On Thursday 01 May 2008 07:35:32 Neil Brown wrote: > On Tuesday April 29, bs@xxxxxxxxx wrote: > > Sorry for not responding earlier.... No problem, we are all always overly busy ;) > > I seem to remember the setting used to be global, but I see in this > patch it is per-array -- which makes much more sense. > Though it brings up an interesting question. If two arrays share a > device, and exactly one of them is flagged for parallel sync, does > that make sense? > > Maybe it does. I suspect the reality is that it is individual devices > that should be flagged for parallel sync or not, and the setting on > the array is just an "and" of the settings for the devices in the > array. So different settings on arrays which share one device can > make sense... > > Note: I'm *not* suggesting that the setting should be moved to the > component devices - that would be too clumsy. I'm just musing. > > If get 3 errors and 2 warnings from ./scripts/checkpatch.pl. If you > fix those I'll take the patch. Below is a new version of the patch. If kmail should break something, I also uploaded it here: http://www.pci.uni-heidelberg.de/tc/usr/bernd/downloads/md/parallel_resync.patch Unfortunately I presently don't have any non-production system I could easily reboot to test the new version on. Thanks, Bernd Allow parallel resync of md-devices. Signed-off-by: Bernd Schubert <bs@xxxxxxxxx> Index: linus/drivers/md/md.c =================================================================== --- linus.orig/drivers/md/md.c +++ linus/drivers/md/md.c @@ -74,6 +74,8 @@ static DEFINE_SPINLOCK(pers_lock); static void md_print_devices(void); +static DECLARE_WAIT_QUEUE_HEAD(resync_wait); + #define MD_BUG(x...) { printk("md: bug in file %s, line %d\n", __FILE__, __LINE__); md_print_devices(); } /* @@ -2979,6 +2981,36 @@ degraded_show(mddev_t *mddev, char *page static struct md_sysfs_entry md_degraded = __ATTR_RO(degraded); static ssize_t +sync_force_parallel_show(mddev_t *mddev, char *page) +{ + return sprintf(page, "%d\n", mddev->parallel_resync); +} + +static ssize_t +sync_force_parallel_store(mddev_t *mddev, const char *buf, size_t len) +{ + long n; + + if (strict_strtol(buf, 10, &n)) + return -EINVAL; + + if (n != 0 && n != 1) + return -EINVAL; + + mddev->parallel_resync = n; + + if (mddev->sync_thread) + wake_up(&resync_wait); + + return len; +} + +/* force parallel resync, even with shared block devices */ +static struct md_sysfs_entry md_sync_force_parallel = +__ATTR(sync_force_parallel, S_IRUGO|S_IWUSR, + sync_force_parallel_show, sync_force_parallel_store); + +static ssize_t sync_speed_show(mddev_t *mddev, char *page) { unsigned long resync, dt, db; @@ -3153,6 +3185,7 @@ static struct attribute *md_redundancy_a &md_sync_min.attr, &md_sync_max.attr, &md_sync_speed.attr, + &md_sync_force_parallel.attr, &md_sync_completed.attr, &md_max_sync.attr, &md_suspend_lo.attr, @@ -5413,8 +5446,6 @@ void md_allow_write(mddev_t *mddev) } EXPORT_SYMBOL_GPL(md_allow_write); -static DECLARE_WAIT_QUEUE_HEAD(resync_wait); - #define SYNC_MARKS 10 #define SYNC_MARK_STEP (3*HZ) void md_do_sync(mddev_t *mddev) @@ -5478,8 +5509,9 @@ void md_do_sync(mddev_t *mddev) for_each_mddev(mddev2, tmp) { if (mddev2 == mddev) continue; - if (mddev2->curr_resync && - match_mddev_units(mddev,mddev2)) { + if (!mddev->parallel_resync + && mddev2->curr_resync + && match_mddev_units(mddev, mddev2)) { DEFINE_WAIT(wq); if (mddev < mddev2 && mddev->curr_resync == 2) { /* arbitrarily yield */ Index: linus/include/linux/raid/md_k.h =================================================================== --- linus.orig/include/linux/raid/md_k.h +++ linus/include/linux/raid/md_k.h @@ -176,6 +176,9 @@ struct mddev_s int sync_speed_min; int sync_speed_max; + /* resync even though the same disks are shared among md-devices */ + int parallel_resync; + int ok_start_degraded; /* recovery/resync flags * NEEDED: we might need to start a resync/recover -- Bernd Schubert Q-Leap Networks GmbH -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
