Re: [PATCH v7 45/49] media: core: Add bitmap manage bufs array entries

[Hans Verkuil]

On 21/09/2023 11:28, Benjamin Gaignard wrote:
> 
> Le 20/09/2023 à 16:56, Hans Verkuil a écrit :
>> On 20/09/2023 16:30, Benjamin Gaignard wrote:
>> <snip>
>>
>>>>>        num_buffers = min_t(unsigned int, num_buffers,
>>>>>                    q->max_allowed_buffers - vb2_get_num_buffers(q));
>>>>>    -    first_index = vb2_get_num_buffers(q);
>>>>> +    first_index = bitmap_find_next_zero_area(q->bufs_map, q->max_allowed_buffers,
>>>>> +                         0, num_buffers, 0);
>>>>>          if (first_index >= q->max_allowed_buffers)
>>>>>            return 0;
>>>>> @@ -675,7 +678,13 @@ static void __vb2_queue_free(struct vb2_queue *q, unsigned int buffers)
>>>>>      struct vb2_buffer *vb2_get_buffer(struct vb2_queue *q, unsigned int index)
>>>>>    {
>>>>> -    if (index < q->num_buffers)
>>>>> +    if (!q->bufs_map || !q->bufs)
>>>>> +        return NULL;
>>>> I don't think this can ever happen.
>>> I got kernel crash without them.
>>> I will keep them.
>> What is the backtrace? How can this happen? It feels wrong that this can be
>> called with a vb2_queue that apparently is not properly initialized.
> 
> I have this log when adding dump_stack() in vb2_get_buffer() if !q->bufs_bitmap:
> 
> [   18.924627] Call trace:
> [   18.927090]  dump_backtrace+0x94/0xec
> [   18.930787]  show_stack+0x18/0x24
> [   18.934137]  dump_stack_lvl+0x48/0x60
> [   18.937833]  dump_stack+0x18/0x24
> [   18.941166]  __vb2_queue_cancel+0x23c/0x2f0
> [   18.945365]  vb2_core_queue_release+0x24/0x6c
> [   18.949740]  vb2_queue_release+0x10/0x1c
> [   18.953677]  v4l2_m2m_ctx_release+0x20/0x40
> [   18.957892]  hantro_release+0x20/0x54
> [   18.961584]  v4l2_release+0x74/0xec
> [   18.965110]  __fput+0xb4/0x274
> [   18.968205]  __fput_sync+0x50/0x5c
> [   18.971626]  __arm64_sys_close+0x38/0x7c
> [   18.975562]  invoke_syscall+0x48/0x114
> [   18.979329]  el0_svc_common.constprop.0+0xc0/0xe0
> [   18.984068]  do_el0_svc+0x1c/0x28
> [   18.987402]  el0_svc+0x40/0xe8
> [   18.990470]  el0t_64_sync_handler+0x100/0x12c
> [   18.994842]  el0t_64_sync+0x190/0x194
> 
> This happen at boot time when hantro driver is open and close without other actions.

Ah, now I see the problem. q->bufs and q->bufs_map are allocated in
vb2_core_create_bufs and vb2_core_reqbufs, but they should be allocated
in vb2_queue_init: that's the counterpart of vb2_core_queue_release.

With that change you shouldn't have to check for q->bufs/bufs_map anymore.

Regards,

	Hans

>    
>>
>>>>> +
>>>>> +    return (bitmap_weight(q->bufs_map, q->max_allowed_buffers) > 0);
>>>> How about:
>>>>
>>>>      return vb2_get_num_buffers(q) > 0;
>>> vb2_get_num_buffers is defined in videobuf2-core.c, I'm not sure that
>>> an inline function could depend of a module function.
>> Not a problem. E.g. v4l2-ctrls.h is full of such static inlines.
>>
>> Regards,
>>
>>     Hans
>>