Re: [videobuf] Query: Condition bytesize limit in videobuf_reqbufs -> buf_setup() call?

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Pawel Osciak wrote:
> Hi,
> 
>> Aguirre, Sergio wrote:
>> Basically, when calling VIDIOC_REQBUFS with a certain buffer
>> Count, we had a software limit for total size, calculated depending on:
>>
>>  Total bytesize = bytesperline x height x count
>>
>> So, we had an arbitrary limit to, say 32 MB, which was generic.
>>
>> Now, we want to condition it ONLY when MMAP buffers will be used.
>> Meaning, we don't want to keep that policy when the kernel is not
>> allocating the space
>>
>> But the thing is that, according to videobuf documentation, buf_setup is
>> the one who should put a RAM usage limit. BUT the memory type passed to
>> reqbufs is not propagated to buf_setup, therefore forcing me to go to a
>> non-standard memory limitation in my reqbufs callback function, instead
>> of doing it properly inside buf_setup.
> 
> buf_setup is called during REQBUFS and is indeed the place to limit the
> size and actually allocate the buffers as well, or at least try to do so,
> as V4L2 API requires. This is not currently the case with videobuf, but
> right now we are working to change it. 

I can't see the problem you're mentioning. Drivers apply (or should apply)
the maximum size limit at buffer setup. For example bttv driver seems to do
the right thing:

static unsigned int gbuffers = 8;
static unsigned int gbufsize = 0x208000;
...
MODULE_PARM_DESC(gbuffers,"number of capture buffers. range 2-32, default 8");
MODULE_PARM_DESC(gbufsize,"size of the capture buffers, default is 0x208000");

...
static int
buffer_setup(struct videobuf_queue *q, unsigned int *count, unsigned int *size)
{
        struct bttv_fh *fh = q->priv_data;

        *size = fh->fmt->depth*fh->width*fh->height >> 3;
        if (0 == *count)
                *count = gbuffers;
        if (*size * *count > gbuffers * gbufsize)
                *count = (gbuffers * gbufsize) / *size;
        return 0;
}

> buf_prepare() is called on QBUF
> and it is definitely too late to do things like that then. It is the
> REQBUFS that should be failing if the requested number of buffers is too
> high.

Yes, it is too late. Restrictions like that should be done at REQBUFS.

-- 

Cheers,
Mauro
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[Index of Archives]     [Linux Input]     [Video for Linux]     [Gstreamer Embedded]     [Mplayer Users]     [Linux USB Devel]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [Yosemite Backpacking]
  Powered by Linux