Re: Regression in suspend to ram in 2.6.31-rc kernels

[Pavel Machek <pavel@ucw.cz>]

> Um..., sorry, I'm not sure what are you talking about. Of course, the
> problem of this is that system freeze on suspend.
> 
> Or are you asking my guess of the cause, or something?  If so, although
> I'm not reading all emails on this thread, from Zdenek's backtrace, the
> sequence would be
> 
>     1) suspend mmc
>     2) mmc generates card removed event
>     3) prepare to invalidate blockdev
>     4) sync fs on invalidating blockdev
>     5) flush buffers on invalidating blockdev (partitions)
>     6) delete blockdev (partitions)
> 
> or like the above. And I can guess some possible issues/root-cause we
> have to handle from it.
> 
>     a) card removed event from mmc for suspend is right design?
>     b) the card can be changed/removed before system was resumed, mmc
>        can be detect/handle it properly?
>     c) flushing buffers on _deleted_ device is right design?
> 
> and I suspect there are more issues in detail and resume process though.

I guess c) is main problem. If device is not there, how do you want to
flush buffers on it?

								Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html