Re: [RFC PATCH 03/13] staging: mmal-vchiq: Fix memory leak in error path

[Maarten <maarten@rmail.be>]

Andrzej Pietrasiewicz schreef op 2024-03-04 08:38:
> Hi Maarten,
> 
> W dniu 3.03.2024 o 16:09, Maarten Vanraes pisze:
>> From: Dave Stevenson <dave.stevenson@raspberrypi.org>
>> 
>> On error, vchiq_mmal_component_init could leave the
>> event context allocated for ports.
>> Clean them up in the error path.
>> 
>> Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
>> 
>> staging: mmal-vchiq: Free the event context for control ports
>> 
>> vchiq_mmal_component_init calls init_event_context for the
>> control port, but vchiq_mmal_component_finalise didn't free
>> it, causing a memory leak..
>> 
>> Add the free call.
>> 
>> Signed-off-by: Dave Stevenson <dave.stevenson@raspberrypi.org>
>> Signed-off-by: Maarten Vanraes <maarten@rmail.be>
>> ---
>>   .../vc04_services/vchiq-mmal/mmal-vchiq.c     | 29 
>> ++++++++++++++-----
>>   1 file changed, 22 insertions(+), 7 deletions(-)
>> 
>> diff --git a/drivers/staging/vc04_services/vchiq-mmal/mmal-vchiq.c 
>> b/drivers/staging/vc04_services/vchiq-mmal/mmal-vchiq.c
>> index 2e616604943d..1209b7db8f30 100644
>> --- a/drivers/staging/vc04_services/vchiq-mmal/mmal-vchiq.c
>> +++ b/drivers/staging/vc04_services/vchiq-mmal/mmal-vchiq.c
>> @@ -1825,9 +1825,26 @@ static void free_event_context(struct 
>> vchiq_mmal_port *port)
>>   {
>>   	struct mmal_msg_context *ctx = port->event_context;
>>   +	if (!ctx)
>> +		return;
>> +
>>   	kfree(ctx->u.bulk.buffer->buffer);
>>   	kfree(ctx->u.bulk.buffer);
>>   	release_msg_context(ctx);
>> +	port->event_context = NULL;
>> +}
>> +
>> +static void release_all_event_contexts(struct vchiq_mmal_component 
>> *component)
>> +{
>> +	int idx;
>> +
>> +	for (idx = 0; idx < component->inputs; idx++)
>> +		free_event_context(&component->input[idx]);
>> +	for (idx = 0; idx < component->outputs; idx++)
>> +		free_event_context(&component->output[idx]);
>> +	for (idx = 0; idx < component->clocks; idx++)
>> +		free_event_context(&component->clock[idx]);
>> +	free_event_context(&component->control);
>>   }
>>     /* Initialise a mmal component and its ports
>> @@ -1925,6 +1942,7 @@ int vchiq_mmal_component_init(struct 
>> vchiq_mmal_instance *instance,
>>     release_component:
>>   	destroy_component(instance, component);
>> +	release_all_event_contexts(component);
>>   unlock:
>>   	if (component)
>>   		component->in_use = false;
>> @@ -1940,7 +1958,7 @@ EXPORT_SYMBOL_GPL(vchiq_mmal_component_init);
>>   int vchiq_mmal_component_finalise(struct vchiq_mmal_instance 
>> *instance,
>>   				  struct vchiq_mmal_component *component)
>>   {
>> -	int ret, idx;
>> +	int ret;
>>     	if (mutex_lock_interruptible(&instance->vchiq_mutex))
>>   		return -EINTR;
>> @@ -1952,12 +1970,9 @@ int vchiq_mmal_component_finalise(struct 
>> vchiq_mmal_instance *instance,
>>     	component->in_use = false;
>>   -	for (idx = 0; idx < component->inputs; idx++)
>> -		free_event_context(&component->input[idx]);
>> -	for (idx = 0; idx < component->outputs; idx++)
>> -		free_event_context(&component->output[idx]);
>> -	for (idx = 0; idx < component->clocks; idx++)
>> -		free_event_context(&component->clock[idx]);
>> +	release_all_event_contexts(component);
> 
> The way I understand this chunk is that you factor out the 3 "for" 
> loops into
> the new function "release_all_event_contexts()", because it is then
> reused elsewhere. "release_all_event_contexts()" already contains
> invocation of
> "free_event_context(&component->control)"...
> 
>> +
>> +	free_event_context(&component->control);
> 
> ... but it is repeated here. Why?


I'm sorry, I messed up in squashing 2 patches, that definately does not 
belong there, will fix in v2.

Thanks!