Hi Marty,
On Sat, 13 Mar 2021 at 12:40, Marty E. Plummer hanetzer@startmail.com wrote:
On Wed, Mar 10, 2021 at 11:52:07PM -0500, Simon Glass wrote:
Hi,
On Thu, 13 Aug 2020 at 13:35, Alper Nebi Yasak alpernebiyasak@gmail.com wrote:
Hi Simon, Marty,
I'm interested in getting U-Boot to work with Kevin as well, but don't have a Servo (or the willingness to open up the case yet), so I've been trying to boot from depthcharge as in README.chromium-chainload.
I don't have a way to see serial output and I see no other signs of life. Can you give me a tested configuration that immediately powers-off or reboots a Kevin so I can confirm what I'm doing works on the chainloading side? I mean I can boot Linux, but trying the same with U-Boot just gives me a blank screen even after accounting for a lot of things.
Meanwhile, I've wrote some code to automate making depthcharge partition images, and to enable the display on Kevin (and perhaps Bob). Since I don't know if chainloading works, I don't know if these are broken or not either. I'm unsure about sending untested patches to the list, so I put them up here if you want to take a look (and maybe test/fix them?):
https://github.com/alpernebbi/u-boot/tree/rk3399-gru-kevin/wip
They're not really things that'd make a non-booting Kevin boot, though. I hope at least some of it can be useful in some way.
I have the em100 working and have got to the same point as you, Marty.
em100 -s -c gd25lq64 -d /tmp/b/chromebook_kevin/u-boot.rom -r
So I suppose that means that SDRAM is running and we just need a SPI driver? I will see if I can figure out what is missing...
Update...it seems to just be missing the ID. I pushed a new patch to:
Christ, its always something small and stupid. Perhaps the failure message should be amended to indicate 'unknown jedec id: %x' or so to be a bit more informative.
It doesn't do that because it is the SPL 'tiny' version.
This looks promising. Built it (away from the machine right now so can't test) and it seems that u-boot-rockchip.bin is just a bit too large to be flashed (8.8mb)? Or judging by your above em100 invocation this image is not to be used? If so, why is it produced at all?
I think it is for booting from eMMC. But I am booting from SPI flash.
Now I see:
Channel 0: LPDDR3, 933MHz BW=32 Col=10 Bk=8 CS0 Row=15 CS1 Row=15 CS=2 Die BW=16 Size=2048MB Channel 1: LPDDR3, 933MHz BW=32 Col=10 Bk=8 CS0 Row=15 CS1 Row=15 CS=2 Die BW=16 Size=2048MB 256B stride
U-Boot SPL 2020.10-rc1-00111-gc31b9b4a3f1-dirty (Mar 10 2021 - 21:48:26 -0700) Trying to boot from SPI
U-Boot 2020.10-rc1-00111-gc31b9b4a3f1-dirty (Mar 10 2021 - 21:48:26 -0700)
Model: Google Kevin DRAM: 3.9 GiB Cannot find regulator pwm init_voltage MMC: mmc@fe320000: 1, sdhci@fe330000: 0 Loading Environment from MMC... *** Warning - bad CRC, using default environment
Got rc -1, expected 100 Failed to probe keyboard 'keyboard-controller' In: serial@ff1a0000 Out: serial@ff1a0000 Err: serial@ff1a0000 Model: Google Kevin Net: No ethernet found. Hit any key to stop autoboot: 0 =>
No display and various errors on the way up, but at least it boots to a prompt.
A much better situation then before. I'll pull your changes into my tree and see what can be done with it.
OK. I added a little patch to fix the EC as well
Regards, Simon