From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 6C60BC43334 for ; Wed, 8 Jun 2022 14:41:20 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S241878AbiFHOlS (ORCPT ); Wed, 8 Jun 2022 10:41:18 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:55312 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241705AbiFHOlE (ORCPT ); Wed, 8 Jun 2022 10:41:04 -0400 Received: from dfw.source.kernel.org (dfw.source.kernel.org [139.178.84.217]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 5CBA5157E83 for ; Wed, 8 Jun 2022 07:40:49 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by dfw.source.kernel.org (Postfix) with ESMTPS id 864BF61B95 for ; Wed, 8 Jun 2022 14:40:48 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id BA7C7C341C8; Wed, 8 Jun 2022 14:40:44 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699248; bh=ns1vdA0/qrQYg4Q4rjwNdkKJ1jBqx0nuAlBJgu/kcNw=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=FftvwcczPCwlYEmoEN+HsJhaAgib1ISbeICzSUDzCi11ZeNCEfw6gIVs69DLI7JX9 Z1TBMDcv09De9frZ5eBu7PcNt/FRZyhwk18mbmb+4zKpRqxxTvlUA8P0tvaGa+4GB6 8/TRSz0A+Vae7JgK18uDyO7x1eiAS2r2sXVEmNF1anh7Ya6PsuB8Z2CrAYsw2x+VL2 jWkeYJd1jzDhhRXiFf5fa8Rc04b3fSWAmUW49KhytZvNFQ7ZlxEwf25moY7hNmhQ+Y WZ7dqDLU6TOXDybqxztnbh4mCIgZ99a3ahNFlvLvkjIfmBU34yRaONKuE+emFWigUE kCOQkxrf2LF+Q== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 01/20] context_tracking: Remove unused context_tracking_in_user() Date: Wed, 8 Jun 2022 16:40:18 +0200 Message-Id: <20220608144037.1765000-2-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" This function is not used and CT_WARN_ON() coupled with ct_state() is the preferred way to assert context tracking state values. Reported-by: Nicolas Saenz Julienne Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Signed-off-by: Frederic Weisbecker Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- include/linux/context_tracking_state.h | 5 ----- 1 file changed, 5 deletions(-) diff --git a/include/linux/context_tracking_state.h b/include/linux/context= _tracking_state.h index 65a60d3313b0..02be3ac2040d 100644 --- a/include/linux/context_tracking_state.h +++ b/include/linux/context_tracking_state.h @@ -41,12 +41,7 @@ static inline bool context_tracking_enabled_this_cpu(voi= d) return context_tracking_enabled() && __this_cpu_read(context_tracking.act= ive); } =20 -static __always_inline bool context_tracking_in_user(void) -{ - return __this_cpu_read(context_tracking.state) =3D=3D CONTEXT_USER; -} #else -static inline bool context_tracking_in_user(void) { return false; } static inline bool context_tracking_enabled(void) { return false; } static inline bool context_tracking_enabled_cpu(int cpu) { return false; } static inline bool context_tracking_enabled_this_cpu(void) { return false;= } --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id D4394C433EF for ; Wed, 8 Jun 2022 14:41:41 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S241732AbiFHOlk (ORCPT ); Wed, 8 Jun 2022 10:41:40 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:56310 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241773AbiFHOlF (ORCPT ); Wed, 8 Jun 2022 10:41:05 -0400 Received: from dfw.source.kernel.org (dfw.source.kernel.org [IPv6:2604:1380:4641:c500::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id EE2E014CDFF for ; Wed, 8 Jun 2022 07:40:52 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by dfw.source.kernel.org (Postfix) with ESMTPS id 3ABFB61B9D for ; Wed, 8 Jun 2022 14:40:52 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 6A431C34116; Wed, 8 Jun 2022 14:40:48 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699251; bh=y5jci0bqRsZ94ArUjZ+9wtURoMvqstdDqOTJ7Eps920=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=jMTVaqAHpvYtuYZbDdq0xRaTEYUoMoxdL9RzYX4fvHhmmIBADbynz/isWETDuuZpr 6rJLLQYsgYy129uWysAmkBqNofZdLzW7C+wmpUEVoCyg17DfCD7JGASvG7FlH+YIT0 fvmpBkcK4PEb8PzrBYx9kQkT7plW75ia0RgIIS7DX1vB2HeOMlTQg2IewoOtb4GjHd tnZQhW+VOqvI45IfoRAZoo91ft0oROL/ZpPHA2UxpsAtbmXUy1FxddX8/4EDfuf4BQ 83Xb6Y3B/i4rijDVLjWO0LzJut9JvBe9AllEVr9eIVkfi3wHbGaBRE6/tjrWgtDovR i7hcI8RcnZEgQ== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 02/20] context_tracking: Add a note about noinstr VS unsafe context tracking functions Date: Wed, 8 Jun 2022 16:40:19 +0200 Message-Id: <20220608144037.1765000-3-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Some context tracking functions enter or exit into/from RCU idle mode while using trace-able and lockdep-aware IRQs (un-)masking. As a result those functions can't get tagged as noinstr. This is unlikely to be fixed since these are obsolete APIs. Drop a note about this matter. Reported-by: Peter Zijlstra Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Signed-off-by: Frederic Weisbecker Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- kernel/context_tracking.c | 35 +++++++++++++++++++++++++++++++++++ 1 file changed, 35 insertions(+) diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index 36a98c48aedc..b8032ebf4314 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -103,6 +103,16 @@ void noinstr __context_tracking_enter(enum ctx_state s= tate) } EXPORT_SYMBOL_GPL(__context_tracking_enter); =20 +/* + * OBSOLETE: + * This function should be noinstr but the below local_irq_restore() is + * unsafe because it involves illegal RCU uses through tracing and lockdep. + * This is unlikely to be fixed as this function is obsolete. The preferred + * way is to call __context_tracking_enter() through user_enter_irqoff() + * or context_tracking_guest_enter(). It should be the arch entry code + * responsibility to call into context tracking with IRQs disabled. + + */ void context_tracking_enter(enum ctx_state state) { unsigned long flags; @@ -125,6 +135,14 @@ void context_tracking_enter(enum ctx_state state) NOKPROBE_SYMBOL(context_tracking_enter); EXPORT_SYMBOL_GPL(context_tracking_enter); =20 +/* + * OBSOLETE: + * This function should be noinstr but it unsafely calls local_irq_restore= (), + * involving illegal RCU uses through tracing and lockdep. + * This is unlikely to be fixed as this function is obsolete. The preferred + * way is to call user_enter_irqoff(). It should be the arch entry code + * responsibility to call into context tracking with IRQs disabled. + */ void context_tracking_user_enter(void) { user_enter(); @@ -168,6 +186,15 @@ void noinstr __context_tracking_exit(enum ctx_state st= ate) } EXPORT_SYMBOL_GPL(__context_tracking_exit); =20 +/* + * OBSOLETE: + * This function should be noinstr but the below local_irq_save() is + * unsafe because it involves illegal RCU uses through tracing and lockdep. + * This is unlikely to be fixed as this function is obsolete. The preferred + * way is to call __context_tracking_exit() through user_exit_irqoff() + * or context_tracking_guest_exit(). It should be the arch entry code + * responsibility to call into context tracking with IRQs disabled. + */ void context_tracking_exit(enum ctx_state state) { unsigned long flags; @@ -182,6 +209,14 @@ void context_tracking_exit(enum ctx_state state) NOKPROBE_SYMBOL(context_tracking_exit); EXPORT_SYMBOL_GPL(context_tracking_exit); =20 +/* + * OBSOLETE: + * This function should be noinstr but it unsafely calls local_irq_save(), + * involving illegal RCU uses through tracing and lockdep. This is unlikely + * to be fixed as this function is obsolete. The preferred way is to call + * user_exit_irqoff(). It should be the arch entry code responsibility to + * call into context tracking with IRQs disabled. + */ void context_tracking_user_exit(void) { user_exit(); --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 94D85C433EF for ; Wed, 8 Jun 2022 14:41:46 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S241780AbiFHOln (ORCPT ); Wed, 8 Jun 2022 10:41:43 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:56066 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241585AbiFHOlF (ORCPT ); Wed, 8 Jun 2022 10:41:05 -0400 Received: from dfw.source.kernel.org (dfw.source.kernel.org [IPv6:2604:1380:4641:c500::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 756AA159054 for ; Wed, 8 Jun 2022 07:40:56 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by dfw.source.kernel.org (Postfix) with ESMTPS id D605D61BA0 for ; Wed, 8 Jun 2022 14:40:55 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 193DEC341C4; Wed, 8 Jun 2022 14:40:51 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699255; bh=cMW9QYUBe2UBJwvy5hnwDPFIzlB2U2C+Ns2kaploG1Y=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=eiwAYK+/vEcbIIU/Il8iZEx4+gqyVopH4vqkKC+a60eujpV9JYdJdc/WA5m0GxvDJ nMTRoRIXvuDOJ4egh0BtVL7BKBP8ijZS2ErF5JizZftiPUuyFQ+pzNm7SCcj8eqwhN v5wJICTqok+Z4cmSoPsduAZVDmWp09J3U0PKxwVM/q4l6dCvSAHQRNKjoH5k7NO4iU P6F8gqi+8PljU6HQ55rk4W+Gms+hOYdjcVdUBnOSGF35jxCR6yWvICBSQjvg3fw1H7 BN7hfyXiivLDbAB6KndbqTBc/yaxJ1SUwC3+JCg8Ot4zeqRUaWWSumWykaBWJXVS2/ 6VBWn9qlV8M6g== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 03/20] context_tracking: Rename __context_tracking_enter/exit() to __ct_user_enter/exit() Date: Wed, 8 Jun 2022 16:40:20 +0200 Message-Id: <20220608144037.1765000-4-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" The context tracking namespace is going to expand and some new functions will require even longer names. Start shrinking the context_tracking prefix to "ct" as is already the case for some existing macros, this will make the introduction of new functions easier. Acked-by: Paul E. McKenney Signed-off-by: Frederic Weisbecker Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- include/linux/context_tracking.h | 12 ++++++------ kernel/context_tracking.c | 20 ++++++++++---------- 2 files changed, 16 insertions(+), 16 deletions(-) diff --git a/include/linux/context_tracking.h b/include/linux/context_track= ing.h index 7a14807c9d1a..773035124bad 100644 --- a/include/linux/context_tracking.h +++ b/include/linux/context_tracking.h @@ -14,8 +14,8 @@ extern void context_tracking_cpu_set(int cpu); =20 /* Called with interrupts disabled. */ -extern void __context_tracking_enter(enum ctx_state state); -extern void __context_tracking_exit(enum ctx_state state); +extern void __ct_user_enter(enum ctx_state state); +extern void __ct_user_exit(enum ctx_state state); =20 extern void context_tracking_enter(enum ctx_state state); extern void context_tracking_exit(enum ctx_state state); @@ -38,13 +38,13 @@ static inline void user_exit(void) static __always_inline void user_enter_irqoff(void) { if (context_tracking_enabled()) - __context_tracking_enter(CONTEXT_USER); + __ct_user_enter(CONTEXT_USER); =20 } static __always_inline void user_exit_irqoff(void) { if (context_tracking_enabled()) - __context_tracking_exit(CONTEXT_USER); + __ct_user_exit(CONTEXT_USER); } =20 static inline enum ctx_state exception_enter(void) @@ -74,7 +74,7 @@ static inline void exception_exit(enum ctx_state prev_ctx) static __always_inline bool context_tracking_guest_enter(void) { if (context_tracking_enabled()) - __context_tracking_enter(CONTEXT_GUEST); + __ct_user_enter(CONTEXT_GUEST); =20 return context_tracking_enabled_this_cpu(); } @@ -82,7 +82,7 @@ static __always_inline bool context_tracking_guest_enter(= void) static __always_inline void context_tracking_guest_exit(void) { if (context_tracking_enabled()) - __context_tracking_exit(CONTEXT_GUEST); + __ct_user_exit(CONTEXT_GUEST); } =20 /** diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index b8032ebf4314..e499ffb1d041 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -51,15 +51,15 @@ static __always_inline void context_tracking_recursion_= exit(void) } =20 /** - * context_tracking_enter - Inform the context tracking that the CPU is go= ing - * enter user or guest space mode. + * __ct_user_enter - Inform the context tracking that the CPU is going + * to enter user or guest space mode. * * This function must be called right before we switch from the kernel * to user or guest space, when it's guaranteed the remaining kernel * instructions to execute won't use any RCU read side critical section * because this function sets RCU in extended quiescent state. */ -void noinstr __context_tracking_enter(enum ctx_state state) +void noinstr __ct_user_enter(enum ctx_state state) { /* Kernel threads aren't supposed to go to userspace */ WARN_ON_ONCE(!current->mm); @@ -101,7 +101,7 @@ void noinstr __context_tracking_enter(enum ctx_state st= ate) } context_tracking_recursion_exit(); } -EXPORT_SYMBOL_GPL(__context_tracking_enter); +EXPORT_SYMBOL_GPL(__ct_user_enter); =20 /* * OBSOLETE: @@ -129,7 +129,7 @@ void context_tracking_enter(enum ctx_state state) return; =20 local_irq_save(flags); - __context_tracking_enter(state); + __ct_user_enter(state); local_irq_restore(flags); } NOKPROBE_SYMBOL(context_tracking_enter); @@ -150,8 +150,8 @@ void context_tracking_user_enter(void) NOKPROBE_SYMBOL(context_tracking_user_enter); =20 /** - * context_tracking_exit - Inform the context tracking that the CPU is - * exiting user or guest mode and entering the ker= nel. + * __ct_user_exit - Inform the context tracking that the CPU is + * exiting user or guest mode and entering the kernel. * * This function must be called after we entered the kernel from user or * guest space before any use of RCU read side critical section. This @@ -161,7 +161,7 @@ NOKPROBE_SYMBOL(context_tracking_user_enter); * This call supports re-entrancy. This way it can be called from any exce= ption * handler without needing to know if we came from userspace or not. */ -void noinstr __context_tracking_exit(enum ctx_state state) +void noinstr __ct_user_exit(enum ctx_state state) { if (!context_tracking_recursion_enter()) return; @@ -184,7 +184,7 @@ void noinstr __context_tracking_exit(enum ctx_state sta= te) } context_tracking_recursion_exit(); } -EXPORT_SYMBOL_GPL(__context_tracking_exit); +EXPORT_SYMBOL_GPL(__ct_user_exit); =20 /* * OBSOLETE: @@ -203,7 +203,7 @@ void context_tracking_exit(enum ctx_state state) return; =20 local_irq_save(flags); - __context_tracking_exit(state); + __ct_user_exit(state); local_irq_restore(flags); } NOKPROBE_SYMBOL(context_tracking_exit); --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 9CB68C43334 for ; Wed, 8 Jun 2022 14:41:35 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S241701AbiFHOld (ORCPT ); Wed, 8 Jun 2022 10:41:33 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:56166 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241801AbiFHOlG (ORCPT ); Wed, 8 Jun 2022 10:41:06 -0400 Received: from dfw.source.kernel.org (dfw.source.kernel.org [139.178.84.217]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id EEF9318492E for ; Wed, 8 Jun 2022 07:40:59 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by dfw.source.kernel.org (Postfix) with ESMTPS id 8AEC561B9D for ; Wed, 8 Jun 2022 14:40:59 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id B8E6AC385A5; Wed, 8 Jun 2022 14:40:55 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699259; bh=B1YiA16i/bsh0gW37/8HDqduEMSfDpjIaUkr1sKqPYc=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=DllcGQ16eEgYazAhFuTI+nQS6HkTz5XLshHmM08y6JP1bK0FFuWKKWWZH9sduXIna zyOoD88tzAfHPVGEkVG9s9B8whwhjHwExa52n2Rn/MD+n6gsVoxxBWHaIr4f6UlRMD xIX65yed5yaHDkWN7Hfrpg/ZUQOealsN0U3RDUJ9fwi1JOzZoRCnpX0GxRtNWM8IZY XLNtNy3jheivTHh0x7Bem0K4VH2aJaYvUX5qogF7jq61379eImRf950UOdRY36/rRQ d2Q/r8dtmiA1Zo/oexhgHCHh/u3IKCTJ9X/nNr+bxcZXCIQhuQA3sFPtJX3jH9L2d/ t6YTxtk2xHEsg== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 04/20] context_tracking: Rename context_tracking_user_enter/exit() to user_enter/exit_callable() Date: Wed, 8 Jun 2022 16:40:21 +0200 Message-Id: <20220608144037.1765000-5-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" context_tracking_user_enter() and context_tracking_user_exit() are ASM callable versions of user_enter() and user_exit() for architectures that didn't manage to check the context tracking static key from ASM. Change those function names to better reflect their purpose. Signed-off-by: Frederic Weisbecker Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- arch/arm/kernel/entry-header.S | 8 ++++---- arch/csky/kernel/entry.S | 4 ++-- arch/riscv/kernel/entry.S | 6 +++--- include/linux/context_tracking.h | 4 ++-- kernel/context_tracking.c | 28 +++++++++++++++++----------- 5 files changed, 28 insertions(+), 22 deletions(-) diff --git a/arch/arm/kernel/entry-header.S b/arch/arm/kernel/entry-header.S index 9a1dc142f782..0c14dcc43e58 100644 --- a/arch/arm/kernel/entry-header.S +++ b/arch/arm/kernel/entry-header.S @@ -370,10 +370,10 @@ ALT_UP_B(.L1_\@) #ifdef CONFIG_CONTEXT_TRACKING .if \save stmdb sp!, {r0-r3, ip, lr} - bl context_tracking_user_exit + bl user_exit_callable ldmia sp!, {r0-r3, ip, lr} .else - bl context_tracking_user_exit + bl user_exit_callable .endif #endif .endm @@ -382,10 +382,10 @@ ALT_UP_B(.L1_\@) #ifdef CONFIG_CONTEXT_TRACKING .if \save stmdb sp!, {r0-r3, ip, lr} - bl context_tracking_user_enter + bl user_enter_callable ldmia sp!, {r0-r3, ip, lr} .else - bl context_tracking_user_enter + bl user_enter_callable .endif #endif .endm diff --git a/arch/csky/kernel/entry.S b/arch/csky/kernel/entry.S index a4ababf25e24..bc734d17c16f 100644 --- a/arch/csky/kernel/entry.S +++ b/arch/csky/kernel/entry.S @@ -23,7 +23,7 @@ mfcr a0, epsr btsti a0, 31 bt 1f - jbsr context_tracking_user_exit + jbsr user_exit_callable ldw a0, (sp, LSAVE_A0) ldw a1, (sp, LSAVE_A1) ldw a2, (sp, LSAVE_A2) @@ -160,7 +160,7 @@ ret_from_exception: cmpnei r10, 0 bt exit_work #ifdef CONFIG_CONTEXT_TRACKING - jbsr context_tracking_user_enter + jbsr user_enter_callable #endif 1: #ifdef CONFIG_PREEMPTION diff --git a/arch/riscv/kernel/entry.S b/arch/riscv/kernel/entry.S index c8b9ce274b9a..ecd132af2edf 100644 --- a/arch/riscv/kernel/entry.S +++ b/arch/riscv/kernel/entry.S @@ -112,11 +112,11 @@ _save_context: #endif =20 #ifdef CONFIG_CONTEXT_TRACKING - /* If previous state is in user mode, call context_tracking_user_exit. */ + /* If previous state is in user mode, call user_exit_callable(). */ li a0, SR_PP and a0, s1, a0 bnez a0, skip_context_tracking - call context_tracking_user_exit + call user_exit_callable skip_context_tracking: #endif =20 @@ -256,7 +256,7 @@ resume_userspace: bnez s1, work_pending =20 #ifdef CONFIG_CONTEXT_TRACKING - call context_tracking_user_enter + call user_enter_callable #endif =20 /* Save unwound kernel stack pointer in thread_info */ diff --git a/include/linux/context_tracking.h b/include/linux/context_track= ing.h index 773035124bad..69532cd18f72 100644 --- a/include/linux/context_tracking.h +++ b/include/linux/context_tracking.h @@ -19,8 +19,8 @@ extern void __ct_user_exit(enum ctx_state state); =20 extern void context_tracking_enter(enum ctx_state state); extern void context_tracking_exit(enum ctx_state state); -extern void context_tracking_user_enter(void); -extern void context_tracking_user_exit(void); +extern void user_enter_callable(void); +extern void user_exit_callable(void); =20 static inline void user_enter(void) { diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index e499ffb1d041..30ae833cad9f 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -135,19 +135,22 @@ void context_tracking_enter(enum ctx_state state) NOKPROBE_SYMBOL(context_tracking_enter); EXPORT_SYMBOL_GPL(context_tracking_enter); =20 -/* - * OBSOLETE: - * This function should be noinstr but it unsafely calls local_irq_restore= (), - * involving illegal RCU uses through tracing and lockdep. +/** + * user_enter_callable() - Unfortunate ASM callable version of user_enter(= ) for + * archs that didn't manage to check the context tracking + * static key from low level code. + * + * This OBSOLETE function should be noinstr but it unsafely calls + * local_irq_restore(), involving illegal RCU uses through tracing and loc= kdep. * This is unlikely to be fixed as this function is obsolete. The preferred * way is to call user_enter_irqoff(). It should be the arch entry code * responsibility to call into context tracking with IRQs disabled. */ -void context_tracking_user_enter(void) +void user_enter_callable(void) { user_enter(); } -NOKPROBE_SYMBOL(context_tracking_user_enter); +NOKPROBE_SYMBOL(user_enter_callable); =20 /** * __ct_user_exit - Inform the context tracking that the CPU is @@ -209,19 +212,22 @@ void context_tracking_exit(enum ctx_state state) NOKPROBE_SYMBOL(context_tracking_exit); EXPORT_SYMBOL_GPL(context_tracking_exit); =20 -/* - * OBSOLETE: - * This function should be noinstr but it unsafely calls local_irq_save(), +/** + * user_exit_callable() - Unfortunate ASM callable version of user_exit() = for + * archs that didn't manage to check the context tracking + * static key from low level code. + * + * This OBSOLETE function should be noinstr but it unsafely calls local_ir= q_save(), * involving illegal RCU uses through tracing and lockdep. This is unlikely * to be fixed as this function is obsolete. The preferred way is to call * user_exit_irqoff(). It should be the arch entry code responsibility to * call into context tracking with IRQs disabled. */ -void context_tracking_user_exit(void) +void user_exit_callable(void) { user_exit(); } -NOKPROBE_SYMBOL(context_tracking_user_exit); +NOKPROBE_SYMBOL(user_exit_callable); =20 void __init context_tracking_cpu_set(int cpu) { --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 7B7ABC43334 for ; Wed, 8 Jun 2022 14:41:52 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S241604AbiFHOlu (ORCPT ); Wed, 8 Jun 2022 10:41:50 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:54896 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241813AbiFHOlK (ORCPT ); Wed, 8 Jun 2022 10:41:10 -0400 Received: from ams.source.kernel.org (ams.source.kernel.org [IPv6:2604:1380:4601:e00::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 1EC5914AF70 for ; Wed, 8 Jun 2022 07:41:05 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ams.source.kernel.org (Postfix) with ESMTPS id CCC2DB827F3 for ; Wed, 8 Jun 2022 14:41:03 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 67A28C385A9; Wed, 8 Jun 2022 14:40:59 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699262; bh=/SWLJXUlZnLxnhIg9IH7Gzyw9PQw2dtLQX5NWgCFG1g=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=MBw5wl/hBrK9JivyeIngxyQVifeGFhA1sNoLbp5n6/t5QVHstu+C2bUS8deTdmGc1 cdf3R1sefUefb1gZ7uHniXZUSv5PchdYJJxaw54B9VK9ngtRUyyCVB7f7vCVyYBycV I57cMKwxXfz+JDOZxwjjtYb4hTH8AMUFLgD30U4+LP7IXGq27sJEB6IwjtqcDk44sb /iykF5a1U/evokUFls/GXh4jhTKrj9onDlfQ/ejUotvyajBvZPo7f7apNFRvjKSmQS V5g740weJKe648Nhy+120RgoptCcmS2AfKBYwrLTil9pqgFrJiMvgaZh78O62QYBLn mUk3wZJOalq9A== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 05/20] context_tracking: Rename context_tracking_enter/exit() to ct_user_enter/exit() Date: Wed, 8 Jun 2022 16:40:22 +0200 Message-Id: <20220608144037.1765000-6-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" context_tracking_enter() and context_tracking_exit() have confusing names that don't explain the fact they are referring to user/guest state. Use more self-explanatory names and shrink to the new context tracking prefix instead. Signed-off-by: Frederic Weisbecker Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- include/linux/context_tracking.h | 13 +++++++------ kernel/context_tracking.c | 12 ++++++------ 2 files changed, 13 insertions(+), 12 deletions(-) diff --git a/include/linux/context_tracking.h b/include/linux/context_track= ing.h index 69532cd18f72..7a5f04ae1758 100644 --- a/include/linux/context_tracking.h +++ b/include/linux/context_tracking.h @@ -17,21 +17,22 @@ extern void context_tracking_cpu_set(int cpu); extern void __ct_user_enter(enum ctx_state state); extern void __ct_user_exit(enum ctx_state state); =20 -extern void context_tracking_enter(enum ctx_state state); -extern void context_tracking_exit(enum ctx_state state); +extern void ct_user_enter(enum ctx_state state); +extern void ct_user_exit(enum ctx_state state); + extern void user_enter_callable(void); extern void user_exit_callable(void); =20 static inline void user_enter(void) { if (context_tracking_enabled()) - context_tracking_enter(CONTEXT_USER); + ct_user_enter(CONTEXT_USER); =20 } static inline void user_exit(void) { if (context_tracking_enabled()) - context_tracking_exit(CONTEXT_USER); + ct_user_exit(CONTEXT_USER); } =20 /* Called with interrupts disabled. */ @@ -57,7 +58,7 @@ static inline enum ctx_state exception_enter(void) =20 prev_ctx =3D this_cpu_read(context_tracking.state); if (prev_ctx !=3D CONTEXT_KERNEL) - context_tracking_exit(prev_ctx); + ct_user_exit(prev_ctx); =20 return prev_ctx; } @@ -67,7 +68,7 @@ static inline void exception_exit(enum ctx_state prev_ctx) if (!IS_ENABLED(CONFIG_HAVE_CONTEXT_TRACKING_OFFSTACK) && context_tracking_enabled()) { if (prev_ctx !=3D CONTEXT_KERNEL) - context_tracking_enter(prev_ctx); + ct_user_enter(prev_ctx); } } =20 diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index 30ae833cad9f..3a264ac38a02 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -113,7 +113,7 @@ EXPORT_SYMBOL_GPL(__ct_user_enter); * responsibility to call into context tracking with IRQs disabled. =20 */ -void context_tracking_enter(enum ctx_state state) +void ct_user_enter(enum ctx_state state) { unsigned long flags; =20 @@ -132,8 +132,8 @@ void context_tracking_enter(enum ctx_state state) __ct_user_enter(state); local_irq_restore(flags); } -NOKPROBE_SYMBOL(context_tracking_enter); -EXPORT_SYMBOL_GPL(context_tracking_enter); +NOKPROBE_SYMBOL(ct_user_enter); +EXPORT_SYMBOL_GPL(ct_user_enter); =20 /** * user_enter_callable() - Unfortunate ASM callable version of user_enter(= ) for @@ -198,7 +198,7 @@ EXPORT_SYMBOL_GPL(__ct_user_exit); * or context_tracking_guest_exit(). It should be the arch entry code * responsibility to call into context tracking with IRQs disabled. */ -void context_tracking_exit(enum ctx_state state) +void ct_user_exit(enum ctx_state state) { unsigned long flags; =20 @@ -209,8 +209,8 @@ void context_tracking_exit(enum ctx_state state) __ct_user_exit(state); local_irq_restore(flags); } -NOKPROBE_SYMBOL(context_tracking_exit); -EXPORT_SYMBOL_GPL(context_tracking_exit); +NOKPROBE_SYMBOL(ct_user_exit); +EXPORT_SYMBOL_GPL(ct_user_exit); =20 /** * user_exit_callable() - Unfortunate ASM callable version of user_exit() = for --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id C1FBCC433EF for ; Wed, 8 Jun 2022 14:41:24 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S241659AbiFHOlW (ORCPT ); Wed, 8 Jun 2022 10:41:22 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:56504 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241894AbiFHOlM (ORCPT ); Wed, 8 Jun 2022 10:41:12 -0400 Received: from ams.source.kernel.org (ams.source.kernel.org [IPv6:2604:1380:4601:e00::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 17150158F16 for ; Wed, 8 Jun 2022 07:41:09 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ams.source.kernel.org (Postfix) with ESMTPS id 85B0FB8284C for ; Wed, 8 Jun 2022 14:41:07 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 13D42C36AFF; Wed, 8 Jun 2022 14:41:02 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699266; bh=wVAhOyUu30Q3+FO0o3QLjnw4Mq751lT+fVw296W0Vkk=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=CcBijD6+TSvEEZx97G6zYWd1mVqh0O4OVJOakXiOldLbop7N9rN+XnOTSxErALCoC jvlwxn1bUez5t52iUNeT902TSQ2/Ad11rR4tTNIZmOvR9sN9GWnORH+oH5Zt80f4Xk PVz3Q9nIfu6HHBtrYRcFnmghQfoWDDhdTG2pL6yfFthWqaSiHgY0TuGF5CkGglSzYp Uv6JeVYHWETpDvKRMYfNqO6aHHZ2vbyajrj1hreTRQ7jTWOU2ohQg6jKQBRrYo7osr 1DIltzVZPXGOZWfL1Ux/hpsZmq/7V9Iy42Z716AxZfykPEoRfu3fmvk8Jm8ITAbjZi HBYfhL4tz4uBQ== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 06/20] context_tracking: Rename context_tracking_cpu_set() to ct_cpu_track_user() Date: Wed, 8 Jun 2022 16:40:23 +0200 Message-Id: <20220608144037.1765000-7-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" context_tracking_cpu_set() is called in order to tell a CPU to track user/kernel transitions. Since context tracking is going to expand in to also track transitions from/to idle/IRQ/NMIs, the scope of this function name becomes too broad and needs to be made more specific. Also shorten the prefix to align with the new namespace. Signed-off-by: Frederic Weisbecker Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- include/linux/context_tracking.h | 2 +- kernel/context_tracking.c | 4 ++-- kernel/time/tick-sched.c | 2 +- 3 files changed, 4 insertions(+), 4 deletions(-) diff --git a/include/linux/context_tracking.h b/include/linux/context_track= ing.h index 7a5f04ae1758..63259fece7c7 100644 --- a/include/linux/context_tracking.h +++ b/include/linux/context_tracking.h @@ -11,7 +11,7 @@ =20 =20 #ifdef CONFIG_CONTEXT_TRACKING -extern void context_tracking_cpu_set(int cpu); +extern void ct_cpu_track_user(int cpu); =20 /* Called with interrupts disabled. */ extern void __ct_user_enter(enum ctx_state state); diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index 3a264ac38a02..ccf18032e742 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -229,7 +229,7 @@ void user_exit_callable(void) } NOKPROBE_SYMBOL(user_exit_callable); =20 -void __init context_tracking_cpu_set(int cpu) +void __init ct_cpu_track_user(int cpu) { static __initdata bool initialized =3D false; =20 @@ -259,6 +259,6 @@ void __init context_tracking_init(void) int cpu; =20 for_each_possible_cpu(cpu) - context_tracking_cpu_set(cpu); + ct_cpu_track_user(cpu); } #endif diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c index 2d76c91b85de..06709885c590 100644 --- a/kernel/time/tick-sched.c +++ b/kernel/time/tick-sched.c @@ -571,7 +571,7 @@ void __init tick_nohz_init(void) } =20 for_each_cpu(cpu, tick_nohz_full_mask) - context_tracking_cpu_set(cpu); + ct_cpu_track_user(cpu); =20 ret =3D cpuhp_setup_state_nocalls(CPUHP_AP_ONLINE_DYN, "kernel/nohz:predown", NULL, --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 9B5A8C433EF for ; Wed, 8 Jun 2022 14:41:29 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S241519AbiFHOl1 (ORCPT ); Wed, 8 Jun 2022 10:41:27 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:56310 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241956AbiFHOlP (ORCPT ); Wed, 8 Jun 2022 10:41:15 -0400 Received: from ams.source.kernel.org (ams.source.kernel.org [145.40.68.75]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 09EF5496A2 for ; Wed, 8 Jun 2022 07:41:12 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ams.source.kernel.org (Postfix) with ESMTPS id 33286B827F3 for ; Wed, 8 Jun 2022 14:41:11 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id B6E3AC3411E; Wed, 8 Jun 2022 14:41:06 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699270; bh=noCvXHT3hTPC6hVQpAboJ87v6eyBLh4OfiIeqvsggx8=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=eJchtFrCpj8XYWOVkkXh7nLJ3qCt88pRqQbB4dckvZusVTRfd0c6r7i7KZEJ5A5xg tVPhF301HkllYzMRXFpWuN/ZzAOMAkhilH6Qr90q/is207JyPopL1VJZNTbV3eYhnx merUu3FpNh6t+rIi6sNVYtYn2DBo82mb8Hr2tkaKEwe1SRc3GY7yG/Lm1zXRypF7Yx Z9Xbaa/XFz2VDwdPJXd+eXm4wolKrPILdRt0pPDpI1xKIcltTU4JjDlPQEjwcQuhvB ICvwT9SMjlxdy62FmPWi3rOJDpiIBRbgh5avwVW92hHIp/UOSIwpDdxIIrXpoBA17Z eQ0cWf2/4qitA== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 07/20] context_tracking: Split user tracking Kconfig Date: Wed, 8 Jun 2022 16:40:24 +0200 Message-Id: <20220608144037.1765000-8-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Context tracking is going to be used not only to track user transitions but also idle/IRQs/NMIs. The user tracking part will then become a separate feature. Prepare Kconfig for that. Signed-off-by: Frederic Weisbecker Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- .../time/context-tracking/arch-support.txt | 6 ++-- arch/Kconfig | 4 +-- arch/arm/Kconfig | 2 +- arch/arm/kernel/entry-common.S | 4 +-- arch/arm/kernel/entry-header.S | 4 +-- arch/arm64/Kconfig | 2 +- arch/csky/Kconfig | 2 +- arch/csky/kernel/entry.S | 4 +-- arch/mips/Kconfig | 2 +- arch/powerpc/Kconfig | 2 +- arch/powerpc/include/asm/context_tracking.h | 2 +- arch/riscv/Kconfig | 2 +- arch/riscv/kernel/entry.S | 6 ++-- arch/sparc/Kconfig | 2 +- arch/sparc/kernel/rtrap_64.S | 2 +- arch/x86/Kconfig | 4 +-- include/linux/context_tracking.h | 12 +++---- include/linux/context_tracking_state.h | 4 +-- init/Kconfig | 4 +-- kernel/context_tracking.c | 6 +++- kernel/sched/core.c | 2 +- kernel/time/Kconfig | 31 ++++++++++++------- 22 files changed, 61 insertions(+), 48 deletions(-) diff --git a/Documentation/features/time/context-tracking/arch-support.txt = b/Documentation/features/time/context-tracking/arch-support.txt index bb1c1801553e..59ec30c84541 100644 --- a/Documentation/features/time/context-tracking/arch-support.txt +++ b/Documentation/features/time/context-tracking/arch-support.txt @@ -1,7 +1,7 @@ # -# Feature name: context-tracking -# Kconfig: HAVE_CONTEXT_TRACKING -# description: arch supports context tracking for NO_HZ_FULL +# Feature name: user-context-tracking +# Kconfig: HAVE_CONTEXT_TRACKING_USER +# description: arch supports user context tracking for NO_HZ_FULL # ----------------------- | arch |status| diff --git a/arch/Kconfig b/arch/Kconfig index 1bf29ce754af..7a8cd4d67d9c 100644 --- a/arch/Kconfig +++ b/arch/Kconfig @@ -769,7 +769,7 @@ config HAVE_ARCH_WITHIN_STACK_FRAMES and similar) by implementing an inline arch_within_stack_frames(), which is used by CONFIG_HARDENED_USERCOPY. =20 -config HAVE_CONTEXT_TRACKING +config HAVE_CONTEXT_TRACKING_USER bool help Provide kernel/user boundaries probes necessary for subsystems @@ -780,7 +780,7 @@ config HAVE_CONTEXT_TRACKING protected inside rcu_irq_enter/rcu_irq_exit() but preemption or signal handling on irq exit still need to be protected. =20 -config HAVE_CONTEXT_TRACKING_OFFSTACK +config HAVE_CONTEXT_TRACKING_USER_OFFSTACK bool help Architecture neither relies on exception_enter()/exception_exit() diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index 2e8091e2d8a8..64eb31946598 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -84,7 +84,7 @@ config ARM select HAVE_ARCH_TRANSPARENT_HUGEPAGE if ARM_LPAE select HAVE_ARM_SMCCC if CPU_V7 select HAVE_EBPF_JIT if !CPU_ENDIAN_BE32 - select HAVE_CONTEXT_TRACKING + select HAVE_CONTEXT_TRACKING_USER select HAVE_C_RECORDMCOUNT select HAVE_BUILDTIME_MCOUNT_SORT select HAVE_DEBUG_KMEMLEAK if !XIP_KERNEL diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S index 90d40f4d56cf..393d3945a43f 100644 --- a/arch/arm/kernel/entry-common.S +++ b/arch/arm/kernel/entry-common.S @@ -28,7 +28,7 @@ #include "entry-header.S" =20 saved_psr .req r8 -#if defined(CONFIG_TRACE_IRQFLAGS) || defined(CONFIG_CONTEXT_TRACKING) +#if defined(CONFIG_TRACE_IRQFLAGS) || defined(CONFIG_CONTEXT_TRACKING_USER) saved_pc .req r9 #define TRACE(x...) x #else @@ -38,7 +38,7 @@ saved_pc .req lr =20 .section .entry.text,"ax",%progbits .align 5 -#if !(IS_ENABLED(CONFIG_TRACE_IRQFLAGS) || IS_ENABLED(CONFIG_CONTEXT_TRACK= ING) || \ +#if !(IS_ENABLED(CONFIG_TRACE_IRQFLAGS) || IS_ENABLED(CONFIG_CONTEXT_TRACK= ING_USER) || \ IS_ENABLED(CONFIG_DEBUG_RSEQ)) /* * This is the fast syscall return path. We do as little as possible here, diff --git a/arch/arm/kernel/entry-header.S b/arch/arm/kernel/entry-header.S index 0c14dcc43e58..8a54ff24bed0 100644 --- a/arch/arm/kernel/entry-header.S +++ b/arch/arm/kernel/entry-header.S @@ -367,7 +367,7 @@ ALT_UP_B(.L1_\@) * between user and kernel mode. */ .macro ct_user_exit, save =3D 1 -#ifdef CONFIG_CONTEXT_TRACKING +#ifdef CONFIG_CONTEXT_TRACKING_USER .if \save stmdb sp!, {r0-r3, ip, lr} bl user_exit_callable @@ -379,7 +379,7 @@ ALT_UP_B(.L1_\@) .endm =20 .macro ct_user_enter, save =3D 1 -#ifdef CONFIG_CONTEXT_TRACKING +#ifdef CONFIG_CONTEXT_TRACKING_USER .if \save stmdb sp!, {r0-r3, ip, lr} bl user_enter_callable diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index 57c4c995965f..6e8757b89a54 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -171,7 +171,7 @@ config ARM64 select HAVE_C_RECORDMCOUNT select HAVE_CMPXCHG_DOUBLE select HAVE_CMPXCHG_LOCAL - select HAVE_CONTEXT_TRACKING + select HAVE_CONTEXT_TRACKING_USER select HAVE_DEBUG_KMEMLEAK select HAVE_DMA_CONTIGUOUS select HAVE_DYNAMIC_FTRACE diff --git a/arch/csky/Kconfig b/arch/csky/Kconfig index 75ef86605d69..171b59901214 100644 --- a/arch/csky/Kconfig +++ b/arch/csky/Kconfig @@ -42,7 +42,7 @@ config CSKY select HAVE_ARCH_AUDITSYSCALL select HAVE_ARCH_MMAP_RND_BITS select HAVE_ARCH_SECCOMP_FILTER - select HAVE_CONTEXT_TRACKING + select HAVE_CONTEXT_TRACKING_USER select HAVE_VIRT_CPU_ACCOUNTING_GEN select HAVE_DEBUG_BUGVERBOSE select HAVE_DEBUG_KMEMLEAK diff --git a/arch/csky/kernel/entry.S b/arch/csky/kernel/entry.S index bc734d17c16f..547b4cd1b24b 100644 --- a/arch/csky/kernel/entry.S +++ b/arch/csky/kernel/entry.S @@ -19,7 +19,7 @@ .endm =20 .macro context_tracking -#ifdef CONFIG_CONTEXT_TRACKING +#ifdef CONFIG_CONTEXT_TRACKING_USER mfcr a0, epsr btsti a0, 31 bt 1f @@ -159,7 +159,7 @@ ret_from_exception: and r10, r9 cmpnei r10, 0 bt exit_work -#ifdef CONFIG_CONTEXT_TRACKING +#ifdef CONFIG_CONTEXT_TRACKING_USER jbsr user_enter_callable #endif 1: diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig index de3b32a507d2..d0048cddd558 100644 --- a/arch/mips/Kconfig +++ b/arch/mips/Kconfig @@ -56,7 +56,7 @@ config MIPS select HAVE_ARCH_TRACEHOOK select HAVE_ARCH_TRANSPARENT_HUGEPAGE if CPU_SUPPORTS_HUGEPAGES select HAVE_ASM_MODVERSIONS - select HAVE_CONTEXT_TRACKING + select HAVE_CONTEXT_TRACKING_USER select HAVE_TIF_NOHZ select HAVE_C_RECORDMCOUNT select HAVE_DEBUG_KMEMLEAK diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig index 174edabb74fa..f2fddc09a345 100644 --- a/arch/powerpc/Kconfig +++ b/arch/powerpc/Kconfig @@ -197,7 +197,7 @@ config PPC select HAVE_ARCH_SECCOMP_FILTER select HAVE_ARCH_TRACEHOOK select HAVE_ASM_MODVERSIONS - select HAVE_CONTEXT_TRACKING if PPC64 + select HAVE_CONTEXT_TRACKING_USER if PPC64 select HAVE_C_RECORDMCOUNT select HAVE_DEBUG_KMEMLEAK select HAVE_DEBUG_STACKOVERFLOW diff --git a/arch/powerpc/include/asm/context_tracking.h b/arch/powerpc/inc= lude/asm/context_tracking.h index f2682b28b050..4b63931c49e0 100644 --- a/arch/powerpc/include/asm/context_tracking.h +++ b/arch/powerpc/include/asm/context_tracking.h @@ -2,7 +2,7 @@ #ifndef _ASM_POWERPC_CONTEXT_TRACKING_H #define _ASM_POWERPC_CONTEXT_TRACKING_H =20 -#ifdef CONFIG_CONTEXT_TRACKING +#ifdef CONFIG_CONTEXT_TRACKING_USER #define SCHEDULE_USER bl schedule_user #else #define SCHEDULE_USER bl schedule diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig index 00fd9c548f26..08316de9fa9f 100644 --- a/arch/riscv/Kconfig +++ b/arch/riscv/Kconfig @@ -83,7 +83,7 @@ config RISCV select HAVE_ARCH_THREAD_STRUCT_WHITELIST select HAVE_ARCH_VMAP_STACK if MMU && 64BIT select HAVE_ASM_MODVERSIONS - select HAVE_CONTEXT_TRACKING + select HAVE_CONTEXT_TRACKING_USER select HAVE_DEBUG_KMEMLEAK select HAVE_DMA_CONTIGUOUS if MMU select HAVE_EBPF_JIT if MMU diff --git a/arch/riscv/kernel/entry.S b/arch/riscv/kernel/entry.S index ecd132af2edf..fba900eaa0b8 100644 --- a/arch/riscv/kernel/entry.S +++ b/arch/riscv/kernel/entry.S @@ -111,7 +111,7 @@ _save_context: call __trace_hardirqs_off #endif =20 -#ifdef CONFIG_CONTEXT_TRACKING +#ifdef CONFIG_CONTEXT_TRACKING_USER /* If previous state is in user mode, call user_exit_callable(). */ li a0, SR_PP and a0, s1, a0 @@ -176,7 +176,7 @@ handle_syscall: */ csrs CSR_STATUS, SR_IE #endif -#if defined(CONFIG_TRACE_IRQFLAGS) || defined(CONFIG_CONTEXT_TRACKING) +#if defined(CONFIG_TRACE_IRQFLAGS) || defined(CONFIG_CONTEXT_TRACKING_USER) /* Recover a0 - a7 for system calls */ REG_L a0, PT_A0(sp) REG_L a1, PT_A1(sp) @@ -255,7 +255,7 @@ resume_userspace: andi s1, s0, _TIF_WORK_MASK bnez s1, work_pending =20 -#ifdef CONFIG_CONTEXT_TRACKING +#ifdef CONFIG_CONTEXT_TRACKING_USER call user_enter_callable #endif =20 diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig index 9200bc04701c..be50eae3ab0d 100644 --- a/arch/sparc/Kconfig +++ b/arch/sparc/Kconfig @@ -71,7 +71,7 @@ config SPARC64 select HAVE_DYNAMIC_FTRACE select HAVE_FTRACE_MCOUNT_RECORD select HAVE_SYSCALL_TRACEPOINTS - select HAVE_CONTEXT_TRACKING + select HAVE_CONTEXT_TRACKING_USER select HAVE_TIF_NOHZ select HAVE_DEBUG_KMEMLEAK select IOMMU_HELPER diff --git a/arch/sparc/kernel/rtrap_64.S b/arch/sparc/kernel/rtrap_64.S index c5fd4b450d9b..eef102765a7e 100644 --- a/arch/sparc/kernel/rtrap_64.S +++ b/arch/sparc/kernel/rtrap_64.S @@ -15,7 +15,7 @@ #include #include =20 -#ifdef CONFIG_CONTEXT_TRACKING +#ifdef CONFIG_CONTEXT_TRACKING_USER # define SCHEDULE_USER schedule_user #else # define SCHEDULE_USER schedule diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index b0142e01002e..368a1c5cbb08 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -185,8 +185,8 @@ config X86 select HAVE_ASM_MODVERSIONS select HAVE_CMPXCHG_DOUBLE select HAVE_CMPXCHG_LOCAL - select HAVE_CONTEXT_TRACKING if X86_64 - select HAVE_CONTEXT_TRACKING_OFFSTACK if HAVE_CONTEXT_TRACKING + select HAVE_CONTEXT_TRACKING_USER if X86_64 + select HAVE_CONTEXT_TRACKING_USER_OFFSTACK if HAVE_CONTEXT_TRACKING_USER select HAVE_C_RECORDMCOUNT select HAVE_OBJTOOL_MCOUNT if STACK_VALIDATION select HAVE_BUILDTIME_MCOUNT_SORT diff --git a/include/linux/context_tracking.h b/include/linux/context_track= ing.h index 63259fece7c7..e35ae66b4794 100644 --- a/include/linux/context_tracking.h +++ b/include/linux/context_tracking.h @@ -10,7 +10,7 @@ #include =20 =20 -#ifdef CONFIG_CONTEXT_TRACKING +#ifdef CONFIG_CONTEXT_TRACKING_USER extern void ct_cpu_track_user(int cpu); =20 /* Called with interrupts disabled. */ @@ -52,7 +52,7 @@ static inline enum ctx_state exception_enter(void) { enum ctx_state prev_ctx; =20 - if (IS_ENABLED(CONFIG_HAVE_CONTEXT_TRACKING_OFFSTACK) || + if (IS_ENABLED(CONFIG_HAVE_CONTEXT_TRACKING_USER_OFFSTACK) || !context_tracking_enabled()) return 0; =20 @@ -65,7 +65,7 @@ static inline enum ctx_state exception_enter(void) =20 static inline void exception_exit(enum ctx_state prev_ctx) { - if (!IS_ENABLED(CONFIG_HAVE_CONTEXT_TRACKING_OFFSTACK) && + if (!IS_ENABLED(CONFIG_HAVE_CONTEXT_TRACKING_USER_OFFSTACK) && context_tracking_enabled()) { if (prev_ctx !=3D CONTEXT_KERNEL) ct_user_enter(prev_ctx); @@ -109,14 +109,14 @@ static inline enum ctx_state ct_state(void) { return = CONTEXT_DISABLED; } static __always_inline bool context_tracking_guest_enter(void) { return fa= lse; } static inline void context_tracking_guest_exit(void) { } =20 -#endif /* !CONFIG_CONTEXT_TRACKING */ +#endif /* !CONFIG_CONTEXT_TRACKING_USER */ =20 #define CT_WARN_ON(cond) WARN_ON(context_tracking_enabled() && (cond)) =20 -#ifdef CONFIG_CONTEXT_TRACKING_FORCE +#ifdef CONFIG_CONTEXT_TRACKING_USER_FORCE extern void context_tracking_init(void); #else static inline void context_tracking_init(void) { } -#endif /* CONFIG_CONTEXT_TRACKING_FORCE */ +#endif /* CONFIG_CONTEXT_TRACKING_USER_FORCE */ =20 #endif diff --git a/include/linux/context_tracking_state.h b/include/linux/context= _tracking_state.h index 02be3ac2040d..913acfee5dbf 100644 --- a/include/linux/context_tracking_state.h +++ b/include/linux/context_tracking_state.h @@ -22,7 +22,7 @@ struct context_tracking { } state; }; =20 -#ifdef CONFIG_CONTEXT_TRACKING +#ifdef CONFIG_CONTEXT_TRACKING_USER extern struct static_key_false context_tracking_key; DECLARE_PER_CPU(struct context_tracking, context_tracking); =20 @@ -45,6 +45,6 @@ static inline bool context_tracking_enabled_this_cpu(void) static inline bool context_tracking_enabled(void) { return false; } static inline bool context_tracking_enabled_cpu(int cpu) { return false; } static inline bool context_tracking_enabled_this_cpu(void) { return false;= } -#endif /* CONFIG_CONTEXT_TRACKING */ +#endif /* CONFIG_CONTEXT_TRACKING_USER */ =20 #endif diff --git a/init/Kconfig b/init/Kconfig index ddcbefe535e9..3d88ecadf771 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -502,11 +502,11 @@ config VIRT_CPU_ACCOUNTING_NATIVE =20 config VIRT_CPU_ACCOUNTING_GEN bool "Full dynticks CPU time accounting" - depends on HAVE_CONTEXT_TRACKING + depends on HAVE_CONTEXT_TRACKING_USER depends on HAVE_VIRT_CPU_ACCOUNTING_GEN depends on GENERIC_CLOCKEVENTS select VIRT_CPU_ACCOUNTING - select CONTEXT_TRACKING + select CONTEXT_TRACKING_USER help Select this option to enable task and CPU time accounting on full dynticks systems. This accounting is implemented by watching every diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index ccf18032e742..9e2797d4c41e 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -22,6 +22,8 @@ #include #include =20 +#ifdef CONFIG_CONTEXT_TRACKING_USER + #define CREATE_TRACE_POINTS #include =20 @@ -253,7 +255,7 @@ void __init ct_cpu_track_user(int cpu) initialized =3D true; } =20 -#ifdef CONFIG_CONTEXT_TRACKING_FORCE +#ifdef CONFIG_CONTEXT_TRACKING_USER_FORCE void __init context_tracking_init(void) { int cpu; @@ -262,3 +264,5 @@ void __init context_tracking_init(void) ct_cpu_track_user(cpu); } #endif + +#endif /* #ifdef CONFIG_CONTEXT_TRACKING_USER */ diff --git a/kernel/sched/core.c b/kernel/sched/core.c index c1749f6561ef..a4419ec8bcbe 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -6515,7 +6515,7 @@ void __sched schedule_idle(void) } while (need_resched()); } =20 -#if defined(CONFIG_CONTEXT_TRACKING) && !defined(CONFIG_HAVE_CONTEXT_TRACK= ING_OFFSTACK) +#if defined(CONFIG_CONTEXT_TRACKING_USER) && !defined(CONFIG_HAVE_CONTEXT_= TRACKING_USER_OFFSTACK) asmlinkage __visible void __sched schedule_user(void) { /* diff --git a/kernel/time/Kconfig b/kernel/time/Kconfig index 27b7868b5c30..41f99bcfe9e6 100644 --- a/kernel/time/Kconfig +++ b/kernel/time/Kconfig @@ -73,6 +73,9 @@ config TIME_KUNIT_TEST =20 If unsure, say N. =20 +config CONTEXT_TRACKING + bool + if GENERIC_CLOCKEVENTS menu "Timers subsystem" =20 @@ -111,7 +114,7 @@ config NO_HZ_FULL # NO_HZ_COMMON dependency # We need at least one periodic CPU for timekeeping depends on SMP - depends on HAVE_CONTEXT_TRACKING + depends on HAVE_CONTEXT_TRACKING_USER # VIRT_CPU_ACCOUNTING_GEN dependency depends on HAVE_VIRT_CPU_ACCOUNTING_GEN select NO_HZ_COMMON @@ -137,31 +140,37 @@ config NO_HZ_FULL =20 endchoice =20 -config CONTEXT_TRACKING - bool +config CONTEXT_TRACKING_USER + bool + depends on HAVE_CONTEXT_TRACKING_USER + select CONTEXT_TRACKING + help + Track transitions between kernel and user on behalf of RCU and + tickless cputime accounting. The former case relies on context + tracking to enter/exit RCU extended quiescent states. =20 -config CONTEXT_TRACKING_FORCE - bool "Force context tracking" - depends on CONTEXT_TRACKING +config CONTEXT_TRACKING_USER_FORCE + bool "Force user context tracking" + depends on CONTEXT_TRACKING_USER default y if !NO_HZ_FULL help The major pre-requirement for full dynticks to work is to - support the context tracking subsystem. But there are also + support the user context tracking subsystem. But there are also other dependencies to provide in order to make the full dynticks working. =20 This option stands for testing when an arch implements the - context tracking backend but doesn't yet fulfill all the + user context tracking backend but doesn't yet fulfill all the requirements to make the full dynticks feature working. Without the full dynticks, there is no way to test the support - for context tracking and the subsystems that rely on it: RCU + for user context tracking and the subsystems that rely on it: RCU userspace extended quiescent state and tickless cputime accounting. This option copes with the absence of the full - dynticks subsystem by forcing the context tracking on all + dynticks subsystem by forcing the user context tracking on all CPUs in the system. =20 Say Y only if you're working on the development of an - architecture backend for the context tracking. + architecture backend for the user context tracking. =20 Say N otherwise, this option brings an overhead that you don't want in production. --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 5C197C433EF for ; Wed, 8 Jun 2022 14:41:58 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S241811AbiFHOlz (ORCPT ); Wed, 8 Jun 2022 10:41:55 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:58068 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241737AbiFHOlV (ORCPT ); Wed, 8 Jun 2022 10:41:21 -0400 Received: from ams.source.kernel.org (ams.source.kernel.org [145.40.68.75]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id A75952AC79 for ; Wed, 8 Jun 2022 07:41:16 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ams.source.kernel.org (Postfix) with ESMTPS id D5661B8284C for ; Wed, 8 Jun 2022 14:41:14 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 67C4DC3411F; Wed, 8 Jun 2022 14:41:10 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699273; bh=M+s3v9ycXzf6D6WU9c1RKDgRGC1CzXreKpFrkqISrYY=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=a8MqeWggb9h1HrzUGdpHw6L/fqXwhkLfw7Iqh9VD5vLt4Nm0odKV7G37rqiR/UXNE BAXthPMN7vkQ3kZybmOEXEEmXWX90zAbRdUHVTnYztyggkLu22Crdryv5aXKIapTNK zddN4+U+KsOLSgyuxs4yd5shpmNLaj7+nTp+DkVESjzgEXTZrGuDSUKsARUTe6xtSg lvDcMxk7sMzKjf5sS5pfbIecDNHUJGgzOU25dHzF+pKTMyvdohzY/noMTN/JRKYUpU YXc5X1H5bYXmlyIESAXZdu2vVOHtZEBVBRdayYCLG/XapiS5t2uPGWiZa9Gf7g+RfI bIlCAVUwp9LQQ== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 08/20] context_tracking: Take idle eqs entrypoints over RCU Date: Wed, 8 Jun 2022 16:40:25 +0200 Message-Id: <20220608144037.1765000-9-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" The RCU dynticks counter is going to be merged into the context tracking subsystem. Start with moving the idle extended quiescent states entrypoints to context tracking. For now those are dumb redirections to existing RCU calls. Signed-off-by: Frederic Weisbecker Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- Documentation/RCU/stallwarn.rst | 4 ++-- arch/arm/mach-imx/cpuidle-imx6q.c | 5 +++-- drivers/acpi/processor_idle.c | 5 +++-- drivers/cpuidle/cpuidle.c | 9 +++++---- include/linux/context_tracking.h | 8 ++++++++ include/linux/rcupdate.h | 2 +- kernel/context_tracking.c | 15 +++++++++++++++ kernel/locking/lockdep.c | 2 +- kernel/rcu/Kconfig | 2 ++ kernel/rcu/tree.c | 2 -- kernel/rcu/update.c | 2 +- kernel/sched/idle.c | 10 +++++----- kernel/sched/sched.h | 1 + kernel/time/Kconfig | 6 ++++++ 14 files changed, 53 insertions(+), 20 deletions(-) diff --git a/Documentation/RCU/stallwarn.rst b/Documentation/RCU/stallwarn.= rst index 794837eb519b..b95bda7755fa 100644 --- a/Documentation/RCU/stallwarn.rst +++ b/Documentation/RCU/stallwarn.rst @@ -97,8 +97,8 @@ warnings: which will include additional debugging information. =20 - A low-level kernel issue that either fails to invoke one of the - variants of rcu_user_enter(), rcu_user_exit(), rcu_idle_enter(), - rcu_idle_exit(), rcu_irq_enter(), or rcu_irq_exit() on the one + variants of rcu_user_enter(), rcu_user_exit(), ct_idle_enter(), + ct_idle_exit(), rcu_irq_enter(), or rcu_irq_exit() on the one hand, or that invokes one of them too many times on the other. Historically, the most frequent issue has been an omission of either irq_enter() or irq_exit(), which in turn invoke diff --git a/arch/arm/mach-imx/cpuidle-imx6q.c b/arch/arm/mach-imx/cpuidle-= imx6q.c index 094337dc1bc7..d086cbae09c3 100644 --- a/arch/arm/mach-imx/cpuidle-imx6q.c +++ b/arch/arm/mach-imx/cpuidle-imx6q.c @@ -3,6 +3,7 @@ * Copyright (C) 2012 Freescale Semiconductor, Inc. */ =20 +#include #include #include #include @@ -24,9 +25,9 @@ static int imx6q_enter_wait(struct cpuidle_device *dev, imx6_set_lpm(WAIT_UNCLOCKED); raw_spin_unlock(&cpuidle_lock); =20 - rcu_idle_enter(); + ct_idle_enter(); cpu_do_idle(); - rcu_idle_exit(); + ct_idle_exit(); =20 raw_spin_lock(&cpuidle_lock); if (num_idle_cpus-- =3D=3D num_online_cpus()) diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c index 32b20efff5f8..935f4113d5f6 100644 --- a/drivers/acpi/processor_idle.c +++ b/drivers/acpi/processor_idle.c @@ -22,6 +22,7 @@ #include #include #include +#include =20 /* * Include the apic definitions for x86 to have the APIC timer related def= ines @@ -648,11 +649,11 @@ static int acpi_idle_enter_bm(struct cpuidle_driver *= drv, raw_spin_unlock(&c3_lock); } =20 - rcu_idle_enter(); + ct_idle_enter(); =20 acpi_idle_do_entry(cx); =20 - rcu_idle_exit(); + ct_idle_exit(); =20 /* Re-enable bus master arbitration */ if (dis_bm) { diff --git a/drivers/cpuidle/cpuidle.c b/drivers/cpuidle/cpuidle.c index ef2ea1b12cd8..62dd956025f3 100644 --- a/drivers/cpuidle/cpuidle.c +++ b/drivers/cpuidle/cpuidle.c @@ -23,6 +23,7 @@ #include #include #include +#include #include =20 #include "cpuidle.h" @@ -150,12 +151,12 @@ static void enter_s2idle_proper(struct cpuidle_driver= *drv, */ stop_critical_timings(); if (!(target_state->flags & CPUIDLE_FLAG_RCU_IDLE)) - rcu_idle_enter(); + ct_idle_enter(); target_state->enter_s2idle(dev, drv, index); if (WARN_ON_ONCE(!irqs_disabled())) local_irq_disable(); if (!(target_state->flags & CPUIDLE_FLAG_RCU_IDLE)) - rcu_idle_exit(); + ct_idle_exit(); tick_unfreeze(); start_critical_timings(); =20 @@ -233,10 +234,10 @@ int cpuidle_enter_state(struct cpuidle_device *dev, s= truct cpuidle_driver *drv, =20 stop_critical_timings(); if (!(target_state->flags & CPUIDLE_FLAG_RCU_IDLE)) - rcu_idle_enter(); + ct_idle_enter(); entered_state =3D target_state->enter(dev, drv, index); if (!(target_state->flags & CPUIDLE_FLAG_RCU_IDLE)) - rcu_idle_exit(); + ct_idle_exit(); start_critical_timings(); =20 sched_clock_idle_wakeup_event(); diff --git a/include/linux/context_tracking.h b/include/linux/context_track= ing.h index e35ae66b4794..01abadb2f993 100644 --- a/include/linux/context_tracking.h +++ b/include/linux/context_tracking.h @@ -119,4 +119,12 @@ extern void context_tracking_init(void); static inline void context_tracking_init(void) { } #endif /* CONFIG_CONTEXT_TRACKING_USER_FORCE */ =20 +#ifdef CONFIG_CONTEXT_TRACKING_IDLE +extern void ct_idle_enter(void); +extern void ct_idle_exit(void); +#else +static inline void ct_idle_enter(void) { } +static inline void ct_idle_exit(void) { } +#endif /* !CONFIG_CONTEXT_TRACKING_IDLE */ + #endif diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h index 08059d8d4f5a..2a76f1a6399c 100644 --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -129,7 +129,7 @@ static inline void rcu_nocb_flush_deferred_wakeup(void)= { } * @a: Code that RCU needs to pay attention to. * * RCU read-side critical sections are forbidden in the inner idle loop, - * that is, between the rcu_idle_enter() and the rcu_idle_exit() -- RCU + * that is, between the ct_idle_enter() and the ct_idle_exit() -- RCU * will happily ignore any such read-side critical sections. However, * things like powertop need tracepoints in the inner idle loop. * diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index 9e2797d4c41e..7289b097434d 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -22,6 +22,21 @@ #include #include =20 + +#ifdef CONFIG_CONTEXT_TRACKING_IDLE +noinstr void ct_idle_enter(void) +{ + rcu_idle_enter(); +} +EXPORT_SYMBOL_GPL(ct_idle_enter); + +void ct_idle_exit(void) +{ + rcu_idle_exit(); +} +EXPORT_SYMBOL_GPL(ct_idle_exit); +#endif /* #ifdef CONFIG_CONTEXT_TRACKING_IDLE */ + #ifdef CONFIG_CONTEXT_TRACKING_USER =20 #define CREATE_TRACE_POINTS diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c index c06cab6546ed..5f0dfe37234b 100644 --- a/kernel/locking/lockdep.c +++ b/kernel/locking/lockdep.c @@ -6546,7 +6546,7 @@ void lockdep_rcu_suspicious(const char *file, const i= nt line, const char *s) =20 /* * If a CPU is in the RCU-free window in idle (ie: in the section - * between rcu_idle_enter() and rcu_idle_exit(), then RCU + * between ct_idle_enter() and ct_idle_exit(), then RCU * considers that CPU to be in an "extended quiescent state", * which means that RCU will be completely ignoring that CPU. * Therefore, rcu_read_lock() and friends have absolutely no diff --git a/kernel/rcu/Kconfig b/kernel/rcu/Kconfig index c05ca52cdf64..d471d22a5e21 100644 --- a/kernel/rcu/Kconfig +++ b/kernel/rcu/Kconfig @@ -8,6 +8,8 @@ menu "RCU Subsystem" config TREE_RCU bool default y if SMP + # Dynticks-idle tracking + select CONTEXT_TRACKING_IDLE help This option selects the RCU implementation that is designed for very large SMP system with hundreds or diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index 6a01a48fe56d..d4ac5538febb 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -668,7 +668,6 @@ void noinstr rcu_idle_enter(void) WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !raw_irqs_disabled()); rcu_eqs_enter(false); } -EXPORT_SYMBOL_GPL(rcu_idle_enter); =20 #ifdef CONFIG_NO_HZ_FULL =20 @@ -908,7 +907,6 @@ void noinstr rcu_idle_exit(void) rcu_eqs_exit(false); raw_local_irq_restore(flags); } -EXPORT_SYMBOL_GPL(rcu_idle_exit); =20 #ifdef CONFIG_NO_HZ_FULL /** diff --git a/kernel/rcu/update.c b/kernel/rcu/update.c index 2e93acad1e31..738842c4886b 100644 --- a/kernel/rcu/update.c +++ b/kernel/rcu/update.c @@ -85,7 +85,7 @@ module_param(rcu_normal_after_boot, int, 0444); * and while lockdep is disabled. * * Note that if the CPU is in the idle loop from an RCU point of view (ie: - * that we are in the section between rcu_idle_enter() and rcu_idle_exit()) + * that we are in the section between ct_idle_enter() and ct_idle_exit()) * then rcu_read_lock_held() sets ``*ret`` to false even if the CPU did an * rcu_read_lock(). The reason for this is that RCU ignores CPUs that are * in such a section, considering these as in extended quiescent state, diff --git a/kernel/sched/idle.c b/kernel/sched/idle.c index 8f8b5020e76a..6de222b23b49 100644 --- a/kernel/sched/idle.c +++ b/kernel/sched/idle.c @@ -53,14 +53,14 @@ static noinline int __cpuidle cpu_idle_poll(void) { trace_cpu_idle(0, smp_processor_id()); stop_critical_timings(); - rcu_idle_enter(); + ct_idle_enter(); local_irq_enable(); =20 while (!tif_need_resched() && (cpu_idle_force_poll || tick_check_broadcast_expired())) cpu_relax(); =20 - rcu_idle_exit(); + ct_idle_exit(); start_critical_timings(); trace_cpu_idle(PWR_EVENT_EXIT, smp_processor_id()); =20 @@ -98,12 +98,12 @@ void __cpuidle default_idle_call(void) * * Trace IRQs enable here, then switch off RCU, and have * arch_cpu_idle() use raw_local_irq_enable(). Note that - * rcu_idle_enter() relies on lockdep IRQ state, so switch that + * ct_idle_enter() relies on lockdep IRQ state, so switch that * last -- this is very similar to the entry code. */ trace_hardirqs_on_prepare(); lockdep_hardirqs_on_prepare(_THIS_IP_); - rcu_idle_enter(); + ct_idle_enter(); lockdep_hardirqs_on(_THIS_IP_); =20 arch_cpu_idle(); @@ -116,7 +116,7 @@ void __cpuidle default_idle_call(void) */ raw_local_irq_disable(); lockdep_hardirqs_off(_THIS_IP_); - rcu_idle_exit(); + ct_idle_exit(); lockdep_hardirqs_on(_THIS_IP_); raw_local_irq_enable(); =20 diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index 58263f90c559..f398f2bf05f4 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -27,6 +27,7 @@ #include #include #include +#include #include #include #include diff --git a/kernel/time/Kconfig b/kernel/time/Kconfig index 41f99bcfe9e6..a41753be1a2b 100644 --- a/kernel/time/Kconfig +++ b/kernel/time/Kconfig @@ -76,6 +76,12 @@ config TIME_KUNIT_TEST config CONTEXT_TRACKING bool =20 +config CONTEXT_TRACKING_IDLE + bool + select CONTEXT_TRACKING + help + Tracks idle state on behalf of RCU. + if GENERIC_CLOCKEVENTS menu "Timers subsystem" =20 --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id C00E9C43334 for ; Wed, 8 Jun 2022 14:42:05 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S241568AbiFHOmD (ORCPT ); Wed, 8 Jun 2022 10:42:03 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:58628 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241674AbiFHOl0 (ORCPT ); Wed, 8 Jun 2022 10:41:26 -0400 Received: from dfw.source.kernel.org (dfw.source.kernel.org [IPv6:2604:1380:4641:c500::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 8A6EC3EF34 for ; Wed, 8 Jun 2022 07:41:18 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by dfw.source.kernel.org (Postfix) with ESMTPS id DCEB861B9A for ; Wed, 8 Jun 2022 14:41:17 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 189BAC341C4; Wed, 8 Jun 2022 14:41:13 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699277; bh=gwMTXfFYn/xzL3xzYXsnXSJ+GmOncAWlD+F56x4qE88=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=XNMYOwHfFbgKe9ZuRZEyHdzNm+U0XDoFEh3LvBqK6xEboFXg97kbxb1bzpfuj3gh/ ZYh1U81PtzIunvpsO8TO2oCYou7HDLniDRFM+ldW0NlPIMO/SDquGDCz+EAC8IlBiz YQi7AElWAk7Eu57VFcDfHyLgmJx0boI3bkaHo82zJxcxwIHoRdlxIfTmk35QwlvhVQ 95yaXznECv6n4ZvMHZm7vj/mFPWc1V7kofRy6k+SGJt4/lY9jWWtQSBX0PKtFy970a z1Aypc7NUR6kJl8XrQ2sO/a+WgpONUmvy9Tv2eyvb8nQ0AmpxULUgryY52thpX+UA3 aMt6xZDizdPqA== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 09/20] context_tracking: Take IRQ eqs entrypoints over RCU Date: Wed, 8 Jun 2022 16:40:26 +0200 Message-Id: <20220608144037.1765000-10-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" The RCU dynticks counter is going to be merged into the context tracking subsystem. Prepare with moving the IRQ extended quiescent states entrypoints to context tracking. For now those are dumb redirection to existing RCU calls. Acked-by: Paul E. McKenney Signed-off-by: Frederic Weisbecker Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- .../RCU/Design/Requirements/Requirements.rst | 10 ++++---- Documentation/RCU/stallwarn.rst | 4 ++-- arch/Kconfig | 2 +- arch/arm64/kernel/entry-common.c | 6 ++--- arch/x86/mm/fault.c | 2 +- drivers/cpuidle/cpuidle-psci.c | 8 +++---- drivers/cpuidle/cpuidle-riscv-sbi.c | 8 +++---- include/linux/context_tracking_irq.h | 17 +++++++++++++ include/linux/context_tracking_state.h | 1 + include/linux/entry-common.h | 10 ++++---- include/linux/rcupdate.h | 5 ++-- include/linux/tracepoint.h | 4 ++-- kernel/context_tracking.c | 24 +++++++++++++++++-- kernel/cpu_pm.c | 8 +++---- kernel/entry/common.c | 12 +++++----- kernel/softirq.c | 4 ++-- kernel/trace/trace.c | 6 ++--- 17 files changed, 85 insertions(+), 46 deletions(-) create mode 100644 include/linux/context_tracking_irq.h diff --git a/Documentation/RCU/Design/Requirements/Requirements.rst b/Docum= entation/RCU/Design/Requirements/Requirements.rst index 04ed8bf27a0e..2a657fd2bca3 100644 --- a/Documentation/RCU/Design/Requirements/Requirements.rst +++ b/Documentation/RCU/Design/Requirements/Requirements.rst @@ -1844,10 +1844,10 @@ that meets this requirement. =20 Furthermore, NMI handlers can be interrupted by what appear to RCU to be normal interrupts. One way that this can happen is for code that -directly invokes rcu_irq_enter() and rcu_irq_exit() to be called +directly invokes ct_irq_enter() and ct_irq_exit() to be called from an NMI handler. This astonishing fact of life prompted the current -code structure, which has rcu_irq_enter() invoking -rcu_nmi_enter() and rcu_irq_exit() invoking rcu_nmi_exit(). +code structure, which has ct_irq_enter() invoking +rcu_nmi_enter() and ct_irq_exit() invoking rcu_nmi_exit(). And yes, I also learned of this requirement the hard way. =20 Loadable Modules @@ -2195,7 +2195,7 @@ scheduling-clock interrupt be enabled when RCU needs = it to be: sections, and RCU believes this CPU to be idle, no problem. This sort of thing is used by some architectures for light-weight exception handlers, which can then avoid the overhead of - rcu_irq_enter() and rcu_irq_exit() at exception entry and + ct_irq_enter() and ct_irq_exit() at exception entry and exit, respectively. Some go further and avoid the entireties of irq_enter() and irq_exit(). Just make very sure you are running some of your tests with @@ -2226,7 +2226,7 @@ scheduling-clock interrupt be enabled when RCU needs = it to be: +-----------------------------------------------------------------------+ | **Answer**: | +-----------------------------------------------------------------------+ -| One approach is to do ``rcu_irq_exit();rcu_irq_enter();`` every so | +| One approach is to do ``ct_irq_exit();ct_irq_enter();`` every so | | often. But given that long-running interrupt handlers can cause other | | problems, not least for response time, shouldn't you work to keep | | your interrupt handler's runtime within reasonable bounds? | diff --git a/Documentation/RCU/stallwarn.rst b/Documentation/RCU/stallwarn.= rst index b95bda7755fa..ce1f58a9d954 100644 --- a/Documentation/RCU/stallwarn.rst +++ b/Documentation/RCU/stallwarn.rst @@ -98,11 +98,11 @@ warnings: =20 - A low-level kernel issue that either fails to invoke one of the variants of rcu_user_enter(), rcu_user_exit(), ct_idle_enter(), - ct_idle_exit(), rcu_irq_enter(), or rcu_irq_exit() on the one + ct_idle_exit(), ct_irq_enter(), or ct_irq_exit() on the one hand, or that invokes one of them too many times on the other. Historically, the most frequent issue has been an omission of either irq_enter() or irq_exit(), which in turn invoke - rcu_irq_enter() or rcu_irq_exit(), respectively. Building your + ct_irq_enter() or ct_irq_exit(), respectively. Building your kernel with CONFIG_RCU_EQS_DEBUG=3Dy can help track down these types of issues, which sometimes arise in architecture-specific code. =20 diff --git a/arch/Kconfig b/arch/Kconfig index 7a8cd4d67d9c..888f9725b20b 100644 --- a/arch/Kconfig +++ b/arch/Kconfig @@ -777,7 +777,7 @@ config HAVE_CONTEXT_TRACKING_USER Syscalls need to be wrapped inside user_exit()-user_enter(), either optimized behind static key or through the slow path using TIF_NOHZ flag. Exceptions handlers must be wrapped as well. Irqs are already - protected inside rcu_irq_enter/rcu_irq_exit() but preemption or signal + protected inside ct_irq_enter/ct_irq_exit() but preemption or signal handling on irq exit still need to be protected. =20 config HAVE_CONTEXT_TRACKING_USER_OFFSTACK diff --git a/arch/arm64/kernel/entry-common.c b/arch/arm64/kernel/entry-com= mon.c index 878c65aa7206..a1490e8bf5d4 100644 --- a/arch/arm64/kernel/entry-common.c +++ b/arch/arm64/kernel/entry-common.c @@ -41,7 +41,7 @@ static __always_inline void __enter_from_kernel_mode(stru= ct pt_regs *regs) =20 if (!IS_ENABLED(CONFIG_TINY_RCU) && is_idle_task(current)) { lockdep_hardirqs_off(CALLER_ADDR0); - rcu_irq_enter(); + ct_irq_enter(); trace_hardirqs_off_finish(); =20 regs->exit_rcu =3D true; @@ -76,7 +76,7 @@ static __always_inline void __exit_to_kernel_mode(struct = pt_regs *regs) if (regs->exit_rcu) { trace_hardirqs_on_prepare(); lockdep_hardirqs_on_prepare(CALLER_ADDR0); - rcu_irq_exit(); + ct_irq_exit(); lockdep_hardirqs_on(CALLER_ADDR0); return; } @@ -84,7 +84,7 @@ static __always_inline void __exit_to_kernel_mode(struct = pt_regs *regs) trace_hardirqs_on(); } else { if (regs->exit_rcu) - rcu_irq_exit(); + ct_irq_exit(); } } =20 diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c index d0074c6ed31a..b781785b1ff3 100644 --- a/arch/x86/mm/fault.c +++ b/arch/x86/mm/fault.c @@ -1526,7 +1526,7 @@ DEFINE_IDTENTRY_RAW_ERRORCODE(exc_page_fault) =20 /* * Entry handling for valid #PF from kernel mode is slightly - * different: RCU is already watching and rcu_irq_enter() must not + * different: RCU is already watching and ct_irq_enter() must not * be invoked because a kernel fault on a user space address might * sleep. * diff --git a/drivers/cpuidle/cpuidle-psci.c b/drivers/cpuidle/cpuidle-psci.c index b51b5df08450..fe31b2d522b3 100644 --- a/drivers/cpuidle/cpuidle-psci.c +++ b/drivers/cpuidle/cpuidle-psci.c @@ -68,12 +68,12 @@ static int __psci_enter_domain_idle_state(struct cpuidl= e_device *dev, return -1; =20 /* Do runtime PM to manage a hierarchical CPU toplogy. */ - rcu_irq_enter_irqson(); + ct_irq_enter_irqson(); if (s2idle) dev_pm_genpd_suspend(pd_dev); else pm_runtime_put_sync_suspend(pd_dev); - rcu_irq_exit_irqson(); + ct_irq_exit_irqson(); =20 state =3D psci_get_domain_state(); if (!state) @@ -81,12 +81,12 @@ static int __psci_enter_domain_idle_state(struct cpuidl= e_device *dev, =20 ret =3D psci_cpu_suspend_enter(state) ? -1 : idx; =20 - rcu_irq_enter_irqson(); + ct_irq_enter_irqson(); if (s2idle) dev_pm_genpd_resume(pd_dev); else pm_runtime_get_sync(pd_dev); - rcu_irq_exit_irqson(); + ct_irq_exit_irqson(); =20 cpu_pm_exit(); =20 diff --git a/drivers/cpuidle/cpuidle-riscv-sbi.c b/drivers/cpuidle/cpuidle-= riscv-sbi.c index b459eda2cd37..759603181d53 100644 --- a/drivers/cpuidle/cpuidle-riscv-sbi.c +++ b/drivers/cpuidle/cpuidle-riscv-sbi.c @@ -115,12 +115,12 @@ static int __sbi_enter_domain_idle_state(struct cpuid= le_device *dev, return -1; =20 /* Do runtime PM to manage a hierarchical CPU toplogy. */ - rcu_irq_enter_irqson(); + ct_irq_enter_irqson(); if (s2idle) dev_pm_genpd_suspend(pd_dev); else pm_runtime_put_sync_suspend(pd_dev); - rcu_irq_exit_irqson(); + ct_irq_exit_irqson(); =20 if (sbi_is_domain_state_available()) state =3D sbi_get_domain_state(); @@ -129,12 +129,12 @@ static int __sbi_enter_domain_idle_state(struct cpuid= le_device *dev, =20 ret =3D sbi_suspend(state) ? -1 : idx; =20 - rcu_irq_enter_irqson(); + ct_irq_enter_irqson(); if (s2idle) dev_pm_genpd_resume(pd_dev); else pm_runtime_get_sync(pd_dev); - rcu_irq_exit_irqson(); + ct_irq_exit_irqson(); =20 cpu_pm_exit(); =20 diff --git a/include/linux/context_tracking_irq.h b/include/linux/context_t= racking_irq.h new file mode 100644 index 000000000000..62f62bbd1a50 --- /dev/null +++ b/include/linux/context_tracking_irq.h @@ -0,0 +1,17 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _LINUX_CONTEXT_TRACKING_IRQ_H +#define _LINUX_CONTEXT_TRACKING_IRQ_H + +#ifdef CONFIG_CONTEXT_TRACKING_IDLE +void ct_irq_enter(void); +void ct_irq_exit(void); +void ct_irq_enter_irqson(void); +void ct_irq_exit_irqson(void); +#else +static inline void ct_irq_enter(void) { } +static inline void ct_irq_exit(void) { } +static inline void ct_irq_enter_irqson(void) { } +static inline void ct_irq_exit_irqson(void) { } +#endif + +#endif diff --git a/include/linux/context_tracking_state.h b/include/linux/context= _tracking_state.h index 913acfee5dbf..7b46b43b8c98 100644 --- a/include/linux/context_tracking_state.h +++ b/include/linux/context_tracking_state.h @@ -4,6 +4,7 @@ =20 #include #include +#include =20 struct context_tracking { /* diff --git a/include/linux/entry-common.h b/include/linux/entry-common.h index ab78bd4c2eb0..2d4cf0b5fe1c 100644 --- a/include/linux/entry-common.h +++ b/include/linux/entry-common.h @@ -357,7 +357,7 @@ void irqentry_exit_to_user_mode(struct pt_regs *regs); /** * struct irqentry_state - Opaque object for exception state storage * @exit_rcu: Used exclusively in the irqentry_*() calls; signals whether = the - * exit path has to invoke rcu_irq_exit(). + * exit path has to invoke ct_irq_exit(). * @lockdep: Used exclusively in the irqentry_nmi_*() calls; ensures that * lockdep state is restored correctly on exit from nmi. * @@ -395,12 +395,12 @@ typedef struct irqentry_state { * * For kernel mode entries RCU handling is done conditional. If RCU is * watching then the only RCU requirement is to check whether the tick has - * to be restarted. If RCU is not watching then rcu_irq_enter() has to be - * invoked on entry and rcu_irq_exit() on exit. + * to be restarted. If RCU is not watching then ct_irq_enter() has to be + * invoked on entry and ct_irq_exit() on exit. * - * Avoiding the rcu_irq_enter/exit() calls is an optimization but also + * Avoiding the ct_irq_enter/exit() calls is an optimization but also * solves the problem of kernel mode pagefaults which can schedule, which - * is not possible after invoking rcu_irq_enter() without undoing it. + * is not possible after invoking ct_irq_enter() without undoing it. * * For user mode entries irqentry_enter_from_user_mode() is invoked to * establish the proper context for NOHZ_FULL. Otherwise scheduling on exit diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h index 2a76f1a6399c..910328f2f51a 100644 --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -29,6 +29,7 @@ #include #include #include +#include =20 #define ULONG_CMP_GE(a, b) (ULONG_MAX / 2 >=3D (a) - (b)) #define ULONG_CMP_LT(a, b) (ULONG_MAX / 2 < (a) - (b)) @@ -144,9 +145,9 @@ static inline void rcu_nocb_flush_deferred_wakeup(void)= { } */ #define RCU_NONIDLE(a) \ do { \ - rcu_irq_enter_irqson(); \ + ct_irq_enter_irqson(); \ do { a; } while (0); \ - rcu_irq_exit_irqson(); \ + ct_irq_exit_irqson(); \ } while (0) =20 /* diff --git a/include/linux/tracepoint.h b/include/linux/tracepoint.h index 28031b15f878..55717a2eda08 100644 --- a/include/linux/tracepoint.h +++ b/include/linux/tracepoint.h @@ -200,13 +200,13 @@ static inline struct tracepoint *tracepoint_ptr_deref= (tracepoint_ptr_t *p) */ \ if (rcuidle) { \ __idx =3D srcu_read_lock_notrace(&tracepoint_srcu);\ - rcu_irq_enter_irqson(); \ + ct_irq_enter_irqson(); \ } \ \ __DO_TRACE_CALL(name, TP_ARGS(args)); \ \ if (rcuidle) { \ - rcu_irq_exit_irqson(); \ + ct_irq_exit_irqson(); \ srcu_read_unlock_notrace(&tracepoint_srcu, __idx);\ } \ \ diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index 7289b097434d..5845ddd4c88c 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -35,6 +35,26 @@ void ct_idle_exit(void) rcu_idle_exit(); } EXPORT_SYMBOL_GPL(ct_idle_exit); + +noinstr void ct_irq_enter(void) +{ + rcu_irq_enter(); +} + +noinstr void ct_irq_exit(void) +{ + rcu_irq_exit(); +} + +void ct_irq_enter_irqson(void) +{ + rcu_irq_enter_irqson(); +} + +void ct_irq_exit_irqson(void) +{ + rcu_irq_exit_irqson(); +} #endif /* #ifdef CONFIG_CONTEXT_TRACKING_IDLE */ =20 #ifdef CONFIG_CONTEXT_TRACKING_USER @@ -90,7 +110,7 @@ void noinstr __ct_user_enter(enum ctx_state state) * At this stage, only low level arch entry code remains and * then we'll run in userspace. We can assume there won't be * any RCU read-side critical section until the next call to - * user_exit() or rcu_irq_enter(). Let's remove RCU's dependency + * user_exit() or ct_irq_enter(). Let's remove RCU's dependency * on the tick. */ if (state =3D=3D CONTEXT_USER) { @@ -137,7 +157,7 @@ void ct_user_enter(enum ctx_state state) /* * Some contexts may involve an exception occuring in an irq, * leading to that nesting: - * rcu_irq_enter() rcu_user_exit() rcu_user_exit() rcu_irq_exit() + * ct_irq_enter() rcu_user_exit() rcu_user_exit() ct_irq_exit() * This would mess up the dyntick_nesting count though. And rcu_irq_*() * helpers are enough to protect RCU uses inside the exception. So * just return immediately if we detect we are in an IRQ. diff --git a/kernel/cpu_pm.c b/kernel/cpu_pm.c index 246efc74e3f3..ba4ba71facf9 100644 --- a/kernel/cpu_pm.c +++ b/kernel/cpu_pm.c @@ -35,11 +35,11 @@ static int cpu_pm_notify(enum cpu_pm_event event) * disfunctional in cpu idle. Copy RCU_NONIDLE code to let RCU know * this. */ - rcu_irq_enter_irqson(); + ct_irq_enter_irqson(); rcu_read_lock(); ret =3D raw_notifier_call_chain(&cpu_pm_notifier.chain, event, NULL); rcu_read_unlock(); - rcu_irq_exit_irqson(); + ct_irq_exit_irqson(); =20 return notifier_to_errno(ret); } @@ -49,11 +49,11 @@ static int cpu_pm_notify_robust(enum cpu_pm_event event= _up, enum cpu_pm_event ev unsigned long flags; int ret; =20 - rcu_irq_enter_irqson(); + ct_irq_enter_irqson(); raw_spin_lock_irqsave(&cpu_pm_notifier.lock, flags); ret =3D raw_notifier_call_chain_robust(&cpu_pm_notifier.chain, event_up, = event_down, NULL); raw_spin_unlock_irqrestore(&cpu_pm_notifier.lock, flags); - rcu_irq_exit_irqson(); + ct_irq_exit_irqson(); =20 return notifier_to_errno(ret); } diff --git a/kernel/entry/common.c b/kernel/entry/common.c index e57a224d6b79..9a89386696ba 100644 --- a/kernel/entry/common.c +++ b/kernel/entry/common.c @@ -321,7 +321,7 @@ noinstr irqentry_state_t irqentry_enter(struct pt_regs = *regs) } =20 /* - * If this entry hit the idle task invoke rcu_irq_enter() whether + * If this entry hit the idle task invoke ct_irq_enter() whether * RCU is watching or not. * * Interrupts can nest when the first interrupt invokes softirq @@ -332,12 +332,12 @@ noinstr irqentry_state_t irqentry_enter(struct pt_reg= s *regs) * not nested into another interrupt. * * Checking for rcu_is_watching() here would prevent the nesting - * interrupt to invoke rcu_irq_enter(). If that nested interrupt is + * interrupt to invoke ct_irq_enter(). If that nested interrupt is * the tick then rcu_flavor_sched_clock_irq() would wrongfully * assume that it is the first interrupt and eventually claim * quiescent state and end grace periods prematurely. * - * Unconditionally invoke rcu_irq_enter() so RCU state stays + * Unconditionally invoke ct_irq_enter() so RCU state stays * consistent. * * TINY_RCU does not support EQS, so let the compiler eliminate @@ -350,7 +350,7 @@ noinstr irqentry_state_t irqentry_enter(struct pt_regs = *regs) * as in irqentry_enter_from_user_mode(). */ lockdep_hardirqs_off(CALLER_ADDR0); - rcu_irq_enter(); + ct_irq_enter(); instrumentation_begin(); trace_hardirqs_off_finish(); instrumentation_end(); @@ -418,7 +418,7 @@ noinstr void irqentry_exit(struct pt_regs *regs, irqent= ry_state_t state) trace_hardirqs_on_prepare(); lockdep_hardirqs_on_prepare(CALLER_ADDR0); instrumentation_end(); - rcu_irq_exit(); + ct_irq_exit(); lockdep_hardirqs_on(CALLER_ADDR0); return; } @@ -436,7 +436,7 @@ noinstr void irqentry_exit(struct pt_regs *regs, irqent= ry_state_t state) * was not watching on entry. */ if (state.exit_rcu) - rcu_irq_exit(); + ct_irq_exit(); } } =20 diff --git a/kernel/softirq.c b/kernel/softirq.c index fac801815554..7db9db832b31 100644 --- a/kernel/softirq.c +++ b/kernel/softirq.c @@ -607,7 +607,7 @@ void irq_enter_rcu(void) */ void irq_enter(void) { - rcu_irq_enter(); + ct_irq_enter(); irq_enter_rcu(); } =20 @@ -659,7 +659,7 @@ void irq_exit_rcu(void) void irq_exit(void) { __irq_exit_rcu(); - rcu_irq_exit(); + ct_irq_exit(); /* must be last! */ lockdep_hardirq_exit(); } diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index f4de111fa18f..85e519cc0af9 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -3104,15 +3104,15 @@ void __trace_stack(struct trace_array *tr, unsigned= int trace_ctx, /* * When an NMI triggers, RCU is enabled via rcu_nmi_enter(), * but if the above rcu_is_watching() failed, then the NMI - * triggered someplace critical, and rcu_irq_enter() should + * triggered someplace critical, and ct_irq_enter() should * not be called from NMI. */ if (unlikely(in_nmi())) return; =20 - rcu_irq_enter_irqson(); + ct_irq_enter_irqson(); __ftrace_trace_stack(buffer, trace_ctx, skip, NULL); - rcu_irq_exit_irqson(); + ct_irq_exit_irqson(); } =20 /** --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id AF088C433EF for ; Wed, 8 Jun 2022 14:42:24 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S241868AbiFHOmX (ORCPT ); Wed, 8 Jun 2022 10:42:23 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:33064 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241823AbiFHOl6 (ORCPT ); Wed, 8 Jun 2022 10:41:58 -0400 Received: from ams.source.kernel.org (ams.source.kernel.org [IPv6:2604:1380:4601:e00::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id BD12A15A846 for ; Wed, 8 Jun 2022 07:41:23 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ams.source.kernel.org (Postfix) with ESMTPS id 45319B8284D for ; Wed, 8 Jun 2022 14:41:22 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id BD6EBC3411E; Wed, 8 Jun 2022 14:41:17 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699281; bh=MpWpEZ+ppbvkk302IXxgp9fwXlIo7bKW5n05qijy2DM=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=lrXfoEZL+Tt2te+tWWtnsLvUbA3kean2UGiSFsq//R3ym5jD9nxrq/PGjONNjy+NN JH2uywlCoFmpLYEmFEL7FcYgLT7KLO+cnjdcXk4uXKqKneT9XS7Asva+L641ih8Cqa pV4sYGua34fb2KNMWKPEIpZ8PqISa4J1dU3b6JjtTvA9es44XLD51XxfV8jmt49yuR rtO9r5mDxoQM21Xmb/Or2i/nQChmXq60Is97awFXu6KtCD4rWhGNJzrNS7g2ZvYISd OXdafWHt1TeWxigmmDgPv1aepigvX9T7Hw7BNtx25k3mzquWdC6FRYhEKrdfKbJKvk W6cEFLb43zzuA== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 10/20] context_tracking: Take NMI eqs entrypoints over RCU Date: Wed, 8 Jun 2022 16:40:27 +0200 Message-Id: <20220608144037.1765000-11-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" The RCU dynticks counter is going to be merged into the context tracking subsystem. Prepare with moving the NMI extended quiescent states entrypoints to context tracking. For now those are dumb redirection to existing RCU calls. Acked-by: Paul E. McKenney Signed-off-by: Frederic Weisbecker Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- Documentation/RCU/Design/Requirements/Requirements.rst | 2 +- arch/Kconfig | 2 +- arch/arm64/kernel/entry-common.c | 8 ++++---- include/linux/context_tracking_irq.h | 4 ++++ include/linux/hardirq.h | 4 ++-- kernel/context_tracking.c | 10 ++++++++++ kernel/entry/common.c | 4 ++-- kernel/extable.c | 4 ++-- kernel/trace/trace.c | 2 +- 9 files changed, 27 insertions(+), 13 deletions(-) diff --git a/Documentation/RCU/Design/Requirements/Requirements.rst b/Docum= entation/RCU/Design/Requirements/Requirements.rst index 2a657fd2bca3..ac323f2af19b 100644 --- a/Documentation/RCU/Design/Requirements/Requirements.rst +++ b/Documentation/RCU/Design/Requirements/Requirements.rst @@ -1847,7 +1847,7 @@ normal interrupts. One way that this can happen is fo= r code that directly invokes ct_irq_enter() and ct_irq_exit() to be called from an NMI handler. This astonishing fact of life prompted the current code structure, which has ct_irq_enter() invoking -rcu_nmi_enter() and ct_irq_exit() invoking rcu_nmi_exit(). +ct_nmi_enter() and ct_irq_exit() invoking ct_nmi_exit(). And yes, I also learned of this requirement the hard way. =20 Loadable Modules diff --git a/arch/Kconfig b/arch/Kconfig index 888f9725b20b..a1150a190bf6 100644 --- a/arch/Kconfig +++ b/arch/Kconfig @@ -792,7 +792,7 @@ config HAVE_CONTEXT_TRACKING_USER_OFFSTACK =20 - Critical entry code isn't preemptible (or better yet: not interruptible). - - No use of RCU read side critical sections, unless rcu_nmi_enter() + - No use of RCU read side critical sections, unless ct_nmi_enter() got called. - No use of instrumentation, unless instrumentation_begin() got called. diff --git a/arch/arm64/kernel/entry-common.c b/arch/arm64/kernel/entry-com= mon.c index a1490e8bf5d4..b8fcdd777a8a 100644 --- a/arch/arm64/kernel/entry-common.c +++ b/arch/arm64/kernel/entry-common.c @@ -161,7 +161,7 @@ static void noinstr arm64_enter_nmi(struct pt_regs *reg= s) __nmi_enter(); lockdep_hardirqs_off(CALLER_ADDR0); lockdep_hardirq_enter(); - rcu_nmi_enter(); + ct_nmi_enter(); =20 trace_hardirqs_off_finish(); ftrace_nmi_enter(); @@ -182,7 +182,7 @@ static void noinstr arm64_exit_nmi(struct pt_regs *regs) lockdep_hardirqs_on_prepare(CALLER_ADDR0); } =20 - rcu_nmi_exit(); + ct_nmi_exit(); lockdep_hardirq_exit(); if (restore) lockdep_hardirqs_on(CALLER_ADDR0); @@ -199,7 +199,7 @@ static void noinstr arm64_enter_el1_dbg(struct pt_regs = *regs) regs->lockdep_hardirqs =3D lockdep_hardirqs_enabled(); =20 lockdep_hardirqs_off(CALLER_ADDR0); - rcu_nmi_enter(); + ct_nmi_enter(); =20 trace_hardirqs_off_finish(); } @@ -218,7 +218,7 @@ static void noinstr arm64_exit_el1_dbg(struct pt_regs *= regs) lockdep_hardirqs_on_prepare(CALLER_ADDR0); } =20 - rcu_nmi_exit(); + ct_nmi_exit(); if (restore) lockdep_hardirqs_on(CALLER_ADDR0); } diff --git a/include/linux/context_tracking_irq.h b/include/linux/context_t= racking_irq.h index 62f62bbd1a50..c50b5670c4a5 100644 --- a/include/linux/context_tracking_irq.h +++ b/include/linux/context_tracking_irq.h @@ -7,11 +7,15 @@ void ct_irq_enter(void); void ct_irq_exit(void); void ct_irq_enter_irqson(void); void ct_irq_exit_irqson(void); +void ct_nmi_enter(void); +void ct_nmi_exit(void); #else static inline void ct_irq_enter(void) { } static inline void ct_irq_exit(void) { } static inline void ct_irq_enter_irqson(void) { } static inline void ct_irq_exit_irqson(void) { } +static inline void ct_nmi_enter(void) { } +static inline void ct_nmi_exit(void) { } #endif =20 #endif diff --git a/include/linux/hardirq.h b/include/linux/hardirq.h index 76878b357ffa..345cdbe9c1b7 100644 --- a/include/linux/hardirq.h +++ b/include/linux/hardirq.h @@ -124,7 +124,7 @@ extern void rcu_nmi_exit(void); do { \ __nmi_enter(); \ lockdep_hardirq_enter(); \ - rcu_nmi_enter(); \ + ct_nmi_enter(); \ instrumentation_begin(); \ ftrace_nmi_enter(); \ instrumentation_end(); \ @@ -143,7 +143,7 @@ extern void rcu_nmi_exit(void); instrumentation_begin(); \ ftrace_nmi_exit(); \ instrumentation_end(); \ - rcu_nmi_exit(); \ + ct_nmi_exit(); \ lockdep_hardirq_exit(); \ __nmi_exit(); \ } while (0) diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index 5845ddd4c88c..5f72300e92e1 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -55,6 +55,16 @@ void ct_irq_exit_irqson(void) { rcu_irq_exit_irqson(); } + +noinstr void ct_nmi_enter(void) +{ + rcu_nmi_enter(); +} + +noinstr void ct_nmi_exit(void) +{ + rcu_nmi_exit(); +} #endif /* #ifdef CONFIG_CONTEXT_TRACKING_IDLE */ =20 #ifdef CONFIG_CONTEXT_TRACKING_USER diff --git a/kernel/entry/common.c b/kernel/entry/common.c index 9a89386696ba..256fb950a603 100644 --- a/kernel/entry/common.c +++ b/kernel/entry/common.c @@ -449,7 +449,7 @@ irqentry_state_t noinstr irqentry_nmi_enter(struct pt_r= egs *regs) __nmi_enter(); lockdep_hardirqs_off(CALLER_ADDR0); lockdep_hardirq_enter(); - rcu_nmi_enter(); + ct_nmi_enter(); =20 instrumentation_begin(); trace_hardirqs_off_finish(); @@ -469,7 +469,7 @@ void noinstr irqentry_nmi_exit(struct pt_regs *regs, ir= qentry_state_t irq_state) } instrumentation_end(); =20 - rcu_nmi_exit(); + ct_nmi_exit(); lockdep_hardirq_exit(); if (irq_state.lockdep) lockdep_hardirqs_on(CALLER_ADDR0); diff --git a/kernel/extable.c b/kernel/extable.c index bda5e9761541..71f482581cab 100644 --- a/kernel/extable.c +++ b/kernel/extable.c @@ -114,7 +114,7 @@ int kernel_text_address(unsigned long addr) =20 /* Treat this like an NMI as it can happen anywhere */ if (no_rcu) - rcu_nmi_enter(); + ct_nmi_enter(); =20 if (is_module_text_address(addr)) goto out; @@ -127,7 +127,7 @@ int kernel_text_address(unsigned long addr) ret =3D 0; out: if (no_rcu) - rcu_nmi_exit(); + ct_nmi_exit(); =20 return ret; } diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 85e519cc0af9..c096da6ea121 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -3102,7 +3102,7 @@ void __trace_stack(struct trace_array *tr, unsigned i= nt trace_ctx, } =20 /* - * When an NMI triggers, RCU is enabled via rcu_nmi_enter(), + * When an NMI triggers, RCU is enabled via ct_nmi_enter(), * but if the above rcu_is_watching() failed, then the NMI * triggered someplace critical, and ct_irq_enter() should * not be called from NMI. --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id F2B8CC43334 for ; Wed, 8 Jun 2022 14:42:30 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S241679AbiFHOm1 (ORCPT ); Wed, 8 Jun 2022 10:42:27 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:33094 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241841AbiFHOl6 (ORCPT ); Wed, 8 Jun 2022 10:41:58 -0400 Received: from ams.source.kernel.org (ams.source.kernel.org [145.40.68.75]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 7DF55154B00 for ; Wed, 8 Jun 2022 07:41:27 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ams.source.kernel.org (Postfix) with ESMTPS id EEF94B82850 for ; Wed, 8 Jun 2022 14:41:25 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 6EA5EC341C4; Wed, 8 Jun 2022 14:41:21 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699284; bh=RMsi/UGAzE8TIHOxZuWIfZgw1udptkzdbC+PbK0k3SQ=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=a7vCdkAOCZx9y0JQnXTRBQp3//15sW2IK/FXCcHN1SDqOeN9mm6ZeERBUfAt+OyYo JoNmJLckQLhBLz91jr2A5mI4GFfv0y+EfhuxbND332FRz2rfc6/EZLf/6JF3g7vE5K QOOD5I/9VHjzwFw7ACJIij3g6PUfc/a8ADUolxJhGHMP0ni1r6/AahvASGZ3qmV3vA 52VyvFwLc4H11/eviKHwWeo2Qavu5vN+A631tm5PE0chFDOkz/JZ0GjSVB6Rjey1U+ wf2LOcG53t2gtzKa3i25OtW90klRAl4q3GoolUFaafGOtBmvUl91LZoovpOEpEFVEr 8i2ddcbi2YoUg== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 11/20] rcu/context-tracking: Remove rcu_irq_enter/exit() Date: Wed, 8 Jun 2022 16:40:28 +0200 Message-Id: <20220608144037.1765000-12-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Now rcu_irq_enter/exit() is an unecessary middle call between ct_irq_enter/exit() and nmi_irq_enter/exit(). Take this opportunity to remove the former functions and move the comments above them to the new entrypoints. Signed-off-by: Frederic Weisbecker Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- include/linux/rcutiny.h | 4 -- include/linux/rcutree.h | 4 -- kernel/context_tracking.c | 71 +++++++++++++++++++++++++++++++-- kernel/rcu/tree.c | 83 --------------------------------------- 4 files changed, 67 insertions(+), 95 deletions(-) diff --git a/include/linux/rcutiny.h b/include/linux/rcutiny.h index e6bb31a0927b..b94e8c8fda5a 100644 --- a/include/linux/rcutiny.h +++ b/include/linux/rcutiny.h @@ -97,10 +97,6 @@ static inline void rcu_cpu_stall_reset(void) { } static inline int rcu_jiffies_till_stall_check(void) { return 21 * HZ; } static inline void rcu_idle_enter(void) { } static inline void rcu_idle_exit(void) { } -static inline void rcu_irq_enter(void) { } -static inline void rcu_irq_exit_irqson(void) { } -static inline void rcu_irq_enter_irqson(void) { } -static inline void rcu_irq_exit(void) { } static inline void rcu_irq_exit_check_preempt(void) { } #define rcu_is_idle_cpu(cpu) \ (is_idle_task(current) && !in_nmi() && !in_hardirq() && !in_serving_softi= rq()) diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h index 20dbaa9a3882..cafe3fbf4272 100644 --- a/include/linux/rcutree.h +++ b/include/linux/rcutree.h @@ -49,10 +49,6 @@ void cond_synchronize_rcu(unsigned long oldstate); =20 void rcu_idle_enter(void); void rcu_idle_exit(void); -void rcu_irq_enter(void); -void rcu_irq_exit(void); -void rcu_irq_enter_irqson(void); -void rcu_irq_exit_irqson(void); bool rcu_is_idle_cpu(int cpu); =20 #ifdef CONFIG_PROVE_RCU diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index 5f72300e92e1..c55de8a12c49 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -36,24 +36,87 @@ void ct_idle_exit(void) } EXPORT_SYMBOL_GPL(ct_idle_exit); =20 +/** + * ct_irq_enter - inform RCU that current CPU is entering irq away from id= le + * + * Enter an interrupt handler, which might possibly result in exiting + * idle mode, in other words, entering the mode in which read-side critical + * sections can occur. The caller must have disabled interrupts. + * + * Note that the Linux kernel is fully capable of entering an interrupt + * handler that it never exits, for example when doing upcalls to user mod= e! + * This code assumes that the idle loop never does upcalls to user mode. + * If your architecture's idle loop does do upcalls to user mode (or does + * anything else that results in unbalanced calls to the irq_enter() and + * irq_exit() functions), RCU will give you what you deserve, good and har= d. + * But very infrequently and irreproducibly. + * + * Use things like work queues to work around this limitation. + * + * You have been warned. + * + * If you add or remove a call to ct_irq_enter(), be sure to test with + * CONFIG_RCU_EQS_DEBUG=3Dy. + */ noinstr void ct_irq_enter(void) { - rcu_irq_enter(); + lockdep_assert_irqs_disabled(); + ct_nmi_enter(); } =20 +/** + * ct_irq_exit - inform RCU that current CPU is exiting irq towards idle + * + * Exit from an interrupt handler, which might possibly result in entering + * idle mode, in other words, leaving the mode in which read-side critical + * sections can occur. The caller must have disabled interrupts. + * + * This code assumes that the idle loop never does anything that might + * result in unbalanced calls to irq_enter() and irq_exit(). If your + * architecture's idle loop violates this assumption, RCU will give you wh= at + * you deserve, good and hard. But very infrequently and irreproducibly. + * + * Use things like work queues to work around this limitation. + * + * You have been warned. + * + * If you add or remove a call to ct_irq_exit(), be sure to test with + * CONFIG_RCU_EQS_DEBUG=3Dy. + */ noinstr void ct_irq_exit(void) { - rcu_irq_exit(); + lockdep_assert_irqs_disabled(); + ct_nmi_exit(); } =20 +/* + * Wrapper for ct_irq_enter() where interrupts are enabled. + * + * If you add or remove a call to ct_irq_enter_irqson(), be sure to test + * with CONFIG_RCU_EQS_DEBUG=3Dy. + */ void ct_irq_enter_irqson(void) { - rcu_irq_enter_irqson(); + unsigned long flags; + + local_irq_save(flags); + ct_irq_enter(); + local_irq_restore(flags); } =20 +/* + * Wrapper for ct_irq_exit() where interrupts are enabled. + * + * If you add or remove a call to ct_irq_exit_irqson(), be sure to test + * with CONFIG_RCU_EQS_DEBUG=3Dy. + */ void ct_irq_exit_irqson(void) { - rcu_irq_exit_irqson(); + unsigned long flags; + + local_irq_save(flags); + ct_irq_exit(); + local_irq_restore(flags); } =20 noinstr void ct_nmi_enter(void) diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index d4ac5538febb..5e4047cdd8f9 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -793,31 +793,6 @@ noinstr void rcu_nmi_exit(void) rcu_dynticks_task_enter(); } =20 -/** - * rcu_irq_exit - inform RCU that current CPU is exiting irq towards idle - * - * Exit from an interrupt handler, which might possibly result in entering - * idle mode, in other words, leaving the mode in which read-side critical - * sections can occur. The caller must have disabled interrupts. - * - * This code assumes that the idle loop never does anything that might - * result in unbalanced calls to irq_enter() and irq_exit(). If your - * architecture's idle loop violates this assumption, RCU will give you wh= at - * you deserve, good and hard. But very infrequently and irreproducibly. - * - * Use things like work queues to work around this limitation. - * - * You have been warned. - * - * If you add or remove a call to rcu_irq_exit(), be sure to test with - * CONFIG_RCU_EQS_DEBUG=3Dy. - */ -void noinstr rcu_irq_exit(void) -{ - lockdep_assert_irqs_disabled(); - rcu_nmi_exit(); -} - #ifdef CONFIG_PROVE_RCU /** * rcu_irq_exit_check_preempt - Validate that scheduling is possible @@ -836,21 +811,6 @@ void rcu_irq_exit_check_preempt(void) } #endif /* #ifdef CONFIG_PROVE_RCU */ =20 -/* - * Wrapper for rcu_irq_exit() where interrupts are enabled. - * - * If you add or remove a call to rcu_irq_exit_irqson(), be sure to test - * with CONFIG_RCU_EQS_DEBUG=3Dy. - */ -void rcu_irq_exit_irqson(void) -{ - unsigned long flags; - - local_irq_save(flags); - rcu_irq_exit(); - local_irq_restore(flags); -} - /* * Exit an RCU extended quiescent state, which can be either the * idle loop or adaptive-tickless usermode execution. @@ -1045,49 +1005,6 @@ noinstr void rcu_nmi_enter(void) barrier(); } =20 -/** - * rcu_irq_enter - inform RCU that current CPU is entering irq away from i= dle - * - * Enter an interrupt handler, which might possibly result in exiting - * idle mode, in other words, entering the mode in which read-side critical - * sections can occur. The caller must have disabled interrupts. - * - * Note that the Linux kernel is fully capable of entering an interrupt - * handler that it never exits, for example when doing upcalls to user mod= e! - * This code assumes that the idle loop never does upcalls to user mode. - * If your architecture's idle loop does do upcalls to user mode (or does - * anything else that results in unbalanced calls to the irq_enter() and - * irq_exit() functions), RCU will give you what you deserve, good and har= d. - * But very infrequently and irreproducibly. - * - * Use things like work queues to work around this limitation. - * - * You have been warned. - * - * If you add or remove a call to rcu_irq_enter(), be sure to test with - * CONFIG_RCU_EQS_DEBUG=3Dy. - */ -noinstr void rcu_irq_enter(void) -{ - lockdep_assert_irqs_disabled(); - rcu_nmi_enter(); -} - -/* - * Wrapper for rcu_irq_enter() where interrupts are enabled. - * - * If you add or remove a call to rcu_irq_enter_irqson(), be sure to test - * with CONFIG_RCU_EQS_DEBUG=3Dy. - */ -void rcu_irq_enter_irqson(void) -{ - unsigned long flags; - - local_irq_save(flags); - rcu_irq_enter(); - local_irq_restore(flags); -} - /* * Check to see if any future non-offloaded RCU-related work will need * to be done by the current CPU, even if none need be done immediately, --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 3AACDC433EF for ; Wed, 8 Jun 2022 14:44:10 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S242069AbiFHOoI (ORCPT ); Wed, 8 Jun 2022 10:44:08 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:33064 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241902AbiFHOmP (ORCPT ); Wed, 8 Jun 2022 10:42:15 -0400 Received: from ams.source.kernel.org (ams.source.kernel.org [145.40.68.75]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 77A14182A61 for ; Wed, 8 Jun 2022 07:41:31 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ams.source.kernel.org (Postfix) with ESMTPS id 8CF44B8284A for ; Wed, 8 Jun 2022 14:41:29 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 214FDC3411D; Wed, 8 Jun 2022 14:41:24 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699288; bh=qGrUIfwXvG7yTxjwej4vdlqsYZxH6ontr3GA8ffj/ig=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=CmjcbQcmu5TVYlkl1YvsSds93nbQpfLMYZ/JXbDZVfg+00EiMeiyPnEmwJ1it53QM I2kCOdO0P65XwYcfRZuwc1PQ4RP3VtkMsQg7EyuxAMohjynecIXkOujB8TBxrixb35 mk5UDmnXJb732fItA6u7belr4WJXZbmm3TYnp5VJums/JuSpwu2v092dvFdlXrs7PG am6GmMoPHNgkLhLPM8OrWvMQcGwXnUaVObbdy3SvkRUCz4DuPHtrpM/jQK0MCwz8E4 0AFdNCjuSqdX5SmX3wb4q8mHFScS4sV3KosujH3MwnkrVF6z+vO3VUPVWbpneLwSC3 E+M+sJCjZBoNA== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 12/20] rcu/context_tracking: Move dynticks counter to context tracking Date: Wed, 8 Jun 2022 16:40:29 +0200 Message-Id: <20220608144037.1765000-13-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" In order to prepare for merging RCU dynticks counter into the context tracking state, move the rcu_data's dynticks field to the context tracking structure. It will later be mixed within the context tracking state itself. Acked-by: Paul E. McKenney Signed-off-by: Frederic Weisbecker Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- include/linux/context_tracking_state.h | 31 +++++++++++++- kernel/context_tracking.c | 10 +++-- kernel/rcu/tree.c | 56 +++++++++++++------------- kernel/rcu/tree.h | 1 - kernel/rcu/tree_exp.h | 2 +- kernel/rcu/tree_stall.h | 4 +- 6 files changed, 66 insertions(+), 38 deletions(-) diff --git a/include/linux/context_tracking_state.h b/include/linux/context= _tracking_state.h index 7b46b43b8c98..651bb9c353f0 100644 --- a/include/linux/context_tracking_state.h +++ b/include/linux/context_tracking_state.h @@ -7,6 +7,7 @@ #include =20 struct context_tracking { +#ifdef CONFIG_CONTEXT_TRACKING_USER /* * When active is false, probes are unset in order * to minimize overhead: TIF flags are cleared @@ -21,11 +22,37 @@ struct context_tracking { CONTEXT_USER, CONTEXT_GUEST, } state; +#endif +#ifdef CONFIG_CONTEXT_TRACKING_IDLE + atomic_t dynticks; /* Even value for idle, else odd. */ +#endif }; =20 -#ifdef CONFIG_CONTEXT_TRACKING_USER -extern struct static_key_false context_tracking_key; +#ifdef CONFIG_CONTEXT_TRACKING DECLARE_PER_CPU(struct context_tracking, context_tracking); +#endif + +#ifdef CONFIG_CONTEXT_TRACKING_IDLE +static __always_inline int ct_dynticks(void) +{ + return atomic_read(this_cpu_ptr(&context_tracking.dynticks)); +} + +static __always_inline int ct_dynticks_cpu(int cpu) +{ + struct context_tracking *ct =3D per_cpu_ptr(&context_tracking, cpu); + return atomic_read(&ct->dynticks); +} + +static __always_inline int ct_dynticks_cpu_acquire(int cpu) +{ + struct context_tracking *ct =3D per_cpu_ptr(&context_tracking, cpu); + return atomic_read_acquire(&ct->dynticks); +} +#endif /* #ifdef CONFIG_CONTEXT_TRACKING_IDLE */ + +#ifdef CONFIG_CONTEXT_TRACKING_USER +extern struct static_key_false context_tracking_key; =20 static __always_inline bool context_tracking_enabled(void) { diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index c55de8a12c49..3475dbfa4d73 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -23,6 +23,13 @@ #include =20 =20 +DEFINE_PER_CPU(struct context_tracking, context_tracking) =3D { +#ifdef CONFIG_CONTEXT_TRACKING_IDLE + .dynticks =3D ATOMIC_INIT(1), +#endif +}; +EXPORT_SYMBOL_GPL(context_tracking); + #ifdef CONFIG_CONTEXT_TRACKING_IDLE noinstr void ct_idle_enter(void) { @@ -138,9 +145,6 @@ noinstr void ct_nmi_exit(void) DEFINE_STATIC_KEY_FALSE(context_tracking_key); EXPORT_SYMBOL_GPL(context_tracking_key); =20 -DEFINE_PER_CPU(struct context_tracking, context_tracking); -EXPORT_SYMBOL_GPL(context_tracking); - static noinstr bool context_tracking_recursion_enter(void) { int recursion; diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index 5e4047cdd8f9..5c9ca5900540 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -77,7 +77,6 @@ static DEFINE_PER_CPU_SHARED_ALIGNED(struct rcu_data, rcu_data) =3D { .dynticks_nesting =3D 1, .dynticks_nmi_nesting =3D DYNTICK_IRQ_NONIDLE, - .dynticks =3D ATOMIC_INIT(1), #ifdef CONFIG_RCU_NOCB_CPU .cblist.flags =3D SEGCBLIST_RCU_CORE, #endif @@ -272,7 +271,7 @@ void rcu_softirq_qs(void) */ static noinline noinstr unsigned long rcu_dynticks_inc(int incby) { - return arch_atomic_add_return(incby, this_cpu_ptr(&rcu_data.dynticks)); + return arch_atomic_add_return(incby, this_cpu_ptr(&context_tracking.dynti= cks)); } =20 /* @@ -328,9 +327,7 @@ static noinstr void rcu_dynticks_eqs_exit(void) */ static void rcu_dynticks_eqs_online(void) { - struct rcu_data *rdp =3D this_cpu_ptr(&rcu_data); - - if (atomic_read(&rdp->dynticks) & 0x1) + if (ct_dynticks() & 0x1) return; rcu_dynticks_inc(1); } @@ -342,17 +339,17 @@ static void rcu_dynticks_eqs_online(void) */ static __always_inline bool rcu_dynticks_curr_cpu_in_eqs(void) { - return !(arch_atomic_read(this_cpu_ptr(&rcu_data.dynticks)) & 0x1); + return !(arch_atomic_read(this_cpu_ptr(&context_tracking.dynticks)) & 0x1= ); } =20 /* * Snapshot the ->dynticks counter with full ordering so as to allow * stable comparison of this counter with past and future snapshots. */ -static int rcu_dynticks_snap(struct rcu_data *rdp) +static int rcu_dynticks_snap(int cpu) { smp_mb(); // Fundamental RCU ordering guarantee. - return atomic_read_acquire(&rdp->dynticks); + return ct_dynticks_cpu_acquire(cpu); } =20 /* @@ -367,9 +364,7 @@ static bool rcu_dynticks_in_eqs(int snap) /* Return true if the specified CPU is currently idle from an RCU viewpoin= t. */ bool rcu_is_idle_cpu(int cpu) { - struct rcu_data *rdp =3D per_cpu_ptr(&rcu_data, cpu); - - return rcu_dynticks_in_eqs(rcu_dynticks_snap(rdp)); + return rcu_dynticks_in_eqs(rcu_dynticks_snap(cpu)); } =20 /* @@ -379,7 +374,7 @@ bool rcu_is_idle_cpu(int cpu) */ static bool rcu_dynticks_in_eqs_since(struct rcu_data *rdp, int snap) { - return snap !=3D rcu_dynticks_snap(rdp); + return snap !=3D rcu_dynticks_snap(rdp->cpu); } =20 /* @@ -388,11 +383,10 @@ static bool rcu_dynticks_in_eqs_since(struct rcu_data= *rdp, int snap) */ bool rcu_dynticks_zero_in_eqs(int cpu, int *vp) { - struct rcu_data *rdp =3D per_cpu_ptr(&rcu_data, cpu); int snap; =20 // If not quiescent, force back to earlier extended quiescent state. - snap =3D atomic_read(&rdp->dynticks) & ~0x1; + snap =3D ct_dynticks_cpu(cpu) & ~0x1; =20 smp_rmb(); // Order ->dynticks and *vp reads. if (READ_ONCE(*vp)) @@ -400,7 +394,7 @@ bool rcu_dynticks_zero_in_eqs(int cpu, int *vp) smp_rmb(); // Order *vp read and ->dynticks re-read. =20 // If still in the same extended quiescent state, we are good! - return snap =3D=3D atomic_read(&rdp->dynticks); + return snap =3D=3D ct_dynticks_cpu(cpu); } =20 /* @@ -624,6 +618,7 @@ EXPORT_SYMBOL_GPL(rcutorture_get_gp_data); static noinstr void rcu_eqs_enter(bool user) { struct rcu_data *rdp =3D this_cpu_ptr(&rcu_data); + struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); =20 WARN_ON_ONCE(rdp->dynticks_nmi_nesting !=3D DYNTICK_IRQ_NONIDLE); WRITE_ONCE(rdp->dynticks_nmi_nesting, 0); @@ -637,12 +632,12 @@ static noinstr void rcu_eqs_enter(bool user) =20 instrumentation_begin(); lockdep_assert_irqs_disabled(); - trace_rcu_dyntick(TPS("Start"), rdp->dynticks_nesting, 0, atomic_read(&rd= p->dynticks)); + trace_rcu_dyntick(TPS("Start"), rdp->dynticks_nesting, 0, ct_dynticks()); WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(c= urrent)); rcu_preempt_deferred_qs(current); =20 // instrumentation for the noinstr rcu_dynticks_eqs_enter() - instrument_atomic_write(&rdp->dynticks, sizeof(rdp->dynticks)); + instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); =20 instrumentation_end(); WRITE_ONCE(rdp->dynticks_nesting, 0); /* Avoid irq-access tearing. */ @@ -744,7 +739,7 @@ noinstr void rcu_user_enter(void) * rcu_nmi_exit - inform RCU of exit from NMI context * * If we are returning from the outermost NMI handler that interrupted an - * RCU-idle period, update rdp->dynticks and rdp->dynticks_nmi_nesting + * RCU-idle period, update ct->dynticks and rdp->dynticks_nmi_nesting * to let the RCU grace-period handling know that the CPU is back to * being RCU-idle. * @@ -753,6 +748,7 @@ noinstr void rcu_user_enter(void) */ noinstr void rcu_nmi_exit(void) { + struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); struct rcu_data *rdp =3D this_cpu_ptr(&rcu_data); =20 instrumentation_begin(); @@ -770,7 +766,7 @@ noinstr void rcu_nmi_exit(void) */ if (rdp->dynticks_nmi_nesting !=3D 1) { trace_rcu_dyntick(TPS("--=3D"), rdp->dynticks_nmi_nesting, rdp->dynticks= _nmi_nesting - 2, - atomic_read(&rdp->dynticks)); + ct_dynticks()); WRITE_ONCE(rdp->dynticks_nmi_nesting, /* No store tearing. */ rdp->dynticks_nmi_nesting - 2); instrumentation_end(); @@ -778,11 +774,11 @@ noinstr void rcu_nmi_exit(void) } =20 /* This NMI interrupted an RCU-idle CPU, restore RCU-idleness. */ - trace_rcu_dyntick(TPS("Startirq"), rdp->dynticks_nmi_nesting, 0, atomic_r= ead(&rdp->dynticks)); + trace_rcu_dyntick(TPS("Startirq"), rdp->dynticks_nmi_nesting, 0, ct_dynti= cks()); WRITE_ONCE(rdp->dynticks_nmi_nesting, 0); /* Avoid store tearing. */ =20 // instrumentation for the noinstr rcu_dynticks_eqs_enter() - instrument_atomic_write(&rdp->dynticks, sizeof(rdp->dynticks)); + instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); instrumentation_end(); =20 // RCU is watching here ... @@ -821,6 +817,7 @@ void rcu_irq_exit_check_preempt(void) */ static void noinstr rcu_eqs_exit(bool user) { + struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); struct rcu_data *rdp; long oldval; =20 @@ -840,9 +837,9 @@ static void noinstr rcu_eqs_exit(bool user) instrumentation_begin(); =20 // instrumentation for the noinstr rcu_dynticks_eqs_exit() - instrument_atomic_write(&rdp->dynticks, sizeof(rdp->dynticks)); + instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); =20 - trace_rcu_dyntick(TPS("End"), rdp->dynticks_nesting, 1, atomic_read(&rdp-= >dynticks)); + trace_rcu_dyntick(TPS("End"), rdp->dynticks_nesting, 1, ct_dynticks()); WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(c= urrent)); WRITE_ONCE(rdp->dynticks_nesting, 1); WARN_ON_ONCE(rdp->dynticks_nmi_nesting); @@ -948,7 +945,7 @@ void __rcu_irq_enter_check_tick(void) /** * rcu_nmi_enter - inform RCU of entry to NMI context * - * If the CPU was idle from RCU's viewpoint, update rdp->dynticks and + * If the CPU was idle from RCU's viewpoint, update ct->dynticks and * rdp->dynticks_nmi_nesting to let the RCU grace-period handling know * that the CPU is active. This implementation permits nested NMIs, as * long as the nesting level does not overflow an int. (You will probably @@ -961,6 +958,7 @@ noinstr void rcu_nmi_enter(void) { long incby =3D 2; struct rcu_data *rdp =3D this_cpu_ptr(&rcu_data); + struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); =20 /* Complain about underflow. */ WARN_ON_ONCE(rdp->dynticks_nmi_nesting < 0); @@ -984,9 +982,9 @@ noinstr void rcu_nmi_enter(void) =20 instrumentation_begin(); // instrumentation for the noinstr rcu_dynticks_curr_cpu_in_eqs() - instrument_atomic_read(&rdp->dynticks, sizeof(rdp->dynticks)); + instrument_atomic_read(&ct->dynticks, sizeof(ct->dynticks)); // instrumentation for the noinstr rcu_dynticks_eqs_exit() - instrument_atomic_write(&rdp->dynticks, sizeof(rdp->dynticks)); + instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); =20 incby =3D 1; } else if (!in_nmi()) { @@ -998,7 +996,7 @@ noinstr void rcu_nmi_enter(void) =20 trace_rcu_dyntick(incby =3D=3D 1 ? TPS("Endirq") : TPS("++=3D"), rdp->dynticks_nmi_nesting, - rdp->dynticks_nmi_nesting + incby, atomic_read(&rdp->dynticks)); + rdp->dynticks_nmi_nesting + incby, ct_dynticks()); instrumentation_end(); WRITE_ONCE(rdp->dynticks_nmi_nesting, /* Prevent store tearing. */ rdp->dynticks_nmi_nesting + incby); @@ -1142,7 +1140,7 @@ static void rcu_gpnum_ovf(struct rcu_node *rnp, struc= t rcu_data *rdp) */ static int dyntick_save_progress_counter(struct rcu_data *rdp) { - rdp->dynticks_snap =3D rcu_dynticks_snap(rdp); + rdp->dynticks_snap =3D rcu_dynticks_snap(rdp->cpu); if (rcu_dynticks_in_eqs(rdp->dynticks_snap)) { trace_rcu_fqs(rcu_state.name, rdp->gp_seq, rdp->cpu, TPS("dti")); rcu_gpnum_ovf(rdp->mynode, rdp); @@ -4237,7 +4235,7 @@ rcu_boot_init_percpu_data(int cpu) rdp->grpmask =3D leaf_node_cpu_bit(rdp->mynode, cpu); INIT_WORK(&rdp->strict_work, strict_work_handler); WARN_ON_ONCE(rdp->dynticks_nesting !=3D 1); - WARN_ON_ONCE(rcu_dynticks_in_eqs(rcu_dynticks_snap(rdp))); + WARN_ON_ONCE(rcu_dynticks_in_eqs(rcu_dynticks_snap(cpu))); rdp->barrier_seq_snap =3D rcu_state.barrier_sequence; rdp->rcu_ofl_gp_seq =3D rcu_state.gp_seq; rdp->rcu_ofl_gp_flags =3D RCU_GP_CLEANED; diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h index 3cdc18997a38..2e46ee108d13 100644 --- a/kernel/rcu/tree.h +++ b/kernel/rcu/tree.h @@ -193,7 +193,6 @@ struct rcu_data { int dynticks_snap; /* Per-GP tracking for dynticks. */ long dynticks_nesting; /* Track process nesting level. */ long dynticks_nmi_nesting; /* Track irq/NMI nesting level. */ - atomic_t dynticks; /* Even value for idle, else odd. */ bool rcu_need_heavy_qs; /* GP old, so heavy quiescent state! */ bool rcu_urgent_qs; /* GP old need light quiescent state. */ bool rcu_forced_tick; /* Forced tick to provide QS. */ diff --git a/kernel/rcu/tree_exp.h b/kernel/rcu/tree_exp.h index f092c7f18a5f..be667583a554 100644 --- a/kernel/rcu/tree_exp.h +++ b/kernel/rcu/tree_exp.h @@ -358,7 +358,7 @@ static void __sync_rcu_exp_select_node_cpus(struct rcu_= exp_work *rewp) !(rnp->qsmaskinitnext & mask)) { mask_ofl_test |=3D mask; } else { - snap =3D rcu_dynticks_snap(rdp); + snap =3D rcu_dynticks_snap(cpu); if (rcu_dynticks_in_eqs(snap)) mask_ofl_test |=3D mask; else diff --git a/kernel/rcu/tree_stall.h b/kernel/rcu/tree_stall.h index 2464b0eccfd0..594a3c5aded1 100644 --- a/kernel/rcu/tree_stall.h +++ b/kernel/rcu/tree_stall.h @@ -465,7 +465,7 @@ static void print_cpu_stall_info(int cpu) } delta =3D rcu_seq_ctr(rdp->mynode->gp_seq - rdp->rcu_iw_gp_seq); falsepositive =3D rcu_is_gp_kthread_starving(NULL) && - rcu_dynticks_in_eqs(rcu_dynticks_snap(rdp)); + rcu_dynticks_in_eqs(rcu_dynticks_snap(cpu)); rcuc_starved =3D rcu_is_rcuc_kthread_starving(rdp, &j); if (rcuc_starved) sprintf(buf, " rcuc=3D%ld jiffies(starved)", j); @@ -478,7 +478,7 @@ static void print_cpu_stall_info(int cpu) rdp->rcu_iw_pending ? (int)min(delta, 9UL) + '0' : "!."[!delta], ticks_value, ticks_title, - rcu_dynticks_snap(rdp) & 0xfff, + rcu_dynticks_snap(cpu) & 0xfff, rdp->dynticks_nesting, rdp->dynticks_nmi_nesting, rdp->softirq_snap, kstat_softirqs_cpu(RCU_SOFTIRQ, cpu), data_race(rcu_state.n_force_qs) - rcu_state.n_force_qs_gpstart, --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 0E2B6C43334 for ; Wed, 8 Jun 2022 14:44:15 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S241781AbiFHOoM (ORCPT ); Wed, 8 Jun 2022 10:44:12 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:33048 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241796AbiFHOmO (ORCPT ); Wed, 8 Jun 2022 10:42:14 -0400 Received: from ams.source.kernel.org (ams.source.kernel.org [IPv6:2604:1380:4601:e00::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 9A337182BA5 for ; Wed, 8 Jun 2022 07:41:34 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ams.source.kernel.org (Postfix) with ESMTPS id 3D573B827F0 for ; Wed, 8 Jun 2022 14:41:33 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id C648BC36AFE; Wed, 8 Jun 2022 14:41:28 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699292; bh=wzqhad6/AJjpIQFihkTTwVoo9VWkN1MztzCxEb7UMRE=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=RmQNR4RNzHsZAnm3FBYn/rvMJXGrXcNJhK8I0KYICqO/drKADCTzk7ZpCjiJ8uQR/ SXuXOkCXk0XN3L89ZF8bSLOYq3C4+MmWUzDWe/CDmPa4/kHbf0uGbn/L2I6rE/Glwk N81NUpIP81xGvm7vlpzMydXkwIvoPXLApbStspTr3yYMSRO0tDzyJ4u8EJNEXnfkic 8YVFEaukTsoUklaD7phLHDJx17R81IF1d8cRBLOnVzZYACtF3WaSW3SDOA826IjXA/ bsxapEVHCmAsFYhTuy1T1AqgCneCNGSZQwpruWNosSCcu/SK2eJjP5zHZfzU1cJO6Y YviF+jf26Hs4A== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 13/20] rcu/context_tracking: Move dynticks_nesting to context tracking Date: Wed, 8 Jun 2022 16:40:30 +0200 Message-Id: <20220608144037.1765000-14-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" The RCU eqs tracking is going to be performed by the context tracking subsystem. The related nesting counters thus need to be moved to the context tracking structure. Acked-by: Paul E. McKenney Signed-off-by: Frederic Weisbecker Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- include/linux/context_tracking_state.h | 12 ++++++++++ kernel/context_tracking.c | 1 + kernel/rcu/tree.c | 31 +++++++++++++------------- kernel/rcu/tree.h | 1 - kernel/rcu/tree_stall.h | 2 +- 5 files changed, 30 insertions(+), 17 deletions(-) diff --git a/include/linux/context_tracking_state.h b/include/linux/context= _tracking_state.h index 651bb9c353f0..164f4673ee74 100644 --- a/include/linux/context_tracking_state.h +++ b/include/linux/context_tracking_state.h @@ -25,6 +25,7 @@ struct context_tracking { #endif #ifdef CONFIG_CONTEXT_TRACKING_IDLE atomic_t dynticks; /* Even value for idle, else odd. */ + long dynticks_nesting; /* Track process nesting level. */ #endif }; =20 @@ -49,6 +50,17 @@ static __always_inline int ct_dynticks_cpu_acquire(int c= pu) struct context_tracking *ct =3D per_cpu_ptr(&context_tracking, cpu); return atomic_read_acquire(&ct->dynticks); } + +static __always_inline long ct_dynticks_nesting(void) +{ + return __this_cpu_read(context_tracking.dynticks_nesting); +} + +static __always_inline long ct_dynticks_nesting_cpu(int cpu) +{ + struct context_tracking *ct =3D per_cpu_ptr(&context_tracking, cpu); + return ct->dynticks_nesting; +} #endif /* #ifdef CONFIG_CONTEXT_TRACKING_IDLE */ =20 #ifdef CONFIG_CONTEXT_TRACKING_USER diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index 3475dbfa4d73..d5ff1de1a24a 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -25,6 +25,7 @@ =20 DEFINE_PER_CPU(struct context_tracking, context_tracking) =3D { #ifdef CONFIG_CONTEXT_TRACKING_IDLE + .dynticks_nesting =3D 1, .dynticks =3D ATOMIC_INIT(1), #endif }; diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index 5c9ca5900540..5e7771c78162 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -75,7 +75,6 @@ /* Data structures. */ =20 static DEFINE_PER_CPU_SHARED_ALIGNED(struct rcu_data, rcu_data) =3D { - .dynticks_nesting =3D 1, .dynticks_nmi_nesting =3D DYNTICK_IRQ_NONIDLE, #ifdef CONFIG_RCU_NOCB_CPU .cblist.flags =3D SEGCBLIST_RCU_CORE, @@ -440,7 +439,7 @@ static int rcu_is_cpu_rrupt_from_idle(void) lockdep_assert_irqs_disabled(); =20 /* Check for counter underflows */ - RCU_LOCKDEP_WARN(__this_cpu_read(rcu_data.dynticks_nesting) < 0, + RCU_LOCKDEP_WARN(ct_dynticks_nesting() < 0, "RCU dynticks_nesting counter underflow!"); RCU_LOCKDEP_WARN(__this_cpu_read(rcu_data.dynticks_nmi_nesting) <=3D 0, "RCU dynticks_nmi_nesting counter underflow/zero!"); @@ -456,7 +455,7 @@ static int rcu_is_cpu_rrupt_from_idle(void) WARN_ON_ONCE(!nesting && !is_idle_task(current)); =20 /* Does CPU appear to be idle from an RCU standpoint? */ - return __this_cpu_read(rcu_data.dynticks_nesting) =3D=3D 0; + return ct_dynticks_nesting() =3D=3D 0; } =20 #define DEFAULT_RCU_BLIMIT (IS_ENABLED(CONFIG_RCU_STRICT_GRACE_PERIOD) ? 1= 000 : 10) @@ -623,16 +622,16 @@ static noinstr void rcu_eqs_enter(bool user) WARN_ON_ONCE(rdp->dynticks_nmi_nesting !=3D DYNTICK_IRQ_NONIDLE); WRITE_ONCE(rdp->dynticks_nmi_nesting, 0); WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && - rdp->dynticks_nesting =3D=3D 0); - if (rdp->dynticks_nesting !=3D 1) { + ct_dynticks_nesting() =3D=3D 0); + if (ct_dynticks_nesting() !=3D 1) { // RCU will still be watching, so just do accounting and leave. - rdp->dynticks_nesting--; + ct->dynticks_nesting--; return; } =20 instrumentation_begin(); lockdep_assert_irqs_disabled(); - trace_rcu_dyntick(TPS("Start"), rdp->dynticks_nesting, 0, ct_dynticks()); + trace_rcu_dyntick(TPS("Start"), ct_dynticks_nesting(), 0, ct_dynticks()); WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(c= urrent)); rcu_preempt_deferred_qs(current); =20 @@ -640,7 +639,7 @@ static noinstr void rcu_eqs_enter(bool user) instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); =20 instrumentation_end(); - WRITE_ONCE(rdp->dynticks_nesting, 0); /* Avoid irq-access tearing. */ + WRITE_ONCE(ct->dynticks_nesting, 0); /* Avoid irq-access tearing. */ // RCU is watching here ... rcu_dynticks_eqs_enter(); // ... but is no longer watching here. @@ -797,7 +796,7 @@ void rcu_irq_exit_check_preempt(void) { lockdep_assert_irqs_disabled(); =20 - RCU_LOCKDEP_WARN(__this_cpu_read(rcu_data.dynticks_nesting) <=3D 0, + RCU_LOCKDEP_WARN(ct_dynticks_nesting() <=3D 0, "RCU dynticks_nesting counter underflow/zero!"); RCU_LOCKDEP_WARN(__this_cpu_read(rcu_data.dynticks_nmi_nesting) !=3D DYNTICK_IRQ_NONIDLE, @@ -823,11 +822,11 @@ static void noinstr rcu_eqs_exit(bool user) =20 WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !raw_irqs_disabled()); rdp =3D this_cpu_ptr(&rcu_data); - oldval =3D rdp->dynticks_nesting; + oldval =3D ct_dynticks_nesting(); WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && oldval < 0); if (oldval) { // RCU was already watching, so just do accounting and leave. - rdp->dynticks_nesting++; + ct->dynticks_nesting++; return; } rcu_dynticks_task_exit(); @@ -839,9 +838,9 @@ static void noinstr rcu_eqs_exit(bool user) // instrumentation for the noinstr rcu_dynticks_eqs_exit() instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); =20 - trace_rcu_dyntick(TPS("End"), rdp->dynticks_nesting, 1, ct_dynticks()); + trace_rcu_dyntick(TPS("End"), ct_dynticks_nesting(), 1, ct_dynticks()); WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(c= urrent)); - WRITE_ONCE(rdp->dynticks_nesting, 1); + WRITE_ONCE(ct->dynticks_nesting, 1); WARN_ON_ONCE(rdp->dynticks_nmi_nesting); WRITE_ONCE(rdp->dynticks_nmi_nesting, DYNTICK_IRQ_NONIDLE); instrumentation_end(); @@ -4229,12 +4228,13 @@ static void rcu_init_new_rnp(struct rcu_node *rnp_l= eaf) static void __init rcu_boot_init_percpu_data(int cpu) { + struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); struct rcu_data *rdp =3D per_cpu_ptr(&rcu_data, cpu); =20 /* Set up local state, ensuring consistent view of global state. */ rdp->grpmask =3D leaf_node_cpu_bit(rdp->mynode, cpu); INIT_WORK(&rdp->strict_work, strict_work_handler); - WARN_ON_ONCE(rdp->dynticks_nesting !=3D 1); + WARN_ON_ONCE(ct->dynticks_nesting !=3D 1); WARN_ON_ONCE(rcu_dynticks_in_eqs(rcu_dynticks_snap(cpu))); rdp->barrier_seq_snap =3D rcu_state.barrier_sequence; rdp->rcu_ofl_gp_seq =3D rcu_state.gp_seq; @@ -4259,6 +4259,7 @@ rcu_boot_init_percpu_data(int cpu) int rcutree_prepare_cpu(unsigned int cpu) { unsigned long flags; + struct context_tracking *ct =3D per_cpu_ptr(&context_tracking, cpu); struct rcu_data *rdp =3D per_cpu_ptr(&rcu_data, cpu); struct rcu_node *rnp =3D rcu_get_root(); =20 @@ -4267,7 +4268,7 @@ int rcutree_prepare_cpu(unsigned int cpu) rdp->qlen_last_fqs_check =3D 0; rdp->n_force_qs_snap =3D READ_ONCE(rcu_state.n_force_qs); rdp->blimit =3D blimit; - rdp->dynticks_nesting =3D 1; /* CPU not up, no tearing. */ + ct->dynticks_nesting =3D 1; /* CPU not up, no tearing. */ raw_spin_unlock_rcu_node(rnp); /* irqs remain disabled. */ =20 /* diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h index 2e46ee108d13..642b11cd8027 100644 --- a/kernel/rcu/tree.h +++ b/kernel/rcu/tree.h @@ -191,7 +191,6 @@ struct rcu_data { =20 /* 3) dynticks interface. */ int dynticks_snap; /* Per-GP tracking for dynticks. */ - long dynticks_nesting; /* Track process nesting level. */ long dynticks_nmi_nesting; /* Track irq/NMI nesting level. */ bool rcu_need_heavy_qs; /* GP old, so heavy quiescent state! */ bool rcu_urgent_qs; /* GP old need light quiescent state. */ diff --git a/kernel/rcu/tree_stall.h b/kernel/rcu/tree_stall.h index 594a3c5aded1..b5195bbd9ebc 100644 --- a/kernel/rcu/tree_stall.h +++ b/kernel/rcu/tree_stall.h @@ -479,7 +479,7 @@ static void print_cpu_stall_info(int cpu) "!."[!delta], ticks_value, ticks_title, rcu_dynticks_snap(cpu) & 0xfff, - rdp->dynticks_nesting, rdp->dynticks_nmi_nesting, + ct_dynticks_nesting_cpu(cpu), rdp->dynticks_nmi_nesting, rdp->softirq_snap, kstat_softirqs_cpu(RCU_SOFTIRQ, cpu), data_race(rcu_state.n_force_qs) - rcu_state.n_force_qs_gpstart, rcuc_starved ? buf : "", --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 4CDF4C433EF for ; Wed, 8 Jun 2022 14:43:49 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S241880AbiFHOnr (ORCPT ); Wed, 8 Jun 2022 10:43:47 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59904 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241873AbiFHOmK (ORCPT ); Wed, 8 Jun 2022 10:42:10 -0400 Received: from dfw.source.kernel.org (dfw.source.kernel.org [139.178.84.217]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 4EAAB18C06A for ; Wed, 8 Jun 2022 07:41:37 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by dfw.source.kernel.org (Postfix) with ESMTPS id 4C29E61B9C for ; Wed, 8 Jun 2022 14:41:36 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 762D1C36AFF; Wed, 8 Jun 2022 14:41:32 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699295; bh=brhVaSfjXzn5ay7jCuchujDwc07f8/vcN+lzcYVVxLk=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=sXqs90bS2LZ9aKcPyPuonqxVw0Z7MH+dwfyuHgzwUDY50jywAUesaJ3t9GjDyaIAa tl4WBWc0UEpuJ5LrX0zQpbwa4OhK9DdmkOyhOiaRw16eIDd/myqt/iELH9NQ0kWmGO pcv5prDIyfkaoldDtgfl+w+MyowlicEylrllJl3ZZHLWzOyvxPf+3b1eJxow1F67Pf yuLnvxy62wsXXpYc4h2Mq7M68sdyxNx9NNI6teKO4FBPhnn7lcVUkm6+O2YxntqdYz CG97qhZkNgAJKxin6UHgGSo0qBSVBXmeNUUKsgQ5woWachrkuMcp2VnF7I0BtdMV9F uCLLetfjpTOtA== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 14/20] rcu/context_tracking: Move dynticks_nmi_nesting to context tracking Date: Wed, 8 Jun 2022 16:40:31 +0200 Message-Id: <20220608144037.1765000-15-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" The RCU eqs tracking is going to be performed by the context tracking subsystem. The related nesting counters thus need to be moved to the context tracking structure. Signed-off-by: Frederic Weisbecker Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- include/linux/context_tracking_state.h | 15 ++++++++ kernel/context_tracking.c | 1 + kernel/rcu/rcu.h | 4 --- kernel/rcu/tree.c | 48 +++++++++++--------------- kernel/rcu/tree.h | 1 - kernel/rcu/tree_stall.h | 2 +- 6 files changed, 38 insertions(+), 33 deletions(-) diff --git a/include/linux/context_tracking_state.h b/include/linux/context= _tracking_state.h index 164f4673ee74..b99e9827cc08 100644 --- a/include/linux/context_tracking_state.h +++ b/include/linux/context_tracking_state.h @@ -6,6 +6,9 @@ #include #include =20 +/* Offset to allow distinguishing irq vs. task-based idle entry/exit. */ +#define DYNTICK_IRQ_NONIDLE ((LONG_MAX / 2) + 1) + struct context_tracking { #ifdef CONFIG_CONTEXT_TRACKING_USER /* @@ -26,6 +29,7 @@ struct context_tracking { #ifdef CONFIG_CONTEXT_TRACKING_IDLE atomic_t dynticks; /* Even value for idle, else odd. */ long dynticks_nesting; /* Track process nesting level. */ + long dynticks_nmi_nesting; /* Track irq/NMI nesting level. */ #endif }; =20 @@ -61,6 +65,17 @@ static __always_inline long ct_dynticks_nesting_cpu(int = cpu) struct context_tracking *ct =3D per_cpu_ptr(&context_tracking, cpu); return ct->dynticks_nesting; } + +static __always_inline long ct_dynticks_nmi_nesting(void) +{ + return __this_cpu_read(context_tracking.dynticks_nmi_nesting); +} + +static __always_inline long ct_dynticks_nmi_nesting_cpu(int cpu) +{ + struct context_tracking *ct =3D per_cpu_ptr(&context_tracking, cpu); + return ct->dynticks_nmi_nesting; +} #endif /* #ifdef CONFIG_CONTEXT_TRACKING_IDLE */ =20 #ifdef CONFIG_CONTEXT_TRACKING_USER diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index d5ff1de1a24a..95ce83904287 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -26,6 +26,7 @@ DEFINE_PER_CPU(struct context_tracking, context_tracking) =3D { #ifdef CONFIG_CONTEXT_TRACKING_IDLE .dynticks_nesting =3D 1, + .dynticks_nmi_nesting =3D DYNTICK_IRQ_NONIDLE, .dynticks =3D ATOMIC_INIT(1), #endif }; diff --git a/kernel/rcu/rcu.h b/kernel/rcu/rcu.h index 15b96f990774..f86584fe90ba 100644 --- a/kernel/rcu/rcu.h +++ b/kernel/rcu/rcu.h @@ -12,10 +12,6 @@ =20 #include =20 -/* Offset to allow distinguishing irq vs. task-based idle entry/exit. */ -#define DYNTICK_IRQ_NONIDLE ((LONG_MAX / 2) + 1) - - /* * Grace-period counter management. */ diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index 5e7771c78162..ef734577b63d 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -75,7 +75,6 @@ /* Data structures. */ =20 static DEFINE_PER_CPU_SHARED_ALIGNED(struct rcu_data, rcu_data) =3D { - .dynticks_nmi_nesting =3D DYNTICK_IRQ_NONIDLE, #ifdef CONFIG_RCU_NOCB_CPU .cblist.flags =3D SEGCBLIST_RCU_CORE, #endif @@ -441,11 +440,11 @@ static int rcu_is_cpu_rrupt_from_idle(void) /* Check for counter underflows */ RCU_LOCKDEP_WARN(ct_dynticks_nesting() < 0, "RCU dynticks_nesting counter underflow!"); - RCU_LOCKDEP_WARN(__this_cpu_read(rcu_data.dynticks_nmi_nesting) <=3D 0, + RCU_LOCKDEP_WARN(ct_dynticks_nmi_nesting() <=3D 0, "RCU dynticks_nmi_nesting counter underflow/zero!"); =20 /* Are we at first interrupt nesting level? */ - nesting =3D __this_cpu_read(rcu_data.dynticks_nmi_nesting); + nesting =3D ct_dynticks_nmi_nesting(); if (nesting > 1) return false; =20 @@ -616,11 +615,10 @@ EXPORT_SYMBOL_GPL(rcutorture_get_gp_data); */ static noinstr void rcu_eqs_enter(bool user) { - struct rcu_data *rdp =3D this_cpu_ptr(&rcu_data); struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); =20 - WARN_ON_ONCE(rdp->dynticks_nmi_nesting !=3D DYNTICK_IRQ_NONIDLE); - WRITE_ONCE(rdp->dynticks_nmi_nesting, 0); + WARN_ON_ONCE(ct_dynticks_nmi_nesting() !=3D DYNTICK_IRQ_NONIDLE); + WRITE_ONCE(ct->dynticks_nmi_nesting, 0); WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && ct_dynticks_nesting() =3D=3D 0); if (ct_dynticks_nesting() !=3D 1) { @@ -738,7 +736,7 @@ noinstr void rcu_user_enter(void) * rcu_nmi_exit - inform RCU of exit from NMI context * * If we are returning from the outermost NMI handler that interrupted an - * RCU-idle period, update ct->dynticks and rdp->dynticks_nmi_nesting + * RCU-idle period, update ct->dynticks and ct->dynticks_nmi_nesting * to let the RCU grace-period handling know that the CPU is back to * being RCU-idle. * @@ -748,7 +746,6 @@ noinstr void rcu_user_enter(void) noinstr void rcu_nmi_exit(void) { struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); - struct rcu_data *rdp =3D this_cpu_ptr(&rcu_data); =20 instrumentation_begin(); /* @@ -756,25 +753,25 @@ noinstr void rcu_nmi_exit(void) * (We are exiting an NMI handler, so RCU better be paying attention * to us!) */ - WARN_ON_ONCE(rdp->dynticks_nmi_nesting <=3D 0); + WARN_ON_ONCE(ct_dynticks_nmi_nesting() <=3D 0); WARN_ON_ONCE(rcu_dynticks_curr_cpu_in_eqs()); =20 /* * If the nesting level is not 1, the CPU wasn't RCU-idle, so * leave it in non-RCU-idle state. */ - if (rdp->dynticks_nmi_nesting !=3D 1) { - trace_rcu_dyntick(TPS("--=3D"), rdp->dynticks_nmi_nesting, rdp->dynticks= _nmi_nesting - 2, + if (ct_dynticks_nmi_nesting() !=3D 1) { + trace_rcu_dyntick(TPS("--=3D"), ct_dynticks_nmi_nesting(), ct_dynticks_n= mi_nesting() - 2, ct_dynticks()); - WRITE_ONCE(rdp->dynticks_nmi_nesting, /* No store tearing. */ - rdp->dynticks_nmi_nesting - 2); + WRITE_ONCE(ct->dynticks_nmi_nesting, /* No store tearing. */ + ct_dynticks_nmi_nesting() - 2); instrumentation_end(); return; } =20 /* This NMI interrupted an RCU-idle CPU, restore RCU-idleness. */ - trace_rcu_dyntick(TPS("Startirq"), rdp->dynticks_nmi_nesting, 0, ct_dynti= cks()); - WRITE_ONCE(rdp->dynticks_nmi_nesting, 0); /* Avoid store tearing. */ + trace_rcu_dyntick(TPS("Startirq"), ct_dynticks_nmi_nesting(), 0, ct_dynti= cks()); + WRITE_ONCE(ct->dynticks_nmi_nesting, 0); /* Avoid store tearing. */ =20 // instrumentation for the noinstr rcu_dynticks_eqs_enter() instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); @@ -798,7 +795,7 @@ void rcu_irq_exit_check_preempt(void) =20 RCU_LOCKDEP_WARN(ct_dynticks_nesting() <=3D 0, "RCU dynticks_nesting counter underflow/zero!"); - RCU_LOCKDEP_WARN(__this_cpu_read(rcu_data.dynticks_nmi_nesting) !=3D + RCU_LOCKDEP_WARN(ct_dynticks_nmi_nesting() !=3D DYNTICK_IRQ_NONIDLE, "Bad RCU dynticks_nmi_nesting counter\n"); RCU_LOCKDEP_WARN(rcu_dynticks_curr_cpu_in_eqs(), @@ -817,11 +814,9 @@ void rcu_irq_exit_check_preempt(void) static void noinstr rcu_eqs_exit(bool user) { struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); - struct rcu_data *rdp; long oldval; =20 WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !raw_irqs_disabled()); - rdp =3D this_cpu_ptr(&rcu_data); oldval =3D ct_dynticks_nesting(); WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && oldval < 0); if (oldval) { @@ -841,8 +836,8 @@ static void noinstr rcu_eqs_exit(bool user) trace_rcu_dyntick(TPS("End"), ct_dynticks_nesting(), 1, ct_dynticks()); WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(c= urrent)); WRITE_ONCE(ct->dynticks_nesting, 1); - WARN_ON_ONCE(rdp->dynticks_nmi_nesting); - WRITE_ONCE(rdp->dynticks_nmi_nesting, DYNTICK_IRQ_NONIDLE); + WARN_ON_ONCE(ct_dynticks_nmi_nesting()); + WRITE_ONCE(ct->dynticks_nmi_nesting, DYNTICK_IRQ_NONIDLE); instrumentation_end(); } =20 @@ -945,7 +940,7 @@ void __rcu_irq_enter_check_tick(void) * rcu_nmi_enter - inform RCU of entry to NMI context * * If the CPU was idle from RCU's viewpoint, update ct->dynticks and - * rdp->dynticks_nmi_nesting to let the RCU grace-period handling know + * ct->dynticks_nmi_nesting to let the RCU grace-period handling know * that the CPU is active. This implementation permits nested NMIs, as * long as the nesting level does not overflow an int. (You will probably * run out of stack space first.) @@ -956,11 +951,10 @@ void __rcu_irq_enter_check_tick(void) noinstr void rcu_nmi_enter(void) { long incby =3D 2; - struct rcu_data *rdp =3D this_cpu_ptr(&rcu_data); struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); =20 /* Complain about underflow. */ - WARN_ON_ONCE(rdp->dynticks_nmi_nesting < 0); + WARN_ON_ONCE(ct_dynticks_nmi_nesting() < 0); =20 /* * If idle from RCU viewpoint, atomically increment ->dynticks @@ -994,11 +988,11 @@ noinstr void rcu_nmi_enter(void) } =20 trace_rcu_dyntick(incby =3D=3D 1 ? TPS("Endirq") : TPS("++=3D"), - rdp->dynticks_nmi_nesting, - rdp->dynticks_nmi_nesting + incby, ct_dynticks()); + ct_dynticks_nmi_nesting(), + ct_dynticks_nmi_nesting() + incby, ct_dynticks()); instrumentation_end(); - WRITE_ONCE(rdp->dynticks_nmi_nesting, /* Prevent store tearing. */ - rdp->dynticks_nmi_nesting + incby); + WRITE_ONCE(ct->dynticks_nmi_nesting, /* Prevent store tearing. */ + ct_dynticks_nmi_nesting() + incby); barrier(); } =20 diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h index 642b11cd8027..12e249fbaf1e 100644 --- a/kernel/rcu/tree.h +++ b/kernel/rcu/tree.h @@ -191,7 +191,6 @@ struct rcu_data { =20 /* 3) dynticks interface. */ int dynticks_snap; /* Per-GP tracking for dynticks. */ - long dynticks_nmi_nesting; /* Track irq/NMI nesting level. */ bool rcu_need_heavy_qs; /* GP old, so heavy quiescent state! */ bool rcu_urgent_qs; /* GP old need light quiescent state. */ bool rcu_forced_tick; /* Forced tick to provide QS. */ diff --git a/kernel/rcu/tree_stall.h b/kernel/rcu/tree_stall.h index b5195bbd9ebc..91e4fd4db12d 100644 --- a/kernel/rcu/tree_stall.h +++ b/kernel/rcu/tree_stall.h @@ -479,7 +479,7 @@ static void print_cpu_stall_info(int cpu) "!."[!delta], ticks_value, ticks_title, rcu_dynticks_snap(cpu) & 0xfff, - ct_dynticks_nesting_cpu(cpu), rdp->dynticks_nmi_nesting, + ct_dynticks_nesting_cpu(cpu), ct_dynticks_nmi_nesting_cpu(cpu), rdp->softirq_snap, kstat_softirqs_cpu(RCU_SOFTIRQ, cpu), data_race(rcu_state.n_force_qs) - rcu_state.n_force_qs_gpstart, rcuc_starved ? buf : "", --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id CAEF7C43334 for ; Wed, 8 Jun 2022 14:44:05 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S242092AbiFHOoB (ORCPT ); Wed, 8 Jun 2022 10:44:01 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:34696 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241923AbiFHOmP (ORCPT ); Wed, 8 Jun 2022 10:42:15 -0400 Received: from dfw.source.kernel.org (dfw.source.kernel.org [IPv6:2604:1380:4641:c500::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 88C8818F85B for ; Wed, 8 Jun 2022 07:41:40 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by dfw.source.kernel.org (Postfix) with ESMTPS id E565161B96 for ; Wed, 8 Jun 2022 14:41:39 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 26DADC3411D; Wed, 8 Jun 2022 14:41:35 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699299; bh=thLW1AdTFUGR2qp/zVtupKcplrWYxLzE4qfZIs11S9o=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=UXMd76+gq41l5m27KnPwIxsF/GEKVechCLIJAg/3KanqL9j0uYerjSneewQKU5RWv K0NQkaFpV7pp0Z9v/r/ARL+KOwlx3eoc0yfig0MdHyth7CwgqXV/mHUcErnyHi4ez0 xyN+fv/H//G5EjJ5AtNtEBTIGKgQRnEvJdvJ1R0zE9w9DVrV9YPVmoElSJh4fJS2oL A++zsqS5RmNBCTpt5V7ZxyZuBlfAndKcNWjzJmZvFAmpaqIrQhl9QoNfwMZjpxhphc UAbyXzMzkGQHk4/ObljyKhqEWkQ3KLxq08ybKT3Hc8W4JGz6t6CZQjq9MecllNVmcf aIsFNS3i6nlGw== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 15/20] rcu/context-tracking: Move deferred nocb resched to context tracking Date: Wed, 8 Jun 2022 16:40:32 +0200 Message-Id: <20220608144037.1765000-16-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" To prepare for migrating the RCU eqs accounting code to context tracking, split the last-resort deferred nocb resched from rcu_user_enter() and move it into a separate call from context tracking. Acked-by: Paul E. McKenney Signed-off-by: Frederic Weisbecker Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- include/linux/rcupdate.h | 6 ++++++ kernel/context_tracking.c | 8 ++++++++ kernel/rcu/tree.c | 15 ++------------- 3 files changed, 16 insertions(+), 13 deletions(-) diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h index 910328f2f51a..1d120dca6548 100644 --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -113,6 +113,12 @@ static inline void rcu_user_enter(void) { } static inline void rcu_user_exit(void) { } #endif /* CONFIG_NO_HZ_FULL */ =20 +#if defined(CONFIG_NO_HZ_FULL) && (!defined(CONFIG_GENERIC_ENTRY) || !defi= ned(CONFIG_KVM_XFER_TO_GUEST_WORK)) +void rcu_irq_work_resched(void); +#else +static inline void rcu_irq_work_resched(void) { } +#endif + #ifdef CONFIG_RCU_NOCB_CPU void rcu_init_nohz(void); int rcu_nocb_cpu_offload(int cpu); diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index 95ce83904287..d782c868f36f 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -177,6 +177,8 @@ static __always_inline void context_tracking_recursion_= exit(void) */ void noinstr __ct_user_enter(enum ctx_state state) { + lockdep_assert_irqs_disabled(); + /* Kernel threads aren't supposed to go to userspace */ WARN_ON_ONCE(!current->mm); =20 @@ -198,6 +200,12 @@ void noinstr __ct_user_enter(enum ctx_state state) vtime_user_enter(current); instrumentation_end(); } + /* + * Other than generic entry implementation, we may be past the last + * rescheduling opportunity in the entry code. Trigger a self IPI + * that will fire and reschedule once we resume in user/guest mode. + */ + rcu_irq_work_resched(); rcu_user_enter(); } /* diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index ef734577b63d..beddb4fb5d07 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -685,7 +685,7 @@ static DEFINE_PER_CPU(struct irq_work, late_wakeup_work= ) =3D * last resort is to fire a local irq_work that will trigger a reschedule = once IRQs * get re-enabled again. */ -noinstr static void rcu_irq_work_resched(void) +noinstr void rcu_irq_work_resched(void) { struct rcu_data *rdp =3D this_cpu_ptr(&rcu_data); =20 @@ -701,10 +701,7 @@ noinstr static void rcu_irq_work_resched(void) } instrumentation_end(); } - -#else -static inline void rcu_irq_work_resched(void) { } -#endif +#endif /* #if !defined(CONFIG_GENERIC_ENTRY) || !defined(CONFIG_KVM_XFER_T= O_GUEST_WORK) */ =20 /** * rcu_user_enter - inform RCU that we are resuming userspace. @@ -719,14 +716,6 @@ static inline void rcu_irq_work_resched(void) { } */ noinstr void rcu_user_enter(void) { - lockdep_assert_irqs_disabled(); - - /* - * Other than generic entry implementation, we may be past the last - * rescheduling opportunity in the entry code. Trigger a self IPI - * that will fire and reschedule once we resume in user/guest mode. - */ - rcu_irq_work_resched(); rcu_eqs_enter(true); } =20 --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id A6FAFC43334 for ; Wed, 8 Jun 2022 14:42:36 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S241774AbiFHOme (ORCPT ); Wed, 8 Jun 2022 10:42:34 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:33102 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241956AbiFHOmP (ORCPT ); Wed, 8 Jun 2022 10:42:15 -0400 Received: from ams.source.kernel.org (ams.source.kernel.org [145.40.68.75]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id B7598179C17 for ; Wed, 8 Jun 2022 07:41:45 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ams.source.kernel.org (Postfix) with ESMTPS id 35FAAB8284B for ; Wed, 8 Jun 2022 14:41:44 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id C9481C3411F; Wed, 8 Jun 2022 14:41:39 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699303; bh=Y6eipDC4/0BUrPTdpSOYE3HZLlVUBNRbpbQ8YfEd5q0=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=FDc8Oa1s8dRZn77kIvMAvJV0wRYMmC1D+JOKODjuSGqGqHp6Uqy6+2HbfHvj4tmzU WAVpN5dpntweRoaew2CJCp4eIGW1BY74aHZ7DhL4NxGPYFf1s9BizGtFSsV5xucRgX lGdjHpkJTFE45aZcjLKX9p96eZ+NgrpGiG1dR/pA2iugyAimsjLUsWZuhUWssJOINg aI6doz+0DyDBGX/VDXm64Fd3yQr2btt0Bb3JE3joKcWe6Jr3fr4Q4Gpkov6VCPf644 5oa5o7cno+5qJDvu491fnLcAnBPH21nE0IR9hQmg6I7QmIG5aFNivyavhevuabHBeu wRq506B0wNkCA== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 16/20] rcu/context-tracking: Move RCU-dynticks internal functions to context_tracking Date: Wed, 8 Jun 2022 16:40:33 +0200 Message-Id: <20220608144037.1765000-17-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Move the core RCU eqs/dynticks functions to context tracking so that we can later merge all that code within context tracking. Acked-by: Paul E. McKenney Signed-off-by: Frederic Weisbecker Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- include/linux/context_tracking.h | 20 ++ include/linux/rcutree.h | 3 + kernel/context_tracking.c | 336 +++++++++++++++++++++++++++++++ kernel/rcu/tree.c | 324 +---------------------------- kernel/rcu/tree.h | 5 - kernel/rcu/tree_plugin.h | 38 +--- 6 files changed, 364 insertions(+), 362 deletions(-) diff --git a/include/linux/context_tracking.h b/include/linux/context_track= ing.h index 01abadb2f993..1f568676bc1d 100644 --- a/include/linux/context_tracking.h +++ b/include/linux/context_tracking.h @@ -122,6 +122,26 @@ static inline void context_tracking_init(void) { } #ifdef CONFIG_CONTEXT_TRACKING_IDLE extern void ct_idle_enter(void); extern void ct_idle_exit(void); + +/* + * Is the current CPU in an extended quiescent state? + * + * No ordering, as we are sampling CPU-local information. + */ +static __always_inline bool rcu_dynticks_curr_cpu_in_eqs(void) +{ + return !(arch_atomic_read(this_cpu_ptr(&context_tracking.dynticks)) & 0x1= ); +} + +/* + * Increment the current CPU's context_tracking structure's ->dynticks fie= ld + * with ordering. Return the new value. + */ +static __always_inline unsigned long rcu_dynticks_inc(int incby) +{ + return arch_atomic_add_return(incby, this_cpu_ptr(&context_tracking.dynti= cks)); +} + #else static inline void ct_idle_enter(void) { } static inline void ct_idle_exit(void) { } diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h index cafe3fbf4272..c31e4e526643 100644 --- a/include/linux/rcutree.h +++ b/include/linux/rcutree.h @@ -57,6 +57,9 @@ void rcu_irq_exit_check_preempt(void); static inline void rcu_irq_exit_check_preempt(void) { } #endif =20 +struct task_struct; +void rcu_preempt_deferred_qs(struct task_struct *t); + void exit_rcu(void); =20 void rcu_scheduler_starting(void); diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index d782c868f36f..bb64c1ed5ab2 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -21,6 +21,7 @@ #include #include #include +#include =20 =20 DEFINE_PER_CPU(struct context_tracking, context_tracking) =3D { @@ -33,6 +34,309 @@ DEFINE_PER_CPU(struct context_tracking, context_trackin= g) =3D { EXPORT_SYMBOL_GPL(context_tracking); =20 #ifdef CONFIG_CONTEXT_TRACKING_IDLE +#define TPS(x) tracepoint_string(x) + +/* Record the current task on dyntick-idle entry. */ +static __always_inline void rcu_dynticks_task_enter(void) +{ +#if defined(CONFIG_TASKS_RCU) && defined(CONFIG_NO_HZ_FULL) + WRITE_ONCE(current->rcu_tasks_idle_cpu, smp_processor_id()); +#endif /* #if defined(CONFIG_TASKS_RCU) && defined(CONFIG_NO_HZ_FULL) */ +} + +/* Record no current task on dyntick-idle exit. */ +static __always_inline void rcu_dynticks_task_exit(void) +{ +#if defined(CONFIG_TASKS_RCU) && defined(CONFIG_NO_HZ_FULL) + WRITE_ONCE(current->rcu_tasks_idle_cpu, -1); +#endif /* #if defined(CONFIG_TASKS_RCU) && defined(CONFIG_NO_HZ_FULL) */ +} + +/* Turn on heavyweight RCU tasks trace readers on idle/user entry. */ +static __always_inline void rcu_dynticks_task_trace_enter(void) +{ +#ifdef CONFIG_TASKS_TRACE_RCU + if (IS_ENABLED(CONFIG_TASKS_TRACE_RCU_READ_MB)) + current->trc_reader_special.b.need_mb =3D true; +#endif /* #ifdef CONFIG_TASKS_TRACE_RCU */ +} + +/* Turn off heavyweight RCU tasks trace readers on idle/user exit. */ +static __always_inline void rcu_dynticks_task_trace_exit(void) +{ +#ifdef CONFIG_TASKS_TRACE_RCU + if (IS_ENABLED(CONFIG_TASKS_TRACE_RCU_READ_MB)) + current->trc_reader_special.b.need_mb =3D false; +#endif /* #ifdef CONFIG_TASKS_TRACE_RCU */ +} + +/* + * Record entry into an extended quiescent state. This is only to be + * called when not already in an extended quiescent state, that is, + * RCU is watching prior to the call to this function and is no longer + * watching upon return. + */ +static noinstr void rcu_dynticks_eqs_enter(void) +{ + int seq; + + /* + * CPUs seeing atomic_add_return() must see prior RCU read-side + * critical sections, and we also must force ordering with the + * next idle sojourn. + */ + rcu_dynticks_task_trace_enter(); // Before ->dynticks update! + seq =3D rcu_dynticks_inc(1); + // RCU is no longer watching. Better be in extended quiescent state! + WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && (seq & 0x1)); +} + +/* + * Record exit from an extended quiescent state. This is only to be + * called from an extended quiescent state, that is, RCU is not watching + * prior to the call to this function and is watching upon return. + */ +static noinstr void rcu_dynticks_eqs_exit(void) +{ + int seq; + + /* + * CPUs seeing atomic_add_return() must see prior idle sojourns, + * and we also must force ordering with the next RCU read-side + * critical section. + */ + seq =3D rcu_dynticks_inc(1); + // RCU is now watching. Better not be in an extended quiescent state! + rcu_dynticks_task_trace_exit(); // After ->dynticks update! + WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !(seq & 0x1)); +} + +/* + * Enter an RCU extended quiescent state, which can be either the + * idle loop or adaptive-tickless usermode execution. + * + * We crowbar the ->dynticks_nmi_nesting field to zero to allow for + * the possibility of usermode upcalls having messed up our count + * of interrupt nesting level during the prior busy period. + */ +static void noinstr rcu_eqs_enter(bool user) +{ + struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); + + WARN_ON_ONCE(ct_dynticks_nmi_nesting() !=3D DYNTICK_IRQ_NONIDLE); + WRITE_ONCE(ct->dynticks_nmi_nesting, 0); + WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && + ct_dynticks_nesting() =3D=3D 0); + if (ct_dynticks_nesting() !=3D 1) { + // RCU will still be watching, so just do accounting and leave. + ct->dynticks_nesting--; + return; + } + + instrumentation_begin(); + lockdep_assert_irqs_disabled(); + trace_rcu_dyntick(TPS("Start"), ct_dynticks_nesting(), 0, ct_dynticks()); + WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(c= urrent)); + rcu_preempt_deferred_qs(current); + + // instrumentation for the noinstr rcu_dynticks_eqs_enter() + instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); + + instrumentation_end(); + WRITE_ONCE(ct->dynticks_nesting, 0); /* Avoid irq-access tearing. */ + // RCU is watching here ... + rcu_dynticks_eqs_enter(); + // ... but is no longer watching here. + rcu_dynticks_task_enter(); +} + +/* + * Exit an RCU extended quiescent state, which can be either the + * idle loop or adaptive-tickless usermode execution. + * + * We crowbar the ->dynticks_nmi_nesting field to DYNTICK_IRQ_NONIDLE to + * allow for the possibility of usermode upcalls messing up our count of + * interrupt nesting level during the busy period that is just now startin= g. + */ +static void noinstr rcu_eqs_exit(bool user) +{ + struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); + long oldval; + + WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !raw_irqs_disabled()); + oldval =3D ct_dynticks_nesting(); + WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && oldval < 0); + if (oldval) { + // RCU was already watching, so just do accounting and leave. + ct->dynticks_nesting++; + return; + } + rcu_dynticks_task_exit(); + // RCU is not watching here ... + rcu_dynticks_eqs_exit(); + // ... but is watching here. + instrumentation_begin(); + + // instrumentation for the noinstr rcu_dynticks_eqs_exit() + instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); + + trace_rcu_dyntick(TPS("End"), ct_dynticks_nesting(), 1, ct_dynticks()); + WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(c= urrent)); + WRITE_ONCE(ct->dynticks_nesting, 1); + WARN_ON_ONCE(ct_dynticks_nmi_nesting()); + WRITE_ONCE(ct->dynticks_nmi_nesting, DYNTICK_IRQ_NONIDLE); + instrumentation_end(); +} + +/** + * rcu_nmi_exit - inform RCU of exit from NMI context + * + * If we are returning from the outermost NMI handler that interrupted an + * RCU-idle period, update ct->dynticks and ct->dynticks_nmi_nesting + * to let the RCU grace-period handling know that the CPU is back to + * being RCU-idle. + * + * If you add or remove a call to rcu_nmi_exit(), be sure to test + * with CONFIG_RCU_EQS_DEBUG=3Dy. + */ +void noinstr rcu_nmi_exit(void) +{ + struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); + + instrumentation_begin(); + /* + * Check for ->dynticks_nmi_nesting underflow and bad ->dynticks. + * (We are exiting an NMI handler, so RCU better be paying attention + * to us!) + */ + WARN_ON_ONCE(ct_dynticks_nmi_nesting() <=3D 0); + WARN_ON_ONCE(rcu_dynticks_curr_cpu_in_eqs()); + + /* + * If the nesting level is not 1, the CPU wasn't RCU-idle, so + * leave it in non-RCU-idle state. + */ + if (ct_dynticks_nmi_nesting() !=3D 1) { + trace_rcu_dyntick(TPS("--=3D"), ct_dynticks_nmi_nesting(), ct_dynticks_n= mi_nesting() - 2, + ct_dynticks()); + WRITE_ONCE(ct->dynticks_nmi_nesting, /* No store tearing. */ + ct_dynticks_nmi_nesting() - 2); + instrumentation_end(); + return; + } + + /* This NMI interrupted an RCU-idle CPU, restore RCU-idleness. */ + trace_rcu_dyntick(TPS("Startirq"), ct_dynticks_nmi_nesting(), 0, ct_dynti= cks()); + WRITE_ONCE(ct->dynticks_nmi_nesting, 0); /* Avoid store tearing. */ + + // instrumentation for the noinstr rcu_dynticks_eqs_enter() + instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); + instrumentation_end(); + + // RCU is watching here ... + rcu_dynticks_eqs_enter(); + // ... but is no longer watching here. + + if (!in_nmi()) + rcu_dynticks_task_enter(); +} + +/** + * rcu_nmi_enter - inform RCU of entry to NMI context + * + * If the CPU was idle from RCU's viewpoint, update ct->dynticks and + * ct->dynticks_nmi_nesting to let the RCU grace-period handling know + * that the CPU is active. This implementation permits nested NMIs, as + * long as the nesting level does not overflow an int. (You will probably + * run out of stack space first.) + * + * If you add or remove a call to rcu_nmi_enter(), be sure to test + * with CONFIG_RCU_EQS_DEBUG=3Dy. + */ +void noinstr rcu_nmi_enter(void) +{ + long incby =3D 2; + struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); + + /* Complain about underflow. */ + WARN_ON_ONCE(ct_dynticks_nmi_nesting() < 0); + + /* + * If idle from RCU viewpoint, atomically increment ->dynticks + * to mark non-idle and increment ->dynticks_nmi_nesting by one. + * Otherwise, increment ->dynticks_nmi_nesting by two. This means + * if ->dynticks_nmi_nesting is equal to one, we are guaranteed + * to be in the outermost NMI handler that interrupted an RCU-idle + * period (observation due to Andy Lutomirski). + */ + if (rcu_dynticks_curr_cpu_in_eqs()) { + + if (!in_nmi()) + rcu_dynticks_task_exit(); + + // RCU is not watching here ... + rcu_dynticks_eqs_exit(); + // ... but is watching here. + + instrumentation_begin(); + // instrumentation for the noinstr rcu_dynticks_curr_cpu_in_eqs() + instrument_atomic_read(&ct->dynticks, sizeof(ct->dynticks)); + // instrumentation for the noinstr rcu_dynticks_eqs_exit() + instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); + + incby =3D 1; + } else if (!in_nmi()) { + instrumentation_begin(); + rcu_irq_enter_check_tick(); + } else { + instrumentation_begin(); + } + + trace_rcu_dyntick(incby =3D=3D 1 ? TPS("Endirq") : TPS("++=3D"), + ct_dynticks_nmi_nesting(), + ct_dynticks_nmi_nesting() + incby, ct_dynticks()); + instrumentation_end(); + WRITE_ONCE(ct->dynticks_nmi_nesting, /* Prevent store tearing. */ + ct_dynticks_nmi_nesting() + incby); + barrier(); +} + +/** + * rcu_idle_enter - inform RCU that current CPU is entering idle + * + * Enter idle mode, in other words, -leave- the mode in which RCU + * read-side critical sections can occur. (Though RCU read-side + * critical sections can occur in irq handlers in idle, a possibility + * handled by irq_enter() and irq_exit().) + * + * If you add or remove a call to rcu_idle_enter(), be sure to test with + * CONFIG_RCU_EQS_DEBUG=3Dy. + */ +void noinstr rcu_idle_enter(void) +{ + WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !raw_irqs_disabled()); + rcu_eqs_enter(false); +} + +/** + * rcu_idle_exit - inform RCU that current CPU is leaving idle + * + * Exit idle mode, in other words, -enter- the mode in which RCU + * read-side critical sections can occur. + * + * If you add or remove a call to rcu_idle_exit(), be sure to test with + * CONFIG_RCU_EQS_DEBUG=3Dy. + */ +void noinstr rcu_idle_exit(void) +{ + unsigned long flags; + + raw_local_irq_save(flags); + rcu_eqs_exit(false); + raw_local_irq_restore(flags); +} +EXPORT_SYMBOL_GPL(rcu_idle_exit); + noinstr void ct_idle_enter(void) { rcu_idle_enter(); @@ -139,6 +443,38 @@ noinstr void ct_nmi_exit(void) } #endif /* #ifdef CONFIG_CONTEXT_TRACKING_IDLE */ =20 +#ifdef CONFIG_NO_HZ_FULL +/** + * rcu_user_enter - inform RCU that we are resuming userspace. + * + * Enter RCU idle mode right before resuming userspace. No use of RCU + * is permitted between this call and rcu_user_exit(). This way the + * CPU doesn't need to maintain the tick for RCU maintenance purposes + * when the CPU runs in userspace. + * + * If you add or remove a call to rcu_user_enter(), be sure to test with + * CONFIG_RCU_EQS_DEBUG=3Dy. + */ +noinstr void rcu_user_enter(void) +{ + rcu_eqs_enter(true); +} + +/** + * rcu_user_exit - inform RCU that we are exiting userspace. + * + * Exit RCU idle mode while entering the kernel because it can + * run a RCU read side critical section anytime. + * + * If you add or remove a call to rcu_user_exit(), be sure to test with + * CONFIG_RCU_EQS_DEBUG=3Dy. + */ +void noinstr rcu_user_exit(void) +{ + rcu_eqs_exit(true); +} +#endif /* #ifdef CONFIG_NO_HZ_FULL */ + #ifdef CONFIG_CONTEXT_TRACKING_USER =20 #define CREATE_TRACE_POINTS diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index beddb4fb5d07..155e8ce3d267 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -62,6 +62,7 @@ #include #include #include +#include #include "../time/tick-internal.h" =20 #include "tree.h" @@ -263,56 +264,6 @@ void rcu_softirq_qs(void) rcu_tasks_qs(current, false); } =20 -/* - * Increment the current CPU's rcu_data structure's ->dynticks field - * with ordering. Return the new value. - */ -static noinline noinstr unsigned long rcu_dynticks_inc(int incby) -{ - return arch_atomic_add_return(incby, this_cpu_ptr(&context_tracking.dynti= cks)); -} - -/* - * Record entry into an extended quiescent state. This is only to be - * called when not already in an extended quiescent state, that is, - * RCU is watching prior to the call to this function and is no longer - * watching upon return. - */ -static noinstr void rcu_dynticks_eqs_enter(void) -{ - int seq; - - /* - * CPUs seeing atomic_add_return() must see prior RCU read-side - * critical sections, and we also must force ordering with the - * next idle sojourn. - */ - rcu_dynticks_task_trace_enter(); // Before ->dynticks update! - seq =3D rcu_dynticks_inc(1); - // RCU is no longer watching. Better be in extended quiescent state! - WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && (seq & 0x1)); -} - -/* - * Record exit from an extended quiescent state. This is only to be - * called from an extended quiescent state, that is, RCU is not watching - * prior to the call to this function and is watching upon return. - */ -static noinstr void rcu_dynticks_eqs_exit(void) -{ - int seq; - - /* - * CPUs seeing atomic_add_return() must see prior idle sojourns, - * and we also must force ordering with the next RCU read-side - * critical section. - */ - seq =3D rcu_dynticks_inc(1); - // RCU is now watching. Better not be in an extended quiescent state! - rcu_dynticks_task_trace_exit(); // After ->dynticks update! - WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !(seq & 0x1)); -} - /* * Reset the current CPU's ->dynticks counter to indicate that the * newly onlined CPU is no longer in an extended quiescent state. @@ -330,16 +281,6 @@ static void rcu_dynticks_eqs_online(void) rcu_dynticks_inc(1); } =20 -/* - * Is the current CPU in an extended quiescent state? - * - * No ordering, as we are sampling CPU-local information. - */ -static __always_inline bool rcu_dynticks_curr_cpu_in_eqs(void) -{ - return !(arch_atomic_read(this_cpu_ptr(&context_tracking.dynticks)) & 0x1= ); -} - /* * Snapshot the ->dynticks counter with full ordering so as to allow * stable comparison of this counter with past and future snapshots. @@ -605,65 +546,7 @@ void rcutorture_get_gp_data(enum rcutorture_type test_= type, int *flags, } EXPORT_SYMBOL_GPL(rcutorture_get_gp_data); =20 -/* - * Enter an RCU extended quiescent state, which can be either the - * idle loop or adaptive-tickless usermode execution. - * - * We crowbar the ->dynticks_nmi_nesting field to zero to allow for - * the possibility of usermode upcalls having messed up our count - * of interrupt nesting level during the prior busy period. - */ -static noinstr void rcu_eqs_enter(bool user) -{ - struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); - - WARN_ON_ONCE(ct_dynticks_nmi_nesting() !=3D DYNTICK_IRQ_NONIDLE); - WRITE_ONCE(ct->dynticks_nmi_nesting, 0); - WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && - ct_dynticks_nesting() =3D=3D 0); - if (ct_dynticks_nesting() !=3D 1) { - // RCU will still be watching, so just do accounting and leave. - ct->dynticks_nesting--; - return; - } - - instrumentation_begin(); - lockdep_assert_irqs_disabled(); - trace_rcu_dyntick(TPS("Start"), ct_dynticks_nesting(), 0, ct_dynticks()); - WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(c= urrent)); - rcu_preempt_deferred_qs(current); - - // instrumentation for the noinstr rcu_dynticks_eqs_enter() - instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); - - instrumentation_end(); - WRITE_ONCE(ct->dynticks_nesting, 0); /* Avoid irq-access tearing. */ - // RCU is watching here ... - rcu_dynticks_eqs_enter(); - // ... but is no longer watching here. - rcu_dynticks_task_enter(); -} - -/** - * rcu_idle_enter - inform RCU that current CPU is entering idle - * - * Enter idle mode, in other words, -leave- the mode in which RCU - * read-side critical sections can occur. (Though RCU read-side - * critical sections can occur in irq handlers in idle, a possibility - * handled by irq_enter() and irq_exit().) - * - * If you add or remove a call to rcu_idle_enter(), be sure to test with - * CONFIG_RCU_EQS_DEBUG=3Dy. - */ -void noinstr rcu_idle_enter(void) -{ - WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !raw_irqs_disabled()); - rcu_eqs_enter(false); -} - -#ifdef CONFIG_NO_HZ_FULL - -#if !defined(CONFIG_GENERIC_ENTRY) || !defined(CONFIG_KVM_XFER_TO_GUEST_WO= RK) +#if defined(CONFIG_NO_HZ_FULL) && (!defined(CONFIG_GENERIC_ENTRY) || !defi= ned(CONFIG_KVM_XFER_TO_GUEST_WORK)) /* * An empty function that will trigger a reschedule on * IRQ tail once IRQs get re-enabled on userspace/guest resume. @@ -701,78 +584,7 @@ noinstr void rcu_irq_work_resched(void) } instrumentation_end(); } -#endif /* #if !defined(CONFIG_GENERIC_ENTRY) || !defined(CONFIG_KVM_XFER_T= O_GUEST_WORK) */ - -/** - * rcu_user_enter - inform RCU that we are resuming userspace. - * - * Enter RCU idle mode right before resuming userspace. No use of RCU - * is permitted between this call and rcu_user_exit(). This way the - * CPU doesn't need to maintain the tick for RCU maintenance purposes - * when the CPU runs in userspace. - * - * If you add or remove a call to rcu_user_enter(), be sure to test with - * CONFIG_RCU_EQS_DEBUG=3Dy. - */ -noinstr void rcu_user_enter(void) -{ - rcu_eqs_enter(true); -} - -#endif /* CONFIG_NO_HZ_FULL */ - -/** - * rcu_nmi_exit - inform RCU of exit from NMI context - * - * If we are returning from the outermost NMI handler that interrupted an - * RCU-idle period, update ct->dynticks and ct->dynticks_nmi_nesting - * to let the RCU grace-period handling know that the CPU is back to - * being RCU-idle. - * - * If you add or remove a call to rcu_nmi_exit(), be sure to test - * with CONFIG_RCU_EQS_DEBUG=3Dy. - */ -noinstr void rcu_nmi_exit(void) -{ - struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); - - instrumentation_begin(); - /* - * Check for ->dynticks_nmi_nesting underflow and bad ->dynticks. - * (We are exiting an NMI handler, so RCU better be paying attention - * to us!) - */ - WARN_ON_ONCE(ct_dynticks_nmi_nesting() <=3D 0); - WARN_ON_ONCE(rcu_dynticks_curr_cpu_in_eqs()); - - /* - * If the nesting level is not 1, the CPU wasn't RCU-idle, so - * leave it in non-RCU-idle state. - */ - if (ct_dynticks_nmi_nesting() !=3D 1) { - trace_rcu_dyntick(TPS("--=3D"), ct_dynticks_nmi_nesting(), ct_dynticks_n= mi_nesting() - 2, - ct_dynticks()); - WRITE_ONCE(ct->dynticks_nmi_nesting, /* No store tearing. */ - ct_dynticks_nmi_nesting() - 2); - instrumentation_end(); - return; - } - - /* This NMI interrupted an RCU-idle CPU, restore RCU-idleness. */ - trace_rcu_dyntick(TPS("Startirq"), ct_dynticks_nmi_nesting(), 0, ct_dynti= cks()); - WRITE_ONCE(ct->dynticks_nmi_nesting, 0); /* Avoid store tearing. */ - - // instrumentation for the noinstr rcu_dynticks_eqs_enter() - instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); - instrumentation_end(); - - // RCU is watching here ... - rcu_dynticks_eqs_enter(); - // ... but is no longer watching here. - - if (!in_nmi()) - rcu_dynticks_task_enter(); -} +#endif /* #if defined(CONFIG_NO_HZ_FULL) && (!defined(CONFIG_GENERIC_ENTRY= ) || !defined(CONFIG_KVM_XFER_TO_GUEST_WORK)) */ =20 #ifdef CONFIG_PROVE_RCU /** @@ -792,77 +604,7 @@ void rcu_irq_exit_check_preempt(void) } #endif /* #ifdef CONFIG_PROVE_RCU */ =20 -/* - * Exit an RCU extended quiescent state, which can be either the - * idle loop or adaptive-tickless usermode execution. - * - * We crowbar the ->dynticks_nmi_nesting field to DYNTICK_IRQ_NONIDLE to - * allow for the possibility of usermode upcalls messing up our count of - * interrupt nesting level during the busy period that is just now startin= g. - */ -static void noinstr rcu_eqs_exit(bool user) -{ - struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); - long oldval; - - WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !raw_irqs_disabled()); - oldval =3D ct_dynticks_nesting(); - WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && oldval < 0); - if (oldval) { - // RCU was already watching, so just do accounting and leave. - ct->dynticks_nesting++; - return; - } - rcu_dynticks_task_exit(); - // RCU is not watching here ... - rcu_dynticks_eqs_exit(); - // ... but is watching here. - instrumentation_begin(); - - // instrumentation for the noinstr rcu_dynticks_eqs_exit() - instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); - - trace_rcu_dyntick(TPS("End"), ct_dynticks_nesting(), 1, ct_dynticks()); - WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(c= urrent)); - WRITE_ONCE(ct->dynticks_nesting, 1); - WARN_ON_ONCE(ct_dynticks_nmi_nesting()); - WRITE_ONCE(ct->dynticks_nmi_nesting, DYNTICK_IRQ_NONIDLE); - instrumentation_end(); -} - -/** - * rcu_idle_exit - inform RCU that current CPU is leaving idle - * - * Exit idle mode, in other words, -enter- the mode in which RCU - * read-side critical sections can occur. - * - * If you add or remove a call to rcu_idle_exit(), be sure to test with - * CONFIG_RCU_EQS_DEBUG=3Dy. - */ -void noinstr rcu_idle_exit(void) -{ - unsigned long flags; - - raw_local_irq_save(flags); - rcu_eqs_exit(false); - raw_local_irq_restore(flags); -} - #ifdef CONFIG_NO_HZ_FULL -/** - * rcu_user_exit - inform RCU that we are exiting userspace. - * - * Exit RCU idle mode while entering the kernel because it can - * run a RCU read side critical section anytime. - * - * If you add or remove a call to rcu_user_exit(), be sure to test with - * CONFIG_RCU_EQS_DEBUG=3Dy. - */ -void noinstr rcu_user_exit(void) -{ - rcu_eqs_exit(true); -} - /** * __rcu_irq_enter_check_tick - Enable scheduler tick on CPU if RCU needs = it. * @@ -925,66 +667,6 @@ void __rcu_irq_enter_check_tick(void) } #endif /* CONFIG_NO_HZ_FULL */ =20 -/** - * rcu_nmi_enter - inform RCU of entry to NMI context - * - * If the CPU was idle from RCU's viewpoint, update ct->dynticks and - * ct->dynticks_nmi_nesting to let the RCU grace-period handling know - * that the CPU is active. This implementation permits nested NMIs, as - * long as the nesting level does not overflow an int. (You will probably - * run out of stack space first.) - * - * If you add or remove a call to rcu_nmi_enter(), be sure to test - * with CONFIG_RCU_EQS_DEBUG=3Dy. - */ -noinstr void rcu_nmi_enter(void) -{ - long incby =3D 2; - struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); - - /* Complain about underflow. */ - WARN_ON_ONCE(ct_dynticks_nmi_nesting() < 0); - - /* - * If idle from RCU viewpoint, atomically increment ->dynticks - * to mark non-idle and increment ->dynticks_nmi_nesting by one. - * Otherwise, increment ->dynticks_nmi_nesting by two. This means - * if ->dynticks_nmi_nesting is equal to one, we are guaranteed - * to be in the outermost NMI handler that interrupted an RCU-idle - * period (observation due to Andy Lutomirski). - */ - if (rcu_dynticks_curr_cpu_in_eqs()) { - - if (!in_nmi()) - rcu_dynticks_task_exit(); - - // RCU is not watching here ... - rcu_dynticks_eqs_exit(); - // ... but is watching here. - - instrumentation_begin(); - // instrumentation for the noinstr rcu_dynticks_curr_cpu_in_eqs() - instrument_atomic_read(&ct->dynticks, sizeof(ct->dynticks)); - // instrumentation for the noinstr rcu_dynticks_eqs_exit() - instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); - - incby =3D 1; - } else if (!in_nmi()) { - instrumentation_begin(); - rcu_irq_enter_check_tick(); - } else { - instrumentation_begin(); - } - - trace_rcu_dyntick(incby =3D=3D 1 ? TPS("Endirq") : TPS("++=3D"), - ct_dynticks_nmi_nesting(), - ct_dynticks_nmi_nesting() + incby, ct_dynticks()); - instrumentation_end(); - WRITE_ONCE(ct->dynticks_nmi_nesting, /* Prevent store tearing. */ - ct_dynticks_nmi_nesting() + incby); - barrier(); -} - /* * Check to see if any future non-offloaded RCU-related work will need * to be done by the current CPU, even if none need be done immediately, diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h index 12e249fbaf1e..d4a97e40ea9c 100644 --- a/kernel/rcu/tree.h +++ b/kernel/rcu/tree.h @@ -435,7 +435,6 @@ static void rcu_cpu_kthread_setup(unsigned int cpu); static void rcu_spawn_one_boost_kthread(struct rcu_node *rnp); static bool rcu_preempt_has_tasks(struct rcu_node *rnp); static bool rcu_preempt_need_deferred_qs(struct task_struct *t); -static void rcu_preempt_deferred_qs(struct task_struct *t); static void zero_cpu_stall_ticks(struct rcu_data *rdp); static struct swait_queue_head *rcu_nocb_gp_get(struct rcu_node *rnp); static void rcu_nocb_gp_cleanup(struct swait_queue_head *sq); @@ -475,10 +474,6 @@ do { \ =20 static void rcu_bind_gp_kthread(void); static bool rcu_nohz_full_cpu(void); -static void rcu_dynticks_task_enter(void); -static void rcu_dynticks_task_exit(void); -static void rcu_dynticks_task_trace_enter(void); -static void rcu_dynticks_task_trace_exit(void); =20 /* Forward declarations for tree_stall.h */ static void record_gp_stall_check_time(void); diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h index f50c1fafc562..b2c01919b92c 100644 --- a/kernel/rcu/tree_plugin.h +++ b/kernel/rcu/tree_plugin.h @@ -595,7 +595,7 @@ static notrace bool rcu_preempt_need_deferred_qs(struct= task_struct *t) * evaluate safety in terms of interrupt, softirq, and preemption * disabling. */ -static notrace void rcu_preempt_deferred_qs(struct task_struct *t) +notrace void rcu_preempt_deferred_qs(struct task_struct *t) { unsigned long flags; =20 @@ -935,7 +935,7 @@ static bool rcu_preempt_need_deferred_qs(struct task_st= ruct *t) // period for a quiescent state from this CPU. Note that requests from // tasks are handled when removing the task from the blocked-tasks list // below. -static void rcu_preempt_deferred_qs(struct task_struct *t) +void rcu_preempt_deferred_qs(struct task_struct *t) { struct rcu_data *rdp =3D this_cpu_ptr(&rcu_data); =20 @@ -1296,37 +1296,3 @@ static void rcu_bind_gp_kthread(void) return; housekeeping_affine(current, HK_TYPE_RCU); } - -/* Record the current task on dyntick-idle entry. */ -static __always_inline void rcu_dynticks_task_enter(void) -{ -#if defined(CONFIG_TASKS_RCU) && defined(CONFIG_NO_HZ_FULL) - WRITE_ONCE(current->rcu_tasks_idle_cpu, smp_processor_id()); -#endif /* #if defined(CONFIG_TASKS_RCU) && defined(CONFIG_NO_HZ_FULL) */ -} - -/* Record no current task on dyntick-idle exit. */ -static __always_inline void rcu_dynticks_task_exit(void) -{ -#if defined(CONFIG_TASKS_RCU) && defined(CONFIG_NO_HZ_FULL) - WRITE_ONCE(current->rcu_tasks_idle_cpu, -1); -#endif /* #if defined(CONFIG_TASKS_RCU) && defined(CONFIG_NO_HZ_FULL) */ -} - -/* Turn on heavyweight RCU tasks trace readers on idle/user entry. */ -static __always_inline void rcu_dynticks_task_trace_enter(void) -{ -#ifdef CONFIG_TASKS_TRACE_RCU - if (IS_ENABLED(CONFIG_TASKS_TRACE_RCU_READ_MB)) - current->trc_reader_special.b.need_mb =3D true; -#endif /* #ifdef CONFIG_TASKS_TRACE_RCU */ -} - -/* Turn off heavyweight RCU tasks trace readers on idle/user exit. */ -static __always_inline void rcu_dynticks_task_trace_exit(void) -{ -#ifdef CONFIG_TASKS_TRACE_RCU - if (IS_ENABLED(CONFIG_TASKS_TRACE_RCU_READ_MB)) - current->trc_reader_special.b.need_mb =3D false; -#endif /* #ifdef CONFIG_TASKS_TRACE_RCU */ -} --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 47BB4C43334 for ; Wed, 8 Jun 2022 14:42:42 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S242018AbiFHOmj (ORCPT ); Wed, 8 Jun 2022 10:42:39 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:34840 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S241964AbiFHOmQ (ORCPT ); Wed, 8 Jun 2022 10:42:16 -0400 Received: from dfw.source.kernel.org (dfw.source.kernel.org [IPv6:2604:1380:4641:c500::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 48B5F198C3E for ; Wed, 8 Jun 2022 07:41:47 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by dfw.source.kernel.org (Postfix) with ESMTPS id 4E54D61A97 for ; Wed, 8 Jun 2022 14:41:47 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 7BC42C341C8; Wed, 8 Jun 2022 14:41:43 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699306; bh=hdQIV88BNOyKOtFeeVMzqR7ZFe/k0qmhBxMWn//m5Q0=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=UKZpLWh6cC62HypnHndjVP+sQ086Dbo9TH/Zm7y9jZBDQpRBZ+7rN3Ae4WJnehQKl rzIesvgo+000n9hKAxgMWh1yxKrR+260JJmAmBK5EGAS95HZ4DEMYPqKEA1dNH4kJV AI/CzmWimJ0GaEQo1o4/8BfWVeSXiIgVxYC7uOM1FNF0LSIFwxEBW0ubMnDyBG4lVS SIC2OSHSUmowEvtU3yqPrEV4kID5QLRwENB4Ltfbfj9VzRrzhwgW3NIbnw2OS0JLic yjlPu48KFSiDO3zhbvKNZX07wqIVFoVc9zW4nSRsrP9Vom559G9viNy7QkxP9zDKH9 gJbncBzbLMTgA== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 17/20] rcu/context-tracking: Remove unused and/or unecessary middle functions Date: Wed, 8 Jun 2022 16:40:34 +0200 Message-Id: <20220608144037.1765000-18-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Some eqs functions are now only used internally by context tracking, so their public declarations can be removed. Also middle functions such as rcu_user_*() and rcu_idle_*() which now directly call to rcu_eqs_enter() and rcu_eqs_exit() can be wiped out as well. Signed-off-by: Frederic Weisbecker Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- Documentation/RCU/stallwarn.rst | 2 +- include/linux/hardirq.h | 8 --- include/linux/rcupdate.h | 8 --- include/linux/rcutiny.h | 2 - include/linux/rcutree.h | 2 - kernel/context_tracking.c | 98 +++++++++------------------------ 6 files changed, 28 insertions(+), 92 deletions(-) diff --git a/Documentation/RCU/stallwarn.rst b/Documentation/RCU/stallwarn.= rst index ce1f58a9d954..e38c587067fc 100644 --- a/Documentation/RCU/stallwarn.rst +++ b/Documentation/RCU/stallwarn.rst @@ -97,7 +97,7 @@ warnings: which will include additional debugging information. =20 - A low-level kernel issue that either fails to invoke one of the - variants of rcu_user_enter(), rcu_user_exit(), ct_idle_enter(), + variants of rcu_eqs_enter(true), rcu_eqs_exit(true), ct_idle_enter(), ct_idle_exit(), ct_irq_enter(), or ct_irq_exit() on the one hand, or that invokes one of them too many times on the other. Historically, the most frequent issue has been an omission diff --git a/include/linux/hardirq.h b/include/linux/hardirq.h index 345cdbe9c1b7..d57cab4d4c06 100644 --- a/include/linux/hardirq.h +++ b/include/linux/hardirq.h @@ -92,14 +92,6 @@ void irq_exit_rcu(void); #define arch_nmi_exit() do { } while (0) #endif =20 -#ifdef CONFIG_TINY_RCU -static inline void rcu_nmi_enter(void) { } -static inline void rcu_nmi_exit(void) { } -#else -extern void rcu_nmi_enter(void); -extern void rcu_nmi_exit(void); -#endif - /* * NMI vs Tracing * -------------- diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h index 1d120dca6548..b6e9cc7352ed 100644 --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -105,14 +105,6 @@ static inline void rcu_sysrq_start(void) { } static inline void rcu_sysrq_end(void) { } #endif /* #else #ifdef CONFIG_RCU_STALL_COMMON */ =20 -#ifdef CONFIG_NO_HZ_FULL -void rcu_user_enter(void); -void rcu_user_exit(void); -#else -static inline void rcu_user_enter(void) { } -static inline void rcu_user_exit(void) { } -#endif /* CONFIG_NO_HZ_FULL */ - #if defined(CONFIG_NO_HZ_FULL) && (!defined(CONFIG_GENERIC_ENTRY) || !defi= ned(CONFIG_KVM_XFER_TO_GUEST_WORK)) void rcu_irq_work_resched(void); #else diff --git a/include/linux/rcutiny.h b/include/linux/rcutiny.h index b94e8c8fda5a..62815c0a2dce 100644 --- a/include/linux/rcutiny.h +++ b/include/linux/rcutiny.h @@ -95,8 +95,6 @@ static inline int rcu_needs_cpu(void) static inline void rcu_virt_note_context_switch(int cpu) { } static inline void rcu_cpu_stall_reset(void) { } static inline int rcu_jiffies_till_stall_check(void) { return 21 * HZ; } -static inline void rcu_idle_enter(void) { } -static inline void rcu_idle_exit(void) { } static inline void rcu_irq_exit_check_preempt(void) { } #define rcu_is_idle_cpu(cpu) \ (is_idle_task(current) && !in_nmi() && !in_hardirq() && !in_serving_softi= rq()) diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h index c31e4e526643..47eaa4cb0df7 100644 --- a/include/linux/rcutree.h +++ b/include/linux/rcutree.h @@ -47,8 +47,6 @@ unsigned long start_poll_synchronize_rcu(void); bool poll_state_synchronize_rcu(unsigned long oldstate); void cond_synchronize_rcu(unsigned long oldstate); =20 -void rcu_idle_enter(void); -void rcu_idle_exit(void); bool rcu_is_idle_cpu(int cpu); =20 #ifdef CONFIG_PROVE_RCU diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index bb64c1ed5ab2..3e93a9a93a84 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -189,17 +189,17 @@ static void noinstr rcu_eqs_exit(bool user) } =20 /** - * rcu_nmi_exit - inform RCU of exit from NMI context + * ct_nmi_exit - inform RCU of exit from NMI context * * If we are returning from the outermost NMI handler that interrupted an * RCU-idle period, update ct->dynticks and ct->dynticks_nmi_nesting * to let the RCU grace-period handling know that the CPU is back to * being RCU-idle. * - * If you add or remove a call to rcu_nmi_exit(), be sure to test + * If you add or remove a call to ct_nmi_exit(), be sure to test * with CONFIG_RCU_EQS_DEBUG=3Dy. */ -void noinstr rcu_nmi_exit(void) +void noinstr ct_nmi_exit(void) { struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); =20 @@ -242,7 +242,7 @@ void noinstr rcu_nmi_exit(void) } =20 /** - * rcu_nmi_enter - inform RCU of entry to NMI context + * ct_nmi_enter - inform RCU of entry to NMI context * * If the CPU was idle from RCU's viewpoint, update ct->dynticks and * ct->dynticks_nmi_nesting to let the RCU grace-period handling know @@ -250,10 +250,10 @@ void noinstr rcu_nmi_exit(void) * long as the nesting level does not overflow an int. (You will probably * run out of stack space first.) * - * If you add or remove a call to rcu_nmi_enter(), be sure to test + * If you add or remove a call to ct_nmi_enter(), be sure to test * with CONFIG_RCU_EQS_DEBUG=3Dy. */ -void noinstr rcu_nmi_enter(void) +void noinstr ct_nmi_enter(void) { long incby =3D 2; struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); @@ -302,32 +302,33 @@ void noinstr rcu_nmi_enter(void) } =20 /** - * rcu_idle_enter - inform RCU that current CPU is entering idle + * ct_idle_enter - inform RCU that current CPU is entering idle * * Enter idle mode, in other words, -leave- the mode in which RCU * read-side critical sections can occur. (Though RCU read-side * critical sections can occur in irq handlers in idle, a possibility * handled by irq_enter() and irq_exit().) * - * If you add or remove a call to rcu_idle_enter(), be sure to test with + * If you add or remove a call to ct_idle_enter(), be sure to test with * CONFIG_RCU_EQS_DEBUG=3Dy. */ -void noinstr rcu_idle_enter(void) +void noinstr ct_idle_enter(void) { WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !raw_irqs_disabled()); rcu_eqs_enter(false); } +EXPORT_SYMBOL_GPL(ct_idle_enter); =20 /** - * rcu_idle_exit - inform RCU that current CPU is leaving idle + * ct_idle_exit - inform RCU that current CPU is leaving idle * * Exit idle mode, in other words, -enter- the mode in which RCU * read-side critical sections can occur. * - * If you add or remove a call to rcu_idle_exit(), be sure to test with + * If you add or remove a call to ct_idle_exit(), be sure to test with * CONFIG_RCU_EQS_DEBUG=3Dy. */ -void noinstr rcu_idle_exit(void) +void noinstr ct_idle_exit(void) { unsigned long flags; =20 @@ -335,18 +336,6 @@ void noinstr rcu_idle_exit(void) rcu_eqs_exit(false); raw_local_irq_restore(flags); } -EXPORT_SYMBOL_GPL(rcu_idle_exit); - -noinstr void ct_idle_enter(void) -{ - rcu_idle_enter(); -} -EXPORT_SYMBOL_GPL(ct_idle_enter); - -void ct_idle_exit(void) -{ - rcu_idle_exit(); -} EXPORT_SYMBOL_GPL(ct_idle_exit); =20 /** @@ -431,50 +420,11 @@ void ct_irq_exit_irqson(void) ct_irq_exit(); local_irq_restore(flags); } - -noinstr void ct_nmi_enter(void) -{ - rcu_nmi_enter(); -} - -noinstr void ct_nmi_exit(void) -{ - rcu_nmi_exit(); -} +#else +static __always_inline void rcu_eqs_enter(bool user) { } +static __always_inline void rcu_eqs_exit(bool user) { } #endif /* #ifdef CONFIG_CONTEXT_TRACKING_IDLE */ =20 -#ifdef CONFIG_NO_HZ_FULL -/** - * rcu_user_enter - inform RCU that we are resuming userspace. - * - * Enter RCU idle mode right before resuming userspace. No use of RCU - * is permitted between this call and rcu_user_exit(). This way the - * CPU doesn't need to maintain the tick for RCU maintenance purposes - * when the CPU runs in userspace. - * - * If you add or remove a call to rcu_user_enter(), be sure to test with - * CONFIG_RCU_EQS_DEBUG=3Dy. - */ -noinstr void rcu_user_enter(void) -{ - rcu_eqs_enter(true); -} - -/** - * rcu_user_exit - inform RCU that we are exiting userspace. - * - * Exit RCU idle mode while entering the kernel because it can - * run a RCU read side critical section anytime. - * - * If you add or remove a call to rcu_user_exit(), be sure to test with - * CONFIG_RCU_EQS_DEBUG=3Dy. - */ -void noinstr rcu_user_exit(void) -{ - rcu_eqs_exit(true); -} -#endif /* #ifdef CONFIG_NO_HZ_FULL */ - #ifdef CONFIG_CONTEXT_TRACKING_USER =20 #define CREATE_TRACE_POINTS @@ -542,7 +492,13 @@ void noinstr __ct_user_enter(enum ctx_state state) * that will fire and reschedule once we resume in user/guest mode. */ rcu_irq_work_resched(); - rcu_user_enter(); + /* + * Enter RCU idle mode right before resuming userspace. No use of RCU + * is permitted between this call and rcu_eqs_exit(). This way the + * CPU doesn't need to maintain the tick for RCU maintenance purposes + * when the CPU runs in userspace. + */ + rcu_eqs_enter(true); } /* * Even if context tracking is disabled on this CPU, because it's outside @@ -580,7 +536,7 @@ void ct_user_enter(enum ctx_state state) /* * Some contexts may involve an exception occuring in an irq, * leading to that nesting: - * ct_irq_enter() rcu_user_exit() rcu_user_exit() ct_irq_exit() + * ct_irq_enter() rcu_eqs_exit(true) rcu_eqs_enter(true) ct_irq_exit() * This would mess up the dyntick_nesting count though. And rcu_irq_*() * helpers are enough to protect RCU uses inside the exception. So * just return immediately if we detect we are in an IRQ. @@ -632,10 +588,10 @@ void noinstr __ct_user_exit(enum ctx_state state) if (__this_cpu_read(context_tracking.state) =3D=3D state) { if (__this_cpu_read(context_tracking.active)) { /* - * We are going to run code that may use RCU. Inform - * RCU core about that (ie: we may need the tick again). + * Exit RCU idle mode while entering the kernel because it can + * run a RCU read side critical section anytime. */ - rcu_user_exit(); + rcu_eqs_exit(true); if (state =3D=3D CONTEXT_USER) { instrumentation_begin(); vtime_user_exit(current); --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 08D29C433EF for ; Wed, 8 Jun 2022 14:42:47 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S241802AbiFHOmn (ORCPT ); Wed, 8 Jun 2022 10:42:43 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:58144 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S242030AbiFHOmR (ORCPT ); Wed, 8 Jun 2022 10:42:17 -0400 Received: from ams.source.kernel.org (ams.source.kernel.org [IPv6:2604:1380:4601:e00::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 8B3C41AFACB for ; Wed, 8 Jun 2022 07:41:53 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ams.source.kernel.org (Postfix) with ESMTPS id 94492B827F3 for ; Wed, 8 Jun 2022 14:41:51 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 2BBE9C341C4; Wed, 8 Jun 2022 14:41:46 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699310; bh=7cRn/Fj9tWxbW/Il7XA+vQwNrK/qxuPSHeS31wD8ijA=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=Mo3U8vk1fLJBkZ1I6U59rrYHlnd8CYEZkmY9NwZU69CrEQ5Z+yth03zTSIxBI73Bb 6WGKZ4t8B0J5/TC2x3K6MU8AFXMj197NRqJz7pDXrLRISgUSsD4M96pNYXliPcFD9K uAQUvCW+cl7+t7uJOVI6DpW6nhDUeUz6QD9g2QBitSP66Ucyvu+5ycOf0jEUEabGKT QFKuAsYmqtCZiE6GLHrz0QiIfPnyI4Tc30/7DKe6Z681vxowtFeSgNAWyc9Sx6onQE cGTEzRee5wtWAeM+K2xuhxF/6Q/r3ipC8iOrQb/FEH7/w2m4rnscoha+L9lI6qan8l obG4tffq5AMFQ== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 18/20] context_tracking: Convert state to atomic_t Date: Wed, 8 Jun 2022 16:40:35 +0200 Message-Id: <20220608144037.1765000-19-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Context tracking's state and dynticks counter are going to be merged in a single field so that both updates can happen atomically and at the same time. Prepare for that with converting the state into an atomic_t. Signed-off-by: Frederic Weisbecker Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- include/linux/context_tracking.h | 24 ++++------------ include/linux/context_tracking_state.h | 40 ++++++++++++++++++++++---- kernel/context_tracking.c | 15 ++++++---- 3 files changed, 49 insertions(+), 30 deletions(-) diff --git a/include/linux/context_tracking.h b/include/linux/context_track= ing.h index 1f568676bc1d..a8c1db0a3f65 100644 --- a/include/linux/context_tracking.h +++ b/include/linux/context_tracking.h @@ -56,7 +56,7 @@ static inline enum ctx_state exception_enter(void) !context_tracking_enabled()) return 0; =20 - prev_ctx =3D this_cpu_read(context_tracking.state); + prev_ctx =3D __ct_state(); if (prev_ctx !=3D CONTEXT_KERNEL) ct_user_exit(prev_ctx); =20 @@ -86,33 +86,21 @@ static __always_inline void context_tracking_guest_exit= (void) __ct_user_exit(CONTEXT_GUEST); } =20 -/** - * ct_state() - return the current context tracking state if known - * - * Returns the current cpu's context tracking state if context tracking - * is enabled. If context tracking is disabled, returns - * CONTEXT_DISABLED. This should be used primarily for debugging. - */ -static __always_inline enum ctx_state ct_state(void) -{ - return context_tracking_enabled() ? - this_cpu_read(context_tracking.state) : CONTEXT_DISABLED; -} +#define CT_WARN_ON(cond) WARN_ON(context_tracking_enabled() && (cond)) + #else static inline void user_enter(void) { } static inline void user_exit(void) { } static inline void user_enter_irqoff(void) { } static inline void user_exit_irqoff(void) { } -static inline enum ctx_state exception_enter(void) { return 0; } +static inline int exception_enter(void) { return 0; } static inline void exception_exit(enum ctx_state prev_ctx) { } -static inline enum ctx_state ct_state(void) { return CONTEXT_DISABLED; } +static inline int ct_state(void) { return -1; } static __always_inline bool context_tracking_guest_enter(void) { return fa= lse; } static inline void context_tracking_guest_exit(void) { } - +#define CT_WARN_ON(cond) #endif /* !CONFIG_CONTEXT_TRACKING_USER */ =20 -#define CT_WARN_ON(cond) WARN_ON(context_tracking_enabled() && (cond)) - #ifdef CONFIG_CONTEXT_TRACKING_USER_FORCE extern void context_tracking_init(void); #else diff --git a/include/linux/context_tracking_state.h b/include/linux/context= _tracking_state.h index b99e9827cc08..1501df6d4cfa 100644 --- a/include/linux/context_tracking_state.h +++ b/include/linux/context_tracking_state.h @@ -9,6 +9,13 @@ /* Offset to allow distinguishing irq vs. task-based idle entry/exit. */ #define DYNTICK_IRQ_NONIDLE ((LONG_MAX / 2) + 1) =20 +enum ctx_state { + CONTEXT_DISABLED =3D -1, /* returned by ct_state() if unknown */ + CONTEXT_KERNEL =3D 0, + CONTEXT_USER, + CONTEXT_GUEST, +}; + struct context_tracking { #ifdef CONFIG_CONTEXT_TRACKING_USER /* @@ -19,12 +26,7 @@ struct context_tracking { */ bool active; int recursion; - enum ctx_state { - CONTEXT_DISABLED =3D -1, /* returned by ct_state() if unknown */ - CONTEXT_KERNEL =3D 0, - CONTEXT_USER, - CONTEXT_GUEST, - } state; + atomic_t state; #endif #ifdef CONFIG_CONTEXT_TRACKING_IDLE atomic_t dynticks; /* Even value for idle, else odd. */ @@ -96,6 +98,32 @@ static inline bool context_tracking_enabled_this_cpu(voi= d) return context_tracking_enabled() && __this_cpu_read(context_tracking.act= ive); } =20 +static __always_inline int __ct_state(void) +{ + return atomic_read(this_cpu_ptr(&context_tracking.state)); +} + +/** + * ct_state() - return the current context tracking state if known + * + * Returns the current cpu's context tracking state if context tracking + * is enabled. If context tracking is disabled, returns + * CONTEXT_DISABLED. This should be used primarily for debugging. + */ +static __always_inline int ct_state(void) +{ + int ret; + + if (!context_tracking_enabled()) + return CONTEXT_DISABLED; + + preempt_disable(); + ret =3D __ct_state(); + preempt_enable(); + + return ret; +} + #else static inline bool context_tracking_enabled(void) { return false; } static inline bool context_tracking_enabled_cpu(int cpu) { return false; } diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index 3e93a9a93a84..810bca217151 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -463,6 +463,7 @@ static __always_inline void context_tracking_recursion_= exit(void) */ void noinstr __ct_user_enter(enum ctx_state state) { + struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); lockdep_assert_irqs_disabled(); =20 /* Kernel threads aren't supposed to go to userspace */ @@ -471,8 +472,8 @@ void noinstr __ct_user_enter(enum ctx_state state) if (!context_tracking_recursion_enter()) return; =20 - if ( __this_cpu_read(context_tracking.state) !=3D state) { - if (__this_cpu_read(context_tracking.active)) { + if (__ct_state() !=3D state) { + if (ct->active) { /* * At this stage, only low level arch entry code remains and * then we'll run in userspace. We can assume there won't be @@ -513,7 +514,7 @@ void noinstr __ct_user_enter(enum ctx_state state) * OTOH we can spare the calls to vtime and RCU when context_tracking.ac= tive * is false because we know that CPU is not tickless. */ - __this_cpu_write(context_tracking.state, state); + atomic_set(&ct->state, state); } context_tracking_recursion_exit(); } @@ -582,11 +583,13 @@ NOKPROBE_SYMBOL(user_enter_callable); */ void noinstr __ct_user_exit(enum ctx_state state) { + struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); + if (!context_tracking_recursion_enter()) return; =20 - if (__this_cpu_read(context_tracking.state) =3D=3D state) { - if (__this_cpu_read(context_tracking.active)) { + if (__ct_state() =3D=3D state) { + if (ct->active) { /* * Exit RCU idle mode while entering the kernel because it can * run a RCU read side critical section anytime. @@ -599,7 +602,7 @@ void noinstr __ct_user_exit(enum ctx_state state) instrumentation_end(); } } - __this_cpu_write(context_tracking.state, CONTEXT_KERNEL); + atomic_set(&ct->state, CONTEXT_KERNEL); } context_tracking_recursion_exit(); } --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 024A3C433EF for ; Wed, 8 Jun 2022 14:42:57 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S241682AbiFHOmx (ORCPT ); Wed, 8 Jun 2022 10:42:53 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:34694 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S242070AbiFHOmS (ORCPT ); Wed, 8 Jun 2022 10:42:18 -0400 Received: from dfw.source.kernel.org (dfw.source.kernel.org [139.178.84.217]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 47AFB2A24B for ; Wed, 8 Jun 2022 07:41:55 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by dfw.source.kernel.org (Postfix) with ESMTPS id 9A1E961BB9 for ; Wed, 8 Jun 2022 14:41:54 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id CD6B0C3411E; Wed, 8 Jun 2022 14:41:50 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699314; bh=HWCFka5weLXGq0bS4SsHRIPkvDGK8ohd+BW3CLeixpI=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=O+/Ps8sxZ/sDoNZ+ZbALC8iZZtYhkuEigRzILqm9c5hbT4VhQVGDe7gbn+Btq++CQ Kg88zvIGNepwtqbonOYpRS+1lIV6/Ew7Ed/GxcR43wZSrw5me3JDJEoxYhTOk5aLBu EvQsCmnPSn+5inaixK4ECvAu/MNTR5TfJT5no6PvFp2a4FrUTGm80gh+WBdm3pq01G gyU+RL3zNflRoEh/QKSCj+9Q3SXZpCJpaodP4UJx5GuyI6a+VUwuCHGd8gWZXBpI48 RC8yyI1EiHjOwubUo4rUHWwrXq+ymOF36W361HYgYZeU5jK7be4VzAObSFuv7NNQUi FL1adr9w/RHiA== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 19/20] rcu/context_tracking: Merge dynticks counter and context tracking states Date: Wed, 8 Jun 2022 16:40:36 +0200 Message-Id: <20220608144037.1765000-20-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Updating the context tracking state and the RCU dynticks counter atomically in a single operation is a first step towards improving CPU isolation. This makes the context tracking state updates fully ordered and therefore allow for later enhancements such as postponing some work while a task is running isolated in userspace until it ever comes back to the kernel. The state field becomes divided in two parts: 1) Two Lower bits for context tracking state: CONTEXT_KERNEL =3D 0 CONTEXT_IDLE =3D 1, CONTEXT_USER =3D 2, CONTEXT_GUEST =3D 3, 2) Higher bits for RCU eqs dynticks counting: RCU_DYNTICKS_IDX =3D 4 The dynticks counting is always incremented by this value. (state & RCU_DYNTICKS_IDX) means we are NOT in an extended quiescent state. This makes the chance for a collision more likely between two RCU dynticks snapshots but wrapping up 28 bits of eqs dynticks increments still takes some bad luck (also rdp.dynticks_snap could be converted from int to long?) Some RCU eqs functions have been renamed to better reflect their broader scope that now include context tracking state. Signed-off-by: Frederic Weisbecker Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- include/linux/context_tracking.h | 8 +- include/linux/context_tracking_state.h | 35 ++++--- kernel/context_tracking.c | 132 ++++++++++++++++--------- kernel/rcu/tree.c | 13 ++- kernel/rcu/tree_stall.h | 4 +- 5 files changed, 121 insertions(+), 71 deletions(-) diff --git a/include/linux/context_tracking.h b/include/linux/context_track= ing.h index a8c1db0a3f65..fd354eaea510 100644 --- a/include/linux/context_tracking.h +++ b/include/linux/context_tracking.h @@ -118,16 +118,16 @@ extern void ct_idle_exit(void); */ static __always_inline bool rcu_dynticks_curr_cpu_in_eqs(void) { - return !(arch_atomic_read(this_cpu_ptr(&context_tracking.dynticks)) & 0x1= ); + return !(arch_atomic_read(this_cpu_ptr(&context_tracking.state)) & RCU_DY= NTICKS_IDX); } =20 /* - * Increment the current CPU's context_tracking structure's ->dynticks fie= ld + * Increment the current CPU's context_tracking structure's ->state field * with ordering. Return the new value. */ -static __always_inline unsigned long rcu_dynticks_inc(int incby) +static __always_inline unsigned long ct_state_inc(int incby) { - return arch_atomic_add_return(incby, this_cpu_ptr(&context_tracking.dynti= cks)); + return arch_atomic_add_return(incby, this_cpu_ptr(&context_tracking.state= )); } =20 #else diff --git a/include/linux/context_tracking_state.h b/include/linux/context= _tracking_state.h index 1501df6d4cfa..580a525bfba7 100644 --- a/include/linux/context_tracking_state.h +++ b/include/linux/context_tracking_state.h @@ -10,12 +10,20 @@ #define DYNTICK_IRQ_NONIDLE ((LONG_MAX / 2) + 1) =20 enum ctx_state { - CONTEXT_DISABLED =3D -1, /* returned by ct_state() if unknown */ - CONTEXT_KERNEL =3D 0, - CONTEXT_USER, - CONTEXT_GUEST, + CONTEXT_DISABLED =3D -1, /* returned by ct_state() if unknown */ + CONTEXT_KERNEL =3D 0, + CONTEXT_IDLE =3D 1, + CONTEXT_USER =3D 2, + CONTEXT_GUEST =3D 3, + CONTEXT_MAX =3D 4, }; =20 +/* Even value for idle, else odd. */ +#define RCU_DYNTICKS_IDX CONTEXT_MAX + +#define CT_STATE_MASK (CONTEXT_MAX - 1) +#define CT_DYNTICKS_MASK (~CT_STATE_MASK) + struct context_tracking { #ifdef CONFIG_CONTEXT_TRACKING_USER /* @@ -26,10 +34,11 @@ struct context_tracking { */ bool active; int recursion; +#endif +#ifdef CONFIG_CONTEXT_TRACKING atomic_t state; #endif #ifdef CONFIG_CONTEXT_TRACKING_IDLE - atomic_t dynticks; /* Even value for idle, else odd. */ long dynticks_nesting; /* Track process nesting level. */ long dynticks_nmi_nesting; /* Track irq/NMI nesting level. */ #endif @@ -37,24 +46,29 @@ struct context_tracking { =20 #ifdef CONFIG_CONTEXT_TRACKING DECLARE_PER_CPU(struct context_tracking, context_tracking); + +static __always_inline int __ct_state(void) +{ + return atomic_read(this_cpu_ptr(&context_tracking.state)) & CT_STATE_MASK; +} #endif =20 #ifdef CONFIG_CONTEXT_TRACKING_IDLE static __always_inline int ct_dynticks(void) { - return atomic_read(this_cpu_ptr(&context_tracking.dynticks)); + return atomic_read(this_cpu_ptr(&context_tracking.state)) & CT_DYNTICKS_M= ASK; } =20 static __always_inline int ct_dynticks_cpu(int cpu) { struct context_tracking *ct =3D per_cpu_ptr(&context_tracking, cpu); - return atomic_read(&ct->dynticks); + return atomic_read(&ct->state) & CT_DYNTICKS_MASK; } =20 static __always_inline int ct_dynticks_cpu_acquire(int cpu) { struct context_tracking *ct =3D per_cpu_ptr(&context_tracking, cpu); - return atomic_read_acquire(&ct->dynticks); + return atomic_read_acquire(&ct->state) & CT_DYNTICKS_MASK; } =20 static __always_inline long ct_dynticks_nesting(void) @@ -98,11 +112,6 @@ static inline bool context_tracking_enabled_this_cpu(vo= id) return context_tracking_enabled() && __this_cpu_read(context_tracking.act= ive); } =20 -static __always_inline int __ct_state(void) -{ - return atomic_read(this_cpu_ptr(&context_tracking.state)); -} - /** * ct_state() - return the current context tracking state if known * diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index 810bca217151..d50bc66f2b1c 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -28,8 +28,8 @@ DEFINE_PER_CPU(struct context_tracking, context_tracking)= =3D { #ifdef CONFIG_CONTEXT_TRACKING_IDLE .dynticks_nesting =3D 1, .dynticks_nmi_nesting =3D DYNTICK_IRQ_NONIDLE, - .dynticks =3D ATOMIC_INIT(1), #endif + .state =3D ATOMIC_INIT(RCU_DYNTICKS_IDX), }; EXPORT_SYMBOL_GPL(context_tracking); =20 @@ -76,7 +76,7 @@ static __always_inline void rcu_dynticks_task_trace_exit(= void) * RCU is watching prior to the call to this function and is no longer * watching upon return. */ -static noinstr void rcu_dynticks_eqs_enter(void) +static noinstr void ct_kernel_exit_state(int offset) { int seq; =20 @@ -86,9 +86,9 @@ static noinstr void rcu_dynticks_eqs_enter(void) * next idle sojourn. */ rcu_dynticks_task_trace_enter(); // Before ->dynticks update! - seq =3D rcu_dynticks_inc(1); + seq =3D ct_state_inc(offset); // RCU is no longer watching. Better be in extended quiescent state! - WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && (seq & 0x1)); + WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && (seq & RCU_DYNTICKS_IDX)= ); } =20 /* @@ -96,7 +96,7 @@ static noinstr void rcu_dynticks_eqs_enter(void) * called from an extended quiescent state, that is, RCU is not watching * prior to the call to this function and is watching upon return. */ -static noinstr void rcu_dynticks_eqs_exit(void) +static noinstr void ct_kernel_enter_state(int offset) { int seq; =20 @@ -105,10 +105,10 @@ static noinstr void rcu_dynticks_eqs_exit(void) * and we also must force ordering with the next RCU read-side * critical section. */ - seq =3D rcu_dynticks_inc(1); + seq =3D ct_state_inc(offset); // RCU is now watching. Better not be in an extended quiescent state! rcu_dynticks_task_trace_exit(); // After ->dynticks update! - WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !(seq & 0x1)); + WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !(seq & RCU_DYNTICKS_IDX= )); } =20 /* @@ -119,7 +119,7 @@ static noinstr void rcu_dynticks_eqs_exit(void) * the possibility of usermode upcalls having messed up our count * of interrupt nesting level during the prior busy period. */ -static void noinstr rcu_eqs_enter(bool user) +static void noinstr ct_kernel_exit(bool user, int offset) { struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); =20 @@ -139,13 +139,13 @@ static void noinstr rcu_eqs_enter(bool user) WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(c= urrent)); rcu_preempt_deferred_qs(current); =20 - // instrumentation for the noinstr rcu_dynticks_eqs_enter() - instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); + // instrumentation for the noinstr ct_kernel_exit_state() + instrument_atomic_write(&ct->state, sizeof(ct->state)); =20 instrumentation_end(); WRITE_ONCE(ct->dynticks_nesting, 0); /* Avoid irq-access tearing. */ // RCU is watching here ... - rcu_dynticks_eqs_enter(); + ct_kernel_exit_state(offset); // ... but is no longer watching here. rcu_dynticks_task_enter(); } @@ -158,7 +158,7 @@ static void noinstr rcu_eqs_enter(bool user) * allow for the possibility of usermode upcalls messing up our count of * interrupt nesting level during the busy period that is just now startin= g. */ -static void noinstr rcu_eqs_exit(bool user) +static void noinstr ct_kernel_enter(bool user, int offset) { struct context_tracking *ct =3D this_cpu_ptr(&context_tracking); long oldval; @@ -173,12 +173,12 @@ static void noinstr rcu_eqs_exit(bool user) } rcu_dynticks_task_exit(); // RCU is not watching here ... - rcu_dynticks_eqs_exit(); + ct_kernel_enter_state(offset); // ... but is watching here. instrumentation_begin(); =20 - // instrumentation for the noinstr rcu_dynticks_eqs_exit() - instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); + // instrumentation for the noinstr ct_kernel_enter_state() + instrument_atomic_write(&ct->state, sizeof(ct->state)); =20 trace_rcu_dyntick(TPS("End"), ct_dynticks_nesting(), 1, ct_dynticks()); WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(c= urrent)); @@ -192,7 +192,7 @@ static void noinstr rcu_eqs_exit(bool user) * ct_nmi_exit - inform RCU of exit from NMI context * * If we are returning from the outermost NMI handler that interrupted an - * RCU-idle period, update ct->dynticks and ct->dynticks_nmi_nesting + * RCU-idle period, update ct->state and ct->dynticks_nmi_nesting * to let the RCU grace-period handling know that the CPU is back to * being RCU-idle. * @@ -229,12 +229,12 @@ void noinstr ct_nmi_exit(void) trace_rcu_dyntick(TPS("Startirq"), ct_dynticks_nmi_nesting(), 0, ct_dynti= cks()); WRITE_ONCE(ct->dynticks_nmi_nesting, 0); /* Avoid store tearing. */ =20 - // instrumentation for the noinstr rcu_dynticks_eqs_enter() - instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); + // instrumentation for the noinstr ct_kernel_exit_state() + instrument_atomic_write(&ct->state, sizeof(ct->state)); instrumentation_end(); =20 // RCU is watching here ... - rcu_dynticks_eqs_enter(); + ct_kernel_exit_state(RCU_DYNTICKS_IDX); // ... but is no longer watching here. =20 if (!in_nmi()) @@ -244,7 +244,7 @@ void noinstr ct_nmi_exit(void) /** * ct_nmi_enter - inform RCU of entry to NMI context * - * If the CPU was idle from RCU's viewpoint, update ct->dynticks and + * If the CPU was idle from RCU's viewpoint, update ct->state and * ct->dynticks_nmi_nesting to let the RCU grace-period handling know * that the CPU is active. This implementation permits nested NMIs, as * long as the nesting level does not overflow an int. (You will probably @@ -275,14 +275,14 @@ void noinstr ct_nmi_enter(void) rcu_dynticks_task_exit(); =20 // RCU is not watching here ... - rcu_dynticks_eqs_exit(); + ct_kernel_enter_state(RCU_DYNTICKS_IDX); // ... but is watching here. =20 instrumentation_begin(); // instrumentation for the noinstr rcu_dynticks_curr_cpu_in_eqs() - instrument_atomic_read(&ct->dynticks, sizeof(ct->dynticks)); - // instrumentation for the noinstr rcu_dynticks_eqs_exit() - instrument_atomic_write(&ct->dynticks, sizeof(ct->dynticks)); + instrument_atomic_read(&ct->state, sizeof(ct->state)); + // instrumentation for the noinstr ct_kernel_enter_state() + instrument_atomic_write(&ct->state, sizeof(ct->state)); =20 incby =3D 1; } else if (!in_nmi()) { @@ -315,7 +315,7 @@ void noinstr ct_nmi_enter(void) void noinstr ct_idle_enter(void) { WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !raw_irqs_disabled()); - rcu_eqs_enter(false); + ct_kernel_exit(false, RCU_DYNTICKS_IDX + CONTEXT_IDLE); } EXPORT_SYMBOL_GPL(ct_idle_enter); =20 @@ -333,7 +333,7 @@ void noinstr ct_idle_exit(void) unsigned long flags; =20 raw_local_irq_save(flags); - rcu_eqs_exit(false); + ct_kernel_enter(false, RCU_DYNTICKS_IDX - CONTEXT_IDLE); raw_local_irq_restore(flags); } EXPORT_SYMBOL_GPL(ct_idle_exit); @@ -421,8 +421,8 @@ void ct_irq_exit_irqson(void) local_irq_restore(flags); } #else -static __always_inline void rcu_eqs_enter(bool user) { } -static __always_inline void rcu_eqs_exit(bool user) { } +static __always_inline void ct_kernel_exit(bool user, int offset) { } +static __always_inline void ct_kernel_enter(bool user, int offset) { } #endif /* #ifdef CONFIG_CONTEXT_TRACKING_IDLE */ =20 #ifdef CONFIG_CONTEXT_TRACKING_USER @@ -493,28 +493,49 @@ void noinstr __ct_user_enter(enum ctx_state state) * that will fire and reschedule once we resume in user/guest mode. */ rcu_irq_work_resched(); + /* * Enter RCU idle mode right before resuming userspace. No use of RCU * is permitted between this call and rcu_eqs_exit(). This way the * CPU doesn't need to maintain the tick for RCU maintenance purposes * when the CPU runs in userspace. */ - rcu_eqs_enter(true); + ct_kernel_exit(true, RCU_DYNTICKS_IDX + state); + + /* + * Special case if we only track user <-> kernel transitions for tickle= ss + * cputime accounting but we don't support RCU extended quiescent state. + * In this we case we don't care about any concurrency/ordering. + */ + if (!IS_ENABLED(CONFIG_CONTEXT_TRACKING_IDLE)) + atomic_set(&ct->state, state); + } else { + /* + * Even if context tracking is disabled on this CPU, because it's outsi= de + * the full dynticks mask for example, we still have to keep track of t= he + * context transitions and states to prevent inconsistency on those of + * other CPUs. + * If a task triggers an exception in userspace, sleep on the exception + * handler and then migrate to another CPU, that new CPU must know where + * the exception returns by the time we call exception_exit(). + * This information can only be provided by the previous CPU when it ca= lled + * exception_enter(). + * OTOH we can spare the calls to vtime and RCU when context_tracking.a= ctive + * is false because we know that CPU is not tickless. + */ + if (!IS_ENABLED(CONFIG_CONTEXT_TRACKING_IDLE)) { + /* Tracking for vtime only, no concurrent RCU EQS accounting */ + atomic_set(&ct->state, state); + } else { + /* + * Tracking for vtime and RCU EQS. Make sure we don't race + * with NMIs. OTOH we don't care about ordering here since + * RCU only requires RCU_DYNTICKS_IDX increments to be fully + * ordered. + */ + atomic_add(state, &ct->state); + } } - /* - * Even if context tracking is disabled on this CPU, because it's outside - * the full dynticks mask for example, we still have to keep track of the - * context transitions and states to prevent inconsistency on those of - * other CPUs. - * If a task triggers an exception in userspace, sleep on the exception - * handler and then migrate to another CPU, that new CPU must know where - * the exception returns by the time we call exception_exit(). - * This information can only be provided by the previous CPU when it cal= led - * exception_enter(). - * OTOH we can spare the calls to vtime and RCU when context_tracking.ac= tive - * is false because we know that CPU is not tickless. - */ - atomic_set(&ct->state, state); } context_tracking_recursion_exit(); } @@ -594,15 +615,36 @@ void noinstr __ct_user_exit(enum ctx_state state) * Exit RCU idle mode while entering the kernel because it can * run a RCU read side critical section anytime. */ - rcu_eqs_exit(true); + ct_kernel_enter(true, RCU_DYNTICKS_IDX - state); if (state =3D=3D CONTEXT_USER) { instrumentation_begin(); vtime_user_exit(current); trace_user_exit(0); instrumentation_end(); } + + /* + * Special case if we only track user <-> kernel transitions for tickle= ss + * cputime accounting but we don't support RCU extended quiescent state. + * In this we case we don't care about any concurrency/ordering. + */ + if (!IS_ENABLED(CONFIG_CONTEXT_TRACKING_IDLE)) + atomic_set(&ct->state, CONTEXT_KERNEL); + + } else { + if (!IS_ENABLED(CONFIG_CONTEXT_TRACKING_IDLE)) { + /* Tracking for vtime only, no concurrent RCU EQS accounting */ + atomic_set(&ct->state, CONTEXT_KERNEL); + } else { + /* + * Tracking for vtime and RCU EQS. Make sure we don't race + * with NMIs. OTOH we don't care about ordering here since + * RCU only requires RCU_DYNTICKS_IDX increments to be fully + * ordered. + */ + atomic_sub(state, &ct->state); + } } - atomic_set(&ct->state, CONTEXT_KERNEL); } context_tracking_recursion_exit(); } diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index 155e8ce3d267..642622f2a6b4 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -276,9 +276,9 @@ void rcu_softirq_qs(void) */ static void rcu_dynticks_eqs_online(void) { - if (ct_dynticks() & 0x1) + if (ct_dynticks() & RCU_DYNTICKS_IDX) return; - rcu_dynticks_inc(1); + ct_state_inc(RCU_DYNTICKS_IDX); } =20 /* @@ -297,7 +297,7 @@ static int rcu_dynticks_snap(int cpu) */ static bool rcu_dynticks_in_eqs(int snap) { - return !(snap & 0x1); + return !(snap & RCU_DYNTICKS_IDX); } =20 /* Return true if the specified CPU is currently idle from an RCU viewpoin= t. */ @@ -325,8 +325,7 @@ bool rcu_dynticks_zero_in_eqs(int cpu, int *vp) int snap; =20 // If not quiescent, force back to earlier extended quiescent state. - snap =3D ct_dynticks_cpu(cpu) & ~0x1; - + snap =3D ct_dynticks_cpu(cpu) & ~RCU_DYNTICKS_IDX; smp_rmb(); // Order ->dynticks and *vp reads. if (READ_ONCE(*vp)) return false; // Non-zero, so report failure; @@ -352,9 +351,9 @@ notrace void rcu_momentary_dyntick_idle(void) int seq; =20 raw_cpu_write(rcu_data.rcu_need_heavy_qs, false); - seq =3D rcu_dynticks_inc(2); + seq =3D ct_state_inc(2 * RCU_DYNTICKS_IDX); /* It is illegal to call this from idle state. */ - WARN_ON_ONCE(!(seq & 0x1)); + WARN_ON_ONCE(!(seq & RCU_DYNTICKS_IDX)); rcu_preempt_deferred_qs(current); } EXPORT_SYMBOL_GPL(rcu_momentary_dyntick_idle); diff --git a/kernel/rcu/tree_stall.h b/kernel/rcu/tree_stall.h index 91e4fd4db12d..c3fbbcc09327 100644 --- a/kernel/rcu/tree_stall.h +++ b/kernel/rcu/tree_stall.h @@ -469,7 +469,7 @@ static void print_cpu_stall_info(int cpu) rcuc_starved =3D rcu_is_rcuc_kthread_starving(rdp, &j); if (rcuc_starved) sprintf(buf, " rcuc=3D%ld jiffies(starved)", j); - pr_err("\t%d-%c%c%c%c: (%lu %s) idle=3D%03x/%ld/%#lx softirq=3D%u/%u fqs= =3D%ld%s%s\n", + pr_err("\t%d-%c%c%c%c: (%lu %s) idle=3D%04x/%ld/%#lx softirq=3D%u/%u fqs= =3D%ld%s%s\n", cpu, "O."[!!cpu_online(cpu)], "o."[!!(rdp->grpmask & rdp->mynode->qsmaskinit)], @@ -478,7 +478,7 @@ static void print_cpu_stall_info(int cpu) rdp->rcu_iw_pending ? (int)min(delta, 9UL) + '0' : "!."[!delta], ticks_value, ticks_title, - rcu_dynticks_snap(cpu) & 0xfff, + rcu_dynticks_snap(cpu) & 0xffff, ct_dynticks_nesting_cpu(cpu), ct_dynticks_nmi_nesting_cpu(cpu), rdp->softirq_snap, kstat_softirqs_cpu(RCU_SOFTIRQ, cpu), data_race(rcu_state.n_force_qs) - rcu_state.n_force_qs_gpstart, --=20 2.25.1 From nobody Tue Apr 28 01:10:07 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 1322DCCA47B for ; Wed, 8 Jun 2022 14:42:50 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S242041AbiFHOms (ORCPT ); Wed, 8 Jun 2022 10:42:48 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:35180 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S242082AbiFHOmT (ORCPT ); Wed, 8 Jun 2022 10:42:19 -0400 Received: from ams.source.kernel.org (ams.source.kernel.org [IPv6:2604:1380:4601:e00::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 592181B9F9B for ; Wed, 8 Jun 2022 07:42:00 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ams.source.kernel.org (Postfix) with ESMTPS id DE09CB827F0 for ; Wed, 8 Jun 2022 14:41:58 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 7C9E2C385A5; Wed, 8 Jun 2022 14:41:54 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1654699317; bh=2CBAllI1hzeZcnfVP9eQWZJba0Jxs5p3K1LdbrCyRPA=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=W3sl6dyKIs1uyuJUT5VLXMVBt8f2dwlCdodC1ae8/F06cmGg1XkiDGsQSXIU5RQDy 9yJWtpDL3Q8lzvp2HgfER9OO5tmNCQEF3kVGpz0o67q61z4r17uZyI1NqtXv5jMEjn 9Oo4zg9FYHvU1HMbm44WNmdnENwSWO3NMO5GMeUCsoYKkjESBNAwhdv33Cl8n71EbF LVI1n3DpKrVO3Xuu9Vk2zpbRzCU4RtOaA7qQWYFERMprjREFhugvnQt8S14bDN2VIK aDLjaCpmrRqavEKb62QCBGsbxdnsXC3sP8e9cYLMkm106ml7ks82N2RsOXrRK7M+pj D5gUGUvcCfAIg== From: Frederic Weisbecker To: LKML Cc: Frederic Weisbecker , Peter Zijlstra , Phil Auld , Alex Belits , Nicolas Saenz Julienne , Xiongfeng Wang , Neeraj Upadhyay , Thomas Gleixner , Yu Liao , Boqun Feng , "Paul E . McKenney" , Marcelo Tosatti , Paul Gortmaker , Uladzislau Rezki , Joel Fernandes Subject: [PATCH 20/20] MAINTAINERS: Add Paul as context tracking maintainer Date: Wed, 8 Jun 2022 16:40:37 +0200 Message-Id: <20220608144037.1765000-21-frederic@kernel.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220608144037.1765000-1-frederic@kernel.org> References: <20220608144037.1765000-1-frederic@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Since most of the bits have been imported from kernel/rcu/tree.c and now that the context tracking code is tightly linked to RCU, add Paul as a context tracking maintainer. Also update the context tracking file header accordingly. Signed-off-by: Frederic Weisbecker Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Neeraj Upadhyay Cc: Uladzislau Rezki Cc: Joel Fernandes Cc: Boqun Feng Cc: Nicolas Saenz Julienne Cc: Marcelo Tosatti Cc: Xiongfeng Wang Cc: Yu Liao Cc: Phil Auld Cc: Paul Gortmaker Cc: Alex Belits Reviewed-by: Nicolas Saenz Julienne Tested-by: Nicolas Saenz Julienne --- MAINTAINERS | 1 + kernel/context_tracking.c | 12 +++++++----- 2 files changed, 8 insertions(+), 5 deletions(-) diff --git a/MAINTAINERS b/MAINTAINERS index fd768d43e048..f49d9e6790d4 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -4976,6 +4976,7 @@ F: include/linux/console* =20 CONTEXT TRACKING M: Frederic Weisbecker +M: "Paul E. McKenney" S: Maintained F: kernel/context_tracking.c F: include/linux/context_tracking* diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c index d50bc66f2b1c..4c5655c612e3 100644 --- a/kernel/context_tracking.c +++ b/kernel/context_tracking.c @@ -1,18 +1,20 @@ // SPDX-License-Identifier: GPL-2.0-only /* - * Context tracking: Probe on high level context boundaries such as kernel - * and userspace. This includes syscalls and exceptions entry/exit. + * Context tracking: Probe on high level context boundaries such as kernel, + * userspace, guest or idle. * * This is used by RCU to remove its dependency on the timer tick while a = CPU - * runs in userspace. + * runs in idle, userspace or guest mode. * - * Started by Frederic Weisbecker: + * User/guest tracking started by Frederic Weisbecker: * - * Copyright (C) 2012 Red Hat, Inc., Frederic Weisbecker + * Copyright (C) 2012 Red Hat, Inc., Frederic Weisbecker * * Many thanks to Gilad Ben-Yossef, Paul McKenney, Ingo Molnar, Andrew Mor= ton, * Steven Rostedt, Peter Zijlstra for suggestions and improvements. * + * RCU extended quiescent state bits imported from kernel/rcu/tree.c + * where the relevant authorship may be found. */ =20 #include --=20 2.25.1