** Description changed: + == SRU REQUEST [Xenial, Yakkety, Zesty] == + Ubuntu 16.04.1 Kernel = 4.4.0-53-generic-74-Ubuntu ppc64le When running the stress-ng "fstat" stressor, it is trying to access the USB bus and giving a call trace and locking up any further USB activity (lsusb hangs). This only seems to occur so far on openpower(Firestone and Garrison) where there is a virtual USB keyboard and mouse built into the BMC. From lsusb(before crashing): Bus 001 Device 004: ID 046b:ff10 American Megatrends, Inc. Virtual Keyboard and Mouse Another openpower server(Briggs) has no virtual usb devices and does not experience the failure. Please see attached kern.log and dmesg output for further details. + + == Fix == + + Quirking the Virtual AMI keyboard and mouse with ALWAYS_POLL addresses + the issue. The patch has been accepted into the upstream queue for + 4.11. + + == Test Case == + + run 10 times: + sudo stress-ng --fstat 128 -t 60 -v + + Without the fix, it will hang, with the fix there is no hang or USB + error messages. + + == Regression Potential == + + This only quirks a specific AMI virtual keyboard and mouse into a poll + mode, so it touches one device. Futhermore, the poll mode shouldn't + affect operation; it just makes the URB handling less efficient.
** Description changed: == SRU REQUEST [Xenial, Yakkety, Zesty] == Ubuntu 16.04.1 Kernel = 4.4.0-53-generic-74-Ubuntu ppc64le When running the stress-ng "fstat" stressor, it is trying to access the USB bus and giving a call trace and locking up any further USB activity (lsusb hangs). This only seems to occur so far on openpower(Firestone and Garrison) where there is a virtual USB keyboard and mouse built into the BMC. From lsusb(before crashing): Bus 001 Device 004: ID 046b:ff10 American Megatrends, Inc. Virtual Keyboard and Mouse Another openpower server(Briggs) has no virtual usb devices and does not experience the failure. Please see attached kern.log and dmesg output for further details. == Fix == Quirking the Virtual AMI keyboard and mouse with ALWAYS_POLL addresses the issue. The patch has been accepted into the upstream queue for - 4.11. + 4.11, see http://www.spinics.net/lists/linux-usb/msg152977.html == Test Case == run 10 times: sudo stress-ng --fstat 128 -t 60 -v Without the fix, it will hang, with the fix there is no hang or USB error messages. == Regression Potential == This only quirks a specific AMI virtual keyboard and mouse into a poll mode, so it touches one device. Futhermore, the poll mode shouldn't affect operation; it just makes the URB handling less efficient. -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to linux in Ubuntu. https://bugs.launchpad.net/bugs/1652132 Title: Call trace when testing fstat stressor on ppc64el with virtual keyboard and mouse present Status in Linux: Unknown Status in linux package in Ubuntu: In Progress Bug description: == SRU REQUEST [Xenial, Yakkety, Zesty] == Ubuntu 16.04.1 Kernel = 4.4.0-53-generic-74-Ubuntu ppc64le When running the stress-ng "fstat" stressor, it is trying to access the USB bus and giving a call trace and locking up any further USB activity (lsusb hangs). This only seems to occur so far on openpower(Firestone and Garrison) where there is a virtual USB keyboard and mouse built into the BMC. From lsusb(before crashing): Bus 001 Device 004: ID 046b:ff10 American Megatrends, Inc. Virtual Keyboard and Mouse Another openpower server(Briggs) has no virtual usb devices and does not experience the failure. Please see attached kern.log and dmesg output for further details. == Fix == Quirking the Virtual AMI keyboard and mouse with ALWAYS_POLL addresses the issue. The patch has been accepted into the upstream queue for 4.11, see http://www.spinics.net/lists/linux-usb/msg152977.html == Test Case == run 10 times: sudo stress-ng --fstat 128 -t 60 -v Without the fix, it will hang, with the fix there is no hang or USB error messages. == Regression Potential == This only quirks a specific AMI virtual keyboard and mouse into a poll mode, so it touches one device. Futhermore, the poll mode shouldn't affect operation; it just makes the URB handling less efficient. To manage notifications about this bug go to: https://bugs.launchpad.net/linux/+bug/1652132/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp
