IRC logs for #automotive for Tuesday, 2021-01-19

*** leon-anavi has quit IRC00:13
*** toscalix_ has quit IRC01:48
*** Wetmelon has quit IRC04:10
*** anjumkaiser has joined #automotive05:25
*** vrubiolo has joined #automotive07:54
*** walzert has joined #automotive08:13
*** walzert has quit IRC09:47
paulbarkerdl9pf: https://www.mail-archive.com/debian-bugs-dist@lists.debian.org/msg1743984.html10:14
paulbarkerdl9pf: Maybe try dropping "cma=64M" from the kernel cmdline10:19
paulbarkerI've just tried adding that here to see if it stops audio from coming up, but instead it completely stops the kernel from booting10:20
paulbarkerOk, so "cma=64M" prevents the SD card from being read. That won't be a problem in CI if you're net booting10:24
*** vrubiolo has quit IRC10:29
*** vrubiolo has joined #automotive10:29
dl9pfbut it might lead to other issues.10:54
dl9pfok we can take cma=64M out ...10:54
dl9pfthe same filesystem works if booted from the sdcard10:56
dl9pfso our kernel cfg seems ok wrt modules and such10:56
paulbarkerdl9pf: Yes, it looks like it breaks MMC, audio, several other drivers which all depend on the vchiq mailboxes10:56
dl9pfthats all due to cma ? ... heh ... that setting was just for things to work wrt gfx back then. well sure. Can remove. But might break older kernels in CI/test then.10:57
dl9pfworth a try10:57
dl9pf<rant mode>undocumented *****</rant mode>10:58
paulbarkerPretty standard for rpi10:58
dl9pfindeed10:59
paulbarkerIt will be annoying if the kernel cmdline in CI needs to depend on the kernel version11:00
dl9pfthat would be 'jikes' but not impossible ...11:01
dl9pfok11:01
dl9pf+ aplay -l11:01
dl9pf**** List of PLAYBACK Hardware Devices ****11:01
dl9pfcard 0: Headphones [bcm2835 Headphones], device 0: bcm2835 Headphones [bcm2835 Headphones]11:01
dl9pf  Subdevices: 8/811:01
dl9pf  Subdevice #0: subdevice #011:01
dl9pf  Subdevice #1: subdevice #111:01
dl9pf  Subdevice #2: subdevice #211:01
paulbarkerhttps://www.raspberrypi.org/forums/viewtopic.php?p=1674488#p1674488 may be relevant11:01
dl9pf  Subdevice #3: subdevice #311:01
dl9pf  Subdevice #4: subdevice #411:01
dl9pf  Subdevice #5: subdevice #511:01
dl9pf  Subdevice #6: subdevice #611:01
dl9pf  Subdevice #7: subdevice #711:01
dl9pf+ set +x11:01
dl9pfhttps://lava.automotivelinux.org/scheduler/job/15850   with cma=64M off11:01
dl9pfeasy to fix11:02
dl9pfread half through that post and so far all just half-baked and no explanation as to why11:04
dl9pfbut thats how it is11:04
paulbarkerdl9pf: For me, using "cma=64M@256M" works, just "cma=64M" breaks11:05
paulbarkerI believe the issue is where in memory that cma block gets allocated, not its size11:05
dl9pfheh ... shouldn't there be known boundaries ... lol .11:06
dl9pfwell cma=64m@256m sound good to me then ... trying11:06
dl9pfhttps://lava.automotivelinux.org/scheduler/job/1585211:07
dl9pfif I recall correctly cma=64 was necessary to render gfx11:08
paulbarkerIt should be set in the device tree these days11:10
dl9pfiirc there are settings in there now11:10
dl9pfhttps://github.com/raspberrypi/linux/blob/88debfb15b3ac9059b72dc1ebc5b82f3394cac87/arch/arm/boot/dts/bcm2711.dtsi#L87111:12
paulbarkerThe "cma=..." cmdline argument overrides what's in the dtb11:14
dl9pfok, so better kill the cmdline as its now in the devicetree and let it handle that11:14
dl9pfless to care about11:14
dl9pfhttps://lava.automotivelinux.org/scheduler/job/15852#L2446   works, thus let me deploy that change and retrigger the 5.4 job11:15
dl9pfnasty little bugger11:15
paulbarkerhttps://github.com/raspberrypi/linux/blob/88debfb15b3ac9059b72dc1ebc5b82f3394cac87/arch/arm/boot/dts/bcm283x.dtsi#L4011:16
paulbarkerThat's the current default size11:16
paulbarker64M11:16
dl9pfok, so should be enough space by default.11:17
dl9pffine11:17
dl9pflet's remove the direct setting and be done11:17
paulbarkerSounds good11:17
paulbarkerThe error you posted late last night put me onto the right track11:18
dl9pfonly left now:11:18
dl9pfextra_kernel_args: "coherent_pool=1M 8250.nr_uarts=1 rw fstab=no "11:18
dl9pfcool - glad it worked out.11:18
paulbarkerThe firmware adds more options by default but of course none of them are documented11:18
paulbarkerI think we just have to say that if the tests pass we go with what we've got11:19
dl9pfwas already thinking 'omg /me old mouldwarp, that can't be rocket science'11:20
dl9pfw/o the additional options we also see:11:22
dl9pfhttps://www.irccloud.com/pastebin/qxo2vCD9/11:22
*** leon-anavi has joined #automotive11:22
leon-anavihi11:23
dl9pfwhats the jira again ?11:25
dl9pfwill use SPEC-370211:30
paulbarkerdl9pf: Yes that's the one11:36
*** AlisonChaiken has quit IRC11:46
*** kooltux has quit IRC11:57
*** kooltux has joined #automotive11:58
*** leon-anavi has quit IRC11:58
dl9pfok one down12:02
*** anjumkaiser has quit IRC13:10
*** anjumkaiser has joined #automotive13:11
*** Newami has joined #automotive14:26
*** Newami has quit IRC14:27
*** eduardas has joined #automotive15:06
*** anjumkaiser has quit IRC15:22
*** anjumkaiser has joined #automotive16:23
dl9pfmvlad: ping16:42
dl9pfmvlad: See https://lava.automotivelinux.org/scheduler/job/15868#L510516:42
dl9pfmvlad: any idea what16:42
dl9pf[  887.275053] vmap allocation for size 8298496 failed: use vmalloc=<size> to increase size16:42
dl9pfmeans we need to do ?16:42
dl9pfthis is qemu-system-arm16:42
*** eduardas has quit IRC16:48
*** eduardas has joined #automotive16:48
*** eduardas has quit IRC16:49
*** vrubiolo has quit IRC16:58
dl9pfmontjoie: try this:17:53
dl9pf           image_arg: '-kernel {kernel} -append "console=ttyAMA0,115200 root=/dev/vda verbose systemd.log_color=false mem=2048M vmalloc=256M"'17:53
dl9pfsee last 2 options added17:53
dl9pfseems slow, need to check17:53
montjoiedl9pf: I have started one with memory 4096 https://lava.automotivelinux.org/scheduler/job/15869 the message goes away17:56
montjoiethe screenshot is still bad17:57
*** toscalix has joined #automotive18:17
dl9pfyes18:30
dl9pfyou need the vmalloc= kernel cmdline ... figuring out the value18:31
dl9pfmvlad: https://lava.automotivelinux.org/scheduler/job/15883#L248918:47
dl9pf"Usable area:" does not show on qemu-system-arm18:48
*** eduardas has joined #automotive19:03
*** eduardas has quit IRC19:11
*** Wetmelon has joined #automotive21:12
*** anjumkaiser has quit IRC21:22
*** Wetmelon has quit IRC21:51
*** Net147 has quit IRC22:05
*** Net147 has joined #automotive22:06
*** bn_mobile has joined #automotive22:13
*** Wetmelon has joined #automotive22:42
*** toscalix has quit IRC22:49
*** Net147 has quit IRC23:15
*** Net147 has joined #automotive23:15

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!