Re: [Bug 1738224] Re: Suspend broken after kernel upgrade

2017-12-20 Thread Mario Limonciello
Information isn't available on what has changed between NVM releases from Intel. It generally improves performance and resolves issues with the ICM stack. Those could lead to for example the controller not going to sleep when it should. On Wed, Dec 20, 2017, 11:26 Jeffrey Bouter <1738...@bugs.laun

Re: [Bug 1738224] Re: Suspend broken after kernel upgrade

2017-12-19 Thread Mario Limonciello
Thunderbolt NVM isn't on LVFS stable but it's on testing. Either turn on fwupd LVFS testing remote or install manually from this file: https://fwupd.org/lvfs/device/dfd51125-338f-56ff-b721-fa3bea8e534e On Tue, Dec 19, 2017, 02:45 Kai-Heng Feng wrote: > I can't reproduce the issue on the XPS 936

Re: [Bug 1738224] Re: Suspend broken after kernel upgrade

2017-12-16 Thread Mario Limonciello
Did you manually set s2idle? That policy shouldn't have been picked unless you were on 4.14. On Sat, Dec 16, 2017, 17:55 Jeffrey Bouter <1738...@bugs.launchpad.net> wrote: > Kai-Heng, > > A cat on /sys/power/mem_sleep already shows: > > [s2idle] deep > > -- > You received this bug notification be

Re: [Bug 1738224] Re: Suspend broken after kernel upgrade

2017-12-16 Thread Mario Limonciello
That particular commit actually breaks proper s2idle. Because of it S3 should be used instead on 9360. It was intentionally broken because of some reported regression by one guy with an nvme SSD of a particular vendor. Can you please check which SSD you have? Nvme info command on it can be helpfu