Do you remove eMMC setting in test 3?
Update: with test-3, emmc not detected after 3 reboots
Now flashing again test-2
that doesnāt help me.
TEST3, whatās the max stable setting?
In test 3 options were removed
I canāt see how, Iām testing here on CRAASK and they show/work correctly. Try a fresh build?
coreboot_edk2-pujjo-mrchromebox_20260918-emm_test-1.zip (4.8 MB)
In fact, yesterday I thought you were going to remove the options and leave just one. Iāll take a look at the one you just posted today in a bit.
the only thing I removed was the confusing auto/unlimited options. I should have removed HS400 too, since thatās the same as auto
This build works fine, eMMC options available, thanks.
Why not left by default the lowest speed? I mean for safe. In OS speed it should be the same or not?
in order to boot as quickly as possible, youād want the controller to use the fastest speed it supports, no? What the OS decides is up to the OS drivers/policy, but itās normally the same.
The difference between the firmware init and the OS init is how complicated the drivers/init sequence are. It would take a non-trivial amount of effort to determine exactly why the eMMC controller is failing under edk2 in some cases, for specific boards. The best I can offer is a workaround (lowering the speed) without being able to reproduce or identify a specific bug from the existing logging.
so all Iām asking for is a clear answer on the max speed at which your board/eMMC will boot reliably.
Is strange because, for example with this latest build, HS400 200MHz DDR option is booting fine
Still not test with āauto (no cap)ā option
Edit: auto option works fine so far (reboot about 5 or 6 times)
Then, seem what i dont have a clear answer, in fact in the first build with hs400 option random failures were presented, with latest build hs400 is booting fine, i dont understend what happen.
In other things, i dont speak english, then for me is difficult explain the behaviour.
in the first build, HS400 did not use HS400-ES (enhanced strobe), which I think was the missing part. Linux shows HS400-ES on my CRAASK, so I figured that might be the issue and added support for it to edk2.
āautoā option in latest build is hs400es? And/or hs400 is really hs400es?
correct. I donāt have an option for HS400 without ES currently, too hard to disentangle
So that was it, you could say its been resolved.
A new problem with this build; I hadnāt noticed but no audio in windows, speaker seem working (volume icon show waves) but none sound can heard. Maybe pujjo vs pujjoteen acpi differences or nothing related?
there is no difference in firmware between the two devices. And no change from previous firmware audio wise. If you roll back, does audio work again?