- linaro-kernel - lists.linaro.org

[PATCH 1/4][resend] cpuidle : handle clockevent notify from the cpuidle framework

by Daniel Lezcano

When a cpu enters a deep idle state, the local timers are stopped and the time framework falls back to the timer device used as a broadcast timer. The different cpuidle drivers are calling clockevents_notify ENTER/EXIT when the idle state stops the local timer. Add a new flag CPUIDLE_FLAG_TIMER_STOP which can be set by the cpuidle drivers. If the flag is set, the cpuidle core code takes care of the notification on behalf of the driver to avoid pointless code duplication. Signed-off-by: Daniel Lezcano <daniel.lezcano(a)linaro.org> Reviewed-by: Thomas Gleixner <tglx(a)linutronix.de> Cc: Len Brown <lenb(a)kernel.org> Cc: Linus Walleij <linus.walleij(a)linaro.org> Cc: Santosh Shilimkar <santosh.shilimkar(a)ti.com> Cc: Rajendra Nayak <rnayak(a)ti.com> Cc: Sascha Hauer <kernel(a)pengutronix.de> Cc: Thomas Gleixner <tglx(a)linutronix.de> --- drivers/cpuidle/cpuidle.c | 9 +++++++++ include/linux/cpuidle.h | 1 + 2 files changed, 10 insertions(+) diff --git a/drivers/cpuidle/cpuidle.c b/drivers/cpuidle/cpuidle.c index eba6929..c500370 100644 --- a/drivers/cpuidle/cpuidle.c +++ b/drivers/cpuidle/cpuidle.c @@ -8,6 +8,7 @@ * This code is licenced under the GPL. */ +#include <linux/clockchips.h> #include <linux/kernel.h> #include <linux/mutex.h> #include <linux/sched.h> @@ -146,12 +147,20 @@ int cpuidle_idle_call(void) trace_cpu_idle_rcuidle(next_state, dev->cpu); + if (drv->states[next_state].flags & CPUIDLE_FLAG_TIMER_STOP) + clockevents_notify(CLOCK_EVT_NOTIFY_BROADCAST_ENTER, + &dev->cpu); + if (cpuidle_state_is_coupled(dev, drv, next_state)) entered_state = cpuidle_enter_state_coupled(dev, drv, next_state); else entered_state = cpuidle_enter_state(dev, drv, next_state); + if (drv->states[next_state].flags & CPUIDLE_FLAG_TIMER_STOP) + clockevents_notify(CLOCK_EVT_NOTIFY_BROADCAST_EXIT, + &dev->cpu); + trace_cpu_idle_rcuidle(PWR_EVENT_EXIT, dev->cpu); /* give the governor an opportunity to reflect on the outcome */ diff --git a/include/linux/cpuidle.h b/include/linux/cpuidle.h index 480c14d..a837b33 100644 --- a/include/linux/cpuidle.h +++ b/include/linux/cpuidle.h @@ -57,6 +57,7 @@ struct cpuidle_state { /* Idle State Flags */ #define CPUIDLE_FLAG_TIME_VALID (0x01) /* is residency time measurable? */ #define CPUIDLE_FLAG_COUPLED (0x02) /* state applies to multiple cpus */ +#define CPUIDLE_FLAG_TIMER_STOP (0x04) /* timer is stopped on this state */ #define CPUIDLE_DRIVER_FLAGS_MASK (0xFFFF0000) -- 1.7.9.5

12 years, 2 months

5
19
0 0

[PATCH 0/4] ARM: add support for context tracking subsystem

by Kevin Hilman

commit 91d1aa43 (context_tracking: New context tracking susbsystem) generalized parts of the RCU userspace extended quiescent state into the context tracking subsystem. Context tracking is then used to implement adaptive tickless (a.k.a full nohz) Mainline currently only includes x86 support for the context tracking susbsystem and the goal of this series is to add ARM support, in order to make adaptive tickless functional on ARM. Depends on the prerequistes series: [PATCH 0/3] ARM: context tracking support prerequisites http://marc.info/?l=linux-kernel&m=136382248131438&w=2 Both of which are combined on top of Frederic's 3.9-rc1-nohz1 branch and available here: git://git.kernel.org/pub/scm/linux/kernel/git/khilman/linux.git arm-nohz-v2/context-tracking Using this, I tested adaptive tickless on a 2 CPU ARM SoC (OMAP4 Panda.) Kevin Hilman (4): ARM: context tracking: add exception support ARM: context tracking: instrument system calls ARM: context tracking: handle post exception/syscall/IRQ work ARM: Kconfig: allow context tracking arch/arm/Kconfig | 1 + arch/arm/include/asm/thread_info.h | 4 +++- arch/arm/kernel/ptrace.c | 11 +++++++++++ arch/arm/kernel/signal.c | 12 +++++++++--- arch/arm/kernel/traps.c | 18 +++++++++++++++++- arch/arm/mm/fault.c | 32 +++++++++++++++++++++++++++----- 6 files changed, 68 insertions(+), 10 deletions(-) -- 1.8.2

12 years, 2 months

2
7
0 0

[PATCH 1/2] cpufreq: Notify all policy->cpus in cpufreq_notify_transition()

by Viresh Kumar

policy->cpus contains all online cpus that have single shared clock line. And their frequencies are always updated together. Many SMP system's cpufreq drivers take care of this in individual drivers but the best place for this code is in cpufreq core. This patch modifies cpufreq_notify_transition() to notify frequency change for all cpus in policy->cpus and hence updates all users of this API. Signed-off-by: Viresh Kumar <viresh.kumar(a)linaro.org> Cc: Sekhar Nori <nsekhar(a)ti.com> Cc: Sascha Hauer <kernel(a)pengutronix.de> Cc: Eric Miao <eric.y.miao(a)gmail.com> Cc: Haojian Zhuang <haojian.zhuang(a)gmail.com> Cc: Ben Dooks <ben-linux(a)fluff.org> Cc: Kukjin Kim <kgene.kim(a)samsung.com> Cc: Stephen Warren <swarren(a)wwwdotorg.org> Cc: Haavard Skinnemoen <hskinnemoen(a)gmail.com> Cc: Hans-Christian Egtvedt <egtvedt(a)samfundet.no> Cc: Mike Frysinger <vapier(a)gentoo.org> Cc: Mikael Starvik <starvik(a)axis.com> Cc: Jesper Nilsson <jesper.nilsson(a)axis.com> Cc: Tony Luck <tony.luck(a)intel.com> Cc: Fenghua Yu <fenghua.yu(a)intel.com> Cc: Arnd Bergmann <arnd(a)arndb.de> Cc: Benjamin Herrenschmidt <benh(a)kernel.crashing.org> Cc: Ralf Baechle <ralf(a)linux-mips.org> Cc: Paul Mundt <lethal(a)linux-sh.org> Cc: David S. Miller <davem(a)davemloft.net> Cc: Thomas Renninger <trenn(a)suse.de> Cc: Borislav Petkov <bp(a)alien8.de> Cc: Guan Xuetao <gxt(a)mprc.pku.edu.cn> Cc: linux-arm-kernel(a)lists.infradead.org Cc: linux-cris-kernel(a)axis.com Cc: linux-ia64(a)vger.kernel.org Cc: linuxppc-dev(a)lists.ozlabs.org Cc: cbe-oss-dev(a)lists.ozlabs.org Cc: linux-mips(a)linux-mips.org Cc: linux-sh(a)vger.kernel.org Cc: sparclinux(a)vger.kernel.org Cc: linux-kernel(a)vger.kernel.org --- arch/arm/mach-davinci/cpufreq.c | 5 +- arch/arm/mach-imx/cpufreq.c | 5 +- arch/arm/mach-integrator/cpu.c | 6 +-- arch/arm/mach-pxa/cpufreq-pxa2xx.c | 5 +- arch/arm/mach-pxa/cpufreq-pxa3xx.c | 5 +- arch/arm/mach-s3c24xx/cpufreq.c | 8 +-- arch/arm/mach-sa1100/cpu-sa1100.c | 5 +- arch/arm/mach-sa1100/cpu-sa1110.c | 5 +- arch/arm/mach-tegra/cpu-tegra.c | 6 +-- arch/avr32/mach-at32ap/cpufreq.c | 5 +- arch/blackfin/mach-common/cpufreq.c | 79 ++++++++++++---------------- arch/cris/arch-v32/mach-a3/cpufreq.c | 20 +++---- arch/cris/arch-v32/mach-fs/cpufreq.c | 17 +++--- arch/ia64/kernel/cpufreq/acpi-cpufreq.c | 22 ++++---- arch/mips/kernel/cpufreq/loongson2_cpufreq.c | 5 +- arch/powerpc/platforms/cell/cbe_cpufreq.c | 5 +- arch/powerpc/platforms/pasemi/cpufreq.c | 5 +- arch/powerpc/platforms/powermac/cpufreq_32.c | 14 ++--- arch/powerpc/platforms/powermac/cpufreq_64.c | 5 +- arch/sh/kernel/cpufreq.c | 5 +- arch/sparc/kernel/us2e_cpufreq.c | 13 ++--- arch/sparc/kernel/us3_cpufreq.c | 13 ++--- arch/unicore32/kernel/cpu-ucv2.c | 5 +- drivers/cpufreq/acpi-cpufreq.c | 11 +--- drivers/cpufreq/cpufreq-cpu0.c | 12 ++--- drivers/cpufreq/cpufreq-nforce2.c | 5 +- drivers/cpufreq/cpufreq.c | 45 +++++++++------- drivers/cpufreq/dbx500-cpufreq.c | 6 +-- drivers/cpufreq/e_powersaver.c | 11 ++-- drivers/cpufreq/elanfreq.c | 10 ++-- drivers/cpufreq/exynos-cpufreq.c | 7 +-- drivers/cpufreq/gx-suspmod.c | 11 ++-- drivers/cpufreq/imx6q-cpufreq.c | 12 ++--- drivers/cpufreq/kirkwood-cpufreq.c | 10 ++-- drivers/cpufreq/longhaul.c | 18 ++++--- drivers/cpufreq/maple-cpufreq.c | 5 +- drivers/cpufreq/omap-cpufreq.c | 11 +--- drivers/cpufreq/p4-clockmod.c | 10 +--- drivers/cpufreq/pcc-cpufreq.c | 5 +- drivers/cpufreq/powernow-k6.c | 12 ++--- drivers/cpufreq/powernow-k7.c | 10 ++-- drivers/cpufreq/powernow-k8.c | 16 +++--- drivers/cpufreq/s3c2416-cpufreq.c | 5 +- drivers/cpufreq/s3c64xx-cpufreq.c | 7 ++- drivers/cpufreq/s5pv210-cpufreq.c | 5 +- drivers/cpufreq/sc520_freq.c | 10 ++-- drivers/cpufreq/spear-cpufreq.c | 7 +-- drivers/cpufreq/speedstep-centrino.c | 24 ++------- drivers/cpufreq/speedstep-ich.c | 12 +---- drivers/cpufreq/speedstep-smi.c | 5 +- include/linux/cpufreq.h | 4 +- 51 files changed, 232 insertions(+), 337 deletions(-) diff --git a/arch/arm/mach-davinci/cpufreq.c b/arch/arm/mach-davinci/cpufreq.c index 4729eaa..55eb870 100644 --- a/arch/arm/mach-davinci/cpufreq.c +++ b/arch/arm/mach-davinci/cpufreq.c @@ -90,7 +90,6 @@ static int davinci_target(struct cpufreq_policy *policy, freqs.old = davinci_getspeed(0); freqs.new = clk_round_rate(armclk, target_freq * 1000) / 1000; - freqs.cpu = 0; if (freqs.old == freqs.new) return ret; @@ -102,7 +101,7 @@ static int davinci_target(struct cpufreq_policy *policy, if (ret) return -EINVAL; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); /* if moving to higher frequency, up the voltage beforehand */ if (pdata->set_voltage && freqs.new > freqs.old) { @@ -126,7 +125,7 @@ static int davinci_target(struct cpufreq_policy *policy, pdata->set_voltage(idx); out: - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return ret; } diff --git a/arch/arm/mach-imx/cpufreq.c b/arch/arm/mach-imx/cpufreq.c index d8c75c3..cfce5e3 100644 --- a/arch/arm/mach-imx/cpufreq.c +++ b/arch/arm/mach-imx/cpufreq.c @@ -87,13 +87,12 @@ static int mxc_set_target(struct cpufreq_policy *policy, freqs.old = clk_get_rate(cpu_clk) / 1000; freqs.new = freq_Hz / 1000; - freqs.cpu = 0; freqs.flags = 0; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); ret = set_cpu_freq(freq_Hz); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return ret; } diff --git a/arch/arm/mach-integrator/cpu.c b/arch/arm/mach-integrator/cpu.c index 590c192..df863c3 100644 --- a/arch/arm/mach-integrator/cpu.c +++ b/arch/arm/mach-integrator/cpu.c @@ -123,14 +123,12 @@ static int integrator_set_target(struct cpufreq_policy *policy, vco = icst_hz_to_vco(&cclk_params, target_freq * 1000); freqs.new = icst_hz(&cclk_params, vco) / 1000; - freqs.cpu = policy->cpu; - if (freqs.old == freqs.new) { set_cpus_allowed(current, cpus_allowed); return 0; } - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); cm_osc = __raw_readl(CM_OSC); @@ -151,7 +149,7 @@ static int integrator_set_target(struct cpufreq_policy *policy, */ set_cpus_allowed(current, cpus_allowed); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return 0; } diff --git a/arch/arm/mach-pxa/cpufreq-pxa2xx.c b/arch/arm/mach-pxa/cpufreq-pxa2xx.c index 6a7aeab..f1ca4da 100644 --- a/arch/arm/mach-pxa/cpufreq-pxa2xx.c +++ b/arch/arm/mach-pxa/cpufreq-pxa2xx.c @@ -311,7 +311,6 @@ static int pxa_set_target(struct cpufreq_policy *policy, new_freq_mem = pxa_freq_settings[idx].membus; freqs.old = policy->cur; freqs.new = new_freq_cpu; - freqs.cpu = policy->cpu; if (freq_debug) pr_debug("Changing CPU frequency to %d Mhz, (SDRAM %d Mhz)\n", @@ -327,7 +326,7 @@ static int pxa_set_target(struct cpufreq_policy *policy, * you should add a notify client with any platform specific * Vcc changing capability */ - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); /* Calculate the next MDREFR. If we're slowing down the SDRAM clock * we need to preset the smaller DRI before the change. If we're @@ -382,7 +381,7 @@ static int pxa_set_target(struct cpufreq_policy *policy, * you should add a notify client with any platform specific * SDRAM refresh timer adjustments */ - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); /* * Even if voltage setting fails, we don't report it, as the frequency diff --git a/arch/arm/mach-pxa/cpufreq-pxa3xx.c b/arch/arm/mach-pxa/cpufreq-pxa3xx.c index b85b4ab..8c45b2b 100644 --- a/arch/arm/mach-pxa/cpufreq-pxa3xx.c +++ b/arch/arm/mach-pxa/cpufreq-pxa3xx.c @@ -184,7 +184,6 @@ static int pxa3xx_cpufreq_set(struct cpufreq_policy *policy, freqs.old = policy->cur; freqs.new = next->cpufreq_mhz * 1000; - freqs.cpu = policy->cpu; pr_debug("CPU frequency from %d MHz to %d MHz%s\n", freqs.old / 1000, freqs.new / 1000, @@ -193,14 +192,14 @@ static int pxa3xx_cpufreq_set(struct cpufreq_policy *policy, if (freqs.old == target_freq) return 0; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); local_irq_save(flags); __update_core_freq(next); __update_bus_freq(next); local_irq_restore(flags); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return 0; } diff --git a/arch/arm/mach-s3c24xx/cpufreq.c b/arch/arm/mach-s3c24xx/cpufreq.c index 5f181e7..3c0e78e 100644 --- a/arch/arm/mach-s3c24xx/cpufreq.c +++ b/arch/arm/mach-s3c24xx/cpufreq.c @@ -204,7 +204,6 @@ static int s3c_cpufreq_settarget(struct cpufreq_policy *policy, freqs.old = cpu_cur.freq; freqs.new = cpu_new.freq; - freqs.freqs.cpu = 0; freqs.freqs.old = cpu_cur.freq.armclk / 1000; freqs.freqs.new = cpu_new.freq.armclk / 1000; @@ -218,9 +217,7 @@ static int s3c_cpufreq_settarget(struct cpufreq_policy *policy, s3c_cpufreq_updateclk(clk_pclk, cpu_new.freq.pclk); /* start the frequency change */ - - if (policy) - cpufreq_notify_transition(&freqs.freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs.freqs, CPUFREQ_PRECHANGE); /* If hclk is staying the same, then we do not need to * re-write the IO or the refresh timings whilst we are changing @@ -264,8 +261,7 @@ static int s3c_cpufreq_settarget(struct cpufreq_policy *policy, local_irq_restore(flags); /* notify everyone we've done this */ - if (policy) - cpufreq_notify_transition(&freqs.freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs.freqs, CPUFREQ_POSTCHANGE); s3c_freq_dbg("%s: finished\n", __func__); return 0; diff --git a/arch/arm/mach-sa1100/cpu-sa1100.c b/arch/arm/mach-sa1100/cpu-sa1100.c index e8f4d1e..3268761 100644 --- a/arch/arm/mach-sa1100/cpu-sa1100.c +++ b/arch/arm/mach-sa1100/cpu-sa1100.c @@ -201,9 +201,8 @@ static int sa1100_target(struct cpufreq_policy *policy, freqs.old = cur; freqs.new = sa11x0_ppcr_to_freq(new_ppcr); - freqs.cpu = 0; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); if (freqs.new > cur) sa1100_update_dram_timings(cur, freqs.new); @@ -213,7 +212,7 @@ static int sa1100_target(struct cpufreq_policy *policy, if (freqs.new < cur) sa1100_update_dram_timings(cur, freqs.new); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return 0; } diff --git a/arch/arm/mach-sa1100/cpu-sa1110.c b/arch/arm/mach-sa1100/cpu-sa1110.c index 48c45b0..38a7733 100644 --- a/arch/arm/mach-sa1100/cpu-sa1110.c +++ b/arch/arm/mach-sa1100/cpu-sa1110.c @@ -258,7 +258,6 @@ static int sa1110_target(struct cpufreq_policy *policy, freqs.old = sa11x0_getspeed(0); freqs.new = sa11x0_ppcr_to_freq(ppcr); - freqs.cpu = 0; sdram_calculate_timing(&sd, freqs.new, sdram); @@ -279,7 +278,7 @@ static int sa1110_target(struct cpufreq_policy *policy, sd.mdcas[2] = 0xaaaaaaaa; #endif - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); /* * The clock could be going away for some time. Set the SDRAMs @@ -327,7 +326,7 @@ static int sa1110_target(struct cpufreq_policy *policy, */ sdram_update_refresh(freqs.new, sdram); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return 0; } diff --git a/arch/arm/mach-tegra/cpu-tegra.c b/arch/arm/mach-tegra/cpu-tegra.c index e3d6e15..3b441d6 100644 --- a/arch/arm/mach-tegra/cpu-tegra.c +++ b/arch/arm/mach-tegra/cpu-tegra.c @@ -128,8 +128,7 @@ static int tegra_update_cpu_speed(unsigned long rate) else clk_set_rate(emc_clk, 100000000); /* emc 50Mhz */ - for_each_online_cpu(freqs.cpu) - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); #ifdef CONFIG_CPU_FREQ_DEBUG printk(KERN_DEBUG "cpufreq-tegra: transition: %u --> %u\n", @@ -143,8 +142,7 @@ static int tegra_update_cpu_speed(unsigned long rate) return ret; } - for_each_online_cpu(freqs.cpu) - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return 0; } diff --git a/arch/avr32/mach-at32ap/cpufreq.c b/arch/avr32/mach-at32ap/cpufreq.c index 18b7656..6544887 100644 --- a/arch/avr32/mach-at32ap/cpufreq.c +++ b/arch/avr32/mach-at32ap/cpufreq.c @@ -61,7 +61,6 @@ static int at32_set_target(struct cpufreq_policy *policy, freqs.old = at32_get_speed(0); freqs.new = (freq + 500) / 1000; - freqs.cpu = 0; freqs.flags = 0; if (!ref_freq) { @@ -69,7 +68,7 @@ static int at32_set_target(struct cpufreq_policy *policy, loops_per_jiffy_ref = boot_cpu_data.loops_per_jiffy; } - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); if (freqs.old < freqs.new) boot_cpu_data.loops_per_jiffy = cpufreq_scale( loops_per_jiffy_ref, ref_freq, freqs.new); @@ -77,7 +76,7 @@ static int at32_set_target(struct cpufreq_policy *policy, if (freqs.new < freqs.old) boot_cpu_data.loops_per_jiffy = cpufreq_scale( loops_per_jiffy_ref, ref_freq, freqs.new); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); pr_debug("cpufreq: set frequency %lu Hz\n", freq); diff --git a/arch/blackfin/mach-common/cpufreq.c b/arch/blackfin/mach-common/cpufreq.c index d88bd31..4e67368 100644 --- a/arch/blackfin/mach-common/cpufreq.c +++ b/arch/blackfin/mach-common/cpufreq.c @@ -127,13 +127,13 @@ unsigned long cpu_set_cclk(int cpu, unsigned long new) } #endif -static int bfin_target(struct cpufreq_policy *poli, +static int bfin_target(struct cpufreq_policy *policy, unsigned int target_freq, unsigned int relation) { #ifndef CONFIG_BF60x unsigned int plldiv; #endif - unsigned int index, cpu; + unsigned int index; unsigned long cclk_hz; struct cpufreq_freqs freqs; static unsigned long lpj_ref; @@ -144,59 +144,48 @@ static int bfin_target(struct cpufreq_policy *poli, cycles_t cycles; #endif - for_each_online_cpu(cpu) { - struct cpufreq_policy *policy = cpufreq_cpu_get(cpu); + if (cpufreq_frequency_table_target(policy, bfin_freq_table, target_freq, + relation, &index)) + return -EINVAL; - if (!policy) - continue; + cclk_hz = bfin_freq_table[index].frequency; - if (cpufreq_frequency_table_target(policy, bfin_freq_table, - target_freq, relation, &index)) - return -EINVAL; + freqs.old = bfin_getfreq_khz(0); + freqs.new = cclk_hz; - cclk_hz = bfin_freq_table[index].frequency; + pr_debug("cpufreq: changing cclk to %lu; target = %u, oldfreq = %u\n", + cclk_hz, target_freq, freqs.old); - freqs.old = bfin_getfreq_khz(0); - freqs.new = cclk_hz; - freqs.cpu = cpu; - - pr_debug("cpufreq: changing cclk to %lu; target = %u, oldfreq = %u\n", - cclk_hz, target_freq, freqs.old); - - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); - if (cpu == CPUFREQ_CPU) { + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); #ifndef CONFIG_BF60x - plldiv = (bfin_read_PLL_DIV() & SSEL) | - dpm_state_table[index].csel; - bfin_write_PLL_DIV(plldiv); + plldiv = (bfin_read_PLL_DIV() & SSEL) | dpm_state_table[index].csel; + bfin_write_PLL_DIV(plldiv); #else - ret = cpu_set_cclk(cpu, freqs.new * 1000); - if (ret != 0) { - WARN_ONCE(ret, "cpufreq set freq failed %d\n", ret); - break; - } + ret = cpu_set_cclk(policy->cpu, freqs.new * 1000); + if (ret != 0) { + WARN_ONCE(ret, "cpufreq set freq failed %d\n", ret); + break; + } #endif - on_each_cpu(bfin_adjust_core_timer, &index, 1); + on_each_cpu(bfin_adjust_core_timer, &index, 1); #if defined(CONFIG_CYCLES_CLOCKSOURCE) - cycles = get_cycles(); - SSYNC(); - cycles += 10; /* ~10 cycles we lose after get_cycles() */ - __bfin_cycles_off += - (cycles << __bfin_cycles_mod) - (cycles << index); - __bfin_cycles_mod = index; + cycles = get_cycles(); + SSYNC(); + cycles += 10; /* ~10 cycles we lose after get_cycles() */ + __bfin_cycles_off += (cycles << __bfin_cycles_mod) - (cycles << index); + __bfin_cycles_mod = index; #endif - if (!lpj_ref_freq) { - lpj_ref = loops_per_jiffy; - lpj_ref_freq = freqs.old; - } - if (freqs.new != freqs.old) { - loops_per_jiffy = cpufreq_scale(lpj_ref, - lpj_ref_freq, freqs.new); - } - } - /* TODO: just test case for cycles clock source, remove later */ - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + if (!lpj_ref_freq) { + lpj_ref = loops_per_jiffy; + lpj_ref_freq = freqs.old; } + if (freqs.new != freqs.old) { + loops_per_jiffy = cpufreq_scale(lpj_ref, + lpj_ref_freq, freqs.new); + } + + /* TODO: just test case for cycles clock source, remove later */ + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); pr_debug("cpufreq: done\n"); return ret; diff --git a/arch/cris/arch-v32/mach-a3/cpufreq.c b/arch/cris/arch-v32/mach-a3/cpufreq.c index ee391ec..ee142c4 100644 --- a/arch/cris/arch-v32/mach-a3/cpufreq.c +++ b/arch/cris/arch-v32/mach-a3/cpufreq.c @@ -27,23 +27,17 @@ static unsigned int cris_freq_get_cpu_frequency(unsigned int cpu) return clk_ctrl.pll ? 200000 : 6000; } -static void cris_freq_set_cpu_state(unsigned int state) +static void cris_freq_set_cpu_state(struct cpufreq_policy *policy, + unsigned int state) { - int i = 0; struct cpufreq_freqs freqs; reg_clkgen_rw_clk_ctrl clk_ctrl; clk_ctrl = REG_RD(clkgen, regi_clkgen, rw_clk_ctrl); -#ifdef CONFIG_SMP - for_each_present_cpu(i) -#endif - { - freqs.old = cris_freq_get_cpu_frequency(i); - freqs.new = cris_freq_table[state].frequency; - freqs.cpu = i; - } + freqs.old = cris_freq_get_cpu_frequency(policy->cpu); + freqs.new = cris_freq_table[state].frequency; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); local_irq_disable(); @@ -57,7 +51,7 @@ static void cris_freq_set_cpu_state(unsigned int state) local_irq_enable(); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); }; static int cris_freq_verify(struct cpufreq_policy *policy) @@ -75,7 +69,7 @@ static int cris_freq_target(struct cpufreq_policy *policy, target_freq, relation, &newstate)) return -EINVAL; - cris_freq_set_cpu_state(newstate); + cris_freq_set_cpu_state(policy, newstate); return 0; } diff --git a/arch/cris/arch-v32/mach-fs/cpufreq.c b/arch/cris/arch-v32/mach-fs/cpufreq.c index d92cf70..1295223 100644 --- a/arch/cris/arch-v32/mach-fs/cpufreq.c +++ b/arch/cris/arch-v32/mach-fs/cpufreq.c @@ -27,20 +27,17 @@ static unsigned int cris_freq_get_cpu_frequency(unsigned int cpu) return clk_ctrl.pll ? 200000 : 6000; } -static void cris_freq_set_cpu_state(unsigned int state) +static void cris_freq_set_cpu_state(struct cpufreq_policy *policy, + unsigned int state) { - int i; struct cpufreq_freqs freqs; reg_config_rw_clk_ctrl clk_ctrl; clk_ctrl = REG_RD(config, regi_config, rw_clk_ctrl); - for_each_possible_cpu(i) { - freqs.old = cris_freq_get_cpu_frequency(i); - freqs.new = cris_freq_table[state].frequency; - freqs.cpu = i; - } + freqs.old = cris_freq_get_cpu_frequency(policy->cpu); + freqs.new = cris_freq_table[state].frequency; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); local_irq_disable(); @@ -54,7 +51,7 @@ static void cris_freq_set_cpu_state(unsigned int state) local_irq_enable(); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); }; static int cris_freq_verify(struct cpufreq_policy *policy) @@ -71,7 +68,7 @@ static int cris_freq_target(struct cpufreq_policy *policy, (policy, cris_freq_table, target_freq, relation, &newstate)) return -EINVAL; - cris_freq_set_cpu_state(newstate); + cris_freq_set_cpu_state(policy, newstate); return 0; } diff --git a/arch/ia64/kernel/cpufreq/acpi-cpufreq.c b/arch/ia64/kernel/cpufreq/acpi-cpufreq.c index f09b174..4700fef 100644 --- a/arch/ia64/kernel/cpufreq/acpi-cpufreq.c +++ b/arch/ia64/kernel/cpufreq/acpi-cpufreq.c @@ -137,7 +137,7 @@ migrate_end: static int processor_set_freq ( struct cpufreq_acpi_io *data, - unsigned int cpu, + struct cpufreq_policy *policy, int state) { int ret = 0; @@ -149,8 +149,8 @@ processor_set_freq ( pr_debug("processor_set_freq\n"); saved_mask = current->cpus_allowed; - set_cpus_allowed_ptr(current, cpumask_of(cpu)); - if (smp_processor_id() != cpu) { + set_cpus_allowed_ptr(current, cpumask_of(policy->cpu)); + if (smp_processor_id() != policy->cpu) { retval = -EAGAIN; goto migrate_end; } @@ -170,12 +170,11 @@ processor_set_freq ( data->acpi_data.state, state); /* cpufreq frequency struct */ - cpufreq_freqs.cpu = cpu; cpufreq_freqs.old = data->freq_table[data->acpi_data.state].frequency; cpufreq_freqs.new = data->freq_table[state].frequency; /* notify cpufreq */ - cpufreq_notify_transition(&cpufreq_freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &cpufreq_freqs, CPUFREQ_PRECHANGE); /* * First we write the target state's 'control' value to the @@ -189,17 +188,20 @@ processor_set_freq ( ret = processor_set_pstate(value); if (ret) { unsigned int tmp = cpufreq_freqs.new; - cpufreq_notify_transition(&cpufreq_freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &cpufreq_freqs, + CPUFREQ_POSTCHANGE); cpufreq_freqs.new = cpufreq_freqs.old; cpufreq_freqs.old = tmp; - cpufreq_notify_transition(&cpufreq_freqs, CPUFREQ_PRECHANGE); - cpufreq_notify_transition(&cpufreq_freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &cpufreq_freqs, + CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &cpufreq_freqs, + CPUFREQ_POSTCHANGE); printk(KERN_WARNING "Transition failed with error %d\n", ret); retval = -ENODEV; goto migrate_end; } - cpufreq_notify_transition(&cpufreq_freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &cpufreq_freqs, CPUFREQ_POSTCHANGE); data->acpi_data.state = state; @@ -240,7 +242,7 @@ acpi_cpufreq_target ( if (result) return (result); - result = processor_set_freq(data, policy->cpu, next_state); + result = processor_set_freq(data, policy, next_state); return (result); } diff --git a/arch/mips/kernel/cpufreq/loongson2_cpufreq.c b/arch/mips/kernel/cpufreq/loongson2_cpufreq.c index 3237c52..bafda70 100644 --- a/arch/mips/kernel/cpufreq/loongson2_cpufreq.c +++ b/arch/mips/kernel/cpufreq/loongson2_cpufreq.c @@ -80,7 +80,6 @@ static int loongson2_cpufreq_target(struct cpufreq_policy *policy, pr_debug("cpufreq: requested frequency %u Hz\n", target_freq * 1000); - freqs.cpu = cpu; freqs.old = loongson2_cpufreq_get(cpu); freqs.new = freq; freqs.flags = 0; @@ -89,7 +88,7 @@ static int loongson2_cpufreq_target(struct cpufreq_policy *policy, return 0; /* notifiers */ - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); set_cpus_allowed_ptr(current, &cpus_allowed); @@ -97,7 +96,7 @@ static int loongson2_cpufreq_target(struct cpufreq_policy *policy, clk_set_rate(cpuclk, freq); /* notifiers */ - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); pr_debug("cpufreq: set frequency %u kHz\n", freq); diff --git a/arch/powerpc/platforms/cell/cbe_cpufreq.c b/arch/powerpc/platforms/cell/cbe_cpufreq.c index d4c39e3..718c6a3 100644 --- a/arch/powerpc/platforms/cell/cbe_cpufreq.c +++ b/arch/powerpc/platforms/cell/cbe_cpufreq.c @@ -156,10 +156,9 @@ static int cbe_cpufreq_target(struct cpufreq_policy *policy, freqs.old = policy->cur; freqs.new = cbe_freqs[cbe_pmode_new].frequency; - freqs.cpu = policy->cpu; mutex_lock(&cbe_switch_mutex); - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); pr_debug("setting frequency for cpu %d to %d kHz, " \ "1/%d of max frequency\n", @@ -169,7 +168,7 @@ static int cbe_cpufreq_target(struct cpufreq_policy *policy, rc = set_pmode(policy->cpu, cbe_pmode_new); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); mutex_unlock(&cbe_switch_mutex); return rc; diff --git a/arch/powerpc/platforms/pasemi/cpufreq.c b/arch/powerpc/platforms/pasemi/cpufreq.c index 890f30e..be1e795 100644 --- a/arch/powerpc/platforms/pasemi/cpufreq.c +++ b/arch/powerpc/platforms/pasemi/cpufreq.c @@ -273,10 +273,9 @@ static int pas_cpufreq_target(struct cpufreq_policy *policy, freqs.old = policy->cur; freqs.new = pas_freqs[pas_astate_new].frequency; - freqs.cpu = policy->cpu; mutex_lock(&pas_switch_mutex); - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); pr_debug("setting frequency for cpu %d to %d kHz, 1/%d of max frequency\n", policy->cpu, @@ -288,7 +287,7 @@ static int pas_cpufreq_target(struct cpufreq_policy *policy, for_each_online_cpu(i) set_astate(i, pas_astate_new); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); mutex_unlock(&pas_switch_mutex); ppc_proc_freq = freqs.new * 1000ul; diff --git a/arch/powerpc/platforms/powermac/cpufreq_32.c b/arch/powerpc/platforms/powermac/cpufreq_32.c index 311b804..3104fad 100644 --- a/arch/powerpc/platforms/powermac/cpufreq_32.c +++ b/arch/powerpc/platforms/powermac/cpufreq_32.c @@ -335,7 +335,8 @@ static int pmu_set_cpu_speed(int low_speed) return 0; } -static int do_set_cpu_speed(int speed_mode, int notify) +static int do_set_cpu_speed(struct cpufreq_policy *policy, int speed_mode, + int notify) { struct cpufreq_freqs freqs; unsigned long l3cr; @@ -343,13 +344,12 @@ static int do_set_cpu_speed(int speed_mode, int notify) freqs.old = cur_freq; freqs.new = (speed_mode == CPUFREQ_HIGH) ? hi_freq : low_freq; - freqs.cpu = smp_processor_id(); if (freqs.old == freqs.new) return 0; if (notify) - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); if (speed_mode == CPUFREQ_LOW && cpu_has_feature(CPU_FTR_L3CR)) { l3cr = _get_L3CR(); @@ -366,7 +366,7 @@ static int do_set_cpu_speed(int speed_mode, int notify) _set_L3CR(prev_l3cr); } if (notify) - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); cur_freq = (speed_mode == CPUFREQ_HIGH) ? hi_freq : low_freq; return 0; @@ -393,7 +393,7 @@ static int pmac_cpufreq_target( struct cpufreq_policy *policy, target_freq, relation, &newstate)) return -EINVAL; - rc = do_set_cpu_speed(newstate, 1); + rc = do_set_cpu_speed(policy, newstate, 1); ppc_proc_freq = cur_freq * 1000ul; return rc; @@ -442,7 +442,7 @@ static int pmac_cpufreq_suspend(struct cpufreq_policy *policy) no_schedule = 1; sleep_freq = cur_freq; if (cur_freq == low_freq && !is_pmu_based) - do_set_cpu_speed(CPUFREQ_HIGH, 0); + do_set_cpu_speed(policy, CPUFREQ_HIGH, 0); return 0; } @@ -458,7 +458,7 @@ static int pmac_cpufreq_resume(struct cpufreq_policy *policy) * is that we force a switch to whatever it was, which is * probably high speed due to our suspend() routine */ - do_set_cpu_speed(sleep_freq == low_freq ? + do_set_cpu_speed(policy, sleep_freq == low_freq ? CPUFREQ_LOW : CPUFREQ_HIGH, 0); ppc_proc_freq = cur_freq * 1000ul; diff --git a/arch/powerpc/platforms/powermac/cpufreq_64.c b/arch/powerpc/platforms/powermac/cpufreq_64.c index 9650c60..7ba4234 100644 --- a/arch/powerpc/platforms/powermac/cpufreq_64.c +++ b/arch/powerpc/platforms/powermac/cpufreq_64.c @@ -339,11 +339,10 @@ static int g5_cpufreq_target(struct cpufreq_policy *policy, freqs.old = g5_cpu_freqs[g5_pmode_cur].frequency; freqs.new = g5_cpu_freqs[newstate].frequency; - freqs.cpu = 0; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); rc = g5_switch_freq(newstate); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); mutex_unlock(&g5_switch_mutex); diff --git a/arch/sh/kernel/cpufreq.c b/arch/sh/kernel/cpufreq.c index e68b45b..2c7bd94 100644 --- a/arch/sh/kernel/cpufreq.c +++ b/arch/sh/kernel/cpufreq.c @@ -69,15 +69,14 @@ static int sh_cpufreq_target(struct cpufreq_policy *policy, dev_dbg(dev, "requested frequency %u Hz\n", target_freq * 1000); - freqs.cpu = cpu; freqs.old = sh_cpufreq_get(cpu); freqs.new = (freq + 500) / 1000; freqs.flags = 0; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); set_cpus_allowed_ptr(current, &cpus_allowed); clk_set_rate(cpuclk, freq); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); dev_dbg(dev, "set frequency %lu Hz\n", freq); diff --git a/arch/sparc/kernel/us2e_cpufreq.c b/arch/sparc/kernel/us2e_cpufreq.c index 489fc15..abe963d 100644 --- a/arch/sparc/kernel/us2e_cpufreq.c +++ b/arch/sparc/kernel/us2e_cpufreq.c @@ -248,8 +248,10 @@ static unsigned int us2e_freq_get(unsigned int cpu) return clock_tick / estar_to_divisor(estar); } -static void us2e_set_cpu_divider_index(unsigned int cpu, unsigned int index) +static void us2e_set_cpu_divider_index(struct cpufreq_policy *policy, + unsigned int index) { + unsigned int cpu = policy->cpu; unsigned long new_bits, new_freq; unsigned long clock_tick, divisor, old_divisor, estar; cpumask_t cpus_allowed; @@ -272,14 +274,13 @@ static void us2e_set_cpu_divider_index(unsigned int cpu, unsigned int index) freqs.old = clock_tick / old_divisor; freqs.new = new_freq; - freqs.cpu = cpu; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); if (old_divisor != divisor) us2e_transition(estar, new_bits, clock_tick * 1000, old_divisor, divisor); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); set_cpus_allowed_ptr(current, &cpus_allowed); } @@ -295,7 +296,7 @@ static int us2e_freq_target(struct cpufreq_policy *policy, target_freq, relation, &new_index)) return -EINVAL; - us2e_set_cpu_divider_index(policy->cpu, new_index); + us2e_set_cpu_divider_index(policy, new_index); return 0; } @@ -335,7 +336,7 @@ static int __init us2e_freq_cpu_init(struct cpufreq_policy *policy) static int us2e_freq_cpu_exit(struct cpufreq_policy *policy) { if (cpufreq_us2e_driver) - us2e_set_cpu_divider_index(policy->cpu, 0); + us2e_set_cpu_divider_index(policy, 0); return 0; } diff --git a/arch/sparc/kernel/us3_cpufreq.c b/arch/sparc/kernel/us3_cpufreq.c index eb1624b..7ceb9c8 100644 --- a/arch/sparc/kernel/us3_cpufreq.c +++ b/arch/sparc/kernel/us3_cpufreq.c @@ -96,8 +96,10 @@ static unsigned int us3_freq_get(unsigned int cpu) return ret; } -static void us3_set_cpu_divider_index(unsigned int cpu, unsigned int index) +static void us3_set_cpu_divider_index(struct cpufreq_policy *policy, + unsigned int index) { + unsigned int cpu = policy->cpu; unsigned long new_bits, new_freq, reg; cpumask_t cpus_allowed; struct cpufreq_freqs freqs; @@ -131,14 +133,13 @@ static void us3_set_cpu_divider_index(unsigned int cpu, unsigned int index) freqs.old = get_current_freq(cpu, reg); freqs.new = new_freq; - freqs.cpu = cpu; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); reg &= ~SAFARI_CFG_DIV_MASK; reg |= new_bits; write_safari_cfg(reg); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); set_cpus_allowed_ptr(current, &cpus_allowed); } @@ -156,7 +157,7 @@ static int us3_freq_target(struct cpufreq_policy *policy, &new_index)) return -EINVAL; - us3_set_cpu_divider_index(policy->cpu, new_index); + us3_set_cpu_divider_index(policy, new_index); return 0; } @@ -192,7 +193,7 @@ static int __init us3_freq_cpu_init(struct cpufreq_policy *policy) static int us3_freq_cpu_exit(struct cpufreq_policy *policy) { if (cpufreq_us3_driver) - us3_set_cpu_divider_index(policy->cpu, 0); + us3_set_cpu_divider_index(policy, 0); return 0; } diff --git a/arch/unicore32/kernel/cpu-ucv2.c b/arch/unicore32/kernel/cpu-ucv2.c index 4a99f62..ba5a71c 100644 --- a/arch/unicore32/kernel/cpu-ucv2.c +++ b/arch/unicore32/kernel/cpu-ucv2.c @@ -52,15 +52,14 @@ static int ucv2_target(struct cpufreq_policy *policy, struct cpufreq_freqs freqs; struct clk *mclk = clk_get(NULL, "MAIN_CLK"); - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); if (!clk_set_rate(mclk, target_freq * 1000)) { freqs.old = cur; freqs.new = target_freq; - freqs.cpu = 0; } - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return 0; } diff --git a/drivers/cpufreq/acpi-cpufreq.c b/drivers/cpufreq/acpi-cpufreq.c index 937bc28..afbef9c 100644 --- a/drivers/cpufreq/acpi-cpufreq.c +++ b/drivers/cpufreq/acpi-cpufreq.c @@ -423,7 +423,6 @@ static int acpi_cpufreq_target(struct cpufreq_policy *policy, struct drv_cmd cmd; unsigned int next_state = 0; /* Index into freq_table */ unsigned int next_perf_state = 0; /* Index into perf table */ - unsigned int i; int result = 0; pr_debug("acpi_cpufreq_target %d (%d)\n", target_freq, policy->cpu); @@ -486,10 +485,7 @@ static int acpi_cpufreq_target(struct cpufreq_policy *policy, freqs.old = perf->states[perf->state].core_frequency * 1000; freqs.new = data->freq_table[next_state].frequency; - for_each_cpu(i, policy->cpus) { - freqs.cpu = i; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); - } + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); drv_write(&cmd); @@ -502,10 +498,7 @@ static int acpi_cpufreq_target(struct cpufreq_policy *policy, } } - for_each_cpu(i, policy->cpus) { - freqs.cpu = i; - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); - } + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); perf->state = next_perf_state; out: diff --git a/drivers/cpufreq/cpufreq-cpu0.c b/drivers/cpufreq/cpufreq-cpu0.c index 4e5b7fb..dff4dbc 100644 --- a/drivers/cpufreq/cpufreq-cpu0.c +++ b/drivers/cpufreq/cpufreq-cpu0.c @@ -45,7 +45,7 @@ static int cpu0_set_target(struct cpufreq_policy *policy, struct cpufreq_freqs freqs; struct opp *opp; unsigned long freq_Hz, volt = 0, volt_old = 0, tol = 0; - unsigned int index, cpu; + unsigned int index; int ret; ret = cpufreq_frequency_table_target(policy, freq_table, target_freq, @@ -65,10 +65,7 @@ static int cpu0_set_target(struct cpufreq_policy *policy, if (freqs.old == freqs.new) return 0; - for_each_online_cpu(cpu) { - freqs.cpu = cpu; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); - } + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); if (cpu_reg) { rcu_read_lock(); @@ -117,10 +114,7 @@ static int cpu0_set_target(struct cpufreq_policy *policy, } } - for_each_online_cpu(cpu) { - freqs.cpu = cpu; - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); - } + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return 0; } diff --git a/drivers/cpufreq/cpufreq-nforce2.c b/drivers/cpufreq/cpufreq-nforce2.c index 13d311e..224a478 100644 --- a/drivers/cpufreq/cpufreq-nforce2.c +++ b/drivers/cpufreq/cpufreq-nforce2.c @@ -263,7 +263,6 @@ static int nforce2_target(struct cpufreq_policy *policy, freqs.old = nforce2_get(policy->cpu); freqs.new = target_fsb * fid * 100; - freqs.cpu = 0; /* Only one CPU on nForce2 platforms */ if (freqs.old == freqs.new) return 0; @@ -271,7 +270,7 @@ static int nforce2_target(struct cpufreq_policy *policy, pr_debug("Old CPU frequency %d kHz, new %d kHz\n", freqs.old, freqs.new); - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); /* Disable IRQs */ /* local_irq_save(flags); */ @@ -286,7 +285,7 @@ static int nforce2_target(struct cpufreq_policy *policy, /* Enable IRQs */ /* local_irq_restore(flags); */ - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return 0; } diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index b02824d..2cc6bc4 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -244,19 +244,9 @@ static inline void adjust_jiffies(unsigned long val, struct cpufreq_freqs *ci) #endif -/** - * cpufreq_notify_transition - call notifier chain and adjust_jiffies - * on frequency transition. - * - * This function calls the transition notifiers and the "adjust_jiffies" - * function. It is called twice on all CPU frequency changes that have - * external effects. - */ -void cpufreq_notify_transition(struct cpufreq_freqs *freqs, unsigned int state) +void __cpufreq_notify_transition(struct cpufreq_policy *policy, + struct cpufreq_freqs *freqs, unsigned int state) { - struct cpufreq_policy *policy; - unsigned long flags; - BUG_ON(irqs_disabled()); if (cpufreq_disabled()) @@ -266,10 +256,6 @@ void cpufreq_notify_transition(struct cpufreq_freqs *freqs, unsigned int state) pr_debug("notification %u of frequency transition to %u kHz\n", state, freqs->new); - spin_lock_irqsave(&cpufreq_driver_lock, flags); - policy = per_cpu(cpufreq_cpu_data, freqs->cpu); - spin_unlock_irqrestore(&cpufreq_driver_lock, flags); - switch (state) { case CPUFREQ_PRECHANGE: @@ -303,6 +289,20 @@ void cpufreq_notify_transition(struct cpufreq_freqs *freqs, unsigned int state) break; } } +/** + * cpufreq_notify_transition - call notifier chain and adjust_jiffies + * on frequency transition. + * + * This function calls the transition notifiers and the "adjust_jiffies" + * function. It is called twice on all CPU frequency changes that have + * external effects. + */ +void cpufreq_notify_transition(struct cpufreq_policy *policy, + struct cpufreq_freqs *freqs, unsigned int state) +{ + for_each_cpu(freqs->cpu, policy->cpus) + __cpufreq_notify_transition(policy, freqs, state); +} EXPORT_SYMBOL_GPL(cpufreq_notify_transition); @@ -1134,16 +1134,23 @@ static void handle_update(struct work_struct *work) static void cpufreq_out_of_sync(unsigned int cpu, unsigned int old_freq, unsigned int new_freq) { + struct cpufreq_policy *policy; struct cpufreq_freqs freqs; + unsigned long flags; + pr_debug("Warning: CPU frequency out of sync: cpufreq and timing " "core thinks of %u, is %u kHz.\n", old_freq, new_freq); - freqs.cpu = cpu; freqs.old = old_freq; freqs.new = new_freq; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + + spin_lock_irqsave(&cpufreq_driver_lock, flags); + policy = per_cpu(cpufreq_cpu_data, cpu); + spin_unlock_irqrestore(&cpufreq_driver_lock, flags); + + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); } diff --git a/drivers/cpufreq/dbx500-cpufreq.c b/drivers/cpufreq/dbx500-cpufreq.c index 72f0c3e..7192a6d 100644 --- a/drivers/cpufreq/dbx500-cpufreq.c +++ b/drivers/cpufreq/dbx500-cpufreq.c @@ -55,8 +55,7 @@ static int dbx500_cpufreq_target(struct cpufreq_policy *policy, return 0; /* pre-change notification */ - for_each_cpu(freqs.cpu, policy->cpus) - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); /* update armss clk frequency */ ret = clk_set_rate(armss_clk, freqs.new * 1000); @@ -68,8 +67,7 @@ static int dbx500_cpufreq_target(struct cpufreq_policy *policy, } /* post change notification */ - for_each_cpu(freqs.cpu, policy->cpus) - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return 0; } diff --git a/drivers/cpufreq/e_powersaver.c b/drivers/cpufreq/e_powersaver.c index 3fffbe6..37380fb 100644 --- a/drivers/cpufreq/e_powersaver.c +++ b/drivers/cpufreq/e_powersaver.c @@ -104,7 +104,7 @@ static unsigned int eps_get(unsigned int cpu) } static int eps_set_state(struct eps_cpu_data *centaur, - unsigned int cpu, + struct cpufreq_policy *policy, u32 dest_state) { struct cpufreq_freqs freqs; @@ -112,10 +112,9 @@ static int eps_set_state(struct eps_cpu_data *centaur, int err = 0; int i; - freqs.old = eps_get(cpu); + freqs.old = eps_get(policy->cpu); freqs.new = centaur->fsb * ((dest_state >> 8) & 0xff); - freqs.cpu = cpu; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); /* Wait while CPU is busy */ rdmsr(MSR_IA32_PERF_STATUS, lo, hi); @@ -162,7 +161,7 @@ postchange: current_multiplier); } #endif - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return err; } @@ -190,7 +189,7 @@ static int eps_target(struct cpufreq_policy *policy, /* Make frequency transition */ dest_state = centaur->freq_table[newstate].index & 0xffff; - ret = eps_set_state(centaur, cpu, dest_state); + ret = eps_set_state(centaur, policy, dest_state); if (ret) printk(KERN_ERR "eps: Timeout!\n"); return ret; diff --git a/drivers/cpufreq/elanfreq.c b/drivers/cpufreq/elanfreq.c index 960671f..658d860 100644 --- a/drivers/cpufreq/elanfreq.c +++ b/drivers/cpufreq/elanfreq.c @@ -117,15 +117,15 @@ static unsigned int elanfreq_get_cpu_frequency(unsigned int cpu) * There is no return value. */ -static void elanfreq_set_cpu_state(unsigned int state) +static void elanfreq_set_cpu_state(struct cpufreq_policy *policy, + unsigned int state) { struct cpufreq_freqs freqs; freqs.old = elanfreq_get_cpu_frequency(0); freqs.new = elan_multiplier[state].clock; - freqs.cpu = 0; /* elanfreq.c is UP only driver */ - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); printk(KERN_INFO "elanfreq: attempting to set frequency to %i kHz\n", elan_multiplier[state].clock); @@ -161,7 +161,7 @@ static void elanfreq_set_cpu_state(unsigned int state) udelay(10000); local_irq_enable(); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); }; @@ -188,7 +188,7 @@ static int elanfreq_target(struct cpufreq_policy *policy, target_freq, relation, &newstate)) return -EINVAL; - elanfreq_set_cpu_state(newstate); + elanfreq_set_cpu_state(policy, newstate); return 0; } diff --git a/drivers/cpufreq/exynos-cpufreq.c b/drivers/cpufreq/exynos-cpufreq.c index 78057a3..c0c4ce5 100644 --- a/drivers/cpufreq/exynos-cpufreq.c +++ b/drivers/cpufreq/exynos-cpufreq.c @@ -70,7 +70,6 @@ static int exynos_cpufreq_scale(unsigned int target_freq) freqs.old = policy->cur; freqs.new = target_freq; - freqs.cpu = policy->cpu; if (freqs.new == freqs.old) goto out; @@ -105,8 +104,7 @@ static int exynos_cpufreq_scale(unsigned int target_freq) } arm_volt = volt_table[index]; - for_each_cpu(freqs.cpu, policy->cpus) - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); /* When the new frequency is higher than current frequency */ if ((freqs.new > freqs.old) && !safe_arm_volt) { @@ -131,8 +129,7 @@ static int exynos_cpufreq_scale(unsigned int target_freq) exynos_info->set_freq(old_index, index); - for_each_cpu(freqs.cpu, policy->cpus) - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); /* When the new frequency is lower than current frequency */ if ((freqs.new < freqs.old) || diff --git a/drivers/cpufreq/gx-suspmod.c b/drivers/cpufreq/gx-suspmod.c index 456bee0..3dfc99b 100644 --- a/drivers/cpufreq/gx-suspmod.c +++ b/drivers/cpufreq/gx-suspmod.c @@ -251,14 +251,13 @@ static unsigned int gx_validate_speed(unsigned int khz, u8 *on_duration, * set cpu speed in khz. **/ -static void gx_set_cpuspeed(unsigned int khz) +static void gx_set_cpuspeed(struct cpufreq_policy *policy, unsigned int khz) { u8 suscfg, pmer1; unsigned int new_khz; unsigned long flags; struct cpufreq_freqs freqs; - freqs.cpu = 0; freqs.old = gx_get_cpuspeed(0); new_khz = gx_validate_speed(khz, &gx_params->on_duration, @@ -266,11 +265,9 @@ static void gx_set_cpuspeed(unsigned int khz) freqs.new = new_khz; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); local_irq_save(flags); - - if (new_khz != stock_freq) { /* if new khz == 100% of CPU speed, it is special case */ switch (gx_params->cs55x0->device) { @@ -317,7 +314,7 @@ static void gx_set_cpuspeed(unsigned int khz) gx_params->pci_suscfg = suscfg; - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); pr_debug("suspend modulation w/ duration of ON:%d us, OFF:%d us\n", gx_params->on_duration * 32, gx_params->off_duration * 32); @@ -397,7 +394,7 @@ static int cpufreq_gx_target(struct cpufreq_policy *policy, tmp_freq = gx_validate_speed(tmp_freq, &tmp1, &tmp2); } - gx_set_cpuspeed(tmp_freq); + gx_set_cpuspeed(policy, tmp_freq); return 0; } diff --git a/drivers/cpufreq/imx6q-cpufreq.c b/drivers/cpufreq/imx6q-cpufreq.c index 54e336d..b78bc35 100644 --- a/drivers/cpufreq/imx6q-cpufreq.c +++ b/drivers/cpufreq/imx6q-cpufreq.c @@ -50,7 +50,7 @@ static int imx6q_set_target(struct cpufreq_policy *policy, struct cpufreq_freqs freqs; struct opp *opp; unsigned long freq_hz, volt, volt_old; - unsigned int index, cpu; + unsigned int index; int ret; ret = cpufreq_frequency_table_target(policy, freq_table, target_freq, @@ -68,10 +68,7 @@ static int imx6q_set_target(struct cpufreq_policy *policy, if (freqs.old == freqs.new) return 0; - for_each_online_cpu(cpu) { - freqs.cpu = cpu; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); - } + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); rcu_read_lock(); opp = opp_find_freq_ceil(cpu_dev, &freq_hz); @@ -166,10 +163,7 @@ static int imx6q_set_target(struct cpufreq_policy *policy, } } - for_each_online_cpu(cpu) { - freqs.cpu = cpu; - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); - } + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return 0; } diff --git a/drivers/cpufreq/kirkwood-cpufreq.c b/drivers/cpufreq/kirkwood-cpufreq.c index 0e83e3c..c6e14fb 100644 --- a/drivers/cpufreq/kirkwood-cpufreq.c +++ b/drivers/cpufreq/kirkwood-cpufreq.c @@ -55,7 +55,8 @@ static unsigned int kirkwood_cpufreq_get_cpu_frequency(unsigned int cpu) return kirkwood_freq_table[0].frequency; } -static void kirkwood_cpufreq_set_cpu_state(unsigned int index) +static void kirkwood_cpufreq_set_cpu_state(struct cpufreq_policy *policy, + unsigned int index) { struct cpufreq_freqs freqs; unsigned int state = kirkwood_freq_table[index].index; @@ -63,9 +64,8 @@ static void kirkwood_cpufreq_set_cpu_state(unsigned int index) freqs.old = kirkwood_cpufreq_get_cpu_frequency(0); freqs.new = kirkwood_freq_table[index].frequency; - freqs.cpu = 0; /* Kirkwood is UP */ - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); dev_dbg(priv.dev, "Attempting to set frequency to %i KHz\n", kirkwood_freq_table[index].frequency); @@ -99,7 +99,7 @@ static void kirkwood_cpufreq_set_cpu_state(unsigned int index) local_irq_enable(); } - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); }; static int kirkwood_cpufreq_verify(struct cpufreq_policy *policy) @@ -117,7 +117,7 @@ static int kirkwood_cpufreq_target(struct cpufreq_policy *policy, target_freq, relation, &index)) return -EINVAL; - kirkwood_cpufreq_set_cpu_state(index); + kirkwood_cpufreq_set_cpu_state(policy, index); return 0; } diff --git a/drivers/cpufreq/longhaul.c b/drivers/cpufreq/longhaul.c index 1180d53..b448638 100644 --- a/drivers/cpufreq/longhaul.c +++ b/drivers/cpufreq/longhaul.c @@ -242,7 +242,8 @@ static void do_powersaver(int cx_address, unsigned int mults_index, * Sets a new clock ratio. */ -static void longhaul_setstate(unsigned int table_index) +static void longhaul_setstate(struct cpufreq_policy *policy, + unsigned int table_index) { unsigned int mults_index; int speed, mult; @@ -267,9 +268,8 @@ static void longhaul_setstate(unsigned int table_index) freqs.old = calc_speed(longhaul_get_cpu_mult()); freqs.new = speed; - freqs.cpu = 0; /* longhaul.c is UP only driver */ - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); pr_debug("Setting to FSB:%dMHz Mult:%d.%dx (%s)\n", fsb, mult/10, mult%10, print_speed(speed/1000)); @@ -386,7 +386,7 @@ retry_loop: } } /* Report true CPU frequency */ - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); if (!bm_timeout) printk(KERN_INFO PFX "Warning: Timeout while waiting for " @@ -648,7 +648,7 @@ static int longhaul_target(struct cpufreq_policy *policy, return 0; if (!can_scale_voltage) - longhaul_setstate(table_index); + longhaul_setstate(policy, table_index); else { /* On test system voltage transitions exceeding single * step up or down were turning motherboard off. Both @@ -663,7 +663,7 @@ static int longhaul_target(struct cpufreq_policy *policy, while (i != table_index) { vid = (longhaul_table[i].index >> 8) & 0x1f; if (vid != current_vid) { - longhaul_setstate(i); + longhaul_setstate(policy, i); current_vid = vid; msleep(200); } @@ -672,7 +672,7 @@ static int longhaul_target(struct cpufreq_policy *policy, else i--; } - longhaul_setstate(table_index); + longhaul_setstate(policy, table_index); } longhaul_index = table_index; return 0; @@ -998,15 +998,17 @@ static int __init longhaul_init(void) static void __exit longhaul_exit(void) { + struct cpufreq_policy *policy = cpufreq_cpu_get(0); int i; for (i = 0; i < numscales; i++) { if (mults[i] == maxmult) { - longhaul_setstate(i); + longhaul_setstate(policy, i); break; } } + cpufreq_cpu_put(policy); cpufreq_unregister_driver(&longhaul_driver); kfree(longhaul_table); } diff --git a/drivers/cpufreq/maple-cpufreq.c b/drivers/cpufreq/maple-cpufreq.c index d4c4989..cdd6291 100644 --- a/drivers/cpufreq/maple-cpufreq.c +++ b/drivers/cpufreq/maple-cpufreq.c @@ -158,11 +158,10 @@ static int maple_cpufreq_target(struct cpufreq_policy *policy, freqs.old = maple_cpu_freqs[maple_pmode_cur].frequency; freqs.new = maple_cpu_freqs[newstate].frequency; - freqs.cpu = 0; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); rc = maple_scom_switch_freq(newstate); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); mutex_unlock(&maple_switch_mutex); diff --git a/drivers/cpufreq/omap-cpufreq.c b/drivers/cpufreq/omap-cpufreq.c index 9128c07..b610edd 100644 --- a/drivers/cpufreq/omap-cpufreq.c +++ b/drivers/cpufreq/omap-cpufreq.c @@ -88,16 +88,12 @@ static int omap_target(struct cpufreq_policy *policy, } freqs.old = omap_getspeed(policy->cpu); - freqs.cpu = policy->cpu; if (freqs.old == freqs.new && policy->cur == freqs.new) return ret; /* notifiers */ - for_each_cpu(i, policy->cpus) { - freqs.cpu = i; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); - } + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); freq = freqs.new * 1000; ret = clk_round_rate(mpu_clk, freq); @@ -157,10 +153,7 @@ static int omap_target(struct cpufreq_policy *policy, done: /* notifiers */ - for_each_cpu(i, policy->cpus) { - freqs.cpu = i; - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); - } + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return ret; } diff --git a/drivers/cpufreq/p4-clockmod.c b/drivers/cpufreq/p4-clockmod.c index 827629c9..4b2e773 100644 --- a/drivers/cpufreq/p4-clockmod.c +++ b/drivers/cpufreq/p4-clockmod.c @@ -125,10 +125,7 @@ static int cpufreq_p4_target(struct cpufreq_policy *policy, return 0; /* notifiers */ - for_each_cpu(i, policy->cpus) { - freqs.cpu = i; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); - } + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); /* run on each logical CPU, * see section 13.15.3 of IA32 Intel Architecture Software @@ -138,10 +135,7 @@ static int cpufreq_p4_target(struct cpufreq_policy *policy, cpufreq_p4_setdc(i, p4clockmod_table[newstate].index); /* notifiers */ - for_each_cpu(i, policy->cpus) { - freqs.cpu = i; - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); - } + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return 0; } diff --git a/drivers/cpufreq/pcc-cpufreq.c b/drivers/cpufreq/pcc-cpufreq.c index 503996a..0de0008 100644 --- a/drivers/cpufreq/pcc-cpufreq.c +++ b/drivers/cpufreq/pcc-cpufreq.c @@ -215,8 +215,7 @@ static int pcc_cpufreq_target(struct cpufreq_policy *policy, (pcch_virt_addr + pcc_cpu_data->input_offset)); freqs.new = target_freq; - freqs.cpu = cpu; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); input_buffer = 0x1 | (((target_freq * 100) / (ioread32(&pcch_hdr->nominal) * 1000)) << 8); @@ -237,7 +236,7 @@ static int pcc_cpufreq_target(struct cpufreq_policy *policy, } iowrite16(0, &pcch_hdr->status); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); pr_debug("target: was SUCCESSFUL for cpu %d\n", cpu); spin_unlock(&pcc_lock); diff --git a/drivers/cpufreq/powernow-k6.c b/drivers/cpufreq/powernow-k6.c index af23e0b..ea0222a 100644 --- a/drivers/cpufreq/powernow-k6.c +++ b/drivers/cpufreq/powernow-k6.c @@ -68,7 +68,8 @@ static int powernow_k6_get_cpu_multiplier(void) * * Tries to change the PowerNow! multiplier */ -static void powernow_k6_set_state(unsigned int best_i) +static void powernow_k6_set_state(struct cpufreq_policy *policy, + unsigned int best_i) { unsigned long outvalue = 0, invalue = 0; unsigned long msrval; @@ -81,9 +82,8 @@ static void powernow_k6_set_state(unsigned int best_i) freqs.old = busfreq * powernow_k6_get_cpu_multiplier(); freqs.new = busfreq * clock_ratio[best_i].index; - freqs.cpu = 0; /* powernow-k6.c is UP only driver */ - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); /* we now need to transform best_i to the BVC format, see AMD#23446 */ @@ -98,7 +98,7 @@ static void powernow_k6_set_state(unsigned int best_i) msrval = POWERNOW_IOPORT + 0x0; wrmsr(MSR_K6_EPMR, msrval, 0); /* disable it again */ - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return; } @@ -136,7 +136,7 @@ static int powernow_k6_target(struct cpufreq_policy *policy, target_freq, relation, &newstate)) return -EINVAL; - powernow_k6_set_state(newstate); + powernow_k6_set_state(policy, newstate); return 0; } @@ -182,7 +182,7 @@ static int powernow_k6_cpu_exit(struct cpufreq_policy *policy) unsigned int i; for (i = 0; i < 8; i++) { if (i == max_multiplier) - powernow_k6_set_state(i); + powernow_k6_set_state(policy, i); } cpufreq_frequency_table_put_attr(policy->cpu); return 0; diff --git a/drivers/cpufreq/powernow-k7.c b/drivers/cpufreq/powernow-k7.c index 334cc2f..53888da 100644 --- a/drivers/cpufreq/powernow-k7.c +++ b/drivers/cpufreq/powernow-k7.c @@ -248,7 +248,7 @@ static void change_VID(int vid) } -static void change_speed(unsigned int index) +static void change_speed(struct cpufreq_policy *policy, unsigned int index) { u8 fid, vid; struct cpufreq_freqs freqs; @@ -263,15 +263,13 @@ static void change_speed(unsigned int index) fid = powernow_table[index].index & 0xFF; vid = (powernow_table[index].index & 0xFF00) >> 8; - freqs.cpu = 0; - rdmsrl(MSR_K7_FID_VID_STATUS, fidvidstatus.val); cfid = fidvidstatus.bits.CFID; freqs.old = fsb * fid_codes[cfid] / 10; freqs.new = powernow_table[index].frequency; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); /* Now do the magic poking into the MSRs. */ @@ -292,7 +290,7 @@ static void change_speed(unsigned int index) if (have_a0 == 1) local_irq_enable(); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); } @@ -546,7 +544,7 @@ static int powernow_target(struct cpufreq_policy *policy, relation, &newstate)) return -EINVAL; - change_speed(newstate); + change_speed(policy, newstate); return 0; } diff --git a/drivers/cpufreq/powernow-k8.c b/drivers/cpufreq/powernow-k8.c index d13a136..52137a3 100644 --- a/drivers/cpufreq/powernow-k8.c +++ b/drivers/cpufreq/powernow-k8.c @@ -928,9 +928,10 @@ static int get_transition_latency(struct powernow_k8_data *data) static int transition_frequency_fidvid(struct powernow_k8_data *data, unsigned int index) { + struct cpufreq_policy *policy; u32 fid = 0; u32 vid = 0; - int res, i; + int res; struct cpufreq_freqs freqs; pr_debug("cpu %d transition to index %u\n", smp_processor_id(), index); @@ -959,10 +960,10 @@ static int transition_frequency_fidvid(struct powernow_k8_data *data, freqs.old = find_khz_freq_from_fid(data->currfid); freqs.new = find_khz_freq_from_fid(fid); - for_each_cpu(i, data->available_cores) { - freqs.cpu = i; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); - } + policy = cpufreq_cpu_get(smp_processor_id()); + cpufreq_cpu_put(policy); + + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); res = transition_fid_vid(data, fid, vid); if (res) @@ -970,10 +971,7 @@ static int transition_frequency_fidvid(struct powernow_k8_data *data, freqs.new = find_khz_freq_from_fid(data->currfid); - for_each_cpu(i, data->available_cores) { - freqs.cpu = i; - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); - } + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return res; } diff --git a/drivers/cpufreq/s3c2416-cpufreq.c b/drivers/cpufreq/s3c2416-cpufreq.c index bcc053b..4f1881e 100644 --- a/drivers/cpufreq/s3c2416-cpufreq.c +++ b/drivers/cpufreq/s3c2416-cpufreq.c @@ -256,7 +256,6 @@ static int s3c2416_cpufreq_set_target(struct cpufreq_policy *policy, goto out; } - freqs.cpu = 0; freqs.flags = 0; freqs.old = s3c_freq->is_dvs ? FREQ_DVS : clk_get_rate(s3c_freq->armclk) / 1000; @@ -274,7 +273,7 @@ static int s3c2416_cpufreq_set_target(struct cpufreq_policy *policy, if (!to_dvs && freqs.old == freqs.new) goto out; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); if (to_dvs) { pr_debug("cpufreq: enter dvs\n"); @@ -287,7 +286,7 @@ static int s3c2416_cpufreq_set_target(struct cpufreq_policy *policy, ret = s3c2416_cpufreq_set_armdiv(s3c_freq, freqs.new); } - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); out: mutex_unlock(&cpufreq_lock); diff --git a/drivers/cpufreq/s3c64xx-cpufreq.c b/drivers/cpufreq/s3c64xx-cpufreq.c index 6f9490b..27cacb5 100644 --- a/drivers/cpufreq/s3c64xx-cpufreq.c +++ b/drivers/cpufreq/s3c64xx-cpufreq.c @@ -84,7 +84,6 @@ static int s3c64xx_cpufreq_set_target(struct cpufreq_policy *policy, if (ret != 0) return ret; - freqs.cpu = 0; freqs.old = clk_get_rate(armclk) / 1000; freqs.new = s3c64xx_freq_table[i].frequency; freqs.flags = 0; @@ -95,7 +94,7 @@ static int s3c64xx_cpufreq_set_target(struct cpufreq_policy *policy, pr_debug("Transition %d-%dkHz\n", freqs.old, freqs.new); - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); #ifdef CONFIG_REGULATOR if (vddarm && freqs.new > freqs.old) { @@ -117,7 +116,7 @@ static int s3c64xx_cpufreq_set_target(struct cpufreq_policy *policy, goto err; } - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); #ifdef CONFIG_REGULATOR if (vddarm && freqs.new < freqs.old) { @@ -141,7 +140,7 @@ err_clk: if (clk_set_rate(armclk, freqs.old * 1000) < 0) pr_err("Failed to restore original clock rate\n"); err: - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return ret; } diff --git a/drivers/cpufreq/s5pv210-cpufreq.c b/drivers/cpufreq/s5pv210-cpufreq.c index a484aae..5c77570 100644 --- a/drivers/cpufreq/s5pv210-cpufreq.c +++ b/drivers/cpufreq/s5pv210-cpufreq.c @@ -229,7 +229,6 @@ static int s5pv210_target(struct cpufreq_policy *policy, } freqs.new = s5pv210_freq_table[index].frequency; - freqs.cpu = 0; if (freqs.new == freqs.old) goto exit; @@ -256,7 +255,7 @@ static int s5pv210_target(struct cpufreq_policy *policy, goto exit; } - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); /* Check if there need to change PLL */ if ((index == L0) || (priv_index == L0)) @@ -468,7 +467,7 @@ static int s5pv210_target(struct cpufreq_policy *policy, } } - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); if (freqs.new < freqs.old) { regulator_set_voltage(int_regulator, diff --git a/drivers/cpufreq/sc520_freq.c b/drivers/cpufreq/sc520_freq.c index e42e073..f740b13 100644 --- a/drivers/cpufreq/sc520_freq.c +++ b/drivers/cpufreq/sc520_freq.c @@ -53,7 +53,8 @@ static unsigned int sc520_freq_get_cpu_frequency(unsigned int cpu) } } -static void sc520_freq_set_cpu_state(unsigned int state) +static void sc520_freq_set_cpu_state(struct cpufreq_policy *policy, + unsigned int state) { struct cpufreq_freqs freqs; @@ -61,9 +62,8 @@ static void sc520_freq_set_cpu_state(unsigned int state) freqs.old = sc520_freq_get_cpu_frequency(0); freqs.new = sc520_freq_table[state].frequency; - freqs.cpu = 0; /* AMD Elan is UP */ - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); pr_debug("attempting to set frequency to %i kHz\n", sc520_freq_table[state].frequency); @@ -75,7 +75,7 @@ static void sc520_freq_set_cpu_state(unsigned int state) local_irq_enable(); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); }; static int sc520_freq_verify(struct cpufreq_policy *policy) @@ -93,7 +93,7 @@ static int sc520_freq_target(struct cpufreq_policy *policy, target_freq, relation, &newstate)) return -EINVAL; - sc520_freq_set_cpu_state(newstate); + sc520_freq_set_cpu_state(policy, newstate); return 0; } diff --git a/drivers/cpufreq/spear-cpufreq.c b/drivers/cpufreq/spear-cpufreq.c index 7e4d773..156829f 100644 --- a/drivers/cpufreq/spear-cpufreq.c +++ b/drivers/cpufreq/spear-cpufreq.c @@ -121,7 +121,6 @@ static int spear_cpufreq_target(struct cpufreq_policy *policy, target_freq, relation, &index)) return -EINVAL; - freqs.cpu = policy->cpu; freqs.old = spear_cpufreq_get(0); newfreq = spear_cpufreq.freq_tbl[index].frequency * 1000; @@ -158,8 +157,7 @@ static int spear_cpufreq_target(struct cpufreq_policy *policy, freqs.new = newfreq / 1000; freqs.new /= mult; - for_each_cpu(freqs.cpu, policy->cpus) - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); if (mult == 2) ret = spear1340_set_cpu_rate(srcclk, newfreq); @@ -172,8 +170,7 @@ static int spear_cpufreq_target(struct cpufreq_policy *policy, freqs.new = clk_get_rate(spear_cpufreq.clk) / 1000; } - for_each_cpu(freqs.cpu, policy->cpus) - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return ret; } diff --git a/drivers/cpufreq/speedstep-centrino.c b/drivers/cpufreq/speedstep-centrino.c index 3a953d5..3dbbcc3 100644 --- a/drivers/cpufreq/speedstep-centrino.c +++ b/drivers/cpufreq/speedstep-centrino.c @@ -457,7 +457,7 @@ static int centrino_target (struct cpufreq_policy *policy, unsigned int msr, oldmsr = 0, h = 0, cpu = policy->cpu; struct cpufreq_freqs freqs; int retval = 0; - unsigned int j, k, first_cpu, tmp; + unsigned int j, first_cpu, tmp; cpumask_var_t covered_cpus; if (unlikely(!zalloc_cpumask_var(&covered_cpus, GFP_KERNEL))) @@ -522,13 +522,8 @@ static int centrino_target (struct cpufreq_policy *policy, pr_debug("target=%dkHz old=%d new=%d msr=%04x\n", target_freq, freqs.old, freqs.new, msr); - for_each_cpu(k, policy->cpus) { - if (!cpu_online(k)) - continue; - freqs.cpu = k; - cpufreq_notify_transition(&freqs, + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); - } first_cpu = 0; /* all but 16 LSB are reserved, treat them with care */ @@ -544,12 +539,7 @@ static int centrino_target (struct cpufreq_policy *policy, cpumask_set_cpu(j, covered_cpus); } - for_each_cpu(k, policy->cpus) { - if (!cpu_online(k)) - continue; - freqs.cpu = k; - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); - } + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); if (unlikely(retval)) { /* @@ -565,12 +555,8 @@ static int centrino_target (struct cpufreq_policy *policy, tmp = freqs.new; freqs.new = freqs.old; freqs.old = tmp; - for_each_cpu(j, policy->cpus) { - if (!cpu_online(j)) - continue; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); - } + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); } retval = 0; diff --git a/drivers/cpufreq/speedstep-ich.c b/drivers/cpufreq/speedstep-ich.c index e29b59a..e2e5aa9 100644 --- a/drivers/cpufreq/speedstep-ich.c +++ b/drivers/cpufreq/speedstep-ich.c @@ -263,7 +263,6 @@ static int speedstep_target(struct cpufreq_policy *policy, { unsigned int newstate = 0, policy_cpu; struct cpufreq_freqs freqs; - int i; if (cpufreq_frequency_table_target(policy, &speedstep_freqs[0], target_freq, relation, &newstate)) @@ -272,7 +271,6 @@ static int speedstep_target(struct cpufreq_policy *policy, policy_cpu = cpumask_any_and(policy->cpus, cpu_online_mask); freqs.old = speedstep_get(policy_cpu); freqs.new = speedstep_freqs[newstate].frequency; - freqs.cpu = policy->cpu; pr_debug("transiting from %u to %u kHz\n", freqs.old, freqs.new); @@ -280,18 +278,12 @@ static int speedstep_target(struct cpufreq_policy *policy, if (freqs.old == freqs.new) return 0; - for_each_cpu(i, policy->cpus) { - freqs.cpu = i; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); - } + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); smp_call_function_single(policy_cpu, _speedstep_set_state, &newstate, true); - for_each_cpu(i, policy->cpus) { - freqs.cpu = i; - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); - } + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return 0; } diff --git a/drivers/cpufreq/speedstep-smi.c b/drivers/cpufreq/speedstep-smi.c index 6a457fc..f5a6b70 100644 --- a/drivers/cpufreq/speedstep-smi.c +++ b/drivers/cpufreq/speedstep-smi.c @@ -252,14 +252,13 @@ static int speedstep_target(struct cpufreq_policy *policy, freqs.old = speedstep_freqs[speedstep_get_state()].frequency; freqs.new = speedstep_freqs[newstate].frequency; - freqs.cpu = 0; /* speedstep.c is UP only driver */ if (freqs.old == freqs.new) return 0; - cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_PRECHANGE); speedstep_set_state(newstate); - cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + cpufreq_notify_transition(policy, &freqs, CPUFREQ_POSTCHANGE); return 0; } diff --git a/include/linux/cpufreq.h b/include/linux/cpufreq.h index a22944c..47be364 100644 --- a/include/linux/cpufreq.h +++ b/include/linux/cpufreq.h @@ -268,8 +268,8 @@ int cpufreq_register_driver(struct cpufreq_driver *driver_data); int cpufreq_unregister_driver(struct cpufreq_driver *driver_data); -void cpufreq_notify_transition(struct cpufreq_freqs *freqs, unsigned int state); - +void cpufreq_notify_transition(struct cpufreq_policy *policy, + struct cpufreq_freqs *freqs, unsigned int state); static inline void cpufreq_verify_within_limits(struct cpufreq_policy *policy, unsigned int min, unsigned int max) { -- 1.7.12.rc2.18.g61b472e

12 years, 2 months

4
7
0 0

[RFC PATCH 0/2] sched: Task placement on mixed cpu_power systems

by Morten Rasmussen

In multiprocessor systems with cpus with different compute capabilities it is essential for performance that heavy tasks are scheduled on the most capable cpus. The current scheduler does not handle such performance heterogeneous systems optimally. This patch set proposes a small set of changes that significantly improves performance on these systems. Looking at the current scheduler design the most obvious way to represent the compute capability of each individual cpu is to use cpu_power as this is already used for load-balancing. The recently included entity load-tracking adds the infrastructure to distinguish between heavy and light tasks. The proposed changes moves heavy tasks to cpus with higher cpu_power to get better performance and fixes load-balancing issues for caused by the cpu_power difference when having one heavy task per cpu. The patches requires load-balancing to be based on entity load-tracking and there uses Alex Shi's patch set as the starting point: https://lkml.org/lkml/2013/1/25/767 The patches are based in 3.9-rc2 and have been tested on an ARM vexpress TC2 big.LITTLE testchip containing five cpus: 2xCortex-A15 + 3xCortex-A7. Additional testing and refinements might be needed later as more sophisticated platforms become available. cpu_power A15: 1441 cpu_power A7: 606 Benchmarks: cyclictest: cyclictest -a -t 2 -n -D 10 hackbench: hackbench (default settings) sysbench_1t: sysbench --test=cpu --num-threads=1 --max-requests=1000 run sysbench_2t: sysbench --test=cpu --num-threads=2 --max-requests=1000 run sysbench_5t: sysbench --test=cpu --num-threads=5 --max-requests=1000 run Mixed cpu_power: Average times over 20 runs normalized to 3.9-rc2 (lower is better): 3.9-rc2 +shi +shi+patches Improvement cyclictest AVG 74.9 74.5 75.75 -1.13% MIN 69 69 69 MAX 88 88 94 hackbench AVG 2.17 2.09 2.09 3.90% MIN 2.10 1.95 2.02 MAX 2.25 2.48 2.17 sysbench_1t AVG 25.13* 16.47' 16.48 34.43% MIN 16.47 16.47 16.47 MAX 33.78 16.48 16.54 sysbench_2t AVG 19.32 18.19 16.51 14.55% MIN 16.48 16.47 16.47 MAX 22.15 22.19 16.61 sysbench_5t AVG 27.22 27.71 24.14 11.31% MIN 25.42 27.66 24.04 MAX 27.75 27.86 24.31 * The unpatched 3.9-rc2 scheduler gives inconsistent performance as tasks may randomly be placed on either A7 or A15 cores. The max/min values reflects this behaviour. A15 and A7 performance are ~16.5 and ~33.5 respectively. ' While Alex Shi's patches appear to solve the performance inconsistency for sysbench_1t, it is not the true picture for all workloads. This can be seen for sysbench_2t. To ensure that the proposed changes does not affect normal SMP systems, the same benchmarks have been run on a 2xCortex-A15 configuration as well: SMP: Average times over 20 runs normalized to 3.9-rc2 (lower is better): 3.9-rc2 +shi +shi+patches Improvement cyclictest AVG 78.6 75.3 77.6 1.34% MIN 69 69 69 MAX 135 98 125 hackbench AVG 3.55 3.54 3.55 0.06% MIN 3.51 3.48 3.49 MAX 3.66 3.65 3.67 sysbench_1t AVG 16.48 16.48 16.48 -0.03% MIN 16.47 16.48 16.48 MAX 16.49 16.48 16.48 sysbench_2t AVG 16.53 16.53 16.54 -0.05% MIN 16.47 16.47 16.48 MAX 16.59 16.57 16.59 sysbench_5t AVG 41.16 41.15 41.15 0.04% MIN 41.14 41.13 41.11 MAX 41.35 41.19 41.17 Note: The cpu_power setup code is already present in 3.9-rc2, but the device tree provided for ARM vexpress TC2 is missing frequency information. Adding this will give the cpu_powers listed above. Morten Morten Rasmussen (1): sched: Pull tasks from cpus with multiple tasks when idle Vincent Guittot (1): sched: Force migration on a better cpu kernel/sched/fair.c | 57 +++++++++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 53 insertions(+), 4 deletions(-) -- 1.7.9.5

12 years, 2 months

1
2
0 0

[PATCH v5 0/5] Add ST-Ericsson AB8500 HWMON driver

by Hongbo Zhang

Guenter, Sorry for the delay due to our internal issues. Please check this v5 patches, thanks for all your reviews/comments for this patch set. Anton Vorontsov, I have add your Acked-by: into patch [1/5] (which was [1/3]); and split the old [2/3] into new [2/5], [3/5] and [4/5] patches, so please have a look them again, thank you. v4 -> v5 changes: - split the old [2/3]-ab8500-re-arrange-ab8500-power-and-temperature-data into new three [2/5], [3/5] and [4/5] patches. - hwmon driver minor coding style clean ups: - {} usage in if-else statement in ab8500_read_sensor function - index error fix in gpadc_monitor function - fix issue of clamp_val() usage - remove unnecessary else in function abx500_attrs_visible - remove redundant print message about irq set up - return the calling function return value directly in probe function v3 -> v4 changes: for patch [3/3] - define delays in HZ - update ab8500_read_sensor function, returning temp by parameter - remove ab8500_is_visible function - use clamp_val in set_min and set_max callback - remove unnecessary locks in remove and suspend functions - let abx500 and ab8500 use its own data structure for patch [2/3] - move the data tables from driver/power/ab8500_bmdata.c to include/linux/power/ab8500.h - rename driver/power/ab8500_bmdata.c to driver/power/ab8500_bm.c - rename these variable names to eliminate CamelCase warnings - add const attribute to these data v2 -> v3 changes: - Add interface for converting voltage to temperature - Remove temp5 sensor since we cannot offer temperature read interface of it - Update hyst to use absolute temperature instead of a difference - Add the 3/3 patch v1 -> v2 changes: - Add Documentation/hwmon/abx500 and Documentation/hwmon/abx500 - Make devices which cannot report milli-Celsius invisible - Add temp5_crit interface - Re-work the old find_active_thresholds() to threshold_updated() - Reset updated_min_alarm and updated_max_alarm at the end of each loop - Update the hyst mechamisn to make it works as real hyst - Remove non-stand attributes - Re-order the operations sequence inside probe and remove functions - Update all the lock usages to eliminate race conditions - Make attibutes index starts from 0 also changes: - Since the old [1/2] "ARM: ux500: rename ab8500 to abx500 for hwmon driver" has been merged by Samuel, so won't send it again. - Add another new patch "ab8500_btemp: export two symblols" as [2/2] of this patch set. Hongbo Zhang (5): ab8500_btemp: make ab8500_btemp_get* interfaces public ab8500: power: eliminate CamelCase warning of some variables ab8500: power: add const attributes to some data arrays ab8500: power: export abx500_res_to_temp tables for hwmon hwmon: add ST-Ericsson ABX500 hwmon driver Documentation/hwmon/ab8500 | 22 ++ Documentation/hwmon/abx500 | 28 ++ drivers/hwmon/Kconfig | 13 + drivers/hwmon/Makefile | 1 + drivers/hwmon/ab8500.c | 206 +++++++++++++++ drivers/hwmon/abx500.c | 491 +++++++++++++++++++++++++++++++++++ drivers/hwmon/abx500.h | 69 +++++ drivers/power/ab8500_bmdata.c | 38 +-- drivers/power/ab8500_btemp.c | 5 +- drivers/power/ab8500_fg.c | 4 +- include/linux/mfd/abx500.h | 6 +- include/linux/mfd/abx500/ab8500-bm.h | 5 + include/linux/power/ab8500.h | 16 ++ 13 files changed, 881 insertions(+), 23 deletions(-) create mode 100644 Documentation/hwmon/ab8500 create mode 100644 Documentation/hwmon/abx500 create mode 100644 drivers/hwmon/ab8500.c create mode 100644 drivers/hwmon/abx500.c create mode 100644 drivers/hwmon/abx500.h create mode 100644 include/linux/power/ab8500.h -- 1.8.0

12 years, 2 months

2
10
0 0

[PATCH v9 0/4] Add DRM FIMD DT support for Exynos4 DT Machines

by Vikas Sajjan

This patch series adds support for DRM FIMD DT for Exynos4 DT Machines, specifically for Exynos4412 SoC. changes since v8: - addressed comments to add missing documentation for clock and clock-names properties as pointed out by Sachin Kamat <sachin.kamat(a)linaro.org> changes since v7: - rebased to kgene's "for-next" - Migrated to Common Clock Framework - removed the patch "ARM: dts: Add FIMD AUXDATA node entry for exynos4 DT", as migration to Common Clock Framework will NOT need this. - addressed the comments raised by Sachin Kamat <sachin.kamat(a)linaro.org> changes since v6: - addressed comments and added interrupt-names = "fifo", "vsync", "lcd_sys" in exynos4.dtsi and re-ordered the interrupt numbering to match the order in interrupt combiner IP as suggested by Sylwester Nawrocki <sylvester.nawrocki(a)gmail.com>. changes since v5: - renamed the fimd binding documentation file name as "samsung-fimd.txt", since it not only talks about exynos display controller but also about previous samsung display controllers. - rephrased an abmigious statement about the interrupt combiner in the fimd binding documentation as pointed out by Sachin Kamat <sachin.kamat(a)linaro.org> changes since v4: - moved the fimd binding documentation to Documentation/devicetree/bindings/video/ as suggested by Sylwester Nawrocki <sylvester.nawrocki(a)gmail.com> - added more fimd compatiblity strings in fimd documentation as discussed at https://patchwork.kernel.org/patch/2144861/ with Sylwester Nawrocki <sylvester.nawrocki(a)gmail.com> and Tomasz Figa <tomasz.figa(a)gmail.com> - modified compatible string for exynos4 fimd as "exynos4210-fimd" exynos5 fimd as "exynos5250-fimd" to stick to the rule that compatible value should be named after first specific SoC model in which this particular IP version was included as discussed at https://patchwork.kernel.org/patch/2144861/ - documented more about the interrupt combiner and their order as suggested by Sylwester Nawrocki <sylvester.nawrocki(a)gmail.com> changes since v3: - rebased on http://git.kernel.org/?p=linux/kernel/git/kgene/linux-samsung.git;a=shortlo… changes since v2: - added alias to 'fimd@11c00000' node (reported by: Rahul Sharma <r.sh.open(a)gmail.com>) - removed 'lcd0_data' node as there was already a similar node lcd_data24 (reported by: Jingoo Han <jg1.han(a)samsung.com> - replaced spaces with tabs in display-timing node changes since v1: - added new patch to add FIMD DT binding Documentation - removed patch enabling SAMSUNG_DEV_BACKLIGHT and SAMSUNG_DEV_PMW for mach-exynos4 DT - added 'status' property to fimd DT node Is based on branch kgene's "for-next" https://git.kernel.org/cgit/linux/kernel/git/kgene/linux-samsung.git/log/?h… Sachin Kamat (1): ARM: dts: Add lcd pinctrl node entries for EXYNOS4412 SoC Vikas Sajjan (3): ARM: dts: Add FIMD node to exynos4 ARM: dts: Add FIMD node and display timing node to exynos4412-origen.dts ARM: dts: Add FIMD DT binding Documentation .../devicetree/bindings/video/samsung-fimd.txt | 68 ++++++++++++++++++++ arch/arm/boot/dts/exynos4.dtsi | 11 ++++ arch/arm/boot/dts/exynos4412-origen.dts | 22 +++++++ arch/arm/boot/dts/exynos4x12-pinctrl.dtsi | 14 ++++ 4 files changed, 115 insertions(+) create mode 100644 Documentation/devicetree/bindings/video/samsung-fimd.txt -- 1.7.9.5

12 years, 2 months

5
10
0 0

[ACTIVITY] 2013-03-16 - 2012-03-22

by David Long

=== David Long === === Highlights === * I've set the uprobe xol/VM issue aside for the moment and have been working on trying to separate the instruction simulation code from the k/u-probe code so it may be more cleanly shared. === Plans === * Continue with the code restructuring, return to the xol issue. === Issues === -dl

12 years, 2 months

1
0
0 0

[PATCH 0/2] Add devfreq runtime pm support

by Rajagopal Venkat

Patch to bind devfreq to runtime pm framework. The device devfreq is automatically suspended with pm_runtime_suspend() and resumed with pm_runtime_resume(). Discussed at http://comments.gmane.org/gmane.linux.linaro.devel/13787 Rajagopal Venkat (2): PM / devfreq: Fix compiler warnings PM / devfreq: tie suspend/resume to runtime-pm drivers/base/power/runtime.c | 18 ++++++++++++++- drivers/devfreq/devfreq.c | 51 ++++++++++++++++++++++++++++++++++++++---- include/linux/devfreq.h | 30 +++++++++---------------- include/linux/pm.h | 2 ++ 4 files changed, 76 insertions(+), 25 deletions(-) -- 1.7.10.4

12 years, 2 months

3
6
0 0

[PATCH] cpufreq: acpi-cpufreq: Don't set policy->related_cpus from .init()

by Viresh Kumar

With the addition of following patch: fcf8058 cpufreq: Simplify cpufreq_add_dev() cpufreq driver's .init() routine must initialize policy->cpus with mask of all possible cpus (Online + Offline) that share the clock. Then the core would copy this mask onto policy->related_cpus and will reset policy->cpus to carry only online cpus. acpi-cpufreq driver wasn't updated with this assumption and so sometimes when we try to hot[un]plug cpus at run time, sysfs directories gets corrupted. This patch fixes acpi-cpufreq driver against this corruption. Signed-off-by: Viresh Kumar <viresh.kumar(a)linaro.org> Tested-by: Maciej Rutecki <maciej.rutecki(a)gmail.com> --- drivers/cpufreq/acpi-cpufreq.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/cpufreq/acpi-cpufreq.c b/drivers/cpufreq/acpi-cpufreq.c index afbef9c..11b8b4b 100644 --- a/drivers/cpufreq/acpi-cpufreq.c +++ b/drivers/cpufreq/acpi-cpufreq.c @@ -723,7 +723,6 @@ static int acpi_cpufreq_cpu_init(struct cpufreq_policy *policy) policy->shared_type == CPUFREQ_SHARED_TYPE_ANY) { cpumask_copy(policy->cpus, perf->shared_cpu_map); } - cpumask_copy(policy->related_cpus, perf->shared_cpu_map); #ifdef CONFIG_SMP dmi_check_system(sw_any_bug_dmi_table); @@ -735,7 +734,6 @@ static int acpi_cpufreq_cpu_init(struct cpufreq_policy *policy) if (check_amd_hwpstate_cpu(cpu) && !acpi_pstate_strict) { cpumask_clear(policy->cpus); cpumask_set_cpu(cpu, policy->cpus); - cpumask_copy(policy->related_cpus, cpu_sibling_mask(cpu)); policy->shared_type = CPUFREQ_SHARED_TYPE_HW; pr_info_once(PFX "overriding BIOS provided _PSD data\n"); } -- 1.7.12.rc2.18.g61b472e

12 years, 2 months

1
0
0 0

Re: linaro-kernel Digest, Vol 26, Issue 60

by vipzrx

On Sat 23 Mar 2013 06:52:36 AM CST, linaro-kernel-request(a)lists.linaro.org wrote: > Send linaro-kernel mailing list submissions to > linaro-kernel(a)lists.linaro.org > > To subscribe or unsubscribe via the World Wide Web, visit > http://lists.linaro.org/mailman/listinfo/linaro-kernel > or, via email, send a message with subject or body 'help' to > linaro-kernel-request(a)lists.linaro.org > > You can reach the person managing the list at > linaro-kernel-owner(a)lists.linaro.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of linaro-kernel digest..." > > > Today's Topics: > > 1. Re: [PATCH V3 5/7] mmc: queue work on any cpu (Viresh Kumar) > 2. Re: [PATCH V3 5/7] mmc: queue work on any cpu (Chris Ball) > 3. Re: [PATCH V3 5/7] mmc: queue work on any cpu (Chris Ball) > 4. Re: [PATCH V3 5/7] mmc: queue work on any cpu (Chris Ball) > 5. Re: [PATCH 2/2] PM / devfreq: tie suspend/resume to > runtime-pm (Kevin Hilman) > 6. [ACTIVITY] (Linus Walleij) 2013-02-22 - 2013-03-22 (Linus Walleij) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Fri, 22 Mar 2013 22:57:49 +0530 > From: Viresh Kumar <viresh.kumar(a)linaro.org> > To: Chris Ball <cjb(a)laptop.org> > Cc: venki(a)google.com, linaro-kernel(a)lists.linaro.org, > suresh.b.siddha(a)intel.com, peterz(a)infradead.org, Liviu.Dudau(a)arm.com, > robin.randhawa(a)arm.com, linux-kernel(a)vger.kernel.org, > rostedt(a)goodmis.org, mingo(a)redhat.com, Steve.Bannister(a)arm.com, > tj(a)kernel.org, linux-mmc(a)vger.kernel.org, Arvind.Chauhan(a)arm.com, > pjt(a)google.com, linux-rt-users(a)vger.kernel.org, > charles.garcia-tobin(a)arm.com > Subject: Re: [PATCH V3 5/7] mmc: queue work on any cpu > Message-ID: > <CAKohpomqkYeMgvTpjrUjGs3md=vzguW5Ae6Snf4hzsqtQxNDOw(a)mail.gmail.com> > Content-Type: text/plain; charset=windows-1252 > > On 22 March 2013 22:56, Chris Ball <cjb(a)laptop.org> wrote: >> On Mon, Mar 18 2013, Viresh Kumar wrote: >> >> /home/cjb/git/mmc/drivers/mmc/core/core.c: In function ?mmc_schedule_delayed_work?: >> /home/cjb/git/mmc/drivers/mmc/core/core.c:88:2: error: implicit declaration of function ?queue_delayed_work_on_any_cpu? [-Werror=implicit-function-declaration] >> >> I've dropped this patch for now. This function doesn't seem to be >> defined in linux-next either. > > Hi chris, > > This patch was part of a bigger patchset which also adds this API. I > don't want you to > apply this one but just Ack here. Probably Tejun or some scheduler > maintainer will > apply it later (if they like all patches). > > > > ------------------------------ > > Message: 2 > Date: Fri, 22 Mar 2013 13:09:13 -0400 > From: Chris Ball <cjb(a)laptop.org> > To: Viresh Kumar <viresh.kumar(a)linaro.org> > Cc: venki(a)google.com, linaro-kernel(a)lists.linaro.org, > suresh.b.siddha(a)intel.com, peterz(a)infradead.org, Liviu.Dudau(a)arm.com, > robin.randhawa(a)arm.com, linux-kernel(a)vger.kernel.org, > rostedt(a)goodmis.org, mingo(a)redhat.com, Steve.Bannister(a)arm.com, > tj(a)kernel.org, linux-mmc(a)vger.kernel.org, Arvind.Chauhan(a)arm.com, > pjt(a)google.com, linux-rt-users(a)vger.kernel.org, > charles.garcia-tobin(a)arm.com > Subject: Re: [PATCH V3 5/7] mmc: queue work on any cpu > Message-ID: <87sj3nl5ty.fsf(a)octavius.laptop.org> > Content-Type: text/plain > > Hi, > > On Mon, Mar 18 2013, Viresh Kumar wrote: >> mmc uses workqueues for running mmc_rescan(). There is no real dependency of >> scheduling these on the cpu which scheduled them. >> >> On a idle system, it is observed that and idle cpu wakes up many times just to >> service this work. It would be better if we can schedule it on a cpu which isn't >> idle to save on power. >> >> By idle cpu (from scheduler's perspective) we mean: >> - Current task is idle task >> - nr_running == 0 >> - wake_list is empty >> >> This patch replaces the queue_delayed_work() with >> queue_delayed_work_on_any_cpu() siblings. >> >> This routine would look for the closest (via scheduling domains) non-idle cpu >> (non-idle from schedulers perspective). If the current cpu is not idle or all >> cpus are idle, work will be scheduled on local cpu. >> >> Cc: Chris Ball <cjb(a)laptop.org> >> Cc: linux-mmc(a)vger.kernel.org >> Signed-off-by: Viresh Kumar <viresh.kumar(a)linaro.org> >> --- >> drivers/mmc/core/core.c | 2 +- >> 1 file changed, 1 insertion(+), 1 deletion(-) >> >> diff --git a/drivers/mmc/core/core.c b/drivers/mmc/core/core.c >> index 9290bb5..adf331a 100644 >> --- a/drivers/mmc/core/core.c >> +++ b/drivers/mmc/core/core.c >> @@ -85,7 +85,7 @@ MODULE_PARM_DESC( >> static int mmc_schedule_delayed_work(struct delayed_work *work, >> unsigned long delay) >> { >> - return queue_delayed_work(workqueue, work, delay); >> + return queue_delayed_work_on_any_cpu(workqueue, work, delay); >> } >> >> /* > > Thanks, pushed to mmc-next for 3.10. > > - Chris. where can i find the kernel source for pandaboard es ?

12 years, 2 months

1
0
0 0

[ACTIVITY] (John Stultz) Mar 18-22

by John Stultz

=== Highlights === * Tixy reported issue with the linaro.android-3.9-experimental branch, and I patched it up. * Sent Tixy's issue along with other fixups I had for the linaro.android-3.9 branch to AOSP via gerrit. Arve reviewed and picked them up, so AOSP now matches our tree. * Talked with Jakub about Jira card process * Prepped and did interview #1 for the week, wrote up my summary afterwards * Did a few iterations of review with Serban on his binder patches. * Provided MikeL w/ Timezone coverage info (Linaro covers 71% of timezones, missing only 7!) * Reviewed blueprints and did weekly Android upstreaming subteam email. * Prepped, interviewed and summarized interview #2 this week. * Started reviewing and testing Minchan's latest vrange patches. Sent some questions and a small cleanup fix his way. * Took a second pass at NTP changes required for timekeeping lock hold reductions & sent to Thomas. === Plans === * Still need to sort out Linaro Connect expense reporting. * Probably more interviews, although hopefully not too much more, as they're taking up a big chunk of my time. * Send out timekeeping lock hold time reductions. * Send pull request to tglx for my 3.10 queue. * Still need to work on earlysuspend blog post. * Focus on volatile range work in prep for lsf-mm === Issues === * NA

12 years, 2 months

1
0
0 0

[ACTIVITY] (Linus Walleij) 2013-02-22 - 2013-03-22

by Linus Walleij

== Linus Walleij linusw == === Highlights === * First activity report since Linaro Connect! * Sent first batch of multiplatform patches after finally getting a few cycles to actually fix this. After that Arnd saw that there was not much left and completed the series. (Yay!) We now have ux500 booting in multiplatform config, but the exact format of the patches need to be discussed and finalized. Some guys working on the PRCMU, CPUidle and suspend drivers need to be coordinated. * Collected pinctrl fixes and new patches. * Started to churn through some of the GPIO review backlog. * Initiated a discussion on PCI and VGA consoles, which really took off. This is one of those open items, ARM PCI needs some work for sure. === Plans === * A short paternity leave 6/5->9/5 in may. * Finalize the multiplatform series, coordinate with stakeholders and send a pull request. * Convert Nomadik pinctrl driver to register GPIO ranges from the gpiochip side. * Test the PL08x patches on the Ericsson Research PB11MPCore and submit platform data for using pl08x DMA on that platform. * Look into other Ux500 stuff in need of mainlining... using an internal tracking sheet for this. * Get hands dirty with regmap. === Issues === * Things have been hectic internally at ST-Ericsson diverting me from Linaro work. Thanks, Linus Walleij

12 years, 2 months

1
0
0 0

[PATCH] cpufreq: stats: do cpufreq_cpu_put() corresponding to cpufreq_cpu_get

by Viresh Kumar

In cpufreq_stats_free_sysfs() we aren't balancing calls to cpufreq_cpu_get() with cpufreq_cpu_put(). This will never let us have ref count to policy->kobj as zero. Signed-off-by: Viresh Kumar <viresh.kumar(a)linaro.org> --- drivers/cpufreq/cpufreq_stats.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/drivers/cpufreq/cpufreq_stats.c b/drivers/cpufreq/cpufreq_stats.c index 2fd779e..bfd6273 100644 --- a/drivers/cpufreq/cpufreq_stats.c +++ b/drivers/cpufreq/cpufreq_stats.c @@ -180,15 +180,19 @@ static void cpufreq_stats_free_sysfs(unsigned int cpu) { struct cpufreq_policy *policy = cpufreq_cpu_get(cpu); - if (!cpufreq_frequency_get_table(cpu)) + if (!policy) return; - if (policy && !policy_is_shared(policy)) { + if (!cpufreq_frequency_get_table(cpu)) + goto put_ref; + + if (!policy_is_shared(policy)) { pr_debug("%s: Free sysfs stat\n", __func__); sysfs_remove_group(&policy->kobj, &stats_attr_group); } - if (policy) - cpufreq_cpu_put(policy); + +put_ref: + cpufreq_cpu_put(policy); } static int cpufreq_stats_create_table(struct cpufreq_policy *policy, -- 1.7.12.rc2.18.g61b472e

12 years, 2 months

3
9
0 0

[ACTIVITY] (Ulf Hansson) 2013-03-16 - 2013-03-22

by Ulf Hansson

== Ulf Hansson == === Highlights === Storage: * Reviewing patches on mmc-list. Pushed a patch related to polling card detect. * Discussing sent patchset to enable runtime pm support for mmc/sd block device. * Rework parts of the HS200 and SDR104 support in the mmc protocol layer. First part for tuning sequence done, patch will be pushed to mmc-list shortly. Clk: * Preparing patchset for upstreaming patches that will add support for abx500 clocks. * Preparing patchset to update different driver's clk support used by ux500. * Resent patches for clk_set_parent fixup. * Patches disable unprepared clocks at late init merged by Mike. * Diving into discussion around doing DVFS through the clock API. Reviewing related patches. === Plans === Storage: * Continue the work for the next item in the mmc power management blueprint. * Push patches for mmci host driver, to support UHS cards/HS200, CMD23. * Push patches for mmci host driver to extend the power management support. Context save/restore are for example missing. * Push patches for mmci host driver to add support for new STE 8540 variant. Clk: * Optimizations and bug fixes for ux500 clk implementations. === Issues === * None.

12 years, 2 months

1
0
0 0

[PATCH V2] cpufreq: ARM big LITTLE: Add generic cpufreq driver and its DT glue

by Viresh Kumar

big LITTLE is ARM's new Architecture focussing power/performance needs of modern world. More information about big LITTLE can be found here: http://www.arm.com/products/processors/technologies/biglittleprocessing.php http://lwn.net/Articles/481055/ In order to keep cpufreq support for all big LITTLE platforms simple/generic, this patch tries to add a generic cpufreq driver layer for all big LITTLE platforms. The driver is divided into two parts: - Core driver: Generic and shared across all big LITTLE SoC's - Glue drivers: Per platform drivers providing ops to the core driver This patch adds in a generic glue driver which would extract information from Device Tree. Future SoC's can either reuse the DT glue or write their own depending on the need. Signed-off-by: Sudeep KarkadaNagesha <sudeep.karkadanagesha(a)arm.com> Signed-off-by: Viresh Kumar <viresh.kumar(a)linaro.org> --- V1->V2: - s/IS_ERR_OR_NULL/IS_ERR for return value of clk_get - Added "depends on HAVE_CLK" in Kconfig - Removed "default n" from Kconfig - Renamed cluster clk to "cpu-cluster.%u" and added it as dev_id instead of con_id and using clk_get_sys() instead of clk_get() - Removed setting clk to freq_table to NULL on put_clk_and_freq_table call. .../bindings/cpufreq/arm_big_little_dt.txt | 29 +++ MAINTAINERS | 11 + drivers/cpufreq/Kconfig.arm | 12 + drivers/cpufreq/Makefile | 4 + drivers/cpufreq/arm_big_little.c | 276 +++++++++++++++++++++ drivers/cpufreq/arm_big_little.h | 40 +++ drivers/cpufreq/arm_big_little_dt.c | 125 ++++++++++ 7 files changed, 497 insertions(+) create mode 100644 Documentation/devicetree/bindings/cpufreq/arm_big_little_dt.txt create mode 100644 drivers/cpufreq/arm_big_little.c create mode 100644 drivers/cpufreq/arm_big_little.h create mode 100644 drivers/cpufreq/arm_big_little_dt.c diff --git a/Documentation/devicetree/bindings/cpufreq/arm_big_little_dt.txt b/Documentation/devicetree/bindings/cpufreq/arm_big_little_dt.txt new file mode 100644 index 0000000..6f534eb --- /dev/null +++ b/Documentation/devicetree/bindings/cpufreq/arm_big_little_dt.txt @@ -0,0 +1,29 @@ +Generic ARM big LITTLE cpufreq driver's DT glue +----------------------------------------------- + +It is DT specific glue layer for generic cpufreq driver for big LITTLE systems. + +Both required and optional properties listed below must be defined under node +cluster*. * can be 0 or 1. + +Required properties: +- freqs: List of all supported frequencies. + +Optional properties: +- clock-latency: Specify the possible maximum transition latency for clock, in + unit of nanoseconds. + +Examples: + +cluster0: cluster@0 { + .. + + freqs = <500000000 600000000 700000000 800000000 900000000 1000000000 1100000000 1200000000>; + clock-latency = <200000>; + + .. + + cores { + .. + }; +}; diff --git a/MAINTAINERS b/MAINTAINERS index 554fd30..b14b749 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2206,6 +2206,17 @@ S: Maintained F: drivers/cpufreq/ F: include/linux/cpufreq.h +CPU FREQUENCY DRIVERS - ARM BIG LITTLE +M: Viresh Kumar <viresh.kumar(a)linaro.org> +M: Sudeep KarkadaNagesha <sudeep.karkadanagesha(a)arm.com> +L: cpufreq(a)vger.kernel.org +L: linux-pm(a)vger.kernel.org +W: http://www.arm.com/products/processors/technologies/biglittleprocessing.php +S: Maintained +F: drivers/cpufreq/arm_big_little.h +F: drivers/cpufreq/arm_big_little.c +F: drivers/cpufreq/arm_big_little_dt.c + CPUID/MSR DRIVER M: "H. Peter Anvin" <hpa(a)zytor.com> S: Maintained diff --git a/drivers/cpufreq/Kconfig.arm b/drivers/cpufreq/Kconfig.arm index 030ddf6..87b7e48 100644 --- a/drivers/cpufreq/Kconfig.arm +++ b/drivers/cpufreq/Kconfig.arm @@ -2,6 +2,18 @@ # ARM CPU Frequency scaling drivers # +config ARM_BIG_LITTLE_CPUFREQ + tristate + depends on ARM_CPU_TOPOLOGY + +config ARM_DT_BL_CPUFREQ + tristate "Generic ARM big LITTLE CPUfreq driver probed via DT" + select ARM_BIG_LITTLE_CPUFREQ + depends on OF && HAVE_CLK + help + This enables the Generic CPUfreq driver for ARM big.LITTLE platform. + This gets frequency tables from DT. + config ARM_OMAP2PLUS_CPUFREQ bool "TI OMAP2+" depends on ARCH_OMAP2PLUS diff --git a/drivers/cpufreq/Makefile b/drivers/cpufreq/Makefile index 863fd18..d1b0832 100644 --- a/drivers/cpufreq/Makefile +++ b/drivers/cpufreq/Makefile @@ -44,6 +44,10 @@ obj-$(CONFIG_X86_INTEL_PSTATE) += intel_pstate.o ################################################################################## # ARM SoC drivers +obj-$(CONFIG_ARM_BIG_LITTLE_CPUFREQ) += arm_big_little.o +# big LITTLE per platform glues. Keep DT_BL_CPUFREQ as the last entry in all big +# LITTLE drivers, so that it is probed last. +obj-$(CONFIG_ARM_DT_BL_CPUFREQ) += arm_big_little_dt.o obj-$(CONFIG_UX500_SOC_DB8500) += dbx500-cpufreq.o obj-$(CONFIG_ARM_S3C2416_CPUFREQ) += s3c2416-cpufreq.o obj-$(CONFIG_ARM_S3C64XX_CPUFREQ) += s3c64xx-cpufreq.o diff --git a/drivers/cpufreq/arm_big_little.c b/drivers/cpufreq/arm_big_little.c new file mode 100644 index 0000000..8b8b07e --- /dev/null +++ b/drivers/cpufreq/arm_big_little.c @@ -0,0 +1,276 @@ +/* + * ARM big.LITTLE Platforms CPUFreq support + * + * Copyright (C) 2013 ARM Ltd. + * Sudeep KarkadaNagesha <sudeep.karkadanagesha(a)arm.com> + * + * Copyright (C) 2013 Linaro. + * Viresh Kumar <viresh.kumar(a)linaro.org> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * This program is distributed "as is" WITHOUT ANY WARRANTY of any + * kind, whether express or implied; without even the implied warranty + * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/clk.h> +#include <linux/cpufreq.h> +#include <linux/cpumask.h> +#include <linux/export.h> +#include <linux/of_platform.h> +#include <linux/slab.h> +#include <linux/topology.h> +#include <linux/types.h> + +#include "arm_big_little.h" + +/* Currently we support only two clusters */ +#define MAX_CLUSTERS 2 + +static struct cpufreq_arm_bL_ops *arm_bL_ops; +static struct clk *clk[MAX_CLUSTERS]; +static struct cpufreq_frequency_table *freq_table[MAX_CLUSTERS]; +static atomic_t cluster_usage[MAX_CLUSTERS] = {ATOMIC_INIT(0), ATOMIC_INIT(0)}; + +static int cpu_to_cluster(int cpu) +{ + return topology_physical_package_id(cpu); +} + +static unsigned int bL_cpufreq_get(unsigned int cpu) +{ + u32 cur_cluster = cpu_to_cluster(cpu); + + return clk_get_rate(clk[cur_cluster]) / 1000; +} + +/* Validate policy frequency range */ +static int bL_cpufreq_verify_policy(struct cpufreq_policy *policy) +{ + u32 cur_cluster = cpu_to_cluster(policy->cpu); + + return cpufreq_frequency_table_verify(policy, freq_table[cur_cluster]); +} + +/* Set clock frequency */ +static int bL_cpufreq_set_target(struct cpufreq_policy *policy, + unsigned int target_freq, unsigned int relation) +{ + struct cpufreq_freqs freqs; + u32 cpu = policy->cpu, freq_tab_idx, cur_cluster; + int ret = 0; + + cur_cluster = cpu_to_cluster(policy->cpu); + + freqs.old = bL_cpufreq_get(policy->cpu); + + /* Determine valid target frequency using freq_table */ + cpufreq_frequency_table_target(policy, freq_table[cur_cluster], + target_freq, relation, &freq_tab_idx); + freqs.new = freq_table[cur_cluster][freq_tab_idx].frequency; + + freqs.cpu = policy->cpu; + + pr_debug("%s: cpu: %d, cluster: %d, oldfreq: %d, target freq: %d, new freq: %d\n", + __func__, cpu, cur_cluster, freqs.old, target_freq, + freqs.new); + + if (freqs.old == freqs.new) + return 0; + + for_each_cpu(freqs.cpu, policy->cpus) + cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + + ret = clk_set_rate(clk[cur_cluster], freqs.new * 1000); + if (ret) { + pr_err("clk_set_rate failed: %d\n", ret); + return ret; + } + + policy->cur = freqs.new; + + for_each_cpu(freqs.cpu, policy->cpus) + cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + + return ret; +} + +/* translate the integer array into cpufreq_frequency_table entries */ +struct cpufreq_frequency_table * +arm_bL_copy_table_from_array(unsigned int *table, int count) +{ + int i; + + struct cpufreq_frequency_table *freq_table; + + pr_debug("%s: table: %p, count: %d\n", __func__, table, count); + + freq_table = kmalloc(sizeof(*freq_table) * (count + 1), GFP_KERNEL); + if (!freq_table) + return NULL; + + for (i = 0; i < count; i++) { + pr_debug("%s: index: %d, freq: %d\n", __func__, i, table[i]); + freq_table[i].index = i; + freq_table[i].frequency = table[i]; /* in kHZ */ + } + + freq_table[i].index = count; + freq_table[i].frequency = CPUFREQ_TABLE_END; + + return freq_table; +} +EXPORT_SYMBOL_GPL(arm_bL_copy_table_from_array); + +void arm_bL_free_freq_table(u32 cluster) +{ + pr_debug("%s: free freq table\n", __func__); + + kfree(freq_table[cluster]); +} +EXPORT_SYMBOL_GPL(arm_bL_free_freq_table); + +static void put_cluster_clk_and_freq_table(u32 cluster) +{ + if (!atomic_dec_return(&cluster_usage[cluster])) { + clk_put(clk[cluster]); + arm_bL_ops->put_freq_tbl(cluster); + pr_debug("%s: cluster: %d\n", __func__, cluster); + } +} + +static int get_cluster_clk_and_freq_table(u32 cluster) +{ + char name[14] = "cpu-cluster."; + int count; + + if (atomic_inc_return(&cluster_usage[cluster]) != 1) + return 0; + + freq_table[cluster] = arm_bL_ops->get_freq_tbl(cluster, &count); + if (!freq_table[cluster]) + goto atomic_dec; + + name[12] = cluster + '0'; + clk[cluster] = clk_get_sys(name, NULL); + if (!IS_ERR(clk[cluster])) { + pr_debug("%s: clk: %p & freq table: %p, cluster: %d\n", + __func__, clk[cluster], freq_table[cluster], + cluster); + return 0; + } + + arm_bL_ops->put_freq_tbl(cluster); + +atomic_dec: + atomic_dec(&cluster_usage[cluster]); + pr_err("%s: Failed to get data for cluster: %d\n", __func__, cluster); + return -ENODATA; +} + +/* Per-CPU initialization */ +static int bL_cpufreq_init(struct cpufreq_policy *policy) +{ + u32 cur_cluster = cpu_to_cluster(policy->cpu); + int result; + + result = get_cluster_clk_and_freq_table(cur_cluster); + if (result) + return result; + + result = cpufreq_frequency_table_cpuinfo(policy, + freq_table[cur_cluster]); + if (result) { + pr_err("CPU %d, cluster: %d invalid freq table\n", policy->cpu, + cur_cluster); + put_cluster_clk_and_freq_table(cur_cluster); + return result; + } + + cpufreq_frequency_table_get_attr(freq_table[cur_cluster], policy->cpu); + + policy->cpuinfo.transition_latency = + arm_bL_ops->get_transition_latency(cur_cluster); + policy->cur = bL_cpufreq_get(policy->cpu); + + cpumask_copy(policy->cpus, topology_core_cpumask(policy->cpu)); + + pr_info("CPU %d initialized\n", policy->cpu); + return 0; +} + +static int bL_cpufreq_exit(struct cpufreq_policy *policy) +{ + put_cluster_clk_and_freq_table(cpu_to_cluster(policy->cpu)); + pr_debug("%s: Exited, cpu: %d\n", __func__, policy->cpu); + + return 0; +} + +/* Export freq_table to sysfs */ +static struct freq_attr *bL_cpufreq_attr[] = { + &cpufreq_freq_attr_scaling_available_freqs, + NULL, +}; + +static struct cpufreq_driver bL_cpufreq_driver = { + .name = "arm-big-little", + .flags = CPUFREQ_STICKY, + .verify = bL_cpufreq_verify_policy, + .target = bL_cpufreq_set_target, + .get = bL_cpufreq_get, + .init = bL_cpufreq_init, + .exit = bL_cpufreq_exit, + .attr = bL_cpufreq_attr, +}; + +int bL_cpufreq_register(struct cpufreq_arm_bL_ops *ops) +{ + int ret; + + if (arm_bL_ops) { + pr_debug("%s: Already registered: %s, exiting\n", __func__, + arm_bL_ops->name); + return -EBUSY; + } + + if (!ops || !strlen(ops->name) || !ops->get_freq_tbl) { + pr_err("%s: Invalid arm_bL_ops, exiting\n", __func__); + return -ENODEV; + } + + arm_bL_ops = ops; + + ret = cpufreq_register_driver(&bL_cpufreq_driver); + if (ret) { + pr_info("%s: Failed registering platform driver: %s, err: %d\n", + __func__, ops->name, ret); + arm_bL_ops = NULL; + } else { + pr_info("%s: Registered platform driver: %s\n", __func__, + ops->name); + } + + return ret; +} +EXPORT_SYMBOL_GPL(bL_cpufreq_register); + +void bL_cpufreq_unregister(struct cpufreq_arm_bL_ops *ops) +{ + if (arm_bL_ops != ops) { + pr_info("%s: Registered with: %s, can't unregister, exiting\n", + __func__, arm_bL_ops->name); + } + + cpufreq_unregister_driver(&bL_cpufreq_driver); + pr_info("%s: Un-registered platform driver: %s\n", __func__, + arm_bL_ops->name); + arm_bL_ops = NULL; +} +EXPORT_SYMBOL_GPL(bL_cpufreq_unregister); diff --git a/drivers/cpufreq/arm_big_little.h b/drivers/cpufreq/arm_big_little.h new file mode 100644 index 0000000..049e15d --- /dev/null +++ b/drivers/cpufreq/arm_big_little.h @@ -0,0 +1,40 @@ +/* + * ARM big.LITTLE platform's CPUFreq header file + * + * Copyright (C) 2013 ARM Ltd. + * Sudeep KarkadaNagesha <sudeep.karkadanagesha(a)arm.com> + * + * Copyright (C) 2013 Linaro. + * Viresh Kumar <viresh.kumar(a)linaro.org> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * This program is distributed "as is" WITHOUT ANY WARRANTY of any + * kind, whether express or implied; without even the implied warranty + * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ +#ifndef CPUFREQ_ARM_BIG_LITTLE_H +#define CPUFREQ_ARM_BIG_LITTLE_H + +#include <linux/cpufreq.h> +#include <linux/types.h> + +struct cpufreq_arm_bL_ops { + char name[CPUFREQ_NAME_LEN]; + u32 (*get_transition_latency)(u32 cluster); + struct cpufreq_frequency_table *(*get_freq_tbl)(u32 cluster, + int *count); + void (*put_freq_tbl)(u32 cluster); +}; + +struct cpufreq_frequency_table * +arm_bL_copy_table_from_array(unsigned int *table, int count); +void arm_bL_free_freq_table(u32 cluster); + +int bL_cpufreq_register(struct cpufreq_arm_bL_ops *ops); +void bL_cpufreq_unregister(struct cpufreq_arm_bL_ops *ops); + +#endif /* CPUFREQ_ARM_BIG_LITTLE_H */ diff --git a/drivers/cpufreq/arm_big_little_dt.c b/drivers/cpufreq/arm_big_little_dt.c new file mode 100644 index 0000000..c1eee50 --- /dev/null +++ b/drivers/cpufreq/arm_big_little_dt.c @@ -0,0 +1,125 @@ +/* + * Generic big.LITTLE CPUFreq Interface driver + * + * It provides necessary ops to arm_big_little cpufreq driver and gets + * Frequency information from Device Tree. Freq table in DT must be in KHz. + * + * Copyright (C) 2013 Linaro. + * Viresh Kumar <viresh.kumar(a)linaro.org> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * This program is distributed "as is" WITHOUT ANY WARRANTY of any + * kind, whether express or implied; without even the implied warranty + * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/cpufreq.h> +#include <linux/export.h> +#include <linux/module.h> +#include <linux/of.h> +#include <linux/slab.h> +#include <linux/types.h> +#include "arm_big_little.h" + +static struct cpufreq_frequency_table *generic_get_freq_tbl(u32 cluster, + int *count) +{ + struct device_node *np = NULL; + const struct property *pp; + unsigned int *table = NULL; + int cluster_id; + struct cpufreq_frequency_table *cpufreq_table; + + while ((np = of_find_node_by_name(np, "cluster"))) { + if (of_property_read_u32(np, "reg", &cluster_id)) + continue; + + if (cluster_id != cluster) + continue; + + pp = of_find_property(np, "freqs", NULL); + if (!pp) + continue; + + *count = pp->length / sizeof(u32); + if (!*count) + continue; + + table = kmalloc(sizeof(*table) * (*count), GFP_KERNEL); + if (!table) { + pr_err("%s: Failed to allocate memory for table\n", + __func__); + return NULL; + } + + of_property_read_u32_array(np, "freqs", table, *count); + break; + } + + if (!table) { + pr_err("%s: Unable to retrieve Freq table from Device Tree", + __func__); + return NULL; + } + + cpufreq_table = arm_bL_copy_table_from_array(table, *count); + kfree(table); + + return cpufreq_table; +} + +static void generic_put_freq_tbl(u32 cluster) +{ + arm_bL_free_freq_table(cluster); +} + +static u32 get_transition_latency(u32 cluster) +{ + struct device_node *np = NULL; + u32 transition_latency = CPUFREQ_ETERNAL; + int cluster_id; + + while ((np = of_find_node_by_name(np, "cluster"))) { + if (of_property_read_u32(np, "reg", &cluster_id)) + continue; + + if (cluster_id != cluster) + continue; + + of_property_read_u32(np, "clock-latency", &transition_latency); + return transition_latency; + } + + pr_err("%s: Unable to get clock-latency from DT, use CPUFREQ_ETERNAL", + __func__); + + return transition_latency; +} + +static struct cpufreq_arm_bL_ops generic_bL_ops = { + .name = "generic-bl", + .get_transition_latency = get_transition_latency, + .get_freq_tbl = generic_get_freq_tbl, + .put_freq_tbl = generic_put_freq_tbl, +}; + +static int generic_bL_init(void) +{ + return bL_cpufreq_register(&generic_bL_ops); +} +module_init(generic_bL_init); + +static void generic_bL_exit(void) +{ + return bL_cpufreq_unregister(&generic_bL_ops); +} +module_exit(generic_bL_exit); + +MODULE_DESCRIPTION("Generic ARM big LITTLE cpufreq driver"); +MODULE_LICENSE("GPL"); -- 1.7.12.rc2.18.g61b472e

12 years, 2 months

4
8
0 0

[PATCH] DMA: OF: Check properties value before running be32_to_cpup() on it

by Viresh Kumar

In of_dma_controller_register() routine we are calling of_get_property() as an parameter to be32_to_cpup(). In case the property doesn't exist we will get a crash. This patch changes this code to check if we got a valid property first and then runs be32_to_cpup() on it. Signed-off-by: Viresh Kumar <viresh.kumar(a)linaro.org> --- My mails are broken, i have pushed this patch here: http://git.linaro.org/gitweb?p=people/vireshk/linux.git;a=shortlog;h=refs/h… drivers/dma/of-dma.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/drivers/dma/of-dma.c b/drivers/dma/of-dma.c index 69d04d2..09c7ad1 100644 --- a/drivers/dma/of-dma.c +++ b/drivers/dma/of-dma.c @@ -93,6 +93,7 @@ int of_dma_controller_register(struct device_node *np, { struct of_dma *ofdma; int nbcells; + const __be32 *prop; if (!np || !of_dma_xlate) { pr_err("%s: not enough information provided\n", __func__); @@ -103,8 +104,11 @@ int of_dma_controller_register(struct device_node *np, if (!ofdma) return -ENOMEM; - nbcells = be32_to_cpup(of_get_property(np, "#dma-cells", NULL)); - if (!nbcells) { + prop = of_get_property(np, "#dma-cells", NULL); + if (prop) + nbcells = be32_to_cpup(prop); + + if (!prop || !nbcells) { pr_err("%s: #dma-cells property is missing or invalid\n", __func__); kfree(ofdma); -- 1.7.12.rc2.18.g61b472e

12 years, 2 months

3
4
0 0

[PATCH] cpuidle : handle clockevent notify from the cpuidle framework

by Daniel Lezcano

When a cpu enters a deep idle state, the local timers are stopped and the time framework falls back to the timer device used as a broadcast timer. The different cpuidle drivers are calling clockevents_notify ENTER/EXIT when the idle state stops the local timer. The proposed patch introduces a new flag CPUIDLE_FLAG_TIMER_STOP to let the cpuidle framework to call clockevents_notify instead of duplicating again and again these lines in all the cpuidle drivers. Signed-off-by: Daniel Lezcano <daniel.lezcano(a)linaro.org> --- drivers/cpuidle/cpuidle.c | 9 +++++++++ include/linux/cpuidle.h | 1 + 2 files changed, 10 insertions(+) diff --git a/drivers/cpuidle/cpuidle.c b/drivers/cpuidle/cpuidle.c index eba6929..c500370 100644 --- a/drivers/cpuidle/cpuidle.c +++ b/drivers/cpuidle/cpuidle.c @@ -8,6 +8,7 @@ * This code is licenced under the GPL. */ +#include <linux/clockchips.h> #include <linux/kernel.h> #include <linux/mutex.h> #include <linux/sched.h> @@ -146,12 +147,20 @@ int cpuidle_idle_call(void) trace_cpu_idle_rcuidle(next_state, dev->cpu); + if (drv->states[next_state].flags & CPUIDLE_FLAG_TIMER_STOP) + clockevents_notify(CLOCK_EVT_NOTIFY_BROADCAST_ENTER, + &dev->cpu); + if (cpuidle_state_is_coupled(dev, drv, next_state)) entered_state = cpuidle_enter_state_coupled(dev, drv, next_state); else entered_state = cpuidle_enter_state(dev, drv, next_state); + if (drv->states[next_state].flags & CPUIDLE_FLAG_TIMER_STOP) + clockevents_notify(CLOCK_EVT_NOTIFY_BROADCAST_EXIT, + &dev->cpu); + trace_cpu_idle_rcuidle(PWR_EVENT_EXIT, dev->cpu); /* give the governor an opportunity to reflect on the outcome */ diff --git a/include/linux/cpuidle.h b/include/linux/cpuidle.h index 480c14d..a837b33 100644 --- a/include/linux/cpuidle.h +++ b/include/linux/cpuidle.h @@ -57,6 +57,7 @@ struct cpuidle_state { /* Idle State Flags */ #define CPUIDLE_FLAG_TIME_VALID (0x01) /* is residency time measurable? */ #define CPUIDLE_FLAG_COUPLED (0x02) /* state applies to multiple cpus */ +#define CPUIDLE_FLAG_TIMER_STOP (0x04) /* timer is stopped on this state */ #define CPUIDLE_DRIVER_FLAGS_MASK (0xFFFF0000) -- 1.7.9.5

12 years, 2 months

2
2
0 0

[PATCH v8 0/4] Add DRM FIMD DT support for Exynos4 DT Machines

by Vikas Sajjan

This patch series adds support for DRM FIMD DT for Exynos4 DT Machines, specifically for Exynos4412 SoC. changes since v7: - rebased to kgene's "for-next" - Migrated to Common Clock Framework - removed the patch "ARM: dts: Add FIMD AUXDATA node entry for exynos4 DT", as migration to Common Clock Framework will NOT need this. - addressed the comments raised by Sachin Kamat <sachin.kamat(a)linaro.org> changes since v6: - addressed comments and added interrupt-names = "fifo", "vsync", "lcd_sys" in exynos4.dtsi and re-ordered the interrupt numbering to match the order in interrupt combiner IP as suggested by Sylwester Nawrocki <sylvester.nawrocki(a)gmail.com>. changes since v5: - renamed the fimd binding documentation file name as "samsung-fimd.txt", since it not only talks about exynos display controller but also about previous samsung display controllers. - rephrased an abmigious statement about the interrupt combiner in the fimd binding documentation as pointed out by Sachin Kamat <sachin.kamat(a)linaro.org> changes since v4: - moved the fimd binding documentation to Documentation/devicetree/bindings/video/ as suggested by Sylwester Nawrocki <sylvester.nawrocki(a)gmail.com> - added more fimd compatiblity strings in fimd documentation as discussed at https://patchwork.kernel.org/patch/2144861/ with Sylwester Nawrocki <sylvester.nawrocki(a)gmail.com> and Tomasz Figa <tomasz.figa(a)gmail.com> - modified compatible string for exynos4 fimd as "exynos4210-fimd" exynos5 fimd as "exynos5250-fimd" to stick to the rule that compatible value should be named after first specific SoC model in which this particular IP version was included as discussed at https://patchwork.kernel.org/patch/2144861/ - documented more about the interrupt combiner and their order as suggested by Sylwester Nawrocki <sylvester.nawrocki(a)gmail.com> changes since v3: - rebased on http://git.kernel.org/?p=linux/kernel/git/kgene/linux-samsung.git;a=shortlo… changes since v2: - added alias to 'fimd@11c00000' node (reported by: Rahul Sharma <r.sh.open(a)gmail.com>) - removed 'lcd0_data' node as there was already a similar node lcd_data24 (reported by: Jingoo Han <jg1.han(a)samsung.com> - replaced spaces with tabs in display-timing node changes since v1: - added new patch to add FIMD DT binding Documentation - removed patch enabling SAMSUNG_DEV_BACKLIGHT and SAMSUNG_DEV_PMW for mach-exynos4 DT - added 'status' property to fimd DT node Is based on branch kgene's "for-next" https://git.kernel.org/cgit/linux/kernel/git/kgene/linux-samsung.git/log/?h… Sachin Kamat (1): ARM: dts: Add lcd pinctrl node entries for EXYNOS4412 SoC Vikas Sajjan (3): ARM: dts: Add FIMD node to exynos4 ARM: dts: Add FIMD node and display timing node to exynos4412-origen.dts ARM: dts: Add FIMD DT binding Documentation .../devicetree/bindings/video/samsung-fimd.txt | 61 ++++++++++++++++++++ arch/arm/boot/dts/exynos4.dtsi | 11 ++++ arch/arm/boot/dts/exynos4412-origen.dts | 22 +++++++ arch/arm/boot/dts/exynos4x12-pinctrl.dtsi | 14 +++++ 4 files changed, 108 insertions(+) create mode 100644 Documentation/devicetree/bindings/video/samsung-fimd.txt -- 1.7.9.5

12 years, 2 months

3
6
0 0

[PATCH] ARM: dts: add interrupt-names property to get interrupt resource by name

by Vikas Sajjan

The FIMD driver expects the "vsync" interrupt to be mentioned as the 1st parameter in the FIMD DT node. So to meet this expectation of the driver, the FIMD DT node was forced to be made by keeping "vsync" as the 1st parameter. this resolves the above mentioned "hack" by introducing "interrupt-names", so that FIMD driver can get the interrupt resource by name as discussed at http://www.mail-archive.com/linux-samsung-soc@vger.kernel.org/msg16211.html patch is dependent on https://patchwork.kernel.org/patch/2184981/ Signed-off-by: Vikas Sajjan <vikas.sajjan(a)linaro.org> --- arch/arm/boot/dts/exynos5250.dtsi | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/arm/boot/dts/exynos5250.dtsi b/arch/arm/boot/dts/exynos5250.dtsi index 0ee4706..76c8911 100644 --- a/arch/arm/boot/dts/exynos5250.dtsi +++ b/arch/arm/boot/dts/exynos5250.dtsi @@ -588,6 +588,7 @@ compatible = "samsung,exynos5-fimd"; interrupt-parent = <&combiner>; reg = <0x14400000 0x40000>; - interrupts = <18 5>, <18 4>, <18 6>; + interrupt-names = "fifo", "vsync", "lcd_sys"; + interrupts = <18 4>, <18 5>, <18 6>; }; }; -- 1.7.9.5

12 years, 2 months

6
11
0 0

[PATCH 1/2] cpuidle / kirkwood: remove redundant Kconfig option

by Daniel Lezcano

When the CPU_IDLE and the ARCH_KIRKWOOD options are set it is pointless to define a new option CPU_IDLE_KIRKWOOD because it is redundant. The Makefile drivers directory contains a condition to compile the cpuidle drivers: obj-$(CONFIG_CPU_IDLE) += cpuidle/ Hence, if CPU_IDLE is not set we won't enter this directory. This patch removes the useless Kconfig option and replaces the condition in the Makefile by CONFIG_ARCH_KIRKWOOD. Signed-off-by: Daniel Lezcano <daniel.lezcano(a)linaro.org> --- arch/arm/configs/kirkwood_defconfig | 1 - drivers/cpuidle/Kconfig | 6 ------ drivers/cpuidle/Makefile | 2 +- 3 files changed, 1 insertion(+), 8 deletions(-) diff --git a/arch/arm/configs/kirkwood_defconfig b/arch/arm/configs/kirkwood_defconfig index 13482ea..93f3794 100644 --- a/arch/arm/configs/kirkwood_defconfig +++ b/arch/arm/configs/kirkwood_defconfig @@ -56,7 +56,6 @@ CONFIG_AEABI=y CONFIG_ZBOOT_ROM_TEXT=0x0 CONFIG_ZBOOT_ROM_BSS=0x0 CONFIG_CPU_IDLE=y -CONFIG_CPU_IDLE_KIRKWOOD=y CONFIG_NET=y CONFIG_PACKET=y CONFIG_UNIX=y diff --git a/drivers/cpuidle/Kconfig b/drivers/cpuidle/Kconfig index 071e2c3..c4cc27e 100644 --- a/drivers/cpuidle/Kconfig +++ b/drivers/cpuidle/Kconfig @@ -39,10 +39,4 @@ config CPU_IDLE_CALXEDA help Select this to enable cpuidle on Calxeda processors. -config CPU_IDLE_KIRKWOOD - bool "CPU Idle Driver for Kirkwood processors" - depends on ARCH_KIRKWOOD - help - Select this to enable cpuidle on Kirkwood processors. - endif diff --git a/drivers/cpuidle/Makefile b/drivers/cpuidle/Makefile index 24c6e7d..0d8bd55 100644 --- a/drivers/cpuidle/Makefile +++ b/drivers/cpuidle/Makefile @@ -6,4 +6,4 @@ obj-y += cpuidle.o driver.o governor.o sysfs.o governors/ obj-$(CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED) += coupled.o obj-$(CONFIG_CPU_IDLE_CALXEDA) += cpuidle-calxeda.o -obj-$(CONFIG_CPU_IDLE_KIRKWOOD) += cpuidle-kirkwood.o +obj-$(CONFIG_ARCH_KIRKWOOD) += cpuidle-kirkwood.o -- 1.7.9.5

12 years, 2 months

3
7
0 0

Re: CPU power management bug -- CPU bound task fails to raise CPU frequency

by Vincent Guittot

The bash thread can wake up on a different CPU depending of which CPU has initiated the wake up. (CPU7 or CPU10 on your screenshoot). What is a bit less normal is why your expr tasks migrate in the middle of their execution. On 18 March 2013 20:07, Bruce Dawson <bruced(a)valvesoftware.com> wrote: > BTW, I just uploaded a screenshot of the shell script running. You can see it here: > > http://www.cygnus-software.com/images/ZoomScreenshot_croppedbig.png > > It was made using http://www.rotateright.com/zoom/. The red blocks are the bash process, the other ones are various invocations of expr. > > -----Original Message----- > From: cpufreq-owner(a)vger.kernel.org [mailto:cpufreq-owner@vger.kernel.org] On Behalf Of Bruce Dawson > Sent: Monday, March 18, 2013 10:01 AM > To: 'Vincent Guittot'; Viresh Kumar > Cc: Dave Jones; cpufreq(a)vger.kernel.org; Rafael J. Wysocki; linaro-kernel(a)lists.linaro.org > Subject: RE: CPU power management bug -- CPU bound task fails to raise CPU frequency > > I guess that makes sense for the scheduler to look for the idlest CPU in the system. That's good to know. > > I had guessed that the scheduler would do something like that and that my test would run on two cores. However I find that bash alternates between two cores, which seems odd. Additionally, each invocation of expr starts on one core and moves to another, which seems odd that each invocation lives for a ms or less. The net effect is that six or more different cores get involved. > > Anyway, it is a pathological case, so maybe it doesn't matter, but given the popularity of $(command) in shell scripts it may not be completely irrelevant either. > > -----Original Message----- > From: Vincent Guittot [mailto:vincent.guittot@linaro.org] > Sent: Monday, March 18, 2013 4:07 AM > To: Viresh Kumar > Cc: Bruce Dawson; Dave Jones; cpufreq(a)vger.kernel.org; Rafael J. Wysocki; linaro-kernel(a)lists.linaro.org > Subject: Re: CPU power management bug -- CPU bound task fails to raise CPU frequency > > On 18 March 2013 06:04, Viresh Kumar <viresh.kumar(a)linaro.org> wrote: >> Let me get in the scheduler expert from Linaro (Vincent Guittot, would >> be available after few hours) >> >> Vincent, please start reading from this mail: >> >> http://permalink.gmane.org/gmane.linux.kernel.cpufreq/9675 >> >> Now, we want to understand how to make this task perform better as >> scheduler is using multiple cpus for it and hence all are staying at >> low freqs, as load isn't enough.. > > Hi, > > Your 1st test creates a task to evaluate each expr, and the fork sequence of the scheduler looks for the idlest CPU in the system. > That's explain why your test is evenly spread on all CPUs and the average load of each CPU is below the threshold of cpufreq At the opposite, your 2nd test uses only one task which stays on one CPU and trig the frequency increase. > > I would say that the scheduler behavior is almost normal : spread to get best performance (even if in this use case, the threads run > sequentially) but you have this side effect on the cpufreq thats sees each core individually as not loaded. This example tends to push in favor of a better cooperation between scheduler and cpufreq for sharing statistics > > Vincent > >> >> -- >> viresh >> >> On 18 March 2013 10:28, Bruce Dawson <bruced(a)valvesoftware.com> wrote: >>> This is with the Ondemand governor. >>> >>> The more I ponder this the more I think that the real issue is not the frequency drivers, but the kernel scheduler. The shell script involves two processes being alive at any given time, and one process running (since bash always waits for expr to finish). Therefore the entire task should run on either one core or on two. Instead I see (from looking at thread scheduling graphed using the Zoom profiler -- http://www.rotateright.com/) that it runs on six different cores. bash alternates between two cores, and each invocation of expr is started on one core and then moves to another. Given that it seems not surprising that the CPU frequency management doesn't trigger. >>> >>>> So, the frequency might not be increased if there are multiple cpus >>>> running for a specific task and none of them has high enough load at >>>> that time >>> >>> Yep, that's what I figured. Each cpu's load is quite low -- 20% or lower -- because the work is so spread out. >>> >>> If I run the entire thing under "taskset 1" then everything runs on one core, the frequency elevation happens, and the entire task runs roughly three times faster. >>> >>> Crazy/cool. >>> >>> -----Original Message----- >>> From: viresh.linux(a)gmail.com [mailto:viresh.linux@gmail.com] On >>> Behalf Of Viresh Kumar >>> Sent: Sunday, March 17, 2013 9:32 PM >>> To: Bruce Dawson >>> Cc: Dave Jones; cpufreq(a)vger.kernel.org; Rafael J. Wysocki; >>> linaro-kernel(a)lists.linaro.org >>> Subject: Re: CPU power management bug -- CPU bound task fails to >>> raise CPU frequency >>> >>> On Sun, Mar 17, 2013 at 6:37 AM, Bruce Dawson <bruced(a)valvesoftware.com> wrote: >>>> Dave/others, I've come up with a simple (and real) scenario where a CPU bound task running on Ubuntu (and presumably other Linux flavors) fails to be detected as CPU bound by the Linux kernel, meaning that the CPU continues to run at low speed, meaning that this CPU bound task takes (on my machines) about three times longer to run than it should. >>>> >>>> I found these e-mail addresses in the MAINTAINERS list under CPU FREQUENCY DRIVERS which I'm hoping is the correct area. >>> >>> Yes, cpufreq mailing list is the right list for this. >>> >>>> The basic problem is that on a multi-core system if you run a shell script that spawns lots of sub processes then the workload ends up distributed across all of the CPUs. Therefore, since none of the CPUs are particularly busy, the Linux kernel doesn't realize that a CPU bound task is running, so it leaves the CPU frequency set to low. I have confirmed the behavior in multiple ways. Specifically, I have used "iostat 1" and "mpstat -P ALL 1" to confirm that a full core's worth of CPU work is being done. mpstat also showed that the work was distributed across multiple cores. Using the zoom profiler UI for perf showed the sub processes (and bash) being spread across multiple cores, and perf stat showed that the CPU frequency was staying low even though the task was CPU bound. >>> >>> There are few things which would be helpful to understand what's going on. >>> What governor is used in your case? Probably Ondemand (My ubuntu uses this). >>> >>> Ideally, cpu frequency is increased only if cpu load is very high (or >>> above threshold, >>> 95 in my ubuntu). So, the frequency might not be increased if there are multiple cpus running for a specific task and none of them has high enough load at that time. >>> >>> Other stuff that i suspect here is a bug which was solved recently by >>> below patch. If >>> policy->cpu (that might be cpu 0 for you) is sleeping, then load is >>> never evaluated even >>> if all other cpus are very busy. If you can try below patch then it might be helpful. BTW, you might not be able to apply it easily as it has got lots of dependencies.. and so you might need to pick all drivers/cpufreq patches from v3.9-rc1. >>> >>> commit 2abfa876f1117b0ab45f191fb1f82c41b1cbc8fe >>> Author: Rickard Andersson <rickard.andersson(a)stericsson.com> >>> Date: Thu Dec 27 14:55:38 2012 +0000 >>> >>> cpufreq: handle SW coordinated CPUs >>> >>> This patch fixes a bug that occurred when we had load on a secondary CPU >>> and the primary CPU was sleeping. Only one sampling timer was spawned >>> and it was spawned as a deferred timer on the primary CPU, so when a >>> secondary CPU had a change in load this was not detected by the cpufreq >>> governor (both ondemand and conservative). >>> >>> This patch make sure that deferred timers are run on all CPUs in the >>> case of software controlled CPUs that run on the same frequency. >>> >>> Signed-off-by: Rickard Andersson <rickard.andersson(a)stericsson.com> >>> Signed-off-by: Fabio Baltieri <fabio.baltieri(a)linaro.org> >>> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki(a)intel.com> >>> --- >>> drivers/cpufreq/cpufreq_conservative.c | 3 ++- >>> drivers/cpufreq/cpufreq_governor.c | 44 >>> ++++++++++++++++++++++++++++++++++++++------ >>> drivers/cpufreq/cpufreq_governor.h | 1 + >>> drivers/cpufreq/cpufreq_ondemand.c | 3 ++- >>> 4 files changed, 43 insertions(+), 8 deletions(-) >>> >>> >>>> I have only reproed this behavior on six-core/twelve-thread systems. I would assume that at least a two-core system would be needed to repro this bug, and perhaps more. The bug will not repro if the system is not relatively idle, since a background CPU hog will force the frequency up. >>>> >>>> The repro is exquisitely simple -- ExprCount() is a simplified version of the repro (portable looping in a shell script) and BashCount() is an optimized and less portable version that runs far faster and also avoids this power management problem -- the CPU frequency is raised appropriately. Running a busy loop in another process is another way to get the frequency up and this makes ExprCount() run ~3x faster. Here is the script: >>>> >>>> -------------------------------------- >>>> #!/bin/bash >>>> function ExprCount() { >>>> i=$1 >>>> while [ $i -gt 0 ]; do >>>> i=$(expr $i - 1) >>> >>> I may be wrong but one cpu is used to run this script and other one would be used to run expr program.. So, 2 cpus should be good enough to reproduce this setup. >>> >>> BTW, i have tried your scripts and was able to reproduce the setup >>> here on a 2 cpu >>> 4 thread system. >>> >>> -- >>> viresh > -- > To unsubscribe from this list: send the line "unsubscribe cpufreq" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html

12 years, 2 months

2
1
0 0

[PATCH] drm/exynos: enable FIMD clocks

by Vikas Sajjan

While migrating to common clock framework (CCF), found that the FIMD clocks were pulled down by the CCF. If CCF finds any clock(s) which has NOT been claimed by any of the drivers, then such clock(s) are PULLed low by CCF. By calling clk_prepare_enable() for FIMD clocks fixes the issue. Signed-off-by: Vikas Sajjan <vikas.sajjan(a)linaro.org> --- drivers/gpu/drm/exynos/exynos_drm_fimd.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimd.c b/drivers/gpu/drm/exynos/exynos_drm_fimd.c index 9537761..d93dd8a 100644 --- a/drivers/gpu/drm/exynos/exynos_drm_fimd.c +++ b/drivers/gpu/drm/exynos/exynos_drm_fimd.c @@ -934,6 +934,9 @@ static int fimd_probe(struct platform_device *pdev) return ret; } + clk_prepare_enable(ctx->lcd_clk); + clk_prepare_enable(ctx->bus_clk); + ctx->vidcon0 = pdata->vidcon0; ctx->vidcon1 = pdata->vidcon1; ctx->default_win = pdata->default_win; -- 1.7.9.5

12 years, 2 months

3
2
0 0

[PATCH 0/4][V2] time: dynamic irq affinity

by Daniel Lezcano

When a cpu goes to a deep idle state where its local timer is shutdown, it notifies the time framework to use the broadcast timer instead. Unfortunately, the broadcast device could wake up any CPU, including an idle one which is not concerned by the wake up at all. This implies, in the worst case, an idle CPU will wake up to send an IPI to another idle cpu. This patch solves this by setting the irq affinity to the cpu concerned by the nearest timer event, by this way, the CPU which is wake up is guarantee to be the one concerned by the next event and we are safe with unnecessary wakeup for another idle CPU. As the irq affinity is not supported by all the archs, a flag is needed to specify which clocksource can handle it : CLOCK_EVT_FEAT_DYNIRQ. Tested on a u8500 board with a test program doing indefinitely usleep 10000 wired on each CPU. With dynamic irq affinity: Log is 10.042298 secs long with 4190 events cpu0/state0, 24 hits, total 2718.00us, avg 113.25us, min 0.00us, max 854.00us cpu0/state1, 994 hits, total 9874827.00us, avg 9934.43us, min 30.00us, max 10346.00us cpu1/state0, 73 hits, total 17001.00us, avg 232.89us, min 0.00us, max 10040.00us cpu1/state1, 1002 hits, total 9883840.00us, avg 9864.11us, min 0.00us, max 10742.00us cluster/state0, 0 hits, total 0.00us, avg 0.00us, min 0.00us, max 0.00us cluster/state1, 1931 hits, total 9762328.00us, avg 5055.58us, min 30.00us, max 9308.00us Without dynamic irq affinity: Log is 10.036834 secs long with 6574 events cpu0/state0, 114 hits, total 20107.00us, avg 176.38us, min 0.00us, max 7233.00us cpu0/state1, 1951 hits, total 9833836.00us, avg 5040.41us, min 0.00us, max 9217.00us cpu1/state0, 223 hits, total 21140.00us, avg 94.80us, min 0.00us, max 2960.00us cpu1/state1, 997 hits, total 9879748.00us, avg 9909.48us, min 0.00us, max 10346.00us cluster/state0, 5 hits, total 5462.00us, avg 1092.40us, min 580.00us, max 2899.00us cluster/state1, 2298 hits, total 9740988.00us, avg 4238.90us, min 30.00us, max 9217.00us Results for the specific test case 'usleep 10000' * reduced by 40% the number of wake up on the system * reduced by 49% the number of wake up for CPU0 * increased by factor two idle time for CPU0 * increase by 16% package idle hits + 16% average package idle time Changelog: V2 : * mentioned CLOCK_EVT_FEAT_DYNIRQ flag name in patch description * added comments for CLOCK_EVT_FEAT_DYNIRQ * replaced tick_broadcast_set_affinity parameter to use a cpumask V1 : initial post Daniel Lezcano (3): time : pass broadcast parameter time : set broadcast irq affinity ARM: nomadik: add dynamic irq flag to the timer Viresh Kumar (1): ARM: timer-sp: Set dynamic irq affinity arch/arm/common/timer-sp.c | 3 ++- drivers/clocksource/nomadik-mtu.c | 3 ++- include/linux/clockchips.h | 5 +++++ kernel/time/tick-broadcast.c | 41 +++++++++++++++++++++++++++++-------- 4 files changed, 42 insertions(+), 10 deletions(-) -- 1.7.9.5

12 years, 2 months

5
17
0 0

[PATCH v5] sched: fix init NOHZ_IDLE flag

by Vincent Guittot

On my smp platform which is made of 5 cores in 2 clusters, I have the nr_busy_cpu field of sched_group_power struct that is not null when the platform is fully idle. The root cause is: During the boot sequence, some CPUs reach the idle loop and set their NOHZ_IDLE flag while waiting for others CPUs to boot. But the nr_busy_cpus field is initialized later with the assumption that all CPUs are in the busy state whereas some CPUs have already set their NOHZ_IDLE flag. More generally, the NOHZ_IDLE flag must be initialized when new sched_domains are created in order to ensure that NOHZ_IDLE and nr_busy_cpus are aligned. This condition can be ensured by adding a synchronize_rcu between the destruction of old sched_domains and the creation of new ones so the NOHZ_IDLE flag will not be updated with old sched_domain once it has been initialized. But this solution introduces a additionnal latency in the rebuild sequence that is called during cpu hotplug. As suggested by Frederic Weisbecker, another solution is to have the same rcu lifecycle for both NOHZ_IDLE and sched_domain struct. I have introduce a new sched_domain_rq struct that is the entry point for both sched_domains and objects that must follow the same lifecycle like NOHZ_IDLE flags. They will share the same RCU lifecycle and will be always synchronized. The synchronization is done at the cost of : - an additional indirection for accessing the first sched_domain level - an additional indirection and a rcu_dereference before accessing to the NOHZ_IDLE flag. Change since v4: - link both sched_domain and NOHZ_IDLE flag in one RCU object so their states are always synchronized. Change since V3; - NOHZ flag is not cleared if a NULL domain is attached to the CPU - Remove patch 2/2 which becomes useless with latest modifications Change since V2: - change the initialization to idle state instead of busy state so a CPU that enters idle during the build of the sched_domain will not corrupt the initialization state Change since V1: - remove the patch for SCHED softirq on an idle core use case as it was a side effect of the other use cases. Signed-off-by: Vincent Guittot <vincent.guittot(a)linaro.org> --- include/linux/sched.h | 6 +++ kernel/sched/core.c | 105 ++++++++++++++++++++++++++++++++++++++++++++----- kernel/sched/fair.c | 35 +++++++++++------ kernel/sched/sched.h | 24 +++++++++-- 4 files changed, 145 insertions(+), 25 deletions(-) diff --git a/include/linux/sched.h b/include/linux/sched.h index d35d2b6..2a52188 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -959,6 +959,12 @@ struct sched_domain { unsigned long span[0]; }; +struct sched_domain_rq { + struct sched_domain *sd; + unsigned long flags; + struct rcu_head rcu; /* used during destruction */ +}; + static inline struct cpumask *sched_domain_span(struct sched_domain *sd) { return to_cpumask(sd->span); diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 7f12624..69e2313 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -5602,6 +5602,15 @@ static void destroy_sched_domains(struct sched_domain *sd, int cpu) destroy_sched_domain(sd, cpu); } +static void destroy_sched_domain_rq(struct sched_domain_rq *sd_rq, int cpu) +{ + if (!sd_rq) + return; + + destroy_sched_domains(sd_rq->sd, cpu); + kfree_rcu(sd_rq, rcu); +} + /* * Keep a special pointer to the highest sched_domain that has * SD_SHARE_PKG_RESOURCE set (Last Level Cache Domain) for this @@ -5632,10 +5641,23 @@ static void update_top_cache_domain(int cpu) * hold the hotplug lock. */ static void -cpu_attach_domain(struct sched_domain *sd, struct root_domain *rd, int cpu) +cpu_attach_domain(struct sched_domain_rq *sd_rq, struct root_domain *rd, + int cpu) { struct rq *rq = cpu_rq(cpu); - struct sched_domain *tmp; + struct sched_domain_rq *tmp_rq; + struct sched_domain *tmp, *sd = NULL; + + /* + * If we don't have any sched_domain and associated object, we can + * directly jump to the attach sequence otherwise we try to degenerate + * the sched_domain + */ + if (!sd_rq) + goto attach; + + /* Get a pointer to the 1st sched_domain */ + sd = sd_rq->sd; /* Remove the sched domains which do not contribute to scheduling. */ for (tmp = sd; tmp; ) { @@ -5658,14 +5680,17 @@ cpu_attach_domain(struct sched_domain *sd, struct root_domain *rd, int cpu) destroy_sched_domain(tmp, cpu); if (sd) sd->child = NULL; + /* update sched_domain_rq */ + sd_rq->sd = sd; } +attach: sched_domain_debug(sd, cpu); rq_attach_root(rq, rd); - tmp = rq->sd; - rcu_assign_pointer(rq->sd, sd); - destroy_sched_domains(tmp, cpu); + tmp_rq = rq->sd_rq; + rcu_assign_pointer(rq->sd_rq, sd_rq); + destroy_sched_domain_rq(tmp_rq, cpu); update_top_cache_domain(cpu); } @@ -5695,12 +5720,14 @@ struct sd_data { }; struct s_data { + struct sched_domain_rq ** __percpu sd_rq; struct sched_domain ** __percpu sd; struct root_domain *rd; }; enum s_alloc { sa_rootdomain, + sa_sd_rq, sa_sd, sa_sd_storage, sa_none, @@ -5935,7 +5962,7 @@ static void init_sched_groups_power(int cpu, struct sched_domain *sd) return; update_group_power(sd, cpu); - atomic_set(&sg->sgp->nr_busy_cpus, sg->group_weight); + atomic_set(&sg->sgp->nr_busy_cpus, 0); } int __weak arch_sd_sibling_asym_packing(void) @@ -6011,6 +6038,8 @@ static void set_domain_attribute(struct sched_domain *sd, static void __sdt_free(const struct cpumask *cpu_map); static int __sdt_alloc(const struct cpumask *cpu_map); +static void __sdrq_free(const struct cpumask *cpu_map, struct s_data *d); +static int __sdrq_alloc(const struct cpumask *cpu_map, struct s_data *d); static void __free_domain_allocs(struct s_data *d, enum s_alloc what, const struct cpumask *cpu_map) @@ -6019,6 +6048,9 @@ static void __free_domain_allocs(struct s_data *d, enum s_alloc what, case sa_rootdomain: if (!atomic_read(&d->rd->refcount)) free_rootdomain(&d->rd->rcu); /* fall through */ + case sa_sd_rq: + __sdrq_free(cpu_map, d); /* fall through */ + free_percpu(d->sd_rq); /* fall through */ case sa_sd: free_percpu(d->sd); /* fall through */ case sa_sd_storage: @@ -6038,9 +6070,14 @@ static enum s_alloc __visit_domain_allocation_hell(struct s_data *d, d->sd = alloc_percpu(struct sched_domain *); if (!d->sd) return sa_sd_storage; + d->sd_rq = alloc_percpu(struct sched_domain_rq *); + if (!d->sd_rq) + return sa_sd; + if (__sdrq_alloc(cpu_map, d)) + return sa_sd_rq; d->rd = alloc_rootdomain(); if (!d->rd) - return sa_sd; + return sa_sd_rq; return sa_rootdomain; } @@ -6466,6 +6503,46 @@ static void __sdt_free(const struct cpumask *cpu_map) } } +static int __sdrq_alloc(const struct cpumask *cpu_map, struct s_data *d) +{ + int j; + + for_each_cpu(j, cpu_map) { + struct sched_domain_rq *sd_rq; + + sd_rq = kzalloc_node(sizeof(struct sched_domain_rq), + GFP_KERNEL, cpu_to_node(j)); + if (!sd_rq) + return -ENOMEM; + + *per_cpu_ptr(d->sd_rq, j) = sd_rq; + } + + return 0; +} + +static void __sdrq_free(const struct cpumask *cpu_map, struct s_data *d) +{ + int j; + + for_each_cpu(j, cpu_map) + if (*per_cpu_ptr(d->sd_rq, j)) + kfree(*per_cpu_ptr(d->sd_rq, j)); +} + +static void build_sched_domain_rq(struct s_data *d, int cpu) +{ + struct sched_domain_rq *sd_rq; + struct sched_domain *sd; + + /* Attach sched_domain to sched_domain_rq */ + sd = *per_cpu_ptr(d->sd, cpu); + sd_rq = *per_cpu_ptr(d->sd_rq, cpu); + sd_rq->sd = sd; + /* Init flags */ + set_bit(NOHZ_IDLE, sched_rq_flags(sd_rq)); +} + struct sched_domain *build_sched_domain(struct sched_domain_topology_level *tl, struct s_data *d, const struct cpumask *cpu_map, struct sched_domain_attr *attr, struct sched_domain *child, @@ -6495,6 +6572,7 @@ static int build_sched_domains(const struct cpumask *cpu_map, struct sched_domain_attr *attr) { enum s_alloc alloc_state = sa_none; + struct sched_domain_rq *sd_rq; struct sched_domain *sd; struct s_data d; int i, ret = -ENOMEM; @@ -6547,11 +6625,18 @@ static int build_sched_domains(const struct cpumask *cpu_map, } } + /* Init objects that must follow the sched_domain lifecycle */ + for_each_cpu(i, cpu_map) { + build_sched_domain_rq(&d, i); + } + /* Attach the domains */ rcu_read_lock(); for_each_cpu(i, cpu_map) { - sd = *per_cpu_ptr(d.sd, i); - cpu_attach_domain(sd, d.rd, i); + sd_rq = *per_cpu_ptr(d.sd_rq, i); + cpu_attach_domain(sd_rq, d.rd, i); + /* claim allocation of sched_domain_rq object */ + *per_cpu_ptr(d.sd_rq, i) = NULL; } rcu_read_unlock(); @@ -6982,7 +7067,7 @@ void __init sched_init(void) rq->last_load_update_tick = jiffies; #ifdef CONFIG_SMP - rq->sd = NULL; + rq->sd_rq = NULL; rq->rd = NULL; rq->cpu_power = SCHED_POWER_SCALE; rq->post_schedule = 0; diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 7a33e59..1c7447e 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -5392,31 +5392,39 @@ static inline void nohz_balance_exit_idle(int cpu) static inline void set_cpu_sd_state_busy(void) { + struct sched_domain_rq *sd_rq; struct sched_domain *sd; int cpu = smp_processor_id(); - if (!test_bit(NOHZ_IDLE, nohz_flags(cpu))) - return; - clear_bit(NOHZ_IDLE, nohz_flags(cpu)); - rcu_read_lock(); - for_each_domain(cpu, sd) + sd_rq = get_sched_domain_rq(cpu); + + if (!sd_rq || !test_bit(NOHZ_IDLE, sched_rq_flags(sd_rq))) + goto unlock; + clear_bit(NOHZ_IDLE, sched_rq_flags(sd_rq)); + + for_each_domain_from_rq(sd_rq, sd) atomic_inc(&sd->groups->sgp->nr_busy_cpus); +unlock: rcu_read_unlock(); } void set_cpu_sd_state_idle(void) { + struct sched_domain_rq *sd_rq; struct sched_domain *sd; int cpu = smp_processor_id(); - if (test_bit(NOHZ_IDLE, nohz_flags(cpu))) - return; - set_bit(NOHZ_IDLE, nohz_flags(cpu)); - rcu_read_lock(); - for_each_domain(cpu, sd) + sd_rq = get_sched_domain_rq(cpu); + + if (!sd_rq || test_bit(NOHZ_IDLE, sched_rq_flags(sd_rq))) + goto unlock; + set_bit(NOHZ_IDLE, sched_rq_flags(sd_rq)); + + for_each_domain_from_rq(sd_rq, sd) atomic_dec(&sd->groups->sgp->nr_busy_cpus); +unlock: rcu_read_unlock(); } @@ -5673,7 +5681,12 @@ static void run_rebalance_domains(struct softirq_action *h) static inline int on_null_domain(int cpu) { - return !rcu_dereference_sched(cpu_rq(cpu)->sd); + struct sched_domain_rq *sd_rq = + rcu_dereference_sched(cpu_rq(cpu)->sd_rq); + struct sched_domain *sd = NULL; + if (sd_rq) + sd = sd_rq->sd; + return !sd; } /* diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index cc03cfd..f589306 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -417,7 +417,7 @@ struct rq { #ifdef CONFIG_SMP struct root_domain *rd; - struct sched_domain *sd; + struct sched_domain_rq *sd_rq; unsigned long cpu_power; @@ -505,21 +505,37 @@ DECLARE_PER_CPU(struct rq, runqueues); #ifdef CONFIG_SMP -#define rcu_dereference_check_sched_domain(p) \ +#define rcu_dereference_check_sched_domain_rq(p) \ rcu_dereference_check((p), \ lockdep_is_held(&sched_domains_mutex)) +#define get_sched_domain_rq(cpu) \ + rcu_dereference_check_sched_domain_rq(cpu_rq(cpu)->sd_rq) + +#define rcu_dereference_check_sched_domain(cpu) ({ \ + struct sched_domain_rq *__sd_rq = get_sched_domain_rq(cpu); \ + struct sched_domain *__sd = NULL; \ + if (__sd_rq) \ + __sd = __sd_rq->sd; \ + __sd; \ +}) + +#define sched_rq_flags(sd_rq) (&sd_rq->flags) + /* - * The domain tree (rq->sd) is protected by RCU's quiescent state transition. + * The domain tree (rq->sd_rq) is protected by RCU's quiescent state transition. * See detach_destroy_domains: synchronize_sched for details. * * The domain tree of any CPU may only be accessed from within * preempt-disabled sections. */ #define for_each_domain(cpu, __sd) \ - for (__sd = rcu_dereference_check_sched_domain(cpu_rq(cpu)->sd); \ + for (__sd = rcu_dereference_check_sched_domain(cpu); \ __sd; __sd = __sd->parent) +#define for_each_domain_from_rq(sd_rq, __sd) \ + for (__sd = sd_rq->sd; __sd; __sd = __sd->parent) + #define for_each_lower_domain(sd) for (; sd; sd = sd->child) /** -- 1.7.9.5

12 years, 2 months

1
0
0 0

[PATCH v4 0/3] Add ST-Ericsson AB8500 HWMON driver

by Hongbo Zhang

Guenter, Please check this v4 patches, thanks for all your reviews/comments for this patch set. Anton Vorontsov, I have add your Acked-by: into patch [1/3]; but for this v4 [2/3], we have to change it much according to Guenter's feedback and our internal discussions, so please have a look at this new [2/3] again, thank you. v3 -> v4 changes: for patch [3/3] - define delays in HZ - update ab8500_read_sensor function, returning temp by parameter - remove ab8500_is_visible function - use clamp_val in set_min and set_max callback - remove unnecessary locks in remove and suspend functions - let abx500 and ab8500 use its own data structure for patch [2/3] - move the data tables from driver/power/ab8500_bmdata.c to include/linux/power/ab8500.h - rename driver/power/ab8500_bmdata.c to driver/power/ab8500_bm.c - rename these variable names to eliminate CamelCase warnings - add const attribute to these data v2 -> v3 changes: - Add interface for converting voltage to temperature - Remove temp5 sensor since we cannot offer temperature read interface of it - Update hyst to use absolute temperature instead of a difference - Add the 3/3 patch v1 -> v2 changes: - Add Documentation/hwmon/abx500 and Documentation/hwmon/abx500 - Make devices which cannot report milli-Celsius invisible - Add temp5_crit interface - Re-work the old find_active_thresholds() to threshold_updated() - Reset updated_min_alarm and updated_max_alarm at the end of each loop - Update the hyst mechamisn to make it works as real hyst - Remove non-stand attributes - Re-order the operations sequence inside probe and remove functions - Update all the lock usages to eliminate race conditions - Make attibutes index starts from 0 also changes: - Since the old [1/2] "ARM: ux500: rename ab8500 to abx500 for hwmon driver" has been merged by Samuel, so won't send it again. - Add another new patch "ab8500_btemp: export two symblols" as [2/2] of this patch set. Hongbo Zhang (3): ab8500_btemp: make ab8500_btemp_get* interfaces public ab8500: re-arrange ab8500 power and temperature data tables hwmon: add ST-Ericsson ABX500 hwmon driver Documentation/hwmon/ab8500 | 22 ++ Documentation/hwmon/abx500 | 28 ++ drivers/hwmon/Kconfig | 13 + drivers/hwmon/Makefile | 1 + drivers/hwmon/ab8500.c | 208 ++++++++++++++ drivers/hwmon/abx500.c | 494 +++++++++++++++++++++++++++++++++ drivers/hwmon/abx500.h | 69 +++++ drivers/power/Makefile | 2 +- drivers/power/ab8500_bm.c | 341 +++++++++++++++++++++++ drivers/power/ab8500_bmdata.c | 519 ----------------------------------- drivers/power/ab8500_btemp.c | 5 +- drivers/power/ab8500_fg.c | 4 +- include/linux/mfd/abx500.h | 6 +- include/linux/mfd/abx500/ab8500-bm.h | 5 + include/linux/power/ab8500.h | 189 +++++++++++++ 15 files changed, 1380 insertions(+), 526 deletions(-) create mode 100644 Documentation/hwmon/ab8500 create mode 100644 Documentation/hwmon/abx500 create mode 100644 drivers/hwmon/ab8500.c create mode 100644 drivers/hwmon/abx500.c create mode 100644 drivers/hwmon/abx500.h create mode 100644 drivers/power/ab8500_bm.c delete mode 100644 drivers/power/ab8500_bmdata.c create mode 100644 include/linux/power/ab8500.h -- 1.8.0

12 years, 2 months

3
11
0 0

RE: CPU power management bug -- CPU bound task fails to raise CPU frequency

by Viresh Kumar

---------- Forwarded message ---------- From: <linaro-kernel-bounces(a)lists.linaro.org> Date: 18 March 2013 22:31 Subject: Auto-discard notification To: linaro-kernel-owner(a)lists.linaro.org The attached message has been automatically discarded by lists.linaro.org and hence i am forwarding it again. -- viresh ---------- Forwarded message ---------- From: Bruce Dawson <bruced(a)valvesoftware.com> To: 'Vincent Guittot' <vincent.guittot(a)linaro.org>, Viresh Kumar <viresh.kumar(a)linaro.org> Cc: Dave Jones <davej(a)redhat.com>, "cpufreq(a)vger.kernel.org" <cpufreq(a)vger.kernel.org>, "Rafael J. Wysocki" <rjw(a)sisk.pl>, "linaro-kernel(a)lists.linaro.org" <linaro-kernel(a)lists.linaro.org> Date: Mon, 18 Mar 2013 17:01:17 +0000 Subject: RE: CPU power management bug -- CPU bound task fails to raise CPU frequency I guess that makes sense for the scheduler to look for the idlest CPU in the system. That's good to know. I had guessed that the scheduler would do something like that and that my test would run on two cores. However I find that bash alternates between two cores, which seems odd. Additionally, each invocation of expr starts on one core and moves to another, which seems odd that each invocation lives for a ms or less. The net effect is that six or more different cores get involved. Anyway, it is a pathological case, so maybe it doesn't matter, but given the popularity of $(command) in shell scripts it may not be completely irrelevant either. -----Original Message----- From: Vincent Guittot [mailto:vincent.guittot@linaro.org] Sent: Monday, March 18, 2013 4:07 AM To: Viresh Kumar Cc: Bruce Dawson; Dave Jones; cpufreq(a)vger.kernel.org; Rafael J. Wysocki; linaro-kernel(a)lists.linaro.org Subject: Re: CPU power management bug -- CPU bound task fails to raise CPU frequency On 18 March 2013 06:04, Viresh Kumar <viresh.kumar(a)linaro.org> wrote: > Let me get in the scheduler expert from Linaro (Vincent Guittot, would > be available after few hours) > > Vincent, please start reading from this mail: > > http://permalink.gmane.org/gmane.linux.kernel.cpufreq/9675 > > Now, we want to understand how to make this task perform better as > scheduler is using multiple cpus for it and hence all are staying at > low freqs, as load isn't enough.. Hi, Your 1st test creates a task to evaluate each expr, and the fork sequence of the scheduler looks for the idlest CPU in the system. That's explain why your test is evenly spread on all CPUs and the average load of each CPU is below the threshold of cpufreq At the opposite, your 2nd test uses only one task which stays on one CPU and trig the frequency increase. I would say that the scheduler behavior is almost normal : spread to get best performance (even if in this use case, the threads run sequentially) but you have this side effect on the cpufreq thats sees each core individually as not loaded. This example tends to push in favor of a better cooperation between scheduler and cpufreq for sharing statistics Vincent > > -- > viresh > > On 18 March 2013 10:28, Bruce Dawson <bruced(a)valvesoftware.com> wrote: >> This is with the Ondemand governor. >> >> The more I ponder this the more I think that the real issue is not the frequency drivers, but the kernel scheduler. The shell script involves two processes being alive at any given time, and one process running (since bash always waits for expr to finish). Therefore the entire task should run on either one core or on two. Instead I see (from looking at thread scheduling graphed using the Zoom profiler -- http://www.rotateright.com/) that it runs on six different cores. bash alternates between two cores, and each invocation of expr is started on one core and then moves to another. Given that it seems not surprising that the CPU frequency management doesn't trigger. >> >>> So, the frequency might not be increased if there are multiple cpus >>> running for a specific task and none of them has high enough load at >>> that time >> >> Yep, that's what I figured. Each cpu's load is quite low -- 20% or lower -- because the work is so spread out. >> >> If I run the entire thing under "taskset 1" then everything runs on one core, the frequency elevation happens, and the entire task runs roughly three times faster. >> >> Crazy/cool. >> >> -----Original Message----- >> From: viresh.linux(a)gmail.com [mailto:viresh.linux@gmail.com] On >> Behalf Of Viresh Kumar >> Sent: Sunday, March 17, 2013 9:32 PM >> To: Bruce Dawson >> Cc: Dave Jones; cpufreq(a)vger.kernel.org; Rafael J. Wysocki; >> linaro-kernel(a)lists.linaro.org >> Subject: Re: CPU power management bug -- CPU bound task fails to >> raise CPU frequency >> >> On Sun, Mar 17, 2013 at 6:37 AM, Bruce Dawson <bruced(a)valvesoftware.com> wrote: >>> Dave/others, I've come up with a simple (and real) scenario where a CPU bound task running on Ubuntu (and presumably other Linux flavors) fails to be detected as CPU bound by the Linux kernel, meaning that the CPU continues to run at low speed, meaning that this CPU bound task takes (on my machines) about three times longer to run than it should. >>> >>> I found these e-mail addresses in the MAINTAINERS list under CPU FREQUENCY DRIVERS which I'm hoping is the correct area. >> >> Yes, cpufreq mailing list is the right list for this. >> >>> The basic problem is that on a multi-core system if you run a shell script that spawns lots of sub processes then the workload ends up distributed across all of the CPUs. Therefore, since none of the CPUs are particularly busy, the Linux kernel doesn't realize that a CPU bound task is running, so it leaves the CPU frequency set to low. I have confirmed the behavior in multiple ways. Specifically, I have used "iostat 1" and "mpstat -P ALL 1" to confirm that a full core's worth of CPU work is being done. mpstat also showed that the work was distributed across multiple cores. Using the zoom profiler UI for perf showed the sub processes (and bash) being spread across multiple cores, and perf stat showed that the CPU frequency was staying low even though the task was CPU bound. >> >> There are few things which would be helpful to understand what's going on. >> What governor is used in your case? Probably Ondemand (My ubuntu uses this). >> >> Ideally, cpu frequency is increased only if cpu load is very high (or >> above threshold, >> 95 in my ubuntu). So, the frequency might not be increased if there are multiple cpus running for a specific task and none of them has high enough load at that time. >> >> Other stuff that i suspect here is a bug which was solved recently by >> below patch. If >> policy->cpu (that might be cpu 0 for you) is sleeping, then load is >> never evaluated even >> if all other cpus are very busy. If you can try below patch then it might be helpful. BTW, you might not be able to apply it easily as it has got lots of dependencies.. and so you might need to pick all drivers/cpufreq patches from v3.9-rc1. >> >> commit 2abfa876f1117b0ab45f191fb1f82c41b1cbc8fe >> Author: Rickard Andersson <rickard.andersson(a)stericsson.com> >> Date: Thu Dec 27 14:55:38 2012 +0000 >> >> cpufreq: handle SW coordinated CPUs >> >> This patch fixes a bug that occurred when we had load on a secondary CPU >> and the primary CPU was sleeping. Only one sampling timer was spawned >> and it was spawned as a deferred timer on the primary CPU, so when a >> secondary CPU had a change in load this was not detected by the cpufreq >> governor (both ondemand and conservative). >> >> This patch make sure that deferred timers are run on all CPUs in the >> case of software controlled CPUs that run on the same frequency. >> >> Signed-off-by: Rickard Andersson <rickard.andersson(a)stericsson.com> >> Signed-off-by: Fabio Baltieri <fabio.baltieri(a)linaro.org> >> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki(a)intel.com> >> --- >> drivers/cpufreq/cpufreq_conservative.c | 3 ++- >> drivers/cpufreq/cpufreq_governor.c | 44 >> ++++++++++++++++++++++++++++++++++++++------ >> drivers/cpufreq/cpufreq_governor.h | 1 + >> drivers/cpufreq/cpufreq_ondemand.c | 3 ++- >> 4 files changed, 43 insertions(+), 8 deletions(-) >> >> >>> I have only reproed this behavior on six-core/twelve-thread systems. I would assume that at least a two-core system would be needed to repro this bug, and perhaps more. The bug will not repro if the system is not relatively idle, since a background CPU hog will force the frequency up. >>> >>> The repro is exquisitely simple -- ExprCount() is a simplified version of the repro (portable looping in a shell script) and BashCount() is an optimized and less portable version that runs far faster and also avoids this power management problem -- the CPU frequency is raised appropriately. Running a busy loop in another process is another way to get the frequency up and this makes ExprCount() run ~3x faster. Here is the script: >>> >>> -------------------------------------- >>> #!/bin/bash >>> function ExprCount() { >>> i=$1 >>> while [ $i -gt 0 ]; do >>> i=$(expr $i - 1) >> >> I may be wrong but one cpu is used to run this script and other one would be used to run expr program.. So, 2 cpus should be good enough to reproduce this setup. >> >> BTW, i have tried your scripts and was able to reproduce the setup >> here on a 2 cpu >> 4 thread system. >> >> -- >> viresh

12 years, 2 months

1
0
0 0

[ACTIVITY] (John Stultz) Mar 11-15

by John Stultz

=== Highlights === * My Linaro connect talk was covered on lwn here: https://lwn.net/Articles/542466/ * Worked with tglx to try to sort out remaining issues on reducing timekeeping lock hold time * Got my CLOCK_TAI patches properly tested and ready for 3.10 * Got a bunch of community timekeeping changes queued for 3.10 * Re-reported git server errors folks in the community were seeing w/ our trees. * Sent out weekly email Android Upstreaming status report (canned hangout, since we all were in HK). * Reviewed Mathieu's sysrq timeout patch. * Pointed Axel at ioctl work being done in trinity, and he pointed out that others are extending it for Android testing. * Preped and setup interviews with two Linaro candidates. * Looked over DmitryP's netfilter patch queue, and sent back some minor comments. * Got the linaro.android branch updated to 3.9, using the experimental/android-3.9 branch on AOSP. Thanks to Tixy for helping with testing! * Took a look at the ion driver to see what shape it was in for upstreaming. It has some arm-specific assumptions, so it may need some rework. * Synced with Greg and AndrzejP on the removal of CCG from staging (since its basically abandoned for the configfs gadget). Benoit from Google chimed in as well here. * Did an initial review of Serban's binder patches === Plans === * Sort out Linaro Connect expense reporting * Another Linaro candidate interview. * Review new volatile range patches from Minchan * Try to finish up timekeeping lock hold time reductions * Send pull request to tglx for my 3.10 queue * Work on earlysuspend blog post * Maybe send out pstore updates from Google? === Issues === * Jet lag returning from Linaro Connect was pretty bad. Think I'm finally over it though.

12 years, 2 months

2
1
0 0

Re: CPU power management bug -- CPU bound task fails to raise CPU frequency

by Viresh Kumar

On Sun, Mar 17, 2013 at 6:37 AM, Bruce Dawson <bruced(a)valvesoftware.com> wrote: > Dave/others, I've come up with a simple (and real) scenario where a CPU bound task running on Ubuntu (and presumably other Linux flavors) fails to be detected as CPU bound by the Linux kernel, meaning that the CPU continues to run at low speed, meaning that this CPU bound task takes (on my machines) about three times longer to run than it should. > > I found these e-mail addresses in the MAINTAINERS list under CPU FREQUENCY DRIVERS which I'm hoping is the correct area. Yes, cpufreq mailing list is the right list for this. > The basic problem is that on a multi-core system if you run a shell script that spawns lots of sub processes then the workload ends up distributed across all of the CPUs. Therefore, since none of the CPUs are particularly busy, the Linux kernel doesn't realize that a CPU bound task is running, so it leaves the CPU frequency set to low. I have confirmed the behavior in multiple ways. Specifically, I have used "iostat 1" and "mpstat -P ALL 1" to confirm that a full core's worth of CPU work is being done. mpstat also showed that the work was distributed across multiple cores. Using the zoom profiler UI for perf showed the sub processes (and bash) being spread across multiple cores, and perf stat showed that the CPU frequency was staying low even though the task was CPU bound. There are few things which would be helpful to understand what's going on. What governor is used in your case? Probably Ondemand (My ubuntu uses this). Ideally, cpu frequency is increased only if cpu load is very high (or above threshold, 95 in my ubuntu). So, the frequency might not be increased if there are multiple cpus running for a specific task and none of them has high enough load at that time. Other stuff that i suspect here is a bug which was solved recently by below patch. If policy->cpu (that might be cpu 0 for you) is sleeping, then load is never evaluated even if all other cpus are very busy. If you can try below patch then it might be helpful. BTW, you might not be able to apply it easily as it has got lots of dependencies.. and so you might need to pick all drivers/cpufreq patches from v3.9-rc1. commit 2abfa876f1117b0ab45f191fb1f82c41b1cbc8fe Author: Rickard Andersson <rickard.andersson(a)stericsson.com> Date: Thu Dec 27 14:55:38 2012 +0000 cpufreq: handle SW coordinated CPUs This patch fixes a bug that occurred when we had load on a secondary CPU and the primary CPU was sleeping. Only one sampling timer was spawned and it was spawned as a deferred timer on the primary CPU, so when a secondary CPU had a change in load this was not detected by the cpufreq governor (both ondemand and conservative). This patch make sure that deferred timers are run on all CPUs in the case of software controlled CPUs that run on the same frequency. Signed-off-by: Rickard Andersson <rickard.andersson(a)stericsson.com> Signed-off-by: Fabio Baltieri <fabio.baltieri(a)linaro.org> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki(a)intel.com> --- drivers/cpufreq/cpufreq_conservative.c | 3 ++- drivers/cpufreq/cpufreq_governor.c | 44 ++++++++++++++++++++++++++++++++++++++------ drivers/cpufreq/cpufreq_governor.h | 1 + drivers/cpufreq/cpufreq_ondemand.c | 3 ++- 4 files changed, 43 insertions(+), 8 deletions(-) > I have only reproed this behavior on six-core/twelve-thread systems. I would assume that at least a two-core system would be needed to repro this bug, and perhaps more. The bug will not repro if the system is not relatively idle, since a background CPU hog will force the frequency up. > > The repro is exquisitely simple -- ExprCount() is a simplified version of the repro (portable looping in a shell script) and BashCount() is an optimized and less portable version that runs far faster and also avoids this power management problem -- the CPU frequency is raised appropriately. Running a busy loop in another process is another way to get the frequency up and this makes ExprCount() run ~3x faster. Here is the script: > > -------------------------------------- > #!/bin/bash > function ExprCount() { > i=$1 > while [ $i -gt 0 ]; do > i=$(expr $i - 1) I may be wrong but one cpu is used to run this script and other one would be used to run expr program.. So, 2 cpus should be good enough to reproduce this setup. BTW, i have tried your scripts and was able to reproduce the setup here on a 2 cpu 4 thread system. -- viresh

12 years, 2 months

3
3
0 0

[ACTIVITY] 2013-3-2 - 2012-03-15

by David Long

=== David Long === === Highlights === * Attended Connect. The keynotes were particulaly good this time. Kudos to all who arranged this and those who arranged the weather (not that I got outside much). * Although the uprobe code is installing and taking the breakpoint properly, it is getting lost somewhere when it goes to execute the probed instruction out-of-line. I've tried to isolate this a couple of different ways, but no success yet. === Plans === * Continue isolating the xol problem and restructing the emulation code. === Issues === -dl

12 years, 2 months

1
0
0 0

LCA13: Video: Scheduler Internals by: Vincent Guittot

by Viresh Kumar

Hi Guys, Below are hangout upstreams of Scheduler Internals by Vincent Guittot done in LCA13. We have got another version of this recording that is done by some other cameras, but its size was 30 GB and so hard to upstream. In case you need that please contact me. Day 1: http://www.youtube.com/watch?v=2yzelou80JE Day 2: http://www.youtube.com/watch?v=fN11Lltx1nQ Thanks to Naresh for arranging for hangouts. -- Viresh

12 years, 2 months

4
5
0 0

lists.linaro.org maintenance completed

by IT Services

This is to let you know that the migration of lists.linaro.org has been successfully completed. As per the email I sent on Wednesday, it may take some time for the new address of the server to be seen by your computer. You can check this by trying to connect to the web site: http://lists.linaro.org/ If you are able to connect and you do not get an error, this means you are connecting to the new server and you can send email to the lists. If you experience any problems after the weekend and you find that you still cannot connect to the server, please reply to this email to let us know. Regards Philip IT Services Manager Linaro

12 years, 2 months

1
0
0 0

[ACTIVITY] (Ulf Hansson) 2013-03-02 - 2013-03-15

by Ulf Hansson

== Ulf Hansson == === Highlights === General: * Last week spent at Linaro Connect Hong Kong. A great week! Storage: * Reviewing patches on mmc-list. * Discussing sent patchset to enable runtime pm support for mmc/sd block device. * Rework parts of the HS200 and SDR104 support in the mmc protocol layer. First part for tuning sequence done, patch will be pushed to mmc-list shortly. Clk: * Preparing patchset for upstreaming patches that will add support for abx500 clocks. * Preparing patchset to update different driver's clk support used by ux500. * Resent patches for clk_set_parent fixup and for disable unprepared clocks at late init. * Diving into discussion around doing DVFS through the clock API. Reviewing related patches. === Plans === Storage: * Speed up work around the mmc power management blueprint so we can finalize this work as soon as possible. * Push patches for mmci host driver, to support UHS cards/HS200, CMD23. * Push patches for mmci host driver to extend the power management support. Context save/restore are for example missing. * Push patches for mmci host driver to add support for new STE 8540 variant. Clk: * Optimizations and bug fixes for ux500 clk implementations. * Send RFC/PATCH for DVFS clock type used by ux500. * Follow up on previously sent patchset. === Issues === * None.

12 years, 2 months

1
0
0 0

Re: cpufreq cowork with scheduler

by Viresh Kumar

On Fri, Mar 15, 2013 at 11:54 AM, Chao Xie <xiechao.mail(a)gmail.com> wrote: > hi > It may be a old topic. > Now the cpufreq governors will sample for system work load. The > schduler knows about the current workload of each cores. So why not > make use of it? The sampling need take some time, so when the cpufreq > increase the frequency , the system has been busy for a period of > time. Making use of the schduler information can reduce time spending > at sampling. Hi Chao, I am working for Linaro Power Management Working Group and we know about this problem or solution. We have a dedicated blueprint towards this goal: https://blueprints.launchpad.net/linaro-big-little-system/+spec/sched-coope… -- viresh

12 years, 2 months

1
0
0 0

[PATCH] drm/exynos: change the method for getting the interrupt resource of FIMD

by Vikas Sajjan

Replaces the "platform_get_resource() for IORESOURCE_IRQ" with platform_get_resource_byname(). Both in exynos4 and exynos5, FIMD IP has 3 interrupts in the order: "fifo", "vsync", and "lcd_sys". But The FIMD driver expects the "vsync" interrupt to be mentioned as the 1st parameter in the FIMD DT node. So to meet this expectation of the driver, the FIMD DT node was forced to be made by keeping "vsync" as the 1st paramter. For example in exynos4, the FIMD DT node has interrupt numbers mentioned as <11, 1> <11, 0> <11, 2> keeping "vsync" as the 1st paramter. This patch fixes the above mentioned "hack" of re-ordering of the FIMD interrupt numbers by getting interrupt resource of FIMD by using platform_get_resource_byname(). Signed-off-by: Vikas Sajjan <vikas.sajjan(a)linaro.org> --- drivers/gpu/drm/exynos/exynos_drm_fimd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimd.c b/drivers/gpu/drm/exynos/exynos_drm_fimd.c index 1ea173a..cd79d38 100644 --- a/drivers/gpu/drm/exynos/exynos_drm_fimd.c +++ b/drivers/gpu/drm/exynos/exynos_drm_fimd.c @@ -945,7 +945,7 @@ static int fimd_probe(struct platform_device *pdev) return -ENXIO; } - res = platform_get_resource(pdev, IORESOURCE_IRQ, 0); + res = platform_get_resource_byname(pdev, IORESOURCE_IRQ, "vsync"); if (!res) { dev_err(dev, "irq request failed.\n"); return -ENXIO; -- 1.7.9.5

12 years, 2 months

1
0
0 0

[PATCH v7 0/5] Add DRM FIMD DT support for Exynos4 DT Machines

by Vikas Sajjan

This patch series adds support for DRM FIMD DT for Exynos4 DT Machines, specifically for Exynos4412 SoC. changes since v6: - addressed comments and added interrupt-names = "fifo", "vsync", "lcd_sys" in exynos4.dtsi and re-ordered the interrupt numbering to match the order in interrupt combiner IP as suggested by Sylwester Nawrocki <sylvester.nawrocki(a)gmail.com>. changes since v5: - renamed the fimd binding documentation file name as "samsung-fimd.txt", since it not only talks about exynos display controller but also about previous samsung display controllers. - rephrased an abmigious statement about the interrupt combiner in the fimd binding documentation as pointed out by Sachin Kamat <sachin.kamat(a)linaro.org> changes since v4: - moved the fimd binding documentation to Documentation/devicetree/bindings/video/ as suggested by Sylwester Nawrocki <sylvester.nawrocki(a)gmail.com> - added more fimd compatiblity strings in fimd documentation as discussed at https://patchwork.kernel.org/patch/2144861/ with Sylwester Nawrocki <sylvester.nawrocki(a)gmail.com> and Tomasz Figa <tomasz.figa(a)gmail.com> - modified compatible string for exynos4 fimd as "exynos4210-fimd" exynos5 fimd as "exynos5250-fimd" to stick to the rule that compatible value should be named after first specific SoC model in which this particular IP version was included as discussed at https://patchwork.kernel.org/patch/2144861/ - documented more about the interrupt combiner and their order as suggested by Sylwester Nawrocki <sylvester.nawrocki(a)gmail.com> changes since v3: - rebased on http://git.kernel.org/?p=linux/kernel/git/kgene/linux-samsung.git;a=shortlo… changes since v2: - added alias to 'fimd@11c00000' node (reported by: Rahul Sharma <r.sh.open(a)gmail.com>) - removed 'lcd0_data' node as there was already a similar node lcd_data24 (reported by: Jingoo Han <jg1.han(a)samsung.com> - replaced spaces with tabs in display-timing node changes since v1: - added new patch to add FIMD DT binding Documentation - removed patch enabling SAMSUNG_DEV_BACKLIGHT and SAMSUNG_DEV_PMW for mach-exynos4 DT - added 'status' property to fimd DT node Is based on branch "for-next-next" http://git.kernel.org/?p=linux/kernel/git/kgene/linux-samsung.git;a=shortlo… Sachin Kamat (1): ARM: dts: Add lcd pinctrl node entries for EXYNOS4412 SoC Vikas Sajjan (4): ARM: dts: Add FIMD node to exynos4 ARM: dts: Add FIMD node and display timing node to exynos4412-origen.dts ARM: dts: Add FIMD AUXDATA node entry for exynos4 DT ARM: dts: Add FIMD DT binding Documentation .../devicetree/bindings/video/samsung-fimd.txt | 58 ++++++++++++++++++++ arch/arm/boot/dts/exynos4.dtsi | 8 +++ arch/arm/boot/dts/exynos4412-origen.dts | 22 ++++++++ arch/arm/boot/dts/exynos4x12-pinctrl.dtsi | 14 +++++ arch/arm/mach-exynos/mach-exynos4-dt.c | 2 + 5 files changed, 104 insertions(+) create mode 100644 Documentation/devicetree/bindings/video/samsung-fimd.txt -- 1.7.9.5

12 years, 2 months

2
7
0 0

Planned maintenance - lists.linaro.org

by IT Services

Hello You are receiving this email because you are subscribed to one or more mailing lists provided by the lists.linaro.org server. IT Services are announcing planned maintenance for this server scheduled for *Friday 15th March 2013, starting at 2pm GMT*. The purpose of the work is to move the service to another server. There will be some disruption during this maintenance. In order to ensure that you do not accidentally try to use the service while it is being moved, the current server will be shut down at 2pm. A further email will be sent on Friday afternoon to confirm that the migration of the service is completed. However, due to the way servers are found, it may take a while before your computer is able to connect to the relocated service. After the old server has been shut down, email sent to any of the lists will be queued, but it is possible that the sending server will still trying to deliver the email to the old server rather than the new one when it is started. It is therefore *strongly* recommended that you do not send any email to an @lists.linaro.org email address until you can connect to the new service, which you will be able to test by trying to use a web browser to connect to http://lists.linaro.org after you receive the email confirming that the migration has been completed. Since the old service will be shut down, if you are able to connect, you can be sure you have connected to the new service. If by Monday you are still unable to connect to the service or you are not able to send email to an @lists.linaro.org email address, please send an email to its(a)linaro.org. Thank you. Regards Philip IT Services Manager Linaro

12 years, 2 months

1
0
0 0

[PATCH 1/3] workqueue: define mod_scheduled_delayed_work[_on]() helpers

by Viresh Kumar

The only difference between schedule_delayed_work[_on]() and queue_delayed_work[_on]() is the workqueue, work is scheduled on. We may need to modify the delay for works queued with schedule_delayed_work[_on]() calls and thus adding these helpers. First users of these new helpers is cpufreq governors which need to modify the delay for its works. Cc: Tejun Heo <tj(a)kernel.org> Signed-off-by: Viresh Kumar <viresh.kumar(a)linaro.org> --- include/linux/workqueue.h | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/include/linux/workqueue.h b/include/linux/workqueue.h index 2b58905..864c2b3 100644 --- a/include/linux/workqueue.h +++ b/include/linux/workqueue.h @@ -412,6 +412,7 @@ extern bool schedule_delayed_work_on(int cpu, struct delayed_work *work, extern bool schedule_delayed_work(struct delayed_work *work, unsigned long delay); extern int schedule_on_each_cpu(work_func_t func); + extern int keventd_up(void); int execute_in_process_context(work_func_t fn, struct execute_work *); @@ -465,6 +466,11 @@ static inline long work_on_cpu(unsigned int cpu, long (*fn)(void *), void *arg) long work_on_cpu(unsigned int cpu, long (*fn)(void *), void *arg); #endif /* CONFIG_SMP */ +#define mod_scheduled_delayed_work_on(cpu, dwork, delay) \ + mod_delayed_work_on(cpu, system_wq, dwork, delay) +#define mod_scheduled_delayed_work(dwork, delay) \ + mod_delayed_work(system_wq, dwork, delay) + #ifdef CONFIG_FREEZER extern void freeze_workqueues_begin(void); extern bool freeze_workqueues_busy(void); -- 1.7.12.rc2.18.g61b472e

12 years, 2 months

3
8
0 0

[PATCH 0/4] time: dynamic irq affinity

by Daniel Lezcano

When a cpu goes to a deep idle state where its local timer is shutdown, it notifies the time framework to use the broadcast timer instead. Unfortunately, the broadcast device could wake up any CPU, including an idle one which is not concerned by the wake up at all. This implies, in the worst case, an idle CPU will wake up to send an IPI to another idle cpu. This patch solves this by setting the irq affinity to the cpu concerned by the nearest timer event, by this way, the CPU which is wake up is guarantee to be the one concerned by the next event and we are safe with unnecessary wakeup for another idle CPU. As the irq affinity is not supported by all the archs, a flag is needed to specify which clocksource can handle it. Daniel Lezcano (3): time : pass broadcast parameter time : set broadcast irq affinity ARM: nomadik: add dynamic irq flag to the timer Viresh Kumar (1): ARM: timer-sp: Set dynamic irq affinity arch/arm/common/timer-sp.c | 3 ++- drivers/clocksource/nomadik-mtu.c | 3 ++- include/linux/clockchips.h | 1 + kernel/time/tick-broadcast.c | 40 +++++++++++++++++++++++++++++-------- 4 files changed, 37 insertions(+), 10 deletions(-) -- 1.7.9.5

12 years, 2 months

8
22
0 0

[PATCH v13 0/2] Add display-timing node parsing to exynos drm fimd

by Vikas Sajjan

Add display-timing node parsing to drm fimd and depends on the display helper patchset at http://lists.freedesktop.org/archives/dri-devel/2013-January/033998.html changes since v12: - Added dependency of "OF" for exynos drm fimd as suggested by Inki Dae <inki.dae(a)samsung.com> changes since v11: - Oops, there was a build error, fixed that. changes since v10: - abandoned the pinctrl patch, as commented by Linus Walleij <linus.walleij(a)linaro.org> - added new patch to enable the OF_VIDEOMODE and FB_MODE_HELPERS for EXYNOS DRM FIMD. changes since v9: - replaced IS_ERR_OR_NULL() with IS_ERR(), since IS_ERR_OR_NULL() will be depreciated, as discussed at http://lists.infradead.org/pipermail/linux-arm-kernel/2013-January/140543.h… http://www.mail-archive.com/linux-omap@vger.kernel.org/msg78030.html changes since v8: - replaced IS_ERR() with IS_ERR_OR_NULL(), because devm_pinctrl_get_select_default can return NULL, If CONFIG_PINCTRL is disabled. - modified the error log, such that it shall NOT cross 80 column. - added Acked-by. changes since v7: - addressed comments from Joonyoung Shim <jy0922.shim(a)samsung.com> to remove a unnecessary variable. changes since v6: addressed comments from Inki Dae <inki.dae(a)samsung.com> to separated out the pinctrl functionality and made a separate patch. changes since v5: - addressed comments from Inki Dae <inki.dae(a)samsung.com>, to remove the allocation of 'fbmode' and replaced '-1'in "of_get_fb_videomode(dev->of_node, fbmode, -1)" with OF_USE_NATIVE_MODE. changes since v4: - addressed comments from Paul Menzel <paulepanter(a)users.sourceforge.net>, to modify the commit message changes since v3: - addressed comments from Sean Paul <seanpaul(a)chromium.org>, to modify the return values and print messages. changes since v2: - moved 'devm_pinctrl_get_select_default' function call under 'if (pdev->dev.of_node)', this makes NON-DT code unchanged. (reported by: Rahul Sharma <r.sh.open(a)gmail.com>) changes since v1: - addressed comments from Sean Paul <seanpaul(a)chromium.org> Vikas Sajjan (2): video: drm: exynos: Add display-timing node parsing using video helper function drm/exynos: enable OF_VIDEOMODE and FB_MODE_HELPERS for exynos drm fimd drivers/gpu/drm/exynos/Kconfig | 4 +++- drivers/gpu/drm/exynos/exynos_drm_fimd.c | 24 ++++++++++++++++++++---- 2 files changed, 23 insertions(+), 5 deletions(-) -- 1.7.9.5

12 years, 3 months

1
2
0 0

ARM architected timers question.

by dpc＠ucore.info

Hi, I'm developing a baremetal hypervisor for Cortex-A15 and I'm using Arndale Board as a reference platform. ATM, I'm working on running Linaro 13.02 as a guest. I've noticed that `arch/arm/kernel/arch_timer.c` Linux prefers the virtual IRQ timers (CNTV_): static bool arch_timer_use_virtual = false; as a timer source and I'm wondering if there's a reason for this. Is it even possible for virtual timer to be available, while physical one is not? I don't think so. So why would Linux want to default to virtual one instead of physical one, especially in the situation when it was started in SVC mode and HYP is not accessible to it? >From Hypervisor point of view, it seems natural that guest does not care about virtualization stuff, does not try to be smarter than necessary and simply uses physical timer (PL1) all the time. It's hypervisor's job to trap guest accesses to physical timer (CNTP_) and route them to virtual timer (CNTV_) so virtual offset (CNTVOFF) value can be easily switched (again, by hypervisor, not the guest) when handling multiple guests, to correctly handle virtual time for them. In opposite direction the hardware virtual timer irq is just renumbered to appear to guest as a physical one. While emulating the CNTVOFF and CNTV_ is entirely possible too and required for completeness, it seems like additional work for no apparent reason. So I wanted to clarify this. Maybe there is something that I don't understand or I misread the intentions described in reference manual. -- Dawid Ciężarkiewicz B-Labs. http://b-labs.com/

12 years, 3 months

3
3
0 0

[PATCH v12 0/2] Add display-timing node parsing to exynos drm fimd

by Vikas Sajjan

Add display-timing node parsing to drm fimd and depends on the display helper patchset at http://lists.freedesktop.org/archives/dri-devel/2013-January/033998.html changes since v11: - Oops, there was a build error, fixed that. changes since v10: - abandoned the pinctrl patch, as commented by Linus Walleij <linus.walleij(a)linaro.org> - added new patch to enable the OF_VIDEOMODE and FB_MODE_HELPERS for EXYNOS DRM FIMD. changes since v9: - replaced IS_ERR_OR_NULL() with IS_ERR(), since IS_ERR_OR_NULL() will be depreciated, as discussed at http://lists.infradead.org/pipermail/linux-arm-kernel/2013-January/140543.h… http://www.mail-archive.com/linux-omap@vger.kernel.org/msg78030.html changes since v8: - replaced IS_ERR() with IS_ERR_OR_NULL(), because devm_pinctrl_get_select_default can return NULL, If CONFIG_PINCTRL is disabled. - modified the error log, such that it shall NOT cross 80 column. - added Acked-by. changes since v7: - addressed comments from Joonyoung Shim <jy0922.shim(a)samsung.com> to remove a unnecessary variable. changes since v6: addressed comments from Inki Dae <inki.dae(a)samsung.com> to separated out the pinctrl functionality and made a separate patch. changes since v5: - addressed comments from Inki Dae <inki.dae(a)samsung.com>, to remove the allocation of 'fbmode' and replaced '-1'in "of_get_fb_videomode(dev->of_node, fbmode, -1)" with OF_USE_NATIVE_MODE. changes since v4: - addressed comments from Paul Menzel <paulepanter(a)users.sourceforge.net>, to modify the commit message changes since v3: - addressed comments from Sean Paul <seanpaul(a)chromium.org>, to modify the return values and print messages. changes since v2: - moved 'devm_pinctrl_get_select_default' function call under 'if (pdev->dev.of_node)', this makes NON-DT code unchanged. (reported by: Rahul Sharma <r.sh.open(a)gmail.com>) changes since v1: - addressed comments from Sean Paul <seanpaul(a)chromium.org> Vikas Sajjan (2): video: drm: exynos: Add display-timing node parsing using video helper function drm/exynos: enable OF_VIDEOMODE and FB_MODE_HELPERS for exynos drm fimd drivers/gpu/drm/exynos/Kconfig | 2 ++ drivers/gpu/drm/exynos/exynos_drm_fimd.c | 27 +++++++++++++++++++++++---- 2 files changed, 25 insertions(+), 4 deletions(-) -- 1.7.9.5

12 years, 3 months

2
4
0 0

[PATCH] cpufreq: ARM big LITTLE: Add generic cpufreq driver and its DT glue

by Viresh Kumar

big LITTLE is ARM's new Architecture focussing power/performance needs of modern world. More information about big LITTLE can be found here: http://www.arm.com/products/processors/technologies/biglittleprocessing.php http://lwn.net/Articles/481055/ In order to keep cpufreq support for all big LITTLE platforms simple/generic, this patch tries to add a generic cpufreq driver layer for all big LITTLE platforms. The driver is divided into two parts: - Core driver: Generic and shared across all big LITTLE SoC's - Glue drivers: Per platform drivers providing ops to the core driver This patch adds in a generic glue driver which would extract information from Device Tree. Future SoC's can either reuse the DT glue or write their own depending on the need. Signed-off-by: Sudeep KarkadaNagesha <sudeep.karkadanagesha(a)arm.com> Signed-off-by: Viresh Kumar <viresh.kumar(a)linaro.org> --- This is pushed here: http://git.linaro.org/gitweb?p=people/vireshk/linux.git;a=shortlog;h=refs/h… .../bindings/cpufreq/arm_big_little_dt.txt | 29 +++ MAINTAINERS | 11 + drivers/cpufreq/Kconfig.arm | 13 + drivers/cpufreq/Makefile | 4 + drivers/cpufreq/arm_big_little.c | 278 +++++++++++++++++++++ drivers/cpufreq/arm_big_little.h | 40 +++ drivers/cpufreq/arm_big_little_dt.c | 125 +++++++++ 7 files changed, 500 insertions(+) create mode 100644 Documentation/devicetree/bindings/cpufreq/arm_big_little_dt.txt create mode 100644 drivers/cpufreq/arm_big_little.c create mode 100644 drivers/cpufreq/arm_big_little.h create mode 100644 drivers/cpufreq/arm_big_little_dt.c diff --git a/Documentation/devicetree/bindings/cpufreq/arm_big_little_dt.txt b/Documentation/devicetree/bindings/cpufreq/arm_big_little_dt.txt new file mode 100644 index 0000000..6f534eb --- /dev/null +++ b/Documentation/devicetree/bindings/cpufreq/arm_big_little_dt.txt @@ -0,0 +1,29 @@ +Generic ARM big LITTLE cpufreq driver's DT glue +----------------------------------------------- + +It is DT specific glue layer for generic cpufreq driver for big LITTLE systems. + +Both required and optional properties listed below must be defined under node +cluster*. * can be 0 or 1. + +Required properties: +- freqs: List of all supported frequencies. + +Optional properties: +- clock-latency: Specify the possible maximum transition latency for clock, in + unit of nanoseconds. + +Examples: + +cluster0: cluster@0 { + .. + + freqs = <500000000 600000000 700000000 800000000 900000000 1000000000 1100000000 1200000000>; + clock-latency = <200000>; + + .. + + cores { + .. + }; +}; diff --git a/MAINTAINERS b/MAINTAINERS index 554fd30..b14b749 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2206,6 +2206,17 @@ S: Maintained F: drivers/cpufreq/ F: include/linux/cpufreq.h +CPU FREQUENCY DRIVERS - ARM BIG LITTLE +M: Viresh Kumar <viresh.kumar(a)linaro.org> +M: Sudeep KarkadaNagesha <sudeep.karkadanagesha(a)arm.com> +L: cpufreq(a)vger.kernel.org +L: linux-pm(a)vger.kernel.org +W: http://www.arm.com/products/processors/technologies/biglittleprocessing.php +S: Maintained +F: drivers/cpufreq/arm_big_little.h +F: drivers/cpufreq/arm_big_little.c +F: drivers/cpufreq/arm_big_little_dt.c + CPUID/MSR DRIVER M: "H. Peter Anvin" <hpa(a)zytor.com> S: Maintained diff --git a/drivers/cpufreq/Kconfig.arm b/drivers/cpufreq/Kconfig.arm index 030ddf6..fdf54a9 100644 --- a/drivers/cpufreq/Kconfig.arm +++ b/drivers/cpufreq/Kconfig.arm @@ -2,6 +2,19 @@ # ARM CPU Frequency scaling drivers # +config ARM_BIG_LITTLE_CPUFREQ + tristate + depends on ARM_CPU_TOPOLOGY + +config ARM_DT_BL_CPUFREQ + tristate "Generic ARM big LITTLE CPUfreq driver probed via DT" + select ARM_BIG_LITTLE_CPUFREQ + depends on OF + default n + help + This enables the Generic CPUfreq driver for ARM big.LITTLE platform. + This gets frequency tables from DT. + config ARM_OMAP2PLUS_CPUFREQ bool "TI OMAP2+" depends on ARCH_OMAP2PLUS diff --git a/drivers/cpufreq/Makefile b/drivers/cpufreq/Makefile index 863fd18..d1b0832 100644 --- a/drivers/cpufreq/Makefile +++ b/drivers/cpufreq/Makefile @@ -44,6 +44,10 @@ obj-$(CONFIG_X86_INTEL_PSTATE) += intel_pstate.o ################################################################################## # ARM SoC drivers +obj-$(CONFIG_ARM_BIG_LITTLE_CPUFREQ) += arm_big_little.o +# big LITTLE per platform glues. Keep DT_BL_CPUFREQ as the last entry in all big +# LITTLE drivers, so that it is probed last. +obj-$(CONFIG_ARM_DT_BL_CPUFREQ) += arm_big_little_dt.o obj-$(CONFIG_UX500_SOC_DB8500) += dbx500-cpufreq.o obj-$(CONFIG_ARM_S3C2416_CPUFREQ) += s3c2416-cpufreq.o obj-$(CONFIG_ARM_S3C64XX_CPUFREQ) += s3c64xx-cpufreq.o diff --git a/drivers/cpufreq/arm_big_little.c b/drivers/cpufreq/arm_big_little.c new file mode 100644 index 0000000..0d6de0e --- /dev/null +++ b/drivers/cpufreq/arm_big_little.c @@ -0,0 +1,278 @@ +/* + * ARM big.LITTLE Platforms CPUFreq support + * + * Copyright (C) 2013 ARM Ltd. + * Sudeep KarkadaNagesha <sudeep.karkadanagesha(a)arm.com> + * + * Copyright (C) 2013 Linaro. + * Viresh Kumar <viresh.kumar(a)linaro.org> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * This program is distributed "as is" WITHOUT ANY WARRANTY of any + * kind, whether express or implied; without even the implied warranty + * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/clk.h> +#include <linux/cpufreq.h> +#include <linux/cpumask.h> +#include <linux/export.h> +#include <linux/of_platform.h> +#include <linux/slab.h> +#include <linux/topology.h> +#include <linux/types.h> + +#include "arm_big_little.h" + +/* Currently we support only two clusters */ +#define MAX_CLUSTERS 2 + +static struct cpufreq_arm_bL_ops *arm_bL_ops; +static struct clk *clk[MAX_CLUSTERS]; +static struct cpufreq_frequency_table *freq_table[MAX_CLUSTERS]; +static atomic_t cluster_usage[MAX_CLUSTERS] = {ATOMIC_INIT(0), ATOMIC_INIT(0)}; + +static int cpu_to_cluster(int cpu) +{ + return topology_physical_package_id(cpu); +} + +static unsigned int bL_cpufreq_get(unsigned int cpu) +{ + u32 cur_cluster = cpu_to_cluster(cpu); + + return clk_get_rate(clk[cur_cluster]) / 1000; +} + +/* Validate policy frequency range */ +static int bL_cpufreq_verify_policy(struct cpufreq_policy *policy) +{ + u32 cur_cluster = cpu_to_cluster(policy->cpu); + + return cpufreq_frequency_table_verify(policy, freq_table[cur_cluster]); +} + +/* Set clock frequency */ +static int bL_cpufreq_set_target(struct cpufreq_policy *policy, + unsigned int target_freq, unsigned int relation) +{ + struct cpufreq_freqs freqs; + u32 cpu = policy->cpu, freq_tab_idx, cur_cluster; + int ret = 0; + + cur_cluster = cpu_to_cluster(policy->cpu); + + freqs.old = bL_cpufreq_get(policy->cpu); + + /* Determine valid target frequency using freq_table */ + cpufreq_frequency_table_target(policy, freq_table[cur_cluster], + target_freq, relation, &freq_tab_idx); + freqs.new = freq_table[cur_cluster][freq_tab_idx].frequency; + + freqs.cpu = policy->cpu; + + pr_debug("%s: cpu: %d, cluster: %d, oldfreq: %d, target freq: %d, new freq: %d\n", + __func__, cpu, cur_cluster, freqs.old, target_freq, + freqs.new); + + if (freqs.old == freqs.new) + return 0; + + for_each_cpu(freqs.cpu, policy->cpus) + cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); + + ret = clk_set_rate(clk[cur_cluster], freqs.new * 1000); + if (ret) { + pr_err("clk_set_rate failed: %d\n", ret); + return ret; + } + + policy->cur = freqs.new; + + for_each_cpu(freqs.cpu, policy->cpus) + cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); + + return ret; +} + +/* translate the integer array into cpufreq_frequency_table entries */ +struct cpufreq_frequency_table * +arm_bL_copy_table_from_array(unsigned int *table, int count) +{ + int i; + + struct cpufreq_frequency_table *freq_table; + + pr_debug("%s: table: %p, count: %d\n", __func__, table, count); + + freq_table = kmalloc(sizeof(*freq_table) * (count + 1), GFP_KERNEL); + if (!freq_table) + return NULL; + + for (i = 0; i < count; i++) { + pr_debug("%s: index: %d, freq: %d\n", __func__, i, table[i]); + freq_table[i].index = i; + freq_table[i].frequency = table[i]; /* in kHZ */ + } + + freq_table[i].index = count; + freq_table[i].frequency = CPUFREQ_TABLE_END; + + return freq_table; +} +EXPORT_SYMBOL_GPL(arm_bL_copy_table_from_array); + +void arm_bL_free_freq_table(u32 cluster) +{ + pr_debug("%s: free freq table\n", __func__); + + kfree(freq_table[cluster]); +} +EXPORT_SYMBOL_GPL(arm_bL_free_freq_table); + +static void put_cluster_clk_and_freq_table(u32 cluster) +{ + if (!atomic_dec_return(&cluster_usage[cluster])) { + clk_put(clk[cluster]); + clk[cluster] = NULL; + arm_bL_ops->put_freq_tbl(cluster); + freq_table[cluster] = NULL; + pr_debug("%s: cluster: %d\n", __func__, cluster); + } +} + +static int get_cluster_clk_and_freq_table(u32 cluster) +{ + char name[9] = "cluster"; + int count; + + if (atomic_inc_return(&cluster_usage[cluster]) != 1) + return 0; + + freq_table[cluster] = arm_bL_ops->get_freq_tbl(cluster, &count); + if (!freq_table[cluster]) + goto atomic_dec; + + name[7] = cluster + '0'; + clk[cluster] = clk_get(NULL, name); + if (!IS_ERR_OR_NULL(clk[cluster])) { + pr_debug("%s: clk: %p & freq table: %p, cluster: %d\n", + __func__, clk[cluster], freq_table[cluster], + cluster); + return 0; + } + + arm_bL_ops->put_freq_tbl(cluster); + +atomic_dec: + atomic_dec(&cluster_usage[cluster]); + pr_err("%s: Failed to get data for cluster: %d\n", __func__, cluster); + return -ENODATA; +} + +/* Per-CPU initialization */ +static int bL_cpufreq_init(struct cpufreq_policy *policy) +{ + u32 cur_cluster = cpu_to_cluster(policy->cpu); + int result; + + result = get_cluster_clk_and_freq_table(cur_cluster); + if (result) + return result; + + result = cpufreq_frequency_table_cpuinfo(policy, + freq_table[cur_cluster]); + if (result) { + pr_err("CPU %d, cluster: %d invalid freq table\n", policy->cpu, + cur_cluster); + put_cluster_clk_and_freq_table(cur_cluster); + return result; + } + + cpufreq_frequency_table_get_attr(freq_table[cur_cluster], policy->cpu); + + policy->cpuinfo.transition_latency = + arm_bL_ops->get_transition_latency(cur_cluster); + policy->cur = bL_cpufreq_get(policy->cpu); + + cpumask_copy(policy->cpus, topology_core_cpumask(policy->cpu)); + + pr_info("CPU %d initialized\n", policy->cpu); + return 0; +} + +static int bL_cpufreq_exit(struct cpufreq_policy *policy) +{ + put_cluster_clk_and_freq_table(cpu_to_cluster(policy->cpu)); + pr_debug("%s: Exited, cpu: %d\n", __func__, policy->cpu); + + return 0; +} + +/* Export freq_table to sysfs */ +static struct freq_attr *bL_cpufreq_attr[] = { + &cpufreq_freq_attr_scaling_available_freqs, + NULL, +}; + +static struct cpufreq_driver bL_cpufreq_driver = { + .name = "arm-big-little", + .flags = CPUFREQ_STICKY, + .verify = bL_cpufreq_verify_policy, + .target = bL_cpufreq_set_target, + .get = bL_cpufreq_get, + .init = bL_cpufreq_init, + .exit = bL_cpufreq_exit, + .attr = bL_cpufreq_attr, +}; + +int bL_cpufreq_register(struct cpufreq_arm_bL_ops *ops) +{ + int ret; + + if (arm_bL_ops) { + pr_debug("%s: Already registered: %s, exiting\n", __func__, + arm_bL_ops->name); + return -EBUSY; + } + + if (!ops || !strlen(ops->name) || !ops->get_freq_tbl) { + pr_err("%s: Invalid arm_bL_ops, exiting\n", __func__); + return -ENODEV; + } + + arm_bL_ops = ops; + + ret = cpufreq_register_driver(&bL_cpufreq_driver); + if (ret) { + pr_info("%s: Failed registering platform driver: %s, err: %d\n", + __func__, ops->name, ret); + arm_bL_ops = NULL; + } else { + pr_info("%s: Registered platform driver: %s\n", __func__, + ops->name); + } + + return ret; +} +EXPORT_SYMBOL_GPL(bL_cpufreq_register); + +void bL_cpufreq_unregister(struct cpufreq_arm_bL_ops *ops) +{ + if (arm_bL_ops != ops) { + pr_info("%s: Registered with: %s, can't unregister, exiting\n", + __func__, arm_bL_ops->name); + } + + cpufreq_unregister_driver(&bL_cpufreq_driver); + pr_info("%s: Un-registered platform driver: %s\n", __func__, + arm_bL_ops->name); + arm_bL_ops = NULL; +} +EXPORT_SYMBOL_GPL(bL_cpufreq_unregister); diff --git a/drivers/cpufreq/arm_big_little.h b/drivers/cpufreq/arm_big_little.h new file mode 100644 index 0000000..049e15d --- /dev/null +++ b/drivers/cpufreq/arm_big_little.h @@ -0,0 +1,40 @@ +/* + * ARM big.LITTLE platform's CPUFreq header file + * + * Copyright (C) 2013 ARM Ltd. + * Sudeep KarkadaNagesha <sudeep.karkadanagesha(a)arm.com> + * + * Copyright (C) 2013 Linaro. + * Viresh Kumar <viresh.kumar(a)linaro.org> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * This program is distributed "as is" WITHOUT ANY WARRANTY of any + * kind, whether express or implied; without even the implied warranty + * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ +#ifndef CPUFREQ_ARM_BIG_LITTLE_H +#define CPUFREQ_ARM_BIG_LITTLE_H + +#include <linux/cpufreq.h> +#include <linux/types.h> + +struct cpufreq_arm_bL_ops { + char name[CPUFREQ_NAME_LEN]; + u32 (*get_transition_latency)(u32 cluster); + struct cpufreq_frequency_table *(*get_freq_tbl)(u32 cluster, + int *count); + void (*put_freq_tbl)(u32 cluster); +}; + +struct cpufreq_frequency_table * +arm_bL_copy_table_from_array(unsigned int *table, int count); +void arm_bL_free_freq_table(u32 cluster); + +int bL_cpufreq_register(struct cpufreq_arm_bL_ops *ops); +void bL_cpufreq_unregister(struct cpufreq_arm_bL_ops *ops); + +#endif /* CPUFREQ_ARM_BIG_LITTLE_H */ diff --git a/drivers/cpufreq/arm_big_little_dt.c b/drivers/cpufreq/arm_big_little_dt.c new file mode 100644 index 0000000..c1eee50 --- /dev/null +++ b/drivers/cpufreq/arm_big_little_dt.c @@ -0,0 +1,125 @@ +/* + * Generic big.LITTLE CPUFreq Interface driver + * + * It provides necessary ops to arm_big_little cpufreq driver and gets + * Frequency information from Device Tree. Freq table in DT must be in KHz. + * + * Copyright (C) 2013 Linaro. + * Viresh Kumar <viresh.kumar(a)linaro.org> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * This program is distributed "as is" WITHOUT ANY WARRANTY of any + * kind, whether express or implied; without even the implied warranty + * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/cpufreq.h> +#include <linux/export.h> +#include <linux/module.h> +#include <linux/of.h> +#include <linux/slab.h> +#include <linux/types.h> +#include "arm_big_little.h" + +static struct cpufreq_frequency_table *generic_get_freq_tbl(u32 cluster, + int *count) +{ + struct device_node *np = NULL; + const struct property *pp; + unsigned int *table = NULL; + int cluster_id; + struct cpufreq_frequency_table *cpufreq_table; + + while ((np = of_find_node_by_name(np, "cluster"))) { + if (of_property_read_u32(np, "reg", &cluster_id)) + continue; + + if (cluster_id != cluster) + continue; + + pp = of_find_property(np, "freqs", NULL); + if (!pp) + continue; + + *count = pp->length / sizeof(u32); + if (!*count) + continue; + + table = kmalloc(sizeof(*table) * (*count), GFP_KERNEL); + if (!table) { + pr_err("%s: Failed to allocate memory for table\n", + __func__); + return NULL; + } + + of_property_read_u32_array(np, "freqs", table, *count); + break; + } + + if (!table) { + pr_err("%s: Unable to retrieve Freq table from Device Tree", + __func__); + return NULL; + } + + cpufreq_table = arm_bL_copy_table_from_array(table, *count); + kfree(table); + + return cpufreq_table; +} + +static void generic_put_freq_tbl(u32 cluster) +{ + arm_bL_free_freq_table(cluster); +} + +static u32 get_transition_latency(u32 cluster) +{ + struct device_node *np = NULL; + u32 transition_latency = CPUFREQ_ETERNAL; + int cluster_id; + + while ((np = of_find_node_by_name(np, "cluster"))) { + if (of_property_read_u32(np, "reg", &cluster_id)) + continue; + + if (cluster_id != cluster) + continue; + + of_property_read_u32(np, "clock-latency", &transition_latency); + return transition_latency; + } + + pr_err("%s: Unable to get clock-latency from DT, use CPUFREQ_ETERNAL", + __func__); + + return transition_latency; +} + +static struct cpufreq_arm_bL_ops generic_bL_ops = { + .name = "generic-bl", + .get_transition_latency = get_transition_latency, + .get_freq_tbl = generic_get_freq_tbl, + .put_freq_tbl = generic_put_freq_tbl, +}; + +static int generic_bL_init(void) +{ + return bL_cpufreq_register(&generic_bL_ops); +} +module_init(generic_bL_init); + +static void generic_bL_exit(void) +{ + return bL_cpufreq_unregister(&generic_bL_ops); +} +module_exit(generic_bL_exit); + +MODULE_DESCRIPTION("Generic ARM big LITTLE cpufreq driver"); +MODULE_LICENSE("GPL"); -- 1.7.12.rc2.18.g61b472e

12 years, 3 months

4
14
0 0

64k page support for ARM v7

by David Betz

Has anyone worked on adding 64k page support to the ARM v7 architecture, in particular the ARM Cortex A9? I know that there is HugeTLB support available but I'm wondering if anyone has tried making the native page size 64k rather than 4k. In other words, make PAGE_SHIFT 16 instead of 12 and all of the corresponding changes to the way page tables are handled. Has this been attempted already? Can someone point me to a patch? Thanks! David Betz

12 years, 3 months

1
0
0 0

[PATCH v11 0/2] Add display-timing node parsing to exynos drm fimd

by Vikas Sajjan

Add display-timing node parsing to drm fimd and depends on the display helper patchset at http://lists.freedesktop.org/archives/dri-devel/2013-January/033998.html changes since v10: - abandoned the pinctrl patch, as commented by Linus Walleij <linus.walleij(a)linaro.org> - added new patch to enable the OF_VIDEOMODE and FB_MODE_HELPERS for EXYNOS DRM FIMD. changes since v9: - replaced IS_ERR_OR_NULL() with IS_ERR(), since IS_ERR_OR_NULL() will be depreciated, as discussed at http://lists.infradead.org/pipermail/linux-arm-kernel/2013-January/140543.h… http://www.mail-archive.com/linux-omap@vger.kernel.org/msg78030.html changes since v8: - replaced IS_ERR() with IS_ERR_OR_NULL(), because devm_pinctrl_get_select_default can return NULL, If CONFIG_PINCTRL is disabled. - modified the error log, such that it shall NOT cross 80 column. - added Acked-by. changes since v7: - addressed comments from Joonyoung Shim <jy0922.shim(a)samsung.com> to remove a unnecessary variable. changes since v6: addressed comments from Inki Dae <inki.dae(a)samsung.com> to separated out the pinctrl functionality and made a separate patch. changes since v5: - addressed comments from Inki Dae <inki.dae(a)samsung.com>, to remove the allocation of 'fbmode' and replaced '-1'in "of_get_fb_videomode(dev->of_node, fbmode, -1)" with OF_USE_NATIVE_MODE. changes since v4: - addressed comments from Paul Menzel <paulepanter(a)users.sourceforge.net>, to modify the commit message changes since v3: - addressed comments from Sean Paul <seanpaul(a)chromium.org>, to modify the return values and print messages. changes since v2: - moved 'devm_pinctrl_get_select_default' function call under 'if (pdev->dev.of_node)', this makes NON-DT code unchanged. (reported by: Rahul Sharma <r.sh.open(a)gmail.com>) changes since v1: - addressed comments from Sean Paul <seanpaul(a)chromium.org> Vikas Sajjan (2): video: drm: exynos: Add display-timing node parsing using video helper function drm/exynos: enable OF_VIDEOMODE and FB_MODE_HELPERS for exynos drm fimd drivers/gpu/drm/exynos/Kconfig | 2 ++ drivers/gpu/drm/exynos/exynos_drm_fimd.c | 27 +++++++++++++++++++++++---- 2 files changed, 25 insertions(+), 4 deletions(-) -- 1.7.9.5

12 years, 3 months

1
2
0 0

[PATCH V2 0/2] cpufreq: governors: Avoid unnecessary per cpu timer interrupts

by Viresh Kumar

Following patch has introduced per cpu timers or works for ondemand and conservative governors. commit 2abfa876f1117b0ab45f191fb1f82c41b1cbc8fe Author: Rickard Andersson <rickard.andersson(a)stericsson.com> Date: Thu Dec 27 14:55:38 2012 +0000 cpufreq: handle SW coordinated CPUs This causes additional unnecessary interrupts on all cpus when the load is recently evaluated by any other cpu. i.e. When load is recently evaluated by cpu x, we don't really need any other cpu to evaluate this load again for the next sampling_rate time. Some sort of code is present to avoid that but we are still getting timer interrupts for all cpus. A good way of avoiding this would be to modify delays for all cpus (policy->cpus) whenever any cpu has evaluated load. This patchset tries to fix this issue. These patches are applied here: http://git.linaro.org/gitweb?p=people/vireshk/linux.git;a=shortlog;h=refs/h… V1->V2: - Dropped Workqueue modifications and use system_wq directly. Viresh Kumar (2): cpufreq: ondemand: Don't update sample_type if we don't evaluate load again cpufreq: governors: Avoid unnecessary per cpu timer interrupts drivers/cpufreq/cpufreq_conservative.c | 8 ++++--- drivers/cpufreq/cpufreq_governor.c | 39 ++++++++++++++++++++++++---------- drivers/cpufreq/cpufreq_governor.h | 2 ++ drivers/cpufreq/cpufreq_ondemand.c | 34 ++++++++++++++--------------- 4 files changed, 52 insertions(+), 31 deletions(-) -- 1.7.12.rc2.18.g61b472e

12 years, 3 months

1
2
0 0

[ACTIVITY] 2013-2-23 - 2012-03-3-1

by David Long

=== David Long === === Highlights === * I'm unfortunately still trying to sort out what chat in basic uprobes support is causing the upleveled ARM uprobe patch to behave incorrectly. I still think it's a small issue, it's just a matter of tracking it down. * Spent a little time getting ready for Connect. === Plans === * Connect-specific activities will occupy most of my time in the coming week. * Find the problem in the upleved patch and move forward. === Issues === -dl

12 years, 3 months

1
0
0 0

[ACTIVITY] (John Stultz) Feb 25 - Mar 1

by John Stultz

=== Highlights === * Got my slides ready for connect * Acked Serban's reduced scope ashmem compat_ioctl patch * Saw the experimental/android-3.8 branch showed up and send Arve a build fix patch * Reviewed the android-3.8 branch and found some staging fixes that should probably go upstream. After consulting the Google devs, sent them on to lkml. Little bit of contention on the list, so I may have to drop one and resend the rest. * Sorted out ABS travel expenses * Pinged Jason Wessel on Anton's KDB patches * Updated blueprints and held bi-weekly Android Upstreaming hangout meeting. * Sent out RFD on submitting the sync driver to staging. Got no objections, so submitted the patches. Greg seemed ok with most of it, likely to be merged for 3.10 * Got invited to, but declined a conference on the future of UTC. * Spent some time testing valid-interval lock concept for timekeeping * Did packing and prep for connect. === Plans === * Linaro Connect mostly. * Still need to review Serban's binder patches * Still need Look into Androids support of large-files with 32bit applications * Need to look into some community bugs I've been ignoring. === Issues === * NA

12 years, 3 months

1
0
0 0

[ACTIVITY] (Ulf Hansson) 2013-02-11 - 2013-03-01

by Ulf Hansson

== Ulf Hansson == === Highlights === Storage: * Reviewing patches on mmc-list. * Patches for fixing signal voltage switch procedure for SD card UHS mode ready and merged by Chris. * Rework parts of the HS200 and SDR104 support in the mmc protocol layer. First part for tuning sequence done, patch will be pushed to mmc-list shortly. * Sent patch to enable runtime pm support for mmc/sd block device. The intention is let Idle time BKOPS build upon this. Clk: * Started to prepare a patchset for upstreaming patches that will add support for abx500 clocks, update different driver's clk support and include ux500 clk optimizations. * Reviewing DVFS clock related patches from Mike Turquette. Interesting work here. === Plans === Storage: * Doing an overall analyse about the eMMC 4.5/4.6 features. Check what can be considered finished, what needs further fixing and point out the new features for which we should spend our focus on in Linaro storage team. * Keep focus on the mmc power management blueprint so we can finalize this work as soon as possible. * Push patches for mmci host driver to support UHS cards. * Push patches for mmci host driver to further extend the power management support. * Push patches for mmci host driver to add new features like CMD23 support and more. * Push patches for mmci host driver to add support for new STE 8540 variant. Clk: * Upstreaming of internal work for ux500. * Follow up on patchset for fixing clk_set_parent API. * Follow up on patchset for disable unsed prepared clks. === Issues === * None. Kind regards Ulf Hansson

12 years, 3 months

1
0
0 0

[PATCH v10 0/2] Add display-timing node parsing to exynos drm fimd

by Vikas Sajjan

Add display-timing node parsing to drm fimd and depends on the display helper patchset at http://lists.freedesktop.org/archives/dri-devel/2013-January/033998.html It also adds pinctrl support for drm fimd. changes since v9: - replaced IS_ERR_OR_NULL() with IS_ERR(), since IS_ERR_OR_NULL() will be depreciated, as discussed at http://lists.infradead.org/pipermail/linux-arm-kernel/2013-January/140543.h… http://www.mail-archive.com/linux-omap@vger.kernel.org/msg78030.html changes since v8: - replaced IS_ERR() with IS_ERR_OR_NULL(), because devm_pinctrl_get_select_default can return NULL, If CONFIG_PINCTRL is disabled. - modified the error log, such that it shall NOT cross 80 column. - added Acked-by. changes since v7: - addressed comments from Joonyoung Shim <jy0922.shim(a)samsung.com> to remove a unnecessary variable. changes since v6: addressed comments from Inki Dae <inki.dae(a)samsung.com> to separated out the pinctrl functionality and made a separate patch. changes since v5: - addressed comments from Inki Dae <inki.dae(a)samsung.com>, to remove the allocation of 'fbmode' and replaced '-1'in "of_get_fb_videomode(dev->of_node, fbmode, -1)" with OF_USE_NATIVE_MODE. changes since v4: - addressed comments from Paul Menzel <paulepanter(a)users.sourceforge.net>, to modify the commit message changes since v3: - addressed comments from Sean Paul <seanpaul(a)chromium.org>, to modify the return values and print messages. changes since v2: - moved 'devm_pinctrl_get_select_default' function call under 'if (pdev->dev.of_node)', this makes NON-DT code unchanged. (reported by: Rahul Sharma <r.sh.open(a)gmail.com>) changes since v1: - addressed comments from Sean Paul <seanpaul(a)chromium.org> Vikas Sajjan (2): video: drm: exynos: Add display-timing node parsing using video helper function video: drm: exynos: Add pinctrl support to fimd drivers/gpu/drm/exynos/exynos_drm_fimd.c | 33 ++++++++++++++++++++++++++---- 1 file changed, 29 insertions(+), 4 deletions(-) -- 1.7.9.5

12 years, 3 months

1
2
0 0

Power Management WG schedule for LCA13

by Amit Kucheria

Hi, Please find a link[1] to some of the things we plan to discuss and work on in Hong Kong next week. If you're interested in some of these topics and are attending in person, please come and say hello. See you in Hong Kong! Regards, Amit ----------------------------------------------- PMWG Tech Lead, Linaro [1] https://wiki.linaro.org/WorkingGroups/PowerManagement/Doc/HK_LCA2013

12 years, 3 months

1
1
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

linaro-kernel