From: Richard Palethorpe <rpalethorpe@suse.de>
To: "xuyang2018.jy@fujitsu.com" <xuyang2018.jy@fujitsu.com>
Cc: "ltp@lists.linux.it" <ltp@lists.linux.it>
Subject: Re: [LTP] [PATCH v2] syscalls/statx01: Add exit condition when parsing /proc/self/mountinfo
Date: Thu, 15 Dec 2022 09:58:18 +0000	[thread overview]
Message-ID: <87y1r9oujj.fsf@suse.de> (raw)
In-Reply-To: <ba8c6bd3-1a74-ec37-4ab5-f299975bd823@fujitsu.com>

Hello,

"xuyang2018.jy@fujitsu.com" <xuyang2018.jy@fujitsu.com> writes:

> Hi Richard
>
>
>> Hello,
>> 
>> Yang Xu <xuyang2018.jy@fujitsu.com> writes:
>> 
>>> When using user filesystem such as overlayfs, the current parsing way can't
>>> work well.
>>>
>>> 63 66 8:3 / /sysroot rw,relatime - ext4 /dev/sda3 rw,seclabel
>>> 43 66 8:3 /ostree/deploy/rhivos/var /var rw,relatime shared:3 - ext4 /dev/sda3 rw,seclabel
>>>
>>> So add the exit condition for statx.mnt_id check so it can skip the
>>> underflying filesystem and parse the correct user fileystem's mnt_id.
>>>
>>> Fixes: #1001
>>> Signed-off-by: Yang Xu <xuyang2018.jy@fujitsu.com>
>>> ---
>>> v1-v2: change the fail message
>>>   testcases/kernel/syscalls/statx/statx01.c | 7 ++++---
>>>   1 file changed, 4 insertions(+), 3 deletions(-)
>>>
>>> diff --git a/testcases/kernel/syscalls/statx/statx01.c b/testcases/kernel/syscalls/statx/statx01.c
>>> index 60b50958b..e9677475a 100644
>>> --- a/testcases/kernel/syscalls/statx/statx01.c
>>> +++ b/testcases/kernel/syscalls/statx/statx01.c
>>> @@ -68,7 +68,8 @@ static void test_mnt_id(struct statx *buf)
>>>   		if (sscanf(line, "%"SCNu64" %*d %d:%d", &mnt_id, &line_mjr, &line_mnr) != 3)
>>>   			continue;
>>>   
>>> -		if (line_mjr == buf->stx_dev_major && line_mnr == buf->stx_dev_minor)
>>> +		if (line_mjr == buf->stx_dev_major && line_mnr == buf->stx_dev_minor &&
>>> +				mnt_id == buf->stx_mnt_id)
>>>   			break;
>>>   	}
>>>   
>>> @@ -80,8 +81,8 @@ static void test_mnt_id(struct statx *buf)
>>>   			mnt_id);
>>>   	else
>>>   		tst_res(TFAIL,
>>> -			"statx.stx_mnt_id(%"PRIu64") is different from mount_id(%"PRIu64") in /proc/self/mountinfo",
>>> -			(uint64_t)buf->stx_mnt_id, mnt_id);
>>> +			"statx.stx_mnt_id(%"PRIu64") doesn't exist in /proc/self/mountinfo",
>>> +			(uint64_t)buf->stx_mnt_id);
>> 
>> The mnt_id may exist in mountinfo, but not the triple (mnt_id,
>> dev_major, dev_minor). So really we should print all three here (unless
>> we already display that somewhere else).
>> 
>
> Yes, so how about the following changes:
>
> diff --git a/testcases/kernel/syscalls/statx/statx01.c 
> b/testcases/kernel/syscalls/statx/statx01.c
> index 60b50958b..8f3b01b32 100644
> --- a/testcases/kernel/syscalls/statx/statx01.c
> +++ b/testcases/kernel/syscalls/statx/statx01.c
> @@ -53,7 +53,7 @@ static void test_mnt_id(struct statx *buf)
>   {
>          FILE *file;
>          char line[PATH_MAX];
> -       int pid;
> +       int pid, flag = 0;
>          unsigned int line_mjr, line_mnr;
>          uint64_t mnt_id;
>
> @@ -68,20 +68,26 @@ static void test_mnt_id(struct statx *buf)
>                  if (sscanf(line, "%"SCNu64" %*d %d:%d", &mnt_id, 
> &line_mjr, &line_mnr) != 3)
>                          continue;
>
> -               if (line_mjr == buf->stx_dev_major && line_mnr == 
> buf->stx_dev_minor)
> -                       break;
> +               if (line_mjr == buf->stx_dev_major && line_mnr == 
> buf->stx_dev_minor) {
> +                       if (buf->stx_mnt_id == mnt_id) {
> +                               flag = 1;
> +                               break
> +                       }
> +                       tst_res(TINFO, "%s doesn't contain (%"PRIu64") 
> %d:%d",
> +                               line, (uint64_t)buf->stx_mnt_id, 
> buf->stx_dev_major, buf->stx_dev_minor);
> +               }
>          }
>
>          SAFE_FCLOSE(file);
>
> -       if (buf->stx_mnt_id == mnt_id)
> +       if (flag)
>                  tst_res(TPASS,
>                          "statx.stx_mnt_id equals to mount_id(%"PRIu64") 
> in /proc/self/mountinfo",
>                          mnt_id);
>          else
>                  tst_res(TFAIL,
> -                       "statx.stx_mnt_id(%"PRIu64") is different from 
> mount_id(%"PRIu64") in /proc/self/mountinfo",
> -                       (uint64_t)buf->stx_mnt_id, mnt_id);
> +                       "statx.stx_mnt_id(%"PRIu64") doesn't exist in 
> /proc/self/mountinfo",
> +                       (uint64_t)buf->stx_mnt_id);

Looks good, (except for the line wrap). With that

Reviewed-by: Richard Palethorpe <rpalethorpe@suse.com>

>
>          pid = getpid();
>          snprintf(line, PATH_MAX, "/proc/%d/fdinfo/%d", pid, file_fd);
>
> Best Regards
> Yang Xu
>>>   
>>>   	pid = getpid();
>>>   	snprintf(line, PATH_MAX, "/proc/%d/fdinfo/%d", pid, file_fd);
>>> -- 
>>> 2.27.0
>> 
>> 


-- 
Thank you,
Richard.

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

  reply	other threads:[~2022-12-15 10:01 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-12-13  3:14 [LTP] [PATCH v2] syscalls/statx01: Add exit condition when parsing /proc/self/mountinfo Yang Xu
2022-12-13 10:59 ` Richard Palethorpe
2022-12-15  7:14   ` xuyang2018.jy
2022-12-15  9:58     ` Richard Palethorpe [this message]
2022-12-16  2:05       ` xuyang2018.jy

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=87y1r9oujj.fsf@suse.de \
    --to=rpalethorpe@suse.de \
    --cc=ltp@lists.linux.it \
    --cc=xuyang2018.jy@fujitsu.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.