[Arm-netbook] RAM, uboot

[David Given]

On 21/05/12 14:12, Mike Thompson wrote:
[...]
> I've also been getting ext4 corruption that I attributed to the errors
> that I'm seeing from the SATA driver.  Also, I'm occasionally get seg
> faults in applications which should not have them.  Again I'm
> attributing these to the SATA errors, but without much evidence.

My box, which mysteriously started crashing for no apparent reason, has
now mysteriously stopped crashing for no apparent reason. I've tried
booting from the SATA disk and it mostly works, but I did once get a
stream of these:

[  116.050000] ata1.00: exception Emask 0x0 SAct 0x7fffffff SErr 0x0
action 0x6 frozen
[  116.060000] ata1.00: failed command: WRITE FPDMA QUEUED
[  116.070000] ata1.00: cmd 61/08:00:b0:1e:e4/00:00:06:00:00/40 tag 0
ncq 4096 out
[  116.070000]          res 40/00:01:00:00:00/00:00:00:00:00/00 Emask
0x4 (timeout)
[  116.080000] ata1.00: status: { DRDY }

...followed by:

[  116.890000] ata1.00: status: { DRDY }
[  116.900000] ata1: hard resetting link
[  117.370000] ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[  117.700000] ata1.00: configured for UDMA/100
[  117.710000] ata1.00: device reported invalid CHS sector 0
[  117.720000] ata1.00: device reported invalid CHS sector 0
...repeated for several seconds...
[  118.020000] ata1: EH complete

And then it was fine. And a bit later I got:

[  368.570000] ata1.00: exception Emask 0x12 SAct 0x2 SErr 0x1280500
action 0x6 frozen
[  368.580000] ata1.00: irq_stat 0x08000000, interface fatal error
[  368.590000] ata1: SError: { UnrecovData Proto 10B8B BadCRC TrStaTrns }
[  368.600000] ata1.00: failed command: READ FPDMA QUEUED
[  368.610000] ata1.00: cmd 60/a8:08:08:6d:65/00:00:06:00:00/40 tag 1
ncq 86016 in
[  368.610000]          res 40/00:0c:08:6d:65/00:00:06:00:00/40 Emask
0x12 (ATA bus error)
[  368.620000] ata1.00: status: { DRDY }
[  368.630000] ata1: hard resetting link
[  369.170000] ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[  369.530000] ata1.00: configured for UDMA/100
[  369.540000] ata1: EH complete

This is all very similar to what you're getting.

This is with the v1 kernel.

Is this sort of thing normal with SATA? Errors aren't that important if
they're properly detected and handled, and it does seem to be
recovering. OTOH... mysterious file system corruption.

I've opened a bug for reference,
https://github.com/amery/linux-allwinner/issues/24 and copied your error
onto it --- hope you don't mind.

[...]
> I'll be out of town for a few days, but when I get back I'll try using
> a more robust power supply than what came with the Mele.  Looking at
> it, it does seem pretty small and maybe the SATA drive attached to the
> system is causing intermittent power issues.

My thoughts as well.

-- 
┌─── ｄｇ＠ｃｏｗｌａｒｋ．ｃｏｍ ───── http://www.cowlark.com ─────
│
│ "Never attribute to malice what can be adequately explained by
│ stupidity." --- Nick Diamos (Hanlon's Razor)

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 254 bytes
Desc: OpenPGP digital signature
Url : http://lists.phcomp.co.uk/pipermail/arm-netbook/attachments/20120521/ee226b62/attachment.bin