Kernel Traffic #317 For 14 Aug 2005

By Zack Brown

Table Of Contents


OK, about time! KT lateness has gotten a bit out of hand lately, and I'm going to make a real effort to get it back on schedule and keep it there. I'd like to thank the folks who emailed me to ask what was up, and whether I planned to keep going with KT or stop.

Basically, the reason for the lateness is that I do KT in my spare time, and I haven't had much of that for quite awhile. It takes something over a day to write a full issue, and I just haven't been able to find that time lately. I started a new day job, a really great one, and I end up spending a lot of extra time there.

That said, as the job shaped together and seemed to reward my putting more and more of my time into it, I ended up giving serious thought to stopping KT. I'd done it for 6 years, and maybe that was enough.

But ultimately, I don't want to stop. I want to keep reading the linux-kernel mailing list, and keep following along with the developers and see what will happen next. A lot of amazing things are happening, what with ditching the old stable/unstable development system, and the advent of git, and so much else. Linux is a problem that just keeps being solved, and I really like that.

So I'm sticking around. I plan to catch up on the issues I've missed, and then put a real schedule back in place, so KT once again comes out each week, and talks about the previous week, instead of several weeks ago. Wish me luck!

Mailing List Stats For This Week

We looked at 1842 posts in 13MB. See the Full Statistics.

There were 641 different contributors. 244 posted more than once. The average length of each message was 121 lines.

The top posters of the week were: The top subjects of the week were:
201 posts in 2MB by greg kh
38 posts in 196KB by domen puncer
38 posts in 154KB by andrew morton
30 posts in 172KB by ingo molnar
28 posts in 103KB by jeff garzik
69 posts in 283KB for "a great idea (tm) about reimplementing nls."
60 posts in 288KB for "preempt_rt vs i-pipe: the numbers, part 2"
47 posts in 472KB for "problem: devices behind pci express-to-pci bridge not mapped"
38 posts in 348KB for "2.6.12-mm1"
36 posts in 127KB for "2.6.12 udev hangs at boot"

These stats generated by mboxstats version 2.8

1. Kernel Debugger Compatibility Issues With RT Patch

13 Jun 2005 - 17 Jun 2005 (7 posts) Archive Link: "RT and kernel debugger ( 2.6.12rc6 + RT > 48-00 )"

Topics: Real-Time

People: Ingo Molnar

Serge Noiraud asked which kernel debugger to use in kernels that include the RT patch. He had been using kgdb, but this had stopped working after some spinlock modifications were accepted into the official tree. Ingo Molnar replied that some folks had been able to get kgdb working, just by renaming some of the spinlock calls. Kus Kusche Klaus said he had been one of these people, and posted a link describing how far he'd come ( to success. He had still encountered unresolved problems, so the issue remained open.

2. Dispute Over inotify Implementation

15 Jun 2005 - 21 Jun 2005 (40 posts) Archive Link: "[patch] inotify."

Topics: Ioctls

People: Christoph HellwigAndrew MortonArnd BergmannJohn McCutchanRobert Love

Robert Love announced a new version of inotify. Arnd Bergmann remarked on the fact that the user interface was very similar to epoll's user interface. But while epoll was implemented as a set of system calls, inotify was a character device. He asked whether one or the other was the correct way. Christoph Hellwig blew his stack immediately, saying, "It's because Robert and John insist on their horrible interface and simply ignore any feedback on how to do a better one." John McCutchan and Robert both replied that Christoph's feedback had indeed been considered, but that they and Andrew Morton had decided on the current approach as best. Christoph continued name-calling, while Robert tried to calm him down. Elsewhere, Andrew remarked:

I don't think I ever really affirmatively agreed to anything. I do recall various things being discussed at various times and various things being changed, but from where I sit it's all spread out and foggy.

I certainly remember that good-sounding recommendations which addressed the things which Christoph doesn't like were convincingly shot down by yourself and by Robert, but I don't recall why.

Look, this stuff is hard. This is why I've asked you and Robert again and again and again to generate some sort of design doc or FAQ which addresses each of these frequently-asked-questions. So the poor rest of us can look through it and say "oh yeah". Because inotify _is_ a tricky thing, and standard kernel interface designs _don't_ fit it well.

So. It's not too late. Please spend an hour and write up the Inofity Implementation FAQ? You probably remember and fully understand what all of our objections are and I know that you have explanations and rebuttals at hand.

Please? Something like:

q: Why does it use an ioctl multiplexer

a: Because ...


I haven't done a detailed review of the patch in months and I intend to do another soon. That FAQ will help! When I ask more silly questions we can update it, so those questions will never again be asked.

I know it's unusual process-wise, but inotify is an unusual feature.

Robert replied that he'd already written up such a document, the first time Andrew had asked for it; and published it. Arnd was not 100% satisfied that this doc answered the objections, and Robert said he was happy to continue adding questions and their answers to this document until everyone was satisfied. The discussion began to peter out at this point, and died shortly thereafter.

3. Linux 2.6.12 Released

17 Jun 2005 - 23 Jun 2005 (24 posts) Archive Link: "Linux 2.6.12"

Topics: CREDITS File, Kernel Release Announcement

People: Linus TorvaldsKeith Owens

Linus Torvalds announced Linux 2.6.12, saying:

As some people may have noticed already, 2.6.12 is out there now.

The full ChangeLog ended up missing, because I only have the history from 2.6.12-rc2 in my git archives, but if you want to, you can puzzle it together by taking the 2.6.12 changelog and merging it with the -rc1 and -rc2 logs in the testing directory. The file that says "ChangeLog-2.6.12" only contains the stuff from -rc2 onward.

Included here in the email are the changes since -rc6, and as you can see from the appended diffstat, most of the things are pretty small (ie it looks like a long list, and then you look at the diffstat and realize that most of the changes end up being just a line or two).

One of the least important changes is still worth pointing out: it was discussed earlier on the kernel mailing list in another thread, but maybe people didn't notice it: the sign-off procedure was clarified to make it clear that the person signing off understands that the project - and thus the patch and the sign-off itself, of course - is public and will be archived.

This may sound silly and obvious - and it is - but it makes people more comfortable about the fact that we obviously save identifying information in the sign-off (that's the whole point), and in general people also submit things like their own email addresses in CREDITS files etc, and so nobody should be expecting any of that to be kept confidential.

I don't think anybody did, of course, but hey, this way it's explicit. So part of the new stuff is this patch:

        diff --git a/Documentation/SubmittingPatches b/Documentation/SubmittingPatches
        --- a/Documentation/SubmittingPatches
        +++ b/Documentation/SubmittingPatches
        @@ -271,7 +271,7 @@ patch, which certifies that you wrote it
         pass it on as a open-source patch.  The rules are pretty simple: if you
         can certify the below:

        -        Developer's Certificate of Origin 1.0
        +        Developer's Certificate of Origin 1.1

                 By making a contribution to this project, I certify that:

        @@ -291,6 +291,12 @@ can certify the below:
                     person who certified (a), (b) or (c) and I have not modified

        +       (d) I understand and agree that this project and the contribution
        +           are public and that a record of the contribution (including all
        +           personal information I submit with it, including my sign-off) is
        +           maintained indefinitely and may be redistributed consistent with
        +           this project or the open source license(s) involved.
         then you just add a line saying

                Signed-off-by: Random J Developer <>

just so that people are aware of this.

Btw, in case anybody ends up wondering about what the actual patches are, if you're a git user (or, more likely, not quite a user yet, but rather wondering what you can do with git), you can start off with doing

git-whatchanged -p v2.6.12-rc6..v2.6.12

and it will do exactly what you think it migth do - it shows every commit between -rc6 and the final 2.6.12 release as a patch ("-p") with the associated commit message.

Keith Owens remarked, "tar xjvf linux-2.6.12.tar.bz2, using SuSE tar-1.13.25-325.3 on IA64, reports "tar: pax_global_header: Unknown file type 'g', extracted as normal file". It does not seem to cause any problems." And Linus said:

Yes, git creates tar-archives that use the extended pax headers, and I think you need tar-1.14 to fully understand them. They should not hurt (apart from the warning) on older versions of tar.

The extended header just contains a hidden comment record that tells the git commit ID that was used to generate the tar-tree.

Because it's extracted as a regular file (instead of tar knowing that it's a comment header), you will now have a file called "pax_global_header" that has the contents

52 comment=9ee1c939d1cb936b1f98e8d81aeffab57bae46ab

in it (where "9ee1c939d1cb936b1f98e8d81aeffab57bae46ab" is the git SHA1 name of the Linux-2.6.12 commit).

So it's not entirely "harmless" in that it causes a bogus file to be created, but it's not like it's a huge problem either, and that bogus file actually does contain real information (although it's not useful unless you're a git user).

4. New Yealink USB-P1K Phone Driver For VOIP

18 Jun 2005 - 20 Jun 2005 (5 posts) Archive Link: "[PATCH] new driver for yealink usb-p1k phone"

Topics: Sound, USB

People: Henk Vergonet

Henk Vergonet said:

In an attempt to make VOIP technology usable by other members of the household, I have created a device driver for the Yealink usb-p1k phone also known as a so called Skype phone.

Basically the device consists of an usb sound-card with keyboard, LCD, speaker and will set you back for about 30 euro's.

The "sound card" is supported by the generic usb-audio driver. This driver adds support for keyboard, LED, dialtone and LCD functions.

5. Linux 2.6.12-mm1 Released

19 Jun 2005 - 23 Jun 2005 (63 posts) Subject: "2.6.12-mm1"

Topics: Disks: SCSI, Kernel Release Announcement

People: Andrew MortonBenjamin Herrenschmidt

Andrew Morton announced Linux 2.6.12-mm1, saying:

Benjamin Herrenschmidt claimed responsibility for the /proc/device-tree situation, but said there wasn't really any breakage. He went on, "The problem is that the "ofpath" script that is part of the yaboot package has a stupid bug where for some reason, when booting from SCSI (or libata in this case), it decides to check wether there are any symlinks in /proc/device-tree, and if not, decides it's broken and aborts. It doesn't actually make any use of the symlinks that were there though (and they were useless and partially broken anyway, which is why I removed them). So it's a bug in "ofpath", a bit annoying, but at the same time, you don't need to run it when changing kernels, so it's not too harmful."

There were other various compilation problems and oopses reported by various folks, and the usual debugging efforts.

6. Too Many Patches Coming In For The -mm Tree

20 Jun 2005 (3 posts) Archive Link: "-mm patch glut"

People: Andrew Morton

Andrew Morton said, "Various people have sent me things in the past few days: thanks, it is queued for now. There's a lot of stuff in -mm to go through (I'll put a summary out soon). For now, I'm concentrating on cleaning up the diffs and changelogs, poking reviewers, stabilisation, testing and, once the various subsystem maintainers (Greg) have done thir bit, merging. I'll be looking for things to drop, too - it's getting a bit crazy."

7. BTTV And V4L Maintainership

21 Jun 2005 (1 post) Subject: "V4L maintainer patch"


People: Gerd Knorr

Mauro Carvalho Chehab posted a patch, updating the MAINTAINERS entries for BTTV and V4L in 2.6, in both cases replacing Gerd Knorr with himself as maintainer.

8. Removing DevFS Configuration Option From 2.6

21 Jun 2005 - 23 Jun 2005 (12 posts) Archive Link: "[PATCH] devfs: remove devfs from Kconfig preventing it from being built"

Topics: FS: devfs

People: Greg KH

Greg KH posted a patch to remove DevFS from Kconfig, preventing it from showing up as an option during kernel configuration. Greg said if there were no big complaints, he'd post another patch to remove DevFS entirely.

9. Linux RealTime Benchmarking Framework (LRTBF) Released

22 Jun 2005 (1 post) Subject: "[ANNOUNCE] Linux RT Benchmarking Framework"

Topics: Real-Time

People: Kristian BenoitPhilippe GerumIngo Molnar

Kristian Benoit said:

As promised, we are finally releasing the Linux RealTime Benchmarking Framework (LRTBF). We hope others will find it useful and even want to add their own testsets. Generally we'll be more than pleased to add contributions to future releases.

As was explained earlier, the Linux RT Benchmarking Framework (LRTBF) is a set of drivers and scripts for evaluating the performance of various real-time additions for the Linux kernel. Specifically, the LRTBF allows measuring the overall load imposed by the RT enhancement and its ability to deterministically respond to incoming interrupts. Initially, the LRTBF was used for evaluating Ingo Molnar's PREEMPT_RT patches and Philippe Gerum's I-pipe, but by releasing it under the GPL we hope its usefullness will extend beyond those initial tests.

The LRTBF and all related information is found here:

10. New Driver For Telecom Expansion Card On ATCA Platform

22 Jun 2005 - 23 Jun 2005 (11 posts) Archive Link: "Patch of a new driver for kernel 2.4.x that need review"

Topics: FS: devfs

People: Jesper Juhl

Sebastien Bouchard submitted a 2.4 driver, to support a telecom expansion card on the ATCA platform. Jesper Juhl noticed that Sebastien's driver included DevFS support, and remarked, "In 2.6 devfs is in the process of being removed.. It'll probably stay in 2.4, but long term (if/when you move the driver to 2.6) it's probably not worth bothering with."

11. Linux Released

22 Jun 2005 - 23 Jun 2005 (4 posts) Archive Link: "Linux"

People: Chris WrightTomasz KloczkoChris Wright:Linus Torvalds

Chris Wright announced Linux, saying:

We (the -stable team) are announcing the release of the kernel which has two security fixes.

The diffstat and short summary of the fixes are below.

I'll also be replying to this message with a copy of the patch between 2.6.12 and, as it is small enough to do so.

The updated 2.6.12.y git tree can be found at: rsync://

and can be browsed at the normal git web browser: (

 Makefile                  |    2 +-
 arch/ia64/kernel/ptrace.c |   15 ++++++++++-----
 arch/ia64/kernel/signal.c |    5 +++--
 fs/exec.c                 |    1 +
 4 files changed, 15 insertions(+), 8 deletions(-)

Summary of changes from v2.6.12 to v2.6.12.1

Chris Wright:

Linus Torvalds:
Clean up subthread exec (CAN-2005-1913)

Matthew Chapman:
ia64 ptrace + sigrestore_context (CAN-2005-1761)

Tomasz Kloczko said, "Qlogic driver still is broken. Patch with minimal set of changes for this was sended to k-l few days ago. Is it something wrong with this fixes?" Chris replied, "Qlogic fix will be in the next -stable release."

12. Linux Games Author John Hall Battles Cancer

23 Jun 2005 (1 post) Archive Link: "Kernel Developers, donate money to help fight cancer!"

People: Patrick McFarland

Patrick McFarland said, "John Hall, the author of Programming Linux Games ( , and someone who I consider a friend, was diagnosed with Stage IV Melanoma, which is very life threatening. He's asking the public to help donate to the American Cancer Society (, he's trying to raise $1,000 for melanoma research. All money donated goes straight to the American Cancer Society."

By KT publication time, John has raised $4,325.00. Maybe we can bump it to $5K.

Note: John Hall should not be confused with Jon 'Maddog' Hall, author of various Linux "For Dummies" books.







Sharon And Joy

Kernel Traffic is grateful to be developed on a computer donated by Professor Greg Benson and Professor Allan Cruse in the Department of Computer Science at the University of San Francisco. This is the same department that invented FlashMob Computing. Kernel Traffic is hosted by the generous folks at All pages on this site are copyright their original authors, and distributed under the terms of the GNU General Public License version 2.0.