On Tue, May 11, 2010 at 11:05:52AM +0800, Xu, Andiry wrote: > Sarah, > > I made the patches based on your last master branch - commit 15ba9138. > And I've verified it can also apply to your latest master branch, commit > 8e6632cd. I don't know what is wrong, but I'll give a check and resend > the patches. What do you use to apply the patches? I end up using `git am --3way`, but it complains about not being able to find objects and gives up. It's possible my incoming mail server mangled the patches somehow, so please resend. :( Sarah Sharp > > -----Original Message----- > > From: Sarah Sharp [mailto:sarah.a.sharp@xxxxxxxxxxxxxxx] > > Sent: Tuesday, May 11, 2010 3:32 AM > > To: Xu, Andiry > > Cc: linux-usb@xxxxxxxxxxxxxxx; gregkh@xxxxxxx; Yang, Libin > > Subject: Re: [PATCH 0/3 v5] ISOC supporting for xHCI > > > > Andiry, > > > > What tree does this apply against? It's obvious you applied it > against > > a tree with streams support, but the first patch is failing to apply > > against all the trees I have tried. I have tried my master branch and > > various commits before the master branch HEAD, up to commit > > 15ba9138ee41a0. I've also tried applying the patches against > linux-next > > snapshots 20100506 and 20100507, and the first patch doesn't apply > there > > either. > > > > Sarah Sharp > > > > On Fri, May 07, 2010 at 04:54:37AM +0800, Andiry Xu wrote: > > > Hi, > > > > > > Here is v5 of isoc supporting for xHCI patchset. > > > > > > Change log from v4: > > > 1. Add initialization of variable len in handle_tx_event() to get > rid of > > > the compile warning. > > > > > > 2. Add bool variable skip for struct xhci_virt_ep. It's used to deal > > > with the Missed Service Error Event for isochronous endpoint. > > > Sometimes the xHC is unable to process the isoc TDs in time, and it > will > > > generate a Missed Service Error Event. In this case some TDs on the > ring > > > are not processed and missed. When encounter a Missed Servce Error > > > Event, set the skip flag of the endpoint, and process the missed TDs > > > until reach the next xHC processed TD, then clear the skip flag. > Missed > > > TDs are treated as short transfer. > > > > > > 3. Add judgment for event_trb in handle_tx_event(). No-op TRB should > not > > > trigger interrupts. If event_trb is a no-op trb, it refers to a > > > cancelled TD so just ignore it. > > > > > > 4. Modify the urb_dequeue process, only add unprocessed TDs of an > URB to > > > endpoint's cancelled TD list. > > > > > > Thanks, > > > Andiry Xu > > > > > -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
