Public bug reported: The autonegotiation fails sporadically for various Lenovo notebooks (e.g. T450s, T550, W540) when resuming inside the docking station. The speed ethtool reports is 10Mb/s, full duplex, even though that is not the speed the connected switch uses. So the systems are unable to send/receive any packages.
Issuing the command "ethtool -r eth0" does not lead to a successful renegotiation. To get a correct link, I have to either undock the notebook for about 15 seconds, and then re-dock it, or disconnect the LAN cable for about 15 seconds, and then reconnect the cable. In my case the problem appears about one time in 10-20 suspend/resume cycles. I found the problem in the 4.15 Ubuntu Bionic kernel, but the problem is still present in kernel 5.0-rc8. I've done a git bisect and the following commit seems to have introduced the problem: commit 04e04009bb01502f8023c707d7569ea42fba903b Author: Rafael J. Wysocki <rafael.j.wyso...@intel.com> Date: Fri Feb 9 22:55:28 2018 +0100 ACPI / EC: Restore polling during noirq suspend/resume phases BugLink: http://bugs.launchpad.net/bugs/1774063 [ Upstream commit 3cd091a773936c54344a519f7ee1379ccb620bee ] Commit 662591461c4b (ACPI / EC: Drop EC noirq hooks to fix a regression) modified the ACPI EC driver so that it doesn't switch over to busy polling mode during noirq stages of system suspend and resume in an attempt to fix an issue resulting from that behavior. However, that modification introduced a system resume regression on Thinkpad X240, so make the EC driver switch over to the polling mode during noirq stages of system suspend and resume again, which effectively reverts the problematic commit. Fixes: 662591461c4b (ACPI / EC: Drop EC noirq hooks to fix a regression) Link: https://bugzilla.kernel.org/show_bug.cgi?id=197863 Reported-by: Markus Demleitner <m...@tfiu.de> Tested-by: Markus Demleitner <m...@tfiu.de> Signed-off-by: Rafael J. Wysocki <rafael.j.wyso...@intel.com> Signed-off-by: Sasha Levin <alexander.le...@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gre...@linuxfoundation.org> Signed-off-by: Kamal Mostafa <ka...@canonical.com> Signed-off-by: Khalid Elmously <khalid.elmou...@canonical.com> I have reverted the commit in the Ubuntu Bionic kernel, and cannot reproduce the problem any longer. I have also opened a corresponding upstream bugreport on bugzilla.kernel.org: https://bugzilla.kernel.org/show_bug.cgi?id=202675 ** Affects: linux (Ubuntu) Importance: Undecided Status: New ** Description changed: The autonegotiation fails sporadically for various Lenovo notebooks (e.g. T450s, T550, W540) when resuming inside the docking station. The speed ethtool reports is 10Mb/s, full duplex, even though that is not the speed the connected switch uses. So the systems are unable to send/receive any packages. Issuing the command "ethtool -r eth0" does not lead to a successful renegotiation. To get a correct link, I have to either undock the notebook for about 15 seconds, and then re-dock it, or disconnect the LAN cable for about 15 seconds, and then reconnect the cable. In my case the problem appears about one time in 10-20 suspend/resume cycles. I found the problem in the 4.15 Ubuntu Bionic kernel, but the problem is still present in kernel 5.0-rc8. - - I've done a git bisect and the following commit seems to have introduced the problem: + I've done a git bisect and the following commit seems to have introduced + the problem: commit 04e04009bb01502f8023c707d7569ea42fba903b Author: Rafael J. Wysocki <rafael.j.wyso...@intel.com> Date: Fri Feb 9 22:55:28 2018 +0100 - ACPI / EC: Restore polling during noirq suspend/resume phases + ACPI / EC: Restore polling during noirq suspend/resume phases - BugLink: http://bugs.launchpad.net/bugs/1774063 + BugLink: http://bugs.launchpad.net/bugs/1774063 - [ Upstream commit 3cd091a773936c54344a519f7ee1379ccb620bee ] + [ Upstream commit 3cd091a773936c54344a519f7ee1379ccb620bee ] - Commit 662591461c4b (ACPI / EC: Drop EC noirq hooks to fix a - regression) modified the ACPI EC driver so that it doesn't switch - over to busy polling mode during noirq stages of system suspend and - resume in an attempt to fix an issue resulting from that behavior. + Commit 662591461c4b (ACPI / EC: Drop EC noirq hooks to fix a + regression) modified the ACPI EC driver so that it doesn't switch + over to busy polling mode during noirq stages of system suspend and + resume in an attempt to fix an issue resulting from that behavior. - However, that modification introduced a system resume regression on - Thinkpad X240, so make the EC driver switch over to the polling mode - during noirq stages of system suspend and resume again, which - effectively reverts the problematic commit. + However, that modification introduced a system resume regression on + Thinkpad X240, so make the EC driver switch over to the polling mode + during noirq stages of system suspend and resume again, which + effectively reverts the problematic commit. - Fixes: 662591461c4b (ACPI / EC: Drop EC noirq hooks to fix a regression) - Link: https://bugzilla.kernel.org/show_bug.cgi?id=197863 - Reported-by: Markus Demleitner <m...@tfiu.de> - Tested-by: Markus Demleitner <m...@tfiu.de> - Signed-off-by: Rafael J. Wysocki <rafael.j.wyso...@intel.com> - Signed-off-by: Sasha Levin <alexander.le...@microsoft.com> - Signed-off-by: Greg Kroah-Hartman <gre...@linuxfoundation.org> - Signed-off-by: Kamal Mostafa <ka...@canonical.com> - Signed-off-by: Khalid Elmously <khalid.elmou...@canonical.com> + Fixes: 662591461c4b (ACPI / EC: Drop EC noirq hooks to fix a regression) + Link: https://bugzilla.kernel.org/show_bug.cgi?id=197863 + Reported-by: Markus Demleitner <m...@tfiu.de> + Tested-by: Markus Demleitner <m...@tfiu.de> + Signed-off-by: Rafael J. Wysocki <rafael.j.wyso...@intel.com> + Signed-off-by: Sasha Levin <alexander.le...@microsoft.com> + Signed-off-by: Greg Kroah-Hartman <gre...@linuxfoundation.org> + Signed-off-by: Kamal Mostafa <ka...@canonical.com> + Signed-off-by: Khalid Elmously <khalid.elmou...@canonical.com> I have reverted the commit in the Ubuntu Bionic kernel, and cannot reproduce the problem any longer. + + + I have also opened a corresponding upstream bugreport on bugzilla.kernel.org: https://bugzilla.kernel.org/show_bug.cgi?id=202675 -- 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/1817562 Title: Autonegotiation fails for various Lenovo notebooks when resuming inside the docking station Status in linux package in Ubuntu: New Bug description: The autonegotiation fails sporadically for various Lenovo notebooks (e.g. T450s, T550, W540) when resuming inside the docking station. The speed ethtool reports is 10Mb/s, full duplex, even though that is not the speed the connected switch uses. So the systems are unable to send/receive any packages. Issuing the command "ethtool -r eth0" does not lead to a successful renegotiation. To get a correct link, I have to either undock the notebook for about 15 seconds, and then re-dock it, or disconnect the LAN cable for about 15 seconds, and then reconnect the cable. In my case the problem appears about one time in 10-20 suspend/resume cycles. I found the problem in the 4.15 Ubuntu Bionic kernel, but the problem is still present in kernel 5.0-rc8. I've done a git bisect and the following commit seems to have introduced the problem: commit 04e04009bb01502f8023c707d7569ea42fba903b Author: Rafael J. Wysocki <rafael.j.wyso...@intel.com> Date: Fri Feb 9 22:55:28 2018 +0100 ACPI / EC: Restore polling during noirq suspend/resume phases BugLink: http://bugs.launchpad.net/bugs/1774063 [ Upstream commit 3cd091a773936c54344a519f7ee1379ccb620bee ] Commit 662591461c4b (ACPI / EC: Drop EC noirq hooks to fix a regression) modified the ACPI EC driver so that it doesn't switch over to busy polling mode during noirq stages of system suspend and resume in an attempt to fix an issue resulting from that behavior. However, that modification introduced a system resume regression on Thinkpad X240, so make the EC driver switch over to the polling mode during noirq stages of system suspend and resume again, which effectively reverts the problematic commit. Fixes: 662591461c4b (ACPI / EC: Drop EC noirq hooks to fix a regression) Link: https://bugzilla.kernel.org/show_bug.cgi?id=197863 Reported-by: Markus Demleitner <m...@tfiu.de> Tested-by: Markus Demleitner <m...@tfiu.de> Signed-off-by: Rafael J. Wysocki <rafael.j.wyso...@intel.com> Signed-off-by: Sasha Levin <alexander.le...@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gre...@linuxfoundation.org> Signed-off-by: Kamal Mostafa <ka...@canonical.com> Signed-off-by: Khalid Elmously <khalid.elmou...@canonical.com> I have reverted the commit in the Ubuntu Bionic kernel, and cannot reproduce the problem any longer. I have also opened a corresponding upstream bugreport on bugzilla.kernel.org: https://bugzilla.kernel.org/show_bug.cgi?id=202675 To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1817562/+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
