ACK: [SRU][N:linux-azure-nvidia][PATCH 0/1] 64K-enabled ARM64 instances failing to boot on Azure
Kevin Becker
kevin.becker at canonical.com
Thu Nov 12 18:52:15 UTC 2026
On Sat, Nov 7, 2026 at 8:42 AM Vinicius Peixoto via kernel-team <
kernel-team at lists.ubuntu.com> wrote:
> SRU Justification
>
> [ Impact ]
>
> * Regression testing for the 2026.10.26 cycle exposed an issue where ARM64
> instances on Azure with 64K pages enabled fail to boot. The kernel log
> hangs early on during boot, indicating an early panic.
>
> [ Fix ]
>
> * Revert commit ffee5b49b764 ("arm64: mm: Don't remap pgtables for
> allocate vs populate"). The upstream 6.6.y stable tree has actually picked
> up and then reverted this commit due to the boot issue.
>
> [ Test Plan ]
>
> * Compile tested
> * Boot-tested on ARM64 instances with 64K page size enabled on Azure (both
> Ampere and Cobalt) to confirm that they don't hang anymore.
>
> [ Where problems could occur ]
>
> * This change will slightly worsen the boot time for 64K-enabled ARM64
> instances, as its removing an optimization. Other issues here could impact
> the early boot of those instances.
>
> Vinicius Peixoto (1):
> Revert "arm64: mm: Don't remap pgtables for allocate vs populate"
>
> arch/arm64/mm/mmu.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> --
>
>
Acked-by: Kevin Becker <kevin.becker at canonical.com>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ubuntu.com/archives/kernel-team/attachments/20261112/664faa33/attachment.html>
More information about the kernel-team
mailing list