ChanServ changed the topic of #linux-sunxi to: Allwinner/sunxi development - Did you try looking at our wiki? https://linux-sunxi.org - Don't ask to ask. Just ask and wait for an answer! - This channel is logged at https://oftc.irclog.whitequark.org/linux-sunxi
apritzel has quit [Ping timeout: 480 seconds]
Danct12 has quit [Remote host closed the connection]
Danct12 has joined #linux-sunxi
ftg has quit [Ping timeout: 480 seconds]
Daanct12 has joined #linux-sunxi
cnxsoft has joined #linux-sunxi
vagrantc has quit [Quit: leaving]
evgeny_boger has joined #linux-sunxi
Daanct12 has quit [Remote host closed the connection]
Daanct12 has joined #linux-sunxi
Daanct12 has quit [Remote host closed the connection]
Daanct12 has joined #linux-sunxi
Daanct12 has quit [Remote host closed the connection]
Daanct12 has joined #linux-sunxi
Daanct12 has quit [Remote host closed the connection]
Daanct12 has joined #linux-sunxi
apritzel has joined #linux-sunxi
Daanct12 has quit [Remote host closed the connection]
Daanct12 has joined #linux-sunxi
apritzel has quit [Ping timeout: 480 seconds]
moteen has joined #linux-sunxi
moteen has quit [Ping timeout: 480 seconds]
Daanct12 has quit [Remote host closed the connection]
Daanct12 has joined #linux-sunxi
paulk has joined #linux-sunxi
bauen1 has quit [Ping timeout: 480 seconds]
Daanct12 has quit [Remote host closed the connection]
apritzel has joined #linux-sunxi
moteen has joined #linux-sunxi
bauen1 has joined #linux-sunxi
moteen has quit [Remote host closed the connection]
moteen has joined #linux-sunxi
moteen has quit [Ping timeout: 480 seconds]
moteen has joined #linux-sunxi
moteen has quit [Remote host closed the connection]
moteen has joined #linux-sunxi
moteen has quit [Ping timeout: 480 seconds]
ftg has joined #linux-sunxi
vpeter has joined #linux-sunxi
bauen1 has quit [Ping timeout: 480 seconds]
Danct12 has quit [Remote host closed the connection]
Danct12 has joined #linux-sunxi
moteen has joined #linux-sunxi
bauen1 has joined #linux-sunxi
bauen1_ has joined #linux-sunxi
moteen has quit [Ping timeout: 480 seconds]
bauen1 has quit [Ping timeout: 480 seconds]
moteen has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
cnxsoft has quit []
vagrantc has joined #linux-sunxi
<jernej> smaeul: with just released U-Boot v2022.0, your Crust patches don't play nicely together anymore. Please check https://github.com/LibreELEC/LibreELEC.tv/pull/6630
<jernej> *v2022.07
paulk has quit [Ping timeout: 480 seconds]
<jernej> it seems that when building for H5, there is some H3 stuff that breaks build
<jernej> namely CONFIG_ARMV7_SECURE_BASE
<MoeIcenowy> BTW when will crust be mainline ready?
<MoeIcenowy> and, as V853 is out, is there a proper way to split HW resource between two processors?
<jernej> MoeIcenowy: I think that Crust already works with mainline U-Boot and TF-A
<jernej> for 64-bit SoCs
<jernej> only H3 needs U-Boot patches
<jernej> and only few Linux patches are needed
<MoeIcenowy> well I mean the Linux ones
<MoeIcenowy> I know TF-A is ready for crust for a long time
<MoeIcenowy> (BTW TF-A R329 support is upstreamed for a long time, but I didn't push Linux/U-Boot a lot yet
<jernej> Currently I have 12 Crust related patches in 5.18
<jernej> I believe one was mainlined in 5.19
<jernej> already much better situation than before, but not just there yet
<jernej> well, you have to ask smaeul about further plans
<jernej> btw, what kind of HW resources you have in mind?
<gamiee> It would be cool to have Crust ported to V853's RISC-V core
<MoeIcenowy> jernej: well I think peripherals on V853 are quite equal to both cores
<MoeIcenowy> by default I think Allwinner BSP allocates a chunk of DRAM and UART3 to RV
<jernej> ok, I'm out of the loop regarding V853
<jernej> is there any manual for it?
<jernej> or at least datasheet
<MoeIcenowy> there should be none
<MoeIcenowy> but I have some leaked thing
<jernej> well, resource allocation is usually HW dependent, so it's impossible to say if there is no available documentation or BSP code
<MoeIcenowy> BSP code gots available 3 hrs ago
<gamiee> BSP is open now? On open allwinnertech?
<apritzel> MoeIcenowy: what would be the problem with the hardware allocation to either core? Sounds like the DT is a good fit for that:
<apritzel> the DT starts with everything available, then firmware (either TF-A, U-Boot, or crust) grab exclusive possession of a device, and disable it in the DT passed down
<MoeIcenowy> gamiee: yes
<MoeIcenowy> apritzel: well it's even a problem on H616, right?
<apritzel> MoeIcenowy: only if you do the wrong thing: explicitly loading a DTB from disk ;-)
<apritzel> looks like we don't want to really support this anymore
<MoeIcenowy> but is it really wrong?
<MoeIcenowy> we do not update U-Boot DT frequently.
* jernej grabs popcorn :)
<MoeIcenowy> and Linux is now the de facto central DT hub
<apritzel> jernej: lol (and sigh)
<apritzel> MoeIcenowy: for multiple reasons, the main blocker being incompatible changes
<MoeIcenowy> I remember Fedora manually syncs DT to U-Boot in order to prevent explicit DT load
<MoeIcenowy> In addition, if you want to add something that has no stable binding
<apritzel> we broke this with syncing the 5.18 DTs in a forward compatible way now, so can more frequently update
<MoeIcenowy> use a manually loaded DT instead of bootloader shipped DT is a good idea
<apritzel> I plan on syncing the 5.20-rc1 DTs
<apritzel> MoeIcenowy: if it's for something experimental or for development: sure, but then you know what you do, and live with the consequences
<apritzel> or you are clever and use overlays
<MoeIcenowy> apritzel: overlays are not easy to use now
<MoeIcenowy> loading an overlay is in fact much more complex than loading a full DT
<MoeIcenowy> and I don't think we build U-Boot DT in a overlay-friendly way
<apritzel> the U-Boot DT is built with -@, if that's your concern
<MoeIcenowy> oh quite new
<apritzel> that patch lets you simply drop your .dtbo files into one directory, and then a generic command applies them all
<MoeIcenowy> apritzel: BTW for the FEL cache issue, I think I may try to implement it on D1 first
<MoeIcenowy> D1 suffers from some inconsistency too (although I am not sure whether it's I cache
<apritzel> was that the same issue there? that's the original problem you reported months ago?
<MoeIcenowy> maybe years ago
<MoeIcenowy> the behavior is similar to V853
<apritzel> jernej: more popcorn, the discussion steers over to caches ;-)
<MoeIcenowy> Cache for headache
<apritzel> MoeIcenowy: I am expecting a T113 any day now, let's see what surprises this one holds ...
<MoeIcenowy> apritzel: I wonder whether T113 will have the same SoC ID with D1
<apritzel> are there differences in that respect between D1 and D1s?
<MoeIcenowy> apritzel: the D1s manual does not mention DSP
<MoeIcenowy> maybe it's masked out
<MoeIcenowy> BTW Allwinner renamed D1 to D1-H per HDMI licensor request
<apritzel> DSP masked out: that would be a shame
<gamiee> I seen many socs on their website with -H suffix
<gamiee> Also looks like their series page is broken
<MoeIcenowy> apritzel: I think they tend to use some worse dies than D1 to make D1s
<MoeIcenowy> gamiee: yes, everything HDMI capable is now with -H
<jernej> did they also rename old SoCs, like A20, H3, etc?
<MoeIcenowy> I am not sure about these two
<MoeIcenowy> but A64 is now A64-H
<apritzel> even worse than D1? :-D
<MoeIcenowy> apritzel: maybe
<MoeIcenowy> the DSP contains a lot of SRAM
<gamiee> Icenowy: please, do you know if H3 is still manufactured?
paulk has joined #linux-sunxi
<MoeIcenowy> gamiee: no idea.
<jernej> If it is anything like A20, you just need to order ~10k pieces and they'll run a batch
<MoeIcenowy> smaeul: BTW have you tried the REed mctl_hal.c on a 2G Nezha?
<MoeIcenowy> apritzel: I am recently thinking about introduce Allwinner's sunXiwY naming scheme to mainline
<MoeIcenowy> considering Allwinner now tends to have a huge number of variants of the same main die
<jernej> what if we just more efficiently share DT, clocks and pinctrl among SoCs with same die?
<jernej> changing name won't make it easier to share, but we can have one pinctrl driver with superset of all pins
<jernej> same with clocks and DT
<MoeIcenowy> jernej: I agree, but I don't know whether having such a super-driver is proper
<jernej> why not?
<MoeIcenowy> jernej: it will be less error-prone
<jernej> if you really want, you can still limit exposed clocks, pins and whatnot based on specific compatible
<MoeIcenowy> we're already doing this
moteen has quit [Remote host closed the connection]
<MoeIcenowy> interestingly the RV processor on V853 is behind an IOMMU
<jernej> is it same old IOMMU or something more advanced?
<MoeIcenowy> I am not familiar with the old IOMMU too
<MoeIcenowy> but I think they do not have any reason to change it a lot
<MoeIcenowy> exactly the old one.
<MoeIcenowy> (by checking register list against H6 one
<jernej> ok, thanks for checking
<apritzel> MoeIcenowy: I don't know if adopting another *naming scheme* will help, because we can always cover this via a the compatible strings
<apritzel> like: "allwinner,sun8i-r528-pinctrl", "allwinner,sun20i-d1-pinctrl";
<apritzel> and we do this die sharing already, look at the H616 pinctrl, which has pins that only the T507(?) exposes
moteen has joined #linux-sunxi
moteen has quit [Ping timeout: 480 seconds]
<gamiee> It looks like Allwinner removed all their SoC pages from website
<gamiee> H3 seems to be still only H3 :/
apritzel has quit [Ping timeout: 480 seconds]
hlauer has joined #linux-sunxi
moteen has joined #linux-sunxi
macromorgan has joined #linux-sunxi
moteen has quit [Remote host closed the connection]
JohnDoe_71Rus has quit []
rajkosto has joined #linux-sunxi
apritzel has joined #linux-sunxi
aggi has quit [Quit: zzz]
Hypfer has joined #linux-sunxi
Hypfer is now known as hypfer
hypfer is now known as Hypfer
rajkosto has quit [Read error: Connection reset by peer]
moteen has joined #linux-sunxi
moteen has quit [Ping timeout: 480 seconds]
hlauer has quit [Ping timeout: 480 seconds]
sunshavi_ has joined #linux-sunxi
sunshavi has quit [Ping timeout: 480 seconds]
Hypfer has quit [Ping timeout: 480 seconds]
Hypfer has joined #linux-sunxi
Hypfer is now known as Guest4762
Hypfer has joined #linux-sunxi
Guest4762 has quit [Ping timeout: 480 seconds]