Re: [PATCH] Btrfs: fix a bug of sleeping in atomic context

[Chris Mason <clm@fb.com>]

On Fri, Nov 20, 2015 at 12:06:33PM -0800, Liu Bo wrote:
> On Fri, Nov 20, 2015 at 09:57:49AM -0800, Liu Bo wrote:
> > On Fri, Nov 20, 2015 at 08:13:58AM -0500, Chris Mason wrote:
> > > On Thu, Nov 19, 2015 at 05:49:37PM -0800, Liu Bo wrote:
> > > > while xfstesting, this bug[1] is spotted by both btrfs/061 and btrfs/063,
> > > > so those sub-stripe writes are gatherred into plug callback list and
> > > > hopefully we can have a full stripe writes.
> > > > 
> > > > However, while processing these plugged callbacks, it's within an atomic
> > > > context which is provided by blk_sq_make_request() because of a get_cpu()
> > > > in blk_mq_get_ctx().
> > > > 
> > > > This changes to always use btrfs_rmw_helper to complete the pending writes.
> > > > 
> > > 
> > > Thanks Liu, but MD raid has the same troubles, we're not atomic in our unplugs.
> > 
> > Yeah, MD also does, but I don't see a way to change mq code at this
> > stage..
> 
> Correct it: MD raid5_unplug runs stripes inside a pair of spinlock (conf->device_lock) and moreover, those writes will be forwarded to raid5d to finish the job.
> 
> So md raid can run fine within atomic context.

Check MD raid10

-chris