On 3/5/25 00:14, Pavel Begunkov wrote:
On 3/4/25 23:49, Christoph Hellwig wrote:
On Tue, Mar 04, 2025 at 04:43:29PM -0700, Jens Axboe wrote:
On 3/4/25 4:26 PM, Christoph Hellwig wrote:
On Tue, Mar 04, 2025 at 10:36:16AM -0700, Jens Axboe wrote:
stable and actual production certainly do. Not that this should drive
upstream development in any way, it's entirely unrelated to the problem
at hand.
And that's exactly what I'm saying. Do the right thing instead of
whining about backports to old kernels.
Yep we agree on that, that's obvious. What I'm objecting to is your
delivery, which was personal rather than factual, which you should imho
apologize for.
I thus sincerly apologize for flaming Pavel for whining about
Again, pretty interesting choice of words, because it was more akin
to a tantrum. I hope you don't think you'd be taken seriously after
that, I can't, and not only from this conversation.
backporting, but I'd still prefer he would have taken up that proposal
on technical grounds as I see absolutely no alternatively to
synchronously returning an error.
The "You have to do X" hardly reveals any details to claim "technical
grounds". Take stable out of the question, and I would still think
that a simple understandable fix for upstream is better than jumping
to a major rework right away.
That is _if_ there is a simple fix, but your argument was that
it's "not very accurate", which is not very useful.
And honestly pretty tiring that this needs to be said, still. Really.
An I'm really tired of folks whining about backporting instead of
staying ontopic.
--
Pavel Begunkov
