Re: Yes, NetBSD still boots on pretty ancient Alphas...

Jason Thorpe <[email protected]> Thu, 29 Jul 2021 07:23:25 -0700
Newsgroups gmane.os.netbsd.ports.alpha
Message-ID <[email protected]>
> On Jul 28, 2021, at 10:08 AM, Istvan Gyenes <[email protected]> wrote:
> 
> 
> On 2021. 07. 28. 17:37, Jason Thorpe wrote:
>>> On Jul 28, 2021, at 12:28 AM, Istvan Gyenes <[email protected]> wrote:
>>> 
>>> [   1.0000000] depca2 at eisa0 slot 4: DEC DE422 Ethernet
>>> [   1.0000000] depca2: shared memory at 0xd0000-0xdffff
>>> [   1.0000000] le0 at depca2
>>> [   1.0000000] le0: failed to clear memory at 0 (0xff != 0xaa)
>> I am suspicious of this bit of code in depca.c:
>> 
>>         bus_space_write_2(dsc->sc_iot, dsc->sc_ioh, DEPCA_CSR,
>>             DEPCA_CSR_DUM | DEPCA_CSR_IEN | DEPCA_CSR_SHE |
>>             (dsc->sc_memsize == 32*1024 ? DEPCA_CSR_LOW32K : 0));
>> 
>> Specifically, I'm suspicious of the DUM bit.  I found an old DEPCA driver for Linux that calls that bit _128KB.
>> 
>> Can you try building a kernel with that bit removed from that bus_space_write_2() call?
> 
> Removed "DEPCA_CSR_DUM | " recompiled, but nothing has changed.

Hrmph.  Ok.  Well, I've managed to track down an ISA version of this card (a DE202).  Once it arrives I'll be able to debug a bit more directly on my AXPpci33.

> 
>    1.0000000] scsibus0 at ahb0: 8 targets, 8 luns per target
> [   1.0000000] depca2 at eisa0 slot 4: DEC DE422 Ethernet
> [   1.0000000] depca2: shared memory at 0xd0000-0xdffff
> [   1.0000000] le0 at depca2
> [   1.0000000] le0: failed to clear memory at 0 (0xff != 0xaa)
> 
> --
> 
> istvan

-- thorpej