mmind00 changed the topic of #linux-rockchip to: Rockchip development discussion | public log at https://libera.irclog.whitequark.org/linux-rockchip
qschulz has quit [Remote host closed the connection]
qschulz has joined #linux-rockchip
nashpa has joined #linux-rockchip
dliviu has quit [Ping timeout: 265 seconds]
naoki has joined #linux-rockchip
naoki has quit [Client Quit]
_whitelogger has joined #linux-rockchip
vagrantc has quit [Quit: leaving]
warpme has joined #linux-rockchip
jakllsch has quit [Ping timeout: 276 seconds]
jakllsch has joined #linux-rockchip
warpme has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
warpme has joined #linux-rockchip
franoosh has joined #linux-rockchip
stikonas has joined #linux-rockchip
ldevulder has joined #linux-rockchip
franoosh has quit [Ping timeout: 244 seconds]
franoosh has joined #linux-rockchip
stikonas has quit [Remote host closed the connection]
raster has joined #linux-rockchip
raster has quit [Remote host closed the connection]
raster has joined #linux-rockchip
<chewitt> is anyone aware of emmc module issues on Rock 5B?
<chewitt> I've formatted/partitioned the module under macOS using a USB adapter, then connected it to the board, and powered on
<chewitt> I'd have expected it to be automagically detected and mounted .. but
<chewitt> ROCK5B:~ # dmesg | grep mmc0
<chewitt> [ 0.442928] mmc0: SDHCI controller on fe2e0000.mmc [fe2e0000.mmc] using ADMA
<chewitt> [ 0.475192] mmc0: Failed to initialize a non-removable card
<chewitt> I'm also noticing that Rock 5B appears to have the SDIO (mmc2) node enabled when the board has no hardware
<chewitt> looks like 5b+ has an onboard wireless module (not sure if SDIO or USB) but 5b uses the m2 slot so it's a pcie device
<chewitt> the emmc module was shipped by Radxa so I'd have expected it to work OOB :)
<chewitt> that said, I don't see much evidence of upstream kernel use in their forums and this is 6.16-rc6 so perhaps that's a factor?
warpme_ has joined #linux-rockchip
<diederik> chewitt: I'm not actively using mine (yet; still on SDcard), but in a 6.15-rc3 kernel log I have, I see it is detected
<chewitt> SD card works fine
<CounterPillow> check the schematic for whether the onboard SPI flash and the eMMC use the same data lines
<diederik> "mmc0: new HS400 Enhanced strobe MMC card at address 0001" and "mmcblk0: mmc0:0001 CUTB42 58.2 GiB" hint that eMMC should also work
<diederik> same messages on 6.16-rc5
<robmur01> "so it's a pcie device" - nope, M.2 E has all sorts on it: PCIe, USB, SDIO, I2S, UART, random GPIOs...
warpme_ has quit [Quit: My MacBook Air has gone to sleep. ZZZzzz…]
<diederik> Wow, the ASoC spam is soo massive, my dmesg lost its first msgs :-O https://paste.sr.ht/~diederik/a104bb1794cce8042ada701d801aeba99dce892a
<diederik> crap, wrong title
<chewitt> robmur01 I stand corrected then .. but wifi/bt both probe and work (although wifi appears to be tempramental)
<chewitt> (can't spell today either..)
<chewitt> the module radxa sent me is 128GB according to the label on the packet .. not sure if that might make it more exotic than smaller capacities
<chewitt> I note that https://docs.radxa.com/en/accessories/emmc_module says it's a Toshiba part vs. the Foresee one that @diederik is prob. using
<diederik> Turns out this card is made by Samsung (KLMCG2UCTB)
<diederik> (but I do also have Foresee eMMC cards)
<diederik> chewitt: That page only tells which cards they tested with, which says ~ nothing about what you actually have
raster has quit [Remote host closed the connection]
<chewitt> true .. I'm trying to read the chip via phone camera (as my eyeballs aren't the best)
<diederik> IOW look at your actual eMMC module to find out what you have
<diederik> that's exactly what I did too ;-P
<robmur01> chewitt: sure, just saying it's not wrong for the board to have SDIO enabled, since it's still wired to the socket even if you do happen to be using a non-SDIO module
raster has joined #linux-rockchip
<chewitt> emmc issue resolved :)
<chewitt> "user error, please replace user"
<chewitt> turns out I connected the ufs module for Rock 4D? instead of the emmc module
<chewitt> too many little packages and boards combined with enthusiasm and bad eyesight
<diederik> This was bound to happen: same size and same connectors
<diederik> I do have a 'special' eMMC adapter and I wonder if I'd need another one if I will use an UFS module ...
<diederik> ... probably. I have one from PINE64 which has 1 connector and https://radxa.com/products/accessories/emmc-ufs-module-reader shows 2 connectors, 1 for eMMC and 1 for UFS
<chewitt> yup, the adapter appears to support both
ungeskriptet has quit [Ping timeout: 260 seconds]
<chewitt> test distro image now installed to eMMC module
ldevulder has quit [Ping timeout: 252 seconds]
warpme has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
ungeskriptet has joined #linux-rockchip
warpme has joined #linux-rockchip
ungeskriptet has quit [Ping timeout: 252 seconds]
ldevulder has joined #linux-rockchip
ldevulder has quit [Read error: Connection reset by peer]
ldevulder has joined #linux-rockchip
warpme has quit [Quit: Textual IRC Client: www.textualapp.com]
jakllsch has quit [Ping timeout: 276 seconds]
jakllsch has joined #linux-rockchip
<detlevc> diederik: The ASoC messages is kinda my fault, I don't have a solution for that yet. The QP bridge won't let some audio related registers be written when no HDMI cable is connected (it will just hang). So if there is nothing connected, I just return -ENODEV. But ASoC doesn't like that. I think my solution will be to just return 0 but do nothing.
<detlevc> diederik: Another idea was to use a cached regmap and write the registers one a cable is plugged in, but that doesn't work as most values are calculated based on the TMDS char rate
<detlevc> which is unavailable if no screen is plugged in
<diederik> detlevc: I may indeed not have plugged in the HDMI cable that time; I booted it up to check if 6.16-rc5 would have a problem with eMMC (see earlier discussion)
<detlevc> Also, there seem to be no way to mark the sound device as disconnected (so that it doesn't show in alsa or pipewire)
<diederik> I dunno where it came from, but all (?) my devices have various ASoC errors now, although they're not all the same. Given that I also got it on a 6.12.21 kernel, my guess is that it's some userspace thing that messes things up?
<diederik> I did a (quick) downgrade of pipewire, but that didn't make a change.
<diederik> Usually I have no idea where ASoC errors come from ... and after some time they seem to disappear ... once again without me knowing/understanding why
* diederik starts up Rock5B again, this time with HDMI connected ...
<detlevc> These won't disappear unless I fix them in the driver. Also on rock 5b, we now have 2 hdmi outputs so you need to plug 2 screens to not see any error
<diederik> Ah, that explains why I'm still seeing the same ASoC error with 1 HDMI connected, but now 'preseeded' by "HDMI: Unknown ELD version 0" error ... and less spammy as dmesg now still had the first boot messages
<mmind00> detlevc: might be a dumb question, but did you compare clock-state between cable plugged and not? Controller hanging for some register write sounds like some clock thingy
warpme_ has joined #linux-rockchip
<detlevc> mmind00: Yes, that's also a solution, but it feels wrong to enable clocks on hdmi to configure something that won't work (there will also still be no TMDS char rate, so configured values will still be wrong)
<diederik> detlevc: AFAIC there's no need to prioritize this (for me). For the foreseeable future I'll likely only use my 5B to build (kernel) packages (ie headless)
<diederik> I do want to try to daily drive it during the Forky cycle though to see how that fares :)
<mmind00> detlevc: was more thinking about, asoc wants some registers written -> enable pclk -> do the write -> disable pclk
<detlevc> mmind00: I can try. That might remove the "HDMI: Unknown ELD version 0" though, I'll give it a go
<chewitt> detlevc I'm experimenting with Kodi support on 3588 for LibreELEC
b0 has quit [Quit: adiós]
<chewitt> which results in this https://imgur.com/a/YgR3Vcm
<detlevc> chewitt v2 is wip and will not work with hevc
<chewitt> it looks like the branch is the same except for the extra wip commit in v2?
b0 has joined #linux-rockchip
<detlevc> yes
<detlevc> chewitt kodi uses ffmpeg to access the hardware ?
<chewitt> thanks :)
<chewitt> yes, kodi is a big fancy wrapper around libavcodec
<chewitt> I spotted your ffmpeg branch too
<detlevc> that's not ready yet either
<chewitt> I noticed that too .. it failed to compile :)
<detlevc> upstream needs changes in the v4l2 api changes, so I need to update that first
<detlevc> chewitt it didn't compile ? Did you update the linux headers with what is found in the add-vdpu381-and-383-to-rkvdec ?
<chewitt> no, I didn't dig very far at the moment
<chewitt> I also (mostly) try to stay within the limits of things I understand
<detlevc> chewitt what is the pixel format of that file video ?
<chewitt> the whole drm vs mesa vs ffmpeg vs kodi pipeline is definitely something I only partly understand :)
<detlevc> cause it doesn't look like a ref frame issue
<chewitt> i'm fairly confident that issue is related to afbc and formats/modifiers
<chewitt> as I've seen something that looks the same when poking Amlogic vdec support a few months ago
<detlevc> The driver doesn't support AFBC
<chewitt> oddly the Amlogic driver does, but the ffmpeg part (v4l2_m2m not v4l2-request) doesn't .. so you get a similar result
warpme_ has quit [Quit: Textual IRC Client: www.textualapp.com]
<chewitt> I dropped the wip commits on the kernel side but the result is the same
dsimic has quit [Ping timeout: 252 seconds]
dsimic has joined #linux-rockchip
<detlevc> chewitt can you send me your test file ?
<chewitt> sure
<chewitt> link shared via pm
<chewitt> it's just 10-bit media that has the issue
<chewitt> digging around I have some 8-bit hevc and that's playing fine
<chewitt> for reference, here's my kernel branch https://github.com/chewitt/linux/commits/rockchip-6.16.y
<chewitt> Kodi runs under GBM, not under Wayland, which I think hides some of the issues I'm seeing
franoosh has quit [Remote host closed the connection]
helene_ has joined #linux-rockchip
helene_ is now known as helene
helene is now known as Guest2293
Guest2293 has quit [Remote host closed the connection]
smaeul_ has joined #linux-rockchip
smaeul has quit [Ping timeout: 260 seconds]
smaeul_ is now known as smaeul
chewitt has quit [Quit: Zzz..]
stikonas has joined #linux-rockchip
ldevulder has quit [Ping timeout: 252 seconds]
ungeskriptet has joined #linux-rockchip
ungeskriptet has quit [Ping timeout: 245 seconds]
ungeskriptet has joined #linux-rockchip
raster has quit [Quit: Gettin' stinky!]
linkmauve has left #linux-rockchip [Error from remote client]
linkmauve has joined #linux-rockchip
franoosh has joined #linux-rockchip
franoosh has quit [Remote host closed the connection]
System_Error has quit [Ping timeout: 244 seconds]
System_Error has joined #linux-rockchip
ungeskriptet has quit [Ping timeout: 276 seconds]
ungeskriptet has joined #linux-rockchip
Danct12 has quit [Ping timeout: 276 seconds]
Danct12 has joined #linux-rockchip