> -----Original Message----- > From: David Brownell [mailto:[EMAIL PROTECTED]] > Sent: Tuesday, February 04, 2003 8:14 PM > To: Christopher Li > Cc: '[EMAIL PROTECTED]' > Subject: Re: [linux-usb-devel] usbdevfs setup URB buffer size limit > > > Christopher Li wrote: > > > > So usbdevfs will refuse to serve the setup packet bigger > than one page. > > > > Is there a reason why is that? I am asking because some > user get some > > devices try to send setup packet bigger than one page. VMware can't > > pass it to linux kernel using usbdevfs. And I can't find this limit > > on the usb spec either. > > Nobody has cared before ... :) > > The main reason I know to keep any limit around is an implementation > limitation of the OHCI driver: it doesn't try to handle more than 4KB > right in its data stages.
Then the right things seems to fix the OHCI driver and increase the buffer size limit to 8K or 64K. > > Patches to remove that limit (2.5 and 2.4), or notice when alignment > allows up to 8K (no updates to TD queueing code), would be > good to have. Let me see what I can do with the ohci driver. If it is not easy to fix, we can make ohci return nice error code. > Certainly you should be able to actually reach that 4KB limit, if > not get to 64KB. The 2.4 OHCI code doesn't report errors if you > hit the "doesn't work" cases, so it would be safest to just make > usbfs work correctly for the up-to-4KB case. Up-to-4K is not good enough. for that scanner, the data stage is 4K, the setup header is 8 byte. So it is 4K+8 bytes total. We definitely need 2 pages (on x86). Chris ------------------------------------------------------- This SF.NET email is sponsored by: SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See! http://www.vasoftware.com _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
