Re: [TESTPATCH] xhci: usb2 hw lpm extra debug

2016-06-16 Thread Christian M . Amsüss
Hello Mathias, On Thu, Jun 16, 2016 at 01:26:25PM +0300, Mathias Nyman wrote: > This should be fixed with patch: > > commit 27a41a83ec54d0edfcaf079310244e7f013a7701 > xhci: Cleanup only when releasing primary hcd > > which is only in Gregs usb-linus branch, but should end up in Linus tree so

Re: [TESTPATCH] xhci: usb2 hw lpm extra debug

2016-06-15 Thread Christian M . Amsüss
On Wed, Jun 15, 2016 at 05:43:56PM +0200, chrysn wrote: > Reviewing the patches, that was because the invalid port_num is used in > `readl(pm_addr)` already, too early for your `portnum %d > > num_usb2_ports` debug line. I'm shoving the line up to make it out > before the segfault, and will report

Re: [TESTPATCH] xhci: usb2 hw lpm extra debug

2016-06-15 Thread Christian M . Amsüss
On Wed, Jun 15, 2016 at 05:32:11PM +0200, chrysn wrote: > I've only spotted one of your debug statements in the resulting dmesg > ("USB2 LPM capable set, port 4 speed 2"), which confuses me, given that > at least the `portnum %d > num_usb2_ports` line should have shown up > (the equivalent debug st

Re: [TESTPATCH] xhci: usb2 hw lpm extra debug

2016-06-15 Thread Christian M . Amsüss
Hi, On Wed, Jun 15, 2016 at 01:56:57PM +0300, Mathias Nyman wrote: > custom testpatch to show more info about touching usb2 hw lpm > setting for usb3 devices Applied, built, booted and hit the oops; most interestingly, the crashes these days don't get me into the kdump kernel anymore (might be a

Re: [PATCH] xhci: Detect out-of-bounds access to usb2_ports

2016-06-14 Thread Christian M . Amsüss
On Tue, Jun 14, 2016 at 06:19:58PM +0300, Mathias Nyman wrote: > If I write a more verbose debugging patch for this specific case could you > try it out? Sure. I wouldn't have known where to start looking, but if you have anything I can apply and boot into, I can give you a full dmesg output or wh

Re: [PATCH] xhci: Detect out-of-bounds access to usb2_ports

2016-06-14 Thread Christian M . Amsüss
Hello USB maintainers, On Wed, Mar 30, 2016 at 10:52:53AM +0200, Christian M. Amsüss wrote: > Out-of-bounds access was triggered on a Thinkpad Yoga during suspend or > by rmmod'ing the xhci-pci module, sometimes leading to a general > protection fault. the problem still exists in 4

[PATCH] xhci: Detect out-of-bounds access to usb2_ports

2016-03-30 Thread Christian M . Amsüss
even after the xhci struct has its num_usb2_ports reduced to zero, or maybe when the udev is assigned to the wrong xhci. Anyway, this keeps nasty crashes from happening in hard to debug situations (during suspend, not even kdump could provide an error message). Signed-off-by: Christian M. A