mmind00 changed the topic of #linux-rockchip to: Rockchip development discussion | public log at https://libera.irclog.whitequark.org/linux-rockchip
System_Error has joined #linux-rockchip
naoki has quit [Ping timeout: 252 seconds]
naoki has joined #linux-rockchip
sjoerd has quit [Ping timeout: 265 seconds]
naoki has quit [Quit: naoki]
naoki1 has joined #linux-rockchip
naoki1 is now known as naoki
sjoerd has joined #linux-rockchip
System_Error has quit [Remote host closed the connection]
stikonas has quit [Quit: Konversation terminated!]
System_Error has joined #linux-rockchip
System_Error has quit [Remote host closed the connection]
System_Error has joined #linux-rockchip
naoki has quit [Quit: naoki]
naoki has joined #linux-rockchip
Stat_headcrabbed has joined #linux-rockchip
Stat_headcrabbed has quit [Quit: Stat_headcrabbed]
Stat_headcrabbed has joined #linux-rockchip
hexdump0815 has quit [Ping timeout: 260 seconds]
hexdump0815 has joined #linux-rockchip
System_Error has quit [Remote host closed the connection]
System_Error has joined #linux-rockchip
Stat_headcrabbed has quit [Remote host closed the connection]
robmur01 has quit [Ping timeout: 272 seconds]
System_Error has quit [Remote host closed the connection]
robmur01 has joined #linux-rockchip
System_Error has joined #linux-rockchip
franoosh has quit [Remote host closed the connection]
franoosh has joined #linux-rockchip
urja has quit [Read error: Connection reset by peer]
valpackett has joined #linux-rockchip
MyNetAz has joined #linux-rockchip
urja has joined #linux-rockchip
Stat_headcrabbed has joined #linux-rockchip
Stat_headcrabbed has quit [Quit: Stat_headcrabbed]
naoki has quit [Read error: Connection reset by peer]
naoki1 has joined #linux-rockchip
naoki1 is now known as naoki
raster has joined #linux-rockchip
Stat_headcrabbed has joined #linux-rockchip
Stat_headcrabbed has quit [Client Quit]
dsimic has quit [Ping timeout: 260 seconds]
dsimic has joined #linux-rockchip
naoki has quit [Quit: naoki]
naoki1 has joined #linux-rockchip
naoki1 is now known as naoki
unkraut has quit [Remote host closed the connection]
unkraut has joined #linux-rockchip
<qschulz> macromorgan: 256K is too small IMO. It also doesn't need to be a "ton of overlay" just one that is bringing a lot of nodes or big array properties.
<qschulz> Kwiboo: so, I am not sure the implem of your pinctrl patches actually is correct. This expects the GPIO indices to be contiguous, and I'm not sure they necessarily are? The naming scheme may be difficult to extrapolate from the actual GPIO index as well
<qschulz> in any case, I don't think the nr_pins in the controller makes much sense to keep
<qschulz> if we plan on keeping index contiguity between banks, then it's rather simple to know how many pins there are in a controller
<qschulz> it's the base of the last bank + the nr of pins in the last bask
<qschulz> bank*
<qschulz> something like this maybe? https://paste.ack.tf/cb6f18
<Kwiboo> qschulz: yeah I know the whole pinctrl <-> gpio bank is a mess :-)
<Kwiboo> my understanding is that the gpio controller can handle 32 pins, however not all pins are routed/in-use, and for some socs the pins are marked as not-in-use and skipped and others they are not :-S
<qschulz> Kwiboo: if there's no contiguity, then simply summing up the nr_pins of each bank would be enough I guess?
<qschulz> Kwiboo: yeah, I vaguely remember some pin in the middle of banks are not routed. E.g. GPIO2_B stops at B4 instead of the typical B7 and the continues at GPIO2_C0 or something like that
<Kwiboo> if we say that bank nr_pins are the number of pins that are usabe/routed for pinctrl, then your diff probably should be fine, thanks, will test it out later
<qschulz> Kwiboo: oh yeah, I see what you mean, this also means we're expecting the pin index to be contiguous within a bank
<qschulz> oof, don't know how we can do in a simple manner except having an array of pins
<qschulz> or mmmmmm
<Kwiboo> yeah, there are those holes in the banks, and I think for rk3288 the last 8 pins for gpio0 is unsused and last 16 in gpio8, and for rk3066 there is no gpio5 yet gpio6 exists :-S
<qschulz> maybe we always have an array of max nr_pins as defined for the bank
<qschulz> but some entries in the bank can be NULL
<qschulz> and you shouldn't stop to iterate over, just skip that one
<qschulz> ah, for non-existing banks.... that'll be interesting to handle
<Kwiboo> https://patchwork.ozlabs.org/project/uboot/patch/20240802225705.2640999-10-jonas@kwiboo.se/ and the linked linux patches tried to define the relation but the information is based on what the gpio/pinctrl driver not the hw pins
<Kwiboo> my understanding is that pinctrl driver should expose X number of pins that can be controlled (should idellay match hw), and then the gpio-ranges can be used to configure what pinctrl pin is mapped to what gpio bank pin
<Kwiboo> so for your example something like gpio-ranges = <&pinctrl 0 64 12>, <&pinctrl 16 76 16>; would mean i.e. pinctrl 64-73 <> gpio2 a0-b3 and pinctrl 74-91 <> gpio2 c0-d7
System_Error has quit [Remote host closed the connection]
System_Error has joined #linux-rockchip
naoki has quit [Ping timeout: 248 seconds]
naoki has joined #linux-rockchip
System_Error has quit [Remote host closed the connection]
punit has quit [Ping timeout: 252 seconds]
System_Error has joined #linux-rockchip
MyNetAz has quit [Remote host closed the connection]
MyNetAz has joined #linux-rockchip
naoki has quit [Ping timeout: 272 seconds]
naoki has joined #linux-rockchip
Stat_headcrabbed has joined #linux-rockchip
Stat_headcrabbed has quit [Client Quit]
Stat_headcrabbed has joined #linux-rockchip
psydroid has joined #linux-rockchip
ungeskriptet has quit [Remote host closed the connection]
ungeskriptet has joined #linux-rockchip
ungeskriptet has quit [Remote host closed the connection]
ungeskriptet has joined #linux-rockchip
<macromorgan> qschulz: I'm not really comfortable recommending anything larger than 384K today, until we increase the size of the A-TF and that seems to be a lot more work that adjusting a few parameters. 512K technically worked too but with a maximum of no more than 16K more code before things started to choke.
naoki has quit [Quit: naoki]
naoki1 has joined #linux-rockchip
naoki1 is now known as naoki
<qschulz> I guess 384K is fine? We should check what the alignment requirements are but I believe it just needs to be 8B aligned
franoosh has quit [Read error: Connection reset by peer]
psydroid has quit [Quit: KVIrc 5.2.6 Quasar http://www.kvirc.net/]
ungeskriptet has quit [Remote host closed the connection]
naoki has quit [Quit: naoki]
naoki1 has joined #linux-rockchip
ungeskriptet has joined #linux-rockchip
naoki1 is now known as naoki
raster has quit [Quit: Gettin' stinky!]
naoki has quit [Quit: naoki]
ungeskriptet has quit [Remote host closed the connection]
ungeskriptet has joined #linux-rockchip
ldevulder has quit [Read error: Connection reset by peer]
ldevulder_ has joined #linux-rockchip
<macromorgan> that's what I see yes
<macromorgan> yes, the code suggested it needed to be aligned
<macromorgan> so I guess I'll try for 384K and see if anyone balks. Now I just need to figure out how to submit upstream patches for A-TF and Optee.
<qschulz> TF-A is through gerrit
<qschulz> you need an account there
<qschulz> then use git-review tool for it
<qschulz> you need to rebase on top of the integration branch
<CounterPillow> did they remove forced -Werror from their Makefile yet? If not, enjoy using any compiler that's not some ancient gcc 4.x by Linaro.
<qschulz> Why would anyone remove -Werror?
<qschulz> fix the warnings :)
<qschulz> I built with gcc and clang from fedora 41 recently
<qschulz> well, not RK3399
<macromorgan> probably did it wrong, but it's my first commit to A-TF
<qschulz> well... the Aarch64 part, the cortex-m0 part is still compiled only with gcc
<qschulz> macromorgan: seems like a reasonable patch to me, not sure about the commit title prefix but we'll see
<macromorgan> going to do the optee commit next
<qschulz> they'll complain about the Copyright year at the top of the file not being updated
<macromorgan> then the Linux one and we should be good to go
<qschulz> I already complained to them this makes 0 sense, they don't care :)
<qschulz> which Linux one?
<macromorgan> okay, well I can always V2 it I guess
<macromorgan> the rk3588-base.dtsi... remove the Optee node from there
<macromorgan> if Optee is enabled U-Boot will add that node... if Optee is not enabled there's no reason to have that node there in the first place
<macromorgan> (and if that node is present the current code won't add the memory holes)
<CounterPillow> qschulz: because "the build doesn't work if my compiler version doesn't precisely match yours" is very annoying for distributors
System_Error has quit [Remote host closed the connection]
Stat_headcrabbed has quit [Remote host closed the connection]
System_Error has joined #linux-rockchip
Stat_headcrabbed has joined #linux-rockchip
Stat_headcrabbed has quit [Client Quit]
Stat_headcrabbed has joined #linux-rockchip
Stat_headcrabbed has quit [Client Quit]
<macromorgan> I've noted basically how all the fixes link together on this comment here... long story short unless you do all the commits and then make 3 config changes to U-Boot, Optee won't work. But if you do, it'll work just fine: https://github.com/OP-TEE/optee_os/pull/7059#issuecomment-2625252037-permalink
MyNetAz has quit [Remote host closed the connection]
System_Error has quit [Remote host closed the connection]
stikonas has joined #linux-rockchip
System_Error has joined #linux-rockchip
MyNetAz has joined #linux-rockchip
UndrWater has quit [Quit: ZNC 1.8.2+deb2ubuntu0.1 - https://znc.in]
UndrWater has joined #linux-rockchip
paulk has quit [Ping timeout: 260 seconds]
paulk has joined #linux-rockchip
paulk has joined #linux-rockchip
<CounterPillow> is the RK3588's AHB bus 64 bits wide by any chance?
System_Error has quit [Remote host closed the connection]
System_Error has joined #linux-rockchip
naoki has joined #linux-rockchip
System_Error has quit [Remote host closed the connection]
System_Error has joined #linux-rockchip
Perflosopher03 has quit [Ping timeout: 260 seconds]
ldevulder_ has quit [Remote host closed the connection]