On 2023-02-20 19:09:22 +0100, Greg KH wrote:
On Mon, Feb 20, 2023 at 05:45:59PM +0100, Janne Grunau wrote:
Hej,
On 2023-02-20 12:40:25 +0100, gregkh@linuxfoundation.org wrote:
The patch below does not apply to the 6.1-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to stable@vger.kernel.org.
Why was this patch even considered for the 6.1-stable tree? Its "Fixes:" tag references a commit which only appeared in v6.2-rc1.
There is no need to backport this to any stable tree.
The message suggest that ignoring this would have been fine but it seems worthwhile to report a potential bug or ommission in the stable backport tooling.
Possible dependencies:
c06ba7b892a5 ("nvme-apple: reset controller during shutdown")
certainly doesn't depend on itself
...
c76b8308e4c9 ("nvme-apple: fix controller shutdown in apple_nvme_disable")
This is the commit that is now in the 6.1-stable queue, which is why I sent the "FAILED" email for the commit that said this was a fix.
I didn't think of that, I wouldn't have considered c76b8308e4c9 as a candidate for stable. At best it would have avoided some unnecessary work. Did it get queued for stanble due to the "fix" in the commit message?
commit from "Fixes:"
...
If this is intended since "Fixes:" tags are in general not relieable enough to decide to which trees regression fixes have to be applied to that's acceptable
Should I just drop c76b8308e4c9 ("nvme-apple: fix controller shutdown in apple_nvme_disable") instead?
yes, I think dropping c76b8308e4c9 ("nvme-apple: fix controller shutdown apple_nvme_disable") from stable is the best solution.
c06ba7b892a5 ("nvme-apple: reset controller during shutdown") was functionally a revert. It wasn't a clean revert since 285b6e9b5717 ("nvme: merge nvme_shutdown_ctrl into nvme_disable_ctrl") modified the same lines substantially.
thanks
Janne