Re: [PATCH v3] mkfs: Break block discard into chunks of 2 GB

["Darrick J. Wong" <darrick.wong@oracle.com>]

On Sat, Nov 30, 2019 at 10:01:17PM +0000, Chaitanya Kulkarni wrote:
> Not an XFS expert, but patch to handle ^C is been discussed on the
> block layer mailing list which includes discard operations. [1]

Heh, I wasn't aware of that. :)

> This solution seems specific to one file system, which will lead to
> code repetition for all the file systems which are in question.
> 
> How about we come up with the generic solution in the block-layer so
> it can be reused for all the file systems ?
> 
> (fyi, I'm not aware of any drawbacks of handling ^C it in the block
> layer and would like to learn if any).

The only one that I can think of is how to signal a partial completion,
but if you're only aborting on *fatal* signals then that doesn't matter.

Fixing the block layer seems like a better answer anyway.

> [1] https://patchwork.kernel.org/patch/11234607/

Though looking through that patch raises the question of whether xfs'
control loops also need to check for fatal signals, similar to what the
online scrub loops do?

--D

> -Chaitanya
> 
> On 11/27/2019 10:21 PM, Pavel Reichl wrote:
> > Some users are not happy about the BLKDISCARD taking too long and at the same
> > time not being informed about that - so they think that the command actually
> > hung.
> >
> > This commit changes code so that progress reporting is possible and also typing
> > the ^C will cancel the ongoing BLKDISCARD.
> >
> > Signed-off-by: Pavel Reichl<preichl@redhat.com>
> > ---
>