From patchwork Tue Sep 19 17:14:29 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Mark Rutland X-Patchwork-Id: 142006 Return-Path: Delivered-To: ouuuleilei@gmail.com Received: by 2002:a05:612c:172:b0:3f2:4152:657d with SMTP id h50csp3556762vqi; Tue, 19 Sep 2023 10:26:05 -0700 (PDT) X-Google-Smtp-Source: AGHT+IGAmpM9Xut9XJ3RPyOMN0B9LsfrOp/EWfJXtXG8XTdbp2+NqSIlZMX6x6zAiOfSiT16g9oK X-Received: by 2002:a05:6a00:cd0:b0:690:15c7:60bf with SMTP id b16-20020a056a000cd000b0069015c760bfmr360621pfv.20.1695144364820; Tue, 19 Sep 2023 10:26:04 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1695144364; cv=none; d=google.com; s=arc-20160816; b=UyqOIuszGNjynQBevF1sHg5ni8siJjt/s9QCGQ5Q7azvfaOGJkFJ1Vm+Cp6zmlE113 OYxNyP8XohbQElMe0sQdWUUkF7kvPZKPh7oeDunBmIC+mTqflvqP7DzDvGbCLl3KiPId uPc0NUNMBp19ZkMN7b2CElGa+hlR+tMcYA4Mz9ciXn9/iNpuNt8m9Ud2dvKfA99QG6p3 6dg+WcZJM9XPuiItuq8YNSxDhAbn6Ze7ZS/IU6sOarjne7dY8HCPdAZgDVckaEECv8rI 9RMTUHfF9szs8A+jJWsbmRV3YtSb6IhZi/r0SLOeiYYPW7oSyIGuFdvHwZw/+QwiXd7s u+Bg== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=list-id:precedence:content-transfer-encoding:mime-version :message-id:date:subject:cc:to:from; bh=xoyHkFnpe2TqmFZ8nYmKzh8VjWViEznnZ/h18e7vuJ8=; fh=gfjMbt8VYQia4qWUxG4jE82HCGLXormtyVOFWklLTvM=; b=h4dvlpcvHSB+tM2Ya3pzA5zDbmu8kbBBg8jf3raZCSFOiGDpH1VNojr6mDAFmSjjzu sZaPRUNkMmoEwSW7nXPyliLhwXOF1L2q8KkFu+zasHo7JNkefrQWLmmm8aTdWH2R7LGv Jsj4dDwSYhdHm1cUizgZZyT0BV0GpPxvBdgnJiq8yG1hLa4qeTlSZvuQDBt85jb5u5OH c4qyudrXUumSOmg2cpTQznIMJhPR5WqD6WxKhkNUbxl8CBnMCNb934us3VdvmQJQNV9u fZ1KWDTlGvseD0qg+RXbb3qtXedIdR6YEFVNuCcV1fnb4Yu5N45x0Ep/FJwEq4xd8XnP 66tQ== ARC-Authentication-Results: i=1; mx.google.com; spf=pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.37 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=fail (p=NONE sp=NONE dis=NONE) header.from=arm.com Received: from snail.vger.email (snail.vger.email. [23.128.96.37]) by mx.google.com with ESMTPS id r22-20020a638f56000000b005648d3f2031si9863994pgn.362.2023.09.19.10.26.04 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Tue, 19 Sep 2023 10:26:04 -0700 (PDT) Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.37 as permitted sender) client-ip=23.128.96.37; Authentication-Results: mx.google.com; spf=pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.37 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=fail (p=NONE sp=NONE dis=NONE) header.from=arm.com Received: from out1.vger.email (depot.vger.email [IPv6:2620:137:e000::3:0]) by snail.vger.email (Postfix) with ESMTP id 5982981890A2; Tue, 19 Sep 2023 10:14:46 -0700 (PDT) X-Virus-Status: Clean X-Virus-Scanned: clamav-milter 0.103.10 at snail.vger.email Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231993AbjISROp (ORCPT + 26 others); Tue, 19 Sep 2023 13:14:45 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:43718 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230203AbjISROn (ORCPT ); Tue, 19 Sep 2023 13:14:43 -0400 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by lindbergh.monkeyblade.net (Postfix) with ESMTP id 50E3ADC for ; Tue, 19 Sep 2023 10:14:37 -0700 (PDT) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 52FF21FB; Tue, 19 Sep 2023 10:15:14 -0700 (PDT) Received: from lakrids.cambridge.arm.com (usa-sjc-imap-foss1.foss.arm.com [10.121.207.14]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id 5CE2F3F5A1; Tue, 19 Sep 2023 10:14:35 -0700 (PDT) From: Mark Rutland To: linux-kernel@vger.kernel.org, Peter Zijlstra Cc: boqun.feng@gmail.com, david@fromorbit.com, djwong@kernel.org, kent.overstreet@linux.dev, libaokun1@huawei.com, linux-arm-kernel@lists.infradead.org, mark.rutland@arm.com, ming.lei@redhat.com, will@kernel.org, yi.zhang@redhat.com Subject: [PATCH] locking/atomic: scripts: fix fallback ifdeffery Date: Tue, 19 Sep 2023 18:14:29 +0100 Message-Id: <20230919171430.2697727-1-mark.rutland@arm.com> X-Mailer: git-send-email 2.30.2 MIME-Version: 1.0 X-Spam-Status: No, score=-4.2 required=5.0 tests=BAYES_00,RCVD_IN_DNSWL_MED, SPF_HELO_NONE,SPF_NONE autolearn=ham autolearn_force=no version=3.4.6 X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on lindbergh.monkeyblade.net Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.6.4 (snail.vger.email [0.0.0.0]); Tue, 19 Sep 2023 10:14:46 -0700 (PDT) X-getmail-retrieved-from-mailbox: INBOX X-GMAIL-THRID: 1777487697257882048 X-GMAIL-MSGID: 1777487697257882048 Since commit: 9257959a6e5b4fca ("locking/atomic: scripts: restructure fallback ifdeffery") The ordering fallbacks for atomic*_read_acquire() and atomic*_set_release() erroneously fall back to the implictly relaxed atomic*_read() and atomic*_set() variants respectively, without any additional barriers. This loses the ACQUIRE and RELEASE ordering semantics, which can result in a wide variety of problems, even on strongly-ordered architectures where the implementation of atomic*_read() and/or atomic*_set() allows the compiler to reorder those relative to other accesses. In practice this has been observed to break bit spinlocks on arm64, resulting in dentry cache corruption. The fallback logic was intended to allow ACQUIRE/RELEASE/RELAXED ops to be defined in terms of FULL ops, but where an op had RELAXED ordering by default, this unintentionally permitted the ACQUIRE/RELEASE ops to be defined in terms of the implicitly RELAXED default. This patch corrects the logic to avoid falling back to implicitly RELAXED ops, resulting in the same behaviour as prior to commit 9257959a6e5b4fca. I've verified the resulting assembly on arm64 by generating outlined wrappers of the atomics. Prior to this patch the compiler generates sequences using relaxed load (LDR) and store (STR) instructions, e.g. | : | ldr x0, [x0] | ret | | : | str x1, [x0] | ret With this patch applied the compiler generates sequences using the intended load-acquire (LDAR) and store-release (STLR) instructions, e.g. | : | ldar x0, [x0] | ret | | : | stlr x1, [x0] | ret To make sure that there were no other victims of the ifdeffery rewrite, I generated outlined copies of all of the {atomic,atomic64,atomic_long} atomic operations before and after commit 9257959a6e5b4fca. A diff of the generated assembly on arm64 shows that only the read_acquire() and set_release() operations were changed, and only lost their intended ordering: | [mark@lakrids:~/src/linux]% diff -u \ | <(aarch64-linux-gnu-objdump -d before-9257959a6e5b4fca.o) | <(aarch64-linux-gnu-objdump -d after-9257959a6e5b4fca.o) | --- /proc/self/fd/11 2023-09-19 16:51:51.114779415 +0100 | +++ /proc/self/fd/16 2023-09-19 16:51:51.114779415 +0100 | @@ -1,5 +1,5 @@ | | -before-9257959a6e5b4fca.o: file format elf64-littleaarch64 | +after-9257959a6e5b4fca.o: file format elf64-littleaarch64 | | | Disassembly of section .text: | @@ -9,7 +9,7 @@ | 4: d65f03c0 ret | | 0000000000000008 : | - 8: 88dffc00 ldar w0, [x0] | + 8: b9400000 ldr w0, [x0] | c: d65f03c0 ret | | 0000000000000010 : | @@ -17,7 +17,7 @@ | 14: d65f03c0 ret | | 0000000000000018 : | - 18: 889ffc01 stlr w1, [x0] | + 18: b9000001 str w1, [x0] | 1c: d65f03c0 ret | | 0000000000000020 : | @@ -1230,7 +1230,7 @@ | 1070: d65f03c0 ret | | 0000000000001074 : | - 1074: c8dffc00 ldar x0, [x0] | + 1074: f9400000 ldr x0, [x0] | 1078: d65f03c0 ret | | 000000000000107c : | @@ -1238,7 +1238,7 @@ | 1080: d65f03c0 ret | | 0000000000001084 : | - 1084: c89ffc01 stlr x1, [x0] | + 1084: f9000001 str x1, [x0] | 1088: d65f03c0 ret | | 000000000000108c : | @@ -2427,7 +2427,7 @@ | 207c: d65f03c0 ret | | 0000000000002080 : | - 2080: c8dffc00 ldar x0, [x0] | + 2080: f9400000 ldr x0, [x0] | 2084: d65f03c0 ret | | 0000000000002088 : | @@ -2435,7 +2435,7 @@ | 208c: d65f03c0 ret | | 0000000000002090 : | - 2090: c89ffc01 stlr x1, [x0] | + 2090: f9000001 str x1, [x0] | 2094: d65f03c0 ret | | 0000000000002098 : I've build tested this with a variety of configs for alpha, arm, arm64, csky, i386, m68k, microblaze, mips, nios2, openrisc, powerpc, riscv, s390, sh, sparc, x86_64, and xtensa, for which I've seen no issues. I was unable to build test for ia64 and parisc due to existing build breakage in v6.6-rc2. Fixes: 9257959a6e5b4fca ("locking/atomic: scripts: restructure fallback ifdeffery") Reported-by: Ming Lei Link: https://lore.kernel.org/all/ZOWFtqA2om0w5Vmz@fedora/ Reported-by: Darrick J. Wong Link: https://lore.kernel.org/linux-fsdevel/20230912173026.GA3389127@frogsfrogsfrogs/ Signed-off-by: Mark Rutland Cc: Baokun Li Cc: Boqun Feng Cc: Darrick J. Wong Cc: Dave Chinner Cc: Kent Overstreet Cc: Ming Lei Cc: Peter Zijlstra Cc: Will Deacon Cc: Yi Zhang Cc: linux-arm-kernel@lists.infradead.org Tested-by: Darrick J. Wong --- include/linux/atomic/atomic-arch-fallback.h | 10 +--------- scripts/atomic/gen-atomic-fallback.sh | 2 +- 2 files changed, 2 insertions(+), 10 deletions(-) Peter, are you happy to queue this in the tip tree? It's a pretty nasty regresssion in v6.5, and I'd like to get this in as a fix for v6.6 ASAP. Thanks, Mark. diff --git a/include/linux/atomic/atomic-arch-fallback.h b/include/linux/atomic/atomic-arch-fallback.h index 18f5744dfb5d8..b83ef19da13de 100644 --- a/include/linux/atomic/atomic-arch-fallback.h +++ b/include/linux/atomic/atomic-arch-fallback.h @@ -459,8 +459,6 @@ raw_atomic_read_acquire(const atomic_t *v) { #if defined(arch_atomic_read_acquire) return arch_atomic_read_acquire(v); -#elif defined(arch_atomic_read) - return arch_atomic_read(v); #else int ret; @@ -508,8 +506,6 @@ raw_atomic_set_release(atomic_t *v, int i) { #if defined(arch_atomic_set_release) arch_atomic_set_release(v, i); -#elif defined(arch_atomic_set) - arch_atomic_set(v, i); #else if (__native_word(atomic_t)) { smp_store_release(&(v)->counter, i); @@ -2575,8 +2571,6 @@ raw_atomic64_read_acquire(const atomic64_t *v) { #if defined(arch_atomic64_read_acquire) return arch_atomic64_read_acquire(v); -#elif defined(arch_atomic64_read) - return arch_atomic64_read(v); #else s64 ret; @@ -2624,8 +2618,6 @@ raw_atomic64_set_release(atomic64_t *v, s64 i) { #if defined(arch_atomic64_set_release) arch_atomic64_set_release(v, i); -#elif defined(arch_atomic64_set) - arch_atomic64_set(v, i); #else if (__native_word(atomic64_t)) { smp_store_release(&(v)->counter, i); @@ -4657,4 +4649,4 @@ raw_atomic64_dec_if_positive(atomic64_t *v) } #endif /* _LINUX_ATOMIC_FALLBACK_H */ -// 202b45c7db600ce36198eb1f1fc2c2d5268ace2d +// 2fdd6702823fa842f9cea57a002e6e4476ae780c diff --git a/scripts/atomic/gen-atomic-fallback.sh b/scripts/atomic/gen-atomic-fallback.sh index c0c8a85d7c81b..a45154cefa487 100755 --- a/scripts/atomic/gen-atomic-fallback.sh +++ b/scripts/atomic/gen-atomic-fallback.sh @@ -102,7 +102,7 @@ gen_proto_order_variant() fi # Allow ACQUIRE/RELEASE/RELAXED ops to be defined in terms of FULL ops - if [ ! -z "${order}" ]; then + if [ ! -z "${order}" ] && ! meta_is_implicitly_relaxed "${meta}"; then printf "#elif defined(arch_${basename})\n" printf "\t${retstmt}arch_${basename}(${args});\n" fi