SATA write woes

Søren Schmidt sos at DeepCore.dk
Fri Feb 6 08:21:53 PST 2004


Jason Andresen wrote:
> Søren Schmidt wrote:
> 
>> Yes, this is a "feature" of the broken silicon in the Sii3112 chip...
> 
> 
> Is there a workaround?

Those timeouts and retries is the workaround, it took quite a bit of 
inginuity to not have it lockup the system solid each time it wedges...

If you saw the errata docs for this piece of crap you wouldn't trust 
your data to it :)

-- 
-Søren



More information about the freebsd-current mailing list