Bug 106850 - Occasionally random uninitialised memory appears to be dumped when applications fullscreen or the VT is switched
Summary: Occasionally random uninitialised memory appears to be dumped when applicatio...
Status: CLOSED WORKSFORME
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: unspecified
Hardware: x86-64 (AMD64) Linux (All)
: medium normal
Assignee: Intel GFX Bugs mailing list
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard: Triaged
Keywords:
Depends on:
Blocks:
 
Reported: 2018-06-07 16:27 UTC by Nicolas Frattaroli
Modified: 2018-06-08 09:32 UTC (History)
1 user (show)

See Also:
i915 platform: KBL
i915 features:

Attachments
the bug occurring on tty2 (1.77 MB, image/jpeg)
2018-06-07 16:27 UTC, Nicolas Frattaroli 		no flags Details
dmesg with drm.debug=0x1e log_buf_len=1M (740.83 KB, text/plain)
2018-06-07 16:28 UTC, Nicolas Frattaroli 		no flags Details
View All

Description Nicolas Frattaroli 2018-06-07 16:27:58 UTC 
Created attachment 140069 [details]
the bug occurring on tty2

Occasionally random uninitialised memory appears to be dumped when applications fullscreen or the VT is switched

Occasionally, a pattern that looks like uninitialised memory is vomited on top of the screen, usually when an application enters fullscreen, or a VT switch is done.
Moving the mouse makes the pattern disappear in a Xorg VT, but not on the tty.

The regular look of the pattern makes me think that this is some area of possibly kernel memory that is being dumped onto the framebuffer. It also appears on top of everything.

Steps to reproduce:
1. fullscreen mpv with a video playing a bunch, or switch VTs a bunch
2. notice how the corruption disappears when the mouse cursor is moved

Occurrence with fullscreening mpv appears to happen in bursts and then not at all for a while, which is why it's hard to put a number on the tries you need to reproduce it.

✓ fratti@archbook ~ $ uname -m
x86_64
✓ fratti@archbook ~ $ uname -r
4.16.13-2-ARCH

Distribution: Arch Linux
Machine: Tuxedo InfinityBook 13" with Intel Corporation HD Graphics 620 (rev 02)
Display connector: eDP
Comment 1 Nicolas Frattaroli 2018-06-07 16:28:49 UTC 
Created attachment 140070 [details]
dmesg with drm.debug=0x1e log_buf_len=1M
Comment 2 Francesco Balestrieri 2018-06-07 16:53:40 UTC 
Would you be able to test this on the latest drm-tip?
Comment 3 Chris Wilson 2018-06-07 16:58:29 UTC 
What I think might be useful would be cat /sys/kernel/debug/dri/0/i915_display_info when switched away. The first thought is that looks like garbage in a sprite plane, hence checking what planes/fb are still active after VT switch.
Comment 4 Nicolas Frattaroli 2018-06-07 17:03:00 UTC 
(In reply to Chris Wilson from comment #3)
> What I think might be useful would be cat
> /sys/kernel/debug/dri/0/i915_display_info when switched away. The first
> thought is that looks like garbage in a sprite plane, hence checking what
> planes/fb are still active after VT switch.

When switched away from the VT in which the bug occurs, or when switched to the VT where the bug occurs?

Here's when it's occurring on tty2 and I'm switched to Xorg:
CRTC info
---------
CRTC 37: pipe: A, active=yes, (size=1920x1080), dither=no, bpp=24
	fb: 80, pos: 0x0, size: 1920x1080
	encoder 58: type: DDI A, connectors:
		connector 59: type: eDP-1, status: connected, mode:
		id 0:"1920x1080" freq 60 clock 138500 hdisp 1920 hss 1968 hse 2000 htot 2080 vdisp 1080 vss 1083 vse 1088 vtot 1111 type 0x48 flags 0xa
	cursor visible? no, position (0, 0), size 0x0, addr 0x00000000
	num_scalers=2, scaler_users=0 scaler_id=-1, scalers[0]: use=no, mode=0, scalers[1]: use=no, mode=0
	--Plane id 28: type=PRI, crtc_pos=   0x   0, crtc_size=1920x1080, src_pos=0.0000x0.0000, src_size=1920.0000x1080.0000, format=XR24 little-endian (0x34325258), rotation=0 (0x00000001)
	--Plane id 31: type=OVL, crtc_pos=   0x   0, crtc_size=   0x   0, src_pos=0.0000x0.0000, src_size=0.0000x0.0000, format=N/A, rotation=0 (0x00000001)
	--Plane id 34: type=CUR, crtc_pos=   0x   0, crtc_size=   0x   0, src_pos=0.0000x0.0000, src_size=0.0000x0.0000, format=N/A, rotation=0 (0x00000001)
	underrun reporting: cpu=yes pch=yes 
CRTC 47: pipe: B, active=no, (size=0x0), dither=no, bpp=0
	underrun reporting: cpu=yes pch=yes 
CRTC 57: pipe: C, active=no, (size=0x0), dither=no, bpp=0
	underrun reporting: cpu=yes pch=yes 

Connector info
--------------
connector 59: type eDP-1, status: connected
	name: 
	physical dimensions: 290x170mm
	subpixel order: Unknown
	CEA rev: 0
	DPCD rev: 12
	audio support: no
	fixed mode:
		id 63:"1920x1080" freq 60 clock 138500 hdisp 1920 hss 1968 hse 2000 htot 2080 vdisp 1080 vss 1083 vse 1088 vtot 1111 type 0x48 flags 0xa
	DP branch device present: no
	modes:
		id 60:"1920x1080" freq 60 clock 138500 hdisp 1920 hss 1968 hse 2000 htot 2080 vdisp 1080 vss 1083 vse 1088 vtot 1111 type 0x48 flags 0xa
		id 61:"1920x1080" freq 48 clock 110920 hdisp 1920 hss 1968 hse 2000 htot 2080 vdisp 1080 vss 1083 vse 1088 vtot 1111 type 0x40 flags 0xa
connector 67: type DP-1, status: disconnected
	DPCD rev: 0
	audio support: no
	DP branch device present: no
	modes:
connector 72: type HDMI-A-1, status: disconnected
	audio support: no
	modes:
connector 75: type HDMI-A-2, status: disconnected
	audio support: no
	modes:

Here's when it's happening on tty2 and I am on tty2:
CRTC info
---------
CRTC 37: pipe: A, active=yes, (size=1920x1080), dither=no, bpp=24
	fb: 78, pos: 0x0, size: 1920x1080
	encoder 58: type: DDI A, connectors:
		connector 59: type: eDP-1, status: connected, mode:
		id 0:"1920x1080" freq 60 clock 138500 hdisp 1920 hss 1968 hse 2000 htot 2080 vdisp 1080 vss 1083 vse 1088 vtot 1111 type 0x48 flags 0xa
	cursor visible? no, position (0, 0), size 0x0, addr 0x00000000
	num_scalers=2, scaler_users=0 scaler_id=-1, scalers[0]: use=no, mode=0, scalers[1]: use=no, mode=0
	--Plane id 28: type=PRI, crtc_pos=   0x   0, crtc_size=1920x1080, src_pos=0.0000x0.0000, src_size=1920.0000x1080.0000, format=XR24 little-endian (0x34325258), rotation=0 (0x00000001)
	--Plane id 31: type=OVL, crtc_pos=   0x   0, crtc_size=   0x   0, src_pos=0.0000x0.0000, src_size=0.0000x0.0000, format=N/A, rotation=0 (0x00000001)
	--Plane id 34: type=CUR, crtc_pos=   0x   0, crtc_size=   0x   0, src_pos=0.0000x0.0000, src_size=0.0000x0.0000, format=N/A, rotation=0 (0x00000001)
	underrun reporting: cpu=yes pch=yes 
CRTC 47: pipe: B, active=no, (size=0x0), dither=no, bpp=0
	underrun reporting: cpu=yes pch=yes 
CRTC 57: pipe: C, active=no, (size=0x0), dither=no, bpp=0
	underrun reporting: cpu=yes pch=yes 

Connector info
--------------
connector 59: type eDP-1, status: connected
	name: 
	physical dimensions: 290x170mm
	subpixel order: Unknown
	CEA rev: 0
	DPCD rev: 12
	audio support: no
	fixed mode:
		id 63:"1920x1080" freq 60 clock 138500 hdisp 1920 hss 1968 hse 2000 htot 2080 vdisp 1080 vss 1083 vse 1088 vtot 1111 type 0x48 flags 0xa
	DP branch device present: no
	modes:
		id 60:"1920x1080" freq 60 clock 138500 hdisp 1920 hss 1968 hse 2000 htot 2080 vdisp 1080 vss 1083 vse 1088 vtot 1111 type 0x48 flags 0xa
		id 61:"1920x1080" freq 48 clock 110920 hdisp 1920 hss 1968 hse 2000 htot 2080 vdisp 1080 vss 1083 vse 1088 vtot 1111 type 0x40 flags 0xa
connector 67: type DP-1, status: disconnected
	DPCD rev: 0
	audio support: no
	DP branch device present: no
	modes:
connector 72: type HDMI-A-1, status: disconnected
	audio support: no
	modes:
connector 75: type HDMI-A-2, status: disconnected
	audio support: no
	modes:
Comment 5 Nicolas Frattaroli 2018-06-07 18:47:28 UTC 
Built drm-tip with the same kernel config.
Can't reproduce on drm-tip anymore. I'll let you know if I do happen to run into it again, but it appears to be gone now from the testing I've done.
Comment 6 Jani Saarinen 2018-06-08 09:32:35 UTC 
Thanks for reporting, closing now, please re-open if seen again.

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.