ChanServ changed the topic of #linux-rockchip to: Rockchip development discussion | IRC log http://irclog.whitequark.org/linux-rockchip | Community GH https://github.com/linux-rockchip | Rockchip GH https://github.com/rockchip-linux | ML https://groups.google.com/group/linux-rockchip
<micken> :)
* micken is dumping hdmi
<micken> from linux
<micken> so far no diff
<micken> I am using busybox devmem, but that only do one address at a time
<micken> is there a tool that can dump a range of iomem?
<anarsoul> modify devmem for your purposes
<micken> true
<anarsoul> I did it once but I don't think that I kept the patch
field^Mop has quit [Ping timeout: 240 seconds]
lopsided98 has quit [Ping timeout: 240 seconds]
vagrantc has joined #linux-rockchip
lopsided98 has joined #linux-rockchip
vagrantc has quit [Quit: leaving]
lkcl has joined #linux-rockchip
inode has quit [Quit: ]
lkcl has quit [Ping timeout: 268 seconds]
marcodiego has joined #linux-rockchip
lkcl has joined #linux-rockchip
lkcl has quit [Ping timeout: 240 seconds]
vstehle has quit [Ping timeout: 265 seconds]
lkcl has joined #linux-rockchip
stikonas has quit [Remote host closed the connection]
<anarsoul> micken: didn't help :(
<anarsoul> so rk808 configuration is not related
<anarsoul> darn
<anarsoul> at this point I'm not sure what else to check...
lkcl has quit [Ping timeout: 246 seconds]
_whitelogger has joined #linux-rockchip
warpme_ has quit [Quit: Connection closed for inactivity]
cristian__c has joined #linux-rockchip
cristian_c has quit [Ping timeout: 240 seconds]
_whitelogger has joined #linux-rockchip
cristian_c has joined #linux-rockchip
cristian__c has quit [Ping timeout: 240 seconds]
vstehle has joined #linux-rockchip
chewitt has quit [Quit: Zzz..]
chewitt has joined #linux-rockchip
chewitt has quit [Quit: Adios!]
marcodiego has quit [Quit: Leaving]
<anarsoul> OK, I found how to fix it for me
<anarsoul> mw 0xff320180 01000100
<anarsoul> that's pmu_io_domains pmu1830-supply = <&vcc_3v0>;
<anarsoul> and in linux rockchip-io-domain driver does it
<anarsoul> u-boot doesn't have one
_whitelogger has joined #linux-rockchip
cristian__c has joined #linux-rockchip
cristian_c has quit [Ping timeout: 246 seconds]
<micken> anarsoul: cool
cristian_c has joined #linux-rockchip
cristian__c has quit [Ping timeout: 245 seconds]
cristian_c has quit [Ping timeout: 268 seconds]
cristian_c has joined #linux-rockchip
inode has joined #linux-rockchip
<micken> anarsoul: excelent
Kwiboo has quit [Quit: .]
Kwiboo has joined #linux-rockchip
<micken> anarsoul: I *think* I have the same problem, seems like running linux first fix my problem
<micken> anarsoul: I tried before to sort out which pmic reg that is connected to vop and or hdmi phy, but never figured it out
adjtm_ has quit [Quit: Leaving]
<micken> anarsoul: bah reading power tree gives me nada
_whitelogger has joined #linux-rockchip
stikonas has joined #linux-rockchip
<micken> anarsoul: 0x10 and forward shows same after reset and from power off
stikonas has quit [Remote host closed the connection]
<micken> I believ that hdmi phy reg is correct, otherwise I shouldnt be able to use it at all
<micken> but vopbig might be wrong
<micken> should be pf_vo
<micken> pd_vo
<micken> and pd_hdcp for hdmi
rkos has quit [Quit: Lost terminal]
ayaka has quit [Ping timeout: 265 seconds]
vicencb has joined #linux-rockchip
ayaka has joined #linux-rockchip
ayaka has quit [Ping timeout: 246 seconds]
ayaka has joined #linux-rockchip
vicencb has quit [Quit: Leaving.]
vicencb has joined #linux-rockchip
ayaka has quit [Ping timeout: 276 seconds]
ayaka has joined #linux-rockchip
field^Mop has joined #linux-rockchip
leah2 has quit [Ping timeout: 250 seconds]
leah2 has joined #linux-rockchip
aalm has joined #linux-rockchip
stikonas has joined #linux-rockchip
<micken> sigh
ayaka has quit [Ping timeout: 265 seconds]
ayaka has joined #linux-rockchip
nashpa has quit [Ping timeout: 246 seconds]
nashpa has joined #linux-rockchip
vicencb has quit [Quit: Leaving.]
jailbox has quit [Ping timeout: 268 seconds]
ldevulder__ has joined #linux-rockchip
ldevulder_ has quit [Ping timeout: 240 seconds]
nashpa has quit [Ping timeout: 268 seconds]
nashpa has joined #linux-rockchip
<anarsoul> mmind00: I'm not sure if anyone reported that but rockchip-iommu driver throws a WARN "WARNING: CPU: 5 PID: 1 at drivers/iommu/rockchip-iommu.c:529 rk_iommu_irq+0x140/0x348" on shutdown/reboot
<anarsoul> also it often oopses, so board can't be reloaded or powered off
<anarsoul> that's on 5.3, I'll test 5.4-rc later today
<mmind00> anarsoul: not that I know of ... there are sometimes issues with the iommu driver, but I can't at least remember something from the 5.4 cycle - at least on my boards
<anarsoul> I hope it's fixed in 5.4
<anarsoul> :)
<anarsoul> btw, I assume that rockchip-drm hanging system when built-in is a known issue? it works fine as a module
<stikonas> anarsoul: I wouldn't assume bugs as known issues...
<mmind00> anarsoul: I guess not widely known ... at least new for me
<anarsoul> I saw some commits that are supposed to fix it but looks like they're not
<anarsoul> *not fixing it
<anarsoul> stikonas: btw, here's the patch to fix usb on rockpro64: https://lists.denx.de/pipermail/u-boot/2019-November/389815.html
<stikonas> anarsoul: oh thanks, I can try to test it too
<stikonas> although, I'm not currently subscribed to that list, so can't reply to your mail
<mmind00> at least the iommu warning comes from the pm_runtime_get_if_in_use ... it returns an error when runtime_pm is disabled ... but so far I fail to see any pm_runtime_disable in the iommu driver
<mmind00> hmm, ok pm_runtime_force_suspend() does disable runtime-pm, but then it should've freed the irq before that already ... maybe a weired race condition with the irq routine?
<anarsoul> mmind00: I'm not familiar with this code, sorry
<micken> while you are here, which pmic regs is for vop/hdmi. I get vd_vo and vd_hdcp , but I don't know where to find them and if those are the *only* regs involved. HDMI phy works , which suggests that it gets power. Image output works good 30% of bootups, but get blurry the rest. If first booting linux, then reset and then riscos th image is perfect..
<anarsoul> so yeah, looks like driver didn't remove irq
<anarsoul> and it fired
<anarsoul> :)
<anarsoul> micken: hdmi is digital, so I don't get how you can get blurry image
<anarsoul> maybe you have some post processing occasionally enabled?
<anarsoul> e.g. deinterlacer or something like that (not sure if rk3399 has it though)
<micken> anarsoul: it is like pixel get doubles
<micken> but not extending length
<micken> it seems random :/
<micken> but always works after linux
<anarsoul> maybe there's some scaler? :)
<micken> anarsoul: so I got the idea of pmic from our recent conversations
<anarsoul> I doubt it's pmic. With hdmi you either get a picture or not
<micken> anarsoul: yea but why kick in it is random
<anarsoul> micken: likely you leave it uniniailized?
<micken> ram timings? , but then most other stuff would go wrong
<micken> basicly I do the frambuffer from : either a os allocated ram area or a fixed address and write that to vop_big yrgb win0
ayaka has quit [Ping timeout: 252 seconds]
<micken> both ways fail the same way, even if the framebuffer just is a bitmap image
ayaka has joined #linux-rockchip
<stikonas> anarsoul: am I missing osmething, or did you forget to git add some files in your patch?
<anarsoul> hm
<anarsoul> let me check it
<stikonas> configs/rockpro64_rk3399.h is not in
<anarsoul> yeah
<anarsoul> sorry
<anarsoul> will resend
<stikonas> and also rebase on top of master maybe...
<stikonas> if you send it
<stikonas> as it has some conflicts
<stikonas> trivial conflicts but still...
<anarsoul> give me a minute, I'll rebase it
<stikonas> ok, thanks
<stikonas> trying
<micken> hard to capture on photo
<anarsoul> it's not blur
<micken> you haven't seen textoutput
<anarsoul> it's like some pixel columns are misplaced
<anarsoul> please show text
<anarsoul> I wonder if it's some kind of pixel format
<anarsoul> tiled?
<anarsoul> interesting
<anarsoul> definitely looks like some pixels are misplaced
<anarsoul> is there any kind of lookup table for reordering pixels?
<anarsoul> stikonas: did it work? :)
<stikonas> anarsoul: no...
<stikonas> just tried
<stikonas> still can't detect my USB hard drive
<anarsoul> are you sure you flashed it?
<stikonas> yes, it reports today's date
<stikonas> in version field
<anarsoul> make sure you ran 'make rockpro64_rk3399_defconfig'
<stikonas> well, I used menuconfig to pick the board, maybe I need to run defconfig...
<anarsoul> (or check what's your target in .config)
<micken> anarsoul: no lut. (but there is a lut setting in win0, dunno what it do) and it works sometimes. The image I linked is just memcopied out to the yrgb framebuffer
<anarsoul> stikonas: you should have "CONFIG_TARGET_ROCKPRO64_RK3399=y" set
<stikonas> yes
<stikonas> I have it set
<anarsoul> hmm
<stikonas> maybe it helps up to certain power level?
<stikonas> hmm
<anarsoul> maybe?
<anarsoul> it works pretty reliably for me now
<stikonas> it still works with my usb flash...
<stikonas> anarsoul: my external SSD also doesn't work
<stikonas> only USB sticks work
<stikonas> it doesn't make it worse though
<micken> just rebooted
<micken> and suddenly text that I can read without guessing :D
<anarsoul> stikonas: :(
<stikonas> anarsoul: usb3 devices?
<anarsoul> no
<stikonas> hmm, maybe related to that?
<anarsoul> my stick is usb 2.0
<stikonas> my stick is also usb 3 though...
<anarsoul> I can test with usb 3.0 hard drive
<anarsoul> give me a moment
<anarsoul> stikonas: my drive doesn't get enough power to start in USB2 ports, but works fine in USB3
<stikonas> hmm
<stikonas> let me run defconfig just in case...
<anarsoul> likely won't help :)
<anarsoul> btw, my drive doesn't work in usb2 even in linux
<anarsoul> so it's not u-boot issue
<stikonas> well, yeah, usb3 needs more power
<stikonas> possibly would work with powered usb2 hub...
<stikonas> but anyway, not u-boot
<stikonas> no, still no workee
<micken> mem in vop
<micken> ok. so new findings... running linux before is no guaranty that it will work.. if just press reset and not powercycle riscos I end up with about 90% success rate (compared to ~30 with powercycle)
<micken> i wish there was a consistant error
<anarsoul> :(
<anarsoul> you need some expert in rk3399 display hardware
<micken> within a week or so I can share some source
<micken> most of it is assembler
<anarsoul> ouch
vicencb has joined #linux-rockchip
BenG83 has joined #linux-rockchip
vagrantc has joined #linux-rockchip
adjtm has joined #linux-rockchip
stikonas has quit [Remote host closed the connection]
stikonas has joined #linux-rockchip
stikonas has quit [Remote host closed the connection]
stikonas has joined #linux-rockchip
<anarsoul> stikonas: do you also have a lot of issues with emmc on rockpro64?
<anarsoul> it boots off emmc 1 out of 4 times
field^Mop has quit [Ping timeout: 268 seconds]
<stikonas> anarsoul: I had, but I swapped it with another emmc and they are gone
<stikonas> I had something like that with emmc shipped by pine64 store
<anarsoul> :\
<stikonas> but the one I had from odroid (my odroid-u2 recently died) works fine
<anarsoul> I have some other emmcs but they're all from pine64 store :)
<stikonas> I tried patches from 5.4 but didn't help...
<anarsoul> yeah
<stikonas> anarsoul: I had some kernel patch that helped when Linux kernel booted
<stikonas> but nothing for u-boot...
<anarsoul> what patch is it?
<stikonas> it's just a quick hack... non-upstreamable
<anarsoul> I wonder what is MMC_CAP2_CQE
<stikonas> I'm not using this patch myself at the moment...
<stikonas> command queues?
<anarsoul> it actually detects the card
<anarsoul> but fails to select hs200 mode
<anarsoul> stikonas: can you show 'dmesg | grep mmc1' with your new emmc?
<anarsoul> (my guess is that it doesn't support high-speed modes and thus it doesn't fail)
<stikonas> anarsoul: sure, one moment
lkcl has joined #linux-rockchip
<anarsoul> nope, it actually runs at hs200
<anarsoul> :)
<stikonas> yeah... clearly stated in dmestg