loop: Only change blocksize when needed.
PATCH 1/1 loop: Only change blocksize when needed.
Return early in loop_set_block_size() if the requested block size is identical to the one we already have; this avoids expensive calls to freeze the block queue.
--- drivers/block/loop.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-)
drivers/block/loop.c +6 -6
@@ -1541,16 +1541,16 @@ static int loop_set_block_size(struct loop_device *lo, unsigned long arg)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_queue->limits.logical_block_size == arg) +return 0 ; + +sync_blockdev (lo->lo_device); +kill_bdev (lo->lo_device);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) { +if (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,
Signature · 1 line
2.25.1.481.gfbce0eb801-goog
Quoted 5 lines · Martijn Coenen
Return early in
loop_set_block_size()if the requested block size is identical to the one we already have; this avoids expensive calls to freeze the block queue.Signed-off-by Martijn Coenen <[email protected]>
Looks good,
Quoted 3 lines · Martijn Coenen
Return early in
loop_set_block_size()if the requested block size is identical to the one we already have; this avoids expensive calls to freeze the block queue.
Applied, thanks.
Signature · 1 line
Jens Axboe
Logically this is a right thing to do, but I wonder how much speedup you are getting with these improvements ?
It will be great if you have some numbers so we all know the speedup.
Irrespective of that, this looks good to me.
Quoted 5 lines · Martijn Coenen
Return early in
loop_set_block_size()if the requested block size is identical to the one we already have; this avoids expensive calls to freeze the block queue.Signed-off-by Martijn Coenen<[email protected]>
Hi Chaitanya,
Quoted 4 lines · Chaitanya Kulkarni
Logically this is a right thing to do, but I wonder how much speedup you are getting with these improvements ?
It will be great if you have some numbers so we all know the speedup.
What makes blk_mq_freeze_queue() relatively expensive is that the implementation of that function calls synchronize_rcu() (for good reasons); on our x86 devices, I've seen that take 15-20ms on average. Recent Android versions configure a lot (~30) of loop devices at boot, and so this saves us about 600ms of boot time. I strongly suspect this benefits other usecases besides Android.
There is another call in loop_set_status() which is harder to remove; eg, if you use loop_set_status() to change the offset of a loop device, you shouldn't do that if there are still requests outstanding, and blk_mq_freeze_queue() ensures that. But in our specific case, we know that there won't be requests outstanding, because during this phase of boot the loop device hasn't been mounted yet. But we can't tell the kernel that. So I will follow up with a patch that tries to address that issue, which will also have some more detailed numbers. If you have ideas or suggestions, feel free to let me know!
Thanks,
Martijn
Quoted 12 lines
Irrespective of that, this looks good to me.
Reviewed-by Chaitanya Kulkarni <[email protected]>Quoted 5 lines · Martijn Coenen
Return early in
loop_set_block_size()if the requested block size is identical to the one we already have; this avoids expensive calls to freeze the block queue.Signed-off-by Martijn Coenen<[email protected]>