[PATCH v2] arm64: signal: Ensure signal delivery failure is recoverable

[Kevin Brodsky]

On 12/12/2024 18:51, Catalin Marinas wrote:
> On Thu, Dec 12, 2024 at 05:26:47PM +0000, Mark Brown wrote:
>> On Thu, Dec 12, 2024 at 05:10:14PM +0000, Catalin Marinas wrote:
>>> On Thu, Dec 12, 2024 at 04:47:46PM +0000, Mark Brown wrote:
>>>>> Fixes: eaf62ce1563b ("arm64/signal: Set up and restore the GCS context for signal handlers")
>>>> Presumably we also need a Fixes for whatever introduced the above copy.
>>> Hmm, it was like this from the beginning. If we are to do a cc stable to
>>> the oldest maintained LTS, I'd rather have a separate patch for
>>> copy_siginfo_to_user(), it doesn't make much sense to backport a GCS
>>> patch to early kernels.
>> Hrm, good point about it triggering backports.  I'm not sure it's worth
>> doing that TBH.
> I had the same thoughts. I think this was mostly triggered by the
> POR_EL0 changes when Will raised the point that the kernel might return
> with a more privileged value in this sysreg.

I don't think backporting would be worth it either, referencing the
original commit implementing signal support felt like overkill. The
copy_siginfo_to_user() issue is such a corner case that I doubt anyone
ever ran into it.

- Kevin