Re: 2.6.15-mm2

[Tejun Heo]

Jens Axboe wrote:
> On Thu, Jan 12 2006, Tejun Heo wrote:
>
> > 4. a REQ_SPECIAL | REQ_BLOCK_PC | REQ_QUIET request gets queued at
> >   the head of the queue.  (I have no idea where this comes from.  sd
> >   driver doesn't even handle PC requests.  It will be just failed.
> >   Some kind of hardware management stuff trying to probe MMC
> >   devices?)
>
>
> But it does, sd understands these just fine (see references to
> blk_pc_request()).
>
> It could be coming from someone doing a blkdev_issue_flush, that will
> even cause sd to queue such a request internally. So it isn't
> necessarily from user space (it would have to be through SG_IO at that
> point), and Reubens boot log doesn't have any evidence of anything of
> that nature being started. So I'm guessing it's the flush, raid1
> propagates these flushes to the bottom devices when it sees one.
>
> Your analysis looks correct though, Reuben looking forward to hearing
> whether this fixes your boot hang!

Ah... you're right.  I was only staring at the !blk_fs_request() test in 
sd_init_command().  It has early exit for blk_pc_request() of course. 
It needs to handle SG_IO's & flushes.  :-p

--
tejun
-
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/