Linux: Testing The Linux Development Kernel

Submitted by Jeremy
on March 27, 2003 - 9:42pm

As the Linux 2.5 development kernel evolves and stabilizes heading towards an eventual 2.6 stable release in future months, it become increasingly important for a wider set of users to test it out and help find the remaining bugs. Scott Robert Ladd recently asked on the lkml, "What do the kernel developers want from testers? What sort of reports are helpful? Is there anything in particular that needs extensive testing?" The earlier linked FAQ includes instructions on reporting kernel Oops as well as other bugs.

As to the question of what needs testing, Alan Cox [interview] suggested, "One of the best things people can do is just use it. OSDL and others run stress tests but often its users configurations that find bugs not stress and coverage runs." Details on the Open Source Development Lab's stabilization efforts can be found here.

In any case, if you're interested in lending a hand to Linux kernel development, you can help out by simply compiling the latest 2.5 development kernel and taking it for a test run. Recent performance improvements should make this an enjoyable experience for you, but remember you're helping to test a beta kernel and may possibly find a bug, so be sure that any important data you have is safely backed up. Of course, this advice brings to mind the infamous Linus Torvalds' quote, "Backups are for wimps. Real men upload their data to an FTP site and have everyone else mirror it."


From: Scott Robert Ladd
To: linux-kernel
Subject: Testing: What do you want?
Date: Mon, 24 Mar 2003 09:46:05 -0500

OSNews and LinuxJournal are urging people to test the Linux kernel. I've 
been doing so for several months now, running the latest kernel live on 
my primary development system. I figure the best way to know if the 
kernel works is to use it when I'm working. ;)

At the moment, my biggest contribution to kernel development is testing, 
given that I'm not yet comfortable hacking kernel code in public. My 
question is: What do the kernel developers want from testers? What sort 
of reports are helpful? Is there anything in particular that needs 
extensive testing?

I haven't found a "Guide to Testing the Kernel" anywhere -- perhaps such 
a document exists, but I've missed it. Such a document could provide 
basic information for "testing newbies."

My experience thus far:
For the most part, the 2.5 series has worked very well for me, albeit 
with a few glitches (radeonfb, for example, as reported last week.) I'll 
build the 2.5.65 kernel on my Sparc later today, and see how well it 
works there.

-- 
Scott Robert Ladd
Coyote Gulch Productions (http://www.coyotegulch.com)
Professional programming for science and engineering;
Interesting and unusual bits of very free code.


From: Alan Cox Subject: Re: Testing: What do you want? Date: 24 Mar 2003 17:01:15 +0000 On Mon, 2003-03-24 at 14:46, Scott Robert Ladd wrote: > question is: What do the kernel developers want from testers? What sort > of reports are helpful? Is there anything in particular that needs > extensive testing? One of the best things people can do is just use it. OSDL and others run stress tests but often its users configurations that find bugs not stress and coverage runs.
From: Craig Thomas Subject: Re: Testing: What do you want? Date: 24 Mar 2003 15:06:26 -0800 I would echo Alan's comments. Testing on various platforms with different configurations will show that linux is capable of running on a larger variety of hardware. OSDL has a set of stress tests that are run on new kernels as they become available. However, the architectures tested are limited (compared with the variety that can be found in the world). There are enough testers in the community to collectively test for correctness of the kernel. Test suites such as LTP and LSB help in this regard. It would be nice to have a collaborative effort of testing and reporting of findings so that everyone can look at particular areas of interest. There is also need to generate test results quickly so that defects can be found and corrected in a timely manner if they exist.

Related Links:

Does anyone have a list of al

Anonymous
on
March 27, 2003 - 10:04pm

Does anyone have a list of all the packages (eg. modutils?) that need to be upgraded so that I can use this kernel? Also, what GCC version is recommended for building the kernel. I hear 2.95.2 or thereabouts is pretty popular, as is RedHat's "2.96".

Same here

Anonymous
on
March 27, 2003 - 11:04pm

Same question: I've been following kernel development closely. Do I need to upgrade glibc for e.g. the sysenter stuff and the likes? What other packages+versions does 2.5 require that 2.4 did not?

Documentation/Changes

schneelocke
on
March 28, 2003 - 1:48am

Documentation/Changes in the kernel tree has a list of current minimum requirements with regard to third-party software.

--
schnee

Help is available.

Anonymous
on
March 28, 2003 - 3:00am

This is a must-read from Dave Jones
http://www.codemonkey.org.uk/post-halloween-2.5.txt

Re: Modutils

Anonymous
on
March 28, 2003 - 5:45am

The new module-init-tools are backwards compatible with the older kernels.
So that means that you can install the module-init-tools and use those to insert your 2.4.x modules as well

Re: modutils

Tomasz T.
on
March 28, 2003 - 8:14am

To be more precise:
module-init-tools during install renames modprobe, insmod and rest programs from modutils to modprobe.old, insmod.old, depmod.old and so on. When module-init-tools cannot handle module loading (with 2.4.x or earlier kernel), those *.old binaries are automatically called.
So technically, you need to have both installed.

--
:wq

HOWTO feedback?

Anonymous
on
March 28, 2003 - 7:27am

Being myself a complete newbie concerning kernel-developing, I find an issue with us (newbies) testing the kernel: how do I give feedback? Sometimes it's not easy to have an idea of where the source of the problem is (i.e., what modules), and even more difficult, to know what data you developers would need in order to find the problem.

As an example, I'll put my own case so you can see more clear the problems that we might have (and by the way, let you know about it ;-)
--------------
Scenario: Laptop with an logitech usb mouse attached to it. Quite heavily loaded machine. Bug proven to happen with 2.5.65, 2.5.66 and 2.5.66-mm1.

Symptoms: If I unplug and plugin again the usb mouse the machine freezes completely. Usually when heavily loaded.

Now the problems that I have to provide you with feedback. Even if I want to give feedback to the kernel maillist, I have no clue on how to do it, since:

1.- I don't know what part of the kernel this could be related to: usb? the new schedulers? I enabled the "preemptive" option, ... could it be related?............ and so on
2.- I don't know what details to give back to the kernel list. Logs? Are there any proper logs written when the machine hungs? Some status at the time this happens? Should I send cat /proc/....?
3.- I don't know if this is a known issue... is there any bugzilla style bug list around?

I hope this helps to see a newbie's point of view, and I think it will be shared by lots of other newbies. I wish there was a "kernel-provide-feedback-HOWTO"

Re: HOWTO feedback?

Tomasz T.
on
March 28, 2003 - 8:18am

less /usr/src/linux/REPORTING-BUGS

http://bugme.osdl.org/.
--
:wq

Thanks!

Anonymous
on
March 28, 2003 - 8:40am

Ooops, I hadn't noticed that file. Thanks very much indeed.

I don't know if this is ment

Anonymous
on
March 28, 2003 - 11:42am

I don't know if this is mentioned in the REPORTING-BUGS doc. If the feature used to work (in 2.4.x or earlier 2.5.x), figuring out which 2.5.x kernel release first introduced the bug is extremely helpful! That narrows done the possible causes of the bug. Of course, building and testing older kernel releases can be a lot of work. To simplify your regression testing, don't test every old kernel. Try using a binary search of the older kernels to locate the bug's injection in O(log n) time! ;-)

Re: I don't know if this is ment

Anonymous
on
March 28, 2003 - 3:08pm

Ah, I'll keep that in mind too. I think I still keep the old 2.4.19 & 2.4.21-presomething compiled.... and the issue existed for sure in 2.5.65. Not too sure in 2.5.64.

nvidia

nero
on
March 28, 2003 - 8:42am

I would use 2.5 daily if the nvidia driver worked with it.

Re: nvidia

Anonymous
on
March 28, 2003 - 9:09am

Yup, although there are tons of patches floating around the net (e.g. on http://www.minion.de/), I have yet not been able to load the NVdriver module on 2.5.

minion.de patches work fine h

zayamut
on
March 28, 2003 - 1:21pm

minion.de patches work fine here on my DELL Inspiron 8100.

if you can successfully compile NVdriver it should also work.
Be sure to disable "module version info".

--
I used to have a sig until the great Kahuna of FOOness
told me to dump it and use /dev/urandom instead.

Er..

Anonymous
on
March 28, 2003 - 3:04pm

Ok, just remember that you'll be laughed at if yo report bugs with a binary only module loaded taht taints the kernel. Use the nv driver..

nvidia and other

Tomasz T.
on
March 29, 2003 - 3:02am

Or mplayer's 'mga_vid' kernel module. And VMware modules.

--
:wq

VMware works

Anonymous
on
March 29, 2003 - 7:05am

I had VMware working on 2.5.64. Not sure about the mplayer's thingy

VMware on 2.5

Tomasz T.
on
March 29, 2003 - 8:33am

In file included from ../common/hostif.h:42,
from .././linux/driver.c:27:
/lib/modules/2.5.66-bk4/build/include/asm/page.h:9: warning: `LARGE_PAGE_MASK' redefined
../include/vm_types.h:72: warning: this is the location of the previous definition
/lib/modules/2.5.66-bk4/build/include/asm/page.h:10: warning: `LARGE_PAGE_SIZE' redefined
../include/vm_types.h:71: warning: this is the location of the previous definition
In file included from /lib/modules/2.5.66-bk4/build/include/linux/irq.h:19,
from /lib/modules/2.5.66-bk4/build/include/asm/hardirq.h:6,
from /lib/modules/2.5.66-bk4/build/include/linux/interrupt.h:9,
from .././linux/driver.c:44:

/lib/modules/2.5.66-bk4/build/include/asm/irq.h:16: irq_vectors.h: No such file or directory

make[2]: *** [driver.o] Error 1
make[2]: Leaving directory `/tmp/vmware-config0/vmmon-only/driver-2.5.66-bk4'
make[1]: *** [driver] Error 2
make[1]: Leaving directory `/tmp/vmware-config0/vmmon-only'
make: *** [auto-build] Error 2
make: Leaving directory `/tmp/vmware-config0/vmmon-only'
Unable to build the vmmon module.

But that's VMware problem. We shouldn't discuss it on this opern source forum.
--
:wq

works for me

gncuster
on
March 29, 2003 - 4:44am

I am running 2.5.65 with the latest nvidia drivers and I am having no issues. Just apply the patch (gentoo's nvidia package checks for 2.5 and applies it for you. Go Gentoo)

Stalls when doing "ls -l" on /dev

Anonymous
on
March 29, 2003 - 9:57am

Hmm...is anybody seeing anything like this?

Using 2.5.66 with the 2.5.66-mm1 patchset. Using devfs and I'm getting intermittent stalls when doing an "ls -l" using gnome-terminal 2.x.

The system won't hang, but suddenly the windows become jerkier to move and ls -l hangs right in the middle...

Maybe this thread on LKML can

Anonymous
on
March 30, 2003 - 5:23am

Maybe this thread on LKML can help you:

http://marc.theaimsgroup.com/?t=104887474800001&r=1&w=2

Ah...

Anonymous
on
March 30, 2003 - 7:04am

Thanks!

So it _is_ a bug in gnome-terminal. I started to suspect that this morning when I tried to recreate this problem on a console, but couldn't.

I have 0.11.0-1 of the vte packages, when a newer version comes out I'll see if it helps.

sb oss driver

Anonymous
on
March 29, 2003 - 12:02pm

ive encountered a problem with 2.5.65, the sb oss driver wont build, i dont know where to post the problem or how to do it, so i didnt.
it accures always in the same thing..

i think linux's "bug" thing is really louse, i mean? WHERE can i post this problem? in what form? ive searched but now found anything usefull..

Reporting Bugs

Anonymous
on
April 9, 2003 - 9:20am

http://bugzilla.kernel.org

There is also a reporting-bugs file in the source tree that will help you give information that may help solve the issue.

-Nathan

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.