ad10: WARNING - READ_DMA UDMA ICRC error (retrying request) LBA=11441599

[Chuck Swiger]

Karl Denninger wrote:
> On Thu, Aug 11, 2005 at 12:46:04AM +0200, S?ren Schmidt wrote:
[ ... ]
>>I've already gone WAY out of my way to try to support the sii3112,  
>>and I'm not inclined to waste more of my precious spare time on it.  
>>However, if it really is that important to enough people to try to  
>>workaround the silicon bugs (which very likely isn't possible), get  
>>together and get me failing HW on my desk and time to work on it.
> 
> Ok, then do the RIGHT THING and document that the SiI chips are declared
> BROKEN by FreeBSD and likely to cause people trouble - including irrevocable 
> data corruption.
> 
> This would have saved me COUNTLESS hours when I first ran into this 
> issue.  Indeed, it was not until someone else started posting excerpts 
> from commit logs (months after I filed the PR originally!) that I was 
> aware FreeBSD developers considered these chipsets "damaged goods".

Look, Karl, we're all as sorry as we can be that you've spent lots of time on 
this issue and/or you've had data get corrupted.  You should not rely on that 
sympathy to be endless.

FreeBSD attempts to document that it works with common hardware which follows 
industry standards and is not otherwise broken.  The information available to 
me suggests the SiI 3112 is broken.  It has multiple hardware defects involving 
ATA request-size handling (SIIBUG in ata_sii_allocate() in 
dev/ata/ata-chipset.c around line ~2300, or what the Linux guys call 
SIL_QUIRK_MOD15WRITE), and with LBA48 if used with various Seagate drives.

I've also gotten the impression that the chipset is prone to locking up the 
entire system under high load, especially under RAID-1 mirroring or other 
parallel access cases, because it mishandles interrupts or some such.

Given that this is the case, I would be looking to get my money back or a 
replacement from the vendor who sold me this crappy hardware, far more than I 
would be looking towards implementing software workarounds which cripple the 
performance of the system in order to safely work around the hardware errata.

> Where is fair warning in the hardware compatability guide?

http://www.freebsd.org/platforms/amd64/motherboards.html ...?

> Second, your requirement for <BOTH> hardware <AND TIME> simply can't be 
> met.  It is not possible for anyone to manufacture or deliver time.  

"Donne-moi, monsieur, tout mais les temps...?"  :-)

> Is it thus necessary for us "mere users" to consider this an issue that 
> will simply not be addressed?  If so, then just say so up front <AND 
> DOCUMENT THAT THE SII CHIPSETS DON'T WORK RIGHT.>
[ ...some 200 (!) lines of ranting at poor Soren deleted :-)... ]

I haven't seen a diatribe like this one since those green JPCON capacitors 
which leaked and fried motherboards everywhere, or maybe even since those old 
KT266 motherboards, which were supposed to do 4x AGP but would lock up hard 
when some slavering gamer tried to make his $500 new 4x AGP card go.

Anyway, yeah, you got it: some SII chipsets don't work right.

FreeBSD tries to compensate; for some people it works OK for what they are 
doing, and for others it doesn't.  Blow $25 and get a cheap 4-port SATA-150 
RAID card using something other than a SiI 3112.  Blow $50 and you can even get 
one from a vendor like Promise or Highpoint that's at least somewhat reputable, 
and/or provides open source drivers and FreeBSD support for their products.

If it makes you feel better, submit this as a PR against the docs category:

--- ata.4~      Tue Apr  5 14:28:00 2005
+++ ata.4       Wed Aug 10 21:43:05 2005
@@ -129,7 +129,7 @@
  .It ServerWorks:
  ROSB4, CSB5, CSB6.
  .It Silicon Image:
-SiI0680, SiI3112, SiI3114, SiI3512.
+SiI0680, SiI3114, SiI3512.  SiI3112 has hardware errata and may not work.
  .It SiS:

-- 
-Chuck