Re: [Regression] 6.10-rc1: Fail to resurrect from suspend.

[Linux regression tracking (Thorsten Leemhuis)]

On 13.06.24 13:04, Pavel Machek wrote:
>> My system fails to resurrect after `systemctl suspend` with 6.10-rc1,
>> when pressing power button, the machine "sounds" starting(fans roaring),
>> but my keyboard/mouse/monitor is not powered, and I have nothing to
>> do but powering cycle the system.
>>
>> I run a bisect session, and narrows it down to following commit:
>>
>> 	commit d74169ceb0d2e32438946a2f1f9fc8c803304bd6
>> 	Author: Dimitri Sivanich <sivanich@xxxxxxx>
>> 	Date:   Wed Apr 24 15:16:29 2024 +0800
>>
>> 	    iommu/vt-d: Allocate DMAR fault interrupts locally
>> 	    
> [...]
>> And I have confirmed that reverting this commit can fix my problem.
> 
> Bisected regression. Should we simply revert the patch?

No need to afaics, as this afaics is fixed by "iommu/amd: Fix panic
accessing amd_iommu_enable_faulting" which Joerg committed earlier today.

https://lore.kernel.org/all/ZljHE%2FR4KLzGU6vx@xxxxxxx/
https://lore.kernel.org/all/31f16e4e.afe2.18fc9d5dad0.Coremail.00107082@xxxxxxx/

Ciao, Thorsten