On Mon, Jun 08, 2020 at 10:10:21PM -0400, Sasha Levin wrote:
On Mon, Jun 08, 2020 at 06:07:27PM -0700, Darrick J. Wong wrote:
On Mon, Jun 08, 2020 at 07:05:37PM -0400, Sasha Levin wrote:
From: "Darrick J. Wong" darrick.wong@oracle.com
[ Upstream commit a5949d3faedf492fa7863b914da408047ab46eb0 ]
When writing to a delalloc region in the data fork, commit the new allocations (of the da reservation) as unwritten so that the mappings are only marked written once writeback completes successfully. This fixes the problem of stale data exposure if the system goes down during targeted writeback of a specific region of a file, as tested by generic/042.
Signed-off-by: Darrick J. Wong darrick.wong@oracle.com Reviewed-by: Christoph Hellwig hch@lst.de Reviewed-by: Brian Foster bfoster@redhat.com Signed-off-by: Sasha Levin sashal@kernel.org
Err, this doesn't have a Fixes: tag attached to it. Does it pass fstests? Because it doesn't look like you've pulled in "xfs: don't fail unwritten extent conversion on writeback due to edquot", which is needed to avoid regressing fstests...
...waitaminute, that whole series lacks Fixes: tags because it wasn't considered a good enough candidate for automatic backport.
AUTOSEL doesn't look just at the Fixes tag :)
Ummm, does the autosel fstests driver turn on quotas? ;)
Uh, apparently not :/ Is it okay to just enable it across all tests?
It should be at this point.
While I go fix that up, would you rather drop the series, or pick up 1edd2c055dff ("xfs: don't fail unwritten extent conversion on writeback due to edquot")?`
Let's drop it for now, please. There might be a few more tweaks needed to get that bit just right.
--D
-- Thanks, Sasha