Greg, Amadeusz, On Tue, Aug 13, 2024 at 04:42:04PM +0200, Amadeusz Sławiński wrote: > On 8/12/2024 4:11 PM, Greg Kroah-Hartman wrote: > > On Mon, Aug 12, 2024 at 01:38:42PM +0300, Vitaly Chikunov wrote: > > > Greg, > > > > > > On Mon, Aug 12, 2024 at 12:25:54PM +0200, Greg Kroah-Hartman wrote: > > > > On Mon, Aug 12, 2024 at 12:01:48PM +0200, Amadeusz Sławiński wrote: > > > > > I guess that for completeness you need to apply both patches: > > > > > > > > > > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/sound/soc/soc-topology.c?id=97ab304ecd95c0b1703ff8c8c3956dc6e2afe8e1 > > > > > > > > This is already in the tree. > > > > > > > > > was an incorrect fix which was later fixed by: > > > > > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/sound/soc/soc-topology.c?id=0298f51652be47b79780833e0b63194e1231fa34 > > > > > > > > This commit will not apply :( > > > > > > It depends upon e0e7bc2cbee9 ("ASoC: topology: Clean up route loading"), > > > which was in the same patchset that didn't get applied. > > > https://lore.kernel.org/stable/?q=ASoC%3A+topology%3A+Clean+up+route+loading > > > > > > I see, Mark Brown said it's not suitable material for stable kernels > > > (since it's code cleanup), and Sasha Levin dropped it, and the dependent > > > commit with real fix. > > > > Ok, then someone needs to provide a working backport please... > > > > Should this be cherry-pick of both (they should apply cleanly): > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/sound/soc/soc-topology.c?id=e0e7bc2cbee93778c4ad7d9a792d425ffb5af6f7 > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/sound/soc/soc-topology.c?id=0298f51652be47b79780833e0b63194e1231fa34 > or just the second one adjusted to apply for stable trees? I think having commit with memory corruption fix is more important to stable kernels than not having the code cleanup commit. So, I would suggest stable policy to be changed a bit, and minor commits like this code cleanup, be allowed in stable if they are dependence of bug fixing commits. Additionally, these neutral commits just make stable trees become closer to mainline trees (which allows more bug fix commits to be applied cleanly). Thanks,
