linaro-dev

linaro-dev@lists.linaro.org

4332 discussions

by Grant Likely

Hi all, With several more engineers working on ARM device tree support, I'm going to start collecting the status of all the work that is going on (and I know about) and posting it in a regular status report, hopefully weekly, for the next few months until the all of the major features are implemented and working on several arm platforms. I'll try to use roughly the following format: 1) latest news and status updates 2) list of tasks with current state and who is responsible for them in the same format as Launchpad blueprint whiteboards[1]. (In fact, I'll probably move much, if not all, of this into Launchpad anyway, in which case these emails will be a summary of all the blueprints. not all of us work with Linaro, but it is a useful method for tracking progress). 3) List of active engineers [1] https://wiki.linaro.org/Process/Blueprints Please read through and reply with comments/corrections. Feel free to add or remove tasks from the list I've given below. Thanks, g. 1 - Latest news --------------- - devicetree/arm on git://git.secretlab.ca/git/linux-2.6 has everything needed to turn on basic device tree support for any platform. - Similarly, u-boot just needs to have the CONFIG_OF_LIBFDT defined to turn on device tree support. - IRQ handling is still a problem and only one interrupt controller can be supported at the moment, but Lennert is working on a solution. - I've posted a patch that will allow dt and non-dt device registration to co-exist peacefully by snooping platform device registrations. I could use some feedback and testing. - hopefully the basic dt support can be merged into Nicolas' tree this week if I get a cleaned up branch pushed out for him quickly. 2 - Task status --------------- Core infrastructure: [glikely] basic infrastructure to enable dt: DONE [r-herring] Allow dtb to be located anywhere in RAM: DONE [bones] Debug dtb corruption during init: INPROGRESS [glikely] OF clock bindings: INPROGRESS [glikely] Correctly handle dtb reserved memory sections: TODO [glikely] merge basic dt support into nico's tree: TODO [buytenh] virtual irq infrastructure: INPROGRESS Remove u-boot restriction of first 16k for dtb: TODO add processing of dtb reserved regions: TODO verify dt gpio infrastructure works: TODO [glikely] Remove early validation of machine type number: INPROGRESS imx51 tasks enable dtb support in u-boot: TODO enable basic kernel dtb support: TODO enable registration of devices form dt: TODO ... add more details here, specific devices, etc... OMAP tasks [buytenh] enable dtb support in u-boot: DONE [buytenh] enable basic kernel dtb support on OMAP3: INPROGRESS enable basic kernel dtb support on OMAP4: TODO [buytenh] enable registration of devices form dt: TODO ... add more details here, specific devices, etc... qemu support tasks: Cleanup and mainline arm dtb support patches: TODO configure qemu emulation from dtb (like microblaze): TODO ... add more details here, specific devices, etc... Samsung S5PV310 tasks: [thomas-ab] enable dtb support in u-boot: DONE? [thomas-ab] enable basic kernel dtb support on S5PV310: DONE? [thomas-ab] make gpio driver dt-aware: TODO [thomas-ab] Register devices from dt: INPROGRESS ... add more details here, specific devices, etc... Tegra tasks: [glikely] enable dtb support in u-boot: DONE [glikely] enable basic kernel dtb support on Harmony board: DONE [glikely] probe irq controllers from device tree: TODO [glikely] make gpio controller driver dt aware: TODO [glikely] register devices from dt: INPROGRESS ... add more details here, specific devices, etc... Versatile tasks: enable dtb support in u-boot: TODO [jk-ozlabs] enable basic kernel dtb support on versatile: DONE [jk-ozlabs] probe clocks from dt: DONE [jk-ozlabs] register devices from dt: DONE decode irqs from dt: TODO Versatile express tasks: [lorenzo-pieralisi] enable dtb support in u-boot: DONE? [lorenzo-pieralisi] enable basic kernel dtb support on versatile express: DONE [lorenzo-pieralisi] probe clocks from dt: DONE? [lorenzo-pieralisi] probe irq controllers from device tree: TODO [lorenzo-pieralisi] register devices from dt: DONE? ... add more details here, specific devices, etc... Xilinx arm tasks: [john linn] enable dtb support in u-boot: TODO? [john linn] enable basic kernel dtb support on xilinx arm devboard: DONE [john linn] probe irq controllers from device tree: TODO [john linn] register devices from dt: INPROGRESS? ... add more details here, specific devices, etc... A15 tasks: nothing yet As far as I know, Rob's dtb anywhere patch is pending mainline, but I'll carry it in devicetree/arm until it is merged. Same for the patch that eliminates early vetting of the machine type number. Active engineering list: Thomas Abraham [thomas-ab] Samsung Lennert Buytenhek [no launchpad id] core infrastructure and OMAP Andy Green [] OMAP Shawn Guo [] imx51 Rob Herring [r-herring] clock bindings, boot code Jason Hui [] imx51 Jeremy Kerr [jk-ozlabs] clock bindings, imx51, versatile John Linn [no launchpad id] Xilinx ARM dev board Grant Likely [glikely] core infrastructure, tegra, versatile Lorenzo Pieralisi [lorenzo-pieralisi] versatile express

13 years, 10 months

Talk: Embedded GPUs at LCA

by Christian Robottom Reis

Hi there, I did a talk at LCA yesterday evening that covered Embedded GPUs and Open Source (or the lack thereof); slides are available here: https://wiki.linaro.org/ChristianReis and a video recording may be available as well later; I'll tack it into the thread if I get a URL for it. I'm happy to answer questions about the presentations or to talk about the problems and possible solutions. One interesting outcome was the idea of a one-day cross-vendor summit that we could tack onto an existing open source event; Plumbers would be an interesting event for that and if enough people are interested I'd be willing to put effort into helping make that happen. Thanks to Jesse and Paul for the help they put into preparing for the presentation. Your comments welcome, -- Christian Robottom Reis | [+55] 16 9112 6430 | http://launchpad.net/~kiko Linaro Engineering VP | [ +1] 612 216 4935 | http://async.com.br/~kiko

13 years, 10 months

Re: problem with linaro-media-create

by Guilherme Salgado

On Mon, 2011-02-07 at 16:55 +0000, David Gilbert wrote: > On 7 February 2011 16:50, Guilherme Salgado > <guilherme.salgado(a)linaro.org> wrote: > > In lmc we already have some code which checks if a given utility is > > present and if not, install (via apt-get) a package which provides that. > > We could extend it to check for a specific version and try to install a > > version greater or equal that (can we tell apt-get to install only if it > > finds a package whose version is greater than X?). > > > > When it fails to install, we would direct the user to the README, which > > has pointers for the things that are not available in the usual > > channels. > > Be careful about forcing the use of apt to install a newer package than is > available in the distro; please check whether there is a working qemu first in > case the user has got one from another source. > Currently the easiest way on Lucid of getting lmc > to work is to copy a qemu-arm-static binary from a natty install; not a pretty > solution but it works. Ok, so ISTM that to keep this solution working we'll need to do something like running qemu-arm-static with no arguments and parse its output for the version string. Unless there's another way to easily trigger the bug without running something inside the chroot (I don't want to wait until we have the chroot ready to find out that the existing qemu-arm-static doesn't work). -- Guilherme Salgado <https://launchpad.net/~salgado>

13 years, 10 months

[PM] cpufreq test

by Vincent Guittot

Hi vishwa, I have passed cpufreq-bench on my platform. The results below have been reached with a sampling rate of 20ms and a sampling_down factor set to 10. I have a question about the sampling_down feature : The ondemand delay value is calculated before calling dbs_check_cpu, but dbs_check_cpu can modify rate_mult. This implies that if the rate_mult is set to 1 in dbs_check_cpu because we set something else than the max frequency, the next sampling period will be sampling_rate*sampling_down_factor (the one used for max freq) but the frequency can be the lowest one. We should rather calculate the delay after dbs_check_cpu or i miss something ? Vincent Test results with current implementation, 20ms sampling rate, sampling_down = 10: _round 1: 59.14% _round 2: 56.87% _round 3: 77.31% _round 4: 88.44% _round 5: 89.56% _round 6: 83.17% _round 7: 89.87% _round 8: 94.33% _round 9: 92.98% _round 10: 91.12% Test results when the delay is calculated after dbs_check_cpu, 20ms sampling rate, sampling_down = 10 : _round 1: 86.75% _round 2: 91.18% _round 3: 93.23% _round 4: 96.70% _round 5: 95.87% _round 6: 96.04% _round 7: 97.39% _round 8: 97.37% _round 9: 98.88% _round 10: 97.78%

13 years, 10 months

OpenOCD support for OMAP4/ PandaBoard

by Nicolas Pitre

I just stumbled across this: http://article.gmane.org/gmane.comp.debugging.openocd.devel/15584 And the mentioned clock issue seems to be fixed here: http://article.gmane.org/gmane.comp.debugging.openocd.devel/15619 This is excellent news, and I think we ought to support this at some point. Nicolas

13 years, 10 months

Notes & Actions: Linaro Graphics Working Group - Feb 02, 2011

by Alexandros Frantzis

Hi, notes and actions from our Wednesday graphics working group call are available on the wiki: + https://wiki.linaro.org/WorkingGroups/Middleware/Graphics/Notes/2011-02-02 Details about when and where of this meeting can be found here: + https://wiki.linaro.org/WorkingGroups/Middleware/Graphics#Weekly%20Public%2… Summary ======= * Cairo GLES work progressing: one patchset accepted, one WIP patchset in review. * New version of glmark2 (11.01) uploaded to natty universe. * Investigation on qtwebkit traces ongoing. * In contact with GLEW upstream to discuss GLES support. * IRC connectivity issues for Samsung Bangalore assignees being worked on by IT department using a clean-room approach. Action Items ============ * Everyone to report access status for git.linaro.org. * Jesse to ensure that we have valid conference call numbers for all members of the gfx WG (in progress). * Jesse to keep track of nux licensing issues. * Everyone to send Jesse a list of the licenses of the projects they are working on (both new and existing). Thanks, -- - Alexandros

13 years, 10 months

Minutes and actions Multimedia working group meeting on 1st Feb 2011

by Sachin Gupta

Hi All, Please find enclosed link to minutes and actions for multimedia wg meeting on 1st Feb 2011. https://wiki.linaro.org/WorkingGroups/Middleware/Multimedia/Notes/2011-02-01 Thanks Sachin

13 years, 10 months

Minutes & actions: Toolchain Working Group meetings of Jan 31 & Feb 02, 2011

by Mounir Bsaibes

Greetings, Enclosed you'll find a link to the agenda, minutes and actions from the Linaro toolchain working group weekly meetings of Feb 03 & Jan 31, 2011. https://wiki.linaro.org/WorkingGroups/ToolChain/Meetings/2011-01-31 https://wiki.linaro.org/WorkingGroups/ToolChain/Meetings/2011-02-02 == Summary == * Benchmark approach continues to solidify ( https://wiki.linaro.org/WorkingGroups/ToolChain/1105GCCPerformancePlan) * Benchmark suites has been chosen (DENBench, Telecom, CoreMark) * Infrastructure is ready to start historical builds * Swing-modulo-scheduling based performance work is continuing * qemu-linaro is almost ready for it's first release. Two rcs have been put out so far. * Perparing to get the string routines into Ubuntu before the feature freeze

13 years, 10 months

Minutes & Actions: Kernel Working Group meeting of Jan 31, 2011

by Mounir Bsaibes

Greetings, Enclosed you'll find a link to the agenda, minutes and actions from the Linaro kernel working group weekly meeting of Jan 31, 2011. https://wiki.linaro.org/WorkingGroups/KernelConsolidation/Meetings/2011-01-… === Summary === * Rebased Linaro tree on 2.6.38-rc1. * Device tree * Have engineers on Samsung, Beagle, Panda, Tegra, Versatile Express and handling backward compatibility for common devices. The goal is to get device-tree turned on in most/all builds. Need device-tree work for standard clock sources. * Close to getting core device-tree patches into Nicolas's tree, but fighting device-tree corruption bug on Versatile Express. Also chasing down an issue that shows up only on qemu. Have a patch that finesses clock sources by requiring a couple of device-tree APIs from SoC-specific clock code, but which can be superseded by real clock support. * Working to allow driver to retrieve device-specific information from device tree in order to make progress while core device-tree code is being refined. The kernel must still be passed a full device tree, but this allows SoCs to implement device tree incrementally as core device-tree support is put in place. Mentor and Xylinx also contributing to device tree. * Device tree support for Samsung platforms. Getting device-support information from device tree. Will post device tree DTS file to ensure that we all agree on format. Grant suggests making DTS filename reflect platform, and also using "include" facility to reduce duplication. * Some clock work in progress for device tree. Locking issues being addressed. * Looking to upstreaming TI BSP, figuring out the best way to get all 2,000 patches upstream. * Linaro Android kernel * Now boots, pushed changes to Linaro git tree. * Reviewing Andorid patches and looking to find minimal set of patches to get Android UI running (ashven and binder). * Pushing second chunk of clock-related patches up to -tip via Thomas Gleixner. * Suspend-blocker interaction with first chunk previously pushed will require some application interaction, but this is all a bit speculative given lack of suspend blockers in mainline. * Working on uboot SPL for OMAP4, making a mini-uboot from uboot to replace xloader. Making clock configuration independent of CPU frequency. Some roadblocks in eMMC support -- excessive memory footprint. * Chasing down tightened real-time checks, which cause failures when forward porting TREE_RCU priority boosting to the 2.6.38 series.

13 years, 10 months

New "linaro-kernel" mailing list

by Paul E. McKenney

Hello! We have created a new linaro-kernel mailing list for kernel-specific discussions. Of course, discussions that are of general interest should continue on linaro-dev. The new linaro-kernel mailing list is open to all comers, both inside and outside Linaro, and the archives are publicly accessible. To join, please point your browser to the following: http://lists.linaro.org/mailman/listinfo/linaro-kernel Thanx, Paul

13 years, 10 months

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

linaro-dev