Please don''t reply to lustre-devel. Instead, comment in Bugzilla by using the following link: https://bugzilla.lustre.org/show_bug.cgi?id=10214 Created an attachment (id=9385) Please don''t reply to lustre-devel. Instead, comment in Bugzilla by using the following link: --> (https://bugzilla.lustre.org/attachment.cgi?id=9385&action=view) implementation o_sync without any extra kernel patches reimplementation O_SYNC without need any extra kernel patchess who usefull for patchless client. In future need rewrite lustre to use PG_writeback bits for lock pages for 2.6 kernel to avoid divide ll_sync_pages to 2.4 and 2.6 parts.
Please don''t reply to lustre-devel. Instead, comment in Bugzilla by
using the following link:
https://bugzilla.lustre.org/show_bug.cgi?id=10214
What |Removed |Added
----------------------------------------------------------------------------
Attachment #9385 is|0 |1
obsolete| |
Created an attachment (id=9407)
Please don''t reply to lustre-devel. Instead, comment in Bugzilla by
using the following link:
--> (https://bugzilla.lustre.org/attachment.cgi?id=9407&action=view)
implementation o_sync without any extra kernel patches
code for 2.4 kernel droped, after vitaly inspection and disscusion with alex.
O_SYNC in 2.4.21-EL3 kernel work fine without any code in lustre.
Please don''t reply to lustre-devel. Instead, comment in Bugzilla by
using the following link:
https://bugzilla.lustre.org/show_bug.cgi?id=10214
What |Removed |Added
----------------------------------------------------------------------------
Attachment #9407|review?(vitaly@clusterfs.com|review+
Flag|) |
(From update of attachment 9407)
the initial idea to have cond_reshed() was probably not so bad.
Please don''t reply to lustre-devel. Instead, comment in Bugzilla by
using the following link:
https://bugzilla.lustre.org/show_bug.cgi?id=10214
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |RESOLVED
Resolution| |FIXED
patch landed to b1_4/b1_5.