Re: [RFC][PATCH] PM: Force GFP_NOIO during suspend/resume (was: Re: [linux-pm] Memory allocations in .suspend became very unreliable)

[Oliver Neukum]

Am Dienstag, 19. Januar 2010 21:37:35 schrieb Rafael J. Wysocki:
> That said, Maxim reported that in his test case the mm subsystem apparently
> attempted to use I/O even if there was a plenty of free memory available and
> I'd like prevent that from happening.

Hi,

it seems to me that this is caused by the mm subsytem maintaing
a share of clean pages precisely so that GFP_NOIO will work.
Perhaps it is a good idea to
a) launder a number of pages if the system is about to be suspended
between the freezer and notifying drivers
b) set the ration of clean pages to dirty pages to 0 while suspending
the system.

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