On Montag 06 September 2004 12:41, David Gümbel wrote: > On Sonntag 05 September 2004 18:49, Jari Ruusu wrote: > > Can you test attached patch? It transforms read-ahead requests from > > file system to normal reads for underlying device. Normal read requests > > should not fail with -EWOULDBLOCK so this should cure the problem. > > Just applied it, will test with 2.6.8.1 ASAP. OK, I tested your patch applied to loop-AES 2.1c on 2.6.8.1, and tried all I did before to trigger the bug. None of the usual suspects worked (no IO error reading encrypted reiserfs /tmp, no ext3 error doing ls in a particualr dir on my encrypted ext3 partition, no error putting the system under load and using KMail. After some hours of work on the machine, I now consider it pretty safe to say that the error has disappeared, which means that the patch v5 you (Jari) posted fixed it. Thanks again for your work, I appreciate it. Are you going to report this error to the kernel mailinglist, and is there going to be a new version of loop-AES (I suspect that ;-) ? Regards, David -- David Gümbel . . . . . . . . . . . . . . . . david.guembel@xxxxxxxxx ITOMIG GbR . . . . . . . . . . . . . . . . . phone: +49 7071 2970493 Sand 14 . . . . . . . . . . . . . . . . . . mobil: +49 179 5400885 D-72076 Tübingen . . . . . . . . . . . . . . . .http://www.itomig.de
Attachment:
pgpN8j1KabxJa.pgp
Description: PGP signature
