From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752616Ab3HZVZp (ORCPT ); Mon, 26 Aug 2013 17:25:45 -0400 Received: from a9-66.smtp-out.amazonses.com ([54.240.9.66]:38889 "EHLO a9-66.smtp-out.amazonses.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752214Ab3HZVZn (ORCPT ); Mon, 26 Aug 2013 17:25:43 -0400 X-Greylist: delayed 564 seconds by postgrey-1.27 at vger.kernel.org; Mon, 26 Aug 2013 17:25:42 EDT Message-ID: <00000140bc7bfb2f-8732df38-4a8e-44cd-9bf1-200f23fc0054-000000@email.amazonses.com> User-Agent: quilt/0.50-1 Date: Mon, 26 Aug 2013 21:16:16 +0000 From: Christoph Lameter To: Tejun Heo Cc: akpm@linuxfoundation.org, Mike Frysinger Cc: linux-arch@vger.kernel.org Cc: Steven Rostedt Cc: linux-kernel@vger.kernel.org Subject: [guv v2 24/31] blackfin: Replace __get_cpu_var uses References: <20130826204351.725357339@linux.com> X-SES-Outgoing: 2013.08.26-54.240.9.66 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org __get_cpu_var() is used for multiple purposes in the kernel source. One of them is address calculation via the form &__get_cpu_var(x). This calculates the address for the instance of the percpu variable of the current processor based on an offset. Others usage cases are for storing and retrieving data from the current processors percpu area. __get_cpu_var() can be used as an lvalue when writing data or on the right side of an assignment. __get_cpu_var() is defined as : #define __get_cpu_var(var) (*this_cpu_ptr(&(var))) __get_cpu_var() always only does a address determination. However, store and retrieve operations could use a segment prefix (or global register on other platforms) to avoid the address calculation. this_cpu_write() and this_cpu_read() can directly take an offset into a percpu area and use optimized assembly code to read and write per cpu variables. This patch converts __get_cpu_var into either and explicit address calculation using this_cpu_ptr() or into a use of this_cpu operations that use the offset. Thereby address calcualtions are avoided and less registers are used when code is generated. At the end of the patchset all uses of __get_cpu_var have been removed so the macro is removed too. The patchset includes passes over all arches as well. Once these operations are used throughout then specialized macros can be defined in non -x86 arches as well in order to optimize per cpu access by f.e. using a global register that may be set to the per cpu base. Transformations done to __get_cpu_var() 1. Determine the address of the percpu instance of the current processor. DEFINE_PER_CPU(int, y); int *x = &__get_cpu_var(y); Converts to int *x = this_cpu_ptr(&y); 2. Same as #1 but this time an array structure is involved. DEFINE_PER_CPU(int, y[20]); int *x = __get_cpu_var(y); Converts to int *x = this_cpu_ptr(y); 3. Retrieve the content of the current processors instance of a per cpu variable. DEFINE_PER_CPU(int, u); int x = __get_cpu_var(y) Converts to int x = __this_cpu_read(y); 4. Retrieve the content of a percpu struct DEFINE_PER_CPU(struct mystruct, y); struct mystruct x = __get_cpu_var(y); Converts to memcpy(this_cpu_ptr(&y), x, sizeof(x)); 5. Assignment to a per cpu variable DEFINE_PER_CPU(int, y) __get_cpu_var(y) = x; Converts to this_cpu_write(y, x); 6. Increment/Decrement etc of a per cpu variable DEFINE_PER_CPU(int, y); __get_cpu_var(y)++ Converts to this_cpu_inc(y) Index: linux/arch/blackfin/include/asm/ipipe.h =================================================================== --- linux.orig/arch/blackfin/include/asm/ipipe.h 2013-08-26 13:25:18.000000000 -0500 +++ linux/arch/blackfin/include/asm/ipipe.h 2013-08-26 13:26:23.578979325 -0500 @@ -157,7 +157,7 @@ static inline unsigned long __ipipe_ffnz } #define __ipipe_do_root_xirq(ipd, irq) \ - ((ipd)->irqs[irq].handler(irq, &__raw_get_cpu_var(__ipipe_tick_regs))) + ((ipd)->irqs[irq].handler(irq, __this_cpu_ptr(&__ipipe_tick_regs))) #define __ipipe_run_irqtail(irq) /* Must be a macro */ \ do { \ Index: linux/arch/blackfin/kernel/perf_event.c =================================================================== --- linux.orig/arch/blackfin/kernel/perf_event.c 2013-08-26 13:25:18.000000000 -0500 +++ linux/arch/blackfin/kernel/perf_event.c 2013-08-26 13:26:23.578979325 -0500 @@ -300,7 +300,7 @@ again: static void bfin_pmu_stop(struct perf_event *event, int flags) { - struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events); struct hw_perf_event *hwc = &event->hw; int idx = hwc->idx; @@ -318,7 +318,7 @@ static void bfin_pmu_stop(struct perf_ev static void bfin_pmu_start(struct perf_event *event, int flags) { - struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events); struct hw_perf_event *hwc = &event->hw; int idx = hwc->idx; @@ -335,7 +335,7 @@ static void bfin_pmu_start(struct perf_e static void bfin_pmu_del(struct perf_event *event, int flags) { - struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events); bfin_pmu_stop(event, PERF_EF_UPDATE); __clear_bit(event->hw.idx, cpuc->used_mask); @@ -345,7 +345,7 @@ static void bfin_pmu_del(struct perf_eve static int bfin_pmu_add(struct perf_event *event, int flags) { - struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events); struct hw_perf_event *hwc = &event->hw; int idx = hwc->idx; int ret = -EAGAIN; @@ -429,7 +429,7 @@ static int bfin_pmu_event_init(struct pe static void bfin_pmu_enable(struct pmu *pmu) { - struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events); struct perf_event *event; struct hw_perf_event *hwc; int i; Index: linux/arch/blackfin/mach-common/ints-priority.c =================================================================== --- linux.orig/arch/blackfin/mach-common/ints-priority.c 2013-08-26 13:25:18.000000000 -0500 +++ linux/arch/blackfin/mach-common/ints-priority.c 2013-08-26 13:26:23.578979325 -0500 @@ -1675,12 +1675,12 @@ asmlinkage int __ipipe_grab_irq(int vec, bfin_write_TIMER_STATUS(1); /* Latch TIMIL0 */ #endif /* This is basically what we need from the register frame. */ - __raw_get_cpu_var(__ipipe_tick_regs).ipend = regs->ipend; - __raw_get_cpu_var(__ipipe_tick_regs).pc = regs->pc; + __this_cpu_write(__ipipe_tick_regs.ipend, regs->ipend); + __this_cpu_write(__ipipe_tick_regs.pc, regs->pc); if (this_domain != ipipe_root_domain) - __raw_get_cpu_var(__ipipe_tick_regs).ipend &= ~0x10; + __this_cpu_and(__ipipe_tick_regs.ipend, ~0x10); else - __raw_get_cpu_var(__ipipe_tick_regs).ipend |= 0x10; + __this_cpu_or(__ipipe_tick_regs.ipend, 0x10); } /* Index: linux/arch/blackfin/mach-common/smp.c =================================================================== --- linux.orig/arch/blackfin/mach-common/smp.c 2013-08-26 13:25:18.000000000 -0500 +++ linux/arch/blackfin/mach-common/smp.c 2013-08-26 13:26:23.578979325 -0500 @@ -146,7 +146,7 @@ static irqreturn_t ipi_handler_int1(int platform_clear_ipi(cpu, IRQ_SUPPLE_1); - bfin_ipi_data = &__get_cpu_var(bfin_ipi); + bfin_ipi_data = this_cpu_ptr(&bfin_ipi); while ((pending = atomic_xchg(&bfin_ipi_data->bits, 0)) != 0) { msg = 0; do { From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Lameter Subject: [guv v2 24/31] blackfin: Replace __get_cpu_var uses Date: Mon, 26 Aug 2013 21:16:16 +0000 Message-ID: <00000140bc7bfb2f-8732df38-4a8e-44cd-9bf1-200f23fc0054-000000@email.amazonses.com> References: <20130826204351.725357339@linux.com> Return-path: Received: from a9-66.smtp-out.amazonses.com ([54.240.9.66]:38889 "EHLO a9-66.smtp-out.amazonses.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752214Ab3HZVZn (ORCPT ); Mon, 26 Aug 2013 17:25:43 -0400 Sender: linux-arch-owner@vger.kernel.org List-ID: To: Tejun Heo Cc: akpm@linuxfoundation.org, Mike Frysinger , linux-arch@vger.kernel.org, Steven Rostedt , linux-kernel@vger.kernel.org __get_cpu_var() is used for multiple purposes in the kernel source. One of them is address calculation via the form &__get_cpu_var(x). This calculates the address for the instance of the percpu variable of the current processor based on an offset. Others usage cases are for storing and retrieving data from the current processors percpu area. __get_cpu_var() can be used as an lvalue when writing data or on the right side of an assignment. __get_cpu_var() is defined as : #define __get_cpu_var(var) (*this_cpu_ptr(&(var))) __get_cpu_var() always only does a address determination. However, store and retrieve operations could use a segment prefix (or global register on other platforms) to avoid the address calculation. this_cpu_write() and this_cpu_read() can directly take an offset into a percpu area and use optimized assembly code to read and write per cpu variables. This patch converts __get_cpu_var into either and explicit address calculation using this_cpu_ptr() or into a use of this_cpu operations that use the offset. Thereby address calcualtions are avoided and less registers are used when code is generated. At the end of the patchset all uses of __get_cpu_var have been removed so the macro is removed too. The patchset includes passes over all arches as well. Once these operations are used throughout then specialized macros can be defined in non -x86 arches as well in order to optimize per cpu access by f.e. using a global register that may be set to the per cpu base. Transformations done to __get_cpu_var() 1. Determine the address of the percpu instance of the current processor. DEFINE_PER_CPU(int, y); int *x = &__get_cpu_var(y); Converts to int *x = this_cpu_ptr(&y); 2. Same as #1 but this time an array structure is involved. DEFINE_PER_CPU(int, y[20]); int *x = __get_cpu_var(y); Converts to int *x = this_cpu_ptr(y); 3. Retrieve the content of the current processors instance of a per cpu variable. DEFINE_PER_CPU(int, u); int x = __get_cpu_var(y) Converts to int x = __this_cpu_read(y); 4. Retrieve the content of a percpu struct DEFINE_PER_CPU(struct mystruct, y); struct mystruct x = __get_cpu_var(y); Converts to memcpy(this_cpu_ptr(&y), x, sizeof(x)); 5. Assignment to a per cpu variable DEFINE_PER_CPU(int, y) __get_cpu_var(y) = x; Converts to this_cpu_write(y, x); 6. Increment/Decrement etc of a per cpu variable DEFINE_PER_CPU(int, y); __get_cpu_var(y)++ Converts to this_cpu_inc(y) Index: linux/arch/blackfin/include/asm/ipipe.h =================================================================== --- linux.orig/arch/blackfin/include/asm/ipipe.h 2013-08-26 13:25:18.000000000 -0500 +++ linux/arch/blackfin/include/asm/ipipe.h 2013-08-26 13:26:23.578979325 -0500 @@ -157,7 +157,7 @@ static inline unsigned long __ipipe_ffnz } #define __ipipe_do_root_xirq(ipd, irq) \ - ((ipd)->irqs[irq].handler(irq, &__raw_get_cpu_var(__ipipe_tick_regs))) + ((ipd)->irqs[irq].handler(irq, __this_cpu_ptr(&__ipipe_tick_regs))) #define __ipipe_run_irqtail(irq) /* Must be a macro */ \ do { \ Index: linux/arch/blackfin/kernel/perf_event.c =================================================================== --- linux.orig/arch/blackfin/kernel/perf_event.c 2013-08-26 13:25:18.000000000 -0500 +++ linux/arch/blackfin/kernel/perf_event.c 2013-08-26 13:26:23.578979325 -0500 @@ -300,7 +300,7 @@ again: static void bfin_pmu_stop(struct perf_event *event, int flags) { - struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events); struct hw_perf_event *hwc = &event->hw; int idx = hwc->idx; @@ -318,7 +318,7 @@ static void bfin_pmu_stop(struct perf_ev static void bfin_pmu_start(struct perf_event *event, int flags) { - struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events); struct hw_perf_event *hwc = &event->hw; int idx = hwc->idx; @@ -335,7 +335,7 @@ static void bfin_pmu_start(struct perf_e static void bfin_pmu_del(struct perf_event *event, int flags) { - struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events); bfin_pmu_stop(event, PERF_EF_UPDATE); __clear_bit(event->hw.idx, cpuc->used_mask); @@ -345,7 +345,7 @@ static void bfin_pmu_del(struct perf_eve static int bfin_pmu_add(struct perf_event *event, int flags) { - struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events); struct hw_perf_event *hwc = &event->hw; int idx = hwc->idx; int ret = -EAGAIN; @@ -429,7 +429,7 @@ static int bfin_pmu_event_init(struct pe static void bfin_pmu_enable(struct pmu *pmu) { - struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events); struct perf_event *event; struct hw_perf_event *hwc; int i; Index: linux/arch/blackfin/mach-common/ints-priority.c =================================================================== --- linux.orig/arch/blackfin/mach-common/ints-priority.c 2013-08-26 13:25:18.000000000 -0500 +++ linux/arch/blackfin/mach-common/ints-priority.c 2013-08-26 13:26:23.578979325 -0500 @@ -1675,12 +1675,12 @@ asmlinkage int __ipipe_grab_irq(int vec, bfin_write_TIMER_STATUS(1); /* Latch TIMIL0 */ #endif /* This is basically what we need from the register frame. */ - __raw_get_cpu_var(__ipipe_tick_regs).ipend = regs->ipend; - __raw_get_cpu_var(__ipipe_tick_regs).pc = regs->pc; + __this_cpu_write(__ipipe_tick_regs.ipend, regs->ipend); + __this_cpu_write(__ipipe_tick_regs.pc, regs->pc); if (this_domain != ipipe_root_domain) - __raw_get_cpu_var(__ipipe_tick_regs).ipend &= ~0x10; + __this_cpu_and(__ipipe_tick_regs.ipend, ~0x10); else - __raw_get_cpu_var(__ipipe_tick_regs).ipend |= 0x10; + __this_cpu_or(__ipipe_tick_regs.ipend, 0x10); } /* Index: linux/arch/blackfin/mach-common/smp.c =================================================================== --- linux.orig/arch/blackfin/mach-common/smp.c 2013-08-26 13:25:18.000000000 -0500 +++ linux/arch/blackfin/mach-common/smp.c 2013-08-26 13:26:23.578979325 -0500 @@ -146,7 +146,7 @@ static irqreturn_t ipi_handler_int1(int platform_clear_ipi(cpu, IRQ_SUPPLE_1); - bfin_ipi_data = &__get_cpu_var(bfin_ipi); + bfin_ipi_data = this_cpu_ptr(&bfin_ipi); while ((pending = atomic_xchg(&bfin_ipi_data->bits, 0)) != 0) { msg = 0; do {