Description of problem: I have a laptop with AMD A8-5557M APU (Radeon HD 8550G, r600) and discrete videocard Radeon HD 8750M (radeonsi). I also have a monitor connected via a hdmi and use it as a main screen. I use PRIME to play games: $ xrandr --setprovideroffloadsink 0x41 0x78 $ DRI_PRIME=1 steam In MGA6 sta1 it works fine, but since I update packages of the X, using steam often leads to system freeze. Sometimes even Magic SysRq keys doesn't work. Freeze often happens when using menu of the steam window, but also can happen even without touching the window. Freezes starts to appear when I install this updates (there are mutual dependencies between them, so I cannot install a single package): - lib64xdmcp6-1.1.2-5.mga6.x86_64 - x11-driver-input-6.0.0-5.mga6.x86_64 - x11-driver-input-acecad-1.5.0-22.mga6.x86_64 - x11-driver-input-evdev-2.10.5-1.mga6.x86_64 - x11-driver-input-evtouch-0.8.8-28.mga6.x86_64 - x11-driver-input-fpit-1.4.0-22.mga6.x86_64 - x11-driver-input-hyperpen-1.4.1-27.mga6.x86_64 - x11-driver-input-joystick-1.6.3-2.mga6.x86_64 - x11-driver-input-libinput-0.25.0-1.mga6.x86_64 - x11-driver-input-mutouch-1.3.0-22.mga6.x86_64 - x11-driver-input-penmount-1.5.0-22.mga6.x86_64 - x11-driver-input-synaptics-1.9.0-2.mga6.x86_64 - x11-driver-input-void-1.4.1-6.mga6.x86_64 - x11-driver-input-wacom-0.34.0-1.mga6.x86_64 - x11-driver-video-amdgpu-1.3.0-1.mga6.x86_64 - x11-driver-video-ati-7.9.0-1.mga6.x86_64 - x11-driver-video-cirrus-1.5.3-7.mga6.x86_64 - x11-driver-video-fbdev-0.4.4-17.mga6.x86_64 - x11-driver-video-glint-1.2.9-1.mga6.x86_64 - x11-driver-video-intel-2.99.917-38.mga6.x86_64 - x11-driver-video-mach64-6.9.5-6.mga6.x86_64 - x11-driver-video-mga-1.6.5-1.mga6.x86_64 - x11-driver-video-nouveau-1.0.14-1.mga6.x86_64 - x11-driver-video-openchrome-0.6.0-1.mga6.x86_64 - x11-driver-video-qxl-0.1.5-11.mga6.x86_64 - x11-driver-video-r128-6.10.2-1.mga6.x86_64 - x11-driver-video-s3-0.6.5-17.mga6.x86_64 - x11-driver-video-s3virge-1.10.7-6.mga6.x86_64 - x11-driver-video-savage-2.3.9-1.mga6.x86_64 - x11-driver-video-sis-0.10.9-1.mga6.x86_64 - x11-driver-video-sisimedia-0.9.1-9.20091203.26.mga6.x86_64 - x11-driver-video-sisusb-0.9.7-1.mga6.x86_64 - x11-driver-video-tdfx-1.4.7-1.mga6.x86_64 - x11-driver-video-trident-1.3.8-1.mga6.x86_64 - x11-driver-video-v4l-0.2.0-31.mga6.x86_64 - x11-driver-video-vboxvideo-5.1.18-6.mga6.x86_64 - x11-driver-video-vesa-2.3.4-8.mga6.x86_64 - x11-driver-video-vmware-13.2.1-3.mga6.x86_64 - x11-server-common-1.19.3-2.mga6.x86_64 - x11-server-xorg-1.19.3-2.mga6.x86_64 If I use snapshot of the system before this updates - I cannot reproduce this freezes. Version-Release number of selected component (if applicable): 1.19.3-2.mga6 How reproducible: Often Steps to Reproduce: 1. Run steam using PRIME 2. Use menu of the steam window: hover mouse over it, click, select entry. 3. The system should hang.
And it happens *only* when using steam, and never when you don't use it?
CC: (none) => marja11, rverschelde, thierry.vignaudSource RPM: x11-server-1.19.3-2.mga6.src.rpm => x11-server-1.19.3-2.mga6.src.rpm, steam
Only with steam on radeonsi using DRI_PRIME.
(In reply to Nikita Krupenko from comment #2) > Only with steam on radeonsi using DRI_PRIME. Assigning to the Steam maintainer, then.
Assignee: bugsquad => rverschelde
If some program can hang the whole system, this is a bug in a system, not in the program. So, this is obviously bug in the driver or in the X server.
(In reply to Nikita Krupenko from comment #4) > If some program can hang the whole system, this is a bug in a system, not in > the program. So, this is obviously bug in the driver or in the X server. I agree: even if there would be a bug in Steam, that should not lead to freezing everything. However, since there is no way to possibly reproduce this except when using Steam, I prefer to keep this bug assigned to the Steam maintainer, anyway: He'll know where to look upstream to see whether others hit this issue, too, and, if it isn't confined to Mageia only, whether a developer found what the (real) culprit is and how it can be fixed or worked around. However, your comment reminds me that I forgot to CC the kernel/drivers maintainers, so adding them now. Of course they and the X11-server maintainer are free to assign this bug report to himself or themselves if they think that is better.
CC: (none) => kernel