From: Chen Gang <gang.chen@asianux.com>
To: Jeff Layton <jlayton@redhat.com>,
Trond.Myklebust@netapp.com, linux-nfs@vger.kernel.org,
linux-kernel@vger.kernel.org
Subject: [Bug fix] nfs-client: fix nfs_inode_attrs_need_update for async read_done comes during truncating to smaller size
Date: Mon, 15 Oct 2012 10:12:01 +0800 [thread overview]
Message-ID: <507B70F1.7040903@asianux.com> (raw)
Hello Trond Myklebust, Jeff Layton:
1) Root Cause:
A) begin truncate to smaller, after async read finish starting.
B) async read done come, after truncate operation change inode size.
C) in nfs_inode_attrs_need_update, nfs_size_need_update return true.
i) the bigger size is the original old size of client itself.
ii) the smaller size is the current true size.
iii) nfs_inode_attrs_need_update not consider this situation.
2) Fix nfs_size_need_update:
A) delete it:
i) it is for performance, not necessary (not for correctness).
ii) if it was necessary, it should use "!=" instead of '>'.
iii) it is the simplest way to fix this bug (maybe not best way).
B) consider this situation in it:
i) it is the best way.
ii) it is a little complex (need think of)
iii) sorry for I do not know how to fix it (at least now).
C) not touch it:
i) correct another place (such as nfs_update_inode)
ii) it is a bad idea (at least, I think it is)
iii) we need keep the source code as clearer as possible.
3) Test Result:
A) it is one client and one server separately, under 3.6-rc5 x86_32.
B) use one process (fsx-linux) test (only one user mode thread).
C) only use read, truncate, llseek, fstat operation for one file.
Before delete nfs_size_need_update, it causes issue.
After delete nfs_size_need_update, it is ok.
User Mode Log:
-------------------------------------------------------------------------
<<<test_start>>>
tag=nfsx-linux stime=1350202875
cmdline="export VERSION SOCKET_TYPE; TCbin=$LTPROOT/testcases/bin fsx.sh"
contacts=""
analysis=exit
<<<test_output>>>
Test Options:
VERSION: 2
RHOST: dhcp122.asianux.net
ITERATIONS: 50000
SOCKET_TYPE: udp
NFS_TYPE: nfs
Setting up remote machine: dhcp122.asianux.net
Mounting NFS filesystem dhcp122.asianux.net:/tmp/fsx1447.testdir on
/opt/ltp/testcases/bin/fsx1447 with options '-o proto=udp,vers=2 '
fsx-linux -N 50000 /opt/ltp/testcases/bin/fsx1447/testfile Starting
truncating to largest ever: 0x13e76
truncating to largest ever: 0x2e52c
truncating to largest ever: 0x3c2c2
truncating to largest ever: 0x3f15f
truncating to largest ever: 0x3fcb9
truncating to largest ever: 0x3fe96
truncating to largest ever: 0x3ff9d
Size error: expected 0x36ef9 stat 0x3bbca seek 0x36ef9
LOG DUMP (5652 total operations):
...
5636: 1350203089.781599 READ 0x143b6 thru 0x21ccb (0xd916 bytes)
5637: 1350203090.028214 READ 0x2a629 thru 0x2d0a1 (0x2a79 bytes)
5638: 1350203090.072029 TRUNCATE DOWN from 0x2d0a2 to 0x1bb35
5639: 1350203090.087401 READ 0x11a05 thru 0x1bb34 (0xa130 bytes)
5640: 1350203090.223985 READ 0x508c thru 0xa9da (0x594f bytes)
5641: 1350203090.245717 TRUNCATE DOWN from 0x1bb35 to 0x8830
5642: 1350203090.353502 READ 0x548f thru 0x882f (0x33a1 bytes)
5643: 1350203090.366596 READ 0x5802 thru 0x882f (0x302e bytes)
5644: 1350203090.366629 TRUNCATE UP from 0x8830 to 0x20011
5645: 1350203090.379476 TRUNCATE DOWN from 0x20011 to 0x134f4
5646: 1350203090.396234 READ 0x124a0 thru 0x134f3 (0x1054 bytes)
5647: 1350203090.401805 READ 0x880b thru 0x1189d (0x9093 bytes)
5648: 1350203090.532050 READ 0x134c7 thru 0x134f3 (0x2d bytes)
5649: 1350203090.532057 TRUNCATE UP from 0x134f4 to 0x3bbca
5650: 1350203090.546373 READ 0x2944c thru 0x2c1d6 (0x2d8b bytes)
5651: 1350203090.561228 READ 0xdbe1 thru 0x16260 (0x8680 bytes)
5652: 1350203090.751937 TRUNCATE DOWN from 0x3bbca to 0x36ef9
Correct content saved for comparison
(maybe hexdump "/opt/ltp/testcases/bin/fsx1447/testfile" vs
"/opt/ltp/testcases/bin/fsx1447/testfile.fsxgood")
fsx-linux -N 50000 /opt/ltp/testcases/bin/fsx1447/testfile Finished
Cleaning up testcase
Unmounting /opt/ltp/testcases/bin/fsx1447
Test Failed: Errors have resulted from this test
incrementing stop
<<<execution_status>>>
initiation_status="ok"
duration=218 termination_type=exited termination_id=1 corefile=no
cutime=43 cstime=82
<<<test_end>>>
-------------------------------------------------------------------------
Kernel Mode Log: (using printk which I add)
-------------------------------------------------------------------------
Time: My Mark: Task ptr: comments (include function name):
[ 280.883701] gchen_tag: f5c30000, nfs_read_done call
nfs_refresh_inode, cur=0x3bbca, new=0x3bbca
[ 280.890677] gchen_tag: f5c30000, nfs_read_done call
nfs_refresh_inode, cur=0x3bbca, new=0x3bbca
[ 280.897437] gchen_tag: f5c30000, nfs_read_done call
nfs_refresh_inode, cur=0x3bbca, new=0x3bbca
[ 280.897441] gchen_tag: f5e48c90, nfs_setattr_update_inode, cur=3bbca,
new=36ef9
[ 280.897450] gchen_tag: f5e48c90, nfs_setattr
[ 280.897462] gchen_tag: hit, f5c30000, nfs_refresh_inode_locked,
cur=36ef9, new=3bbca
[ 280.897469] gchen_tag: f5c30000, nfs_update_inode, change size,
cur=36ef9, new=3bbca
[ 280.898129] gchen_tag: f5e48c90, nfs_update_inode, change size,
cur=3bbca, new=36ef9
[ 280.977915] gchen_tag: f5c30000, nfs_update_inode, not change size,
cur=36ef9, new=2000, pages=55
[ 281.019879] gchen_tag: f5c30000, nfs_update_inode, not change size,
cur=36ef9, new=a000, pages=53
[ 281.070325] gchen_tag: f5c30000, nfs_update_inode, not change size,
cur=36ef9, new=e000, pages=45
[ 281.087103] gchen_tag: f5c30000, nfs_update_inode, not change size,
cur=36ef9, new=12000, pages=43
[ 281.129061] gchen_tag: f5c30000, nfs_update_inode, not change size,
cur=36ef9, new=16000, pages=41
[ 281.163012] gchen_tag: f5c30000, nfs_update_inode, not change size,
cur=36ef9, new=18000, pages=37
[ 281.213481] gchen_tag: f5c30000, nfs_update_inode, not change size,
cur=36ef9, new=1c000, pages=33
[ 281.255727] gchen_tag: f5c30000, nfs_update_inode, not change size,
cur=36ef9, new=22000, pages=31
[ 281.306177] gchen_tag: f5c30000, nfs_update_inode, not change size,
cur=36ef9, new=24000, pages=27
[ 281.356888] gchen_tag: f5c30000, nfs_update_inode, not change size,
cur=36ef9, new=2e000, pages=21
[ 281.398859] gchen_tag: f5c30000, nfs_update_inode, not change size,
cur=36ef9, new=36000, pages=19
[ 281.585491] gchen_tag: f5c30000, nfs_update_inode, not change size,
cur=104d8, new=2000, pages=17
[ 281.644207] gchen_tag: f5c30000, nfs_update_inode, not change size,
cur=104d8, new=10000, pages=15
-------------------------------------------------------------------------
Thanks.
--
Chen Gang
Asianux Corporation
next reply other threads:[~2012-10-15 2:29 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-10-15 2:12 Chen Gang [this message]
2012-10-15 4:27 ` [Bug fix] nfs-client: fix nfs_inode_attrs_need_update for async read_done comes during truncating to smaller size Myklebust, Trond
2012-10-15 4:27 ` Myklebust, Trond
2012-10-15 4:52 ` Chen Gang
2012-10-15 5:39 ` Chen Gang
2012-10-15 12:32 ` Myklebust, Trond
2012-10-15 12:32 ` Myklebust, Trond
2012-10-16 1:37 ` Chen Gang
2012-10-16 2:51 ` Myklebust, Trond
2012-10-16 2:51 ` Myklebust, Trond
2012-10-16 4:13 ` Chen Gang
2012-10-16 10:33 ` Jeff Layton
2012-10-16 11:44 ` Chen Gang
2012-10-16 12:13 ` Jeff Layton
2012-10-17 1:37 ` Chen Gang
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=507B70F1.7040903@asianux.com \
--to=gang.chen@asianux.com \
--cc=Trond.Myklebust@netapp.com \
--cc=jlayton@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-nfs@vger.kernel.org \
/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.