On Sat, Jun 15, 2024 at 07:28:30PM -0400, Ronnie Sahlberg wrote:

Manually checked list of commits in upstream that mention NULL pointers and that reference that they fix a commit that is in linux-running-stable.

a9b9741854a9fe9df948 c4ab9da85b9df3692f86 c44711b78608c98a3e6b 0dcc53abf58d572d34c5

All already grabbed.

445c0b69c72903528fdf

Note, this one references a 6.10-rc1 commit, so it can't be included. Don't know if you want to add that to your script to check.

97ab3e8eec0ce79d9e26

Same, 6.10-rc1. And also, the check is wrong, that's not what this commit does, you might want to read the commits :)

47558cbaa842c4561d08

Fixes 6.10-rc1

62cbabc6fd228e62daff

Fixes 6.10-rc1

02367f52901932674ff2

Welcome to my hell :(

This already is in the tree, as a totally different commit, and backported properly, but the AMD developers seem to like to apply it to different branches at the same time, AND give us no hints that they did so. Ugh, it's a mess and one reason I dread reviewing amdgpu driver patches for stable these days...

thanks,

greg k-h