FreeBSD-13.0-CURRENT-arm64-aarch64-ROCKPRO64-20201210-7578a4862f0 broken ?

[Søren Schmidt]

> On 14 Dec 2020, at 22.45, Søren Schmidt <soren.schmidt at gmail.com <mailto:soren.schmidt at gmail.com>> wrote:
> 
> Went to the official u-boot sources and tried the newest shiniest 2021.01rc3 and it behaves a little bit better, it will boot 1 out of 3 times :)
> 
> Probe looks like this now:
> 
> U-Boot TPL 2021.01-rc3 (Dec 14 2020 - 23:17:52)
> Channel 0: LPDDR4, 50MHz
> BW=32 Col=10 Bk=8 CS0 Row=16/15 CS=1 Die BW=16 Size=2048MB
> Channel 1: LPDDR4, 50MHz
> BW=32 Col=10 Bk=8 CS0 Row=16/15 CS=1 Die BW=16 Size=2048MB
> 256B stride
> lpddr4_set_rate: change freq to 400000000 mhz 0, 1
> lpddr4_set_rate: change freq to 800000000 mhz 1, 0
> 
> 
> The exact same as before. There is no difference in output from the working boots to the failing ones.
> I’ll get on with adding some debug to what parameters get set during memory setup etc.
> However I have this gut felling it is timing related and my board might just be too slow/fast for some operation that the ayufan version does slightly different (it boots every time).
> 
> Anyhow, when it boots it runs rock stable, it builds worlds etc with no issues..

OK, I got a little further, it may not be memory related after all, from dozens ff boots with same u-boot etc I found that those that fails sees the SDcard like this:

mmcsd0: 16GB <SDHC SMI   0.0 SN 00000000 MFG 03/2020 by 0 0x0000> at mmc0 50.0MHz/4bit/2048-block

And those boots that works:

mmcsd0: 16GB <SDHC SMI   0.0 SN 00000000 MFG 03/2020 by 0 0x0000> at mmc0 20.0MHz/4bit/2048-block

So it suggests that either our code for setting the SDcard speed I flaky or marginal or the SD card / my pinebookpro is...
However the exact same SD card can boot netbsd/openbsd dozens of times with no issues at all with 50MHz timings.

Back to debugging :)

--
Søren Schmidt
sos at deepcore.dk <mailto:sos at deepcore.dk> / sos at freebsd.org <mailto:sos at freebsd.org>
"So much code to hack, so little time”