... that there was a slowdown on request_queue->lock in function >> generic_unplug_device. The slowdown is ... , find that the queue is >> already unplugged, release the lock and exit. >> >> With the below patch, performance of the benchmark was increased by ...

linux-kernel - Mikulas Patocka - Apr 29 2008 - 16:02

... , find that the queue is > >>already unplugged, release the lock and exit. > >> ... you have this situation > > dm-linear(unplugged) -> physical-disk(plugged) > > then uplung should ... driver, not in upper layers. Fair enough, I'll put the patch under ...

linux-kernel - Jens Axboe - Apr 29 2008 - 16:05

... there was a slowdown on request_queue->lock in function >>>> generic_unplug_device. The slowdown is ... , find that the queue is >>>> already unplugged, release the lock and exit. >>>> >>>> With the below patch, performance of the benchmark was increased ...

linux-kernel - Mikulas Patocka - Apr 30 2008 - 09:54

... that there was a slowdown on request_queue->lock in function > > > generic_unplug_device. The slowdown is ... find that the queue is > > > already unplugged, release the lock and exit. > > > > > > With the below patch, performance of the benchmark was ...

linux-kernel - Jens Axboe - Apr 30 2008 - 03:14

... there was a slowdown on request_queue->lock in function > >>>>generic_unplug_device. The slowdown is ... find that the queue is > >>>>already unplugged, release the lock and exit. > >>>> > >>>>With the below patch, performance of the benchmark was increased ...

linux-kernel - Jens Axboe - May 4 2008 - 15:11

... that there was a slowdown on request_queue->lock in function > generic_unplug_device. The slowdown is caused ... find that the queue is > already unplugged, release the lock and exit. > > With the below patch, performance of the benchmark was increased ...

linux-kernel - Jens Axboe - Apr 29 2008 - 15:25

... there was a slowdown on request_queue->lock in function > > generic_unplug_device. The slowdown is ... , find that the queue is > > already unplugged, release the lock and exit. > > > > With the below patch, performance of the benchmark was increased ...

linux-kernel - Mike Anderson - Apr 29 2008 - 16:29

... > >>dm-linear(unplugged) -> disk(plugged) > >> > >>then, with your patch, a call to ... and that calls generic_unplug_device on disk > >>queue. > > > >That is because the md/dm ... 's at least a worth while optimization in the mean time. -- Jens Axboe ...

linux-kernel - Jens Axboe - May 7 2008 - 03:45

... >> dm-linear(unplugged) -> disk(plugged) >> >> then, with your patch, a call to ... and that calls generic_unplug_device on disk >> queue. > > That is because the md/dm ... them again to smaller pieces in raid0). Mikulas > -- > Jens Axboe > -- To ...

linux-kernel - Mikulas Patocka - May 5 2008 - 00:01

... > 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 ... the line "unsubscribe linux-kernel" in the body of a message to ...

linux-kernel - Alasdair G Kergon - Apr 30 2008 - 06:38