<apritzel>
at least for SCMI (the successor of SCPI)
Net147 has quit [Read error: Connection reset by peer]
<smaeul>
is mali-supply touched by runtime PM? if so, you'll need to hook up the gating bit to the Linux driver. otherwise you only care during suspend
<smaeul>
which for H616 would probably all be done in TF-A
Net147 has joined #linux-sunxi
<jernej>
I think power domain construct was invented for exactly this cases
<jernej>
smaeul: sure, mali-supply is used for dvfs, so it's actively used by driver
<apritzel>
jernej: so what about a generic power-domain driver, that takes a regulator and a regmap, and exports a power domain that can be easily used with the generic power-domain property?
Net147 has quit [Read error: Connection reset by peer]
<jernej>
ah, I only started looking at power domain concept, not sure how everything fits together yet
<smaeul>
apritzel: if you want to have Linux set the bit, you have to make all of the gating nonsecure
Net147 has joined #linux-sunxi
<apritzel>
personally I am all for abstracting that in firmware
<smaeul>
I agree
matthias_bgg has quit [Quit: Leaving]
zoobab has joined #linux-sunxi
<jernej>
so, bsp doesn't make gating secure? this bit is toggled from gpu driver
<smaeul>
probably not? the ATF dump from way back when had them setting the PRCM security switch to all non-secure
tuxd3v has quit [Ping timeout: 240 seconds]
zoobab has quit [Ping timeout: 264 seconds]
<apritzel>
their understanding of "secure" is somewhat weird
zoobab has joined #linux-sunxi
zoobab has quit [Ping timeout: 246 seconds]
<jernej>
there is "arm,scpi-power-domains" but no DT uses it...
MangyDog has joined #linux-sunxi
Mangy_Dog has quit [Ping timeout: 246 seconds]
<apritzel>
jernej: the Juno is using it
<apritzel>
jernej: so another U-Boot patch later, to tweak the BL31 load address for H616, and "make" create a working u-boot-sunxi-with-spl.bin, just dd to sector 16 and it boots
<apritzel>
with a 40KB SPL
<apritzel>
smaeul: and a sad little empty SCP node :-(
<jernej>
nice, I'll try tomorrow, if you push it on git
zoobab has joined #linux-sunxi
Net147 has quit [Read error: Connection reset by peer]
Net147 has joined #linux-sunxi
ChriChri_ has joined #linux-sunxi
ChriChri has quit [Ping timeout: 240 seconds]
ChriChri_ is now known as ChriChri
Net147 has quit [Read error: Connection reset by peer]
Net147 has joined #linux-sunxi
montjoie has quit [Ping timeout: 264 seconds]
MangyDog has quit [Ping timeout: 260 seconds]
montjoie has joined #linux-sunxi
cnxsoft has joined #linux-sunxi
Net147 has quit [Read error: Connection reset by peer]
Net147 has joined #linux-sunxi
gnarface has quit [Remote host closed the connection]
cnxsoft has quit [Read error: Connection reset by peer]
iamfrankenstein has joined #linux-sunxi
linkmauve has joined #linux-sunxi
<Tooniis>
alright so i'll have to get serial
victhor has joined #linux-sunxi
lurchi_ is now known as lurchi__
lucascastro has joined #linux-sunxi
lucascastro has quit [Ping timeout: 256 seconds]
lucascastro has joined #linux-sunxi
lucascastro has quit [Ping timeout: 240 seconds]
lucascastro has joined #linux-sunxi
lucascastro has quit [Ping timeout: 272 seconds]
lurchi__ is now known as lurchi_
lucascastro has joined #linux-sunxi
apritzel has joined #linux-sunxi
lucascastro has quit [Remote host closed the connection]
lucascastro has joined #linux-sunxi
<apritzel>
jernej: U-Boot patches> FEL64 and mkimage are waiting for review (hint hint), but I am quite confident to merge them in the January merge window
<apritzel>
and I am quite hopeful that we can get H616 support in the same merge window
<apritzel>
jernej: so yes, please base them on top
lucascastro has quit [Ping timeout: 256 seconds]
Mangy_Dog has joined #linux-sunxi
<apritzel>
jernej: arisc code> I know, it's just that the common code relies on those files, and I haven't got around to remove that or make this dependent on the SoC build
<apritzel>
jernej: and since you asked for it, I just pushed what I have ;-)
<apritzel>
jernej: btw: if you just build the orangepi_zero2_defconfig (with $BL31 pointing to bl31.bin), the u-boot-sunxi-with-spl.bin file works on the SD card
<jernej>
I know, but I'm not there yet
<jernej>
(I'm doing code refactoring)
JohnDoe_71Rus has joined #linux-sunxi
lurchi_ is now known as lurchi__
<apritzel>
jernej: I was wondering if we should have a SUN50I_H6_GEN symbol or the like, to cover the different clock style and memory map
<apritzel>
I think all newer SoCs will use that, the A100 and the A63 do, for instance, and I saw even one new 32-bit part with that memory map as well
<jernej>
ok, I'll look into that
lurchi__ is now known as lurchi_
<MoeIcenowy>
apritzel: I agree with that
<wens>
apritzel: did you get access to u-boot custodian repo?
<apritzel>
wens: yes
lucascastro has joined #linux-sunxi
lucascastro has quit [Read error: Connection reset by peer]
gaston1980 has quit [Quit: Konversation terminated!]
lurchi_ is now known as lurchi__
netlynx has quit [Quit: Ex-Chat]
<jernej>
apritzel: any luck with USB on Linux?
<apritzel>
jernej: "usb usb4-port1: Cannot enable. Maybe the USB cable is bad?"
<apritzel>
U-Boot had a generic message as well: "scanning bus usb@5200000 for devices... cannot reset port 1!? 1 USB Device(s) found"
<apritzel>
so there must be more
<apritzel>
the BSP USB PHY driver is completely different from mainline, of course, so it takes a while to dig through
<jernej>
there are some H616 specifics in Linux driver, like low power mode
<jernej>
and some HW bug workaround
<apritzel>
so I scanned for quirk for sun50iw9p1, but those were the same as sun50iw6p1 (H6), so I think that's not it
<jernej>
skip "p1"
<jernej>
apritzel: you should search for CONFIG_ARCH_SUN50IW9
<apritzel>
jernej: I forgot to CC: you on one reply to the MMC DMA shift patch, it should be on the list: "Re: [RESEND PATCH 17/19] mmc: sunxi: add support for A100 mmc controller"
<jernej>
ok, I'll take a look
<apritzel>
I think the BSP has another MMC bug affecting endianess ...
vagrantc has joined #linux-sunxi
tuxd3v has joined #linux-sunxi
JohnDoe7 has joined #linux-sunxi
JohnDoe_71Rus has quit [Ping timeout: 246 seconds]