From: Vineet Gupta <Vineet.Gupta1@synopsys.com> To: "Peter Zijlstra (Intel)" <peterz@infradead.org> Cc: lkml <linux-kernel@vger.kernel.org>, <linux-arch@vger.kernel.org>, <arc-linux-dev@synopsys.com>, Vineet Gupta <Vineet.Gupta1@synopsys.com>, Will Deacon <will.deacon@arm.com> Subject: [PATCH v2 20/28] ARCv2: barriers Date: Tue, 23 Jun 2015 13:28:03 +0530 [thread overview] Message-ID: <1435046283-2073-1-git-send-email-vgupta@synopsys.com> (raw) In-Reply-To: <20150622133656.GG1583@arm.com> ARCv2 based HS38 cores are weakly ordered and thus explicit barriers for kernel proper. SMP barrier is provided by DMB instruction which also guarantees local barrier hence used as backend of smp_*mb() as well as *mb() APIs Also hookup barriers into MMIO accessors to avoid ordering issues in IO Cc: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Will Deacon <will.deacon@arm.com> Signed-off-by: Vineet Gupta <vgupta@synopsys.com> --- Changes since v1 * Better changelog and comments * local/mandatory barriers to NOT use DSYNC, but DMB * define DMB based mandatory barriers even for !SMP --- Signed-off-by: Vineet Gupta <vgupta@synopsys.com> --- arch/arc/include/asm/Kbuild | 1 - arch/arc/include/asm/barrier.h | 48 ++++++++++++++++++++++++++++++++++++++++++ arch/arc/include/asm/io.h | 42 +++++++++++++++++++++++++++++++++--- 3 files changed, 87 insertions(+), 4 deletions(-) create mode 100644 arch/arc/include/asm/barrier.h diff --git a/arch/arc/include/asm/Kbuild b/arch/arc/include/asm/Kbuild index be0c39e76f7c..59e2dd1d434f 100644 --- a/arch/arc/include/asm/Kbuild +++ b/arch/arc/include/asm/Kbuild @@ -1,5 +1,4 @@ generic-y += auxvec.h -generic-y += barrier.h generic-y += bitsperlong.h generic-y += bugs.h generic-y += clkdev.h diff --git a/arch/arc/include/asm/barrier.h b/arch/arc/include/asm/barrier.h new file mode 100644 index 000000000000..a7209983ee64 --- /dev/null +++ b/arch/arc/include/asm/barrier.h @@ -0,0 +1,48 @@ +/* + * Copyright (C) 2014-15 Synopsys, Inc. (www.synopsys.com) + * + * 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. + */ + +#ifndef __ASM_BARRIER_H +#define __ASM_BARRIER_H + +#ifdef CONFIG_ISA_ARCV2 + +/* + * ARCv2 based HS38 cores are in-order issue, but still weakly ordered + * due to micro-arch buffering/queuing of load/store, cache hit vs. miss ... + * + * Explicit barrier provided by DMB instruction + * - Operand supports fine grained load/store/load+store semantics + * - Ensures that selected memory operation issued before it will complete + * before any subsequent memory operation of same type + * - DMB guarantees SMP as well as local barrier semantics + * (asm-generic/barrier.h ensures sane smp_*mb if not defined here, i.e. + * UP: barrier(), SMP: smp_*mb == *mb) + * - DSYNC provides DMB+completion_of_cache_bpu_maintenance_ops hence not needed + * in the general case. Plus it only provides full barrier. + */ + +#define mb() asm volatile("dmb 3\n" : : : "memory") +#define rmb() asm volatile("dmb 1\n" : : : "memory") +#define wmb() asm volatile("dmb 2\n" : : : "memory") + +#endif + +#ifdef CONFIG_ISA_ARCOMPACT + +/* + * ARCompact based cores (ARC700) only have SYNC instruction which is super + * heavy weight as it flushes the pipeline as well. + * There are no real SMP implementations of such cores. + */ + +#define mb() asm volatile("sync\n" : : : "memory") +#endif + +#include <asm-generic/barrier.h> + +#endif diff --git a/arch/arc/include/asm/io.h b/arch/arc/include/asm/io.h index cabd518cb253..18c685bf9748 100644 --- a/arch/arc/include/asm/io.h +++ b/arch/arc/include/asm/io.h @@ -98,9 +98,45 @@ static inline void __raw_writel(u32 w, volatile void __iomem *addr) } -#define readb_relaxed readb -#define readw_relaxed readw -#define readl_relaxed readl +#ifdef CONFIG_ISA_ARCV2 +#include <asm/barrier.h> +#define __iormb() rmb() +#define __iowmb() wmb() +#else +#define __iormb() do { } while (0) +#define __iowmb() do { } while (0) +#endif + +/* + * MMIO can also get buffered/optimized in micro-arch, so barriers needed + * Based on ARM model for the typical use case + * + * <writel_relaxed DMA buffer> + * <writel MMIO "go" reg> + * or: + * <readl MMIO "status" reg> + * <readl_relaxed DMA buffer> + * + * http://www.spinics.net/lists/kernel/msg2018397.html + */ +#define readb(c) ({ u8 __v = readb_relaxed(c); __iormb(); __v; }) +#define readw(c) ({ u16 __v = readw_relaxed(c); __iormb(); __v; }) +#define readl(c) ({ u32 __v = readl_relaxed(c); __iormb(); __v; }) + +#define writeb(v,c) ({ __iowmb(); writeb_relaxed(v,c); }) +#define writew(v,c) ({ __iowmb(); writew_relaxed(v,c); }) +#define writel(v,c) ({ __iowmb(); writel_relaxed(v,c); }) + +/* + * Relaxed API for drivers which can handle any ordering themselves + */ +#define readb_relaxed(c) __raw_readb(c) +#define readw_relaxed(c) __raw_readw(c) +#define readl_relaxed(c) __raw_readl(c) + +#define writeb_relaxed(v,c) __raw_writeb(v,c) +#define writew_relaxed(v,c) __raw_writew(v,c) +#define writel_relaxed(v,c) __raw_writel(v,c) #include <asm-generic/io.h> -- 1.9.1
WARNING: multiple messages have this Message-ID (diff)
From: Vineet Gupta <Vineet.Gupta1@synopsys.com> To: "Peter Zijlstra (Intel)" <peterz@infradead.org> Cc: lkml <linux-kernel@vger.kernel.org>, linux-arch@vger.kernel.org, arc-linux-dev@synopsys.com, Vineet Gupta <Vineet.Gupta1@synopsys.com>, Will Deacon <will.deacon@arm.com> Subject: [PATCH v2 20/28] ARCv2: barriers Date: Tue, 23 Jun 2015 13:28:03 +0530 [thread overview] Message-ID: <1435046283-2073-1-git-send-email-vgupta@synopsys.com> (raw) In-Reply-To: <20150622133656.GG1583@arm.com> ARCv2 based HS38 cores are weakly ordered and thus explicit barriers for kernel proper. SMP barrier is provided by DMB instruction which also guarantees local barrier hence used as backend of smp_*mb() as well as *mb() APIs Also hookup barriers into MMIO accessors to avoid ordering issues in IO Cc: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Will Deacon <will.deacon@arm.com> Signed-off-by: Vineet Gupta <vgupta@synopsys.com> --- Changes since v1 * Better changelog and comments * local/mandatory barriers to NOT use DSYNC, but DMB * define DMB based mandatory barriers even for !SMP --- Signed-off-by: Vineet Gupta <vgupta@synopsys.com> --- arch/arc/include/asm/Kbuild | 1 - arch/arc/include/asm/barrier.h | 48 ++++++++++++++++++++++++++++++++++++++++++ arch/arc/include/asm/io.h | 42 +++++++++++++++++++++++++++++++++--- 3 files changed, 87 insertions(+), 4 deletions(-) create mode 100644 arch/arc/include/asm/barrier.h diff --git a/arch/arc/include/asm/Kbuild b/arch/arc/include/asm/Kbuild index be0c39e76f7c..59e2dd1d434f 100644 --- a/arch/arc/include/asm/Kbuild +++ b/arch/arc/include/asm/Kbuild @@ -1,5 +1,4 @@ generic-y += auxvec.h -generic-y += barrier.h generic-y += bitsperlong.h generic-y += bugs.h generic-y += clkdev.h diff --git a/arch/arc/include/asm/barrier.h b/arch/arc/include/asm/barrier.h new file mode 100644 index 000000000000..a7209983ee64 --- /dev/null +++ b/arch/arc/include/asm/barrier.h @@ -0,0 +1,48 @@ +/* + * Copyright (C) 2014-15 Synopsys, Inc. (www.synopsys.com) + * + * 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. + */ + +#ifndef __ASM_BARRIER_H +#define __ASM_BARRIER_H + +#ifdef CONFIG_ISA_ARCV2 + +/* + * ARCv2 based HS38 cores are in-order issue, but still weakly ordered + * due to micro-arch buffering/queuing of load/store, cache hit vs. miss ... + * + * Explicit barrier provided by DMB instruction + * - Operand supports fine grained load/store/load+store semantics + * - Ensures that selected memory operation issued before it will complete + * before any subsequent memory operation of same type + * - DMB guarantees SMP as well as local barrier semantics + * (asm-generic/barrier.h ensures sane smp_*mb if not defined here, i.e. + * UP: barrier(), SMP: smp_*mb == *mb) + * - DSYNC provides DMB+completion_of_cache_bpu_maintenance_ops hence not needed + * in the general case. Plus it only provides full barrier. + */ + +#define mb() asm volatile("dmb 3\n" : : : "memory") +#define rmb() asm volatile("dmb 1\n" : : : "memory") +#define wmb() asm volatile("dmb 2\n" : : : "memory") + +#endif + +#ifdef CONFIG_ISA_ARCOMPACT + +/* + * ARCompact based cores (ARC700) only have SYNC instruction which is super + * heavy weight as it flushes the pipeline as well. + * There are no real SMP implementations of such cores. + */ + +#define mb() asm volatile("sync\n" : : : "memory") +#endif + +#include <asm-generic/barrier.h> + +#endif diff --git a/arch/arc/include/asm/io.h b/arch/arc/include/asm/io.h index cabd518cb253..18c685bf9748 100644 --- a/arch/arc/include/asm/io.h +++ b/arch/arc/include/asm/io.h @@ -98,9 +98,45 @@ static inline void __raw_writel(u32 w, volatile void __iomem *addr) } -#define readb_relaxed readb -#define readw_relaxed readw -#define readl_relaxed readl +#ifdef CONFIG_ISA_ARCV2 +#include <asm/barrier.h> +#define __iormb() rmb() +#define __iowmb() wmb() +#else +#define __iormb() do { } while (0) +#define __iowmb() do { } while (0) +#endif + +/* + * MMIO can also get buffered/optimized in micro-arch, so barriers needed + * Based on ARM model for the typical use case + * + * <writel_relaxed DMA buffer> + * <writel MMIO "go" reg> + * or: + * <readl MMIO "status" reg> + * <readl_relaxed DMA buffer> + * + * http://www.spinics.net/lists/kernel/msg2018397.html + */ +#define readb(c) ({ u8 __v = readb_relaxed(c); __iormb(); __v; }) +#define readw(c) ({ u16 __v = readw_relaxed(c); __iormb(); __v; }) +#define readl(c) ({ u32 __v = readl_relaxed(c); __iormb(); __v; }) + +#define writeb(v,c) ({ __iowmb(); writeb_relaxed(v,c); }) +#define writew(v,c) ({ __iowmb(); writew_relaxed(v,c); }) +#define writel(v,c) ({ __iowmb(); writel_relaxed(v,c); }) + +/* + * Relaxed API for drivers which can handle any ordering themselves + */ +#define readb_relaxed(c) __raw_readb(c) +#define readw_relaxed(c) __raw_readw(c) +#define readl_relaxed(c) __raw_readl(c) + +#define writeb_relaxed(v,c) __raw_writeb(v,c) +#define writew_relaxed(v,c) __raw_writew(v,c) +#define writel_relaxed(v,c) __raw_writel(v,c) #include <asm-generic/io.h> -- 1.9.1
next prev parent reply other threads:[~2015-06-23 7:59 UTC|newest] Thread overview: 109+ messages / expand[flat|nested] mbox.gz Atom feed top 2015-06-09 11:48 [PATCH 00/28] ARCv2 port to Linux - (B) ISA / Core / platform support Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 01/28] ARCv2: [intc] HS38 core interrupt controller Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 02/28] ARCv2: Support for ARCv2 ISA and HS38x cores Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 03/28] ARCv2: STAR 9000793984: Handle return from intr to Delay Slot Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 04/28] ARCv2: STAR 9000808988: signals involving " Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 05/28] ARCv2: STAR 9000814690: Really Re-enable interrupts to avoid deadlocks Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 06/28] ARCv2: MMUv4: TLB programming Model changes Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 07/28] ARCv2: MMUv4: cache programming model changes Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 08/28] ARCv2: MMUv4: support aliasing icache config Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 09/28] ARCv2: optimised string/mem lib routines Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 10/28] ARCv2: Adhere to Zero Delay loop restriction Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 11/28] ARCv2: extable: Enable sorting at build time Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-24 5:51 ` Vineet Gupta 2015-06-24 5:51 ` Vineet Gupta 2015-06-29 20:38 ` David Daney 2015-06-30 4:41 ` Vineet Gupta 2015-06-30 4:41 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 12/28] ARCv2: clocksource: Introduce 64bit local RTC counter Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 13/28] ARC: make plat_smp_ops weak to allow over-rides Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 14/28] ARCv2: SMP: ARConnect debug/robustness Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 15/28] ARCv2: SMP: clocksource: Enable Global Real Time counter Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 16/28] ARCv2: SMP: intc: IDU 2nd level intc for dynamic IRQ distribution Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 17/28] ARC: add compiler barrier to LLSC based cmpxchg Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 12:23 ` Peter Zijlstra 2015-06-09 11:48 ` [PATCH 18/28] ARC: add smp barriers around atomics per memory-barrriers.txt Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 12:30 ` Peter Zijlstra 2015-06-10 9:17 ` Vineet Gupta 2015-06-10 10:53 ` Peter Zijlstra 2015-06-11 13:03 ` Vineet Gupta 2015-06-12 12:15 ` [PATCH v2] ARC: add smp barriers around atomics per Documentation/atomic_ops.txt Vineet Gupta 2015-06-12 12:15 ` Vineet Gupta 2015-06-12 13:04 ` Peter Zijlstra 2015-06-12 13:16 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 19/28] arch: conditionally define smp_{mb,rmb,wmb} Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 12:32 ` Peter Zijlstra 2015-06-09 11:48 ` [PATCH 20/28] ARCv2: barriers Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 12:40 ` Peter Zijlstra 2015-06-10 9:34 ` Vineet Gupta 2015-06-10 10:58 ` Peter Zijlstra 2015-06-10 13:01 ` Will Deacon 2015-06-11 12:13 ` Vineet Gupta 2015-06-11 13:39 ` Will Deacon 2015-06-19 13:13 ` Vineet Gupta 2015-06-19 13:13 ` Vineet Gupta 2015-06-19 13:13 ` Vineet Gupta 2015-06-22 13:36 ` Will Deacon 2015-06-22 13:36 ` Will Deacon 2015-06-22 13:36 ` Will Deacon 2015-06-23 7:58 ` Vineet Gupta [this message] 2015-06-23 7:58 ` [PATCH v2 " Vineet Gupta 2015-06-23 8:49 ` Will Deacon 2015-06-23 9:03 ` Vineet Gupta 2015-06-23 9:26 ` Will Deacon 2015-06-23 9:52 ` [PATCH v3 22/28] " Vineet Gupta 2015-06-23 9:52 ` Vineet Gupta 2015-06-23 16:28 ` Will Deacon 2015-06-23 9:25 ` [PATCH v2 20/28] " Peter Zijlstra 2015-06-23 8:02 ` [PATCH " Vineet Gupta 2015-06-09 11:48 ` [PATCH 21/28] ARC: Reduce bitops lines of code using macros Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-12 12:20 ` [PATCH v2] " Vineet Gupta 2015-06-12 12:20 ` Vineet Gupta 2015-06-12 13:05 ` Peter Zijlstra 2015-06-09 11:48 ` [PATCH 22/28] ARCv2: STAR 9000837815 workaround hardware exclusive transactions livelock Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 12:35 ` Peter Zijlstra 2015-06-10 10:01 ` Vineet Gupta 2015-06-10 11:02 ` Peter Zijlstra 2015-06-19 9:55 ` [PATCH v2 " Vineet Gupta 2015-06-19 9:55 ` Vineet Gupta 2015-06-19 9:59 ` Will Deacon 2015-06-19 10:09 ` Vineet Gupta 2015-06-23 7:59 ` Vineet Gupta 2015-06-23 7:59 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 23/28] ARCv2: SLC: Handle explcit flush for DMA ops (w/o IO-coherency) Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 24/28] ARCv2: All bits in place, allow ARCv2 builds Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 25/28] ARCv2: [nsim*hs*] Support simulation platforms for HS38x cores Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 26/28] ARC: [axs101] Prepare for AXS103 Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 27/28] ARCv2: [axs103] Support ARC SDP FPGA platform for HS38x cores Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta 2015-06-09 11:48 ` [PATCH 28/28] ARCv2: [vdk] dts files and defconfig for HS38 VDK Vineet Gupta 2015-06-09 11:48 ` Vineet Gupta
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=1435046283-2073-1-git-send-email-vgupta@synopsys.com \ --to=vineet.gupta1@synopsys.com \ --cc=arc-linux-dev@synopsys.com \ --cc=linux-arch@vger.kernel.org \ --cc=linux-kernel@vger.kernel.org \ --cc=peterz@infradead.org \ --cc=will.deacon@arm.com \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.