<apritzel>
jemk: apparently the DC5SET pin isn't properly tied to GND on the Pine64, because the PMIC initializes DCDC5 with 1.24V instead of 1.5V
<apritzel>
jemk: (DCDC5 is the DRAM voltage)
wens has joined #linux-sunxi
<apritzel>
according to the data sheet DC5SET left floating resets DCDC5 to 1.24V
<apritzel>
and connecting it to GND results in a 1.5V reset, which is true for the BPi-M64
popolon has quit [Quit: WeeChat 1.4]
<ssvb>
apritzel: hmm, have you measured it?
<ssvb>
or checked the voltage level reported by the PMIC via software means?
<ssvb>
I also had some questions earlier about the FEL pin status on Pine64 because I could not find any pull-up resistor for it in the schematics document and leaving it floating is a very bad idea
<apritzel>
ssvb: no, not measured, just dumped some PMIC and clock registers after SPL and boot0 run for comparison
<ssvb>
is it a pre-production board?
<apritzel>
possibly
<apritzel>
I guess so, it reads 2015-12-17
<ssvb>
it would be best to verify this with some board from the final production batch
<apritzel>
yeah, don't have such a beast here atm (but should have one in the office)
<apritzel>
isn't a big deal, really, since we can set the voltage
<ssvb>
yes, sure
<apritzel>
boot0 does
<apritzel>
was more a heads up for jemk, since he was experimenting with the SPL DRAM code as well
<apritzel>
ssvb: FEL> indeed, the schematics just connects it to a test pin
<apritzel>
and despite it's not desperately needed on the Pine64, a FEL header pin or button would be nice to have, as it would ease automated control
<ssvb>
I had asked TL Lim at least twice about how to find this test pin on the board (admittedly mixed with a ton of other questions), but never got any reply
<ssvb>
because I was considering to maybe solder some button to it :-)
<apritzel>
yeah, having it on the BPi makes me miss it on the Pine
<ssvb>
regarding the DCDC5 voltage, so does it look like we actually need a basic AXP803 PMIC support in the SPL for A64?
<ssvb>
because the ATF takes control a bit too late
<apritzel>
ssvb: yeah, though I don't like the idea so much
<ssvb>
why not?
<apritzel>
because ATF could then be the only code controlling the AXP
<apritzel>
but I guess this would be fine if it's SPL only
<apritzel>
and looking at the U-Boot SPL code we probably just need to set some #define's
<ssvb>
yes, maybe
<apritzel>
oh, btw: are there any H3 boards with an external _100_ MBit PHY?
<apritzel>
(I was wondering if the Pine64 512MB is the only board around)
<apritzel>
well, it works with 1.24V ;-)
<ssvb>
I guess by pure luck, and maybe not on all boards
<apritzel>
at least U-Boot proper is loaded and run properly
<apritzel>
yeah, possibly
<ssvb>
my understanding is that DDR3L (1.35V) and DDR3 (1.5V) chips are manufactured on the same production line, but then binned and labeled based on what voltage they can tolerate
<ssvb>
and with modern yields, almost all the chips are in fact DDR3L compatible
<ssvb>
apritzel: still if I remember correctly, I tried to measure voltage at different test points and one of them was 1.5V, I could be wrong though
<apritzel>
but the universal image works by detecting the board, then selecting the right code, in contrast to one-code-to-rule-them-all, right?
<apritzel>
ssvb: well, that's just the reset value, boot0 sets up the AXP to 1.5V
<apritzel>
so it's really an SPL issue for now
<ssvb>
well, the idea with the universal image was that we are initially booting the system with a very minimal generic set of peripherals (basically only MMC, USB and HDMI), and then can make a guess about what is the actual board or ask the user
<apritzel>
so this would require DRAM setup or not?
<apritzel>
because this differs already by quite a bit between the AW SoCs, right?
<ssvb>
the SoC differences do not matter because we can detect them at runtime
<ssvb>
the DRAM voltage is a potentially problematic thing, because we need to rely on having a reasonable PMIC default configuration
<ssvb>
which is almost always the case, with maybe some rare exceptions such as the (pre-production?) Pine64 board
Andy-D has quit [Ping timeout: 264 seconds]
<ssvb>
another problematic this is USB power, because sometimes it needs to be enabled via GPIO
<ssvb>
s/this/thing
<wens>
given the pmic default should match the dram type, we should be able to move pmic init to main u-boot?
<ssvb>
wens: well, with the Pine64 board we need to have it in the SPL before the DRAM init
<apritzel>
also not sure if the DCDC1 3.0V reset value creates issues
<ssvb>
I would assume that if somebody has it as 3.0V by default, that's a very unsafe configuration
<apritzel>
lets hope that eMMC, SD cards and NOR flashes can cope with that
<apritzel>
does the AW stack actually sets it to 3.3V at some point?
<ssvb>
unless the DRAM chip can somehow survive getting powered by 3.0V, and only fry when somebody tries to initialize while running at this voltage
<apritzel>
DCDC1 is (hopefully) not connected to the DRAM
<apritzel>
its really for all the other peripherals and the GPIO headers
<ssvb>
still I would assume that having a 3.0V PMIC default as the DRAM power is a defective board design in any case
<ssvb>
apritzel: regarding 3.0V vs. 3.3 for GPIO, I think that 3.0V is still within acceptable limits
<apritzel>
yeah, but some custom extension circuits connected via the headers may choke on this
<apritzel>
if they take the voltage for instance
<apritzel>
but I agree that logic levels should be fine
<ssvb>
in fact 2.7V seems to be the minimal voltage to be considered a logical 1
<apritzel>
wens: I think the DC5SET pin is meant to cover exactly that: allow a board vendor to reset to the DRAM voltage that matches the soldered DRAM chips
<apritzel>
but as usual software has to come for the rescue when someone messes up ;-)
<apritzel>
oh dear, bed time over here ...
<apritzel>
tty l8er
ninolein_ has joined #linux-sunxi
apritzel has quit [Ping timeout: 244 seconds]
Andy-D has joined #linux-sunxi
ninolein has quit [Ping timeout: 264 seconds]
<wens>
ssvb: why is it needed?
cnxsoft has joined #linux-sunxi
<ssvb>
wens: DRAM needs to be properly powered before we try to initialize and use it
<ssvb>
and if the PMIC default is wrong (apritzel suspects that the Pine64 board uses 1.24V for DCDC5 instead of 1.5V after reset) then we need to adjust it
<wens>
crap
sarietta has quit [Remote host closed the connection]
sarietta has joined #linux-sunxi
sarietta has quit [Ping timeout: 264 seconds]
orly_owl has joined #linux-sunxi
_mamalala has joined #linux-sunxi
orly_owl has quit [Quit: leaving]
_mamalala_ has quit [Ping timeout: 272 seconds]
pg12 has quit [Ping timeout: 264 seconds]
pg12_ has joined #linux-sunxi
cnxsoft1 has joined #linux-sunxi
cnxsoft has quit [Ping timeout: 264 seconds]
cnxsoft1 is now known as cnxsoft
sarietta has joined #linux-sunxi
fire219 has quit [Read error: Connection reset by peer]
vagrantc has quit [Quit: leaving]
[7] has quit [Disconnected by services]
TheSeven has joined #linux-sunxi
Andy-D has quit [Ping timeout: 264 seconds]
maldata has left #linux-sunxi [#linux-sunxi]
netlynx has joined #linux-sunxi
netlynx has quit [Changing host]
netlynx has joined #linux-sunxi
|Jeroen| has joined #linux-sunxi
petr has quit [Ping timeout: 240 seconds]
petr has joined #linux-sunxi
|Jeroen| has quit [Quit: dada]
gzamboni has joined #linux-sunxi
reinforce has joined #linux-sunxi
arete74 has joined #linux-sunxi
kaspter has joined #linux-sunxi
premoboss has joined #linux-sunxi
imcsk8 has joined #linux-sunxi
premoboss has quit [Ping timeout: 264 seconds]
* Wizzup
wonders if it's hard to kexec to u-boot from Linux
premoboss has joined #linux-sunxi
kaspter has quit [Ping timeout: 264 seconds]
cnxsoft has quit [Ping timeout: 244 seconds]
<vpeter>
Isn't kexec meant for booting directly another kernel without bootloader?
IgorPec118 has joined #linux-sunxi
<Wizzup>
I think it's meant to halt the kernel and start executing a new piece of code
<Wizzup>
I am not sure if that has to be a linux kernel
ninolein_ has quit [Quit: No Ping reply in 180 seconds.]
premoboss has quit [Ping timeout: 272 seconds]
<KotCzarny>
ring ring ring?
<mdsrv>
banana phone
ninolein has joined #linux-sunxi
cnxsoft has joined #linux-sunxi
ricardocrudo has joined #linux-sunxi
IgorPec118 has quit [Ping timeout: 264 seconds]
sarietta has quit [Remote host closed the connection]
jstein_ has joined #linux-sunxi
jstein_ is now known as jstein
premoboss has joined #linux-sunxi
apritzel has joined #linux-sunxi
<apritzel>
vpeter: well, this bootloader scenario is one of kexec's use cases
<apritzel>
vpeter: petitboot is a bootloader using this technique
<jemk>
apritzel: you are right, i measure 1.24V at the dcdc5 inductor too. since pine has ddr3l chips that might be enough to not crash immediately
<apritzel>
jemk: is it right that you could run LPDDR3 at 1.5V as well?
<apritzel>
that you would just loose the power saving that they offer?
<apritzel>
so 1.5V would be a safe default in some generic code?
<jemk>
lpddr3 needs two voltages, for both 1.5v is outside the specification, but absolute maximum ratings allow 1.6v
<jemk>
low current 1.8v and high current 1.2V, so i guess dcdc5 would be responsible for 1.2v, 1.8v from some ldo
<apritzel>
jemk: yeah, ELDO1 seems to be the 1.8V source in this case
<apritzel>
the schematic labels it: VCC18-LPDDR
<apritzel>
and ELDO1 resets to 1.8V, so this is covered
<jemk>
apritzel: maybe they use 1.24v to do some type auto detection, first try to initialize as lpddr3, if that fails increase voltage and try to initialize as ddr3. the other way around could hurt the lpddr3, but ddr3 won't die because of 1.2v
uwe__ is now known as uwe_
<apritzel>
jemk: can you differentiate between LPDDR3 and DDR3 by looking at the registers?
<apritzel>
or by writing some and reading back / reading others
<apritzel>
?
<jemk>
apritzel: i don't know, but i'd guess initialization will be different, so using the wrong type will just fail
<jemk>
whether that is a good idea or could destroy something? i don't know, would need some research
<apritzel>
jemk: right, boot0 expects the type in one of its parameters
ricardocrudo has quit [Remote host closed the connection]
netlynx has quit [Quit: Ex-Chat]
utente_ has joined #linux-sunxi
fl__0 is now known as fl_0
ricardocrudo has joined #linux-sunxi
lemonzest has joined #linux-sunxi
sarietta has joined #linux-sunxi
sarietta has quit [Ping timeout: 244 seconds]
apritzel has quit [Ping timeout: 264 seconds]
cnxsoft has quit [Quit: cnxsoft]
cnxsoft has joined #linux-sunxi
avph has quit [Ping timeout: 272 seconds]
BenG83 has joined #linux-sunxi
ricardocrudo has quit [Ping timeout: 264 seconds]
ricardocrudo has joined #linux-sunxi
nemunaire has quit [Quit: quit]
nemunaire has joined #linux-sunxi
sarietta has joined #linux-sunxi
sarietta has quit [Ping timeout: 244 seconds]
iamfrankenstein1 has joined #linux-sunxi
iamfrankenstein has quit [Ping timeout: 244 seconds]
iamfrankenstein1 is now known as iamfrankenstein
iamfrankenstein has quit [Ping timeout: 244 seconds]
iamfrankenstein has joined #linux-sunxi
avph has joined #linux-sunxi
ricardocrudo has quit [Ping timeout: 264 seconds]
ricardocrudo has joined #linux-sunxi
<as0d70apf>
i receiver my orange pi 2 plus today but i'm wondering what's up with the otg port, do i need to do anything special? i kind of expect a /dev/ttyUSBx on my linux machines but nothing shows up even with lsusb
<as0d70apf>
received*
ricardocrudo has quit [Ping timeout: 272 seconds]
ricardocrudo has joined #linux-sunxi
<as0d70apf>
oh, i was looking at the wrong port, i need the uart one
apritzel has joined #linux-sunxi
al1o has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
vishnup has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
Andy-D has joined #linux-sunxi
apritzel has quit [Ping timeout: 264 seconds]
theone has quit [Ping timeout: 272 seconds]
fire219 has joined #linux-sunxi
theone has joined #linux-sunxi
sarietta has joined #linux-sunxi
sarietta has quit [Ping timeout: 244 seconds]
afaerber has joined #linux-sunxi
pg12_ has quit [Ping timeout: 272 seconds]
pg12 has joined #linux-sunxi
reinforce has quit [Quit: Leaving.]
afaerber has quit [Ping timeout: 272 seconds]
reinforce has joined #linux-sunxi
reinforce has quit [Client Quit]
afaerber has joined #linux-sunxi
cnxsoft has quit [Remote host closed the connection]
pg12 has quit [Ping timeout: 272 seconds]
pg12 has joined #linux-sunxi
ricardocrudo has quit [Ping timeout: 244 seconds]
pg12 has quit [Ping timeout: 272 seconds]
pg12 has joined #linux-sunxi
jstein__ has joined #linux-sunxi
jstein is now known as Guest82634
jstein__ is now known as jstein
Guest82634 has quit [Ping timeout: 272 seconds]
orly_owl has joined #linux-sunxi
ricardocrudo has joined #linux-sunxi
Andy-D has quit [Remote host closed the connection]
iamfrankenstein has quit [Ping timeout: 244 seconds]
iamfrankenstein1 has joined #linux-sunxi
lemonzest has quit [Quit: Leaving]
iamfrankenstein1 is now known as iamfrankenstein
Ganesh has joined #linux-sunxi
Ganesh is now known as Guest36832
apritzel has joined #linux-sunxi
solarnetone has joined #linux-sunxi
<ssvb>
apritzel: there isn't much diversity in the used DRAM types, almost everyone uses just DDR3/DDR3L rather than DDR2, LPDDR2 or LPDDR3
<ssvb>
still the generic board independent installer idea is mostly useful for using with devices which don't have any kind of non-removable storage, where some guesswork based on the DRAM size, SoC type and user input is required
<ssvb>
and that's something that we want to change in the mid- and long- term by encouraging every development board vendor to least add SPI flash when there is no eMMC :-)
Da_Coynul has joined #linux-sunxi
petr has quit [Ping timeout: 244 seconds]
<Da_Coynul>
having a problem with ethernet disconnecting on Orange Pi PC after updating my kernel to 4.7.2 with Montjoie's latest patches
petr has joined #linux-sunxi
<Da_Coynul>
strange thing is I have 2 OPI PC's, updated both at the same time and one has the problem but the other doesn't
<Da_Coynul>
no errors in dmesg
<Da_Coynul>
if I disconnect the ethernet cable and reconnect, it works for a while, then drops again.
<Da_Coynul>
tried with different power supplies and no change.
sarietta has joined #linux-sunxi
<Da_Coynul>
i think it has something to do with the PHY going idle because if I keep a ping going from a USB serial terminal, the connection stays up.
sarietta has quit [Ping timeout: 252 seconds]
<Da_Coynul>
i think my other OPI is not having the problem because darkice is streaming on it all the time so the PHY is always active.
iamfrankenstein1 has joined #linux-sunxi
scott_ has joined #linux-sunxi
scott_ is now known as Guest81087
solarnetone has quit [Ping timeout: 272 seconds]
premoboss has joined #linux-sunxi
iamfrankenstein1 has quit [Quit: iamfrankenstein1]
Putti has joined #linux-sunxi
premoboss has quit [Quit: Sto andando via]
apritzel has quit [Ping timeout: 264 seconds]
Putti has quit [Ping timeout: 248 seconds]
BenG83 has quit [Ping timeout: 244 seconds]
sarietta has joined #linux-sunxi
sarietta has quit [Ping timeout: 272 seconds]
jernej has joined #linux-sunxi
Guest81087 has quit [Quit: Leaving]
solarnetone has joined #linux-sunxi
solarnetone has quit [Client Quit]
Nacho__ has quit [Remote host closed the connection]
Nacho__ has joined #linux-sunxi
<KotCzarny>
so, is cedrus working on mainline/a20 ?
Guest36832 has quit [Ping timeout: 240 seconds]
<montjoie>
Da_Coynul: I will try to setup an idle opipc. could you paste ethtool -S eth0 ?
sarietta has joined #linux-sunxi
sarietta has quit [Ping timeout: 244 seconds]
<Da_Coynul>
montjoie: I have to step out for a while. will try again later (or tomorrow).
Da_Coynul has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
mossroy has quit [Quit: Leaving]
<zoobab>
I hit this error trying to compile uboot for pine64:
<apritzel>
zoobab: please don't tell me you are surprised ;-)
<zoobab>
I was expecting some DRM with "trusted"
<apritzel>
yeah, it's a marketing term, really
<zoobab>
going to zzz
<zoobab>
can you remember what I should change to add tftp support?
<zoobab>
some DT change?
<apritzel>
trusted refers to both TrustZone and to the possibility of having some secure firmware, not easily being hackable from Linux
<apritzel>
zoobab: add "ethernet0 = &emac;" to the aliases section of the DT
<apritzel>
or wait for my series to hit U-Boot ML hopefully later tonight
<zoobab>
will wait then
popolon has joined #linux-sunxi
sarietta has joined #linux-sunxi
orly_owl has joined #linux-sunxi
sarietta has quit [Ping timeout: 265 seconds]
<apritzel>
zoobab: for the records: ARM Trusted Firmware is an ARM driven effort to standardise firmware on ARM machines
<apritzel>
zoobab: at the moment it is purely AArch64, and is in fact used on quite a lot arm64 machines out there
<apritzel>
Allwinner took the 1.0 release from about two years ago and "massaged" it in their usual way, with both violating rules and introducing security holes
<apritzel>
sounds familiar? ;-)
<apritzel>
but at least they used it and even provided the source in the BSP
Putti has joined #linux-sunxi
dev1990 has quit [Quit: Konversation terminated!]
ikmaak has quit [Remote host closed the connection]
orly_owl has quit [Quit: leaving]
ikmaak has joined #linux-sunxi
vagrantc has joined #linux-sunxi
|jbrown| is now known as jbrown
jernej has quit [Ping timeout: 264 seconds]
Da_Coynul has joined #linux-sunxi
sarietta has joined #linux-sunxi
Da_Coynul has quit [Quit: My Mac has gone to sleep. ZZZzzz…]