Skip to contentSkip to conversation

Optimize lock in queue unplugging

PATCH In review Mikulas Patocka 4 participants 11 messages Apr 29, 2008 – May 7, 2008 Block layer

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 Optimize lock in queue unplugging

Hi

Mike Anderson was doing an OLTP benchmark on a computer with 48 physical disks mapped to one logical device via device mapper.

He found that there was a slowdown on request_queue->lock in function generic_unplug_device. The slowdown is caused by the fact that when some code calls unplug on the device mapper, device mapper calls unplug on all physical disks. These unplug calls take the lock, find that the queue is already unplugged, release the lock and exit.

With the below patch, performance of the benchmark was increased by 18% (the whole OLTP application, not just block layer microbenchmarks).

So I'm submitting this patch for upstream. I think the patch is correct, because when more threads call simultaneously plug and unplug, it is unspecified, if the queue is or isn't plugged (so the patch can't make this worse). And the caller that plugged the queue should unplug it anyway. (if it doesn't, there's 3ms timeout).

Mikulas

---
 block/blk-core.c |    3 +++
 1 file changed, 3 insertions(+)
block/blk-core.c +3 -0
Index: linux-2.6.25/block/blk-core.c
===================================================================
@@ -271,6 +271,9 @@ EXPORT_SYMBOL(__generic_unplug_device);
  **/
 void generic_unplug_device(struct request_queue *q)
 {
+	if (likely(!blk_queue_plugged(q)))
+		return;
+
 	spin_lock_irq(q->queue_lock);
 	__generic_unplug_device(q);
 	spin_unlock_irq(q->queue_lock);
#1 lore raw
Quoted 11 lines · Mikulas Patocka

Hi

Mike Anderson was doing an OLTP benchmark on a computer with 48 physical disks mapped to one logical device via device mapper.

He found that there was a slowdown on request_queue->lock in function generic_unplug_device. The slowdown is caused by the fact that when some code calls unplug on the device mapper, device mapper calls unplug on all physical disks. These unplug calls take the lock, find that the queue is already unplugged, release the lock and exit.

Quoted 8 lines

With the below patch, performance of the benchmark was increased by 18% (the whole OLTP application, not just block layer microbenchmarks).

So I'm submitting this patch for upstream. I think the patch is correct, because when more threads call simultaneously plug and unplug, it is unspecified, if the queue is or isn't plugged (so the patch can't make this worse). And the caller that plugged the queue should unplug it anyway. (if it doesn't, there's 3ms timeout).

Where were these unplug calls coming from? The block layer will generally only unplug when it is already unplugged, so if you are seeing so many unplug calls that the patch redues overhead by as much described, perhaps the callsite is buggy?

Signature · 1 line
Jens Axboe
#2 lore raw
Quoted 21 lines · Jens Axboe
Quoted 11 lines · Mikulas Patocka

Hi

Mike Anderson was doing an OLTP benchmark on a computer with 48 physical disks mapped to one logical device via device mapper.

He found that there was a slowdown on request_queue->lock in function generic_unplug_device. The slowdown is caused by the fact that when some code calls unplug on the device mapper, device mapper calls unplug on all physical disks. These unplug calls take the lock, find that the queue is already unplugged, release the lock and exit.

Quoted 8 lines

With the below patch, performance of the benchmark was increased by 18% (the whole OLTP application, not just block layer microbenchmarks).

So I'm submitting this patch for upstream. I think the patch is correct, because when more threads call simultaneously plug and unplug, it is unspecified, if the queue is or isn't plugged (so the patch can't make this worse). And the caller that plugged the queue should unplug it anyway. (if it doesn't, there's 3ms timeout).

Quoted 7 lines

Where were these unplug calls coming from? The block layer will generally only unplug when it is already unplugged, so if you are seeing so many unplug calls that the patch redues overhead by as much described, perhaps the callsite is buggy?

Signature · 1 line
Jens Axboe

unplug is called on any wait_on_buffer (and similar calls) __wait_on_buffer -> sync_buffer -> blk_run_address_space -> blk_run_backing_dev -> bdi->unplug_io_fn(bdi, page);

(I'm not sure that this was the IBM's case, I'm just guessing - this is the most obvious example where unplug is called repeatedly)

There is not any test that the queue is plugged and there shouldn't be. If you have this situation

dm-linear(unplugged) -> physical-disk(plugged)

then uplung should be called on dm-linear (that will call dm-unplug method dm_unplug_all and that will unplug the disk). If you add the test of plugged queue to the upper layer, you mess this situation with stacked drivers completely.

The test for already plugged queue should be at the lowest physical device driver, not in upper layers.

Mikulas

#3 lore raw
Quoted 46 lines · Mikulas Patocka
Quoted 21 lines · Jens Axboe
Quoted 11 lines · Mikulas Patocka

Hi

Mike Anderson was doing an OLTP benchmark on a computer with 48 physical disks mapped to one logical device via device mapper.

He found that there was a slowdown on request_queue->lock in function generic_unplug_device. The slowdown is caused by the fact that when some code calls unplug on the device mapper, device mapper calls unplug on all physical disks. These unplug calls take the lock, find that the queue is already unplugged, release the lock and exit.

Quoted 8 lines

With the below patch, performance of the benchmark was increased by 18% (the whole OLTP application, not just block layer microbenchmarks).

So I'm submitting this patch for upstream. I think the patch is correct, because when more threads call simultaneously plug and unplug, it is unspecified, if the queue is or isn't plugged (so the patch can't make this worse). And the caller that plugged the queue should unplug it anyway. (if it doesn't, there's 3ms timeout).

Quoted 7 lines

Where were these unplug calls coming from? The block layer will generally only unplug when it is already unplugged, so if you are seeing so many unplug calls that the patch redues overhead by as much described, perhaps the callsite is buggy?

Signature · 1 line
Jens Axboe

unplug is called on any wait_on_buffer (and similar calls) __wait_on_buffer -> sync_buffer -> blk_run_address_space -> blk_run_backing_dev -> bdi->unplug_io_fn(bdi, page);

(I'm not sure that this was the IBM's case, I'm just guessing - this is the most obvious example where unplug is called repeatedly)

There is not any test that the queue is plugged and there shouldn't be. If you have this situation

dm-linear(unplugged) -> physical-disk(plugged)

Quoted 7 lines

then uplung should be called on dm-linear (that will call dm-unplug method dm_unplug_all and that will unplug the disk). If you add the test of plugged queue to the upper layer, you mess this situation with stacked drivers completely.

The test for already plugged queue should be at the lowest physical device driver, not in upper layers.

Fair enough, I'll put the patch under closer scrutiny and queue it up. Thanks!

Signature · 1 line
Jens Axboe
#4 lore raw
In reply to Jens Axboe
Quoted 21 lines · Jens Axboe
Quoted 11 lines · Mikulas Patocka

Hi

Mike Anderson was doing an OLTP benchmark on a computer with 48 physical disks mapped to one logical device via device mapper.

He found that there was a slowdown on request_queue->lock in function generic_unplug_device. The slowdown is caused by the fact that when some code calls unplug on the device mapper, device mapper calls unplug on all physical disks. These unplug calls take the lock, find that the queue is already unplugged, release the lock and exit.

Quoted 8 lines

With the below patch, performance of the benchmark was increased by 18% (the whole OLTP application, not just block layer microbenchmarks).

So I'm submitting this patch for upstream. I think the patch is correct, because when more threads call simultaneously plug and unplug, it is unspecified, if the queue is or isn't plugged (so the patch can't make this worse). And the caller that plugged the queue should unplug it anyway. (if it doesn't, there's 3ms timeout).

Quoted 4 lines

Where were these unplug calls coming from? The block layer will generally only unplug when it is already unplugged, so if you are seeing so many unplug calls that the patch redues overhead by as much described, perhaps the callsite is buggy?

I do not have direct access the the benchmark setup, but here is the data I have received.

The oprofile data was showing ll_rw_blk::generic_unplug_device() as a top routine at 13% of the samples. Annotation of the samples shows hits on spin_lock_irq(q->queue_lock).

Here are some sample call traces:

Call trace #1

kernel:  [<ffffffff80058c6c>] generic_unplug_device+0x5d/0xc6
kernel:  [<ffffffff8820ea3e>] :dm_mod:dm_table_unplug_all+0x33/0x41
kernel:  [<ffffffff8820cc85>] :dm_mod:dm_unplug_all+0x1d/0x28
kernel:  [<ffffffff8005a78a>] blk_backing_dev_unplug+0x56/0x5b
kernel:  [<ffffffff80014cdc>] sync_buffer+0x36/0x3f
kernel:  [<ffffffff800629a4>] __wait_on_bit+0x40/0x6f
kernel:  [<ffffffff80014ca6>] sync_buffer+0x0/0x3f
kernel:  [<ffffffff80062a3f>] out_of_line_wait_on_bit+0x6c/0x78
kernel:  [<ffffffff8009c474>] wake_bit_function+0x0/0x23
kernel:  [<ffffffff88034c85>] :jbd:journal_commit_transaction+0x91f/0x1086
kernel:  [<ffffffff8003d038>] lock_timer_base+0x1b/0x3c
kernel:  [<ffffffff8803840e>] :jbd:kjournald+0xc1/0x213
kernel:  [<ffffffff8009c446>] autoremove_wake_function+0x0/0x2e
kernel:  [<ffffffff8009c283>] keventd_create_kthread+0x0/0x61
kernel:  [<ffffffff8803834d>] :jbd:kjournald+0x0/0x213
kernel:  [<ffffffff8009c283>] keventd_create_kthread+0x0/0x61
kernel:  [<ffffffff800321d5>] kthread+0xfe/0x132
kernel:  [<ffffffff8005cfb1>] child_rip+0xa/0x11
kernel:  [<ffffffff8009c283>] keventd_create_kthread+0x0/0x61
kernel:  [<ffffffff800320d7>] kthread+0x0/0x132
kernel:  [<ffffffff8005cfa7>] child_rip+0x0/0x11                       

Call trace #2

kernel:  [<ffffffff80058c6c>] generic_unplug_device+0x5d/0xc6
kernel:  [<ffffffff8820ea3e>] :dm_mod:dm_table_unplug_all+0x33/0x41
kernel:  [<ffffffff8820cc85>] :dm_mod:dm_unplug_all+0x1d/0x28
kernel:  [<ffffffff8005a78a>] blk_backing_dev_unplug+0x56/0x5b
kernel:  [<ffffffff800e8bfe>] __blockdev_direct_IO+0x889/0xaa2
kernel:  [<ffffffff88050800>] :ext3:ext3_direct_IO+0xf3/0x18b
kernel:  [<ffffffff8804ec84>] :ext3:ext3_get_block+0x0/0xe3
kernel:  [<ffffffff800be6bb>] generic_file_direct_IO+0xbd/0xfb
kernel:  [<ffffffff8001e637>] generic_file_direct_write+0x60/0xf2
kernel:  [<ffffffff80015cfd>] __generic_file_aio_write_nolock+0x2b7/0x3b8
kernel:  [<ffffffff8002134f>] generic_file_aio_write+0x65/0xc1
kernel:  [<ffffffff8804c192>] :ext3:ext3_file_write+0x16/0x91
kernel:  [<ffffffff80017944>] do_sync_write+0xc7/0x104
kernel:  [<ffffffff8009c446>] autoremove_wake_function+0x0/0x2e
kernel:  [<ffffffff80111400>] free_msg+0x22/0x3c
kernel:  [<ffffffff800161c4>] vfs_write+0xce/0x174
kernel:  [<ffffffff8004194c>] sys_pwrite64+0x50/0x70
kernel:  [<ffffffff8005cde9>] error_exit+0x0/0x84
kernel:  [<ffffffff8005c116>] system_call+0x7e/0x83

-andmike
--
Michael Anderson
[email protected]

#5 lore raw
Quoted 72 lines · Mike Anderson
Quoted 21 lines · Jens Axboe
Quoted 11 lines · Mikulas Patocka

Hi

Mike Anderson was doing an OLTP benchmark on a computer with 48 physical disks mapped to one logical device via device mapper.

He found that there was a slowdown on request_queue->lock in function generic_unplug_device. The slowdown is caused by the fact that when some code calls unplug on the device mapper, device mapper calls unplug on all physical disks. These unplug calls take the lock, find that the queue is already unplugged, release the lock and exit.

Quoted 8 lines

With the below patch, performance of the benchmark was increased by 18% (the whole OLTP application, not just block layer microbenchmarks).

So I'm submitting this patch for upstream. I think the patch is correct, because when more threads call simultaneously plug and unplug, it is unspecified, if the queue is or isn't plugged (so the patch can't make this worse). And the caller that plugged the queue should unplug it anyway. (if it doesn't, there's 3ms timeout).

Quoted 4 lines

Where were these unplug calls coming from? The block layer will generally only unplug when it is already unplugged, so if you are seeing so many unplug calls that the patch redues overhead by as much described, perhaps the callsite is buggy?

I do not have direct access the the benchmark setup, but here is the data I have received.

The oprofile data was showing ll_rw_blk::generic_unplug_device() as a top routine at 13% of the samples. Annotation of the samples shows hits on spin_lock_irq(q->queue_lock).

Here are some sample call traces:

Call trace #1

kernel:  [<ffffffff80058c6c>] generic_unplug_device+0x5d/0xc6
kernel:  [<ffffffff8820ea3e>] :dm_mod:dm_table_unplug_all+0x33/0x41
kernel:  [<ffffffff8820cc85>] :dm_mod:dm_unplug_all+0x1d/0x28
kernel:  [<ffffffff8005a78a>] blk_backing_dev_unplug+0x56/0x5b
kernel:  [<ffffffff80014cdc>] sync_buffer+0x36/0x3f
kernel:  [<ffffffff800629a4>] __wait_on_bit+0x40/0x6f
kernel:  [<ffffffff80014ca6>] sync_buffer+0x0/0x3f
kernel:  [<ffffffff80062a3f>] out_of_line_wait_on_bit+0x6c/0x78
kernel:  [<ffffffff8009c474>] wake_bit_function+0x0/0x23
kernel:  [<ffffffff88034c85>] :jbd:journal_commit_transaction+0x91f/0x1086
kernel:  [<ffffffff8003d038>] lock_timer_base+0x1b/0x3c
kernel:  [<ffffffff8803840e>] :jbd:kjournald+0xc1/0x213
kernel:  [<ffffffff8009c446>] autoremove_wake_function+0x0/0x2e
kernel:  [<ffffffff8009c283>] keventd_create_kthread+0x0/0x61
kernel:  [<ffffffff8803834d>] :jbd:kjournald+0x0/0x213
kernel:  [<ffffffff8009c283>] keventd_create_kthread+0x0/0x61
kernel:  [<ffffffff800321d5>] kthread+0xfe/0x132
kernel:  [<ffffffff8005cfb1>] child_rip+0xa/0x11
kernel:  [<ffffffff8009c283>] keventd_create_kthread+0x0/0x61
kernel:  [<ffffffff800320d7>] kthread+0x0/0x132
kernel:  [<ffffffff8005cfa7>] child_rip+0x0/0x11                       

Call trace #2

kernel:  [<ffffffff80058c6c>] generic_unplug_device+0x5d/0xc6
kernel:  [<ffffffff8820ea3e>] :dm_mod:dm_table_unplug_all+0x33/0x41
kernel:  [<ffffffff8820cc85>] :dm_mod:dm_unplug_all+0x1d/0x28
kernel:  [<ffffffff8005a78a>] blk_backing_dev_unplug+0x56/0x5b
kernel:  [<ffffffff800e8bfe>] __blockdev_direct_IO+0x889/0xaa2
kernel:  [<ffffffff88050800>] :ext3:ext3_direct_IO+0xf3/0x18b
kernel:  [<ffffffff8804ec84>] :ext3:ext3_get_block+0x0/0xe3
kernel:  [<ffffffff800be6bb>] generic_file_direct_IO+0xbd/0xfb
kernel:  [<ffffffff8001e637>] generic_file_direct_write+0x60/0xf2
kernel:  [<ffffffff80015cfd>] __generic_file_aio_write_nolock+0x2b7/0x3b8
kernel:  [<ffffffff8002134f>] generic_file_aio_write+0x65/0xc1
Quoted 8 lines
kernel:  [<ffffffff8804c192>] :ext3:ext3_file_write+0x16/0x91
kernel:  [<ffffffff80017944>] do_sync_write+0xc7/0x104
kernel:  [<ffffffff8009c446>] autoremove_wake_function+0x0/0x2e
kernel:  [<ffffffff80111400>] free_msg+0x22/0x3c
kernel:  [<ffffffff800161c4>] vfs_write+0xce/0x174
kernel:  [<ffffffff8004194c>] sys_pwrite64+0x50/0x70
kernel:  [<ffffffff8005cde9>] error_exit+0x0/0x84
kernel:  [<ffffffff8005c116>] system_call+0x7e/0x83

So it's basically dm calling into blk_unplug() all the time, which doesn't check if the queue is plugged. The reason why I didn't like the initial patch is that ->unplug_fn() really should not be called unless the queue IS plugged. So how about this instead:

http://git.kernel.dk/?p=linux-2.6-block.git;a=commit;h=c44993018887e82abd49023e92e8d8b6000e03ed

That's a lot more appropriate, imho.

Signature · 1 line
Jens Axboe
#6 lore raw
Quoted 8 lines · Jens Axboe

So it's basically dm calling into blk_unplug() all the time, which doesn't check if the queue is plugged. The reason why I didn't like the initial patch is that ->unplug_fn() really should not be called unless the queue IS plugged. So how about this instead:

http://git.kernel.dk/?p=linux-2.6-block.git;a=commit;h=c44993018887e82abd49023e92e8d8b6000e03ed

That's a lot more appropriate, imho.

Makes sense to me, ack.

Alasdair

Signature · 1 line
[email protected]
#7 lore raw
In reply to Jens Axboe
Quoted 82 lines · Jens Axboe
Quoted 72 lines · Mike Anderson
Quoted 21 lines · Jens Axboe
Quoted 11 lines · Mikulas Patocka

Hi

Mike Anderson was doing an OLTP benchmark on a computer with 48 physical disks mapped to one logical device via device mapper.

He found that there was a slowdown on request_queue->lock in function generic_unplug_device. The slowdown is caused by the fact that when some code calls unplug on the device mapper, device mapper calls unplug on all physical disks. These unplug calls take the lock, find that the queue is already unplugged, release the lock and exit.

Quoted 8 lines

With the below patch, performance of the benchmark was increased by 18% (the whole OLTP application, not just block layer microbenchmarks).

So I'm submitting this patch for upstream. I think the patch is correct, because when more threads call simultaneously plug and unplug, it is unspecified, if the queue is or isn't plugged (so the patch can't make this worse). And the caller that plugged the queue should unplug it anyway. (if it doesn't, there's 3ms timeout).

Quoted 4 lines

Where were these unplug calls coming from? The block layer will generally only unplug when it is already unplugged, so if you are seeing so many unplug calls that the patch redues overhead by as much described, perhaps the callsite is buggy?

I do not have direct access the the benchmark setup, but here is the data I have received.

The oprofile data was showing ll_rw_blk::generic_unplug_device() as a top routine at 13% of the samples. Annotation of the samples shows hits on spin_lock_irq(q->queue_lock).

Here are some sample call traces:

Call trace #1

kernel:  [<ffffffff80058c6c>] generic_unplug_device+0x5d/0xc6
kernel:  [<ffffffff8820ea3e>] :dm_mod:dm_table_unplug_all+0x33/0x41
kernel:  [<ffffffff8820cc85>] :dm_mod:dm_unplug_all+0x1d/0x28
kernel:  [<ffffffff8005a78a>] blk_backing_dev_unplug+0x56/0x5b
kernel:  [<ffffffff80014cdc>] sync_buffer+0x36/0x3f
kernel:  [<ffffffff800629a4>] __wait_on_bit+0x40/0x6f
kernel:  [<ffffffff80014ca6>] sync_buffer+0x0/0x3f
kernel:  [<ffffffff80062a3f>] out_of_line_wait_on_bit+0x6c/0x78
kernel:  [<ffffffff8009c474>] wake_bit_function+0x0/0x23
kernel:  [<ffffffff88034c85>] :jbd:journal_commit_transaction+0x91f/0x1086
kernel:  [<ffffffff8003d038>] lock_timer_base+0x1b/0x3c
kernel:  [<ffffffff8803840e>] :jbd:kjournald+0xc1/0x213
kernel:  [<ffffffff8009c446>] autoremove_wake_function+0x0/0x2e
kernel:  [<ffffffff8009c283>] keventd_create_kthread+0x0/0x61
kernel:  [<ffffffff8803834d>] :jbd:kjournald+0x0/0x213
kernel:  [<ffffffff8009c283>] keventd_create_kthread+0x0/0x61
kernel:  [<ffffffff800321d5>] kthread+0xfe/0x132
kernel:  [<ffffffff8005cfb1>] child_rip+0xa/0x11
kernel:  [<ffffffff8009c283>] keventd_create_kthread+0x0/0x61
kernel:  [<ffffffff800320d7>] kthread+0x0/0x132
kernel:  [<ffffffff8005cfa7>] child_rip+0x0/0x11

Call trace #2

kernel:  [<ffffffff80058c6c>] generic_unplug_device+0x5d/0xc6
kernel:  [<ffffffff8820ea3e>] :dm_mod:dm_table_unplug_all+0x33/0x41
kernel:  [<ffffffff8820cc85>] :dm_mod:dm_unplug_all+0x1d/0x28
kernel:  [<ffffffff8005a78a>] blk_backing_dev_unplug+0x56/0x5b
kernel:  [<ffffffff800e8bfe>] __blockdev_direct_IO+0x889/0xaa2
kernel:  [<ffffffff88050800>] :ext3:ext3_direct_IO+0xf3/0x18b
kernel:  [<ffffffff8804ec84>] :ext3:ext3_get_block+0x0/0xe3
kernel:  [<ffffffff800be6bb>] generic_file_direct_IO+0xbd/0xfb
kernel:  [<ffffffff8001e637>] generic_file_direct_write+0x60/0xf2
kernel:  [<ffffffff80015cfd>] __generic_file_aio_write_nolock+0x2b7/0x3b8
kernel:  [<ffffffff8002134f>] generic_file_aio_write+0x65/0xc1
Quoted 8 lines
kernel:  [<ffffffff8804c192>] :ext3:ext3_file_write+0x16/0x91
kernel:  [<ffffffff80017944>] do_sync_write+0xc7/0x104
kernel:  [<ffffffff8009c446>] autoremove_wake_function+0x0/0x2e
kernel:  [<ffffffff80111400>] free_msg+0x22/0x3c
kernel:  [<ffffffff800161c4>] vfs_write+0xce/0x174
kernel:  [<ffffffff8004194c>] sys_pwrite64+0x50/0x70
kernel:  [<ffffffff8005cde9>] error_exit+0x0/0x84
kernel:  [<ffffffff8005c116>] system_call+0x7e/0x83
Quoted 11 lines

So it's basically dm calling into blk_unplug() all the time, which doesn't check if the queue is plugged. The reason why I didn't like the initial patch is that ->unplug_fn() really should not be called unless the queue IS plugged. So how about this instead:

http://git.kernel.dk/?p=linux-2.6-block.git;a=commit;h=c44993018887e82abd49023e92e8d8b6000e03ed

That's a lot more appropriate, imho.

Signature · 1 line
Jens Axboe

This doesn't seem correct to me. The difference between blk_unplug and generic_unplug_device is that blk_unplug is called on every type of device and generic_unplug_device (pointed to by q->unplug_fn) is a method that is called on low-level disk devices.

dm and md redefine q->unplug_fn to point to their own method. On dm and md, blk_unplug is called, but generic_unplug_device is not.

So if you have this setup
dm-linear(unplugged) -> disk(plugged)

then, with your patch, a call to blk_unplug(dm-linear) will not unplug the disk. With my patch, a call to blk_unplug(dm-linear) will unplug the disk --- it calls q->unplug_fn that points to dm_unplug_all, that calls blk_unplug again on the disk and that calls generic_unplug_device on disk queue.

Mikulas

#8 lore raw
Quoted 107 lines · Mikulas Patocka
Quoted 83 lines · Jens Axboe
Quoted 73 lines · Mike Anderson
Quoted 22 lines · Jens Axboe
Quoted 12 lines · Mikulas Patocka

Hi

Mike Anderson was doing an OLTP benchmark on a computer with 48 physical disks mapped to one logical device via device mapper.

He found that there was a slowdown on request_queue->lock in function generic_unplug_device. The slowdown is caused by the fact that when some code calls unplug on the device mapper, device mapper calls unplug on all
physical disks. These unplug calls take the lock, find that the queue is already unplugged, release the lock and exit.

Quoted 8 lines

With the below patch, performance of the benchmark was increased by 18% (the whole OLTP application, not just block layer microbenchmarks).

So I'm submitting this patch for upstream. I think the patch is correct, because when more threads call simultaneously plug and unplug, it is unspecified, if the queue is or isn't plugged (so the patch can't make this worse). And the caller that plugged the queue should unplug it anyway. (if it doesn't, there's 3ms timeout).

Quoted 4 lines

Where were these unplug calls coming from? The block layer will generally only unplug when it is already unplugged, so if you are seeing so many unplug calls that the patch redues overhead by as much described, perhaps the callsite is buggy?

I do not have direct access the the benchmark setup, but here is the data I have received.

The oprofile data was showing ll_rw_blk::generic_unplug_device() as a top routine at 13% of the samples. Annotation of the samples shows hits on spin_lock_irq(q->queue_lock).

Here are some sample call traces:

Call trace #1

kernel:  [<ffffffff80058c6c>] generic_unplug_device+0x5d/0xc6
kernel:  [<ffffffff8820ea3e>] :dm_mod:dm_table_unplug_all+0x33/0x41
kernel:  [<ffffffff8820cc85>] :dm_mod:dm_unplug_all+0x1d/0x28
kernel:  [<ffffffff8005a78a>] blk_backing_dev_unplug+0x56/0x5b
kernel:  [<ffffffff80014cdc>] sync_buffer+0x36/0x3f
kernel:  [<ffffffff800629a4>] __wait_on_bit+0x40/0x6f
kernel:  [<ffffffff80014ca6>] sync_buffer+0x0/0x3f
kernel:  [<ffffffff80062a3f>] out_of_line_wait_on_bit+0x6c/0x78
kernel:  [<ffffffff8009c474>] wake_bit_function+0x0/0x23
kernel:  [<ffffffff88034c85>] :jbd:journal_commit_transaction+0x91f/0x1086
kernel:  [<ffffffff8003d038>] lock_timer_base+0x1b/0x3c
kernel:  [<ffffffff8803840e>] :jbd:kjournald+0xc1/0x213
kernel:  [<ffffffff8009c446>] autoremove_wake_function+0x0/0x2e
kernel:  [<ffffffff8009c283>] keventd_create_kthread+0x0/0x61
kernel:  [<ffffffff8803834d>] :jbd:kjournald+0x0/0x213
kernel:  [<ffffffff8009c283>] keventd_create_kthread+0x0/0x61
kernel:  [<ffffffff800321d5>] kthread+0xfe/0x132
kernel:  [<ffffffff8005cfb1>] child_rip+0xa/0x11
kernel:  [<ffffffff8009c283>] keventd_create_kthread+0x0/0x61
kernel:  [<ffffffff800320d7>] kthread+0x0/0x132
kernel:  [<ffffffff8005cfa7>] child_rip+0x0/0x11

Call trace #2

kernel:  [<ffffffff80058c6c>] generic_unplug_device+0x5d/0xc6
kernel:  [<ffffffff8820ea3e>] :dm_mod:dm_table_unplug_all+0x33/0x41
kernel:  [<ffffffff8820cc85>] :dm_mod:dm_unplug_all+0x1d/0x28
kernel:  [<ffffffff8005a78a>] blk_backing_dev_unplug+0x56/0x5b
kernel:  [<ffffffff800e8bfe>] __blockdev_direct_IO+0x889/0xaa2
kernel:  [<ffffffff88050800>] :ext3:ext3_direct_IO+0xf3/0x18b
kernel:  [<ffffffff8804ec84>] :ext3:ext3_get_block+0x0/0xe3
kernel:  [<ffffffff800be6bb>] generic_file_direct_IO+0xbd/0xfb
kernel:  [<ffffffff8001e637>] generic_file_direct_write+0x60/0xf2
kernel:  [<ffffffff80015cfd>] __generic_file_aio_write_nolock+0x2b7/0x3b8
kernel:  [<ffffffff8002134f>] generic_file_aio_write+0x65/0xc1
Quoted 8 lines
kernel:  [<ffffffff8804c192>] :ext3:ext3_file_write+0x16/0x91
kernel:  [<ffffffff80017944>] do_sync_write+0xc7/0x104
kernel:  [<ffffffff8009c446>] autoremove_wake_function+0x0/0x2e
kernel:  [<ffffffff80111400>] free_msg+0x22/0x3c
kernel:  [<ffffffff800161c4>] vfs_write+0xce/0x174
kernel:  [<ffffffff8004194c>] sys_pwrite64+0x50/0x70
kernel:  [<ffffffff8005cde9>] error_exit+0x0/0x84
kernel:  [<ffffffff8005c116>] system_call+0x7e/0x83
Quoted 11 lines

So it's basically dm calling into blk_unplug() all the time, which doesn't check if the queue is plugged. The reason why I didn't like the initial patch is that ->unplug_fn() really should not be called unless the queue IS plugged. So how about this instead:

http://git.kernel.dk/?p=linux-2.6-block.git;a=commit;h=c44993018887e82abd49023e92e8d8b6000e03ed

That's a lot more appropriate, imho.

Signature · 1 line
Jens Axboe

This doesn't seem correct to me. The difference between blk_unplug and generic_unplug_device is that blk_unplug is called on every type of device and generic_unplug_device (pointed to by q->unplug_fn) is a method that is called on low-level disk devices.

dm and md redefine q->unplug_fn to point to their own method. On dm and md, blk_unplug is called, but generic_unplug_device is not.

Quoted 8 lines

So if you have this setup
dm-linear(unplugged) -> disk(plugged)

then, with your patch, a call to blk_unplug(dm-linear) will not unplug the disk. With my patch, a call to blk_unplug(dm-linear) will unplug the disk --- it calls q->unplug_fn that points to dm_unplug_all, that calls blk_unplug again on the disk and that calls generic_unplug_device on disk queue.

That is because the md/dm don't set the plugged flag which I think they should. So we fix that instead so that plugging works the same from the block core or from a driver instead of adding work-arounds in the block unplug handler. Adding a check for plugged in the plug handler is a hack, I don't see how you can argue against that.

Signature · 1 line
Jens Axboe
#9 lore raw
Quoted 29 lines
Quoted 20 lines
Quoted 11 lines

So it's basically dm calling into blk_unplug() all the time, which doesn't check if the queue is plugged. The reason why I didn't like the initial patch is that ->unplug_fn() really should not be called unless the queue IS plugged. So how about this instead:

http://git.kernel.dk/?p=linux-2.6-block.git;a=commit;h=c44993018887e82abd49023e92e8d8b6000e03ed

That's a lot more appropriate, imho.

--
Jens Axboe

This doesn't seem correct to me. The difference between blk_unplug and generic_unplug_device is that blk_unplug is called on every type of device and generic_unplug_device (pointed to by q->unplug_fn) is a method that is called on low-level disk devices.

dm and md redefine q->unplug_fn to point to their own method. On dm and md, blk_unplug is called, but generic_unplug_device is not.

Quoted 8 lines

So if you have this setup
dm-linear(unplugged) -> disk(plugged)

then, with your patch, a call to blk_unplug(dm-linear) will not unplug the disk. With my patch, a call to blk_unplug(dm-linear) will unplug the disk --- it calls q->unplug_fn that points to dm_unplug_all, that calls blk_unplug again on the disk and that calls generic_unplug_device on disk queue.

Quoted 4 lines

That is because the md/dm don't set the plugged flag which I think they should. So we fix that instead so that plugging works the same from the block core or from a driver instead of adding work-arounds in the block unplug handler.

When should dm/md set the plugged flag? It has several disks (or more dm/md layers), some of them may be plugged, some not --- furthermore, the disk queues are shared by other devices --- there may be more devices on different partitions.

So the question: when do you want dm/md to set the plugged bit? Do you really want to plug the queue at the top layer and merge requests there?

Quoted 2 lines

Adding a check for plugged in the plug handler is a hack, I don't see how you can argue against that.

I changed generic_unplug_device from:
spin_lock()
if (plugged bit is set) unplug...;
spin_unlock()

to:
if (plugged bit is set)
{
spin_lock()
if (plugged bit is set) unplug...;
spin_unlock()
}

--- I don't see anything wrong with it. At least, the change is so trivial that we can be sure that it won't have negative effect on performance.

What you propose is complete rewrite of dm/md plugging mechanism to plug the queue and merge requests at upper layer --- the questions are:

- what exactly you are proposing? (plugging at upper layer? lower layer? both layers? don't plug and just somehow propagate the plugged bit?)
- why are you proposing that?

Note that for some dm targets it would be benefical to join the requests at upper layer (dm-linear, raid1), for others (raid0, dm-snapshots) it damages performance (you merge the requests before passing them to raid0 and you chop them again to smaller pieces in raid0).

Mikulas

Quoted 3 lines
Signature · 1 line
Jens Axboe
#10 lore raw
Quoted 64 lines · Mikulas Patocka
Quoted 29 lines
Quoted 20 lines
Quoted 11 lines

So it's basically dm calling into blk_unplug() all the time, which doesn't check if the queue is plugged. The reason why I didn't like the initial patch is that ->unplug_fn() really should not be called unless the queue IS plugged. So how about this instead:

http://git.kernel.dk/?p=linux-2.6-block.git;a=commit;h=c44993018887e82abd49023e92e8d8b6000e03ed

That's a lot more appropriate, imho.

--
Jens Axboe

This doesn't seem correct to me. The difference between blk_unplug and generic_unplug_device is that blk_unplug is called on every type of device and generic_unplug_device (pointed to by q->unplug_fn) is a method that is called on low-level disk devices.

dm and md redefine q->unplug_fn to point to their own method. On dm and md, blk_unplug is called, but generic_unplug_device is not.

Quoted 8 lines

So if you have this setup
dm-linear(unplugged) -> disk(plugged)

then, with your patch, a call to blk_unplug(dm-linear) will not unplug the disk. With my patch, a call to blk_unplug(dm-linear) will unplug the disk --- it calls q->unplug_fn that points to dm_unplug_all, that calls blk_unplug again on the disk and that calls generic_unplug_device on disk queue.

Quoted 4 lines

That is because the md/dm don't set the plugged flag which I think they should. So we fix that instead so that plugging works the same from the block core or from a driver instead of adding work-arounds in the block unplug handler.

When should dm/md set the plugged flag? It has several disks (or more dm/md layers), some of them may be plugged, some not --- furthermore, the disk queues are shared by other devices --- there may be more devices on different partitions.

So the question: when do you want dm/md to set the plugged bit? Do you really want to plug the queue at the top layer and merge requests there?

Quoted 2 lines

Adding a check for plugged in the plug handler is a hack, I don't see how you can argue against that.

I changed generic_unplug_device from:
spin_lock()
if (plugged bit is set) unplug...;
spin_unlock()

to:
if (plugged bit is set)
{
spin_lock()
if (plugged bit is set) unplug...;
spin_unlock()
}

--- I don't see anything wrong with it. At least, the change is so trivial that we can be sure that it won't have negative effect on performance.

What you propose is complete rewrite of dm/md plugging mechanism to plug the queue and merge requests at upper layer --- the questions are:

Quoted 8 lines

- what exactly you are proposing? (plugging at upper layer? lower layer? both layers? don't plug and just somehow propagate the plugged bit?)
- why are you proposing that?

Note that for some dm targets it would be benefical to join the requests at upper layer (dm-linear, raid1), for others (raid0, dm-snapshots) it damages performance (you merge the requests before passing them to raid0 and you chop them again to smaller pieces in raid0).

It will indeed get a lot more involved. Sigh. Well I'll apply your patch to generic_unplug_device(), it's at least a worth while optimization in the mean time.

Signature · 1 line
Jens Axboe