Bug 919036 - Tumbleweed kernel 3.19.0 nouveau - Grub and X Server hang
Summary: Tumbleweed kernel 3.19.0 nouveau - Grub and X Server hang
Status: RESOLVED FIXED
Alias: None
Product: openSUSE Tumbleweed
Classification: openSUSE
Component: Kernel (show other bugs)
Version: 201503*
Hardware: x86-64 Linux
: P3 - Medium : Critical with 5 votes (vote)
Target Milestone: ---
Assignee: E-mail List
QA Contact: E-mail List
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2015-02-23 11:37 UTC by Forgotten User b49zM5D78q
Modified: 2015-06-26 10:23 UTC (History)
2 users (show)

See Also:
Found By: ---
Services Priority:
Business Priority:
Blocker: ---
Marketing QA Status: ---
IT Deployment: ---


Attachments
Xorg.0.log after booting (vga=884) (9.21 KB, text/plain)
2015-02-23 12:04 UTC, Forgotten User b49zM5D78q
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Forgotten User b49zM5D78q 2015-02-23 11:37:59 UTC
OpenSUSE Tumbleweed - snapshot 20140220
HP ZBook 5, 32 MB RAM
Graphics: NVIDIA Corporation GK208GLM [Quadro K610M]

After updating to the latest kernel 3.19.0 my system doesn't boot any longer.

The first problem appeared in Grub - the boot parameter vga=default resulted in hanging after the 3rd line in the boot messages.

If I set vga=884 at least the boot continued, but the X Server didn't come up.
The problem seems to be in the nouveau driver, Xorg.0.log reports: No screen(s) found.

The last working kernel was 3.18.3 (using the unchanged X Server).
Currently I boot my system just using the kernel 3.16.7-7-desktop from the OpenSUSE 13.2 update.
Comment 1 Forgotten User b49zM5D78q 2015-02-23 12:04:49 UTC
Created attachment 624162 [details]
Xorg.0.log after booting (vga=884)
Comment 2 Stefan Dirsch 2015-02-23 13:07:02 UTC
Hmm. Also no generic framebuffer, no vesa Extension at all? 

> vga=0x361

Try first with something more common like 0x317 (1024x768@16bpp)
Comment 3 Forgotten User b49zM5D78q 2015-02-23 13:07:42 UTC
Typo above: Tumbleweed snapshot 20150220
Comment 4 Forgotten User b49zM5D78q 2015-02-23 13:29:11 UTC
The previous kernel worked, also the one from the 13.2 update repository.
Could be also an issue for the kernel maintainers, I haven't been sure where to assign it.

rkrell:~ # zcat /proc/config.gz | grep -v '^#' | grep FB
CONFIG_NET_SCH_SFB=m
CONFIG_IFB=m
CONFIG_VIDEO_FB_IVTV=m
CONFIG_DRM_KMS_FB_HELPER=y
CONFIG_DRM_I915_FBDEV=y
CONFIG_DRM_VMWGFX_FBCON=y
CONFIG_FB=y
CONFIG_FB_DDC=m
CONFIG_FB_BOOT_VESA_SUPPORT=y
CONFIG_FB_CFB_FILLRECT=y
CONFIG_FB_CFB_COPYAREA=y
CONFIG_FB_CFB_IMAGEBLIT=y
CONFIG_FB_SYS_FILLRECT=m
CONFIG_FB_SYS_COPYAREA=m
CONFIG_FB_SYS_IMAGEBLIT=m
CONFIG_FB_SYS_FOPS=m
CONFIG_FB_DEFERRED_IO=y
CONFIG_FB_BACKLIGHT=y
CONFIG_FB_MODE_HELPERS=y
CONFIG_FB_TILEBLITTING=y
CONFIG_FB_UVESA=m
CONFIG_FB_VESA=y
CONFIG_FB_EFI=y
CONFIG_FB_OPENCORES=m
CONFIG_FB_I740=m
CONFIG_FB_SMSCUFX=m
CONFIG_FB_VIRTUAL=m
CONFIG_FB_METRONOME=m
CONFIG_FB_MB862XX=m
CONFIG_FB_MB862XX_PCI_GDC=y
CONFIG_FB_MB862XX_I2C=y
CONFIG_FB_BROADSHEET=m
CONFIG_FB_AUO_K190X=m
CONFIG_FB_AUO_K1900=m
CONFIG_FB_AUO_K1901=m
CONFIG_FB_HYPERV=m
CONFIG_HID_PICOLCD_FB=y
CONFIG_FB_XGI=m
Comment 5 Forgotten User b49zM5D78q 2015-02-23 13:39:12 UTC
For completeness, the output above was for the running 3.16.7 kernel from the OpenSUSE 13.2 update repo.

This is the one for the failing kernel 3.19.0:
CONFIG_NET_SCH_SFB=m
CONFIG_IFB=m
CONFIG_VIDEO_FB_IVTV=m
CONFIG_DRM_KMS_FB_HELPER=y
CONFIG_DRM_I915_FBDEV=y
CONFIG_DRM_VMWGFX_FBCON=y
CONFIG_FB=y
CONFIG_FB_CMDLINE=y
CONFIG_FB_DDC=m
CONFIG_FB_BOOT_VESA_SUPPORT=y
CONFIG_FB_CFB_FILLRECT=y
CONFIG_FB_CFB_COPYAREA=y
CONFIG_FB_CFB_IMAGEBLIT=y
CONFIG_FB_SYS_FILLRECT=m
CONFIG_FB_SYS_COPYAREA=m
CONFIG_FB_SYS_IMAGEBLIT=m
CONFIG_FB_SYS_FOPS=m
CONFIG_FB_DEFERRED_IO=y
CONFIG_FB_BACKLIGHT=y
CONFIG_FB_MODE_HELPERS=y
CONFIG_FB_TILEBLITTING=y
CONFIG_FB_UVESA=m
CONFIG_FB_VESA=y
CONFIG_FB_EFI=y
CONFIG_FB_OPENCORES=m
CONFIG_FB_I740=m
CONFIG_FB_SMSCUFX=m
CONFIG_FB_VIRTUAL=m
CONFIG_FB_METRONOME=m
CONFIG_FB_MB862XX=m
CONFIG_FB_MB862XX_PCI_GDC=y
CONFIG_FB_MB862XX_I2C=y
CONFIG_FB_BROADSHEET=m
CONFIG_FB_AUO_K190X=m
CONFIG_FB_AUO_K1900=m
CONFIG_FB_AUO_K1901=m
CONFIG_FB_HYPERV=m
CONFIG_HID_PICOLCD_FB=y
CONFIG_FB_XGI=m

Furthermore, vga=0x361 did not work at all. The boot hangs with it here:
[ OK ] Reached target Basic System

Mode vga=884 boots without X Server and reaches at least a terminal for analyzing.
Comment 6 Forgotten User b49zM5D78q 2015-02-23 13:46:09 UTC
/var/log/messages when booting with vga=0x317 in kernel 3.19.0:

2015-02-23T14:31:12.997405+01:00 rkrell kernel: [    2.938711] vesafb: mode is 1024x768x16, linelength=2048, pages=0
2015-02-23T14:31:12.997405+01:00 rkrell kernel: [    2.938711] vesafb: scrolling: redraw
2015-02-23T14:31:12.997405+01:00 rkrell kernel: [    2.938713] vesafb: Truecolor: size=0:5:6:5, shift=0:11:5:0
2015-02-23T14:31:12.997406+01:00 rkrell kernel: [    2.938723] vesafb: framebuffer at 0x81000000, mapped to 0xffffc90013600000, using 1536k, total 1536k
2015-02-23T14:31:12.997406+01:00 rkrell kernel: [    2.966902] Console: switching to colour frame buffer device 128x48
2015-02-23T14:31:12.997408+01:00 rkrell kernel: [    2.995005] fb0: VESA VGA frame buffer device

...

2015-02-23T14:31:12.997574+01:00 rkrell kernel: [    4.473110] fb: switching to nouveaufb from VESA VGA
2015-02-23T14:31:12.997574+01:00 rkrell kernel: [    4.473130] Console: switching to colour dummy device 80x25
2015-02-23T14:31:12.997575+01:00 rkrell kernel: [    4.473509] nouveau  [  DEVICE][0000:01:00.0] BOOT0  : 0x108390a1
2015-02-23T14:31:12.997576+01:00 rkrell kernel: [    4.473511] nouveau  [  DEVICE][0000:01:00.0] Chipset: GK208 (NV108)
2015-02-23T14:31:12.997577+01:00 rkrell kernel: [    4.473512] nouveau  [  DEVICE][0000:01:00.0] Family : NVE0
2015-02-23T14:31:12.997577+01:00 rkrell kernel: [    4.474404] nouveau ![   VBIOS][0000:01:00.0] unable to locate usable image
2015-02-23T14:31:12.997577+01:00 rkrell kernel: [    4.474407] nouveau E[  DEVICE][0000:01:00.0] failed to create 0x10000001, -22
2015-02-23T14:31:12.997578+01:00 rkrell kernel: [    4.474409] nouveau E[     DRM] failed to create 0x00000080, -22
2015-02-23T14:31:12.997578+01:00 rkrell kernel: [    4.474581] nouveau: probe of 0000:01:00.0 failed with error -22
...
Comment 7 Forgotten User b49zM5D78q 2015-02-23 13:47:48 UTC
Again a typo, correct:
Furthermore, vga=0x317 did not work at all. The boot hangs with it here:
[ OK ] Reached target Basic System
Comment 8 Stefan Dirsch 2015-02-23 13:53:19 UTC
Ok. Seems nouveau break vesafb during load (and fails). With boot option 'nomodeset' you should at least have vesafb available. And yes, it the nouveau kernel driver, which is broken here.
Comment 9 Forgotten User b49zM5D78q 2015-02-23 13:59:47 UTC
Ok, thank you. To be reassigned to 'Basesystem'?
Comment 10 Stefan Dirsch 2015-02-23 14:30:19 UTC
No, component is fine. I would call this a regression in the kernel.
Comment 11 Forgotten User b49zM5D78q 2015-02-23 14:43:01 UTC
Seems to be already reported at a "higher instance": https://bugs.freedesktop.org/show_bug.cgi?id=89047
http://lists.freedesktop.org/archives/dri-devel/2015-February/077588.html
Comment 12 Forgotten User b49zM5D78q 2015-03-19 12:17:31 UTC
Doesn anyone have an idea of a patch or has anyone heard about some progress on this issue? The issue hasn't been fixed in kernel 3.19.1 neither.
Comment 13 Takashi Iwai 2015-06-26 10:23:14 UTC
The fix went int 4.1 upstream in commit 4195f40685a5f2783b4decece13ed740b61ee038
    drm/nouveau/bios: fix fetching from acpi on certain systems

Since the stable branch also moved to 4.1, let's close this bug.