From: Ying Han <yinghan@google.com>
To: linux-kernel <linux-kernel@vger.kernel.org>,
linux-mm <linux-mm@kvack.org>,
Andrew Morton <akpm@linux-foundation.org>,
guichaz@gmail.com, Alex Khesin <alexk@google.com>,
Mike Waychison <mikew@google.com>,
Rohit Seth <rohitseth@google.com>
Subject: ftruncate-mmap: pages are lost after writing to mmaped file.
Date: Wed, 18 Mar 2009 12:44:08 -0700 [thread overview]
Message-ID: <604427e00903181244w360c5519k9179d5c3e5cd6ab3@mail.gmail.com> (raw)
We triggered the failure during some internal experiment with
ftruncate/mmap/write/read sequence. And we found that some pages are
"lost" after writing to the mmaped file. which in the following test
cases (count >= 0).
First we deployed the test cases into group of machines and see about
>20% failure rate on average. Then, I did couple of experiment to try
to reproduce it on a single machine. what i found is that:
1. add a fsync after write the file, i can not reproduce this issue.
2. add memory pressure(mmap/mlock) while run the test in infinite
loop, the failure is reproduced quickly. ( background flushing ? )
The "bad pages" count differs each time from one digit to 4,5 digit
for 128M ftruncated file. and what i also found that the bad page
number are contiguous for each segment which total bad pages container
several segments. ext "1-4, 9-20, 48-50" ( batch flushing ? )
(The failure is reproduced based on 2.6.29-rc8, also happened on
2.6.18 kernel. . Here is the simple test case to reproduce it with
memory pressure. )
#include <sys/mman.h>
#include <sys/types.h>
#include <fcntl.h>
#include <unistd.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
long kMemSize = 128 << 20;
int kPageSize = 4096;
int main(int argc, char **argv) {
int status;
int count = 0;
int i;
char *fname = "/root/test.mmap";
char *mem;
unlink(fname);
int fd = open(fname, O_CREAT | O_EXCL | O_RDWR, 0600);
status = ftruncate(fd, kMemSize);
mem = mmap(0, kMemSize, PROT_READ | PROT_WRITE, MAP_SHARED, fd, 0);
// Fill the memory with 1s.
memset(mem, 1, kMemSize);
for (i = 0; i < kMemSize; i++) {
int byte_good = mem[i] != 0;
if (!byte_good && ((i % kPageSize) == 0)) {
//printf("%d ", i / kPageSize);
count++;
}
}
munmap(mem, kMemSize);
close(fd);
unlink(fname);
if (count > 0) {
printf("Running %d bad page\n", count);
return 1;
}
return 0;
}
--Ying
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next reply other threads:[~2009-03-18 19:44 UTC|newest]
Thread overview: 60+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-03-18 19:44 Ying Han [this message]
2009-03-18 22:11 ` ftruncate-mmap: pages are lost after writing to mmaped file Andrew Morton
2009-03-18 22:40 ` Linus Torvalds
2009-03-18 23:18 ` Ying Han
2009-03-18 23:36 ` Linus Torvalds
2009-03-18 23:54 ` Ying Han
2009-03-19 15:48 ` Nick Piggin
2009-03-19 16:16 ` Peter Zijlstra
2009-03-19 16:36 ` Nick Piggin
2009-03-19 16:20 ` Linus Torvalds
2009-03-19 16:34 ` Nick Piggin
2009-03-19 16:51 ` Linus Torvalds
2009-03-19 17:03 ` Jan Kara
2009-03-19 17:06 ` Jan Kara
2009-03-19 20:05 ` Linus Torvalds
2009-03-19 20:21 ` Linus Torvalds
2009-03-19 21:17 ` Ying Han
2009-03-19 22:16 ` Jan Kara
2009-03-19 16:46 ` Jan Kara
2009-03-24 7:44 ` Nick Piggin
2009-03-24 10:27 ` Nick Piggin
2009-03-24 10:32 ` Andrew Morton
2009-03-24 15:35 ` Nick Piggin
2009-03-26 18:29 ` Jan Kara
2009-03-26 0:03 ` Ying Han
2009-03-24 12:39 ` Jan Kara
2009-03-24 12:55 ` Jan Kara
2009-03-24 13:26 ` Jan Kara
2009-03-24 14:01 ` Chris Mason
2009-03-24 14:07 ` Jan Kara
2009-03-26 8:18 ` Aneesh Kumar K.V
2009-03-24 14:30 ` Nick Piggin
2009-03-24 14:47 ` Jan Kara
2009-03-24 14:56 ` Peter Zijlstra
2009-03-24 15:29 ` Jan Kara
2009-03-24 20:14 ` OGAWA Hirofumi
2009-03-26 8:47 ` Aneesh Kumar K.V
2009-03-26 11:37 ` Jan Kara
2009-03-26 23:02 ` Linus Torvalds
2009-03-24 15:03 ` Nick Piggin
2009-03-24 15:48 ` Jan Kara
2009-03-24 17:35 ` Jan Kara
2009-04-01 22:36 ` Ying Han
2009-04-02 10:11 ` Jan Kara
2009-04-02 11:24 ` Nick Piggin
2009-04-02 11:34 ` Jan Kara
2009-04-02 15:51 ` Nick Piggin
2009-04-02 17:44 ` Ying Han
2009-04-02 22:52 ` Ying Han
2009-04-02 23:39 ` Jan Kara
2009-04-03 0:25 ` Ying Han
2009-04-03 1:29 ` Ying Han
2009-04-03 9:41 ` Jan Kara
2009-04-03 21:34 ` Ying Han
2009-04-03 0:13 ` Ying Han
2009-03-27 20:35 ` Ying Han
2009-03-20 0:34 ` Ying Han
2009-03-20 0:49 ` Linus Torvalds
2009-03-20 7:00 ` Ying Han
2009-03-25 23:15 ` Ying Han
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=604427e00903181244w360c5519k9179d5c3e5cd6ab3@mail.gmail.com \
--to=yinghan@google.com \
--cc=akpm@linux-foundation.org \
--cc=alexk@google.com \
--cc=guichaz@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mikew@google.com \
--cc=rohitseth@google.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;
as well as URLs for NNTP newsgroup(s).