| ↑back Search ←Prev date Next date→ Show only urls | (Click on time to select a line by its url) |
| Who | What | When |
|---|---|---|
| *** | lexano has quit IRC (Ping timeout: 480 seconds)
mrpops2ko_ has quit IRC (Ping timeout: 480 seconds) | [00:08] |
| ...... (idle for 26mn) | ||
| mrpops2ko has joined #linux-media | [00:38] | |
| ....... (idle for 33mn) | ||
| xecutercmd has joined #linux-media | [01:11] | |
| xecuter has quit IRC (Ping timeout: 480 seconds) | [01:17] | |
| .... (idle for 15mn) | ||
| ten157237743246305066 has quit IRC (Remote host closed the connection)
ten157237743246305066 has joined #linux-media | [01:32] | |
| ................ (idle for 1h19mn) | ||
| eelstrebor has joined #linux-media | [02:51] | |
| .................. (idle for 1h27mn) | ||
| eelstrebor has quit IRC (Quit: Ex-Chat)
eelstrebor has joined #linux-media eelstrebor has quit IRC () | [04:18] | |
| .... (idle for 19mn) | ||
| ten1572377432463050664 has joined #linux-media
ten157237743246305066 has quit IRC (Quit: Ping timeout (120 seconds)) ten1572377432463050664 is now known as ten157237743246305066 | [04:38] | |
| ......... (idle for 44mn) | ||
| mvchtz has quit IRC (Server closed connection)
mvchtz has joined #linux-media | [05:22] | |
| .... (idle for 17mn) | ||
| yang2 has quit IRC (Server closed connection) | [05:40] | |
| yang2 has joined #linux-media | [05:45] | |
| jkqxz has quit IRC (Server closed connection)
jkqxz has joined #linux-media | [05:52] | |
| ..... (idle for 24mn) | ||
| djrscally has joined #linux-media | [06:17] | |
| .......... (idle for 47mn) | ||
| eloy_ has quit IRC (Server closed connection)
eloy_ has joined #linux-media | [07:04] | |
| larsc has quit IRC (Server closed connection)
larsc has joined #linux-media | [07:14] | |
| ....... (idle for 32mn) | ||
| _0andriy_ has quit IRC (Server closed connection)
_0andriy_ has joined #linux-media | [07:46] | |
| ........ (idle for 39mn) | ||
| Mo has joined #linux-media | [08:25] | |
| tomba | How does system suspend work with v4l2? Each device in the pipeline individually suspends/resumes itself, and these happen in "random" order? | [08:26] |
| Found something in camera-sensors.rts. It refers to "bridge drivers". What is that? In DRM, bridge drivers are devices between the display controller and the display. Is bridge driver here the driver for the v4l2 video device? | [08:31] | |
| *** | andrzej_p3 has quit IRC (Server closed connection)
andrzej_p3 has joined #linux-media ndufresne has quit IRC (Server closed connection) ndufresne has joined #linux-media bingbu has quit IRC (Ping timeout: 480 seconds) bingbu has joined #linux-media ocrete has quit IRC (Server closed connection) ocrete has joined #linux-media SebastianCollabora has quit IRC (Server closed connection) SebastianCollabora has joined #linux-media | [08:37] |
| javierm | tomba: AFAIK "bridge drivers" are the drivers for a CSI-2 or ISP, the drivers that register a media device basically | [08:52] |
| pinchartl | tomba: "bridge" is a term I dislike very much in V4L2, it's confusing
it's the "main" driver the "main" driver stops streaming on the whole pipeline at suspend time and restarts it at resume it. the other drivers don't do anything special beside turning the clocks/regulators off and on of course | [08:54] |
| javierm | tomba: e.g: https://elixir.bootlin.com/linux/v6.10/source/drivers/media/platform/renesas/renesas-ceu.c#L1074
pinchartl: maybe should be "media device driver" and "v4l2 sub device driver" ? | [08:55] |
| pinchartl | in many cases that's what it is, yes | [08:57] |
| *** | mal has quit IRC (Server closed connection)
mal has joined #linux-media | [08:57] |
| ........ (idle for 36mn) | ||
| bingbu has quit IRC (Ping timeout: 480 seconds)
bingbu has joined #linux-media cyrinux has quit IRC (Server closed connection) cyrinux has joined #linux-media | [09:33] | |
| sgruszka has joined #linux-media | [09:43] | |
| detlev has quit IRC (Server closed connection)
detlev has joined #linux-media tyalie has quit IRC (Server closed connection) tyalie has joined #linux-media | [09:53] | |
| lplc_ | sailus, pinchartl: hi, I have a question about line-based metadata video nodes: is the width, height and bytesperline in v4l2_format.meta supposed to be returned by the driver in VIDIOC_S_FMT? | [09:59] |
| sailus | lplc_: Generally yes, at least when it comes to sensor embedded data.
See https://www.retiisi.eu/~sailus/v4l2/tmp/streams-doc/userspace-api/media/drivers/camera-sensor.html#id1 . Note that this isn't in upstream yet. | [10:08] |
| lplc_ | sailus: what about devices that sit downstream and receive both pixel data and metadata from a sensor? We don't know in advance what sensor might be connected and what the width and height of the metadata is.
for example: sensor->CSI2->ISI (for an iMX chip) | [10:12] |
| .... (idle for 19mn) | ||
| *** | lplc_ has quit IRC ()
lplc has joined #linux-media | [10:31] |
| pinchartl | lplc: the subsev .get_frame_desc() operation is used to handle that
it allows a downstream device to query the upstream device, and get information about the different streams that flow through the pipeline | [10:34] |
| *** | bemug has quit IRC (Server closed connection)
bemug has joined #linux-media | [10:39] |
| tomba | pinchartl: I don't think that's related. get_frame_desc can only be used at stream-start time. VIDIOC_S_FMT needs to operate independently. | [10:40] |
| lplc | pinchartl: also, with get_frame_desc() we don't get the width/height of a stream... :/
am I missing something? | [10:42] |
| tomba | lplc: I don't think I understand your original question, though. width, height and bytesperline are supposed to be set by the userspace when calling VIDIOC_S_FMT. what do you mean with "to be returned by the driver"? | [10:43] |
| pinchartl | lplc: indeed, I thought width and height were there too, I misremembered | [10:44] |
| lplc | tomba: that's actually my question: is the userspace supposed to set the width and height and the driver returns the bytesperline based on the metadata format depth? | [10:45] |
| pinchartl | lplc: it's the same as with pixel formats. userspace can set the values, the kernel will adjust them based on whatever constraints the device has, and return the adjusted values | [10:46] |
| *** | Jacmet has quit IRC (Server closed connection)
Jacmet has joined #linux-media | [10:48] |
| lplc | pinchartl: I guess my confusion was born from my attempts to use v4l2-ctl --device /dev/video1 --set-fmt-meta=width=1920,height=2,pixelformat=MET8 to see that the width and height are not actually passed over from userspace by the v4l2-ctl app :/ | [10:49] |
| tomba | lplc: if the userspace leaves the bytesperline 0, the driver should adjust it to (probably) the minimum allowed. but the userspace can set bytesperline to a bigger amount too, thus allowing wider buffers than what the actual width is (if the hw allows that). | [10:50] |
| pinchartl | lplc: I'm not sure support for line-based metadata format has been merged in v4l-utils already | [10:52] |
| lplc | pinchartl, tomba: thanks | [10:56] |
| *** | nst has quit IRC (Server closed connection)
nst has joined #linux-media | [11:00] |
| pinchartl | lplc: does this mean I should expect patches for the ISI driver to support capturing metadata ? :-) | [11:02] |
| lplc | pinchartl: I actually had a POC working, using sailus's separate metadata patches, using one ISI instance for pixel data and one for non-pixel data, but I realized I did wrong (even though I was able to capture the video and metadata separately) as both nodes were configured as pixel-data nodes... | [11:13] |
| pinchartl | which SoC has two ISI instances ? or is it one ISI instance with multiple channels ? | [11:15] |
| lplc | now I'm revisiting that topic and I'm actually trying to make it work with using just one ISI instance, after I learned that it could actually be possible to have pixel-data and metadata saved in separate buffers using only one ISI instance
pinchartl: poor wording, I meant ISI channels | [11:15] |
| pinchartl | s/channels/pipelines/ if I recall the ISI terminology correctly
ok | [11:15] |
| *** | xecutercmd has quit IRC (Remote host closed the connection) | [11:16] |
| pinchartl | pinchartl wish he could capture embedded data using the ISI on i.MX8MP | [11:16] |
| *** | q66 has quit IRC (Quit: WeeChat 4.3.5)
q66 has joined #linux-media tleb has quit IRC (Server closed connection) tleb has joined #linux-media xecuter has joined #linux-media | [11:17] |
| ...... (idle for 25mn) | ||
| lexano has joined #linux-media | [11:47] | |
| jernej has quit IRC (Server closed connection)
jernej has joined #linux-media | [11:52] | |
| ....... (idle for 32mn) | ||
| sgruszka has quit IRC (Ping timeout: 480 seconds) | [12:24] | |
| .... (idle for 16mn) | ||
| wallacer has quit IRC (Server closed connection)
wallacer has joined #linux-media | [12:40] | |
| .... (idle for 17mn) | ||
| Mo has quit IRC (Remote host closed the connection)
Mo has joined #linux-media Mo has quit IRC (Remote host closed the connection) Mo has joined #linux-media | [12:57] | |
| .......... (idle for 47mn) | ||
| Mo has quit IRC (Remote host closed the connection) | [13:45] | |
| mtretter has quit IRC (Server closed connection)
mtretter has joined #linux-media | [13:54] | |
| ............ (idle for 59mn) | ||
| charrus has quit IRC (Server closed connection)
charrus has joined #linux-media | [14:53] | |
| ..... (idle for 23mn) | ||
| tmerciai1 has quit IRC (Server closed connection)
tmerciai1 has joined #linux-media Moarc has quit IRC (Server closed connection) Moarc has joined #linux-media | [15:16] | |
| ..... (idle for 23mn) | ||
| cyrinux has quit IRC ()
cyrinux has joined #linux-media | [15:42] | |
| ..... (idle for 22mn) | ||
| eelstrebor has joined #linux-media | [16:04] | |
| ............................... (idle for 2h34mn) | ||
| eelstrebor has quit IRC (Quit: Ex-Chat)
jaimin has joined #linux-media | [18:38] | |
| ....................... (idle for 1h54mn) | ||
| jmassot has quit IRC (Quit: Leaving) | [20:33] | |
| ............... (idle for 1h10mn) | ||
| jaimin has quit IRC (Remote host closed the connection) | [21:43] | |
| ................... (idle for 1h32mn) | ||
| djrscally has quit IRC (Ping timeout: 480 seconds) | [23:15] | |
| darkapex has quit IRC (Remote host closed the connection)
darkapex has joined #linux-media | [23:29] | |
| ↑back Search ←Prev date Next date→ Show only urls | (Click on time to select a line by its url) |