Re: IDE DMA on AXP & barriers

From: Kurt Garloff (garloff@suse.de)
Date: Fri Dec 07 2001 - 14:18:27 EST


Hi Jay,

thanks for your response!

On Fri, Dec 07, 2001 at 01:50:31PM -0500, Jay Estabrook wrote:
> "Identification" may need to be done via "probe and see"... :-\
>
> IIRC, there was only ever one rev of PYXIS, which has a number of
> problems.

OK.

> The first MIATAs went out with (nearly) no fixes, except for the
> console cruft to prevent "unknown" cards from being placed in the
> 64-bit slots (unkown cards == unknown drivers == possibly susceptible
> to the 8K boundary DMA error).

Which is really crufty. A warning would be sufficient.
I guess I have one of those first generation MIATAs.
<4>pci: passed tb register update test
<4>pci: passed sg loopback i/o read test
<4>pci: passed pte write cache snoop test
<4>pci: failed valid tag invalid pte reload test (mcheck; workaround available)
<4>pci: passed pci machine check test
<4>pci: tbia workaround enabled

Maybe it's not that bad. I figure there could even be more bugs ;-)
Just two out of six. Reading the file is scary!

OTOH, I always get one of those when booting:
tulip0: MII transceiver #5 config 1000 status 786f advertising 01e1.
eth0: Digital DS21143 Tulip rev 48 at 0x8000, 00:00:F8:75:B3:C3, IRQ 24.
eth0: Setting full-duplex based on MII#5 link partner capability of 45e1.
CIA machine check: vector=0x670 pc=0x120071f30 code=0x98
machine check type: processor detected hard error
pc = [<0000000120071f30>] ra = [<0000000120071200>] ps = 0008 Not tainted
v0 = 0000008800000000 t0 = 0000000000000000 t1 = 0000000000000000
t2 = 000000241300c014 t3 = 0000000000000008 t4 = 000000007ffffffe
t5 = 00000001208cbfc0 t6 = 000002000024c310 t7 = 0000000000000060
a0 = 0000000000000000 a1 = 0000000000000080 a2 = 0000000000000001
a3 = 0000000000000000 a4 = 0000000000000005 a5 = 0000008809840000
t8 = fffffc0002b3c000 t9 = 00000200001e3828 t10= 0000000000000008
t11= 000002000024c260 pv = 00000200001e7a50 at = 0000000000000000
gp = 000000012023bb18 sp = fffffc0002b3c000
reiserfs: checking transaction log (device 3a:06) ...

> LX and SX introduced off-chip PYXIS fixes, which fixed at least the
> 8K boundary DMA problem, and some others, but NOT all, most notably
> the PCI READ prefetch.

And the PYXIS manual also mentions the LOCK bug; but it should hang
the machine, if I understand correctly, so I figure that QuickSwitch
is there.

> Late MIATAs, aka MIATA-GL, have off-chip logic that fix ALL the PYXIS
> problems, AFAIK. These are distinguished from the earlier model by the
> presence of a QLogic ISP-1040C(?) SCSI chip built into the motherboard.

My QLogic is on a PCI card.

BTW, I wonder, why the Digital engineers didn't fix the pyxis rather than
doing workarounds in hardware.

Is there something like a MIATA technical reference? Thanks to Ivan, I got a
PYXIS manual :-)

Regards,

-- 
Kurt Garloff  <garloff@suse.de>                          Eindhoven, NL
GPG key: See mail header, key servers         Linux kernel development
SuSE GmbH, Nuernberg, DE                                SCSI, Security


- To unsubscribe from this list: send the line "unsubscribe linux-alpha" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html



This archive was generated by hypermail 2b29 : Fri Dec 07 2001 - 21:00:44 EST