linaro-dev

linaro-dev@lists.linaro.org

4332 discussions

Minutes & actions: Kernel Working Group meeting of Nov 29, 2010

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 November 29, 2010. https://wiki.linaro.org/WorkingGroups/KernelConsolidation/Meetings/2010-11-… == Summary == * Focusing on completing of the blueprints Definitions (Specifications). * Started the implementation phase of the blueprints. * Working on pushing patches upstream: * u-boot patches * PIO driver * 5500 flash related code * One Android patch submitted by John was accepted into -tip * Thumb2 SMP/Uniprocessor runtime switch on Versatile Express, also on multiple ARM platforms with single kernel. * Fixed and opened bugs * Fixed two Linaro bugs on iMX51 uboot * opened a bug against toolchain code sourcery. * Packaged a kernel, 2.6.37-rc3 made all the Ubuntu sauce apply cleanly!!! packaged kernel will be done every few weeks in general. Loic does non-packaged nightly kernel builds for more frequent testing. * Working on definition document on what the Linaro tree really is: what is accepted, what it can be used for, what it really is. People seem to be getting more comfortable with the Linaro tree. Regards, Mounir

15 years

Re: [PATCH][ ARM cpu hotplug 1/2 ] extract common code for arm cpu hotplug

by Vincent Guittot

The goal of this patch is to remove as much duplicated code as possible in each platform hotplug file. I have also tried to keep in mind that current platform upstreamed code make nearly no power management in their current implementation. I have added a new interface and a new file in order to -Keep the current interface as it is. So each platform could move to common code when they want -Have a dedicated file for arm hotplug function in which we can add all code that must be executed on an arm core whatever the platform (flushing cache, SCU disabling and handling spurious wake up as a staring point). I agree that the current code is quite small for now and we can wonder if a dedicated file is useful, code might be put in kernel/smp.c file. The next step is also to add some hotplug tracepoints and I would prefer to add the tracepoints only in a platform independent code. With the common code, a new arm platform can have the hotplug feature with less than 20 lines (power management not included) and can be sure that minimal actions will be handled by the common Arm code. We can also keep the platform_cpu_die has the platform entry point and move the common part into kernel/smp.c file. We still have few duplicated code (spurious wake-up) but this seems to be acceptable. I have taken your example into account and have updated the patch accordingly Vincent On 29 November 2010 11:41, Russell King - ARM Linux <linux(a)arm.linux.org.uk> wrote: > On Mon, Nov 29, 2010 at 10:54:35AM +0100, Vincent Guittot wrote: >> This patch extracts the common code of the cpu hotplug feature across >> arm platforms. The goal is to only keep the specific stuff of the >> platform in the sub-architecture. I have created a hotplug.c file in >> the arm/common directory after studying the cpu hotplug code of >> omap2, realview, s5pv310, ux500 and tegra. I have extracted 3 main >> platform dependent functions: >> -platform_enter_lowpower which prepares the platform for low power. >> -platform_do_lowpower on which the cpu will loop until it becomes >> really plugged (spurious wake up). This function must returned the cpu >> Id in order to leave the unplug state. >> -platform_leave_lowpower which restore the platform context. > > I still do not like this patch. The only thing that is worth doing is > this. This leaves less than 256 bytes of object code in the Realview > hotplug.c, most of which is the stuff to handle the low power mode > which you haven't dealt with in your patch either. > > I see no point in adding another API on top of the already existing > and simple API. > Signed-off-by: Vincent Guittot <vincent.guittot(a)linaro.org> --- arch/arm/kernel/smp.c | 35 ++++++++++++++++++++---- arch/arm/mach-omap2/omap-hotplug.c | 28 ------------------- arch/arm/mach-realview/hotplug.c | 29 -------------------- arch/arm/mach-s5pv310/hotplug.c | 29 -------------------- arch/arm/mach-tegra/hotplug.c | 30 --------------------- arch/arm/mach-ux500/hotplug.c | 51 +++++------------------------------- 6 files changed, 36 insertions(+), 166 deletions(-) diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c index 8c19595..03042db 100644 --- a/arch/arm/kernel/smp.c +++ b/arch/arm/kernel/smp.c @@ -24,6 +24,7 @@ #include <linux/irq.h> #include <linux/percpu.h> #include <linux/clockchips.h> +#include <linux/completion.h> #include <asm/atomic.h> #include <asm/cacheflush.h> @@ -213,11 +214,13 @@ int __cpu_disable(void) { unsigned int cpu = smp_processor_id(); struct task_struct *p; - int ret; - ret = platform_cpu_disable(cpu); - if (ret) - return ret; + /* + * we don't allow CPU 0 to be shutdown (it is still too special + * e.g. clock tick interrupts) + */ + if (cpu == 0) + return -EPERM; /* * Take this CPU offline. Once we clear this, we can't return, @@ -252,14 +255,16 @@ int __cpu_disable(void) return 0; } +static DECLARE_COMPLETION(cpu_killed); + /* * called on the thread which is asking for a CPU to be shutdown - * waits until shutdown has completed, or it is timed out. */ void __cpu_die(unsigned int cpu) { - if (!platform_cpu_kill(cpu)) - printk("CPU%u: unable to kill\n", cpu); + if (!wait_for_completion_timeout(&cpu_killed, 5000)) + printk(KERN_NOTICE "CPU%u: cpu didn't die\n", cpu); } /* @@ -273,14 +278,32 @@ void __cpu_die(unsigned int cpu) void __ref cpu_die(void) { unsigned int cpu = smp_processor_id(); +#ifdef DEBUG + unsigned int this_cpu = hard_smp_processor_id(); +#endif local_irq_disable(); idle_task_exit(); + printk(KERN_NOTICE "CPU%u: shutdown\n", cpu); + complete(&cpu_killed); + +#ifdef DEBUG + if (cpu != this_cpu) { + printk(KERN_CRIT "Eek! platform_cpu_die is going to run on %u, should be %u\n", + this_cpu, cpu); + BUG(); + } +#endif + + flush_cache_all(); + wmb(); + /* * actual CPU shutdown procedure is at least platform (if not * CPU) specific */ + platform_cpu_die(cpu); /* diff --git a/arch/arm/mach-omap2/omap-hotplug.c b/arch/arm/mach-omap2/omap-hotplug.c index 6cee456..00f4190 100644 --- a/arch/arm/mach-omap2/omap-hotplug.c +++ b/arch/arm/mach-omap2/omap-hotplug.c @@ -17,35 +17,15 @@ #include <linux/kernel.h> #include <linux/errno.h> #include <linux/smp.h> -#include <linux/completion.h> -#include <asm/cacheflush.h> #include <mach/omap4-common.h> -static DECLARE_COMPLETION(cpu_killed); - -int platform_cpu_kill(unsigned int cpu) -{ - return wait_for_completion_timeout(&cpu_killed, 5000); -} - /* * platform-specific code to shutdown a CPU * Called with IRQs disabled */ void platform_cpu_die(unsigned int cpu) { - unsigned int this_cpu = hard_smp_processor_id(); - - if (cpu != this_cpu) { - pr_crit("platform_cpu_die running on %u, should be %u\n", - this_cpu, cpu); - BUG(); - } - pr_notice("CPU%u: shutdown\n", cpu); - complete(&cpu_killed); - flush_cache_all(); - dsb(); /* * we're ready for shutdown now, so do it @@ -69,11 +49,3 @@ void platform_cpu_die(unsigned int cpu) } } -int platform_cpu_disable(unsigned int cpu) -{ - /* - * we don't allow CPU 0 to be shutdown (it is still too special - * e.g. clock tick interrupts) - */ - return cpu == 0 ? -EPERM : 0; -} diff --git a/arch/arm/mach-realview/hotplug.c b/arch/arm/mach-realview/hotplug.c index f95521a..580cf7c 100644 --- a/arch/arm/mach-realview/hotplug.c +++ b/arch/arm/mach-realview/hotplug.c @@ -11,19 +11,16 @@ #include <linux/kernel.h> #include <linux/errno.h> #include <linux/smp.h> -#include <linux/completion.h> #include <asm/cacheflush.h> extern volatile int pen_release; -static DECLARE_COMPLETION(cpu_killed); static inline void cpu_enter_lowpower(void) { unsigned int v; - flush_cache_all(); asm volatile( " mcr p15, 0, %1, c7, c5, 0\n" " mcr p15, 0, %1, c7, c10, 4\n" @@ -93,11 +90,6 @@ static inline void platform_do_lowpower(unsigned int cpu) } } -int platform_cpu_kill(unsigned int cpu) -{ - return wait_for_completion_timeout(&cpu_killed, 5000); -} - /* * platform-specific code to shutdown a CPU * @@ -105,18 +97,6 @@ int platform_cpu_kill(unsigned int cpu) */ void platform_cpu_die(unsigned int cpu) { -#ifdef DEBUG - unsigned int this_cpu = hard_smp_processor_id(); - - if (cpu != this_cpu) { - printk(KERN_CRIT "Eek! platform_cpu_die running on %u, should be %u\n", - this_cpu, cpu); - BUG(); - } -#endif - - printk(KERN_NOTICE "CPU%u: shutdown\n", cpu); - complete(&cpu_killed); /* * we're ready for shutdown now, so do it @@ -130,12 +110,3 @@ void platform_cpu_die(unsigned int cpu) */ cpu_leave_lowpower(); } - -int platform_cpu_disable(unsigned int cpu) -{ - /* - * we don't allow CPU 0 to be shutdown (it is still too special - * e.g. clock tick interrupts) - */ - return cpu == 0 ? -EPERM : 0; -} diff --git a/arch/arm/mach-s5pv310/hotplug.c b/arch/arm/mach-s5pv310/hotplug.c index 03652c3..59eafad 100644 --- a/arch/arm/mach-s5pv310/hotplug.c +++ b/arch/arm/mach-s5pv310/hotplug.c @@ -13,19 +13,16 @@ #include <linux/kernel.h> #include <linux/errno.h> #include <linux/smp.h> -#include <linux/completion.h> #include <asm/cacheflush.h> extern volatile int pen_release; -static DECLARE_COMPLETION(cpu_killed); static inline void cpu_enter_lowpower(void) { unsigned int v; - flush_cache_all(); asm volatile( " mcr p15, 0, %1, c7, c5, 0\n" " mcr p15, 0, %1, c7, c10, 4\n" @@ -96,11 +93,6 @@ static inline void platform_do_lowpower(unsigned int cpu) } } -int platform_cpu_kill(unsigned int cpu) -{ - return wait_for_completion_timeout(&cpu_killed, 5000); -} - /* * platform-specific code to shutdown a CPU * @@ -108,19 +100,6 @@ int platform_cpu_kill(unsigned int cpu) */ void platform_cpu_die(unsigned int cpu) { -#ifdef DEBUG - unsigned int this_cpu = hard_smp_processor_id(); - - if (cpu != this_cpu) { - printk(KERN_CRIT "Eek! platform_cpu_die running on %u, should be %u\n", - this_cpu, cpu); - BUG(); - } -#endif - - printk(KERN_NOTICE "CPU%u: shutdown\n", cpu); - complete(&cpu_killed); - /* * we're ready for shutdown now, so do it */ @@ -134,11 +113,3 @@ void platform_cpu_die(unsigned int cpu) cpu_leave_lowpower(); } -int platform_cpu_disable(unsigned int cpu) -{ - /* - * we don't allow CPU 0 to be shutdown (it is still too special - * e.g. clock tick interrupts) - */ - return cpu == 0 ? -EPERM : 0; -} diff --git a/arch/arm/mach-tegra/hotplug.c b/arch/arm/mach-tegra/hotplug.c index 8e7f115..7e5eb3d 100644 --- a/arch/arm/mach-tegra/hotplug.c +++ b/arch/arm/mach-tegra/hotplug.c @@ -11,17 +11,14 @@ #include <linux/kernel.h> #include <linux/errno.h> #include <linux/smp.h> -#include <linux/completion.h> #include <asm/cacheflush.h> -static DECLARE_COMPLETION(cpu_killed); static inline void cpu_enter_lowpower(void) { unsigned int v; - flush_cache_all(); asm volatile( " mcr p15, 0, %1, c7, c5, 0\n" " mcr p15, 0, %1, c7, c10, 4\n" @@ -92,11 +89,6 @@ static inline void platform_do_lowpower(unsigned int cpu) } } -int platform_cpu_kill(unsigned int cpu) -{ - return wait_for_completion_timeout(&cpu_killed, 5000); -} - /* * platform-specific code to shutdown a CPU * @@ -104,19 +96,6 @@ int platform_cpu_kill(unsigned int cpu) */ void platform_cpu_die(unsigned int cpu) { -#ifdef DEBUG - unsigned int this_cpu = hard_smp_processor_id(); - - if (cpu != this_cpu) { - printk(KERN_CRIT "Eek! platform_cpu_die running on %u, should be %u\n", - this_cpu, cpu); - BUG(); - } -#endif - - printk(KERN_NOTICE "CPU%u: shutdown\n", cpu); - complete(&cpu_killed); - /* * we're ready for shutdown now, so do it */ @@ -129,12 +108,3 @@ void platform_cpu_die(unsigned int cpu) */ cpu_leave_lowpower(); } - -int platform_cpu_disable(unsigned int cpu) -{ - /* - * we don't allow CPU 0 to be shutdown (it is still too special - * e.g. clock tick interrupts) - */ - return cpu == 0 ? -EPERM : 0; -} diff --git a/arch/arm/mach-ux500/hotplug.c b/arch/arm/mach-ux500/hotplug.c index b782a03..37b443b 100644 --- a/arch/arm/mach-ux500/hotplug.c +++ b/arch/arm/mach-ux500/hotplug.c @@ -6,23 +6,22 @@ * Based on ARM realview platform * * Author: Sundar Iyer <sundar.iyer(a)stericsson.com> + * Author: vincent.guittot <vincent.guittot(a)stericsson.com> * */ #include <linux/kernel.h> #include <linux/errno.h> #include <linux/smp.h> -#include <linux/completion.h> - -#include <asm/cacheflush.h> extern volatile int pen_release; -static DECLARE_COMPLETION(cpu_killed); - -static inline void platform_do_lowpower(unsigned int cpu) +/* + * platform-specific code to shutdown a CPU + * + * Called with IRQs disabled + */ +void platform_cpu_die(unsigned int cpu) { - flush_cache_all(); - /* we put the platform to just WFI */ for (;;) { __asm__ __volatile__("dsb\n\t" "wfi\n\t" @@ -36,40 +35,4 @@ static inline void platform_do_lowpower(unsigned int cpu) } } -int platform_cpu_kill(unsigned int cpu) -{ - return wait_for_completion_timeout(&cpu_killed, 5000); -} - -/* - * platform-specific code to shutdown a CPU - * - * Called with IRQs disabled - */ -void platform_cpu_die(unsigned int cpu) -{ -#ifdef DEBUG - unsigned int this_cpu = hard_smp_processor_id(); - - if (cpu != this_cpu) { - printk(KERN_CRIT "Eek! platform_cpu_die running on %u, should be %u\n", - this_cpu, cpu); - BUG(); - } -#endif - - printk(KERN_NOTICE "CPU%u: shutdown\n", cpu); - complete(&cpu_killed); - - /* directly enter low power state, skipping secure registers */ - platform_do_lowpower(cpu); -} -int platform_cpu_disable(unsigned int cpu) -{ - /* - * we don't allow CPU 0 to be shutdown (it is still too special - * e.g. clock tick interrupts) - */ - return cpu == 0 ? -EPERM : 0; -} -- 1.7.0.4

15 years

NEON power consumption

by Michael Hope

I sat down and measured the power consumption of the NEON unit on an OMAP3. Method and results are here: https://wiki.linaro.org/MichaelHope/Sandbox/NEONPower The board takes 2.37 W and the NEON unit adds an extra 120 mW. Assuming the core takes 1 W, then the code needs to run 12 % faster with NEON on to be a net power win. Note that the results are inaccurate but valid enough. -- Michael

15 years

Situation of armel cross compiler in Ubuntu

by Marcin Juszkiewicz

Hi During Ubuntu 'natty' cycle I am working on backporting armel cross compiler to 10.04 'lucid' and 10.10 'maverick' releases. I have a spec [1] for it and go though steps. Cross compiler is built from two source packages: armel-cross-toolchain-base and gcc-4.5-armel-cross. There are two additional ones: gcc-4.4-armel-cross (may be dropped when gcc 4.6 get released) and gcc-defaults-armel-cross. For now 'natty' is using same versions as maverick but I am working [2][3][4] on updates. Next step is to get rid of update-alternatives use from packages as this is what gcc-defaults-armel-cross has to do. I have new version of this package ready [5] but it has to wait until I will drop u-a from toolchain packages (this can be tracked in bug 676454). Then next step will be providing backports for lucid and maverick releases. For now I have unsigned manually created repository for lucid [6]. Maverick has small update (in proposed) which fixes few bugs. To make backports I need first to patch all components (binutils, eglibc, gcc-4.5, linux) to have a way to build just *-source package from them (who wants libc6 upgrade in LTS?). I have patches [7][8][9][10] for it. I also need advice how to provide backported cross compiler packages. For now there are two ways: - Linaro maintainers PPA - ubuntu-backports repository First one requires less work as there is no cooperation required with Ubuntu backports team [11]. And there will be usually monthly updates as Linaro has montly releases which are taken by Ubuntu packaging (during development cycle) This week I will test lucid/maverick/natty builds of recent cross compiler and can provide packages for testing (no warranty that it works). Where do I need help: - review my gcc-4.[45]-armel-cross branches - review my patches in [7][8][9][10] - test my packages 1. https://blueprints.launchpad.net/ubuntu/+spec/other-linaro-n-cross- compilers 2. https://code.launchpad.net/~hrw/ubuntu/natty/gcc-4.4-armel-cross/1.37-for- review/+merge/41582 3. https://code.launchpad.net/~hrw/ubuntu/natty/gcc-4.5-armel-cross/1.41-for- review/+merge/41583 4. https://code.launchpad.net/~hrw/ubuntu/natty/armel-cross-toolchain- base/1.53-fixed/+merge/41577 5. https://code.launchpad.net/~hrw/ubuntu/natty/gcc-defaults-armel-cross/use- proper-symlinks/+merge/41040 676454. https://bugs.launchpad.net/ubuntu/+source/gcc-4.4-armel- cross/+bug/676454 6. https://wiki.linaro.org/WorkingGroups/ToolChain/CrossCompilerOnLucid 7. https://bugs.launchpad.net/ubuntu/+source/gcc-4.4/+bug/682650 8. https://bugs.launchpad.net/ubuntu/+source/eglibc/+bug/682648 9. https://bugs.launchpad.net/ubuntu/+source/binutils/+bug/682646 10. https://bugs.launchpad.net/ubuntu/+source/linux/+bug/682681 11. https://help.ubuntu.com/community/UbuntuBackports Regards, -- JID: hrw(a)jabber.org Website: http://marcin.juszkiewicz.com.pl/ LinkedIn: http://www.linkedin.com/in/marcinjuszkiewicz

15 years

[PATCH][ ARM cpu hotplug 2/2 ] Use common cpu hotplug code for UX500

by Vincent Guittot

This patch extracts the common code of the cpu hotplug feature across arm platforms. The goal is to only keep the specific stuff of the platform in the sub-architecture. I have created a hotplug.c file in the arm/common directory after studying the cpu hotplug code of omap2, realview, s5pv310, ux500 and tegra. I have extracted 3 main platform dependent functions: -platform_enter_lowpower which prepares the platform for low power. -platform_do_lowpower on which the cpu will loop until it becomes really plugged (spurious wake up). This function must returned the cpu Id in order to leave the unplug state. -platform_leave_lowpower which restore the platform context. An ux500 patch is available which uses the common/hotplug.c code. This patch is quite short because the idle / power down functions are not yet upstreamed Signed-off-by: Vincent Guittot <vincent.guittot(a)stericsson.com> --- arch/arm/Kconfig | 1 + arch/arm/mach-ux500/Makefile | 2 +- arch/arm/mach-ux500/hotplug.c | 68 +++++++++------------------------------- arch/arm/mach-ux500/platsmp.c | 2 + 4 files changed, 20 insertions(+), 53 deletions(-) diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index ca96a44..ee74cf9 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -797,6 +797,7 @@ config ARCH_U8500 select GENERIC_CLOCKEVENTS select COMMON_CLKDEV select ARCH_REQUIRE_GPIOLIB + select USE_COMMON_ARM_HOTPLUG if HOTPLUG_CPU help Support for ST-Ericsson's Ux500 architecture diff --git a/arch/arm/mach-ux500/Makefile b/arch/arm/mach-ux500/Makefile index 0612013..cb86197 100644 --- a/arch/arm/mach-ux500/Makefile +++ b/arch/arm/mach-ux500/Makefile @@ -8,7 +8,7 @@ obj-$(CONFIG_UX500_SOC_DB8500) += cpu-db8500.o devices-db8500.o prcmu.o obj-$(CONFIG_MACH_U8500_MOP) += board-mop500.o board-mop500-sdi.o obj-$(CONFIG_MACH_U5500) += board-u5500.o obj-$(CONFIG_SMP) += platsmp.o headsmp.o -obj-$(CONFIG_HOTPLUG_CPU) += hotplug.o +obj-$(CONFIG_USE_COMMON_ARM_HOTPLUG) += hotplug.o obj-$(CONFIG_LOCAL_TIMERS) += localtimer.o obj-$(CONFIG_REGULATOR_AB8500) += board-mop500-regulators.o obj-$(CONFIG_U5500_MODEM_IRQ) += modem_irq.o diff --git a/arch/arm/mach-ux500/hotplug.c b/arch/arm/mach-ux500/hotplug.c index b782a03..14a614f 100644 --- a/arch/arm/mach-ux500/hotplug.c +++ b/arch/arm/mach-ux500/hotplug.c @@ -6,70 +6,34 @@ * Based on ARM realview platform * * Author: Sundar Iyer <sundar.iyer(a)stericsson.com> + * Author: Vincent Guittot <vincent.guittot(a)stericsson.com> * + * We only keep platform related code in the subarchitecture directory. + * The common code across platform has been moved into kernel/common directory */ -#include <linux/kernel.h> -#include <linux/errno.h> -#include <linux/smp.h> -#include <linux/completion.h> -#include <asm/cacheflush.h> +#include <asm/hotplug.h> extern volatile int pen_release; -static DECLARE_COMPLETION(cpu_killed); - -static inline void platform_do_lowpower(unsigned int cpu) -{ - flush_cache_all(); - - /* we put the platform to just WFI */ - for (;;) { - __asm__ __volatile__("dsb\n\t" "wfi\n\t" - : : : "memory"); - if (pen_release == cpu) { - /* - * OK, proper wakeup, we're done - */ - break; - } - } -} - -int platform_cpu_kill(unsigned int cpu) +/* Default implementation that will be used if not override + * by power specific implementation + */ +void __attribute__((weak)) platform_enter_lowpower(unsigned int cpu) { - return wait_for_completion_timeout(&cpu_killed, 5000); + return; } -/* - * platform-specific code to shutdown a CPU - * - * Called with IRQs disabled - */ -void platform_cpu_die(unsigned int cpu) +int __attribute__((weak)) platform_do_lowpower(unsigned int cpu) { -#ifdef DEBUG - unsigned int this_cpu = hard_smp_processor_id(); - - if (cpu != this_cpu) { - printk(KERN_CRIT "Eek! platform_cpu_die running on %u, should be %u\n", - this_cpu, cpu); - BUG(); - } -#endif - - printk(KERN_NOTICE "CPU%u: shutdown\n", cpu); - complete(&cpu_killed); + __asm__ __volatile__("dsb\n\t" "wfi\n\t" + : : : "memory"); - /* directly enter low power state, skipping secure registers */ - platform_do_lowpower(cpu); + return pen_release; } -int platform_cpu_disable(unsigned int cpu) +void __attribute__((weak)) platform_leave_lowpower(unsigned int cpu) { - /* - * we don't allow CPU 0 to be shutdown (it is still too special - * e.g. clock tick interrupts) - */ - return cpu == 0 ? -EPERM : 0; + return; } + diff --git a/arch/arm/mach-ux500/platsmp.c b/arch/arm/mach-ux500/platsmp.c index 9e4c678..3074260 100644 --- a/arch/arm/mach-ux500/platsmp.c +++ b/arch/arm/mach-ux500/platsmp.c @@ -51,6 +51,7 @@ void __cpuinit platform_secondary_init(unsigned int cpu) * pen, then head off into the C entry point */ pen_release = -1; + smp_wmb(); /* * Synchronise with the boot thread. @@ -82,6 +83,7 @@ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle) timeout = jiffies + (1 * HZ); while (time_before(jiffies, timeout)) { + smp_rmb(); if (pen_release == -1) break; } -- 1.7.0.4

15 years

Notes & Actions: Linaro User Platform Weekly Status Meeting 11/24

by Tom Gall

Greetings, Enclosed you'll find a link to the agenda, notes and actions from the Linaro User Platforms Weekly Status meeting dated November 24th held in #linaro-meeting on irc.freenode.net at 13:00 UTC. https://wiki.linaro.org/Platform/UserPlatforms/WeeklyStatus/2010-11-24 Status Summary: * Good progress on discussions with upstream Cairo team * Confirmed with upstream Unity team that there is no plan to add gles2 support to compiz/nux * Focus continues in refining Blueprints and Specifications across Multimedia and Graphics WGs * Seeds for linaro-development, linaro-multimedia-engineering and linaro-graphical-engineering either merged or in a proposed state in prep for natty-alpha-1 Regards, Tom (tgall_foo) User Platforms Team "We want great men who, when fortune frowns will not be discouraged." - Colonel Henry Knox w) tom.gall att linaro.org w) tom_gall att vnet.ibm.com h) tom_gall att mac.com

15 years

Issues with ondemand governor

by Vishwanath Sripathy

Hi, I was trying to investigate performance issues that we were seeing with some usecases like Video playback on OMAP Platforms with ondemand governor. As part of this, I found a tool called cpufreq-bench (http://lwn.net/Articles/339862) which can be used determine the performance impact of ondemand governor compared to performacne governor. When I ran this tool on OMAP3 (ZOOM3) platform using 2.6.36 kernel with below command, the worstcase ondemand performance is 35% compared to performance governor. cpufreq-bench -l 50000 -s 100000 -x 50000 -y 100000 -g ondemand -r 5 -n 5 -v I tried the same on x86 platforms and there the worstcase performance is around 88%. Attached are the cpufreq-bench logs for x86 and omap3. Questions: 1. Is this is known limitaiton of ondemand governor? 2. How do we support system usecases (like video playback etc) with ondemand governor if governor is not able to scale the frequencies in realtime? Are applications expected to play with scaling_min_freq to increase mpu frequency? Regards Vishwa

15 years

cpufreq-bench test on X86 platforms

by Sripathy, Vishwanath

All, I am trying to investigate ondemand governor's limitation in Linux kernel. As part of that, I have found a tool called cpufreq-bench which can be used to determine the performance degradation due to ondemand governor compared to performance governor. I have used this tool on OMAP platforms and able to see the issue. Now I would like to demonstrate this issue on x86 platform to conclude that it's a generic governor problem and nothing specific to ARM based SOCs. I would need some of your help to run this test bench on some of recent X86 platforms that have support for many P states. (My PC is little old which has only 2 p-states (2.8GHz & 3.4 GHz), so performance degradation is not much visible). More details on the tool and test procedure can be found at https://blueprints.edge.launchpad.net/linaro-pm-wg/+spec/cpufreq-ondemand-g… Pls ping me if you need any assistance in testing. Looking forward for your test results. Regards Vishwanath

15 years

[PATCH][ ARM cpu hotplug 1/2 ] extract common code for arm cpu hotplug

by Vincent Guittot

This patch extracts the common code of the cpu hotplug feature across arm platforms. The goal is to only keep the specific stuff of the platform in the sub-architecture. I have created a hotplug.c file in the arm/common directory after studying the cpu hotplug code of omap2, realview, s5pv310, ux500 and tegra. I have extracted 3 main platform dependent functions: -platform_enter_lowpower which prepares the platform for low power. -platform_do_lowpower on which the cpu will loop until it becomes really plugged (spurious wake up). This function must returned the cpu Id in order to leave the unplug state. -platform_leave_lowpower which restore the platform context. An ux500 patch is available which uses the common/hotplug.c code. This patch is quite short because the idle / power down functions are not yet upstreamed Signed-off-by: Vincent Guittot <vincent.guittot(a)linaro.org> --- arch/arm/common/Kconfig | 4 ++ arch/arm/common/Makefile | 1 + arch/arm/common/hotplug.c | 85 ++++++++++++++++++++++++++++++++++++++++ arch/arm/include/asm/hotplug.h | 8 ++++ 4 files changed, 98 insertions(+), 0 deletions(-) create mode 100644 arch/arm/common/hotplug.c create mode 100644 arch/arm/include/asm/hotplug.h diff --git a/arch/arm/common/Kconfig b/arch/arm/common/Kconfig index 0a34c81..b1c5f6b 100644 --- a/arch/arm/common/Kconfig +++ b/arch/arm/common/Kconfig @@ -41,3 +41,7 @@ config SHARP_SCOOP config COMMON_CLKDEV bool select HAVE_CLK + +config USE_COMMON_ARM_HOTPLUG + bool + depends on HOTPLUG_CPU diff --git a/arch/arm/common/Makefile b/arch/arm/common/Makefile index e6e8664..d8a66d7 100644 --- a/arch/arm/common/Makefile +++ b/arch/arm/common/Makefile @@ -17,3 +17,4 @@ obj-$(CONFIG_ARCH_IXP2000) += uengine.o obj-$(CONFIG_ARCH_IXP23XX) += uengine.o obj-$(CONFIG_PCI_HOST_ITE8152) += it8152.o obj-$(CONFIG_COMMON_CLKDEV) += clkdev.o +obj-$(CONFIG_USE_COMMON_ARM_HOTPLUG) += hotplug.o diff --git a/arch/arm/common/hotplug.c b/arch/arm/common/hotplug.c new file mode 100644 index 0000000..05f39f9 --- /dev/null +++ b/arch/arm/common/hotplug.c @@ -0,0 +1,85 @@ +/* + * Copyright (C) STMicroelectronics 2009 + * Copyright (C) ST-Ericsson SA 2010 + * + * License Terms: GNU General Public License v2 + * Based on ARM realview platform + * + * Author: Sundar Iyer <sundar.iyer(a)stericsson.com> + * Author: Vincent Guittot <vincent.guittot(a)stericsson.com> + * + */ +#include <linux/kernel.h> +#include <linux/errno.h> +#include <linux/smp.h> +#include <linux/completion.h> + +#include <asm/cacheflush.h> +#include <asm/hotplug.h> + +static DECLARE_COMPLETION(cpu_killed); + +int platform_cpu_kill(unsigned int cpu) +{ + int err; + err = wait_for_completion_timeout(&cpu_killed, 5000); +#ifdef DEBUG + printk(KERN_INFO "platform_cpu_kill %d err %d\n", cpu, err); +#endif + return err; +} + +/* + * architecture-specific code to shutdown a CPU + * + * Called with IRQs disabled + */ +void platform_cpu_die(unsigned int cpu) +{ + unsigned long spurious = 0; +#ifdef DEBUG + unsigned int this_cpu = hard_smp_processor_id(); + + if (cpu != this_cpu) { + printk(KERN_CRIT "Eek! platform_cpu_die running on %u, should be %u\n", + this_cpu, cpu); + BUG(); + } + + printk(KERN_NOTICE "CPU%u: shutdown\n", cpu); +#endif + complete(&cpu_killed); + + flush_cache_all(); + wmb(); + + platform_enter_lowpower(cpu); + + for (;;) { + + if (platform_do_lowpower(cpu) == cpu) { + /* + * OK, proper wakeup, we're done + */ + break; + } + spurious++; + } + + platform_leave_lowpower(cpu); + +#ifdef DEBUG + printk(KERN_NOTICE "CPU%u: wake up (%lu spurious wake up)\n", + cpu, spurious); +#endif + +} + +int platform_cpu_disable(unsigned int cpu) +{ + /* + * we don't allow CPU 0 to be shutdown (it is still too special + * e.g. clock tick interrupts) + */ + return cpu == 0 ? -EPERM : 0; +} diff --git a/arch/arm/include/asm/hotplug.h b/arch/arm/include/asm/hotplug.h new file mode 100644 index 0000000..b38e737 --- /dev/null +++ b/arch/arm/include/asm/hotplug.h @@ -0,0 +1,8 @@ +#ifndef __ASM_MACH_HOTPLUG_H +#define __ASM_MACH_HOTPLUG_H + +extern void platform_enter_lowpower(unsigned int cpu); +extern int platform_do_lowpower(unsigned int cpu); +extern void platform_leave_lowpower(unsigned int cpu); + +#endif -- 1.7.0.4

15 years

linux-meta-linaro-natty pull request

by John Rigby

Loïc, And here is the meta package pull request. Thanks, John The following changes since commit 9af23910a5f4134e4ffc4485d3780bbbe53019c8: LINARO: Linaro-2.6.36.1000.1 (2010-11-19 09:45:12 -0700) are available in the git repository at: git://git.linaro.org/ubuntu/linux-meta-linaro-natty.git master John Rigby (1): LINARO: Linaro-2.6.37.1000.1 meta-source/debian/changelog | 6 ++++++ 1 files changed, 6 insertions(+), 0 deletions(-)

15 years

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

linaro-dev