Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756047AbZKQS5s (ORCPT ); Tue, 17 Nov 2009 13:57:48 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1755560AbZKQS5s (ORCPT ); Tue, 17 Nov 2009 13:57:48 -0500 Received: from cavan.codon.org.uk ([93.93.128.6]:39374 "EHLO cavan.codon.org.uk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754365AbZKQS5r (ORCPT ); Tue, 17 Nov 2009 13:57:47 -0500 Date: Tue, 17 Nov 2009 18:57:42 +0000 From: Matthew Garrett To: David Zeuthen Cc: Kay Sievers , linux-kernel@vger.kernel.org, axboe@kernel.dk, linux-hotplug@vger.kernel.org Subject: Re: [PATCH] [RFC] Add support for uevents on block device idle changes Message-ID: <20091117185742.GA19829@srcf.ucam.org> References: <1258468659-5446-1-git-send-email-mjg@redhat.com> <1258474180.16176.62.camel@localhost.localdomain> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1258474180.16176.62.camel@localhost.localdomain> User-Agent: Mutt/1.5.18 (2008-05-17) X-SA-Exim-Connect-IP: X-SA-Exim-Mail-From: mjg59@cavan.codon.org.uk X-SA-Exim-Scanned: No (on cavan.codon.org.uk); SAEximRunCond expanded to false Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 8077 Lines: 244 Ok. How about something like this? It adds an extra field to the stat file and introduces David's suggestion of making it pollable. commit ba6d4c7ab7940ae8dc11a884281d0a36b20455b9 Author: Matthew Garrett Date: Mon Nov 16 17:44:03 2009 -0500 [RFC] Add support for uevents on block device idle changes Userspace may wish to know whether a given disk is active or idle, for example to modify power management policy based on access patterns. This patch adds a deferrable timer to the block layer which will fire if the disk is idle for a user-definable period of time, generating a uevent. A uevent will also be generated if an access is received while the disk is classified as idle. diff --git a/Documentation/ABI/testing/sysfs-block b/Documentation/ABI/testing/sysfs-block index 5f3beda..8747f42 100644 --- a/Documentation/ABI/testing/sysfs-block +++ b/Documentation/ABI/testing/sysfs-block @@ -3,7 +3,7 @@ Date: February 2008 Contact: Jerome Marchand Description: The /sys/block//stat files displays the I/O - statistics of disk . They contain 11 fields: + statistics of disk . They contain 12 fields: 1 - reads completed succesfully 2 - reads merged 3 - sectors read @@ -15,6 +15,7 @@ Description: 9 - I/Os currently in progress 10 - time spent doing I/Os (ms) 11 - weighted time spent doing I/Os (ms) + 12 - 1 if the disk is idle (determined by idle_hysteresis) For more details refer Documentation/iostats.txt @@ -128,3 +129,12 @@ Description: preferred request size for workloads where sustained throughput is desired. If no optimal I/O size is reported this file contains 0. + +What: /sys/block//idle_hysteresis +Date: November 2009 +Contact: Matthew Garrett +Description: + Contains the number of milliseconds to wait after an access + before declaring that a disk is idle. Any accesses during + this time will reset the timer. "0" (the default) indicates + that no events will be generated. \ No newline at end of file diff --git a/block/blk-core.c b/block/blk-core.c index 71da511..f278817 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -1452,6 +1452,15 @@ static inline void __generic_make_request(struct bio *bio) if (should_fail_request(bio)) goto end_io; + if (bio->bi_bdev->bd_disk->hysteresis_time && + bio_has_data(bio) && + !mod_timer(&bio->bi_bdev->bd_disk->hysteresis_timer, + jiffies+msecs_to_jiffies + (bio->bi_bdev->bd_disk->hysteresis_time))) { + bio->bi_bdev->bd_disk->idle = 0; + schedule_work(&bio->bi_bdev->bd_disk->idle_notify); + } + /* * If this device has partitions, remap block n * of partition p to block n+start(p) of the disk. diff --git a/block/genhd.c b/block/genhd.c index 517e433..ea37e48 100644 --- a/block/genhd.c +++ b/block/genhd.c @@ -504,6 +504,21 @@ static int exact_lock(dev_t devt, void *data) return 0; } +static void disk_idle(unsigned long data) +{ + struct gendisk *gd = (struct gendisk *)data; + + gd->idle = 1; + schedule_work(&gd->idle_notify); +} + +static void disk_idle_notify_thread(struct work_struct *work) +{ + struct gendisk *gd = container_of(work, struct gendisk, idle_notify); + + sysfs_notify(&disk_to_dev(gd)->kobj, NULL, "stat"); +} + /** * add_disk - add partitioning information to kernel list * @disk: per-device partitioning information @@ -543,6 +558,10 @@ void add_disk(struct gendisk *disk) blk_register_region(disk_devt(disk), disk->minors, NULL, exact_match, exact_lock, disk); + + init_timer(&disk->hysteresis_timer); + setup_timer(&disk->hysteresis_timer, disk_idle, (unsigned long)disk); + register_disk(disk); blk_register_queue(disk); @@ -861,6 +880,32 @@ static ssize_t disk_alignment_offset_show(struct device *dev, return sprintf(buf, "%d\n", queue_alignment_offset(disk->queue)); } +static ssize_t disk_idle_hysteresis_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct gendisk *disk = dev_to_disk(dev); + + return sprintf(buf, "%d\n", disk->hysteresis_time); +} + +static ssize_t disk_idle_hysteresis_store(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t count) +{ + struct gendisk *disk = dev_to_disk(dev); + unsigned long timeout; + int res; + + res = strict_strtoul(buf, 10, &timeout); + if (res) + return -EINVAL; + + disk->hysteresis_time = timeout; + + return count; +} + static DEVICE_ATTR(range, S_IRUGO, disk_range_show, NULL); static DEVICE_ATTR(ext_range, S_IRUGO, disk_ext_range_show, NULL); static DEVICE_ATTR(removable, S_IRUGO, disk_removable_show, NULL); @@ -870,6 +915,8 @@ static DEVICE_ATTR(alignment_offset, S_IRUGO, disk_alignment_offset_show, NULL); static DEVICE_ATTR(capability, S_IRUGO, disk_capability_show, NULL); static DEVICE_ATTR(stat, S_IRUGO, part_stat_show, NULL); static DEVICE_ATTR(inflight, S_IRUGO, part_inflight_show, NULL); +static DEVICE_ATTR(idle_hysteresis, 0644, disk_idle_hysteresis_show, + disk_idle_hysteresis_store); #ifdef CONFIG_FAIL_MAKE_REQUEST static struct device_attribute dev_attr_fail = __ATTR(make-it-fail, S_IRUGO|S_IWUSR, part_fail_show, part_fail_store); @@ -890,6 +937,7 @@ static struct attribute *disk_attrs[] = { &dev_attr_capability.attr, &dev_attr_stat.attr, &dev_attr_inflight.attr, + &dev_attr_idle_hysteresis.attr, #ifdef CONFIG_FAIL_MAKE_REQUEST &dev_attr_fail.attr, #endif @@ -1183,6 +1231,8 @@ struct gendisk *alloc_disk_node(int minors, int node_id) device_initialize(disk_to_dev(disk)); INIT_WORK(&disk->async_notify, media_change_notify_thread); + INIT_WORK(&disk->idle_notify, + disk_idle_notify_thread); } return disk; } diff --git a/fs/partitions/check.c b/fs/partitions/check.c index 7b685e1..cccfb7d 100644 --- a/fs/partitions/check.c +++ b/fs/partitions/check.c @@ -230,6 +230,7 @@ ssize_t part_stat_show(struct device *dev, struct device_attribute *attr, char *buf) { struct hd_struct *p = dev_to_part(dev); + struct gendisk *gd = dev_to_disk(dev); int cpu; cpu = part_stat_lock(); @@ -238,7 +239,7 @@ ssize_t part_stat_show(struct device *dev, return sprintf(buf, "%8lu %8lu %8llu %8u " "%8lu %8lu %8llu %8u " - "%8u %8u %8u" + "%8u %8u %8u %1u" "\n", part_stat_read(p, ios[READ]), part_stat_read(p, merges[READ]), @@ -250,7 +251,8 @@ ssize_t part_stat_show(struct device *dev, jiffies_to_msecs(part_stat_read(p, ticks[WRITE])), part_in_flight(p), jiffies_to_msecs(part_stat_read(p, io_ticks)), - jiffies_to_msecs(part_stat_read(p, time_in_queue))); + jiffies_to_msecs(part_stat_read(p, time_in_queue)), + gd->idle); } ssize_t part_inflight_show(struct device *dev, @@ -652,6 +654,9 @@ void del_gendisk(struct gendisk *disk) struct disk_part_iter piter; struct hd_struct *part; + del_timer_sync(&disk->hysteresis_timer); + cancel_work_sync(&disk->idle_notify); + /* invalidate stuff */ disk_part_iter_init(&piter, disk, DISK_PITER_INCL_EMPTY | DISK_PITER_REVERSE); diff --git a/include/linux/genhd.h b/include/linux/genhd.h index 297df45..7e969a5 100644 --- a/include/linux/genhd.h +++ b/include/linux/genhd.h @@ -12,6 +12,7 @@ #include #include #include +#include #ifdef CONFIG_BLOCK @@ -163,10 +164,15 @@ struct gendisk { atomic_t sync_io; /* RAID */ struct work_struct async_notify; + struct work_struct idle_notify; #ifdef CONFIG_BLK_DEV_INTEGRITY struct blk_integrity *integrity; #endif int node_id; + + bool idle; + int hysteresis_time; + struct timer_list hysteresis_timer; }; static inline struct gendisk *part_to_disk(struct hd_struct *part) -- Matthew Garrett | mjg59@srcf.ucam.org -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/