hi all
hi all
sailus: pinchartl: syoung: ping
FYI, I merged the main media_tree.git Kernel builds at Jenkins into a single pipelined one:
https://builder.linuxtv.org/job/kernel_media_pipeline/
hi all
it builds allmod/allyes for x86_64, and allyes for: i386, arm32, arm64. It also builds the Kernel docs (both PDF and html)
nice
my plan is to change my workflow to apply patches on a temp tree, letting the pipeline to push back at the master tree only in case of success
this way, if something goes wrong on a build for an arch, we can fix the tree without needing to do rebase
(this would add a merge delay of ~30 mins or so)
ok
hverkuil: with regards to the 3 PRs you sent this week, I applied just one
the other two have some checkpatch warnings that sounds worth to take a look
in particular, one has a problem with a SOB from Ricardo
(I didn't look at the PR contents - just read the jenkins results)
I missed that one in the jenkins results. I'll take a look.
Hmm, "Unknown commit id": I'll look at that as well.
The HEVC pull request is ready, the checkpatch warnings are all expected and they make sense.
still looking for reviews of this series: https://patchwork.linuxtv.org/cover/58918/  (ioctl validation)
It's blocking the merge of Vandana's vivid metadata patches.
If I don't get any reviews by this weekend, then I'll just post a PR for it.
I have some fixes for 5.4 pending, but I'll take those next week (some appear to be still under discussion as well)
ok
Anything else?
fyi, I've been very busy this month and probably the beginning of the next one
so, I may eventually delay more complex stuff that would require me more time to review
good to know
syoung: I applied the changes to support a si2168 board, as there was some mess at the ML with my patches plus the ones written by the guy who tested...
but I probably won't have more time for doing much related to dvb... so feel free to handle any followup stuff - and the two patches I wrote myself for DVB
that are at patchwork (the ones that aren't delegated to myself)
ok
I'll probably order that dvb device on ebay and see if I can figure out what's going on. There must be a better solution.
syoung: yeah, I strongly suspect that it has a custom firmware
antti wrote some extracting tools for the firmware...
not sure if they're at the Kernel tree or not
I could see if I can dump the i2c comms and figure it out from there
hmm... scripts/get_dvb_firmware has only si2165
not sure if the firmware for the other si devices do the same
probably not
well, try to ping Antti
he is not too active those days, but he may eventually answer you via e-mail or if you ping him on IRC
(his nick on IRC is "crope")
anything else for today's meeting?
I'll report the PR for ti-vpe and V4L2_CTRL_TYPE_AREA/V4L2_CID_UNIT_CELL_SIZE
repost
You'll still get warnings, but those are expected (long tables)
Done.
ok
what does this jenkins error mean? error: FETCH_HEAD: cannot verify a non-tag object of type commit.
Do I have to do anything to avoid that?
mchehab: I've ordered the dvb with the si2168.
syoung: nice!
hverkuil: that a git error. it means that it can't check your signature
in thesis it occurs if you send a PR for a branch instead of a tag
basically, you can sign a tag, but you can't sign a branch
hverkuil: don't want to be picky, but why can't HEVC headers use the BIT() macros?
Subject: media: v4l: Add definitions for HEVC stateless decoding
hverkuil: with regards to that FETCH_HEAD error:
https://git.linuxtv.org/hverkuil/media_tree.git/log/?h=for-v5.5h:
media: ti-vpe: vpe: don't rely on colorspace member for conversion    - br-v5.5h is a branch, for-v5.5h is the tag
you sent the PR with br-v5.5h and not with for-v5.5h
that's why it complained
I'll handle it anyway
Oops, my fault. I did make a signed tag br-v5.5h, so check out from that tag.
Updated my script to prevent this from happening again.