From: Cyril Hrubis <chrubis@suse.cz>
To: ltp@lists.linux.it
Subject: [LTP] [PATCH 2/2] syscalls/lseek11.c: fix undefined syncfs() && SEEK_DATA
Date: Wed, 12 Apr 2017 13:42:25 +0200 [thread overview]
Message-ID: <20170412114225.GA25790@rei> (raw)
In-Reply-To: <1491968042-5173-2-git-send-email-yangx.jy@cn.fujitsu.com>
Hi!
> 1) Compilation failed on RHEL6.9GA because syncfs() was not defined.
> this function was introduced since linux 2.6.39, so we could use
> sync() instead of syncfs().
Looking at this again, why don't we use fsync(fd) instead?
> 2) We should Add a check if SEEK_DATA was implemented in get_blocksize().
Missed the lseek() in get_blocksize(), sorry.
> Signed-off-by: Xiao Yang <yangx.jy@cn.fujitsu.com>
> ---
> testcases/kernel/syscalls/lseek/lseek11.c | 17 ++++++++++++-----
> 1 file changed, 12 insertions(+), 5 deletions(-)
>
> diff --git a/testcases/kernel/syscalls/lseek/lseek11.c b/testcases/kernel/syscalls/lseek/lseek11.c
> index fe226b5..01b9fa4 100644
> --- a/testcases/kernel/syscalls/lseek/lseek11.c
> +++ b/testcases/kernel/syscalls/lseek/lseek11.c
> @@ -116,8 +116,15 @@ static void get_blocksize(void)
> offset <<= 1;
> SAFE_FTRUNCATE(fd, 0);
> SAFE_PWRITE(1, fd, "a", 1, offset);
> - syncfs(fd);
> - pos = SAFE_LSEEK(fd, 0, SEEK_DATA);
> + sync();
> + pos = lseek(fd, 0, SEEK_DATA);
> + if (pos == -1) {
> + if (errno == EINVAL) {
> + tst_brk(TCONF | TERRNO,
> + "SEEK_DATA not implemented");
> + }
> + tst_brk(TBROK | TERRNO, "SEEK_DATA failed");
> + }
> }
>
> /* bisect for double check */
> @@ -125,7 +132,7 @@ static void get_blocksize(void)
> while (shift && offset < (st.st_blksize * 2)) {
> SAFE_FTRUNCATE(fd, 0);
> SAFE_PWRITE(1, fd, "a", 1, offset);
> - syncfs(fd);
> + sync();
> pos = SAFE_LSEEK(fd, 0, SEEK_DATA);
> offset += pos ? -shift : shift;
> shift >>= 1;
> @@ -178,7 +185,7 @@ static void setup(void)
> if (lseek(fd, 0, SEEK_HOLE) < 0) {
> if (errno == EINVAL) {
> tst_brk(TCONF | TERRNO,
> - "SEEK_DATA and SEEK_HOLE not implemented");
> + "SEEK_HOLE not implemented");
> }
> tst_brk(TBROK | TERRNO, "SEEK_HOLE failed");
> }
We should get rid of this whole lseek() test in setup() since we call
tst_brk() in the get_blocksize() the lseek() will never fail here.
> @@ -192,7 +199,7 @@ static void setup(void)
> SAFE_LSEEK(fd, -128, SEEK_END);
> write_data(fd, i + 1);
>
> - syncfs(fd);
> + sync();
> SAFE_LSEEK(fd, 0, SEEK_SET);
> }
>
> --
> 1.8.3.1
>
>
>
>
> --
> Mailing list info: https://lists.linux.it/listinfo/ltp
--
Cyril Hrubis
chrubis@suse.cz
next prev parent reply other threads:[~2017-04-12 11:42 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-04-12 3:34 [LTP] [PATCH 1/2] lapi/seek.h: fix typo for SEEK_DATA Xiao Yang
2017-04-12 3:34 ` [LTP] [PATCH 2/2] syscalls/lseek11.c: fix undefined syncfs() && SEEK_DATA Xiao Yang
2017-04-12 11:42 ` Cyril Hrubis [this message]
2017-04-13 1:29 ` Xiao Yang
2017-04-13 2:31 ` [LTP] [PATCH v2] syscalls/lseek11.c: fix " Xiao Yang
2017-04-13 10:04 ` Cyril Hrubis
2017-04-12 6:55 ` [LTP] [PATCH 1/2] lapi/seek.h: fix typo for SEEK_DATA Cyril Hrubis
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=20170412114225.GA25790@rei \
--to=chrubis@suse.cz \
--cc=ltp@lists.linux.it \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox