------- Comment From gbert...@br.ibm.com 2016-08-08 15:09 EDT------- (In reply to comment #13) > (In reply to comment #9) > > I Have Tested this with , Test Kernel available at > > http://people.canonical.com/~rtg/eeh-lp1602724/ . on Ubuntu 16.04.1 > > > > Test Kernel : > > root@everest-lp13-leaf:~# uname -a > > Linux everest-lp13-leaf 4.4.0-32-generic #51 SMP Tue Jul 19 21:41:04 UTC > > 2016 ppc64le ppc64le ppc64le GNU/Linux > > > > > > > > Nvme (Leaf) is getting recovered till 5 times on triggering the EEH, But > > "hitting a kernel crash" after on 6th time trigger of EEH. > > > > This is most likely fixed by > > http://lists.infradead.org/pipermail/linux-nvme/2016-August/005670.html > ("[PATCH v2] nvme: Suspend all queues before deletion") > > Which is not upstream yet. Once it gets accepted, it should be pushed to > Ubuntu via another bugzilla. When that happens, we'll need a new test > kernel for this one.
Canonical, For a little more context, we have identified an issue in the test kernel you provided. After a sequence of 6 EEHs, DD will attempt to remove the adapter, which ends up hitting a BUG_ON. We think the above patch is a fix, but it's still not confirmed. Can you provide a kernel with that patch also applied for verification? It's still not upstream yet, but it has already been ack-ed by the driver maintainer, Keith Busch. -- 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/1602724 Title: Ubuntu 16.04 - Full EEH Recovery Support for NVMe devices Status in linux package in Ubuntu: Fix Released Status in linux source package in Xenial: Fix Committed Bug description: == Comment: #0 - Heitor Ricardo Alves de Siqueira <heit...@br.ibm.com> - 2016-07-12 12:54:27 == Current nvme driver in Ubuntu 16.04 kernel does not handle error recovery; we are missing some patches from the upstream nvme driver. We would like to ask Canonical to cherry pick the following patches for the 16.04 kernel, if possible: * 9396dec916c0 ("nvme: use a work item to submit async event requests") * 79f2b358c9ba ("nvme: don't poll the CQ from the kthread") * 2d55cd5f511d ("nvme: replace the kthread with a per-device watchdog timer") * 9bf2b972afea ("NVMe: Fix reset/remove race") * c875a7093f04 ("nvme: Avoid reset work on watchdog timer function during error recovery") * a5229050b69c ("NVMe: Always use MSI/MSI-x interrupts") To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1602724/+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
