RPI4 Install r357606 test 09.02.2020 failed

Mark Millard marklmi at yahoo.com
Sun Feb 9 22:48:23 UTC 2020



On 2020-Feb-9, at 14:19, Bakul Shah <bakul at bitblocks.com> wrote:

> On Feb 9, 2020, at 12:26 PM, Mike Karels <mike at karels.net> wrote:
>> 
>>> On Sun, Feb 9, 2020, 1:10 PM <ag at x86.ch> wrote:
>> 
>>>> ok I understand  not all opensource  as always claimed  Too bad
>>>> 
>> 
>>> The Raspberry Pi isn't really a totally open design.
>> 
>>> Warner
>> 
>> Indeed.  In this case, the GPL Linux driver is available as a substitute
>> for documentation, but it isn't the clearest description of the (5 versions)
>> of the Ethernet section.
> 
> Perhaps the plan9 drivers would help? They are considerably
> simpler and I suspect you mainly want h/w parameters. 9pi (arm)
> & 9front (arm64) have ethernet, wifi & xhci working.
> 
> Re: documentation. I asked on their forum about their long
> promised documentation and was told it is hung up due to some
> sort of personal issue. The RPF engineers hang out on their
> forum and they do help provided one asks specific questions.
> 
> I will be happy to help once I re-bootstrap into freebsd kernel
> hacking. First I need to figure out why it boots partially and
> then hangs with repeated controller timeout messages:
> 
> ....
> Timecounters tick every 1.000 msec
> Obsolete code will be removed soon: random(9) is the obsolete Park-Miller LCG f8
> usb_needs_explore_all: no devclass
> mmcsd0: 16GB <SDHC SU16G 8.0 SN 0054AA0F MFG 01/2015 by 3 SD> at mmc0 50.0MHz/4k
> sdhci_bcm1-slot0: Got command interrupt 0x00030000, but there is no active comm.
> sdhci_bcm1-slot0: ============== REGISTER DUMP ==============
> dhci_bcm1-slot0: Sys addr: 0x00000000 | Version:  0x00009902
> ...

It is messed up long before that. The earliest that I
know of is when it first outputs:

psci0: PSCI version number mismatched with DT
device_attach: psci0 attach returned 6

By then the PSCI version number comes back as 0 (both
major and minor).


> I built -current locally a couple days back and it has the same
> issue. 
> 

If you want a build that works, use the kernel from head
-r356767 . The next head version ( -r356776 ) is where
the broken status starts for RPi4 and RPi3. RPi4B's with
4 GiBYTes of RAM fail to boot. As I understand, at least
some forms of RPi3 boot but are messed up, such as only
running 1 core. (Not surprising given the PSCI version
problem which happens there as well.)

===
Mark Millard
marklmi at yahoo.com
( dsl-only.net went
away in early 2018-Mar)



More information about the freebsd-arm mailing list