This message is from the T13 list server.
Hi Hale,
I can address your second concern. The access of data on an NV
Cache drive is no different than the accessing of the data of a non NV
Cache drive. There is no 'read from NV Cache' command. There is only
the normal read commands. This was done because it isn't reasonable to
expect BIOSes to support the NV Cache feature set. DOS or Ecos or Linux
accesses will be no different than the BIOS accesses on a system with an
OS that supports NV Cache.
The data integrity of the drive is not changed from when the way
it was before. The last thing written to a sector must be what you get
back when you read from the sector. The read operation makes no
preference as to where the data is read back from. If the data read
back is incorrect, then there is a flaw in the drive that is independent
of the host accessing that data.
Nathan Obr
ATA driver developer
Microsoft Corporation
-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Hale
Landis
Sent: Wednesday, July 12, 2006 2:11 PM
To: [email protected]
Subject: RE: [t13] READ VERIFY SECTOR(S) (EXT) and NV cache command set
This message is from the T13 list server.
Joseph Chen said:
> When the data had been placed on two locations we face issue
> of the consistency.
1) Aren't there really three locations? The NV cache, the normal drive
cache/buffer and the media?
2) When is someone going to explain what happens if a drive that has an
enabled NV cache with pinned data and it is accessed by a OS that
doesn't
know anything about NV cache? For example: If I boot one of those 'virus
scan CDs' that really runs DOS or Ecos or Linux or some other OS... What
happens if the virus scanner cleans a virus from sectors that are in the
NV cache? Will those sectors also be updated on the media?
3) Will drives have a jumper that will enable/disable this extreme
threat
to data integrity?
Hale
--
Hale Landis -- [EMAIL PROTECTED]
