But it was backported on 6.6 and the performance impact is pretty high on some workloads. As this patch is only providing a perf benefit and to keep consistency with other stable kernels, would it be possible to get it merged ?
Le jeu. 28 nov. 2024 à 16:52, Sasha Levin sashal@kernel.org a écrit :
On Thu, Nov 28, 2024 at 12:59:24PM +0100, Borislav Petkov wrote:
Hey folks,
On Mon, Jan 15, 2024 at 06:26:56PM -0500, Sasha Levin wrote:
From: "Borislav Petkov (AMD)" bp@alien8.de
[ Upstream commit 04c3024560d3a14acd18d0a51a1d0a89d29b7eb5 ]
AMD does not have the requirement for a synchronization barrier when acccessing a certain group of MSRs. Do not incur that unnecessary penalty there.
Erwan just mentioned that this one is not in 6.1 and in 5.15. And I have mails about it getting picked up by AUTOSEL.
Did the AI reconsider in the meantime?
You've missed the 5.10 mail :)
Pavel objected to it so I've dropped it: https://lore.kernel.org/all/Zbli7QIGVFT8EtO4@sashalap/
-- Thanks, Sasha