LKML Summary Podcast

2010/07/04 Linux Kernel Podcast

July 12th, 2010 jcm No comments

Audio: COMING SOON

For the weekend of the 4th of July 2010, I’m Jon Masters with a summary of today’s LKML traffic.

In today’s issue: Linux 2.6.35-rc4, Btrfs, Defconfig kernel configs, GDB, Timekeeping, and the VM.

*). Linux 2.6.35-rc4. Linus Torvalds announced the release of Linux 2.6.35-rc4 on July 4th 2010 at 8:44pm Best Coast Time (PDT). Linus says he’s been back online for a week and is happy at the relatively small number of changes building up, “having been strict for -rc3″, in his absence. He obviously sees the increased rigidity in enforcing the merge window has been a success, and considers that there will likely be an on time 2.6.35 release, “despite my vacation”. Linus says his vacation was very enjoyable and was the longest time away from the kernel in many years – apparently he did take a cellphone for email, but didn’t do any compiles while he was having “a great time under water.”

*). Btrfs. Edward Shishkin posted a rather scathing technical review of btrfs internal design, criticising variable record size allocations, file system utilization, the balancing algorithms used, and even suggesting that engineers leave the algorithm design up to academics, rather than re-inventing things for their programs. Edward performed various benchmarks and published his results in a thread entitled (variously), “Unbound(?) Internal fragmentation in Btrfs”, “Btrfs: broken file system design”, and “Balancing leaves when walking from top to down”. For his part, Chris Mason was very civil in his reply on a number of occasions, saying that he didn’t see a fundamental design problem existing in Btrfs. Edward “NACKed” Btrfs anyway for enterprise use (even though it’s been in tree for a while).

*). Defconfig kernel configs. Linus Torvalds (in a thread renamed to “ARM defconfig files”) essentially conveyed his discomfort with the continued existance of many dozens (or perhaps hundreds) of “defconfig” files in the architecture directories. These are reference files which are based upon copies of “known good” configuration files. They worked well back in the day, but as Linus says, times have changed and nobody is really making these files by hand any more without using Kconfig. So he proposes replacing them – eating the pain – with single config files per machine type that use Kconfig and source in particulars for the various chip and architecture family particulars. Russell King pointed out that this is basically what already happens, but the point of the defconfig files is to also handle stuff outside of the architecture – for example, choosing not to use certain “IDE” options on particular boards or systems – as Daniel Walker also pointed out. Daniel noted that those setting up e.g. a BeagleBoard or a Nexus One don’t really want to troll through thousands of possible kernel options if a good reference set is available to begin with. Daniel also point out a previous posting for a boolean SATisfiability solver in the kernel config. Linus thought that was interesting but ‘At the same time, “SAT solver” does scream “over-engineering failure” to me’. Linus later explained that he was looking to either kill the defconfigs or replace them with some templates and a means to generate them, but otherwise prefered them to live some place outside of the kernel.

*). GDB. David Howells posted a patch implementing GDB remote protocol support for the “p” command on FRV. The “p” command is used to transfer information about a single register, as opposed to the “g” command, that transfers data on several. But when a gdb client connects, it will attempt to use “p” or “g” and will then stick with that choice without varying. For this reason, Linus wondered aloud if using single reads would actually slow down clients connecting (since they usually will request a number of registers at a time). Jason Wessel said he had actualy done some fairly detailed benchmarking and would share his findings at a later point.

*). Timekeeping. Oleg Nesterov posted a thread entitled “Q: sys_futex() && timespec_valid()”, in which he attempted to summarize some concerns that the glibc folks were having with the Linux implementation of timespec timeouts. Ulrich Drepper replied, explaining that his point was that a negative value for tv_sec in the case of an absolute timeout should not return -EINVAL, but instead -ETIMEDOUT. He contends that a negative relative time in the 1960s is not an invalid time. Linus strongly disagreed, saying, “Ulrich – you’re wrong. Go away.” and then clarified, ‘In the end, it’s quite simple: the kernel doesn’t accept invalid timevals. And negative tv_secs are invalid. It’s that simple. If somebody gives the kernel a timeout from before the epoch [January 1st 1970], that somebody is being a total idiot. We know it’s not a valid absolute timeout, since there’s no way somebody is “waiting” for something that happened in the sixties. Yeah, yeah, maybe you’re waiting for flower power and and free sec. Good for you. But if you are, don’t ask the Linux kernel to wait with you. Ok?’ This author wonders what those still waiting for Elvis will do now that this is clarified.

*). VM. Larry Woodman posted a patch entitled “Call cond_resched() at bottom of main loo[sic: s/k/p/] in balance_pgdat()”, which handles a situation on small single CPU systems wherein a task should OOM (Out Of Memory) and call the OOM-killer, but it does not because kswapd is constantly running due to at least one system RAM zone being below the high page watermark. Larry adds a single cond_resched() call that will allow the watchdog, tasks, and OOM killer to run, freeing up the affected resources. Andrew Morton didn’t like this approach – implying he prefered something more specific than a cond_resched and waiting for the OOM killer to get chance to run – but he could live with it if there were a giant FIXME and/or some documentation at least explaining the essential nature of the specific cond_resched() call as opposed to a regular point of voluntary kernel preemption.

In today’s miscellaneous items:

*). Patrick Pannuto proposed a usleep API for the kernel to augment the existing msleep one, and be used as an alternative to udelay so as to allow the CPU to go into lower power C-states. After some dialogue between Patrick and Daniel Walker, in which Walker pointed out that some stats were needed to prove that this was power beneficial for small delays, it seemed that there was a small improvement for 50us delay values.

*). Ronny Tschuter had some issues with tracing power_start events when using the cpuidle framework with a menu governor and an cpi-based driver to handle idle states. There wer no instrumentation points in the processor_idle code, so he posted a patch, but Arjan van de Ven pointed out that the ACPI STATE type is pretty much “useless random garbage” so the posted should set their system to use mwait idle.

*). Dave Jones raised a concern with crypto and device-mapper. A potential regression was introduced somewhere between 2.6.32 and now, and the details are available in Red Hat Bugzilla 610278. Nobody replied to the posting on the list, but the Bugzilla says that one should be using LUKS, and in the case of not using it the default encryption options were changed due to a vulnerability. It is possible to mount the existing device using the instructions provided.

In today’s announcements:

*). Jeff Merkey announced the latest version of his MDB “Merkey’s Kernel Debugger” x86_64 2.6.34 07-01-2010 Release 4. It’s available on googlecode.com. There has been no community discussion thereof. Jeff also posted his Open Cworthy Libraries 07-01-2010.

*). Junio C Hamano announced Git version 1.7.1.1 is now available at: http://www.kernel.org/pub/software/scm/git/ He also announced Git 1.7.2.rc1 is available for review.

*). Karel Zak announced the latest stable release of util-linux-ng 2.18 is now available: http://www.kernel.org/pub/linux/utils/util-linux-ng/

*). Subrata Modak announced that the Linux Test Project for June 2010 has been released. http://ltp.sourceforge.net/

The latest kernel release is 2.6.35-rc4.

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

Categories: episodes Tags:

2010/06/27 Linux Kernel Podcast

July 11th, 2010 jcm No comments

Audio: COMING SOON

For the weekend of June 27th 2010, I’m Jon Masters with a summary of today’s LKML traffic.

In today’s issue: Concurrent coredumps, OpenFirmware, and Power management policy.

*). Concurrent coredumps. Edward Allcutt posted, inquiring about placing a limit on the number of concurrent process coredumps that should be allowed to take place on a system. He cited an example Apache-based webserver in which large numbers of CGI processes were crashing, each with a 150-200MB core file that needed writing to disk. He was using a custom patch that would cease dumping cores after a certain number were already concurrently taking place. Roland McGrath and Andrew Morton did not favor this approach, instead prefering either that core dumps would begin to block (but not consuming resources) after a point, or that the blkio_cgroup IO controller be used to limit the IO being consumed. Hiroyuki Kamezawa suggested that distributions like Fedora – which in that case has its own dumping tool called abrt that manages coredumps – could wire up the blkio cgroup prior to beginning the dump process.

*). OpenFirmware. Andres Salomon posted a patch implementing support for making calls into OpenFirmware on x86 OLPC XO systems. The patch works by preserving the necessary page mappings for the OpenFirmware (OFW), which remains in memory at a virtual address. Just the minimum number of mappings are retained, but this does allow calls into the firmware even after Linux has booted. It’s always been interesting to see the XO using OpenFirmware as one of the only x86-based devices doing so.

*). Power management policy. Len Brown posted an RFC patch implementing a new centralized location for userspace to express its power management vs. performance policy preferences to the kernel. In the patch, such expression occurs through the new /sys/power/policy_preference file, which contains 5 different possible levels – ranging from “max_performance”, through “balanced” (the new default), to the “max_powersave” option on the other extreme. The idea is to centralize setting scheduler, cpuidle, governor, and other options.

In today’s miscellaneous items:

*). Dave Chinner posted a 5 part patch series implementing some fixes for emergency filesystem thawing (via sysrq control).

*). Michael Kerrisk posted some man-pages text for the MADV_MERGEABLE and MADV_UNMERGEABLE flags added in 2.6.32 for use with KSM (Kernel Samepage Mapping – the kernel support for detecting duplicate pages in guest virtual machines and mapping them to a single shared page instance).

*). Paul E. McKenney concluded that it was sufficient to turn off the CONFIG_PROVE_RCU option in Fedora rawhide kernels since it’s mostly a developer tool, rather than change licensing or otherwise make it available to non-GPL modules with which it is not compatible.

*). Luis R. Rodriguez posted a script and some documentation to implement some rudimentary ASPM (a PCI extension that allows devices to go to an entirely electrically idle bus state) support. For further information: http://wireless.kernel.org/en/users/Documentation/ASPM

*). Konrad Rzeszutek Wilk posted a 19 part patch series implementing PCI pass-through for Paravirtualizaed Xen guests, using SWIOTLB support.

*). Mike McCormack wasn’t happy with the 32 (NGROUPS_SMALL) group limit on the number shown in /proc/ /status for a given process ID. He and others discussed various ways those who really want more than 32 groups assigned to a process could get the full data through various API changes.

*). Rusty Russell posted the last (hopefully) of his cpumask patches which he says now also means that everyone should be using the cpumask_functions. At least, everyone in kernel is, according to his tests on 32-bit.

In today’s announcements:

*). Mathieu Desnoyers announced that LTTng 0.218 for kernel 2.6.34 is now available. For further information: http://www.lttng.org/

*). Henrik Rydberg announced version 1.0.1 of the mtdev Multitouch Translation Library is now available (releaseed under the MIT license). mtdev does all of the necessary finger tracking pieces in userspace, and separate from the Xorg driver from which it came, as a means to further adoption. This author is still waiting for his Apple Multitouch keypad to work on a Fedora system without having to patch the kernel with a kludge. mtdev is available at: http://bitmath.org/code/mtdev/

*). Len Brown announced the Boston Linux Power Management Mini-Summit will take place concurrently with the Linux Foundation LinuxCon 2010, on the day immediately prior to the beginning of the main events, August 9th. For further information: http://events.linuxfoundation.org/

The latest kernel release was 2.6.35-rc3.

Finally today, Piotr Hosowicz wondered aloud why Linus’ git repository was not being updated, asking if it’s because he’s on vacation. As mentioned before, Linus was indeed on a (well deserved) vacation.

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

Categories: episodes Tags:

2010/06/20 Linux Kernel Podcast

July 11th, 2010 jcm No comments

Audio: COMING SOON

For the weekend of June 20th, I’m Jon Masters with a summary of today’s LKML traffic.

In today’s issue: Panic, Performance Events, Slow-work, and Timekeeping.

*). Panic. Shoichi Tamuki posted version 2 of a patch intended to fix keyboard LED blinking on panic. Existing systems will call mdelay to handle the reboot timeout post-panic, during which time the keyboard LEDs well blink. When a hypervisor is being used, those mdelay calls of 1 second or more will be implemented as spins, in order to avoid timeout accuracy slips, but the side effect is that the keyboard LEDs won’t blink properly. The patch will call panic_blink_enter() between every mdelay call, and it also fixes up the longer mdelays so that the blinking still occurs.

*). Performance Events. Nils Carlson, Andi Kleen, Eric W. Biederman, Tony Luck, and others, discussed the “Hardware Error Kernel Mini-Summit” followup in which it had been proposed to introduce a new hardware error subsystem. They pondered what (mostly) Andi saw as failings of EDAC and the need for a better way to find such things as which DIMM has failed without doing a binary search removal of individual modules (”the way of the 21st century”). Tony Luck proposed some further ideas for a generic subsystem.

*). Slow-work. Ted Ts’o reported that recent 2.6.35 kernels with an Ubuntu userspace would periodically get into a state in which large amounts of CPU time was spent in the kslowd worker threads. It turned out that this was caused by a change to the DRM/KMS code to pull polling of the display connectors into the DRM core. Reverting a specific commit fixed the issue for Nick Bowler, who had also been experiencing this problem.

*). Timekeeping. Suresh Rajashekara inquired as to what appeared to be a problem with timekeeping on his OMAP1 platform with a 2.6.29 kernel. It seemed odd that certain timers were not expiring immediately upon resume on a system that tries to spend most of its time in a suspend state (waking for 35 milliseconds every 4 seconds, apparently). Thomas Gleixner replied, saying that during such suspend operations, only the CLOCK_REALTIME based timers are kept correct (aligned to real time), whereas others won’t expire the moment the system resumes because there may otherwise be a thundering hurd problem as many timers expire at the point that the system wakes up from the suspend state.

In today’s miscellaneous items:

*). R. F. Burns inquired as to whether it was possible to “write a kernel module which, when loaded, will blow the PC speaker?”. Alan Cox replied that this wasn’t really likely, and in the absence of the root password and proper expertise, “throwing it out of the window or feeding it iron filings will work just as well.”

*). Lai Jiangshan posted a patch removing the use a default write bit with EPT page allocations under KVM virtualization. It wasn’t causing a problem now since get_user_pages is always called with write=1 at the moment.

*). Adrian Hunter posted MMC patches adding support for secure erase, trim, and secure trim – all now variants of erase in eMMC v4.4 cards.

*). Peter Zijlstra noted that the historical uses of perf_disable to prevent NMI races in the PMU code were basically now done per-arch, so he suggested that he would remove perf_disable as it did not seem to be really needed.

*). Christoph Hellwig posted the XFS status update for May 2010, in which he noted several of the important features that lands in 2.6.34 (including new inode and quota flushing code). Christoph also posted a patch (not entirely related to XFS) that removed the 4K stacks option on 32-bit x86 systems as it is deemed “too small” these days, even with now mandatory split IRQ/kernel stacks, given the depth of many kernel call chains.

*). A number of objections to the new automated addition of a “+” to the localversion for modified kernel trees, if no other is set. Mark Hills pointed out that this triggers a lengthy modpost step even when doing “casual kernel development” to test out some simple patch.

*). Dan Carpenter posted a patch that changes the output of kernel oops messages such that the previous “cut here” is replaced with a message asking for the entirity of the oops to be sent in to kernel folks.

*). Zachary Amsden (who has been working on this for some time) posted some TSC cleanup patches and documentation for KVM. This should help resolve many of the issues that have been affecting some TSC users under KVM. On that note, Hagen Paul Pfeifer sent a patch that effectively allows for deliberate speeding-up of time for certain guests for testing use.

*). Huan Ying posted a three-part “Unified NMI delayed call mechanism”, which essentially allows the deferment of certain NMI-time processing until the NMI context has been left. Ingo Molnar prefered that the solution be to re-use the existing unified NMI watchdog code. Sadly, the rest of the thread turned into a bit of a flamewar between Andi Kleen and Ingo.

In today’s announcements:

*). Jeff Merkey announced Open CWorth Libraries 06-19-2010, and ranted about wanting larger stack sizes. He also posted version 2.6.34-06-17-2010 of his “MDB” or “Merkey Debugger”. Nobody replied to any of these threads.

*). Etienne Lorrain announced version 2.8.2 of the gujin GPL bootloader. It contains several bugfixs and improvements – http://gujin.org/

*). James Morris announced the Program Schedule for the Linux Security Summit that will run in conjunction with the 2010 LinuxCon in Boston, on August 9. Further information is available at http://www.linuxfoundation.org/

*). Karel Zak announced that the second util-linux-ng 2.18 release candidate is now available. It contains lots of fixes (e.g. disable DOS mode and cylinders by default now in fdisk). Further information is available at: http://www.kernel.org/pub/linux/utils/util-linux-ng/v2.18/

*). Mathieu Desnoyers announced the release of Userspace RCU 0.4.6. The latest release includes added ARMv7l support. Further information is available at: http://www.lttng.org/urcu/

The latest kernel release was 2.6.35-rc3.

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

Categories: episodes Tags:

M	T	W	T	F	S	S
« Jun				Aug »
			1	2	3	4
5	6	7	8	9	10	11
12	13	14	15	16	17	18
19	20	21	22	23	24	25
26	27	28	29	30	31

Archive

2010/07/04 Linux Kernel Podcast

2010/06/27 Linux Kernel Podcast

2010/06/20 Linux Kernel Podcast

Random Posts

Categories

Blogroll

Recent Posts

Highlighted events

Meta

LKML Summary Podcast

Archive

2010/07/04 Linux Kernel Podcast

2010/06/27 Linux Kernel Podcast

2010/06/20 Linux Kernel Podcast

Random Posts

Tag Cloud

Categories

Blogroll

Recent Posts

Highlighted events

Meta