Universal_binary.pdf doesn't have much to say about I/O.
Likely because it isn't affected much at all (unless you are a driver
developer).
I'm an application-level writer who performs CDB-level SCSI I/O to
parallel
SCSI devices, via ScsiTaskUserClient, using calls such as
IOCreatePlugInInterfaceForService and
(*scsiTaskDeviceInterface)->CreateSCSITask, and so forth.
All I really want to know for now is: in the brave new Intel world,
according
to the plan, to a first approximation, WHICH IS TRUE:
a) All that architecture is still present, it's just a recompile
and it should
all "just work," except of course that a lot of fine-grained
details that I
shouldn't be depending on but probably am may change
They will be running Mac OS X 10.4.x it will have IOKit and friends
just like the PPC version. Apple says they have been building and
using Mac OS X and related projects on Intel hardware for 5 years.
b) There will be a totally new SCSI API I will need to write to.
No it will be Mac OS X 10.4.x running on these system. Why would a
new API be needed?
c) Parallel SCSI won't be supported at all Intel Macs--not even on
Intel Macs
with PCI busses using HBA's Mac-savvy vendors like ATTO.
I expect the first systems to get Intel will be portables, Mac mini,
and iMac so in a way your statement is true in that those systems
don't have much of way to connect to parallel/LVD SCSI devices (at
least not easily). PowerMacs will change over a little later I
believe and those will likely start using PCIe (assume a PPC system
before that with PCIe... at least it stands to reason). You will need
PCIe adapters and drivers but nothing will prevent parallel SCSI from
existing on those systems.
-Shawn
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Ata-scsi-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden
