On Wed, Jun 05, 2002 at 07:11:23PM -0700, David Brownell wrote:
> >>But before I do that, is the gist of the patch OK?  I'm thinking
> >>specifically  the export of those symbols out of usb/core/hcd.c.
> 
> I think that's fine, and was going to merge that into a separate
> patch (you should see it on linux-usb-devel later tonight).  That
> was one of those issues I was waiting for someone who used non-PCI
> host controllers to address.
> 
> 
> >I would _much_ rather see the pci specific pieces of hcd.c split out
> >into a separate file (hcd-pci.c?)
> 
> Those are actually two separate issues.  The bus glue needs some
> symbols available, whether it's PCI, SA-1111 custom, or whatever.
> 
> The PCI glue was originally in a separate file in any case, only
> put into a largish "hcd.c" because that was the style established
> at that point for the usbcore code.

Sorry, I didn't mean to imply that I had any displeasure in the current
incantation of hcd.c.  It was a start in splitting up the monolithic
usb.c file into smaller pieces.  Pat took that one setp further and
divided it up even more (he also did the same for the pci code), and
after working with it a while, it makes sense.  We also now have a
separate subdirectory to play in, so it doesn't look so bad if we divide
things up further.

That being said, moving the pci specific parts of hcd.c into a different
file, removes some #ifdefs from the file, and we can just add the file
to the build based on the CONFIG_PCI variable.

Hm, maybe I'll try that out later tonight...

thanks,

greg k-h

_______________________________________________________________

Don't miss the 2002 Sprint PCS Application Developer's Conference
August 25-28 in Las Vegas -- http://devcon.sprintpcs.com/adp/index.cfm

_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

Reply via email to