Re: [RESEND] x86/bus_lock: Don't assume the init value of DEBUGCTLMSR.BUS_LOCK_DETECT to be zero

[Ingo Molnar]

* Chenyi Qiang <chenyi.qiang@xxxxxxxxx> wrote:

> 
> 
> On 8/2/2022 6:51 PM, Ingo Molnar wrote:
> > 
> > * Chenyi Qiang <chenyi.qiang@xxxxxxxxx> wrote:
> > 
> > > It's possible that BIOS/firmware has set DEBUGCTLMSR_BUS_LOCK_DETECT, or
> > > this kernel has been kexec'd from a kernel that enabled bus lock
> > > detection.
> > > 
> > > Disable bus lock detection explicitly if not wanted.
> > 
> > Makes sense.
> > 
> > Just curious: in what circumstances does the BIOS/firmware set
> > DEBUGCTLMSR_BUS_LOCK_DETECT? Does it use it, or does it enable it for some
> > spurious reason, without really using the feature? (Assuming you are aware
> > of instances where this happened - or was this simply a hypothetical?)
> 
> Yes, It's just a hypothetical for BIOS/firmware. Kexec is the real case I
> met with this problem.

Fair enough, I've tweaked the changelog a bit to de-emphasize the firmware 
angle, and applied your fix to tip:x86/urgent.

Thanks,

	Ingo