Note that this was a real fix, but the fix only matters if commit
aaec5a95d596 ("pipe_read: don't wake up the writer if the pipe is
still full") is in the tree.
Now, the bug was pre-existing, and *maybe* it could be hit without
that commit aaec5a95d596, but nobody has ever reported it, so it's
very very unlikely.
Also, this fix then had some fall-out, and while I think you've queued
all the fallout fixes too, I think it might be a good idea to wait for
more reports from the development tree before considering these for
stable.
Put another way: this fix caused some pain. It might not be worth
back-porting to stable at all, and if it is, it might be worth waiting
to see that there's no other fallout.
Linus
On Sun, 9 Mar 2025 at 09:52, Sasha Levin <sashal@xxxxxxxxxx> wrote:
>
> This is a note to let you know that I've just added the patch titled
>
> fs/pipe: Read pipe->{head,tail} atomically outside pipe->mutex
