Bug 1516584

Summary: KVM with 'vga' / 'std' graphics, and VirtualBox, hangs at startup of graphical environment with kernel-4.15.0-0.rc0.git6.1.fc28
Product: [Fedora] Fedora Reporter: Adam Williamson <awilliam>
Component: kernelAssignee: Kernel Maintainer List <kernel-maint>
Status: CLOSED RAWHIDE QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: high Docs Contact:
Priority: unspecified    
Version: rawhideCC: airlied, ajax, bskeggs, ewk, gkrithi8, hdegoede, ichavero, itamar, jarodwilson, jeremy, jglisse, john.j5live, jonathan, josef, kernel-maint, linville, mchehab, mjg59, steved
Target Milestone: ---   
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2017-11-24 22:44:28 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
first trace (looks storage-y, probably unrelated)
none
second trace (more likely related)
none
Boot hang in virtualbox after upgrading to rawhide
none
virtualbox boots fine in kernel-4.15.0-0.rc0.git7.2.fc28 none

Description Adam Williamson 2017-11-23 01:03:15 UTC
kernel-4.15.0-0.rc0.git6.1.fc28 landed in Rawhide compose 20171121.n.0 , and all openQA tests except the 'text install' one failed.

We have openQA set to use the 'std' video adapter in qemu ('-vga std'). I can reproduce the problem using a virt-manager VM with the graphics adapter set to what virt-manager calls 'vga', which results in '-device VGA,id=video0,vgamem_mb=16,bus=pci.0,addr=0x2' in the qemu params. With both those configurations, when a graphical environment should start up (anaconda on an installer image, GNOME or KDE on a live image) the system hangs. If I use 'qxl' or 'virtio' as the adapter, the VM boots fine.

Two kernel traces can be found in the boot messages. The first seems to be storage related, and may not be fatal or relevant to this bug. The second looks like this:

[   30.124968] Call Trace:
[   30.125186]  ttm_pool_populate+0x19b/0x400 [ttm]
[   30.125578]  ttm_bo_vm_fault+0x325/0x570 [ttm]
[   30.125964]  __do_fault+0x19/0x11e
[   30.126255]  __handle_mm_fault+0xcd3/0x1260
[   30.126609]  handle_mm_fault+0x14c/0x310
[   30.126947]  __do_page_fault+0x28c/0x530
[   30.127282]  do_page_fault+0x32/0x270
[   30.127593]  async_page_fault+0x22/0x30

I'll attach the full contents of both traces.

Comment 1 Adam Williamson 2017-11-23 01:03:59 UTC
Created attachment 1357947 [details]
first trace (looks storage-y, probably unrelated)

Comment 2 Adam Williamson 2017-11-23 01:04:48 UTC
Created attachment 1357948 [details]
second trace (more likely related)

Comment 3 gkrithi8 2017-11-23 06:23:22 UTC
Created attachment 1358009 [details]
Boot hang in virtualbox after upgrading to rawhide

Virtualbox Version 5.1.30 r118389 (Qt5.9.2)

Comment 4 Adam Williamson 2017-11-23 07:03:06 UTC
Yep, that looks like the same thing.

Comment 5 Adam Williamson 2017-11-24 22:44:28 UTC
Looks to be fixed in kernel-4.15.0-0.rc0.git7.2.fc28 , in the 20171124.n.0 compose. Thanks for the quick fix!

Comment 6 gkrithi8 2017-11-25 06:22:51 UTC
Created attachment 1358895 [details]
virtualbox boots fine in kernel-4.15.0-0.rc0.git7.2.fc28

as suggested in previous comment.