From: Dave Chinner <david@fromorbit.com>
To: "Zhong, Xin" <xin.zhong@intel.com>
Cc: xfs@oss.sgi.com
Subject: Re: [PATCH v2] xfstests: pwrite hang when writing from mmaped buffer of the same page
Date: Tue, 14 Dec 2010 15:42:59 +1100 [thread overview]
Message-ID: <20101214044259.GA16267@dastard> (raw)
In-Reply-To: <1292295652-9508-1-git-send-email-xin.zhong@intel.com>
On Tue, Dec 14, 2010 at 11:00:52AM +0800, Zhong, Xin wrote:
> The problem is found in meego testing on btrfs:
> http://bugs.meego.com/show_bug.cgi?id=6672
>
> Reviewed-by: Hellwig, Christoph <hch@infradead.org>
> Signed-off-by: Zhong, Xin <xin.zhong@intel.com>
> ---
> 248 | 58 +++++++++++++++++++++++++++++++++
> 248.out | 2 +
> group | 1 +
> src/Makefile | 2 +-
> src/pwrite_mmap_blocked.c | 78 +++++++++++++++++++++++++++++++++++++++++++++
> 5 files changed, 140 insertions(+), 1 deletions(-)
> create mode 100755 248
> create mode 100644 248.out
> create mode 100644 src/pwrite_mmap_blocked.c
>
> diff --git a/248 b/248
> new file mode 100755
> index 0000000..a4664d9
> --- /dev/null
> +++ b/248
> @@ -0,0 +1,58 @@
> +#! /bin/bash
> +# FS QA Test No. 248
> +#
> +# Test for pwrite hang problem when writing from mmaped buffer of the same page
> +#
> +#-----------------------------------------------------------------------
> +# Copyright (c) 2010 YOUR NAME HERE. All Rights Reserved.
^^^^^^^^^^^^^^
Needs updating, I think. ;)
While I'm here, a two minute review...
....
> +_cleanup()
> +{
> + cd /
> + rm -f $tmp.*
> +}
> +
> +# get standard environment, filters and checks
> +. ./common.rc
> +. ./common.filter
> +
> +# real QA test starts here
> +
> +# Modify as appropriate.
> +_supported_fs generic
> +_supported_os Linux
> +
> +TESTFILE=$TEST_DIR/test_file
> +TEST_PROG=$here/src/pwrite_mmap_blocked
> +
> +$TEST_PROG $TESTFILE
> +rm -f $TESTFILE
This rm should be in the _cleanup() function so that it is removed
even when the test is aborted.
.....
> diff --git a/group b/group
> index 0f94dd9..ddda26c 100644
> --- a/group
> +++ b/group
> @@ -361,3 +361,4 @@ deprecated
> 245 auto quick dir
> 246 auto quick rw
> 247 auto quick rw
> +248 other quick rw
Why not the auto group?
> +char *progname;
> +loff_t size;
> +int fd;
> +void *mapped_mem;
None of these need to be global.
> +
> +
> +int main(int argc, char *argv[])
> +{
> + int ret;
> +
> + progname = argv[0];
> + size = 5;
> + fd = open(argv[1], O_RDWR|O_TRUNC|O_CREAT, 0666);
> + if (fd < 0) {
> + fprintf(stderr, "%s: Cannot open `%s': %s\n",
> + progname, argv[1], strerror(errno));
> + exit(1);
> + }
> +
> + char *cc = "01234";
Don't mix declarations in the the code. Better would be:
int ret;
char *cc = "01234";
progname = argv[0];
size = strlen(cc);
....
> + if ((ret = pwrite(fd, (const char *)cc,
> + size, 0)) != size) {
ret = pwrite(fd, cc, size, 0);
if (ret != size) {
> + fprintf(stderr, "%s: pwrite returned %d\n",
> + __FUNCTION__, ret);
> + perror("pwrite");
> + exit(1);
> + }
No need for both fprintf and perror calls.
> +
> + mapped_mem = mmap(0, size, PROT_READ|PROT_WRITE, MAP_SHARED, fd, 0);
> + if (mapped_mem == MAP_FAILED) {
> + perror("mmap");
> + exit(1);
> + }
> + loff_t amount = 1;
> + loff_t from = 2;
> + loff_t to = 3;
> + printf("pwrite %Ld bytes from %Ld to %Ld\n", amount, from, to);
> + if ((ret = pwrite(fd, (char *)mapped_mem + from, amount, to)) != amount) {
ret = pwrite(fd, (char *)mapped_mem + from, amount, to));
if (ret != amount) {
> + printf("%s: pwrite returned %d, not %Ld\n",
> + __FUNCTION__, ret, amount);
> + if (errno == EFAULT) {
> + printf("pwrite: EFAULT\n");
> + } else if (ret < 0) {
> + perror("pwrite");
> + exit(1);
> + }
A single call to perror() will be sufficient here...
> + }
> + munmap(mapped_mem,0);
> + close(fd);
> + exit(0);
> +}
Cheers,
Dave.
--
Dave Chinner
david@fromorbit.com
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
prev parent reply other threads:[~2010-12-14 4:41 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-12-14 3:00 [PATCH v2] xfstests: pwrite hang when writing from mmaped buffer of the same page Zhong, Xin
2010-12-14 4:42 ` Dave Chinner [this message]
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=20101214044259.GA16267@dastard \
--to=david@fromorbit.com \
--cc=xfs@oss.sgi.com \
--cc=xin.zhong@intel.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox