From: Jan Kara <jack@suse.cz>
To: Christoph Hellwig <hch@infradead.org>
Cc: LKML <linux-kernel@vger.kernel.org>,
linux-mm@kvack.org, linux-fsdevel@vger.kernel.org,
npiggin@suse.de
Subject: Re: [PATCH 0/10] Fix page_mkwrite() for blocksize < pagesize (version 3)
Date: Tue, 16 Jun 2009 16:42:17 +0200 [thread overview]
Message-ID: <20090616144217.GA18063@duck.suse.cz> (raw)
In-Reply-To: <20090616143424.GA22002@infradead.org>
[-- Attachment #1: Type: text/plain, Size: 739 bytes --]
On Tue 16-06-09 10:34:24, Christoph Hellwig wrote:
> On Mon, Jun 15, 2009 at 07:59:47PM +0200, Jan Kara wrote:
> >
> > patches below are an attempt to solve problems filesystems have with
> > page_mkwrite() when blocksize < pagesize (see the changelog of the second patch
> > for details).
>
> It would be useful if you had a test case reproducing these issues,
> so that I can verify how well your patches work in various scenarios.
Good point, I should have mentioned in the changelog: fsx-linux is able
to trigger the problem quite quickly.
I have also written a simple program for initial testing of the fix
(works only for 1K blocksize and 4K pagesize) - it's attached.
Honza
--
Jan Kara <jack@suse.cz>
SUSE Labs, CR
[-- Attachment #2: test-mkwrite.c --]
[-- Type: text/x-c++src, Size: 1242 bytes --]
#define _XOPEN_SOURCE 500
#define _GNU_SOURCE
#include <unistd.h>
#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#include <fcntl.h>
#include <sys/mman.h>
#define MMAP_THREADS 4
void map_file(int fd)
{
char *addr = mmap(NULL, 4096, PROT_READ | PROT_WRITE, MAP_SHARED, fd, 0);
int i;
if (addr == MAP_FAILED) {
perror("mmap");
exit(1);
}
while (1) {
for (i = 0; i < 128; i++)
addr[i + 1024] = 'c';
}
}
char buf2[4096] __attribute__ ((aligned (4096)));
int main(int argc, char **argv)
{
int i;
int ret;
int fd;
char buf[1024];
if (argc != 2) {
printf("Usage: test-mkwrite <file>\n");
return 1;
}
fd = open(argv[1], O_RDWR | O_CREAT | O_TRUNC, 0644);
if (fd < 0) {
perror("open");
return 1;
}
memset(buf, 'a', sizeof(buf));
pwrite(fd, buf, sizeof(buf), 0);
for (i = 0; i < MMAP_THREADS; i++) {
ret = fork();
if (ret < 0) {
perror("fork");
return 1;
} else if (ret == 0)
map_file(fd);
}
close(fd);
memset(buf2, 'b', sizeof(buf2));
fd = open(argv[1], O_RDWR | O_DIRECT);
if (fd < 0) {
perror("dopen");
return 1;
}
while (1) {
if (pwrite(fd, buf2, 4096, 4096) < 0)
perror("pwrite");
usleep(10000000);
ftruncate(fd, 1024);
usleep(10000000);
}
return 0;
}
next prev parent reply other threads:[~2009-06-16 14:42 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-06-15 17:59 [PATCH 0/10] Fix page_mkwrite() for blocksize < pagesize (version 3) Jan Kara
2009-06-15 17:59 ` [PATCH 01/11] ext3: Get rid of extenddisksize parameter of ext3_get_blocks_handle() Jan Kara
2009-06-17 10:28 ` Nick Piggin
2009-06-17 11:49 ` Jan Kara
2009-06-15 17:59 ` [PATCH 02/11] vfs: Add better VFS support for page_mkwrite when blocksize < pagesize Jan Kara
2009-06-25 16:17 ` Nick Piggin
2009-06-25 16:43 ` Nick Piggin
2009-06-25 17:47 ` Christoph Hellwig
2009-06-26 8:42 ` Nick Piggin
2009-06-30 17:37 ` Christoph Hellwig
2009-07-02 7:22 ` Nick Piggin
2009-07-04 15:18 ` Christoph Hellwig
2009-07-06 9:08 ` Nick Piggin
2009-07-06 10:35 ` Christoph Hellwig
2009-07-06 11:49 ` Nick Piggin
2009-06-26 12:21 ` Jan Kara
2009-06-26 12:55 ` Nick Piggin
2009-06-26 16:08 ` Jan Kara
2009-06-29 5:54 ` Nick Piggin
2009-06-15 17:59 ` [PATCH 03/11] ext2: Allocate space for mmaped file on page fault Jan Kara
2009-06-15 17:59 ` [PATCH 04/11] ext4: Make sure blocks are properly allocated under mmaped page even when blocksize < pagesize Jan Kara
2009-06-15 17:59 ` [PATCH 05/11] ext3: Allocate space for mmaped file on page fault Jan Kara
2009-06-15 17:59 ` [PATCH 06/11] vfs: Implement generic per-cpu counters for delayed allocation Jan Kara
2009-06-15 17:59 ` [PATCH 07/11] vfs: Unmap underlying metadata of new data buffers only when buffer is mapped Jan Kara
2009-06-17 10:35 ` Nick Piggin
2009-06-17 12:05 ` Jan Kara
2009-06-17 13:53 ` Nick Piggin
2009-06-18 12:00 ` Theodore Tso
2009-06-18 11:51 ` OGAWA Hirofumi
2009-06-15 17:59 ` [PATCH 08/11] fs: Don't clear dirty bits in block_write_full_page() Jan Kara
2009-06-15 17:59 ` [PATCH 09/11] vfs: Export wakeup_pdflush Jan Kara
2009-06-15 17:59 ` [PATCH 10/11] ext3: Implement delayed allocation on page_mkwrite time Jan Kara
2009-06-15 18:02 ` [PATCH 0/10] Fix page_mkwrite() for blocksize < pagesize (version 3) Jan Kara
2009-06-15 18:17 ` Aneesh Kumar K.V
2009-06-16 10:28 ` Jan Kara
2009-06-16 14:34 ` Christoph Hellwig
2009-06-16 14:42 ` Jan Kara [this message]
2009-06-30 17:44 ` Christoph Hellwig
2009-07-01 10:29 ` Aneesh Kumar K.V
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=20090616144217.GA18063@duck.suse.cz \
--to=jack@suse.cz \
--cc=hch@infradead.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=npiggin@suse.de \
/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;
as well as URLs for NNTP newsgroup(s).