Re: [EXT] Technologic Systems TS-7200 Testing, Bugs, and Patches

[Jonathan Chapman <lists%glitchwrks.com@localhost>]

> For issue #2, a workaround in Redboot is to send a SOFT_RESET (lowest bit) to the epe selfCtl register, right before you hand off to NetBSD.   At Reboot (or in your Redboot startup script), issue this:
> 
> mfill -b 0x80010020 -l 4 -p 0x1 -4
> 
> Then you can issue a "go 0x6066000" to boot NetBSD from Redboot.

That's how we isolated the issue and worked around it.

> This has worked for me for a long time and eliminates the vm page corruption when ethernet is physically connected at boot.    Presumably an unconditional epe0 reset when NetBSD loads should also work, but my first attempt a bunch of years ago didn't work first attempt, and I never circled back to it.

Had to nail it down pretty early as we usually netboot boards while doing development work!

> Great to hear about the TS-4600 work.   If it helps, I have TS-4900 support working (arm7, i.MX6) with device tree adjustments, and am working in the background on getting the aarch64 TS-4300 (i.MX93) working.

I'll have to look at it to see what overlap there is. This has been my first real adventure into *kernel* work on ARM. For most of my personal projects, and all of the $day_job work, we've just chosen platforms that already had mature NetBSD support.

Thanks,
Jonathan