From patchwork Fri Jun 1 12:24:03 2012 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Anton Vorontsov X-Patchwork-Id: 9076 Return-Path: X-Original-To: patchwork@peony.canonical.com Delivered-To: patchwork@peony.canonical.com Received: from fiordland.canonical.com (fiordland.canonical.com [91.189.94.145]) by peony.canonical.com (Postfix) with ESMTP id 8CF1A23E1B for ; Fri, 1 Jun 2012 12:26:31 +0000 (UTC) Received: from mail-gg0-f180.google.com (mail-gg0-f180.google.com [209.85.161.180]) by fiordland.canonical.com (Postfix) with ESMTP id 5C808A186C3 for ; Fri, 1 Jun 2012 12:26:31 +0000 (UTC) Received: by mail-gg0-f180.google.com with SMTP id f1so1894261ggn.11 for ; Fri, 01 Jun 2012 05:26:31 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=x-forwarded-to:x-forwarded-for:delivered-to:received-spf:from:to:cc :subject:date:message-id:x-mailer:in-reply-to:references :x-gm-message-state; bh=iCqgvsBUR2DA9+EayC8hW0bLhNlG2ZC4HMw2la2GDjU=; b=oQLtu53+Nkij9THNy1JqptV3DVP1Hca9noe4VHoBnYTHLWvlRaP7O94pYMXK/4tQPq f1/fdOnNgBQPEMP+rCNV0Dn23xPi9LZ/l142cu5HuIZZG4o70+Jp89OzabELom+CVwin RV8fhT0VCKc0MioT4Fv7TNcMrlmKvRtaULTbDMLCOfplWSonfVojx1d7Upl+lYAE4c0/ MAqNtG593RyeG3WucnzcMfbH82Gj+y3TINvCiUxEtd91KktSX1fY/758WTC/1Fvyb9kJ 3MrRVjH8ShKRLFkN2wiqx9Yhv0U70Dzo5mATYtv4BV7lb5Zk+OT43HYoA+4rdwIs4vob QVZg== Received: by 10.42.210.193 with SMTP id gl1mr1392173icb.57.1338553590660; Fri, 01 Jun 2012 05:26:30 -0700 (PDT) X-Forwarded-To: linaro-patchwork@canonical.com X-Forwarded-For: patch@linaro.org linaro-patchwork@canonical.com Delivered-To: patches@linaro.org Received: by 10.231.24.148 with SMTP id v20csp313874ibb; Fri, 1 Jun 2012 05:26:30 -0700 (PDT) Received: by 10.68.134.71 with SMTP id pi7mr8120330pbb.91.1338553589954; Fri, 01 Jun 2012 05:26:29 -0700 (PDT) Received: from mail-pb0-f50.google.com (mail-pb0-f50.google.com [209.85.160.50]) by mx.google.com with ESMTPS id kg6si4294258pbc.185.2012.06.01.05.26.29 (version=TLSv1/SSLv3 cipher=OTHER); Fri, 01 Jun 2012 05:26:29 -0700 (PDT) Received-SPF: neutral (google.com: 209.85.160.50 is neither permitted nor denied by best guess record for domain of anton.vorontsov@linaro.org) client-ip=209.85.160.50; Authentication-Results: mx.google.com; spf=neutral (google.com: 209.85.160.50 is neither permitted nor denied by best guess record for domain of anton.vorontsov@linaro.org) smtp.mail=anton.vorontsov@linaro.org Received: by mail-pb0-f50.google.com with SMTP id rr4so3360376pbb.37 for ; Fri, 01 Jun 2012 05:26:29 -0700 (PDT) Received: by 10.68.231.195 with SMTP id ti3mr9805088pbc.96.1338553589727; Fri, 01 Jun 2012 05:26:29 -0700 (PDT) Received: from localhost (c-71-204-165-222.hsd1.ca.comcast.net. [71.204.165.222]) by mx.google.com with ESMTPS id qm6sm2624184pbc.10.2012.06.01.05.26.28 (version=TLSv1/SSLv3 cipher=OTHER); Fri, 01 Jun 2012 05:26:28 -0700 (PDT) From: Anton Vorontsov To: Pekka Enberg Cc: Leonid Moiseichuk , KOSAKI Motohiro , Minchan Kim , Bartlomiej Zolnierkiewicz , John Stultz , linux-mm@kvack.org, linux-kernel@vger.kernel.org, linaro-kernel@lists.linaro.org, patches@linaro.org, kernel-team@android.com Subject: [PATCH 2/5] vmevent: Convert from deferred timer to deferred work Date: Fri, 1 Jun 2012 05:24:03 -0700 Message-Id: <1338553446-22292-2-git-send-email-anton.vorontsov@linaro.org> X-Mailer: git-send-email 1.7.9.2 In-Reply-To: <20120601122118.GA6128@lizard> References: <20120601122118.GA6128@lizard> X-Gm-Message-State: ALoCoQlb/Q5KnY5hwOx9o5Crkyuj+3mMpzxhh81gd/ohj/+TXoSGoY2mMArWnivSoNCmbcicQHAp We'll need to use smp_function_call() in the sampling routines, and the call is not supposed to be called from the bottom halves. So, let's convert vmevent to dffered workqueues. As a side effect, we also fix the swap reporting (we cannot call si_swapinfo from the interrupt context), i.e. the following oops should be fixed now: ================================= [ INFO: inconsistent lock state ] 3.4.0-rc1+ #37 Not tainted --------------------------------- inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage. swapper/0/0 [HC0[0]:SC1[1]:HE1:SE0] takes: (swap_lock){+.?...}, at: [] si_swapinfo+0x1d/0x90 {SOFTIRQ-ON-W} state was registered at: [] mark_irqflags+0x15f/0x1b0 [] __lock_acquire+0x493/0x9d0 [] lock_acquire+0x9e/0x200 [] _raw_spin_lock+0x41/0x50 [] si_swapinfo+0x1d/0x90 [] meminfo_proc_show+0x38/0x3f0 [] seq_read+0x139/0x3f0 [] proc_reg_read+0x86/0xc0 [] vfs_read+0xac/0x160 [] sys_read+0x4a/0x90 [] system_call_fastpath+0x16/0x1b Signed-off-by: Anton Vorontsov --- mm/vmevent.c | 49 ++++++++++++++++++++++++++++--------------------- 1 file changed, 28 insertions(+), 21 deletions(-) diff --git a/mm/vmevent.c b/mm/vmevent.c index 381e9d1..4ca2a04 100644 --- a/mm/vmevent.c +++ b/mm/vmevent.c @@ -3,7 +3,7 @@ #include #include #include -#include +#include #include #include #include @@ -34,7 +34,7 @@ struct vmevent_watch { struct vmevent_attr *config_attrs[VMEVENT_CONFIG_MAX_ATTRS]; /* sampling */ - struct timer_list timer; + struct delayed_work work; /* poll */ wait_queue_head_t waitq; @@ -146,15 +146,13 @@ static bool vmevent_match(struct vmevent_watch *watch) } /* - * This function is called from the timer context, which has the same - * guaranties as an interrupt handler: it can have only one execution - * thread (unlike bare softirq handler), so we don't need to worry - * about racing w/ ourselves. + * This function is called from a workqueue, which can have only one + * execution thread, so we don't need to worry about racing w/ ourselves. * - * We also don't need to worry about several instances of timers - * accessing the same vmevent_watch, as we allocate vmevent_watch - * together w/ the timer instance in vmevent_fd(), so there is always - * one timer per vmevent_watch. + * We also don't need to worry about several instances of us accessing + * the same vmevent_watch, as we allocate vmevent_watch together w/ the + * work instance in vmevent_fd(), so there is always one work per + * vmevent_watch. * * All the above makes it possible to implement the lock-free logic, * using just the atomic watch->pending variable. @@ -178,26 +176,35 @@ static void vmevent_sample(struct vmevent_watch *watch) atomic_set(&watch->pending, 1); } -static void vmevent_timer_fn(unsigned long data) +static void vmevent_schedule_watch(struct vmevent_watch *watch) { - struct vmevent_watch *watch = (struct vmevent_watch *)data; + schedule_delayed_work(&watch->work, + nsecs_to_jiffies64(watch->config.sample_period_ns)); +} + +static struct vmevent_watch *work_to_vmevent_watch(struct work_struct *work) +{ + struct delayed_work *wk = to_delayed_work(work); + + return container_of(wk, struct vmevent_watch, work); +} + +static void vmevent_timer_fn(struct work_struct *work) +{ + struct vmevent_watch *watch = work_to_vmevent_watch(work); vmevent_sample(watch); if (atomic_read(&watch->pending)) wake_up(&watch->waitq); - mod_timer(&watch->timer, jiffies + - nsecs_to_jiffies64(watch->config.sample_period_ns)); + + vmevent_schedule_watch(watch); } static void vmevent_start_timer(struct vmevent_watch *watch) { - init_timer_deferrable(&watch->timer); - watch->timer.data = (unsigned long)watch; - watch->timer.function = vmevent_timer_fn; - watch->timer.expires = jiffies + - nsecs_to_jiffies64(watch->config.sample_period_ns); - add_timer(&watch->timer); + INIT_DELAYED_WORK_DEFERRABLE(&watch->work, vmevent_timer_fn); + vmevent_schedule_watch(watch); } static unsigned int vmevent_poll(struct file *file, poll_table *wait) @@ -259,7 +266,7 @@ static int vmevent_release(struct inode *inode, struct file *file) { struct vmevent_watch *watch = file->private_data; - del_timer_sync(&watch->timer); + cancel_delayed_work_sync(&watch->work); kfree(watch);