On Wed, Jul 13, 2022 at 12:49 AM Dave Chinner <david@xxxxxxxxxxxxx> wrote:
>
> Fix this by constraining the EOF block matching to only match
> against other EOF blocks that have identical EOF offsets and data.
So I agree that this patch fixes the bug, but I just spent a long time
writing an email about how I think it makes oddly written and fairly
incomprehensible code even harder to read.
This is likely a minimal patch and as such good for backporting (and
just for "current stage of release cycle too" for that matter), but I
*really* think the code here could do with a nice chunk of cleanup.
Linus
