Hi all, just a minor point of critique on v4l2 - why isn't there a field '__u32 size' in the struct v4l2_requestbuffers? For raw devices, calculating the required minimum size isn't that much of a problem - just width * height * bpp, but for hardware-encoding devices, you might want a specific buffersize that is in some way independent of the image size. Higher buffersizes can account for higher quality here, so buffersize is pretty important. Just mmap()'ing a certain buffersize and hoping that that's okay doesn't look like the way to go to me ;-). Is there something I'm completely overseeing here? Ronald -- - .-. - /V\ | Ronald Bultje <rbultje@xxxxxxxxxxxxxxxxxxx> - // \\ | Running: Linux 2.4.17-XFS and OpenBSD 3.0 - /( )\ | http://ronald.bitfreak.net/ - ^^-^^
