[PATCH v2] panic, kexec: Make __crash_kexec() NMI safe

[Baoquan He]

On 06/26/22 at 06:37pm, Baoquan He wrote:
> On 06/24/22 at 02:37pm, Valentin Schneider wrote:
> > On 24/06/22 09:30, Baoquan He wrote:
> > > On 06/20/22 at 12:15pm, Valentin Schneider wrote:
> > >> @@ -94,14 +94,20 @@ static int do_kexec_load(unsigned long entry, unsigned long nr_segments,
> > >>  	/*
> > >>  	 * Because we write directly to the reserved memory region when loading
> > >>  	 * crash kernels we need a mutex here to prevent multiple crash kernels
> > >> -	 * from attempting to load simultaneously, and to prevent a crash kernel
> > >> -	 * from loading over the top of a in use crash kernel.
> > >> -	 *
> > >> -	 * KISS: always take the mutex.
> > >> +	 * from attempting to load simultaneously.
> > >>  	 */
> > >>  	if (!mutex_trylock(&kexec_mutex))
> > >>  		return -EBUSY;
> > >
> > > So kexec_mutex is degenerated to only avoid simultaneous loading,
> > > should we rename to reflect that?, e.g kexec_load_mutex.
> > >
> > 
> > It's also serializing crash_get_memory_size() and crash_shrink_memory();
> > more generally it should still be the preferred serialization mechanism as
> > it's a "proper" lock visible by instrumentation, the atomic variable is a
> > side character for the NMI case.
> 
> You are right. I only checked the code comment in this place. Then this
> patch looks good to me, thx.
> 
> Acked-by: Baoquan He <bhe at redhat.com>

OK, just saw Eric's comment after I replied.