| ↑back Search ←Prev date Next date→ Show only urls | (Click on time to select a line by its url) |
| Who | What | When |
|---|---|---|
| *** | ChanServ sets mode: +v mchehab` | [05:49] |
| ........................................................................... (idle for 6h11mn) | ||
| hverkuil | hi all | [12:00] |
| pinchartl | hello | [12:00] |
| hverkuil | mchehab: ping | [12:02] |
| mchehab: is there still time for another pull request with some small cec patches? There is an extra rc8 so we have another week.
I don't have anything else to discuss, so since mchehab seems absent I'll continue with other stuff :-) | [12:10] | |
| syoung | hi | [12:11] |
| mchehab | sorry, didn't hear the first ping | [12:12] |
| hverkuil | ok | [12:13] |
| mchehab | If we're going to have a delay at -rc, I don't see why not adding small patches | [12:13] |
| hverkuil | OK, I'll prepare a pull request tomorrow. | [12:14] |
| mchehab | assuming that I'll have time to apply them - we're starting anual budget planning
and there will be the trip to EU for ELCE so, I guess next week will be somewhat busy mkrufky: I remember you mentioned something along the week - don't remember what... when I was about to reply to it, I noticed that you were gone from IRC pinchartl, sailus, syoung: anything for today's meeting? | [12:14] |
| hverkuil | mchehab: when will you arrive in Edinburgh
? | [12:18] |
| mchehab | Sunday | [12:18] |
| pinchartl | mchehab: nothing particular for me | [12:18] |
| hverkuil | afternoon?
Basically I am looking for others to go to dinner with. | [12:19] |
| mchehab | I might
I guess I'll arrive at 6:30pm or something | [12:19] |
| hverkuil | arrive at the airport or the hotel at that time? | [12:20] |
| mchehab | no, 7pm - at airport
probably it is better for you to not wait for me | [12:21] |
| hverkuil | yeah, that's going to be a bit late. | [12:22] |
| mchehab | yep | [12:22] |
| hverkuil | Anyone else interested in dinner around (say) 7pm? | [12:22] |
| mkrufky | hello
which day, dinner? | [12:23] |
| hverkuil | Sunday before the ELCE starts
21st | [12:24] |
| mkrufky | mchehab: i thought i had patches for you last week, but I believe that you have already picked them yourself
so i never sent anything, i think there is nothing pending thats urgent | [12:24] |
| mchehab | ok | [12:24] |
| mkrufky | i try to stay on irc, but sometimes the computer gets rebooted. i need a better solution | [12:24] |
| pinchartl | I'll land at 18:55 on the 21st | [12:25] |
| mchehab | here, I use ZNC | [12:25] |
| mkrufky | maybe ill move my irc to a screen session on another box, or something
znc, ill try it | [12:25] |
| mchehab | running on a RPi (or something)
znc is at the init, so, reboots shouldn't harm | [12:25] |
| mkrufky | nice | [12:27] |
| mchehab | anything else for today? | [12:27] |
| mkrufky | i arrive on Oct 20, 3pm
15:00 | [12:27] |
| mchehab | not sure if it would be worth having a meeting next week...
with our merge window closed, and just before a trip for ELCE | [12:27] |
| hverkuil | mchehab: probably not. I vote for canceling it. Same of course for the week after. | [12:28] |
| pinchartl | maybe to discuss the final arrangements for the media summit, but I think that can be done in a ad-hoc fashion | [12:28] |
| mchehab | I'd propose to skip it. If one feels the need, just ping me
yeah, I think so | [12:28] |
| hverkuil | And I might miss the meeting on Nov 1 as well since I have vacation that week.
I plan to post the final (?) media summit agenda on Monday. Not much changed from the current draft. pinchartl: that reminds me: do you want to fold your short dim presentation into the Automated Testing section, or keep it separate? | [12:28] |
| pinchartl | hverkuil: I'm in two minds about that. it's not about just dim, but about different tooling we could/should use
and it's related to automated testing in the sense that any multi-committer model requires safeguards | [12:32] |
| hverkuil | Perhaps schedule it as a separate presentation, but right before the Automated Testing section? | [12:33] |
| pinchartl | I was going to propose that
it would be a good transition | [12:33] |
| hverkuil | OK. Makes sense. | [12:34] |
| pinchartl | mchehab: do you know already what exactly you will talk about in your presentation ? I'm wondering how to best make sure we won't talk about the exact same thing | [12:38] |
| mchehab | no, I don't
thanks for reminding me :-) didn't they change anything? I remember you wanted to ask them to do a different presentation mchehab didn't have any time yet to look at the schedule | [12:40] |
| pinchartl | I contacted Tim Bird to ask for clarifications
especially to understand why two talks were accepted on the same topic | [12:41] |
| mchehab | I have already a presentation I did internally - basically explaining the issues | [12:41] |
| pinchartl | he was open to changes
so I could propose another topic but I have things to say about cameras and I don't really want to ask *you* to pick another topic :-) | [12:41] |
| mchehab | well, I'm not asking you to change the topic... it is just that you said you had some other topics you proposed for ELC
If it would be ok for you, IMO presenting a different topic would be more productive | [12:45] |
| pinchartl | no no, I know, don't worry | [12:45] |
| mchehab | as this way, we can cover more media-related themes at ELC | [12:46] |
| pinchartl | there's another topic that was submitted by a student I mentor that hasn't been accepted and could be interesting (but it's not linux-media development), but on the other hand I also have quite a bit of material for a presentation about cameras | [12:46] |
| mchehab | I didn't think myself on some other theme | [12:47] |
| pinchartl | it's a vast area, so maybe we could use this as an opportunity to cover more | [12:47] |
| mchehab | yeah, sure
the way I'm thinking to talk about complex camera is more focused on compatibility for generic camera apps describing the issues we have due to different ways to talk to a V4L2 and how we're planning to solve via a library | [12:47] |
| hverkuil | mchehab: is there any progress on that? I haven't heard anything since the Tokyo meeting. | [12:49] |
| syoung | hverkuil: apropos dinner I won't arrive until Monday morning (the 22nd), I am afraid | [12:49] |
| mchehab | Google/Intel are working with that...
but they don't have any code yet they're working on allocating human resources for the project | [12:49] |
| hverkuil | Ah. So no progress, in other words. | [12:50] |
| mchehab | (btw, I suspect that the 50' slot we reserved for it won't be used) | [12:50] |
| pinchartl | hverkuil: I've discussed it further with them. if stars aligned there could be something by ELCE time
s/aligned/align/ | [12:50] |
| mchehab | yeah, it is something like that
they're *about* to start | [12:50] |
| pinchartl | not as a usable project, but something public | [12:50] |
| mchehab | It may happen in time for ELCE (or may happen shortly after it) | [12:51] |
| hverkuil | We'll see what happens. | [12:52] |
| mchehab | I'll could probably use maybe 10-15 mins to talk about camorama - I did lots of effots on it, in order to make it easier to migrate to a library
and getting rid of libv4l version 1 I suspect it is the simplest generic app, in terms of V4L2 usage there's nothing there for tuners and no need of gstreamer | [12:53] |
| hverkuil | mchehab: unrelated to ELCE: what's the status of the strncpy and strcpy conversions? I liked it very much, and one of the three patches was merged, but the other two needed a bit more work. But were close as well. | [12:55] |
| mchehab | it implements just the ioctls used on cameras
completely forgot about that I *guess* two patches got merged but need to double check I'll put it on my todo list to check what got applied and what is missing Hmm... I guess *you* said you would be reviewing the strlcpy patch in order to check if the string fields that are supposed to be filled by the driver were zeroed | [12:55] |
| hverkuil | The strlcpy patch is merged. I had some comments for the other two.
(I acked the strlcpy patch as well) It would be nice to get it in for 4.20. I really like the idea of using strscpy everywhere in media. | [12:58] |
| mchehab | yes, agreed | [12:59] |
| hverkuil | Ah, the strcpy patch I marked as Reviewed-by. | [13:00] |
| mchehab | I'll rebase and send again | [13:00] |
| hverkuil | Ah, my strncpy review ends with this: | [13:00] |
| mchehab | yes, that's why I *think* that two patches got merged, and there's just one missing | [13:00] |
| hverkuil | "The 'strcpy' patch uses sizeof everywhere, but I saw a few other occurrences of this in the strlcpy patch. I think the strlcpy patch needs a follow-up that replaces any hardcoded sizes with the safer sizeof()."
strncpy wasn't merged, but the other two were. But I think the strlcpy one needs a follow-up for the reasons stated above. | [13:01] |
| mchehab | ok, I'll handle it | [13:02] |
| hverkuil | I'll prioritize reviewing those patches when you post them. | [13:03] |
| mchehab | Ok
cc1e6315e83d media: replace strcpy() by strscpy() c0decac19da3 media: use strscpy() instead of strlcpy() the third patch still applies upstream There are two new occurrences of strlcpy that require fixes too: drivers/media/platform/imx-pxp.c: strlcpy(cap->driver, MEM2MEM_NAME, sizeof(cap->driver)); drivers/media/platform/imx-pxp.c: strlcpy(cap->card, MEM2MEM_NAME, sizeof(cap->card)); | [13:03] |
| hverkuil | $ git grep strcpy drivers/media/
drivers/media/rc/imon.c: Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â strcpy(buf, "associating\n"); drivers/media/rc/imon.c: Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â strcpy(buf, "closed\n"); | [13:08] |
| mchehab | $ git grep -E str.?cpy drivers/media/ drivers/staging/media/|grep -v strscpy
is actually a better way to get it drivers/media/pci/ttpci/av7110_hw.c: if (strncpy_from_user(textbuf, dc->data, 240) < 0) { drivers/media/platform/imx-pxp.c: strlcpy(cap->driver, MEM2MEM_NAME, sizeof(cap->driver)); drivers/media/platform/imx-pxp.c: strlcpy(cap->card, MEM2MEM_NAME, sizeof(cap->card)); drivers/media/rc/imon.c: strcpy(buf, "associating\n"); drivers/media/rc/imon.c: strcpy(buf, "closed\n"); (after the pending patch) anyway, I'll address it hverkuil: in the specific case of imon, not sure how to fix I remember I faced this issue (not sure if I fixed or just skip) | [13:09] |
| hverkuil | I see we use sprintf elsewhere for sysfs 'show' callbacks.
(I seem to remember that 'buf' points to a full PAGE) Anyway, I suggest that strcpy is replaced by sprintf so we can still easily grep for str[ln]?cpy without getting false positives. | [13:15] |
| syoung | yes, that's right. You can use PAGE_SIZE. See https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/usb/host/ehci-sysfs.c for example | [13:19] |
| mchehab | yes, PAGE_SIZE is a way to solve it
as sysfs is limited to 1 page (I guess debug fs too) I'll probably be replacing <constant> by sizeof() on a separate patch I may use coccinelle to do such change, and then manually review | [13:24] |
| hverkuil | mchehab: make sure you get the strscpy in pulse8_cec_adap_log_addr() right:
it starts at cmd + 1, so the sizeof must be sizeof(cmd) - 1. | [13:28] |
| ......... (idle for 40mn) | ||
| mchehab: since you suspect that your 'Complex Cameras' topic during the media summit won't get used, I propose that I move it to the end of the day. If it is canceled, then it won't affect the remainder of the agenda.
is that ok? | [14:09] | |
| ..... (idle for 22mn) | ||
| mchehab | hverkuil: sure | [14:31] |
| ↑back Search ←Prev date Next date→ Show only urls | (Click on time to select a line by its url) |