lkml.org 
[lkml]   [2000]   [Aug]   [9]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [2.4.x] Loop deadlock workaround?
On Tue, 8 Aug 2000, Aaron Tiensivu wrote:

> So far so good.. I've mounted my loopback ext2 partition with 'sync' mount options..
>
> Am I just lucky or is this a valid workaround until loop gets fixed?

Hi Aaron,

AFAIKT, Linus fixed the deadlock problem. Do you have a formula for
locking the thing up with a post test6-pre5 kernel?

The only problems I'm aware of with the loop device atm are:

1. Strange problem whereby read errors occur (only under vm stress)
iff bs=1k and mount -oloop blkdev is used.. mount -oremount,loop
previously mounted blkdev works fine 100% of the time. [1]

2. Lack of large file support for filebacked device.

-Mike

1. If someone would confirm/deny this, I'd be grateful. This is too
weird to be real :-/ trouble is it's 100% repeatable here.

HowToReproduce:
create 1k bs fs;mount it;copy kernel source into same
unmount fs;mount -oloop fs
make dep;make -jN bzImage [N is enough to make your box swap heartily]
(here: watch cpp bitch like crazy after vm pressure builds)
umount fs
mount fs;mount -oremount,loop fs
make dep;make -jN bzImage
(here: watch it always work fine)
repeat many times in any order you like, then try same with 4k bs.
(here: watch it always work fine with 4k bs.. smoke pours out ears;)



-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:58    [W:0.028 / U:0.160 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site