Gerd Knorr wrote: > So IMHO it would be useful to either: > - require driver do support *both* non-blocking I/O and partial read()s > or > - allow drivers to support only blocking, full-frame read()s. Understood. I'd still vote for the non-blocking support on grounds of select based IO being highly desireable. However I'd note it's only going to useful for portable applications if the video API guarantees that *all* drivers will support it. Ideally bttv would implement non-blocking read by direct dma if there's a pending read, and bounce-buffer if not, but I realize that's in the area of diminishing returns. Purely personally I'm happy with mmap, and just miss select! Ben _______________________________________________ Video4linux-list mailing list [EMAIL PROTECTED] https://listman.redhat.com/mailman/listinfo/video4linux-list
- Re: [V4L] Re: v4l2 + select() + read() Ben Bridgwater
- Re: [V4L] Re: v4l2 + select() + read() Ben Bridgwater
- [V4L] Re: v4l2 + select() + read() Gerd Knorr
- Re: [V4L] Re: v4l2 + select() + read() Ben Bridgwater
- [V4L] Re: v4l2 + select() + read() Gerd Knorr
- Re: [V4L] Re: v4l2 + select() + read() Alan Cox
- Re: [V4L] Re: v4l2 + select() + read() Alan Cox
- Re: [V4L] Re: v4l2 + select() + read() Ben Bridgwater
- [V4L] Re: v4l2 + select() + read() Gerd Knorr
- Re: [V4L] Re: v4l2 + select() + read() Ben Bridgwater
- Re: [V4L] Re: v4l2 + select() + read() Ben Bridgwater
- Re: [V4L] Re: v4l2 + select() + read() Ben Bridgwater
- [V4L] Re: v4l2 + select() + read() Gerd Knorr
- Re: [V4L] Re: v4l2 + select() + read() Mark McClelland
- RE: [V4L] Re: v4l2 + select() + read() Nemosoft Unv.
- Re: [V4L] Re: v4l2 + select() + read() Bill Dirks
- Re: [V4L] Re: v4l2 + select() + read() Justin Schoeman
- [V4L] Re: v4l2 + select() + read() Gerd Knorr
- [V4L] Re: v4l2 + select() + read() Gerd Knorr
- [V4L] Re: v4l2 + select() + read() Gerd Knorr
- RE: [V4L] Re: v4l2 + select() + read() Nemosoft Unv.
