Re: [PATCH] Crash: add lock to serialize crash hotplug handling

[Baoquan He <bhe@redhat.com>]

On 09/23/23 at 07:10am, Eric DeVolder wrote:
> 
> 
> On 9/22/23 18:54, Baoquan He wrote:
> > Eric reported that handling corresponding crash hotplug event can be
> > failed easily when many momery hotplug event are notified in a short period.
> > They failed because failing to take __kexec_lock.
> > 
> > =======
> > [   78.714569] Fallback order for Node 0: 0
> > [   78.714575] Built 1 zonelists, mobility grouping on.  Total pages: 1817886
> > [   78.717133] Policy zone: Normal
> > [   78.724423] crash hp: kexec_trylock() failed, elfcorehdr may be inaccurate
> > [   78.727207] crash hp: kexec_trylock() failed, elfcorehdr may be inaccurate
> > [   80.056643] PEFILE: Unsigned PE binary
> > =======
> > 
> > The memory hotplug events are notified very quickly and very many,
> > while the handling of crash hotplug is much slower relatively. So the
> > atomic variable __kexec_lock and kexec_trylock() can't guarantee the
> > serialization of crash hotplug handling.
> > 
> > Here, add a new mutex lock __crash_hotplug_lock to serialize crash
> > hotplug handling specifically. This doesn't impact the usage of
> > __kexec_lock.
> > 
> > Signed-off-by: Baoquan He <bhe@redhat.com>
> > ---
> >   kernel/crash_core.c     |  3 +++
> >   kernel/kexec_core.c     |  1 +
> >   kernel/kexec_internal.h | 11 +++++++++++
> >   3 files changed, 15 insertions(+)
> > 
> > diff --git a/kernel/crash_core.c b/kernel/crash_core.c
> > index 03a7932cde0a..e8851724a530 100644
> > --- a/kernel/crash_core.c
> > +++ b/kernel/crash_core.c
> > @@ -783,9 +783,11 @@ static void crash_handle_hotplug_event(unsigned int hp_action, unsigned int cpu)
> >   {
> >   	struct kimage *image;
> > +	crash_hotplug_lock();
> >   	/* Obtain lock while changing crash information */
> >   	if (!kexec_trylock()) {
> >   		pr_info("kexec_trylock() failed, elfcorehdr may be inaccurate\n");
> > +		crash_hotplug_unlock();
> >   		return;
> >   	}
> > @@ -852,6 +854,7 @@ static void crash_handle_hotplug_event(unsigned int hp_action, unsigned int cpu)
> >   out:
> >   	/* Release lock now that update complete */
> >   	kexec_unlock();
> > +	crash_hotplug_unlock();
> >   }
> 
> The crash_check_update_elfcorehdr() also has kexec_trylock() and needs similar treatment.
> 
> >   static int crash_memhp_notifier(struct notifier_block *nb, unsigned long val, void *v)
> > diff --git a/kernel/kexec_core.c b/kernel/kexec_core.c
> > index 9dc728982d79..b95a73f35d9a 100644
> > --- a/kernel/kexec_core.c
> > +++ b/kernel/kexec_core.c
> > @@ -48,6 +48,7 @@
> >   #include "kexec_internal.h"
> >   atomic_t __kexec_lock = ATOMIC_INIT(0);
> > +DEFINE_MUTEX(__crash_hotplug_lock);
> >   /* Flag to indicate we are going to kexec a new kernel */
> >   bool kexec_in_progress = false;
> > diff --git a/kernel/kexec_internal.h b/kernel/kexec_internal.h
> > index 74da1409cd14..1db31625ef20 100644
> > --- a/kernel/kexec_internal.h
> > +++ b/kernel/kexec_internal.h
> > @@ -28,6 +28,17 @@ static inline void kexec_unlock(void)
> >   	atomic_set_release(&__kexec_lock, 0);
> >   }
> > +/*
> > + * Different than kexec/kdump loading/unloading/crash or kexec jumping/shrinking
> > + * which usually rarely happen, there will be many crash hotplug events notified
> > + * during one short period, e.g one memory board is hot added and memory regions
> > + * are online. So mutex lock  __crash_hotplug_lock is used to serialize the crash
> > + * hotplug handling specificially.
> > + * */
> > +extern struct mutex __crash_hotplug_lock;
> > +#define crash_hotplug_lock() mutex_lock(&__crash_hotplug_lock)
> > +#define crash_hotplug_unlock() mutex_unlock(&__crash_hotplug_lock)
> > +
> >   #ifdef CONFIG_KEXEC_FILE
> >   #include <linux/purgatory.h>
> >   void kimage_file_post_load_cleanup(struct kimage *image);
> 
> The new content for kexec_internal.h and kexec_core.c could/should probably be
> moved into crash_core.c, within the CONFIG_CRASH_HOTPLUG?

That makes sense, I will spin v2 and post.