On Mon, 12 May 2014 10:45:44 +0200
Olliver Schinagl <[email protected]> wrote:

> 
> On 12-05-14 09:54, Siarhei Siamashka wrote:
> > On Sun, 11 May 2014 22:47:03 +0200
> > Hans de Goede <[email protected]> wrote:
> >
> >> Hi,
> >>
> >> On 05/11/2014 12:31 PM, Siarhei Siamashka wrote:
> >>> On Sun, 11 May 2014 10:36:25 +0200
> >>> Hans de Goede <[email protected]> wrote:
> >>>
> >>>> We've several bug reports indicating that this causes stability issues,
> >>>> so lets just drop it.
> So not just me? jikes ...
> >>> I'm afraid that the reporters still have not done their homework
> >>> properly. And suspect that the FAST_MBUS just happens to be the last
> >>> straw to break the camel's back on slightly unstable systems.
> FAST_MBUS isn't supposed to fix stability problems though does it? It 
> offers a performance/speed boost (if stable), or atleast that was my 
> understanding.
> >> Well at least Oliver has done quite a bit of testing, with dcdc3 set
> >> to 1.3V not only in u-boot but also in the fex file.
> Well not quite a bit, I ran into the problem that my board became very 
> unstable after enabeling (and boosting the voltage). Otherwise it ran 
> perfectly fine (cb3 btw)
> 
> > I don't quite trust the tests, which need tens of hours to run
> > before detecting a problem. The main issue here is that we may have
> > a significant variance in the problem detection time (let's say,
> Tests to see if it is stable? Absolutely. In my case, the board couldn't 
> even reliably boot

It is interesting to check if it could be something related to:
    https://www.mail-archive.com/[email protected]/msg04700.html

I have also recently found that the Cubieboard fork of u-boot-sunxi
has a similar fix, but additionally introduces a bunch of axp209_write
calls:
    https://github.com/cubieboard/u-boot-sunxi/commit/fb357b4e1f07

> and in the rare occasion it did manage to boot, a tar 
> xJvf job would crash due to errors after a few minutes or had corrupted 
> results.

I believe that lima-memtester is a much stronger dram reliability test
than a tar job regardless of the MBUS setup.

It would be still very interesting to check if your configuration
passes the lima-memtester test after you remove FAST_MBUS and tar
failures disappear.

> > something between 10 and 30 hours for example). And if the person
> > running the test is having some expectations (like assuming that
> > lower MBUS clock provides better stability), then this person may
> > stop some of the test runs prematurely and come to biased conclusions.
> >
> > Also we need to confirm that the fex changes have really taken effect.
> > It is possible to grep the dmesg log for "buck3" string after booting
> > up to find the dcdc3 voltage there.
> That is a fair point, that I may not have properly tripple checked ...
> >
> >> Oliver, can you retest your cubietruck breakage with FAST_MBUS, with
> >> the dcdc voltage traised to 1.425 V in u-boot + Siarhei's fex file
> >> changes he send earlier to day, and see if your FAST_MBUS triggered
> >> regression then goes away?
> I can in the next week or so; I"ll try to find Siarhei's Fex changes and 
> run it at 1.425 v
> 
> I don't recall if i kept the mbus in sync with the dram clock however.

If you are clocking dram at just 432MHz in your Cubietruck, then your
situation indeed looks very similar to:

> We also had a report about some A20 device not being able to run stable
> with dram clocked at 432MHz. Unless MBUS was also synchronously clocked
> at 432MHz (using PLL5 instead of PLL6):
>      http://irclog.whitequark.org/linux-sunxi/2013-11-26#5690482

So just a random speculation. Could it be that you (and hglm) somehow
have a "bad" PLL6 in your boards, which causes stability issues if used
for clocking MBUS?

And there is also a magic trick (discovered by jemk) to significantly
improve dram overclocking potential on the Cubietruck. You can set
dram_tpr3 to 0x72222 in the dram_para struct in the u-boot. But if
your MBUS related problems have a different root cause, then it might
not help much.

In any case, you seem to be lucky to have a very unique device with
interesting properties :-) Just be sure to report all your findings.

-- 
Best regards,
Siarhei Siamashka

-- 
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.

Reply via email to