[dm-devel] "dm: prefer kmap_local_page() instead of deprecated kmap_atomic()" is broken

[dm-devel] "dm: prefer kmap_local_page() instead of deprecated kmap_atomic()" is broken

[Eric Biggers]

Hi Mike and Heinz,

The following commit in linux-next is causing "BUG: scheduling while atomic"
bugs.  Probably because it is still using kunmap_atomic(), not kunmap_local().

	commit 621d8807771e430b44136e48dfe3c53540c9953d
	Author: Heinz Mauelshagen <heinzm@redhat.com>
	Date:   Tue Feb 7 20:22:58 2023 +0100

	    dm: prefer kmap_local_page() instead of deprecated kmap_atomic()

This patch does not appear to have been sent to the mailing list (why not?), so
I started a new thread.

- Eric

--
dm-devel mailing list
dm-devel@redhat.com
https://listman.redhat.com/mailman/listinfo/dm-devel

Re: [dm-devel] "dm: prefer kmap_local_page() instead of deprecated kmap_atomic()" is broken

[Mike Snitzer]

On Mon, Feb 13 2023 at  2:50P -0500,
Eric Biggers <ebiggers@kernel.org> wrote:

> Hi Mike and Heinz,
> 
> The following commit in linux-next is causing "BUG: scheduling while atomic"
> bugs.  Probably because it is still using kunmap_atomic(), not kunmap_local().
> 
> 	commit 621d8807771e430b44136e48dfe3c53540c9953d
> 	Author: Heinz Mauelshagen <heinzm@redhat.com>
> 	Date:   Tue Feb 7 20:22:58 2023 +0100
> 
> 	    dm: prefer kmap_local_page() instead of deprecated kmap_atomic()
> 
> This patch does not appear to have been sent to the mailing list (why not?), so
> I started a new thread.

Thanks for letting me know.  And sorry for the trouble.

I meant to send the patchbomb after rebasing Heinz's series that he
shared with me privately.  I obviously forgot to do so which is both
embarassing and problematic.

Will email what I have just staged via rebase (after fixing to use
kunmap_local).

Mike

--
dm-devel mailing list
dm-devel@redhat.com
https://listman.redhat.com/mailman/listinfo/dm-devel