Re: [RFC 0/9] xHCI debugging removal.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Greg,

Is it too late for an xHCI 3.3 pull request?  This patchset is mostly
debugging cleanup, no new features, so should be harmless.  If you don't
mind the late request, please pull from my for-usb-next branch.

The HEAD commit ID is 01acaa8537bc5e8ccb3d0befa86e9c729391443b.

Sarah Sharp

On Fri, Dec 23, 2011 at 05:39:35PM +0800, Andiry Xu wrote:
> On 12/23/2011 08:24 AM, Sarah Sharp wrote:
> > It's cleanup time for the xHCI driver.
> > 
> > Several users (including Linus) have been concerned about the "WARN:
> > short packet" messages that are pretty much a part of any USB device
> > enumeration.  Other folks have been worried about the "failed to enable
> > MSI-X" message that appears on some xHCI hosts with MSI-only support.
> > 
> > Patches 2 and 3 downgrade these severity of these messages so they don't
> > appear unless CONFIG_USB_XHCI_HCD_DEBUGGING is turned on.  The xHCI
> > driver will happily carry on when they occur, so users shouldn't be
> > concerned about them.
> > 
> > Patches 3-8 remove a large chunk of the xHCI driver debugging.
> > Basically, the host controller will be silent during USB transfers, but
> > still print messages about xHCI command submissions completions,
> > cancellation handling, and some initialization/shutdown messages.
> > 
> > It's just too difficult to have debugging turned on without these
> > patches when trying to figure out issues with the ring expansion code.
> > Similarly, debugging UAS transfers is just a pain.
> > 
> > The last patch (#9) adds better debugging for when something goes
> > horribly wrong with the xHCI driver.
> > 
> > 
> > Andiry, please let me know if you're comfortable with the level of debug
> > after these patches are applied, since you're about the only other
> > person who uses CONFIG_USB_XHCI_HCD_DEBUGGING.  The patches on my
> > for-usb-next branch if you'd like to pull them that way.
> > 
> 
> That's good:) Thanks for the effort. The patches look OK to me.
> I also have some debug patches during development. Perhaps you can look
> at them and pick up any of them you want.
> 
> Thanks,
> Andiry
> 
> > 
> > The following changes since commit 1b41c8321e495337e877ca02d0b9680bc4112eff:
> > 
> >   usbfs: Fix oops related to user namespace conversion. (2011-12-22 14:07:09 -0800)
> > 
> > are available in the git repository at:
> >   git://git.kernel.org/pub/scm/linux/kernel/git/sarah/xhci.git for-usb-next
> > 
> > Sarah Sharp (9):
> >       Trivial: xhci: Fix copy-paste error.
> >       xhci: Remove scary warnings about transfer issues.
> >       xhci: Remove warnings about MSI and MSI-X capabilities.
> >       xhci: Remove useless sg-list debugging.
> >       xhci: Remove debugging for individual transfers.
> >       xhci: Remove debugging about toggling cycle bits.
> >       xhci: Remove debugging about ring structure allocation.
> >       xhci: Be less verbose during URB cancellation.
> >       xhci: Better debugging for critical host errors.
> > 
> >  drivers/usb/host/xhci-mem.c  |   14 ------
> >  drivers/usb/host/xhci-ring.c |   95 ++++++++++++------------------------------
> >  drivers/usb/host/xhci.c      |   27 +++++++-----
> >  drivers/usb/host/xhci.h      |    2 +-
> >  4 files changed, 43 insertions(+), 95 deletions(-)
> > 
> 
> 

Attachment: signature.asc
Description: Digital signature


[Index of Archives]     [Linux Media]     [Linux Input]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [Old Linux USB Devel Archive]

  Powered by Linux