Bug 1305772

Summary: lspci chokes on open("/sys/bus/pci/devices/0000:01:00.0/label in kernels 4.3+
Product: [Fedora] Fedora Reporter: Marc <marc_vd_meer>
Component: xorg-x11-drv-nouveauAssignee: Ben Skeggs <bskeggs>
Status: CLOSED DUPLICATE QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 23CC: airlied, ajax, amoosbr, bskeggs, chunnayya, mhlavink, RichardWFJarvis+BugsRedhat
Target Milestone: ---   
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2016-03-14 07:51:18 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
output of strace -r -f lspci
none
strace of "lspci -vvxxx" none

Description Marc 2016-02-09 08:33:12 UTC 
Created attachment 1122354 [details]
output of strace -r -f lspci

Description of problem:
lspci chokes on open("/sys/bus/pci/devices/0000:01:00.0/label in kernels 4.3+, not on kernels 4.2.x and earlier

Version-Release number of selected component (if applicable):
pciutils-3.3.1-2.fc23.x86_64

How reproducible:
enter command lspci
or
enter "sudo -i", which hangs on lspci


Steps to Reproduce:
1.
2.
3.

Actual results:
lspci finishes but takes about 9 seconds


Expected results:
lspci should finish within one second

Additional info:
see bug https://bugzilla.redhat.com/show_bug.cgi?id=1302998
strace -r -f lspci output (attached) shows 
3625       8.728309 open("/sys/bus/pci/devices/0000:01:00.0/label", O_RDONLY) = -1 ENOENT (No such file or directory)

On kernel 4.2.8 max time observed is 0.7 seconds.
This run on a Lenovo W540, SSD, 32GB ram
Comment 1 Rich Jarvis 2016-02-11 16:03:17 UTC 
Created attachment 1123214 [details]
strace of "lspci -vvxxx"

Same behaviour witnessed on F22:

$ time lspci

...

real	0m8.735s
user	0m0.006s
sys	0m0.728s


$ uname -a
Linux obsidian 4.3.4-200.fc22.x86_64 #1 SMP Mon Jan 25 13:37:15 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux


Again a W540. Attached is the result:

$ strace -r -f lspci -vvxxx

in line with the request in the man page for those options when submitting bugs for lspci.
Comment 2 Marc 2016-02-27 12:31:28 UTC 
(In reply to Rich Jarvis from comment #1)
> Created attachment 1123214 [details]
> strace of "lspci -vvxxx"
> 
> Same behaviour witnessed on F22:
> 
> $ time lspci
> 
> ...
> 
> real	0m8.735s
> user	0m0.006s
> sys	0m0.728s
> 
> 
> $ uname -a
> Linux obsidian 4.3.4-200.fc22.x86_64 #1 SMP Mon Jan 25 13:37:15 UTC 2016
> x86_64 x86_64 x86_64 GNU/Linux
> 
> 
> Again a W540. Attached is the result:
> 
> $ strace -r -f lspci -vvxxx
> 
> in line with the request in the man page for those options when submitting
> bugs for lspci.

Not much movement here. What is your bios level of the W540?

BIOS Information
        Vendor: LENOVO
        Version: GNET71WW (2.19 )
        Release Date: 02/05/2015
Comment 3 Marc 2016-03-14 07:23:59 UTC 
Since no reaction on this bug I had another look at /var/log/messages while doing lspci - I see Nouveau is causing the delays:

 Mar 14 08:05:20 marchost kernel: nouveau 0000:01:00.0: DRM: resuming kernel object tree...
Mar 14 08:05:21 marchost kernel: nouveau 0000:01:00.0: devinit: 0x64da[0]: script needs connector type
Mar 14 08:05:23 marchost kernel: nouveau 0000:01:00.0: DRM: resuming client object trees...
Mar 14 08:05:27 marchost kernel: nouveau 0000:01:00.0: DRM: resuming display...
Mar 14 08:05:27 marchost kernel: nouveau 0000:01:00.0: DRM: resuming console...
Mar 14 08:05:32 marchost kernel: ACPI Warning: \_SB_.PCI0.PEG_.VID_._DSM: Argument #4 type mismatch - Found [Buffer], ACPI requires [Package] (20150930/nsarguments-95)
Mar 14 08:05:32 marchost kernel: ACPI Warning: \_SB_.PCI0.PEG_.VID_._DSM: Argument #4 type mismatch - Found [Buffer], ACPI requires [Package] (20150930/nsarguments-95)
Mar 14 08:05:32 marchost kernel: nouveau 0000:01:00.0: DRM: suspending console...
Mar 14 08:05:32 marchost kernel: nouveau 0000:01:00.0: DRM: suspending display...
Mar 14 08:05:32 marchost kernel: nouveau 0000:01:00.0: DRM: evicting buffers...
Mar 14 08:05:32 marchost kernel: nouveau 0000:01:00.0: DRM: waiting for kernel channels to go idle...
Mar 14 08:05:32 marchost kernel: nouveau 0000:01:00.0: DRM: suspending client object trees...
Mar 14 08:05:36 marchost kernel: nouveau 0000:01:00.0: DRM: suspending kernel object tree...

The string  "nouveau 0000:01:00.0:"

matches strace : 8.728309 open("/sys/bus/pci/devices/0000:01:00.0/label", O_RDONLY) = -1 ENOENT (No such file or directory) 

Don't know if changing the component of this bug to Nouveau is the way to do it, but since noone responded to this bug for over a month I guess I give it a try.

Marc
Comment 4 Dave Airlie 2016-03-14 07:51:18 UTC 

*** This bug has been marked as a duplicate of bug 1302847 ***