Hi Lawrence, Thanks for your work. I may take you on your word some day. I've been also working on reverse engineering the arisc firmware and figuring out what it does. It's a bit bloated with various debug functionality, IR code, clock setup, regulator setup, etc. I hope the core functionality could be implemented without spilling half of the arisc code into DRAM, so that it could run from SRAM and do the suspend from there.
Some of the work is here (it's for H3 though, not A33): https://github.com/megous/h3-ar100-firmware-decompiler https://github.com/megous/h3-firmware regards, o. Lawrence Yu píše v St 12. 04. 2017 v 08:42 -0700: > I have been able to combine the A33 suspend to ram code from the > Allwinner 3.4 SDK into a recent mainline kernel and make suspend to > ram work if anyone would like to try it out. I tested it on a A33 > GA10H-v1.1 tablet and a Sinlinx SIN-A33 dev board. The goal was to > see if suspend to ram would work (which it does), so the code is not > clean enough for submission to the kernel. It uses the binary arisc > firmware from Allwinner which there is no source code for, and the > DRAM parameters from the FEX file have to be explicitly defined for > the device in the dts file for the arisc to use. > > The patch is too big to post to the mailing list (over 1.5MB) and the > build instructions are more complicated than building a normal > mainline kernel so I have written up instructions and posted them on > the sunxi wiki at https://linux-sunxi.org/A33_Suspend > > I don't foresee doing much more development on this, so my hope is > that this patch and instructions can help some effort in the future > to get suspend to ram into the mainline kernel. > > -- You received this message because you are subscribed to the Google Groups "linux-sunxi" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
signature.asc
Description: This is a digitally signed message part
