Fwd: Ping²: [PATCH] x86: enable interrupts around dump_execstate()

[Jan Beulich <jbeulich@xxxxxxxx>]

Henry,

-------- Forwarded Message --------
Subject: Ping²: [PATCH] x86: enable interrupts around dump_execstate()
Date: Tue, 5 Jul 2022 18:19:38 +0200
From: Jan Beulich <jbeulich@xxxxxxxx>
To: Andrew Cooper <andrew.cooper3@xxxxxxxxxx>, Roger Pau Monné 
<roger.pau@xxxxxxxxxx>
CC: Wei Liu <wl@xxxxxxx>, xen-devel@xxxxxxxxxxxxxxxxxxxx 
<xen-devel@xxxxxxxxxxxxxxxxxxxx>

>On 11.01.2022 11:08, Jan Beulich wrote:
>> On 16.12.2021 14:33, Jan Beulich wrote:
>>> On 16.12.2021 12:54, Andrew Cooper wrote:
>>>> On 13/12/2021 15:12, Jan Beulich wrote:
>>>>> show_hvm_stack() requires interrupts to be enabled to avoids triggering
>>>>> the consistency check in check_lock() for the p2m lock. To do so in
>>>>> spurious_interrupt() requires adding reentrancy protection / handling
>>>>> there.
>>>>>
>>>>> Fixes: adb715db698b ("x86/HVM: also dump stacks from 
>>>>> show_execution_state()")
>>>>> Signed-off-by: Jan Beulich <jbeulich@xxxxxxxx>
>> 
>> There's a bug here which we need to deal with one way or another.
>> May I please ask for a response to the issues pointed out with
>> what you said in your earlier reply?
>
>I sincerely hope we won't ship another major version with this
>issue unfixed. The only option beyond applying this patch that I'm
>aware of is to revert the commit pointed at by Fixes:, which imo
>would be a shame (moving us further away from proper PVH support,
>including Dom0).

perhaps another item for the list of things needing resolution for
the release.

Jan