Here we are, still hacking away at 440bx raminit, while I accidentally
dropped the list from the trail. :\
Branden told me this:
"I did some more testing of the patch train, but didn't manage to
figure out where it would boot but with an unstable config.
I skipped over 73887, since it was already merged (and master tests okay).
73951, 73952, 74032, and 74033 booted and ran memtest for a few
minutes without errors
But 74034 would fail with or without debug raminit enabled.
Most of this testing was done with the pair of 128mb capable of
2-2-2-5 timings, but showing 3-3-3-6 as in use in memtest86."
Which seems to reflect my tests as well. Testing continues.
Meanwhile, libreboot made another release with this cruel reality:
"Removed boards
asus p2b_ls/p3b_f - they didn’t boot properly in pcbox, and the real
hardware is basically useless / impossible to find"
:(
(And they adopted the WIP Haswell NRI we still haven't fully.)
On Mon, 27 Mar 2023 at 20:45, Branden Waldner wrote:
>
> Hi Keith
>
> So far I've only tested the complete patch train (73888) like
> previously mentioned and didn't have my serial console set up at the
> time.
>
> I dug out the cable from behind my desk and got it run to my work
> bench again, so that should help out.
>
> I'm assuming you would want logs with raminit debug enabled though and
> I didn't have that set last time.
>
> You had mentioned you wanted me to test 74036 specifically?
>
> I checked out the ram I was using with decode-dimms, but didn't check
> any other ram yet. It looks like they are both rated for CL3 @100mhz
> and the 133mhz stick seems to run at it's timings for 133 instead of
> 100? I'm running a 100mhz fsb coppermine cpu and the memory is
> supposed to run at the fsb speed right? It doesn't seem to use (or at
> least display it is using) the 100mhz timing on the oem bios either -
> though I haven't done enough testing to be confident of what's going
> on.
>
> I have some high density 256mb memory that runs at 128mb with this
> chipset, so I could check what the timings are on that yet. I really
> should have just tried to find a lot of 512mb high density to just
> fill out whatever sdram usage I could have.
>
>
> Branden
___
coreboot mailing list -- coreboot@coreboot.org
To unsubscribe send an email to coreboot-le...@coreboot.org