On Wed, Feb 19, 2025 at 05:16:19PM +0530, Naresh Kamboju wrote:
On Mon, 17 Feb 2025 at 17:07, Greg Kroah-Hartman gregkh@linuxfoundation.org wrote:
On Mon, Feb 17, 2025 at 05:00:43PM +0530, Naresh Kamboju wrote:
On Sat, 8 Feb 2025 at 16:54, Naresh Kamboju naresh.kamboju@linaro.org wrote:
On Thu, 6 Feb 2025 at 21:36, Greg Kroah-Hartman gregkh@linuxfoundation.org wrote:
This is the start of the stable review cycle for the 6.6.76 release. There are 389 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.
Responses should be made by Sat, 08 Feb 2025 15:51:12 +0000. Anything received after that time might be too late.
The whole patch series can be found in one patch at: https://www.kernel.org/pub/linux/kernel/v6.x/stable-review/patch-6.6.76-rc2.... or in the git tree and branch at: git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-6.6.y and the diffstat can be found below.
thanks,
greg k-h
There are three different regressions found and reporting here, We are working on bisecting and investigating these issues,
We observed a kernel warning on QEMU-ARM64 and FVP while running the newly added selftest: arm64: check_hugetlb_options. This issue appears on 6.6.76 onward and 6.12.13 onward, as reported in the stable review [1]. However, the test case passes successfully on stable 6.13.
The selftests: arm64: check_hugetlb_options test was introduced following the recent upgrade of kselftest test sources to the stable 6.13 branch. As you are aware, LKFT runs the latest kselftest sources (from stable 6.13.x) on 6.12.x, 6.6.x, and older kernels for validation purposes.
From Anders' bisection results, we identified that the missing patch on
6.12 is likely causing this regression:
First fixed commit: [25c17c4b55def92a01e3eecc9c775a6ee25ca20f] hugetlb: arm64: add MTE support
Could you confirm whether this patch is eligible for backporting to 6.12 and 6.6 kernels? If backporting is not an option, we will need to skip running this test case on older kernels.
The test case itself should properly "skip" if the feature is not present in the kernel. Why not fix that up instead?
The reported test gets PASS at the end, but generates kernel warning while running the test case (always reproducible) on 6.12 and 6.6.
The reported warning was not seen on stable 6.13.
So this implies that userspace can cause a kernel warning? That means it can cause a DoS, that's not good at all.
So the commit you mention actually fixes a bug then? Otherwise this feels really odd, as that means that any kernel without that change can crash this way. What changed to cause this to happen?
thanks,
greg k-h