| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-Id: <20190814103244.92518-1-maco@android.com>
Date: Wed, 14 Aug 2019 12:32:44 +0200
From: Martijn Coenen <maco@...roid.com>
To: axboe@...nel.dk
Cc: linux-block@...r.kernel.org, linux-kernel@...r.kernel.org,
gregkh@...uxfoundation.org, kernel-team@...roid.com,
narayan@...gle.com, dariofreni@...gle.com, ioffe@...gle.com,
jiyong@...gle.com, maco@...gle.com,
Martijn Coenen <maco@...roid.com>
Subject: [PATCH] RFC: loop: Avoid calling blk_mq_freeze_queue() when possible.
Since Android Q, the creation and configuration of loop devices is in
the critical path of device boot. We found that the configuration of
loop devices is pretty slow, because many ioctl()'s involve freezing the
block queue, which in turn needs to wait for an RCU grace period. On
Android devices we've observed up to 60ms for the creation and
configuration of a single loop device; as we anticipate creating many
more in the future, we'd like to avoid this delay.
This allows LOOP_SET_BLOCK_SIZE to be called before the loop device has
been bound; since the block queue is not running at that point, we can
avoid the expensive freezing of the queue.
On a recent x86, this patch yields the following results:
===
Call LOOP_SET_BLOCK_SIZE on /dev/loop0 before being bound
===
~# time ./set_block_size
real 0m0.002s
user 0m0.000s
sys 0m0.002s
===
Call LOOP_SET_BLOCK_SIZE on /dev/loop0 after being bound
===
~# losetup /dev/loop0 fs.img
~# time ./set_block_size
real 0m0.008s
user 0m0.000s
sys 0m0.002s
Over many runs, this is a 4x improvement.
This is RFC because technically it is a change in behavior; before,
calling LOOP_SET_BLOCK_SIZE on an unbound device would return ENXIO, and
userspace programs that left it in their code despite the returned
error, would now suddenly see the requested value effectuated. I'm not
sure whether this is acceptable.
An alternative might be a CONFIG option to set the default block size to
another value than 512. Another alternative I considered is allowing the
block device to be created with a "frozen" queue, where we can manually
unfreeze the queue when all the configuration is done. This would be a
much larger code change, though.
Signed-off-by: Martijn Coenen <maco@...roid.com>
---
drivers/block/loop.c | 42 +++++++++++++++++++++++-------------------
1 file changed, 23 insertions(+), 19 deletions(-)
diff --git a/drivers/block/loop.c b/drivers/block/loop.c
index ab7ca5989097a..d4348a4fdd7a6 100644
--- a/drivers/block/loop.c
+++ b/drivers/block/loop.c
@@ -214,7 +214,8 @@ static void __loop_update_dio(struct loop_device *lo, bool dio)
* LO_FLAGS_READ_ONLY, both are set from kernel, and losetup
* will get updated by ioctl(LOOP_GET_STATUS)
*/
- blk_mq_freeze_queue(lo->lo_queue);
+ if (lo->lo_state == Lo_bound)
+ blk_mq_freeze_queue(lo->lo_queue);
lo->use_dio = use_dio;
if (use_dio) {
blk_queue_flag_clear(QUEUE_FLAG_NOMERGES, lo->lo_queue);
@@ -223,7 +224,8 @@ static void __loop_update_dio(struct loop_device *lo, bool dio)
blk_queue_flag_set(QUEUE_FLAG_NOMERGES, lo->lo_queue);
lo->lo_flags &= ~LO_FLAGS_DIRECT_IO;
}
- blk_mq_unfreeze_queue(lo->lo_queue);
+ if (lo->lo_state == Lo_bound)
+ blk_mq_unfreeze_queue(lo->lo_queue);
}
static int
@@ -621,6 +623,8 @@ static int do_req_filebacked(struct loop_device *lo, struct request *rq)
static inline void loop_update_dio(struct loop_device *lo)
{
+ if (lo->lo_state != Lo_bound)
+ return;
__loop_update_dio(lo, io_is_direct(lo->lo_backing_file) |
lo->use_dio);
}
@@ -1510,27 +1514,26 @@ static int loop_set_block_size(struct loop_device *lo, unsigned long arg)
{
int err = 0;
- if (lo->lo_state != Lo_bound)
- return -ENXIO;
-
if (arg < 512 || arg > PAGE_SIZE || !is_power_of_2(arg))
return -EINVAL;
- if (lo->lo_queue->limits.logical_block_size != arg) {
- sync_blockdev(lo->lo_device);
- kill_bdev(lo->lo_device);
- }
+ if (lo->lo_state == Lo_bound) {
+ if (lo->lo_queue->limits.logical_block_size != arg) {
+ sync_blockdev(lo->lo_device);
+ kill_bdev(lo->lo_device);
+ }
- blk_mq_freeze_queue(lo->lo_queue);
+ blk_mq_freeze_queue(lo->lo_queue);
- /* kill_bdev should have truncated all the pages */
- if (lo->lo_queue->limits.logical_block_size != arg &&
- lo->lo_device->bd_inode->i_mapping->nrpages) {
- err = -EAGAIN;
- pr_warn("%s: loop%d (%s) has still dirty pages (nrpages=%lu)\n",
- __func__, lo->lo_number, lo->lo_file_name,
- lo->lo_device->bd_inode->i_mapping->nrpages);
- goto out_unfreeze;
+ /* kill_bdev should have truncated all the pages */
+ if (lo->lo_queue->limits.logical_block_size != arg &&
+ lo->lo_device->bd_inode->i_mapping->nrpages) {
+ err = -EAGAIN;
+ pr_warn("%s: loop%d (%s) has still dirty pages (nrpages=%lu)\n",
+ __func__, lo->lo_number, lo->lo_file_name,
+ lo->lo_device->bd_inode->i_mapping->nrpages);
+ goto out_unfreeze;
+ }
}
blk_queue_logical_block_size(lo->lo_queue, arg);
@@ -1538,7 +1541,8 @@ static int loop_set_block_size(struct loop_device *lo, unsigned long arg)
blk_queue_io_min(lo->lo_queue, arg);
loop_update_dio(lo);
out_unfreeze:
- blk_mq_unfreeze_queue(lo->lo_queue);
+ if (lo->lo_state == Lo_bound)
+ blk_mq_unfreeze_queue(lo->lo_queue);
return err;
}
--
2.23.0.rc1.153.gdeed80330f-goog
Powered by blists - more mailing lists