Re: [smartmontools-support]darwin version: checksum errors etc

[Bruce A. <ba...@gr...>]

On Fri, 10 Sep 2004, Cameron Hayne wrote:

> On 9/9/04 11:42 PM, "Bruce Allen" <ba...@gr...> wrote:
> 
> > I'm afraid that something has gotten screwed up with your disk's firmware.
> > It's list of SMART Attribute Thresholds contains a bunch of inconsistent
> > data (which is why you get a complaint about the checksum error for that
> > data structure).  It's possible that it was this way when the disk was
> > manufactured -- it's also (more likely) that this has gotten corrupted at
> > some later time.  In most disks this data is stored in some special area
> > of the disk surface which is reserved (users can't touch it).
> > 
> > As a result of this 'mess up' some of the SMART data is being hidden.  Try
> > doing the following modification to line 801 of ataprint.c, and
> > rebuilding/rerunning the code.  Change:
> >   if (disk->id && thre->id){
> > to
> >   if (disk->id){
> > and again send the output of '-a -r ioctl,2'.
> 
> I made the change tp ataprint.c suggested by Bruce,
> did 'make && make install'
> and ran 'smartctl -r ioctl,2 -a /dev/disk0s6
> I attach the results - they aren't too different.

They are: see lines starting Vendor Specific SMART Attributes with
Thresholds for hidden data.

Cameron, for the moment I suspect something wrong with your disk -- if
there is a way to run Toshiba's SMART utilities on it, that would be a
good idea, but I don't know if they distribute a utility that works with
PPC.  Geoff, alternatively, it IS possible that this is some strange OS
bug, and if so we'll hear more about it from other users in the future.

Cheers,
	Bruce