This message is from the T13 list server.
Andre, I don't agree with you. A SATA device agregates the link/phy errors into the task file error register and set the status error bit. If a host wants to know more about the nature of the link errors, then the phy registers would be nice, but in no way required. So I say "Whaaaay Jack"! MKE. -----Original Message----- From: Andre Hedrick [mailto:[EMAIL PROTECTED]] Sent: Friday, July 26, 2002 5:40 PM To: Gana Pat Cc: [EMAIL PROTECTED] Subject: Re: [t13] SATA This message is from the T13 list server. Well first off the most basic kludges to make SATA work use an add-a-chip to PATA. Since SATA 1.0 specifies and requires the issue of SATA/PHY MIB registers, the kludges fail to properly support the transport layer. Now this means, the HBA has no clue how to test if the device interrupt or if the PHY interrupt was generated and any recovery path for the HOST driver is not possible! This really stinks. Now the complete proper solution HBA can still suffer from the brain death of the add-a-chip, if they do not have proper HOST drivers's to use the extented SATA registers. Yet if the HOST's driver is that broken, then it deserves to die and lockup. So have fun playing with the new toys and remember that intq is not just a device anymore. So in short, the answer is YES. SATA 1.0 is/should be compatable with PATA. The long version is no way jack :-(. Cheers, Andre Hedrick LAD Storage Consulting Group On Thu, 25 Jul 2002, Gana Pat wrote: > Hi , > Can I discuss regarding Serial ATA here ? > > If yes here goes my questions. > > I heard that the SATA is software compatible with Parallel ATA. But how far it holds good in terms of devices per channel (port) ? I heard that they are also working on the Hubs with which you can connect more devices per channel. Is this true ? > > Anyone knows any info on this ? > > Your help is highly appreciated. > > Thanks, > Gana >
