Skip to contentSkip to conversation

block: Add test-iosched scheduler

PATCH Superseded Maya Erez 2 participants 5 messages Jun 6, 2012 – Jun 11, 2012

Subsystems (4)

v2 of 4 Superseded Latest is v4 →

lore
Share & tools
lore

Maya Erez

author
#0 lore raw

v2 cover block: Add test-iosched scheduler

The test scheduler allows testing a block device by dispatching specific requests according to the test case and declare PASS/FAIL according to the requests completion error code

Changes in v2:
- Export test-iosched functionality to allow definition of the block device tests under the block device layer
- Add registration of block device tests utilities

Maya Erez (1):
block: Add test-iosched scheduler

 Documentation/block/test-iosched.txt |   39 ++
 block/Kconfig.iosched                |    8 +
 block/Makefile                       |    1 +
 block/blk-core.c                     |    3 +-
 block/test-iosched.c                 | 1025 ++++++++++++++++++++++++++++++++++
 include/linux/test-iosched.h         |  218 +++++++
 6 files changed, 1292 insertions(+), 2 deletions(-)
 create mode 100644 Documentation/block/test-iosched.txt
 create mode 100644 block/test-iosched.c
 create mode 100644 include/linux/test-iosched.h

--
1.7.3.3

Signature · 2 lines
Sent by a consultant of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
author
#1 lore raw

PATCH v2 1/1 block: Add test-iosched scheduler

The test scheduler allows testing a block device by dispatching specific requests according to the test case and declare PASS/FAIL according to the requests completion error code

Documentation/block/test-iosched.txt new file +39 -0
new file mode 100644
index 0000000..75d8134
@@ -0,0 +1,39 @@
+Test IO scheduler
+==================
+
+The test scheduler allows testing a block device by dispatching
+specific requests according to the test case and declare PASS/FAIL
+according to the requests completion error code.
+
+The test IO scheduler implements the no-op scheduler operations, and uses
+them in order to dispatch the non-test requests when no test is running.
+This will allow to keep a normal FS operation in parallel to the test
+capability.
+The test IO scheduler keeps two different queues, one for real-world requests
+(inserted by the FS) and the other for the test requests.
+The test IO scheduler chooses the queue for dispatch requests according to the
+test state (IDLE/RUNNING).
+
+the test IO scheduler is compiled by default as a dynamic module and enabled
+only if CONFIG_DEBUG_FS is defined.
+
+Each block device test utility that would like to use the test-iosched test
+services, should register as a blk_dev_test_type and supply an init and exit
+callbacks. Those callback are called upon selection (or removal) of the
+test-iosched as the active scheduler. From that point the block device test
+can start a test and supply its own callbacks for preparing, running, result
+checking and cleanup of the test.
+
+Each test is exposed via debugfs and can be triggered by writing to
+the debugfs file. In order to add a new test one should expose a new debugfs
+file for the new test.
+
+Selecting IO schedulers
+-----------------------
+Refer to Documentation/block/switching-sched.txt for information on
+selecting an io scheduler on a per-device basis.
+
+
+May 10 2012, maya Erez <[email protected]>
+
+
block/Kconfig.iosched +8 -0
index 421bef9..34a1f9e 100644
@@ -12,6 +12,14 @@ config IOSCHED_NOOP
 	  that do their own scheduling and require only minimal assistance from
 	  the kernel.
 
+config IOSCHED_TEST
+	tristate "Test I/O scheduler"
+	depends on DEBUG_FS
+	default m
+	---help---
+	  The test I/O scheduler is duplicate of the noop scheduler with
+	  test ability.
+
 config IOSCHED_DEADLINE
 	tristate "Deadline I/O scheduler"
 	default y
block/Makefile +1 -0
index 39b76ba..436b220 100644
@@ -15,6 +15,7 @@ obj-$(CONFIG_BLK_DEV_THROTTLING)	+= blk-throttle.o
 obj-$(CONFIG_IOSCHED_NOOP)	+= noop-iosched.o
 obj-$(CONFIG_IOSCHED_DEADLINE)	+= deadline-iosched.o
 obj-$(CONFIG_IOSCHED_CFQ)	+= cfq-iosched.o
+obj-$(CONFIG_IOSCHED_TEST)	+= test-iosched.o
 
 obj-$(CONFIG_BLOCK_COMPAT)	+= compat_ioctl.o
 obj-$(CONFIG_BLK_DEV_INTEGRITY)	+= blk-integrity.o
block/blk-core.c +1 -2
index 3c923a7..f3c3b9e 100644
@@ -1072,8 +1072,6 @@ struct request *blk_get_request(struct request_queue *q, int rw, gfp_t gfp_mask)
 {
 	struct request *rq;
 
-	BUG_ON(rw != READ && rw != WRITE);
-
 	spin_lock_irq(q->queue_lock);
 	if (gfp_mask & __GFP_WAIT)
 		rq = get_request_wait(q, rw, NULL);
@@ -1406,6 +1404,7 @@ void init_request_from_bio(struct request *req, struct bio *bio)
 	req->ioprio = bio_prio(bio);
 	blk_rq_bio_prep(req->q, req, bio);
 }
+EXPORT_SYMBOL(init_request_from_bio);
 
 void blk_queue_bio(struct request_queue *q, struct bio *bio)
 {
block/test-iosched.c new file +1025 -0

1,025 lines not on this page · raw message

include/linux/test-iosched.h new file +218 -0

218 lines not on this page · raw message

--
1.7.3.3

Signature · 2 lines
Sent by a consultant of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
#2 lore raw
In reply to Maya Erez
Quoted 4 lines · Maya Erez

The test scheduler allows testing a block device by dispatching specific requests according to the test case and declare PASS/FAIL according to the requests completion error code

I can't get the point. Isn't this possible purely from userspace using IOCTLs ? Even otherwise, requiring to modify the scheduler for each test case is definitely not scalable.

Quoted 19 lines

Changes in v2:
   - Export test-iosched functionality to allow definition of the block device tests under the block device layer
   - Add registration of block device tests utilities

Maya Erez (1):
 block: Add test-iosched scheduler

 Documentation/block/test-iosched.txt |   39 ++ block/Kconfig.iosched                |    8 +
 block/Makefile                       |    1 +
 block/blk-core.c                     |    3 +- block/test-iosched.c                 | 1025 ++++++++++++++++++++++++++++++++++ include/linux/test-iosched.h         |  218 +++++++

 6 files changed, 1292 insertions(+), 2 deletions(-)

 create mode 100644 Documentation/block/test-iosched.txt create mode 100644 block/test-iosched.c
 create mode 100644 include/linux/test-iosched.h

#3 lore raw
Quoted 10 lines · Maya Erez
Quoted 4 lines

The test scheduler allows testing a block device by dispatching specific requests according to the test case and declare PASS/FAIL according to the requests completion error code

I can't get the point. Isn't this possible purely from userspace using IOCTLs ?
Even otherwise, requiring to modify the scheduler for each test case is definitely not scalable.

The main benefit of the test-iosched is the ability to determine the timing of each request that is being dispatched and to put on hold the real FS requests so that they won't affect the tests scenario. It also allows each block device to determine pass/fail result taking into account the expected behavior and the actual result.
The scheduler doesn't have to be changed per tests case. What made you think it should be?
Currently we use the test-iosched to test the eMMC4.5 features (such as BKOPs, packed commands and sanitize). I hope that after we will release the tests later this week it will be clearer.

Quoted 10 lines
Quoted 10 lines

Changes in v2:
   - Export test-iosched functionality to allow definition of the block device
     tests under the block device layer
   - Add registration of block device tests utilities

Maya Erez (1):
 block: Add test-iosched scheduler

Quoted 13 lines
Quoted 7 lines

 Documentation/block/test-iosched.txt |   39 ++ block/Kconfig.iosched                |    8 + block/Makefile                       |    1 + block/blk-core.c                     |    3 +- block/test-iosched.c                 | 1025
++++++++++++++++++++++++++++++++++
 include/linux/test-iosched.h         |  218 +++++++

Quoted 5 lines
 6 files changed, 1292 insertions(+), 2 deletions(-)

 create mode 100644 Documentation/block/test-iosched.txt create mode 100644 block/test-iosched.c
 create mode 100644 include/linux/test-iosched.h

Thanks,
Maya Erez
Consultant for Qualcomm Innovation Center, Inc. Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum

#4 lore raw
Quoted 6 lines · On Sat, Jun 9, 2012 at 8:24 PM, <[email protected]> wrote:
Quoted 6 lines · Maya Erez
Quoted 4 lines

The test scheduler allows testing a block device by dispatching specific requests according to the test case and declare PASS/FAIL according to the requests completion error code

Quoted 7 lines
Quoted 4 lines

I can't get the point. Isn't this possible purely from userspace using IOCTLs ?
Even otherwise, requiring to modify the scheduler for each test case is definitely not scalable.

The main benefit of the test-iosched is the ability to determine the timing of each request that is being dispatched and to put on hold the real FS requests so that they won't affect the tests scenario.

Then a potentially long running test can block any useful work that can be done on the device. no-op is not the right scheduler for the example you mentioned (eMMC), so such device has to be mounted only for the purpose of running the tests.
So using standard noop + debugfs would be sufficient for 99% of the cases ?

Quoted 6 lines

It also allows each block device to determine pass/fail result taking into account the expected behavior and the actual result.

The scheduler doesn't have to be changed per tests case. What made you think it should be?

Err.. I misread this section of documentation. I read is as sysfs instead of debugfs.
My mistake..
<Quote>
+Each test is exposed via debugfs and can be triggered by writing to +the debugfs file. In order to add a new test one should expose a new debugfs +file for the new test.
</Quote>

Quoted 4 lines

Currently we use the test-iosched to test the eMMC4.5 features (such as BKOPs, packed commands and sanitize). I hope that after we will release the tests later this week it will be clearer.

Quoted 1 line

Sure. It'd be useful.