Re: [PATCH] perf/arm_cspmu: fix device leaks on module unload

[Johan Hovold <johan@kernel.org>]

On Mon, Nov 24, 2025 at 04:03:29PM +0000, Will Deacon wrote:
> On Fri, Nov 21, 2025 at 12:52:13PM +0100, Johan Hovold wrote:
> > Make sure to drop the references taken when looking up the backend
> > devices during vendor module unload.
> > 
> > Fixes: bfc653aa89cb ("perf: arm_cspmu: Separate Arm and vendor module")
> > Cc: stable@vger.kernel.org	# 6.7
> > Cc: Besar Wicaksono <bwicaksono@nvidia.com>
> > Signed-off-by: Johan Hovold <johan@kernel.org>
> > ---
> >  drivers/perf/arm_cspmu/arm_cspmu.c | 4 +++-
> >  1 file changed, 3 insertions(+), 1 deletion(-)
> > 
> > diff --git a/drivers/perf/arm_cspmu/arm_cspmu.c b/drivers/perf/arm_cspmu/arm_cspmu.c
> > index efa9b229e701..e0d4293f06f9 100644
> > --- a/drivers/perf/arm_cspmu/arm_cspmu.c
> > +++ b/drivers/perf/arm_cspmu/arm_cspmu.c
> > @@ -1365,8 +1365,10 @@ void arm_cspmu_impl_unregister(const struct arm_cspmu_impl_match *impl_match)
> >  
> >  	/* Unbind the driver from all matching backend devices. */
> >  	while ((dev = driver_find_device(&arm_cspmu_driver.driver, NULL,
> > -			match, arm_cspmu_match_device)))
> > +			match, arm_cspmu_match_device))) {
> >  		device_release_driver(dev);
> > +		put_device(dev);
> > +	}
> 
> There's already a fix queued for this; please take a look:
>
> https://git.kernel.org/pub/scm/linux/kernel/git/will/linux.git/commit/?h=for-next/perf&id=970e1e41805f0bd49dc234330a9390f4708d097d

Ah, forgot to check linux-next here, sorry.

The diff of that commit is identical even if the commit summary is a bit
misleading as this is not an error path.

Johan