Re: [xfs-masters] [patch 2/2] fs/xfs_super: replace schedule_timeout() with msleep_interruptible()

[Nathan Scott <nathans@xxxxxxx>]

On Mon, Mar 07, 2005 at 04:20:31PM -0800, Nishanth Aravamudan wrote:
> While youre interface may be in human-sensible units, the internal timer
> subsystem is not (jiffies only exist in the kernel).
> msleep_interruptible() changes this, clearly. My patch doesn't really
> change anything. It really shouldn't really result in any different
> behavior as far as I can tell. msleep_interruptible() takes a timeout

Actually, discussing with Christoph a bit he's pointed out this is
going to cause problems for the cases (both xfsbufd and xfssyncd)
where we manually wakeup those processes - and there are several
situations where we need to do that.

So, unless I've missed something, these patches shouldn't be merged.

cheers.

-- 
Nathan