ChanServ changed the topic of #linux-sunxi to: Allwinner/sunxi development - Did you try looking at our wiki? - Don't ask to ask. Just ask and wait for an answer! - This channel is logged at
vagrantc has quit [Quit: leaving]
ChanServ changed the topic of #linux-sunxi to: Allwinner/sunxi development - Did you try looking at our wiki? - Don't ask to ask. Just ask and wait for an answer! - This channel is logged at
cnxsoft has joined #linux-sunxi
pabs has quit [Remote host closed the connection]
pabs has joined #linux-sunxi
grming has quit [Quit: Konversation terminated!]
junari has joined #linux-sunxi
<MoeIcenowy> anarsoul: software developers are of international price
<MoeIcenowy> or slightly less
<MoeIcenowy> Xogium: what do you use to power up the V3LP DRAM?
<Xogium> I don't actually know, I didn't get them because I don't know what the blobs sochip sent were for
<Xogium> didn't want to risk it yet
JohnDoe_71Rus has joined #linux-sunxi
vagrantc has joined #linux-sunxi
<MoeIcenowy> what's your board?
<MoeIcenowy> some V3s board uses dedicated DCDC, which needs to have a few resistors replaced for a different DRAM voltage
<MoeIcenowy> some use AXP20x, which can be adjusted via I2C
<Xogium> was based on lichee zero
<Xogium> but don't know much, since I only did system bring up
codekipper has joined #linux-sunxi
JohnDoe_71Rus has quit []
JohnDoe_71Rus has joined #linux-sunxi
codekipp2r has joined #linux-sunxi
codekipper has quit [Ping timeout: 480 seconds]
vagrantc has quit [Quit: leaving]
junari_ has joined #linux-sunxi
junari has quit [Ping timeout: 480 seconds]
sh1 has quit [Read error: Connection reset by peer]
gsz has joined #linux-sunxi
<Xogium> hmmm the h616 wiki page is a bit vague
<Xogium> saying basic support was added in 6.0, but just how basic are we talking about here ?
junari__ has joined #linux-sunxi
junari_ has quit [Ping timeout: 480 seconds]
enick_7 has quit []
enick_7 has joined #linux-sunxi
enick_7 has quit []
DasChaos has joined #linux-sunxi
<DasChaos> Hi people o/
<gamiee> Welcome back :)
<DasChaos> Yeah had some issues with the matrix <-> irc bridge. It seems like you can only change nick if you are disconnected. But the command doesn't say that :/
<DasChaos> I'm currently trying to get the running on my mango pi T113 devboard. I compiled it successfully, but when trying to boot from it, it doesn't work. I tried executing it with xfel and I also tried it via SD Card. In both cases the blue led lights up, but I can't see any logs on the serial with 115200 bauds. Any information how to further debug the issue?
<buZz> are you sure you're using the right serial port? there might be >1 uart
<buZz> never heard of awboot before :)
<Xogium> same here
<DasChaos> <buZz> "are you sure you're using the..." <- yes I tried multiple serial boot
<buZz> what?
<DasChaos> For the T113 u-boot isn't ported yet.
<buZz> whats a 'multiple serial boot' ?
<DasChaos> buZz: sorry typo, i wanted to say "I tried multiple serial ports"
<buZz> oh, right, ok
<buZz> well, i mean on the SBC
<DasChaos> yes
<DasChaos> Its also for sure working, because a precompiled image exists and with that I get a output on serial
<buZz> ah ok, did you enable the serial logging on your build? :P
<buZz> it seems to use 'uart 5'
<buZz> whatever that might be
<DasChaos> Where did you find it?
<buZz> also lists the pin numbers for that _dbg uart
<buZz> :)
hramrach has quit [Ping timeout: 480 seconds]
bauen1_ has quit [Ping timeout: 480 seconds]
JohnDoe_71Rus has quit []
junari_ has joined #linux-sunxi
bauen1 has joined #linux-sunxi
junari__ has quit [Ping timeout: 480 seconds]
JohnDoe_71Rus has joined #linux-sunxi
cnxsoft has quit []
dliviu has quit [Ping timeout: 480 seconds]
junari_ has quit [Remote host closed the connection]
dliviu has joined #linux-sunxi
grming has joined #linux-sunxi
<mnemoc> has anyone heard of libv? I was hoping to do the final swap of wiki servers before xmas
ftg has joined #linux-sunxi
vagrantc has joined #linux-sunxi
MrMichael has joined #linux-sunxi
MrMichael has quit [Remote host closed the connection]
bauen1 has quit [Ping timeout: 480 seconds]
JohnDoe_71Rus has quit []
bauen1 has joined #linux-sunxi
apritzel has joined #linux-sunxi
<apritzel> Xogium: from all we know the V3LP should be a drop in replacement, minus 1.5V for VccDRAM instead of 1.8V
gsz has quit [Ping timeout: 480 seconds]
<Xogium> apritzel: right, that was also my understanding, but if that's the case why did sochip give me a libram and a bunch of binary blobs ?
<Xogium> this is headache inducing
<apritzel> The DRAM timing parameters could be different, though, but it should still be DDR2 (just the DDR2L variant)
<apritzel> So our H3 DRAM driver in U-Boot should be good, and that is pretty easy on the parameters
<apritzel> Because nobody (including AW) seems to have a real clue about DRAM, they all pass on code
<Xogium> apritzel: could I send you the files, do you think ? Hmm. I don't want any license trouble or etc, or to bother you if you don't want to check this out, that said
<apritzel> Do you have a boot0 binary? They have timing parameters in the first 300 bytes
<Xogium> let me check real quick what I have
<Xogium> yeah I have this, for 3 sort of storage ? Sdcard, spino and... I'm not quite sure what the last one is for. Also a libdram folder with a libdram blob with no file extension
<Xogium> *spinor, even
<apritzel> It seems like AW works out some DRAM timings for each board vendor,and they just use those value
<Xogium> charming
<Xogium> so those are what, u-boot blobs ?
<apritzel> Sdcard or SPI shouldn't matter, just check the hexdump at 0x38
<apritzel> Boot0 is Allwinners SPL replacement
<Xogium> ah, I see. So proprietary bootloader, sort of
<apritzel> Only the first part, doing Dram plus clk setup
<apritzel> Plus loading of other components like proper Uboot
<Xogium> right
<Xogium> let me check for 0x38
<apritzel> Should have the clock freq in MHZ
<apritzel> Next word is DRAM type, likely just 0x00000002
<apritzel> Then comes ZQ, I think, like 0x3f3b3b
<apritzel> I have a decoder for that, just not here
<Xogium> what hexdump command should I use exactly ?
<apritzel> Is your libdram an ELF? (ask "file")
<Xogium> aye, it is. Very nice, not stripped and with debug_info
<apritzel> hexdump -C boot0xxxx.bin | head -20
<apritzel> Good, but still tedious to decipher
<apritzel> And we know the DRAM controller already, so not really helpful
<Xogium> 00000380 d0 f8 00 c0 02 91 24 d5 83 6a c3 f3 42 52 03 92 |......$..j..BR..|
<Xogium> hmm
<apritzel> But keep it around, just in case they changed more than just the DRAM chips
<apritzel> Look at 00000030
<apritzel> 0x38 is in the middle
<Xogium> 00000300 1f 00 c0 e3 13 00 80 e3 c0 00 80 e3 02 0c c0 e3 |................|
<Xogium> sorry its a bit akward ;p never delt with those blobs before
<Xogium> *awkward even
<apritzel> Can you give 00000030, not 300?
<Xogium> oh sorry
<Xogium> my tts said 30 :/
<Xogium> 00000030 seems to be just a bunche of 00
<apritzel> Is that from the boot0 sdcard blob? We need that, not libdram
<Xogium> yeah boot0 sdcard blob
<apritzel> Does ist say eGON in the first line?
<Xogium> |....eGON.BT03.[.| to be exact
<apritzel> Yeah ok, thats the one
<apritzel> Weird that its empty at 30...
<Xogium> I'll check the spinor but I expect the same
<Xogium> yeah same thing there
<apritzel> Is there some board involved? Or is this just about the SoC?
<Xogium> just the SoC. I want to be sure that the v3lp won't be a problem since the v3s is being phased out
<Xogium> and having received those blobs with absolutely no explanation other than, it's a 'patch' from sochip, I got concerned
<apritzel> Board vendors seem to get some tailored boot0 from AW, for their Dram setup
<apritzel> Try to talk to Peter (check the logs from Sunday) he should know more
<Xogium> I suppose the SoC itself is actually more of a SiP than a SoC, given it has integrated dram just like the v3s did, so... Not quite a board, but still valid I suppose
<apritzel> Yeah, good point
<apritzel> The boot0 should be board independent then
<Xogium> yeah I've seen those logs, but he didn't appear too concerned by the blobs. Still, I was, because if I get some random blobs from a vendor my first thought is, oh dear god
<Xogium> :D
<apritzel> Thats common for AW, though
<Xogium> yeah I'm huh
<Xogium> more experienced with st
<apritzel> People reveng'ed the whole DRAM
<apritzel> Based on livdra
<Xogium> makes sense
<apritzel> *libdram disassembly
<apritzel> (Typing on a tablet ... :-(
<Xogium> I figure the sunxi folks might have done that, but I was concerned that maybe, just maybe those blobs meant a new kernel driver, or u-boot driver, or...
<apritzel> I doubt that
<Xogium> who knows what really
<apritzel> Its probably the same die still
<Xogium> yeah, hopefully
<apritzel> So same DRAM *controller*, but different DRAM *chips*
<Xogium> yeah that makes sense
<Xogium> well, thanks for explaining this to an AW newbie :D
<apritzel> Peter is quite far on a board and is using the V3LP
<Xogium> I'm usually fine with stm32, so yeah going from one extreme to the other here, aka super open vendor to extremely close and secretive one, that's quite destabilizing
<apritzel> I tend to ignore source from AW
<Xogium> seems he is
<apritzel> Maybe just peek in for info not in the manuals
<Xogium> yeah that does sound like a good plan :p
<Xogium> all it does it bring more confusion into the mix
<apritzel> High chances are that V3s compiled mainline UBoot just works
<apritzel> Kernel wise it should be identical
<apritzel> I guess they just couldnt source the DDR2 dies anymore
<Xogium> I was eyeing the h616 for another project, but given the current mainline support, I'm not sure about it. I mean the wiki doesn't say much, only that basic support has been done in 6.0 kernel, but just how basic ? Sort of vague
<Xogium> yep it was partly that, partly due to instabilities they found with the other chip
<apritzel> Check the mainlining effort page
<Xogium> ah, so that's where it's hidden
<apritzel> USB MMC GigabitEthernet works
<Xogium> :D
<Xogium> hmm, not that bad for basic support
<apritzel> USB just got merged last week
<Xogium> woah
<apritzel> No graphics merged,but working patches are floating around
<Xogium> wait so merged, but for 6.1 or whichever next release, not down to 6.0, correct ? Unless this sort of stuff is backported to stable
<apritzel> Integrated PHY is WIP
<apritzel> USB wil be in 6.2 only
<Xogium> righto
<apritzel> -rc1 due anytime soon now
<Xogium> I assume things like spi, i2c, uart, alredy there, though
<apritzel> Yep
<Xogium> wow, awesome bringup as always
<Xogium> :)
<apritzel> Low hanging fruits all picked
<Xogium> very nice
<apritzel> 6.0 contains everything minus USB
<Xogium> seems like I just need to wait a little bit more
<apritzel> What do you need?
<Xogium> usb gadgets for a couple things, actually
<apritzel> The USB patches are quite easy to backport
<Xogium> I hope the h616 isn't on the dumber side about mmc boot though ;) I get the impression it's on the higher end range compared to the v3, so there's that
<Xogium> I wouldn't have used the v3s if it had been up to me, but...
vagrantc has quit [Quit: leaving]
<Xogium> you know how that sort of things go
<apritzel> h616 is a different league compared to the V3s
<Xogium> yeah it seems so, it could make a decent android phone :D
<apritzel> In 2016: maybe
<Xogium> ^^
<apritzel> Otherwise 4 A53 cores and max 4GB are out of fashion
<Xogium> yeah android is so ram hungry these days, crazy
<apritzel> Well,others can do much newer (and big) cores, and at least *support* more RAM easily
<apritzel> H616 is popular in cheap Android TV boxes
<apritzel> It has some decent video decoders, and intgerated PHY saves pennies and avoids external chips that are hard to get these days
<Xogium> makes you wonder what they wanted to target with v3s, initially
<apritzel> Camera related devices, I think
<Xogium> oh
<Xogium> hmm, yes I can see it being used for that
<apritzel> Stuff like dash cams
<apritzel> Basically a DMA controller from cam logic to MMC
<apritzel> Plus some GUI
<Xogium> I can see why they'd use it for that, certainly it is cheap enough
<apritzel> Indeed
<apritzel> Inetegrated DRAM saves again
<Xogium> aye
<Xogium> but cheaper doesn't always mean good
<apritzel> Cheaper means cheaper ;-)
<Xogium> I got quite literally scared of cheap sbc and allwinner for a time by an orange pi zero that almost literally caught on fire on my desk while I had my back turned for about a minute
<Xogium> it melted its case... Wait, what case ? Oh yeah that sludge over there
<apritzel> H3 was a bit on the warm side, if used in anger
<Xogium> yeah... And so was h2+
<apritzel> Same thing
<Xogium> I was just letting it hang around in u-boot while I grabbed a cable to hook up to it, and it did that
<apritzel> Just a lower bin of H3, apperently
<Xogium> honest to god scared me
<apritzel> Thats weird though
<Xogium> I know right ?
<apritzel> UBoot is polling, so no idling, but it sits in UBoot for hours on my desk
<apritzel> Maybe the case was the problem?
<Xogium> plus with my disability, I never realized the board was overheating to a dangerous level and was quite nervous searching all over the room to find where that awfull smell came from
<Xogium> my hand landed straight down on the super heated board. I have a scar in the shape of the case to this day :O
<apritzel> Indiana Jones part 1 like?
<Xogium> that's a possibility, certainly it had no heatsink of any sort, and the case was not having a whole lot of airflow except for a few holes on the sides
<apritzel> Might have been a faulty board or chip though
<Xogium> yeah sorta like that :D I never thought of it that way
<apritzel> H3 in UBoot is typically fine, since not SMP nor Mali nor heavy traffic
<Xogium> so yep melted its case into a semi liquid plastic mess that was no fun to clean up either
<Xogium> oh and the fun part, the plastic around the micro sd card melted down and shut the slot, in a permanent way
<Xogium> of course the SoC itself got damaged from that and it won't even power up these days
<apritzel> Just thinking: H2+ bin probably also means worse thermal behaviour
<Xogium> but at least its the only time a board died for me so far :D
<Xogium> yup
<Xogium> most likely
<Xogium> it also had no, huh how's that called again, voltage-controlled cpufreq scaling
<Xogium> it always ran at the max voltage, all the time
<apritzel> Yeah, thats not uncommon for those cheap boards though
<Xogium> yeah... so I've learned
<Xogium> no pmic either, shame
<apritzel> You get what you pay for
<Xogium> indeed
<apritzel> PMIC is good for battery charging, plus supporting many external peripherals
<apritzel> Which you dont need on a small dev board
<Xogium> true, but it can also cut power if thermal go way too high
<apritzel> But it costs and needs pcb space
<Xogium> to be fair I think I've been too used to awesome stuff with stm32mp1 dev boards ^^
<apritzel> Yeah, you think for too reasonable and grown up for AW
<Xogium> I have high expectation for boards and yeah it's quite the rough landing when those aren't met because cheaper :p
<apritzel> Well, I think the OPi 0 sold for like ~12 USD ..
<Xogium> yep, used to, these days apparently its 18
<apritzel> Indeed
<Xogium> because h3 replaced h2+, which is probably not a bad thing, all things considered
<Xogium> compared to that the cheapest stm32 board I got was $50 :D
ftg has quit [Read error: Connection reset by peer]
apritzel has quit [Remote host closed the connection]
<Xogium> before I got to bed, can you tell me who works on the h616 ? Is it more like the entire community may work on the porting, or do you typically assign people to a SoC that they take over if they got the hardware to work on porting it ?
<Xogium> sorry for the probably newb question
apritzel has joined #linux-sunxi
apritzel has quit []
warpme_____ has quit []