LKML Summary Podcast

Audio: COMING SOON

For Monday, December 14th, 2009, I’m Jon Masters with a summary of today’s LKML traffic.

In today’s issue: cpu_clock, first time posting, ponies, and the VM.

cpu_clock. David Miller and Peter Zijlstra debated cpu_clock() and how NMI safe it may or may not be. Apparently, on SPARC systems (where multiple IRQ levels can exist), David emulates a kind of NMI through running code at different IRQ levels (14 and 15), but that approach isn’t safe with respect to regular calls to local_irq_save/restore because those also affect IRQ level. David thinks it’s a bug to be calling those functions from cpu_clock in the first place and that the code isn’t fully NMI safe, while Peter is not sure. David took the time to also provide a number of useful insights for how various platforms might usefully help to fix cpu_clock.

First time. Someone called Masa started a mail thread entitled “How to upload our driver to “kernel.org” in which they requested advice from the community concerning the best way to work with them in getting their driver into the official kernel, refered to as “kernel.org” in the email message. Several people provided pointers to documentation contained within the kernel source in the “Documentation” directory.

Ponies. The nouveau code finally having been merged (in a thread entitled “drm nouveau pony for Xmas” posted by Dave Airlie), various comments were made, including a success report from Benjamin Herrenschmidt, who was able to specify the correct additional firmware bits (nouveau in-tree has been modified to use the standard kernel request_firmware infrastructure) via CONFIG_EXTRA_FIRMWARE_DIR and CONFIG_EXTRA_FIRMWARE for one of his PowerPC systems. He posted an (big) endian fix for the 6600-series chipset. Dave took the time to followup with his assertion that he would post the VMWare DRM driver to the staging tree next by posting the code for the staging tree.

VM. Rik van Riel and Mincham Kim had a conversation about the behavior of the page reclaim code when there are a large number of tasks active in the reclaim code. Lock contention is one issue, but so is the reclaim algorithm attempting to free up to SWAP_CLUSTER_MAX pages per task. Then there is the issue that Mincham was concerned about, namely that heavy VM pressure coupled with a large number of tasks in reclaim could result in a large number of tasks in NR_UNINTERRUPTIBLE and thus a subsequent OOM would target innocent others. Rik posted a patch limiting the number of tasks allowed to be running page reclaim code from a given memory zone simulataneously to 8.

In today’s announcements: Linux 2.6.31.8 and 2.6.32.1. Greg Kroah-Hartman announced the release of the latest stable kernel versions. These are based on patches that Greg had posted over the previous weekend, and included a large number of ext4 updates, amongst other things.

The latest kernel release is 2.6.32.

These is a report of incorrect atimes being generated for files from Petr Titera.

Stephen Rothwell posted a linux-next tree for December 14th. Since Friday, the net-current tree lost its conflict, the md-current tree gained a conflict against Linus’ tree, the 52xx-and-virtex tree gained a build failure for which Stephen applieda patch, the pci tree lost its conflicts, the kbuild tree lost its merge fix, the acpi tree gained a build failure so the version from Friday was used, the cpufreq tree lost its conflict, the devicetree tree gained a build failure for which Stephen reverted several commits, the limits tree gained 3 conflicts against Linus’ tree, the omap_dss2 tree lost its conflict, the percpu tree gained a conflict against the tip tree, the tty tree lost its conflict, the usb tree lost its conflict and build failure, and the staging tree lost its conflicts, while Stephen repeated his usual “call for calm”, asking that readers not post 2.6.34 patches until after 2.6.33-rc1. The total subtree count remained unchanged at 155 trees for Monday.

Finally today, Linus Torvalds ranted a little in response to Rafael J Wysocki’s latest asynchronous suspend/resume patches, suggesting that Rafael is “overdesigning” the API, and that “[t]o a first approximation, THE ONLY THING THAT MATTERS IS USB”. That maybe true, but a generic interface will certainly provide for longer term alternatives also.

That’s a summary of today’s Linux Kernel Mailing List traffic, for further information visit www.kernel.org. I’m Jon Masters.

Folks,

Just a note that the (Xen) guest kernelpodcast.org is running on occasionally OOMs, and that this is being fixed with a virtual RAM upgrade. Meanwhile, on the few occasions this has happened, it has taken down MySQL, leading to crashed database tables. Wordpress then assumes it has just been newly installed…and one ends up with a database filled with a new “test page”, “test post”, “test comment”, etc. for each time someone tries to open the site. Not a particularly helpful situation.

If anyone knows how to tell wordpress, in case this ever happens again, “I am really installed but MySQL might have crashed uncleanly and need a manual myisam check, and under no circumstances should you flood the database with crap”, I would love to hear from you

Jon.

Audio: http://media.libsyn.com/media/jcm/linux_kernel_podcast_20091213.mp3

Wondering if the latest Linus kernel tree builds? Why not follow @kernelbuild on twitter. I’ll be refining the format and adding additional tracking to it over time, and I welcome feedback. It builds every 3 hours at the moment.

For the weekend of December 13th 2009, I’m Jon Masters with a summary of today’s LKML traffic.

In today’s issue: ATA, BFS, dm and md, FBAC-LSM, idle disks, nouveau, and the vger outage is now complete.

ATA. Apparently the long-awaited general support in hardware for 4K sector sizes is now becoming a reality. Mathew wilcox, James Andrewartha, and Peter Anvin all had things to say about this, especially James, who posted a number of links to manufacturer specific documents on the matter.

BFS. Con Kolivas announced that version 0.311 of his “Brain Fuck Scheduler” (BFS) is now available for kernel release 2.6.32. He included a summary of a number of changes, including code to measure cache locality and determine the best task to wakup (complete with a preference list).

dm and md. Alasdair Kergon and Neil Brown both announced updates for dm and md respectively. Both featured barrier updates (Alasdair announcing that dm now supports barriers on all devices, while Neil Brown noted that barriers are supported on all RAID levels – including RAID5 now). It is now possible to convert a 2-disk RAID5 md to RAID1 and vice versa (as possible before).

FBAC-LSM. Z. Cliffe Schreuders posted an initial version of a new LSM (Linux Security Module) intended to limit applications based on the specific features that each provides. He cites a number of items of documentation produced as part of his PhD research, including a comparison to SELinux. More will be available in his LCA (LinuxConf AU) presentation on the topic.

Idle disks. Mathew Garrett posted an RFC patch intended to add an event on block device idle change occurance. The idea is that userspace can monitor when disks are becoming idle (and when not) in order to adjust the power management policy on the fly in accordance with usage.

Nouveau. Following various ongoing debate (especially stired by Linus in response to earlier DRM tree updates, as also reported on LWN), Dave Airlie posted a tree entitled “drm-nouveau-pony” in an email thread entitled “drm pony for Xmas”. In the posting, Dave indirectly references the reasoning behind the driver not being available in mainline sooner (that it requires firmware or other code extract from the device, which isn’t included), notes that the driver alone is 36K lines of code (”bigger than most subsystems we carry, so hopefully ppl realise the monumental scale of writing a driver for these things”), and adds that he’ll be pushing the VMWare virtual GPU KMS driver “early next week” for good measure. There are some other legal topics being discussed surrounding this whole issue: refer to LWN for detail.

Vger. Remember that the mailing list server vger.kernel.org was moved over the weekend, and so consequently list traffic was significantly reduced. On a related note, David Miller noted in the thread discussing kernel names that he “should just remove” the hard-coded block list of words that cannot be said on LKML since the list does “dynamic spam filtering these days”.

In today’s announcements: Git version 1.6.5.6. Junio C Hamano announde version 1.6.5.6 of the GIT SCM as used by the kernel development community has been released. This version includes a number of minor fixes, as well as the removal of the “post-upload-hook” script run in response to “git fetch” due to some security concerns.

The latest kernel release is 2.6.32.

Stephen Rothwell posted a linux-next tree for December 11th. Since Thursday, the origin tree lost its build failure, the microblaze tree lost its conflict, the ext4 tree lost its build failure, the rr tree lost its conflict but gained a build failure for which Stephen applied a patch, the trivial tree lost its conflicts, and the usb tree still has its build failure for which he reverted a commit. The total number of sub-trees remained steady at 155, and Stephen repeated his usual “call for calm” in not pushing updates intended for 2.6.34 until after 2.6.33-rc1.

That’s a summary of today’s Linux Kernel Mailing List traffic, for further information visit www.kernel.org. I’m Jon Masters.

Audio: http://media.libsyn.com/media/jcm/linux_kernel_podcast_20091210.mp3

For my 28th birthday (December 10th 2009), I’m Jon Masters with a summary of today’s LKML traffic.

In today’s issue: BKL, email clients, RCU, and VM.

BKL. Arnd Bergmann posted some compat_ioctl cleanups that also included BKL ellimination. He noted though, that some of these might want to wait for 2.6.34 if Linus felt that they should go through a linux-next cycle.

Email clients. Alan Jenkins posted some helpful updates to the Documentation/email-clients.txt file intended to clarify the situation for those who are using the “Thunderbird” mail client, I assume in light of the recent changes in behavior that have been reported, for which a fix may be forthcoming (and for which Jim Owens noted he had already taken measures to ensure that his modified config would continue to work after any “fix”).

RCU. Thomas Gleixner posted a 9 part patch series entitled “Fix various __task_cred related invalid RCU assumptions” in which he pointed out a number of incorrect assumptions (that happened to be true, but were not generally correct nor guaranteed to hold true forever) in RCU usage along with fixes.

VM. Hiroyuki Kamezawa posted a number of patches for the VM, including several intended to reduce a lock contention problem in vmscan that Larry Woodman had noticed, and several more that introduced an explicit entry for the number of swap entries in use by a given process within its /proc/self/status file.

In today’s announcements: Git version 1.6.6.rc2. Junio C Hamano announced the latest release of the GIT SCM as used by the kernel development community. This update had a large number of minor fixes. On a related note, do take a look at the excellent GIT Manual, available in PDF format. I love it.

Userspace RCU 0.3.2. Mathieu Desnoyers announced release 0.3.2 of his userspace RCU library, “which includes a complete rework of s390/s390x uatomic_* operations”. It is available from http://www.lttng.org/urcu.

The latest kernel release is 2.6.32.

Andrew Morton posted an mm-of-the-moment (mmotm) for 2009-12-10-17-19.

Greg Kroah-Hartman posted a series of review patches for the forthcoming 2.6.31.8 and 2.6.32.1 stable kernels, including a large number of ext4 fixes, and an obscurely unlikely stack corruption on 2.6.32.

Ingo Molnar reported a regression in the SLOB allocator code on an IA32 system for which he posted a complete kernel config. The system hung without any usable logs available.

Jens Axboe dissected an issue with Nehalem-EX failing to boot due to a faulty commit entitled “x86: Move find_smp_config() earlier and avoid bootmem usage”.

Xose Vazquez Perez pointed out some recent benchmarks from Phoronix in which, allegedly, “Solaris is 2.4X times faster than Linux” (in some configuration) when measuring OpenSSL performance. As was noted indirectly by Andi Kleen and others, these benchmarks were performed using vendor kernels (Fedora and Ubuntu – although apparently the tests were also done on a vanilla 2.6.32) and vendor compiled OpenSSL binaries that likely have issues. Andi said the best thing was to report any regressions (especially with OpenSSL) to Fedora directly as they were unlikely to be a kernel problem.

Stephen Rothwell posted a linux-next tree for December 10th. Since Wednesday, the origin tree gained a build failure for which Stephen applied a patch, the powerpc tree lost its conflict, the ext4 tree gained a build failure for which Stephen reverted a commit, the cpufreq tree lost its conflict, the trivial tree gained a conflict against the net-current tree, the tip tree lost its conflicts, and the usb tree still has its build failure for which Stephen reverted a commit. The total sub-tree count remained steady at 155 trees in the latest compose, and Stephen repeated his usual “call for calm” the folks not push patches intended for 2.6.34 until after 2.6.33-rc1.

That’s a summary of today’s Linux Kernel Mailing List traffic, for further information visit www.kernel.org. I’m Jon Masters.

Audio: http://media.libsyn.com/media/jcm/linux_kernel_podcast_20091209.mp3

For Wednesday, December 9th 2009, I’m Jon Masters with a summary of today’s LKML traffic.

In today’s issue: Audio, hardware breakpoints, kbuild, and tracing.

Audio. Mark Brown, Linus Torvalds, Alan Stern, and others had a conversation concerning audio chipsets requiring reference voltages in which the hardware isn’t wired with a convenient reference to use and so various ramping up and down is required at suspend and resume time. Apparently, a delay is necessary in order to avoid a loud noise coming from the sound device on such hardware.

Hardware Breakpoints. Frederic Weisbecker noted a design flaw with the existing hardware breakpoint code. It turns out that the current design, which allows one to modify an existing breakpoint entry, does so by first unregistering and then re-registering the modified form thereof. This is racy with respect to another task creating a similar breakpoint. Frederic’s fix is to instead mark a breakpoint disabled during modification, then re-enable it.

Kbuild. Nir Tzachar pointed out some issues with the new ncurses based menu system that is being worked on. Specifically that lxdialog/check-lxdialog.sh always includes the “wide” (ncursesw) version of the library even if only the “narrow” version is needed, as in the case with the kernel. This can result (on systems with both versions installed) in non-color menus. Nir’s fix is essentially to ignore the “wide” version since it is not being used.

Tracing. Tim Bird posted a variant of the ftrace function tracer (build upon it) that can be used to filter on function duration, especially useful for embedded devices that wish to measure function duration during boot.

In today’s announcements: util-linux-ng version 2.17-rc2. Karel Zak announced the second RC of the forthcoming 2.17 release of util-linux-ng. As Karel previously mentioned, this version introduces a number of overhauls.

The latest kernel release is 2.6.32.

Americo Wang noted a regression in the current git tree, in the VM. He can’t reproduce it at the moment but it triggers in swap_free. Hugh Dickens said it looked like something had corrupted the start of a page table.

Stephen Rothwell posted a linux-next tree for December 9th. Since Tuesday, the omap tree lost its conflict, the samsung tree no longer needed a fixup and lost all its conflicts, the powerpc tree gained a conflict against Linus’ tree, the ext3 tree lost its conflict, the kbuild tree inherited a merge fixup from the net tree, the kvm tree lost its conflicts, the net tree lost its merge fixups, the kgdb tree lost its conflict, the slab tree lost its build failure, the omap_dss2 tree gained a conflict against the omap tree, the tip tree gained a conflict against Linus’ tree, the percpu tree lost 2 conflicts and its merge fixup (due to changes in the kgdb and tip trees), the hwpoison tree lost its build failure, the sysctl tree lost its conflicts, the tty tree gained a conflict against the trivial tree, the usb tree gained a build failure for which Stephen revered a commit, and the staging tree gained a conflict against Linus’ tree. The total sub-tree count remained steady at 155 trees in the latest compose and Stephen repeated his usual “call for calm” that people not merge items intended for 2.6.34 until after 2.6.33-rc1.

Finally today, Johannes Stezen asked Linus why the kernel name was no longer being updated post-2.6.32. Linus eventually replied (a day later, but since this update is late anyway, I’ll add it now) saying that he updates the name totally randomly, usually after he reads something that strikes him as funny.

That’s a summary of today’s Linux Kernel Mailing List traffic, for further information visit www.kernel.org. I’m Jon Masters.

Audio: http://media.libsyn.com/media/jcm/linux_kernel_podcast_20091208.mp3

For Tuesday, December 8th, 2009, I’m Jon Masters with a summary of today’s LKML traffic.

In today’s issue: Async resume, HWPOISON, perl and vger.kernel.org is moving.

Async resume. Rafael J. Wysocki posted a patch intended to implement asynchronous *resume*, as a counterpart to the asynchronous suspend patch, which he also posted an update for (along with rollback support to undo a suspend if part of it fails to complete). The main issue, as Rafael mentions, is in locking and knowing which devices should wait for which other devices, in what order, and handling this in reverse during wakeup. The patches were only “lightly tested”, according to the posting by Rafael.

HWPOISON. Andi Kleen posted some updates to HWPOISON, including a nifty feature entitled “soft offlining” in which pages are selectively offlined without killing a process (when they have too many corrected errors), implemented through userspace policy on kernel reporting.

Perl. Rob Landley posted a series of three patches intended to remove perl from the kernel build system and consequently do away with the dependency upon it. Not to play favorites here, but this author is extremely happy to see any use of perl go away, especially within the kernel community. There are of course a few ancillary functions that will need cleanup after Rob’s work.

Vger. The mail server that hosts the LKML will be taking an outage over the weekend while the server is relocated from one co-location facility to another. Consequently there will at least be downtime on the 12th and 13th, giving everyone an opportunity to take a rest from the LKML In noting the move, David Miller added that the IP address of the server will change.

In today’s announcements: Michal Marek posted a patch formally taking over kbuild maintainership from Sam Ravnborg, who no longer has time for it. Almost immediately, Michal posted his first round of updates, for kernel 2.6.33.

The latest kernel release is 2.6.32. Patches continue to accumulate in Linus’ tree. The latest pull includes DRBD as well as other gems. There was some good news also on the latest IO Controller (V4) patches, which Vivek asked to be repeated now that the patches have hit the blk tree and are in 2.6.33. Alan D. Brunelle (HP) stated that he would repeat his runs with the update.

Andrew Morton posted an mm-of-the-moment (mmotm) for 2009-12-08-17-45.

Greg Kroah-Hartman announced the latest stable kernels: 2.6.27.41 and 2.6.31.7. These contain a number of fixes. Note that there was a 2.6.27.40 kernel for a few hours, which missed a single fix for a driver that many users likely do not have installed.

Stephen Rothwell posted a linux-next tree for December 8th. Since Monday, the samsung tree needed a fixup for left over merge conflict artifacts, the nfs tree lost its built failure, the kbuild tree lost its conflict, the input tree lost its build failure, the block tree lost its build failure, the pcmcia tree lost its conflicts, the slab tree lost its build failure but gained another so the version from Friday was used, the percpu tree gained a build failure due to an interaction with the tip tree for which a patch was applied, and the hwpoison tree gained a build failure so the previous days’ version was used. The total sub-tree count remains steady at 155, and Stephen repeats his usual call for folks not to push items intended for 2.6.34 until after 2.6.33-rc1.

That’s a summary of today’s Linux Kernel Mailing List traffic, for further information visit www.kernel.org. I’m Jon Masters.

Audio: http://media.libsyn.com/media/jcm/linux_kernel_podcast_20091207.mp3

For Monday, December 7th, 2009, I’m Jon Masters with a summary of today’s LKML traffic.

In today’s issue: Ceph, errors, perf and VM.

Ceph. Sage Weil requested that Linus pull his “Ceph” distributed file system client, which has “made a half dozen rounds on linux-fsdevel, and has been in linux-next for the last month or so. Although review has been sparse, Andrew said the code looks reasonable for 2.6.33.”. Ceph apparently is intended for both high availability (HA) and reliability by having no single points of failure and strong data consistency between clients using replicated data stored across N storage nodes. For more, visit: http://ceph.newdream.net.

Errors. Roel Kluin posted a number of patches changing various use of PTR_ERR to use the correct pointer in a number of places within the kernel.

Perf. Mitake Hitoshi is working on a new round of updates for “perf lock”, the ability to profile the performance of spinlocks. He asked Steven Rostedt and Tom Zanussi whether raw_field_ptr() supports __data_loc (which is used to determine that a spinlock has been acquired in the “perf lock” code). He posted a temporary hack patch, and also noted that the meta-data surrounding held locks currently doesn’t include a numeric identifier, making it hard to tell the difference in “perf lock” between an rq lock on CPU0 and CPU1. Finally, Mitake pointed out a lack of information being provided by the trace events component even though it’s available (e.g. lock held time data). On a related note, Frederic Weisbecker sent posted some “urgent” hw-breakpoint fixes and asked that they be pulled.

VM. Al Viro sent one of his classical round of updates, this time for the VM implementation of “do_mremap”. Mostly, his code split out existing stuff into helper functions and added some cleanups. It was light on the usual sorts of “discriptive” language, however Not even one “turd” in there!

Finally today, Thomas Gleixner noted that a one line change to a header file (including a reference to an unused “extern int bla(void)” on an x86_64 system running the GCC 4.x compiler suite seemingly generated a bunch of “random” code changes in the assembler output. He (and Peter Zijlstra) provided other examples and Thomas wondered aloud if there were ways to turn off this “random number generator” or otherwise figure out what the heck is going on with GCC. Jakub Jelinek (of the GCC tools team) noted that the source had changed, which can result in various indexed hash tables changing (with the addition of the new variable reference), and could well explain this. He offered to look further if Thomas repeated his experiments with GCC 4.4 or 4.5.

The latest kernel release is 2.6.32.

Andrew Morton posted an mm-of-the-moment (mmotm) for 2009-12-07-17-01.

Someone identifying themself as merely “walt” (and who doesn’t bother to use a full name in other postings either, according to Google) noted that a recent commit from Frederic Weisbecker affecting hardware breakpoints seems to have also broken gdb.

Stephen Rothwell posted a linux-next tree for December 7th. Since Friday, the microblaze tree lost a conflict, the nfs tree lost a build failure but gained another for which a patch was applied, the pci tree gained 2 conflicts against Linus’ tree, the net tree lost its conflict and a merge fixup but gained a build failure due to an interaction with the kbuild tree for which Stephen applied a patch, the input tree gained a build failure for which Stephen reverted a commit, the block tree lost its build failure but gained another for which Stephen applied a patch, the pcmcia tree lost its conflicts, the slab tree gained a build failure so the version from Friday was used, the trivial tree gained a trivial conflict against the v4l-dvb tree, the i7core_edac tree lost its merge fixup and the tip tree lost 5 conflicts. The total subtree count remained steady at 155 trees, and Stephen repeated his usual call for calm in not pulling items intended for 2.6.34 until 2.6.33-rc1 has been released. The reason for so many conflicts at the moment in large part stems from Linus beginning to pull things into his treef for .33.

That’s a summary of today’s Linux Kernel Mailing List traffic, for further information visit www.kernel.org. I’m Jon Masters.

Audio: http://media.libsyn.com/media/jcm/linux_kernel_podcast_20091206.mp3

For the weekend of December 6th, 2009, I’m Jon Masters with a summary of today’s LKML traffic.

In today’s issue: BKL, cgroups, constants, sh, and tracing.

BKL. Thomas Gleixner posted a series of “BKL removal” patches for drivers, core, and individual architectures. As a side effect of this work, Thomas introduces the raw_spinlock as used in the RT kernel, which on non-RT platforms are identical to the non-raw versions (the non-raw versions map directly onto the new raw_spinlock). The Big Kernel Lock will outlive 2.6.33, but hopefully Thomas and others will decrease the in-kernel usage thereof.

Cgroups. Ben Blum (CMU) posted an interesting series of patches intended to implement modularization support for “building, loading, and unloading subsystems [cgroups] as modules”. He notes this is an iteration of a previous patch for which he provides a link. On a tangent, Jens Axboe noted that the latest block bits for 2.6.33 include the cgroup-based blk-io controller (in addition to other cool stuff, such as the final mainlining of DRBD – the Distributed Replicating Block Device that has existed out-of-tree and has been shipped by at least one Linux vendor for a number of years).

Constants. Emese Revfy posted thirty one patches intended to make various structure elements constants, through use of the “const” keyword. The idea is that many kernel structures are already intended to be filled with consts because they contain mostly unmodified function pointers, and explicitly stating this helps the compiler to enforce it.

Super-H. Paul Mundt posted an update concerning features forthcoming for the Super-H (sh) architecture in 2.6.33. A lot of the work concerns newer parts with “relatively more complex and configurable cache features/topologies”, as well as an ongoing effort to migrate away from oprofile towards perf events.

Tracing. Jiri Olsa, obviously interested in speeding up kernel build time, took a look at Steven Rostedt’s “recordmcount.pl” scripts, that are used at compilation time to annotate kernel code with mcount callsite metadata. He decided to experiment with replacing the perl code with a compiled binary and has already achieved a two minute kernel compile time speedup. He isn’t sure whether he will continue with the work, but wanted to share his data anyway. Steven Rostedt followed up saying that John Reiser had already sent him some code that “convert[s] the entire shebang” and noting that he “like[s] that approach better”, and that he had just gotten back to looking into it.

In today’s announcements: Git version 1.6.5.5. Junio C Hamano announced the latest version of the GIT SCM favored by the kernel community has been released. Since the previous version, a number of issues have been addressed, including path handling “fixes”, diff behavior fixes, and so forth.

LTTng 0.181. Mathieu Desnoyers announced release 0.181 of LTTng targeting kernel 2.6.32 is now available at http://www.lttng.org/.

Pahole and the dwarves 1.8. Arnaldo Carvalho de Melo posted to let everyone know that the latest versions of his ELF binary examination utilities are now available in the pahole git tree on kernel.org.

Stable kernels 2.6.27.40 and 2.6.31.7. Greg Kroah-Hartman announced the availability of review patches for two forthcoming stable kernels. Comments should be provided no later than 20:00:00UTC on Tuesday December 8th.

The latest kernel release is 2.6.32.

Andrew Morton posted an mm-of-the-moment (mmotm) for 2009-12-03-22-57.

Stephen Rothwell posted a linux-next tree for December 4th. Since Thursday, the ext3 tree lost its build failure, the nfs tree gained a build failure so the previous day’s version was used, the block tree gained a build failures so the previous day’s version was used, the pcmcia tree gained a conflict against the net tree, the irda tree gained a conflict against the net tree, the limits tree gained conflicts against the net and fsnotify trees, the tip tree gained a conflict against the limits tree, and the sysctl tree gained 2 conflicts against the net tree. The total sub-tree count remained at 155, while Stephen also repeated his previous “call for calm” in not introducing items intended for 2.6.34 before 2.6.33-rc1 has been released.

Torok Edwin posted concerning some apparent bugs in the “perf” utilities, in particular with regard to the capture and replay of existing runs leading to a segmentation fault for which an example was provided.

That’s a summary of today’s Linux Kernel Mailing List traffic, for further information visit www.kernel.org. I’m Jon Masters.

Figure: Over 100,000 downloads of the LKML podcast. These are some of the latest download statistics, and do not include readers of the transcripts, or those who downloaded the earliest versions (about 9-10K downloads) prior to the CDN.

Thanks for listening!

Jon.

Audio: http://media.libsyn.com/media/jcm/linux_kernel_podcast_20091203.mp3

For Thursday, December 3rd, 2009, I’m Jon Masters with a summary of today’s LKML traffic.

In today’s issue: The 2.6.33 merge window, oops, and perf.

Merge window. The 2.6.32 release on Wednesday signaled the openeing of the 2.6.33 merge window. There have been a number of git tree postings so far, including security updates from James Morris, and GFS2 updates (previously announced though in pre-post warnings) from Steven Whitehouse. The trivial tree has what looks like a larger number of fixes than it does, but there are certainly a good number of cleanups present in Jiri Kosina’s tree. There were also the usual crazy number of git tree updates from Ingo Molnar (which at this point take up an entire screen in mutt) – including RCU-tiny and default RCU stall-detector enabled in Kconfig, and also a new feature in user-return-facility for fast returns to kernel mode only in KVM. Grant Likely also pushed a round of OpenFirmware device tree patches that everyone should get really excited about, because that’s cool stuff.

Oops. Donming Jin posted a question concerning kmsg_dump, a function called to ensure that an oops message is written synchronously before continuing with, for example, a panic inspired crash dump. He wonders if it should be called in oops_end rather than in oops_exit, to ensure ordering.

Perf. Part of the first round of pull requests for 2.6.33 (and amongst his many other git tree updates) were some perf patches from Ingo Molnar. These include support for a new “perf kmem” tool for SLAB allocator analysis, a new “perf probe” tool to utilize kprobes to dynamically probe kernel functions, “perf bench” to run micro-benchmarks in a “uniform way”, and new hardware breakpoints asbstraction and support in the perf tools. Perf really is becomming some kind of “swiss army knife” in the kernel at this point.

Finally today, someone finally answered Michael Gilbert concerning the state of fixes for a couple of CVEs first reported in 2004. It would seem that the cryptoloop code has some “major design/implementation weaknessess” vs. the two other forms of disk encryption already shipping in-kernel.

In today’s announcements: GIT version 1.6.5.4. Junio C Hamano announced the latest release of the GIT SCM as used by the kernel community. It includes a number of fixes, mostly fairly trivial (for example that “git help” used to require the current working directory to be under GIT control.

Lsscsi. Douglas Gilbert announced the latest version of lsscsi (0.23). The latest release of this utility will assume /sys as the mountpoint for sysfs rather than poking around in /proc/mounts by default. The utility is mostly used by those wishing to understand SCSI topology in their systems.

The Linux Test Project. Subrata Modak announced that the Linux Test Project for November 2009 has been released. The latest version features a number of highlights, including cleanup and stabilization of the new build system for distributions with older kernel releases and the addition of “EXECLTP”, a “more user friendly” front end script written in python.

The latest kernel release is 2.6.32.

Stephen Rothwell posted a linux-next tree for December 3rd. Since Wednesday, the new kbuild-current tree (with a new Kbuild maintainer) has been added, the powerpc tree lost its build failure, the ext3 tree gained a build failure for which Stephen reverted a commit, the net tree lost a conflict, the mtd tree gained a conflict against the mips tree, the tip tree lost a conflict, and the workqueues tree lost its conflicts. The total sub-tree count remains stead at 154.

Stephen notes his “usual call for calm” in saying that developers should not put stuff in their git trees intended to linux-next that targets 2.6.34 inclusion until after the first 2.6.33 -rc has gone out.

That’s a summary of today’s Linux Kernel Mailing List traffic, for further information visit www.kernel.org. I’m Jon Masters.