Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

[PATCH] Documentation: ksymoops should no longer be used to decode Oops messages

310 views
Skip to first unread message

Jesper Juhl

unread,
Oct 5, 2005, 4:37:41 PM10/5/05
to linux-kernel, Chris Ricker, chris....@genetics.utah.edu, Linus Torvalds
Document the fact that ksymoops should no longer be used to decode Oops
messages.


Signed-off-by: Jesper Juhl <jespe...@gmail.com>
---

Documentation/Changes | 7 +++----
1 files changed, 3 insertions(+), 4 deletions(-)

--- linux-2.6.14-rc3-git5-orig/Documentation/Changes 2005-10-03 21:54:50.000000000 +0200
+++ linux-2.6.14-rc3-git5/Documentation/Changes 2005-10-05 22:35:44.000000000 +0200
@@ -31,7 +31,7 @@
Eine deutsche Version dieser Datei finden Sie unter
<http://www.stefan-winter.de/Changes-2.4.0.txt>.

-Last updated: October 29th, 2002
+Last updated: October 05th, 2005

Chris Ricker (kab...@gatech.edu or chris....@genetics.utah.edu).

@@ -139,9 +139,8 @@
Ksymoops
--------

-If the unthinkable happens and your kernel oopses, you'll need a 2.4
-version of ksymoops to decode the report; see REPORTING-BUGS in the
-root of the Linux source for more information.
+With a 2.4 kernel you need ksymoops to decode a kernel Oops message. With
+2.6 kernels ksymoops is no longer needed and should not be used.

Module-Init-Tools
-----------------
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majo...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Alexey Dobriyan

unread,
Oct 5, 2005, 5:18:03 PM10/5/05
to Jesper Juhl, linux-...@vger.kernel.org, Chris Ricker, chris....@genetics.utah.edu, Linus Torvalds
On Wed, Oct 05, 2005 at 10:39:43PM +0200, Jesper Juhl wrote:
> Document the fact that ksymoops should no longer be used to decode Oops
> messages.

If it's considered harmful, better remove all references to ksymoops.
2.4 users will happily grab their copy of this file from 2.4 tree.

> Ksymoops
> --------
>
> -If the unthinkable happens and your kernel oopses, you'll need a 2.4
> -version of ksymoops to decode the report; see REPORTING-BUGS in the
> -root of the Linux source for more information.
> +With a 2.4 kernel you need ksymoops to decode a kernel Oops message. With
> +2.6 kernels ksymoops is no longer needed and should not be used.

-

Jesper Juhl

unread,
Oct 5, 2005, 5:26:17 PM10/5/05
to Alexey Dobriyan, linux-...@vger.kernel.org, Chris Ricker, chris....@genetics.utah.edu, Linus Torvalds
On Wednesday 05 October 2005 23:28, Alexey Dobriyan wrote:
> On Wed, Oct 05, 2005 at 10:39:43PM +0200, Jesper Juhl wrote:
> > Document the fact that ksymoops should no longer be used to decode Oops
> > messages.
>
> If it's considered harmful, better remove all references to ksymoops.
> 2.4 users will happily grab their copy of this file from 2.4 tree.
>

I opted to keep the entry but make an explicit note that ksymoops should not
be used for 2.6 kernels to make it clear. There are still people who use
ksymoops on 2.6 Oops messages, so I thought it would make sense to keep the
entry but make it clear that you should /not/ do that.

But, I'm fine with removing it as well.

> > Ksymoops
> > --------
> >
> > -If the unthinkable happens and your kernel oopses, you'll need a 2.4
> > -version of ksymoops to decode the report; see REPORTING-BUGS in the
> > -root of the Linux source for more information.
> > +With a 2.4 kernel you need ksymoops to decode a kernel Oops message. With
> > +2.6 kernels ksymoops is no longer needed and should not be used.
>

Here's a patch to remove the Ksymoops entry instead.

I'll leave it up to the powers that be to sprinkle holy penguin pee on the
preferred version.

Ksymoops should not be used on 2.6 Oops messages, so remove the ksymoops entry
from Documentation/Changes

Signed-off-by: Jesper Juhl <jespe...@gmail.com>
---

Documentation/Changes | 9 +--------
1 files changed, 1 insertion(+), 8 deletions(-)

--- linux-2.6.14-rc3-git5-orig/Documentation/Changes 2005-10-03 21:54:50.000000000 +0200

+++ linux-2.6.14-rc3-git5/Documentation/Changes 2005-10-05 23:21:45.000000000 +0200


@@ -31,7 +31,7 @@
Eine deutsche Version dieser Datei finden Sie unter
<http://www.stefan-winter.de/Changes-2.4.0.txt>.

-Last updated: October 29th, 2002
+Last updated: October 05th, 2005

Chris Ricker (kab...@gatech.edu or chris....@genetics.utah.edu).

@@ -136,13 +136,6 @@
types, have a fdformat which works with 2.4 kernels, and similar goodies.
You'll probably want to upgrade.

-Ksymoops
---------


-
-If the unthinkable happens and your kernel oopses, you'll need a 2.4
-version of ksymoops to decode the report; see REPORTING-BUGS in the
-root of the Linux source for more information.

-
Module-Init-Tools
-----------------

Kalin KOZHUHAROV

unread,
Oct 6, 2005, 11:54:41 AM10/6/05
to linux-...@vger.kernel.org

OK, but what should I use then with 2.6??
And since when is ksymoops not usable with it?
I have reported quite a few times here, alaways with 2.6.x and nobody said anything about it...

Kalin.

--
|[ ~~~~~~~~~~~~~~~~~~~~~~ ]|
+-> http://ThinRope.net/ <-+
|[ ______________________ ]|

Randy.Dunlap

unread,
Oct 6, 2005, 12:00:28 PM10/6/05
to Kalin KOZHUHAROV, linux-...@vger.kernel.org

You should just enable CONFIG_KALLSYMS so that the kernel can
report (decode) the symbols with the oops message.

Yes, ksymoops will still work, but it should only be used if
the kernel was not configured with CONFIG_KALLSYMS, and some
people say that the kernel oops + symbols are better than
the ksymoops output. I haven't compared them so can't say.

--
~Randy

Jesper Juhl

unread,
Oct 6, 2005, 12:10:20 PM10/6/05
to Kalin KOZHUHAROV, linux-...@vger.kernel.org
On 10/6/05, Kalin KOZHUHAROV <ka...@thinrope.net> wrote:
> Jesper Juhl wrote:
> > Document the fact that ksymoops should no longer be used to decode Oops
> > messages.
> >
[snip]

>
> OK, but what should I use then with 2.6??

Nothing at all.
Just post the raw Oops message as-is (from console or dmesg).

> And since when is ksymoops not usable with it?

At least since 2003 and the 2.5 kernel series.
The oldest reference I could find in the archives is an email by Dave
Jones dating back to Mon Apr 14 2003 - 14:31:44 EST
http://www.uwsg.iu.edu/hypermail/linux/kernel/0304.1/1533.html

The interresting part of that mail is this bit

"Merging of kksymoops means that the kernel will now spit out
automatically decoded oopses (no more feeding them to ksymoops).
For this reason, you should always enable the option in the
kernel hacking menu labelled "Load all symbols for debugging/kksymoops".
"

> I have reported quite a few times here, alaways with 2.6.x and nobody said anything about it...

They should have.

Please read a recent Documentation/oops-tracing.txt (here's a link:
http://sosdg.org/~coywolf/lxr/source/Documentation/oops-tracing.txt?v=2.6.14-rc2
)

See the note at the top :

"NOTE: ksymoops is useless on 2.6. Please use the Oops in its original format
(from dmesg, etc). Ignore any references in this or other docs to "decoding
the Oops" or "running it through ksymoops". If you post an Oops fron 2.6 that
has been run through ksymoops, people will just tell you to repost it.
"


--
Jesper Juhl <jespe...@gmail.com>
Don't top-post http://www.catb.org/~esr/jargon/html/T/top-post.html
Plain text mails only, please http://www.expita.com/nomime.html

Alexey Dobriyan

unread,
Oct 6, 2005, 12:37:40 PM10/6/05
to Kalin KOZHUHAROV, linux-...@vger.kernel.org
On Fri, Oct 07, 2005 at 12:49:48AM +0900, Kalin KOZHUHAROV wrote:
> > Ksymoops
> > --------
> >
> >-If the unthinkable happens and your kernel oopses, you'll need a 2.4
> >-version of ksymoops to decode the report; see REPORTING-BUGS in the
> >-root of the Linux source for more information.
> >+With a 2.4 kernel you need ksymoops to decode a kernel Oops message. With
> >+2.6 kernels ksymoops is no longer needed and should not be used.

> OK, but what should I use then with 2.6??

Raw Oops output. From logs, serial console, monitor photoed with
digital camera, copied to paper with pencil...

0 new messages