| ↑back Search ←Prev date Next date→ Show only urls | (Click on time to select a line by its url) |
| Who | What | When |
|---|---|---|
| gnurou | hverkuil: I will resend a new version after looking at a few other issues we have on MTK | [03:34] |
| .......................... (idle for 2h7mn) | ||
| tfiga | hverkuil: I think the fact that we take plane lengths from the userspace v4l2_plane in __verify_length() is a bug
https://elixir.bootlin.com/linux/v5.5-rc2/source/drivers/media/common/videobuf2/videobuf2-v4l2.c#L102 because the spec clarly says the field is set by the driver https://www.kernel.org/doc/html/latest/media/uapi/v4l/buffer.html#struct-v4l2-plane | [05:41] |
| ................................................. (idle for 4h2mn) | ||
| *** | hfr has left | [09:44] |
| .... (idle for 19mn) | ||
| Kwiboo | bbrezillon: I did not test your rkvdec / hantro branch this past weekend, instead I played around with your bus format series and narmstrong yuv420 series and got yuv and 10-bit output to work on my rk3328 to be ready for testing some 10-bit decoding action ;)
see https://github.com/Kwiboo/linux-rockchip/commits/next-20191213-bus-format for work-in-progress patches, I hope I can test your hantro/rkvdec branch later this week | [10:03] |
| bbrezillon | Kwiboo: s/test/debug/ :)
and thanks for working on 10-bit support | [10:06] |
| hverkuil | tfiga: the code is right, the spec is outdated. Userspace should set length for USERPTR and DMABUF since that is not under control of vb2. | [10:10] |
| .................. (idle for 1h29mn) | ||
| tfiga | hverkuil: DMABUF is definitely under control of vb2, as the buffer is kernel managed and for USERPTR it's arguable that anything else than bytesused is needed for OUTPUT buffers
I guess for CAPTURE buffers one would to set the limit indeed to avoid overwriting some data would want to* but I think that's already defined by sizeimage and checked by drivers queue_setup | [11:39] |
| ....................... (idle for 1h50mn) | ||
| mort | so, I just had the venus driver end up stalling a CPU while waiting for a spinlock: https://pastebin.com/va3aS6bZ
probably in the video decoder part of the driver; the video encoder was running and encoding frames just fine | [13:32] |
| ........................................ (idle for 3h16mn) | ||
| *** | benjiG has left | [16:49] |
| ................... (idle for 1h33mn) | ||
| larkfisherman | Hey guys
I need to use video4linux2 for writing some applications for frame capturing/processing for my usb camera I've been reading through wiki for the last few days but v4l2 seems to be quite huge - can anyone point me in the right direction or give me some resources to start with? I'm particularly only interested in usb camera stuff - capabilites, settings, frame capturing, etc... | [18:22] |
| .......... (idle for 49mn) | ||
| *** | larkfisherman has left | [19:13] |
| ↑back Search ←Prev date Next date→ Show only urls | (Click on time to select a line by its url) |