On Thu, Jan 29, 2009 at 12:57:13PM -0500, Alan Stern wrote: > On Thu, 29 Jan 2009, David Vrabel wrote: > > > Alan Stern wrote: > > > > > > Pushing scatterlists down to the HCD level would involve several > > > nontrivial changes to the code. The issue of the buffer lengths is > > > just one of them, although it is perhaps the most vexing. > > > > I don't see where the complexity is. The HCD simply fills the > > hardware's s-g list structures based on the s-g list in the URB. > > I didn't say it was complex; I said it was nontrivial. That's because > the existing HCD code is already complicated and difficult to decipher. > > > > Obviously, none of the existing HCDs can support s-g lists in this > > manner and should continue to only accept URBs without s-g lists. > > It's a little difficult to comprehend the intent behind this statement. > Are you saying that if somebody does implement s-g lists for URBs, none > of the existing HCDs should be changed to accept them? In that case, > what use would they be? I think what he's saying (at least, what I read it to mean) was this: Existing HCDs would use the existing mechanism for s-g lists. The new HCD to support USB 3.0, which (apparently) has more capability, would process the buffer list directly. The usb_sg_* code would need to be changed to recognize a more capable HCD and pass the s-g list to that HCD for "direct" processing. Matt -- Matthew Dharm Home: mdharm-usb@xxxxxxxxxxxxxxxxxx Maintainer, Linux USB Mass Storage Driver A female who groks UNIX? My universe is collapsing. -- Mike User Friendly, 10/11/1998
Attachment:
pgpS8iUQKjv4e.pgp
Description: PGP signature
