From patchwork Thu May 8 18:11:26 2014 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Christoph Lameter (Ampere)" X-Patchwork-Id: 4138131 X-Patchwork-Delegate: lethal@linux-sh.org Return-Path: X-Original-To: patchwork-linux-sh@patchwork.kernel.org Delivered-To: patchwork-parsemail@patchwork1.web.kernel.org Received: from mail.kernel.org (mail.kernel.org [198.145.19.201]) by patchwork1.web.kernel.org (Postfix) with ESMTP id 0073F9F1E1 for ; Thu, 8 May 2014 18:11:38 +0000 (UTC) Received: from mail.kernel.org (localhost [127.0.0.1]) by mail.kernel.org (Postfix) with ESMTP id 85C0D20115 for ; Thu, 8 May 2014 18:11:37 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 5669220170 for ; Thu, 8 May 2014 18:11:32 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751885AbaEHSLb (ORCPT ); Thu, 8 May 2014 14:11:31 -0400 Received: from qmta09.emeryville.ca.mail.comcast.net ([76.96.30.96]:38099 "EHLO qmta09.emeryville.ca.mail.comcast.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752428AbaEHSL3 (ORCPT ); Thu, 8 May 2014 14:11:29 -0400 Received: from omta08.emeryville.ca.mail.comcast.net ([76.96.30.12]) by qmta09.emeryville.ca.mail.comcast.net with comcast id zQZ51n0030FhH24A9WBVZl; Thu, 08 May 2014 18:11:29 +0000 Received: from gentwo.org ([98.213.233.247]) by omta08.emeryville.ca.mail.comcast.net with comcast id zWBS1n00U5Lw0ES8UWBTxZ; Thu, 08 May 2014 18:11:29 +0000 Received: by gentwo.org (Postfix, from userid 1001) id 882EC27B1; Thu, 8 May 2014 13:11:26 -0500 (CDT) Received: from localhost (localhost [127.0.0.1]) by gentwo.org (Postfix) with ESMTP id 8438F168; Thu, 8 May 2014 13:11:26 -0500 (CDT) Date: Thu, 8 May 2014 13:11:26 -0500 (CDT) From: Christoph Lameter X-X-Sender: cl@gentwo.org To: akpm@linux-foundation.org cc: Geert Uytterhoeven , Paul Mundt , Linux-sh list Subject: Re: sh: Replace __get_cpu_var uses In-Reply-To: Message-ID: References: DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=comcast.net; s=q20140121; t=1399572689; bh=pfNOt2fQ4ld2Zrj6kQqDmDLWZJiPs5vHyXKAms7fDUk=; h=Received:Received:Received:Received:Date:From:To:Subject: Message-ID:Content-Type; b=CvJCEV7kbNRUcx8+s8TpySudO3d6b5pbywxKrdKImEI5m02LmrUa0CUitX9fHwy6b bX1rjYpeof0R8Yo+EXgG/kip6ve5s/eFq9SNDefza+pl1dBiW8AsB13G4iIL8/Iw+O MG/Fl6MK7xx8ANujtvp7CfOUS5SSr/lBTZozR6f+n2bAMDrYVT69xBlaaNyKDXwSjL lr5y7bY3Iw5u2N/w5Yo2PDFXchzYYTNAS4vo9O4SfzVvRdA3yqFpZYbAlHJ4vSPQoT PpPzUfWxIOrCjKKiLxJM0Qv3KGENnkHVMw+bipkMgHGaXWd9D7hoEy+qlPcbxfbkCz pL0QyrkWnZgUQ== Sender: linux-sh-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-sh@vger.kernel.org X-Spam-Status: No, score=-7.4 required=5.0 tests=BAYES_00,DKIM_SIGNED, RCVD_IN_DNSWL_HI,RP_MATCHES_RCVD,T_DKIM_INVALID,UNPARSEABLE_RELAY autolearn=ham version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on mail.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP On Tue, 6 May 2014, Geert Uytterhoeven wrote: > >> > Could you see if this compiles and works? > >> > >> shx3_defconfig (one of the SMP defconfigs) still compiles fine. > > > > Could you merge the patch? > > SH is orphan. Please send to akpm. Andrew here is the patch: Subject: sh: Replace __get_cpu_var uses __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. Other use 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 an 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 an explicit address calculation using this_cpu_ptr() or into a use of this_cpu operations that use the offset. Thereby address calculations are avoided and less registers are used when code is generated. At the end of the patch set all uses of __get_cpu_var have been removed so the macro is removed too. The patch set 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, y); 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(&x, this_cpu_ptr(&y), 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) Cc: Paul Mundt CC: linux-sh@vger.kernel.org Compile-Tested-by: Geert Uytterhoeven Signed-off-by: Christoph Lameter --- To unsubscribe from this list: send the line "unsubscribe linux-sh" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Index: linux/arch/sh/kernel/hw_breakpoint.c =================================================================== --- linux.orig/arch/sh/kernel/hw_breakpoint.c 2014-02-25 13:26:10.850735155 -0600 +++ linux/arch/sh/kernel/hw_breakpoint.c 2014-02-25 13:26:10.850735155 -0600 @@ -52,7 +52,7 @@ int i; for (i = 0; i < sh_ubc->num_events; i++) { - struct perf_event **slot = &__get_cpu_var(bp_per_reg[i]); + struct perf_event **slot = this_cpu_ptr(&bp_per_reg[i]); if (!*slot) { *slot = bp; @@ -84,7 +84,7 @@ int i; for (i = 0; i < sh_ubc->num_events; i++) { - struct perf_event **slot = &__get_cpu_var(bp_per_reg[i]); + struct perf_event **slot = this_cpu_ptr(&bp_per_reg[i]); if (*slot == bp) { *slot = NULL; Index: linux/arch/sh/kernel/kprobes.c =================================================================== --- linux.orig/arch/sh/kernel/kprobes.c 2014-02-25 13:26:10.850735155 -0600 +++ linux/arch/sh/kernel/kprobes.c 2014-02-25 13:26:10.850735155 -0600 @@ -102,7 +102,7 @@ void __kprobes arch_remove_kprobe(struct kprobe *p) { - struct kprobe *saved = &__get_cpu_var(saved_next_opcode); + struct kprobe *saved = this_cpu_ptr(&saved_next_opcode); if (saved->addr) { arch_disarm_kprobe(p); @@ -111,7 +111,7 @@ saved->addr = NULL; saved->opcode = 0; - saved = &__get_cpu_var(saved_next_opcode2); + saved = this_cpu_ptr(&saved_next_opcode2); if (saved->addr) { arch_disarm_kprobe(saved); @@ -129,14 +129,14 @@ static void __kprobes restore_previous_kprobe(struct kprobe_ctlblk *kcb) { - __get_cpu_var(current_kprobe) = kcb->prev_kprobe.kp; + __this_cpu_write(current_kprobe, kcb->prev_kprobe.kp); kcb->kprobe_status = kcb->prev_kprobe.status; } static void __kprobes set_current_kprobe(struct kprobe *p, struct pt_regs *regs, struct kprobe_ctlblk *kcb) { - __get_cpu_var(current_kprobe) = p; + __this_cpu_write(current_kprobe, p); } /* @@ -146,15 +146,15 @@ */ static void __kprobes prepare_singlestep(struct kprobe *p, struct pt_regs *regs) { - __get_cpu_var(saved_current_opcode).addr = (kprobe_opcode_t *)regs->pc; + __this_cpu_write(saved_current_opcode.addr, (kprobe_opcode_t *)regs->pc); if (p != NULL) { struct kprobe *op1, *op2; arch_disarm_kprobe(p); - op1 = &__get_cpu_var(saved_next_opcode); - op2 = &__get_cpu_var(saved_next_opcode2); + op1 = this_cpu_ptr(&saved_next_opcode); + op2 = this_cpu_ptr(&saved_next_opcode2); if (OPCODE_JSR(p->opcode) || OPCODE_JMP(p->opcode)) { unsigned int reg_nr = ((p->opcode >> 8) & 0x000F); @@ -249,7 +249,7 @@ kcb->kprobe_status = KPROBE_REENTER; return 1; } else { - p = __get_cpu_var(current_kprobe); + p = __this_cpu_read(current_kprobe); if (p->break_handler && p->break_handler(p, regs)) { goto ss_probe; } @@ -336,9 +336,9 @@ continue; if (ri->rp && ri->rp->handler) { - __get_cpu_var(current_kprobe) = &ri->rp->kp; + __this_cpu_write(current_kprobe, &ri->rp->kp); ri->rp->handler(ri, regs); - __get_cpu_var(current_kprobe) = NULL; + __this_cpu_write(current_kprobe, NULL); } orig_ret_address = (unsigned long)ri->ret_addr; @@ -383,19 +383,19 @@ cur->post_handler(cur, regs, 0); } - p = &__get_cpu_var(saved_next_opcode); + p = this_cpu_ptr(&saved_next_opcode); if (p->addr) { arch_disarm_kprobe(p); p->addr = NULL; p->opcode = 0; - addr = __get_cpu_var(saved_current_opcode).addr; - __get_cpu_var(saved_current_opcode).addr = NULL; + addr = __this_cpu_read(saved_current_opcode.addr); + __this_cpu_write(saved_current_opcode.addr, NULL); p = get_kprobe(addr); arch_arm_kprobe(p); - p = &__get_cpu_var(saved_next_opcode2); + p = this_cpu_ptr(&saved_next_opcode2); if (p->addr) { arch_disarm_kprobe(p); p->addr = NULL; @@ -511,7 +511,7 @@ if (kprobe_handler(args->regs)) { ret = NOTIFY_STOP; } else { - p = __get_cpu_var(current_kprobe); + p = __this_cpu_read(current_kprobe); if (p->break_handler && p->break_handler(p, args->regs)) ret = NOTIFY_STOP; Index: linux/arch/sh/kernel/localtimer.c =================================================================== --- linux.orig/arch/sh/kernel/localtimer.c 2014-02-25 13:26:10.850735155 -0600 +++ linux/arch/sh/kernel/localtimer.c 2014-02-25 13:26:10.850735155 -0600 @@ -32,7 +32,7 @@ */ void local_timer_interrupt(void) { - struct clock_event_device *clk = &__get_cpu_var(local_clockevent); + struct clock_event_device *clk = this_cpu_ptr(&local_clockevent); irq_enter(); clk->event_handler(clk); Index: linux/arch/sh/kernel/perf_event.c =================================================================== --- linux.orig/arch/sh/kernel/perf_event.c 2014-02-25 13:26:10.850735155 -0600 +++ linux/arch/sh/kernel/perf_event.c 2014-02-25 13:26:10.850735155 -0600 @@ -227,7 +227,7 @@ static void sh_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; @@ -245,7 +245,7 @@ static void sh_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; @@ -262,7 +262,7 @@ static void sh_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); sh_pmu_stop(event, PERF_EF_UPDATE); __clear_bit(event->hw.idx, cpuc->used_mask); @@ -272,7 +272,7 @@ static int sh_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; Index: linux/arch/sh/kernel/smp.c =================================================================== --- linux.orig/arch/sh/kernel/smp.c 2014-02-25 13:26:10.850735155 -0600 +++ linux/arch/sh/kernel/smp.c 2014-02-25 13:26:10.850735155 -0600 @@ -111,7 +111,7 @@ irq_ctx_exit(raw_smp_processor_id()); mb(); - __get_cpu_var(cpu_state) = CPU_DEAD; + __this_cpu_write(cpu_state, CPU_DEAD); local_irq_disable(); }