https://bugzilla.kernel.org/show_bug.cgi?id=95911 --- Comment #19 from gitne@xxxxxxxxxxxx --- Just tested kernel version 4.1.4, and I have also started playing with some of the radeon module parameters. It turns out that when I set radeon.msi=0 everything works as expected. No ring test failures anymore and hibernation works perfect too. However, when I set radeon.msi to 1 or -1 (auto) I get exactly the same behavior as described earlier. As it turns out the message signaled interrupts code in the radeon device driver module is broken. I have had similar behavior with the Radeon Mobility 2600 HD (R630). The difference was that hibernation worked but when radeon.msi was set to 1 or -1 (auto) the radeon device driver module caused initially seemingly random lockups, which over time, as interrupts got generated from input devices, the lockups have been released for a couple hundred milliseconds until another interrupt had to be handled by the radeon device driver module. So, I would say that the message signaling interrupt code still needs some major work. -- You are receiving this mail because: You are watching the assignee of the bug. _______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/dri-devel
