I believe the source of this Western Digital thing is Andre Hedric's
discovery:
http://www.uwsg.indiana.edu/hypermail/linux/kernel/0001.3/1207.html
"WDC drives blow off the CRC check of UDMA.........This is BAD and STUPID.
Several of the OEM chipset makers have allowed this crap to exist.
ATA-2 (style) can not handle ATA-3/4 transfer rates without the CRC
checks, you end up continuing the DMA writing regardless if you lost data
that would have been saved if the UDMA CRC was intact.
"This is a pure hardware issue........I do not know yet if there is a way
to check/recover with out death to DMAing for all WDC products.
The best case at the moment is to consider blocking all WDC products from
using the ATA-3 and newere standard."
-pete
On Thu, 26 Apr 2001, Jose M. Sanchez wrote:
>
>
> -----Original Message-----
> From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of
> Todd Lyons
> Sent: Wednesday, April 25, 2001 4:24 AM
> To: [EMAIL PROTECTED]
> Subject: Offlist: [expert] 8.0 final --brakes MANY applications
> (Software Installeris first on that list)
>
>
> "Jose M. Sanchez" wrote:
>
> > This was not a Windows/Linux issue at all. You are asserting that somehow
> > the drives depend upon Windows for proper functioning. They do not.
>
> They depend on drivers being modified for the functionality that was
> removed from firmware.
>
> ---
> Firmware in the DRIVE or in the BIOS?
>
> If Bios this is a non-sequitor...
>
> If it's drive firmware you are talking about (I.E. actual firmware
> controlling drive electronics "below" the IDE interface level)to date this
> has -NOT- been done.
>
> The OS does not communicate with the controller/drive mechanism at a low
> enough level for this to be doable.
>
> While this may change in the future, doing so will require revamping the
> entire IDE design which purposely isolates the user/OS from controlling
> anything lower than sector structure access...
>
> ---
>
> This methodology prompted people to call software modems winmodems.
>
> ---
>
> Yes with all the associated horrific...
>
> ---
>
>
> They are beginning to apply this same methodology to drives, but in much
> smaller proportions.
>
> ---
>
> Not so.
>
> There is a vast distinction between drive IDE "firmware" updates, versus
> driver replacement of actual hardware functions as in Winmodems.
>
> The former may mean some addition difficulty for Linux programmers trying to
> wring the best performance out of a device, but there is no inherent
> dependence upon the OS itself.
>
> The latter represents a complete departure from the IDE/ATA standard,
> effectively creating a new class of "non compatible" (or WinCompatible if
> you talk to Bill G.) devices...
>
> In turn this is a VERY bad move for companies such as WD and others. IDE
> drives are being used in everything from imbedded systems to MP3 players
> (something which I've been recently playing with from a design level).
>
> WD and others do not (yet) produce a "WinDrive" as mis-stated in this
> discussion nor are they even close to this...
>
> ---
>
> I don't think it's fair to call them WinDrives either, but if you look at
> WD's
> website, it does specifically state for Microsoft Windows products. It
> does not exclude others, but it does not support nor offer to support
> them (them being us).
>
> ---
>
> This is more evasion than design. WD's drives have no more or less reliance
> on the OS than any other manufacturer's devices...
>
> It is more likely that this is merely a ploy to avoid fielding questions
> from non MS$ shops...
>
> ---
>
> All in all, it looks like WD has placed their bet that M$ will be the
> 800 lb gorilla forever and a day. We aim to disappoint them.
>
> ---
>
> There is no evidence of this (yet). Just speculation on the part of some
> people that one thing means another.
>
> ---
>
> Civilime is the one who can best explain it. Unfortunately, he won't be
> back until next week.
>
> ---
>
> Civilime began this several months ago when he complained of trouble with
> WD's when mixed with other types of drives.
>
> This had nothing to do with OS dependence vis-a-vis Linux... in his original
> postings he correctly stated that this was more an issue with controller
> timings with certain integrated controllers used on some motherboards. Linux
> had a tendency to push the drive chain differently causing problems...
> though this was not a WD or Linux problem per se... more of a design flaw in
> the integrated IDE controller configuration used by some manufacturers.
>
> Pushing the bus speeds, as done by overclockers, tended to exacerbate the
> problem... as did other slower devices on the same IDE chain (such as
> CD-ROM's, etc.).
>
> For some reason the original discussion has degenerated into "WD is
> producing WinDrives.." being a profound truth...
>
> If this were so, I'd be an outspoken critic... however given that I've been
> playing with their specs due to a project I'm working on, it's obvious to me
> that these assertions are erroneous...
>
> To further cloud the issue (as postings to boards often do) one person
> offers up how putting in a Seagate onto a SCSI chain resolved his problems
> so obviously it's a Windrive issue...
>
> ... eh, no.
>
> Still I like the discussion...
>
> -JMS
> [EMAIL PROTECTED]
>
>
>
