Linux patch for reading ufs2
Niraj Kumar
niraj17 at iitbombay.org
Mon Feb 16 23:27:21 PST 2004
Hi ,
Okay ,I was able to reproduce the problem at my machine.
I have a feeling that the this error can be avoided if you use
CONFIG_LBD in your kernel configuration.
( Device Drivers -> Block Devices - > Support for large block devices )
Please use this option and see whether the problem is still there or not .
Meanwhile , I am trying to figure out a better/complete solution.
Niraj
Vladimir Kushnir wrote:
>Hi,
>This is /usr/ports/distfiles/emovix-0.8.0.tar.gz - regular file, "ls -li":
>1842130 -rw-r--r-- 1 vkushnir wheel 8636010 Jan 5
>23:43 /usr/ports/distfiles/emovix-0.8.0.tar.gz
>
>I don't know precisely what's relevant but after (lengthy indeed) debugging
>information Linux gives:
>......
>Feb 15 18:33:58 localhost kernel: (fs/ufs/inode.c, 543), ufs_read_inode:
>ENTER, ino 1842130
>Feb 15 18:33:58 localhost kernel: (fs/ufs/inode.c, 636), ufs_read_inode:
>Reading ufs2 inode, ino 184213
>Feb 15 18:33:58 localhost kernel: (fs/ufs/inode.c, 704), ufs_read_inode: EXIT
>Feb 15 18:34:06 localhost kernel: attempt to access beyond end of device
>Feb 15 18:34:06 localhost kernel: hde10: rw=0, want=2147483880, limit=42732580
>Feb 15 18:34:06 localhost kernel: Buffer I/O error on device hde10, logical
>block 536870969
>Feb 15 18:34:06 localhost kernel: attempt to access beyond end of device
>Feb 15 18:34:06 localhost kernel: hde10: rw=0, want=2147483884, limit=42732580
>Feb 15 18:34:06 localhost kernel: Buffer I/O error on device hde10, logical
>block 536870970
>Feb 15 18:34:06 localhost kernel: attempt to access beyond end of device
>Feb 15 18:34:06 localhost kernel: hde10: rw=0, want=2147483888, limit=42732580
>Feb 15 18:34:06 localhost kernel: Buffer I/O error on device hde10, logical
>block 536870971
>Feb 15 18:34:06 localhost kernel: attempt to access beyond end of device
>Feb 15 18:34:06 localhost kernel: hde10: rw=0, want=2147483892, limit=42732580
>Feb 15 18:34:06 localhost kernel: Buffer I/O error on device hde10, logical
>block 536870972
>and so on - several blocks in this region (here hd10 == ad0s2e, /usr, mounted
>under Linux with "-r -t ufs -o ufstype=ufs2"). And finally it refuses to copy
>or untar this file. This is the only information I cn see there.
>At the same time when I boot into FreeBSD - no problems whatsoever.
>
>On Monday 16 February 2004 06:22, Niraj Kumar wrote:
>
>
>>Hi,
>>
>>Thanks for testing this stuff.
>>
>>Vladimir Kushnir wrote:
>>
>>
>>>>First off - thanks a bunch, now I don't need to reboot every time I
>>>>forgot to copy some files to Linux.
>>>>Tested and works here: ad0: 57241MB <WDC WD600JB-00CRA1>;
>>>>ad0s2 - current -CURRENT (everything in UFS2);
>>>>ad0s5 - Mandrake 9.2 + kernel 2.6.2 + p{1,2}.txt
>>>>
>>>>There is a file though (emovix distro in ports/distfiles) which Linux
>>>>gives an input/output error on. More precisely, it gives me
>>>>"attempt to access beyond end of device
>>>>Buffer I/O error on device hde10, logical block ..." (several blocks).
>>>>Under FreeBSD - no errors on this file so it doesn't seem like HD
>>>>problem.
>>>>
>>>>
>>Well , what kind of file is this ? Is it a normal file or some device
>>file , or symobolic link etc ...?
>>Can you send me the 'dmesg' output for this error (if there is any) ?
>>FYI , I have enabled some debugging messages in this patch , so you
>>should be getting a lot
>>of debugging messages in 'dmesg' .
>>
>>Any other input on this problem is welcome.
>>
>>Niraj
>>
>>
>>
>
>Regards,
>Vladimir
>_______________________________________________
>freebsd-questions at freebsd.org mailing list
>http://lists.freebsd.org/mailman/listinfo/freebsd-questions
>To unsubscribe, send any mail to "freebsd-questions-unsubscribe at freebsd.org"
>
>
More information about the freebsd-questions
mailing list