Messages in this thread | | | Subject | Re: [PATCH] coredump: Retry writes where appropriate | From | Paul Smith <> | Date | Tue, 23 Jun 2009 13:31:18 -0400 |
| |
On Thu, 2009-06-04 at 10:14 -0700, Roland McGrath wrote: > > Fortunately, this doesn't look hard. Whatever we do, we should modify > > dump_write/seek to check fatal_signal_pending() anyway. Because we can't > > know if f_ops->write() pays attention to signals. > > Yes, that sounds fine. > > > This means we can just add try_to_freeze(). > > Right. > > > As for exit_mm(), we can use freezer_do_not_count() + freezer_count() > > around the "for (;;)" loop. > > Ah yes, sure.
Hi Oleg;
Did you have any more time to look into this? I'm currently using my patch and it's OK for my purposes but I'll be happy to test any proposal you come up with, if you like, so I can drop my patch in the future.
Let me know, thanks!
| |