On Tue, Jan 04, 2011 at 11:37:40PM -0800, Sarah Sharp wrote: > Hi Greg, > > Here's some xHCI patches for 2.6.38, including the late bug fixes for > suspend and resume (which are marked for stable). > > There's also some patches from Andiry to fix some isochronous transfer > bugs, ring ownership issues, and some cleanup patches. Matthew Wilcox > also submitted a patch to shave off another register read on the URB > submission path, so that should give us an additional performance boost. > I've stress-tested this batch, and they look stable. > > Full patches will be sent to the linux-usb mailing list. > > Sarah Sharp > > > The following changes since commit 2af10844eb6ed104f9505bf3a7ba3ceb02264f31: > Greg Kroah-Hartman (1): > USB: Merge 2.6.37-rc5 into usb-next > > are available in the git repository at: > > git+ssh://master.kernel.org/pub/scm/linux/kernel/git/sarah/xhci.git for-usb-next > > Andiry Xu (6): > xHCI: synchronize irq in xhci_suspend() > xHCI: fix queue_trb in isoc transfer > xHCI: remove redundant parameter in giveback_first_trb() > xHCI: fix cycle bit set in giveback_first_trb() > xHCI: replace dev_dbg() with xhci_dbg() > xHCI: fix printk_ratelimit() usage > > Matthew Wilcox (1): > xhci: Remove more doorbell-related reads > > Sarah Sharp (3): > xhci: Resume bus on any port status change. > usb: Realloc xHCI structures after a hub is verified. > xhci: Use GFP_NOIO during device reset. > > Zhang Rui (1): > xhci: Do not run xhci_cleanup_msix with irq disabled > > drivers/usb/core/hcd-pci.c | 7 +++- > drivers/usb/core/hub.c | 19 +++++++++ > drivers/usb/host/xhci-ring.c | 91 +++++++++++++++++++++++------------------ > drivers/usb/host/xhci.c | 60 +++++++++++---------------- > drivers/usb/host/xhci.h | 16 +++----- > include/linux/usb/hcd.h | 1 + > 6 files changed, 108 insertions(+), 86 deletions(-) Something's really wrong here, I get a much different diffstat here, one that includes lots of non-usb patches (like the 2.6.37-rc7 release from Linus). Are you sure you generated this correctly? Care to retry it in a format that doesn't also merge with Linus's tree? Also, as these are a bit late for the -rc1 merge (i.e. they needed to be in linux-next a few weeks ago), can they wait for after the -rc1 merge because they really are just bugfixes? That way we can test them better and I don't have to worry about dealing with untested code sent to Linus at the last minute... thanks, greg k-h -- 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