From mboxrd@z Thu Jan 1 00:00:00 1970 From: bugzilla-daemon@freedesktop.org Subject: [Bug 92974] Fiji Nano long boot up and long X startup with amdgpu-powerplay enabled Date: Tue, 17 Nov 2015 02:11:51 +0000 Message-ID: Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="===============1663547307==" Return-path: Received: from culpepper.freedesktop.org (unknown [131.252.210.165]) by gabe.freedesktop.org (Postfix) with ESMTP id 1FDB86E281 for ; Mon, 16 Nov 2015 18:11:51 -0800 (PST) List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dri-devel-bounces@lists.freedesktop.org Sender: "dri-devel" To: dri-devel@lists.freedesktop.org List-Id: dri-devel@lists.freedesktop.org --===============1663547307== Content-Type: multipart/alternative; boundary="1447726310.bA1aE0.29246"; charset="UTF-8" --1447726310.bA1aE0.29246 Date: Tue, 17 Nov 2015 02:11:50 +0000 MIME-Version: 1.0 Content-Type: text/plain; charset="UTF-8" https://bugs.freedesktop.org/show_bug.cgi?id=92974 Bug ID: 92974 Summary: Fiji Nano long boot up and long X startup with amdgpu-powerplay enabled Product: DRI Version: DRI git Hardware: Other OS: All Status: NEW Severity: normal Priority: medium Component: DRM/AMDgpu Assignee: dri-devel@lists.freedesktop.org Reporter: bug0xa3d2@hushmail.com Created attachment 119727 --> https://bugs.freedesktop.org/attachment.cgi?id=119727&action=edit Output of dmesg after startup. With AMD Powerplay enabled in the kernel (cloned from http://cgit.freedesktop.org/~agd5f/linux/?h=amdgpu-powerplay)--when the computer boots there is text, as usual, before modesetting tries to switch to the native resolution of the monitor. At this point, the monitor backlight turns off for about 5 seconds and then stays on for approximately 2 minutes displaying nothing as if the computer locked up with no other boot activity occuring. The Alt+PrintScreen+E/I/U/R button combo still works at this time. If left alone the system finishes booting normally after about 2 minutes. dmesg gives these new error/messages: "Failed to send Previous Message." and ..."[drm] ib test on ring 12 succeeded" that I have not noticed when AMD Powerplay is disabled. Trying to run, "startx" at the console also simulates a system freeze for about 2 minutes or so then X starts normally. To work around the long boot and long X startup bug the kernel option, "Device Drivers/Graphics support/Direct Rendering Manager/Enable legacy fbdev support for your modesetting driver" must be set with, "n". This gives a black screen or no console after the machine boots to log in blind with. X can also be started blind and displayed normally with no startup delay. Another option is to autologin with KDM or such like to boot directly into X after legacy fbdev has been disabled in the kernel options. With AMD Powerplay disabled in the kernel there are no startup issues. Toggling or including/excluding these kernel boot, "lilo.conf" options have no effect on the bug: "amdgpu.enable_scheduler=0 ; radeon.modeset=1 ; radeon.hw_i2c=1 ; radeon.disp_priority=2 ; radeon.fastfb=1 ; radeon.backlight=1 ; radeon.pcie_gen2=-1 ; radeon.hard_reset=0" These patches (applied and not applied) have no effect on the bug: http://people.freedesktop.org/~agd5f/0001-radeonsi-fix-fiji-raster-config.patch http://people.freedesktop.org/~agd5f/0001-drm-amdgpu-update-Fiji-s-tiling-mode-table.patch Hardware: Fury Nano git: http://cgit.freedesktop.org/~agd5f/linux/commit/?h=amdgpu-powerplay&id=256ca780500bc445752763f32b246dc9ee396b62 Sidenote: hang check timer is disabled in the kernel. -- You are receiving this mail because: You are the assignee for the bug. --1447726310.bA1aE0.29246 Date: Tue, 17 Nov 2015 02:11:50 +0000 MIME-Version: 1.0 Content-Type: text/html; charset="UTF-8"
Bug ID 92974
Summary Fiji Nano long boot up and long X startup with amdgpu-powerplay enabled
Product DRI
Version DRI git
Hardware Other
OS All
Status NEW
Severity normal
Priority medium
Component DRM/AMDgpu
Assignee dri-devel@lists.freedesktop.org
Reporter bug0xa3d2@hushmail.com 

Created attachment 119727 [details]
Output of dmesg after startup.

With AMD Powerplay enabled in the kernel (cloned from
http://cgit.freedesktop.org/~agd5f/linux/?h=amdgpu-powerplay)--when the
computer boots there is text, as usual, before modesetting tries to switch to
the native resolution of the monitor.  At this point, the monitor backlight
turns off for about 5 seconds and then stays on for approximately 2 minutes
displaying nothing as if the computer locked up with no other boot activity
occuring.  The Alt+PrintScreen+E/I/U/R button combo still works at this time. 
If left alone the system finishes booting normally after about 2 minutes. 
dmesg gives these new error/messages: "Failed to send Previous Message." and
..."[drm] ib test on ring 12 succeeded" that I have not noticed when AMD
Powerplay is disabled.

Trying to run, "startx" at the console also simulates a system freeze for about
2 minutes or so then X starts normally.

To work around the long boot and long X startup bug the kernel option, "Device
Drivers/Graphics support/Direct Rendering Manager/Enable legacy fbdev support
for your modesetting driver" must be set with, "n".  This gives a black screen
or no console after the machine boots to log in blind with.  X can also be
started blind and displayed normally with no startup delay.  Another option is
to autologin with KDM or such like to boot directly into X after legacy fbdev
has been disabled in the kernel options.

With AMD Powerplay disabled in the kernel there are no startup issues.

Toggling or including/excluding these kernel boot, "lilo.conf" options have no
effect on the bug: "amdgpu.enable_scheduler=0 ; radeon.modeset=1 ;
radeon.hw_i2c=1 ; radeon.disp_priority=2 ; radeon.fastfb=1 ; radeon.backlight=1
; radeon.pcie_gen2=-1 ; radeon.hard_reset=0"

These patches (applied and not applied) have no effect on the bug:
http://people.freedesktop.org/~agd5f/0001-radeonsi-fix-fiji-raster-config.patch
http://people.freedesktop.org/~agd5f/0001-drm-amdgpu-update-Fiji-s-tiling-mode-table.patch

Hardware: Fury Nano
git:
http://cgit.freedesktop.org/~agd5f/linux/commit/?h=amdgpu-powerplay&id=256ca780500bc445752763f32b246dc9ee396b62

Sidenote: hang check timer is disabled in the kernel.

You are receiving this mail because:
  • You are the assignee for the bug.
--1447726310.bA1aE0.29246-- --===============1663547307== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: base64 Content-Disposition: inline X19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX18KZHJpLWRldmVs IG1haWxpbmcgbGlzdApkcmktZGV2ZWxAbGlzdHMuZnJlZWRlc2t0b3Aub3JnCmh0dHA6Ly9saXN0 cy5mcmVlZGVza3RvcC5vcmcvbWFpbG1hbi9saXN0aW5mby9kcmktZGV2ZWwK --===============1663547307==--