Add a test that measures how the mmc bandwidth depends on the numbers of
sg elements in the sg list. The transfer size if fixed and sg length goes
from a few up to 512. The purpose is to measure overhead caused by
multiple sg elements.
Signed-off-by: NPer Forlin <per.forlin@linaro.org>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Acked-by: NArnd Bergmann <arnd@arndb.de>
Reviewed-by: NVenkatraman S <svenkatr@ti.com>
Tested-by: NSourav Poddar <sourav.poddar@ti.com>
Tested-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NChris Ball <cjb@laptop.org>

Add four tests for read and write performance per
different transfer size, 4k to 4M.
 * Read using blocking mmc request
 * Read using non-blocking mmc request
 * Write using blocking mmc request
 * Write using non-blocking mmc request

The host driver must support pre_req() and post_req()
in order to run the non-blocking test cases.
Signed-off-by: NPer Forlin <per.forlin@linaro.org>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Acked-by: NArnd Bergmann <arnd@arndb.de>
Reviewed-by: NVenkatraman S <svenkatr@ti.com>
Tested-by: Sourav Poddar<sourav.poddar@ti.com>
Tested-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NChris Ball <cjb@laptop.org>

Add a debugfs file "testlist" to print all available tests.
Signed-off-by: NPer Forlin <per.forlin@linaro.org>
Acked-by: NKyungmin Park <kyungmin.park@samsung.com>
Acked-by: NArnd Bergmann <arnd@arndb.de>
Reviewed-by: NVenkatraman S <svenkatr@ti.com>
Tested-by: Sourav Poddar<sourav.poddar@ti.com>
Tested-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NChris Ball <cjb@laptop.org>

Signed-off-by: NAndy Shevchenko <ext-andriy.shevchenko@nokia.com>
Signed-off-by: NChris Ball <cjb@laptop.org>

Converts from:
	struct mmc_request mrq;
	memset(&mrq, 0, sizeof(struct mmc_request));

to:
	struct mmc_request mrq = {0};

because it's shorter, as performant, and easier to work out whether
initialization has happened.
Signed-off-by: NChris Ball <cjb@laptop.org>

Converts from:
	struct mmc_data data;
	memset(&data, 0, sizeof(struct mmc_data));

to:
	struct mmc_data data = {0};

because it's shorter, as performant, and easier to work out whether
initialization has happened.
Signed-off-by: NChris Ball <cjb@laptop.org>

Converts from:
	struct mmc_command cmd;
	memset(&cmd, 0, sizeof(struct mmc_command));

to:
	struct mmc_command cmd = {0};

because it's shorter, as performant, and easier to work out whether
initialization has happened.
Signed-off-by: NChris Ball <cjb@laptop.org>

Fixes generated by 'codespell' and manually reviewed.
Signed-off-by: NLucas De Marchi <lucas.demarchi@profusion.mobi>

Fixes:

drivers/mmc/card/mmc_test.c: In function ‘mmc_test_seq_perf’:
drivers/mmc/card/mmc_test.c:1878:28: warning: variable ‘ts’ set but not
used [-Wunused-but-set-variable]

There's no reason to be calling timespec_sub() here, because
mmc_test_print_avg_rate() is going to do that itself.
Signed-off-by: NChris Ball <cjb@laptop.org>
Cc: Adrian Hunter <adrian.hunter@nokia.com>

Add two large sequential I/O performance tests:
        35. Large sequential read into scattered pages
        36. Large sequential write from scattered pages

The tests measure transfer times for 10MiB, 100MiB, 1000MiB.
Signed-off-by: NAdrian Hunter <adrian.hunter@nokia.com>
Signed-off-by: NChris Ball <cjb@laptop.org>

Existing performance tests measure single or sequential I/O speed.
Add two random I/O tests:
        33. Random read performance by transfer size
        34. Random write performance by transfer size
Signed-off-by: NAdrian Hunter <adrian.hunter@nokia.com>
Signed-off-by: NChris Ball <cjb@laptop.org>

The test area size was set to the preferred erase size but for comparison
purposes it is better if it is the same size for different devices.  Make
it a multiple of preferred erase size that is greater than or equal to 4MiB.
Signed-off-by: NAdrian Hunter <adrian.hunter@nokia.com>
Signed-off-by: NChris Ball <cjb@laptop.org>

If the MMC host controller does not support waiting for card signaling
busy state (MMC_CAP_WAIT_WHILE_BUSY cap), there is no point in prining
the relevant warning message.
Signed-off-by: NPawel Moll <pawel.moll@arm.com>
Reviewed-by: NMatt Fleming <matt@console-pimps.org>
Signed-off-by: NChris Ball <cjb@laptop.org>

The DDR support patch needs the following fixes:

- The block driver does not need to know about DDR, any more
  than it needs to know about bus width.
- Not only the card must be switched to DDR mode.  The host
  controller must also be configured, which is done through
  the 'set_ios()' function.
- Do not set the DDR mode state until after the switch command
  is successful.
- Setting block length is not supported in DDR mode.  Make that
  a core function and change the other place it is used (mmc_test)
  also.
Signed-off-by: NAdrian Hunter <adrian.hunter@nokia.com>
Signed-off-by: NChris Ball <cjb@laptop.org>

Append .5 to KiB display when there are an odd number of sectors.
Signed-off-by: NAdrian Hunter <adrian.hunter@nokia.com>
Signed-off-by: NChris Ball <cjb@laptop.org>

Correctly allocate memory to meet the host controller
driver's maximum segment size and count limits.
Signed-off-by: NAdrian Hunter <adrian.hunter@nokia.com>
Signed-off-by: NChris Ball <cjb@laptop.org>

As proposed by Greg K-H it is more logical to keep files for the mmc_test
driver under debugfs.

Additionally this patch brings seq_file API for show() method.  It allows
to write unlimited data to the file.

Example of usage:
  # mount -t debugfs none /sys/kernel/debug
  # modprobe mmc_test
    [  581.395843] mmc_test mmc0:0001: Card claimed for testing.
  # echo 25 > /sys/kernel/debug/mmc0/mmc0\:0001/test
    [  604.568542] mmc0: Starting tests of card mmc0:0001...
    [  604.582733] mmc0: Test case 25. Best-case read performance into scattered pages...
    [  604.923553] mmc0: Transfer of 8192 sectors (4096 KiB) took 0.124664314 seconds (33644 kB/s, 32856 KiB/s)
    [  604.933227] mmc0: Result: OK
    [  604.936248] mmc0: Tests completed.
  # cat /sys/kernel/debug/mmc0/mmc0\:0001/test
    Test 25: 0
    1 8192 0.124664314 33644784
Signed-off-by: NAndy Shevchenko <ext-andriy.shevchenko@nokia.com>
Cc: Greg KH <greg@kroah.com>
Cc: Adrian Hunter <adrian.hunter@nokia.com>
Cc: Chris Ball <cjb@laptop.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NChris Ball <cjb@laptop.org>

Make it possible to get test results via sysfs.  It helps to do tests
non-interactively.  We have the file created under sysfs already and
can use it to show test results.

Prior to this patch, the "test" file under each card's sysfs node was
write-only, and results were obtained by looking at dmesg.  This patch
improves programmatic access to the test results, making them available by
reading back from the same "test" file:

[root@host mmc0:e624]# echo 6 > test
[root@host mmc0:e624]# cat test
Test 6: 2

[cjb@laptop.org: changelog improvements]
Signed-off-by: NAndy Shevchenko <ext-andriy.shevchenko@nokia.com>
Cc: Chris Ball <cjb@laptop.org>
Cc: <linux-mmc@vger.kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NChris Ball <cjb@laptop.org>

It's better to use strict_strtol() to convert user's input and strictly
check it. At least it forbids to interpret wrong input as a 0 and
prevents to run all tests.
Signed-off-by: NAndy Shevchenko <ext-andriy.shevchenko@nokia.com>
Cc: <linux-mmc@vger.kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NChris Ball <cjb@laptop.org>

There are methods to check card type. Let's use them instead of direct checking
type bits.
Signed-off-by: NAndy Shevchenko <ext-andriy.shevchenko@nokia.com>
Cc: <linux-mmc@vger.kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NChris Ball <cjb@laptop.org>

The host controller driver limits I/O transfers to maximum
transfer size, maximum block count, maximum segment size
and maximum segment count.  The performance tests were
not obeying these limits which meant they would not work
with some drivers.  This patch fixes that.
Signed-off-by: NAdrian Hunter <adrian.hunter@nokia.com>
Signed-off-by: NChris Ball <cjb@laptop.org>

Global symbols should use their subsystem name in a prefixed fashion.
Signed-off-by: NAndy Shevchenko <ext-andriy.shevchenko@nokia.com>
Cc: <linux-mmc@vger.kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NChris Ball <cjb@laptop.org>

Conversion from struct device to struct mmc_card is used more than in one
place.  Due to this it's better to have public macro for such thing.
Signed-off-by: NAndy Shevchenko <ext-andriy.shevchenko@nokia.com>
Cc: Adrian Hunter <adrian.hunter@nokia.com>
Cc: <linux-mmc@vger.kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NChris Ball <cjb@laptop.org>

- Fix mmc_test_alloc_mem.

- Use nr_free_buffer_pages() instead of sysinfo.totalram to determine
  total lowmem pages.

- Change variables containing memory sizes to unsigned long.

- Limit maximum test area size to 128MiB because that is the maximum MMC
  high capacity erase size (the maxmium SD allocation unit size is just
  4MiB)
Signed-off-by: NAdrian Hunter <adrian.hunter@nokia.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

mmc_test provides tests aimed at testing SD/MMC hosts.  This patch adds
performance tests.

It is advantageous to have performance tests in a kernel
module like mmc_test for the following reasons:
	- transfer times can be measured very accurately
	- arbitrarily large transfers are possible
	- the effect of contiguous vs scattered pages
	can be determined

The new tests are:

	23. Best-case read performance
	24. Best-case write performance
	25. Best-case read performance into scattered pages
	26. Best-case write performance from scattered pages
	27. Single read performance by transfer size
	28. Single write performance by transfer size
	29. Single trim performance by transfer size
	30. Consecutive read performance by transfer size
	31. Consecutive write performance by transfer size
	32. Consecutive trim performance by transfer size
Signed-off-by: NAdrian Hunter <adrian.hunter@nokia.com>
Cc: <linux-mmc@vger.kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h

percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files.  percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.

percpu.h -> slab.h dependency is about to be removed.  Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability.  As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.

  http://userweb.kernel.org/~tj/misc/slabh-sweep.py

The script does the followings.

* Scan files for gfp and slab usages and update includes such that
  only the necessary includes are there.  ie. if only gfp is used,
  gfp.h, if slab is used, slab.h.

* When the script inserts a new include, it looks at the include
  blocks and try to put the new include such that its order conforms
  to its surrounding.  It's put in the include block which contains
  core kernel includes, in the same order that the rest are ordered -
  alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
  doesn't seem to be any matching order.

* If the script can't find a place to put a new include (mostly
  because the file doesn't have fitting include block), it prints out
  an error message indicating which .h file needs to be added to the
  file.

The conversion was done in the following steps.

1. The initial automatic conversion of all .c files updated slightly
   over 4000 files, deleting around 700 includes and adding ~480 gfp.h
   and ~3000 slab.h inclusions.  The script emitted errors for ~400
   files.

2. Each error was manually checked.  Some didn't need the inclusion,
   some needed manual addition while adding it to implementation .h or
   embedding .c file was more appropriate for others.  This step added
   inclusions to around 150 files.

3. The script was run again and the output was compared to the edits
   from #2 to make sure no file was left behind.

4. Several build tests were done and a couple of problems were fixed.
   e.g. lib/decompress_*.c used malloc/free() wrappers around slab
   APIs requiring slab.h to be added manually.

5. The script was run on all .h files but without automatically
   editing them as sprinkling gfp.h and slab.h inclusions around .h
   files could easily lead to inclusion dependency hell.  Most gfp.h
   inclusion directives were ignored as stuff from gfp.h was usually
   wildly available and often used in preprocessor macros.  Each
   slab.h inclusion directive was examined and added manually as
   necessary.

6. percpu.h was updated not to include slab.h.

7. Build test were done on the following configurations and failures
   were fixed.  CONFIG_GCOV_KERNEL was turned off for all tests (as my
   distributed build env didn't work with gcov compiles) and a few
   more options had to be turned off depending on archs to make things
   build (like ipr on powerpc/64 which failed due to missing writeq).

   * x86 and x86_64 UP and SMP allmodconfig and a custom test config.
   * powerpc and powerpc64 SMP allmodconfig
   * sparc and sparc64 SMP allmodconfig
   * ia64 SMP allmodconfig
   * s390 SMP allmodconfig
   * alpha SMP allmodconfig
   * um on x86_64 SMP allmodconfig

8. percpu.h modifications were reverted so that it could be applied as
   a separate patch and serve as bisection point.

Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: NTejun Heo <tj@kernel.org>
Guess-its-ok-by: NChristoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>

This patch fixes a bug in the multiblock write tests where the written
data is read back for verifying one block at a time.  The tests in
mmc_test assumes that all cards are byte addressable.

This will cause the multi block write tests to fail, leading the user of
the mmc_test driver thinking there is something wrong with the sdhci
driver they are testing.

The start address for the block is calculated as: blocknum * 512. For
block addressable cards the blocknum alone should be used.
Signed-off-by: NJohan Kristell <johan.kristell@axis.com>
Cc: <linux-mmc@vger.kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Due to a typo in the Basic Read test, it's currently identical to the
Basic Write test.  Fix this.
Signed-off-by: NRabin Vincent <rabin@rab.in>
Signed-off-by: NPierre Ossman <drzeus@drzeus.cx>

The mutex mmc_test_lock is initialized at every time mmc_test device
is probed. Probing another mmc_test device may break the mutex, if
the probe function is called while the mutex is locked.

This patch fixes it by statically initializing mmc_test_lock.
Signed-off-by: NAkinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: NPierre Ossman <drzeus@drzeus.cx>

This reverts commit 48b5352e. Oversized
sg lists are not allowed anymore, and the core even checks for them in
debug mode, so this test is entirely incorrect.
Signed-off-by: NPierre Ossman <drzeus@drzeus.cx>

Make it a bit more obvious that the card has been claimed by the
mmc_test driver so that people don't have to wonder why their block
device never shows up.
Signed-off-by: NPierre Ossman <drzeus@drzeus.cx>

Add tests that make sure the driver properly checks the blocks and
blksz fields and doesn't assume the sg list has a size that perfectly
matches the current request.
Signed-off-by: NPierre Ossman <drzeus@drzeus.cx>

Add a couple of tests to make sure the host driver handles highmem
memory pages properly. Unfortunately there is no way to guarantee an
allocation below 4 GB in i386, so it might give you addresses that
are out of reach for the hardware (OTOH, so will any other highmem
allocation in the kernel).
Signed-off-by: NPierre Ossman <drzeus@drzeus.cx>

Clean up and reorganise the mmc_test driver so that it (hopefully)
is easier to extend with more complex tests.
Signed-off-by: NPierre Ossman <drzeus@drzeus.cx>

We can only perform the tests on MMC and SD cards, so avoid binding
to any other type.
Signed-off-by: NPierre Ossman <drzeus@drzeus.cx>

Add the ability to run just a single test case by writing the test
case number into the sysfs "test" file.
Signed-off-by: NPierre Ossman <drzeus@drzeus.cx>

A dummy driver that performs a series of requests that are often mis-
handled by host drivers.
Signed-off-by: NPierre Ossman <drzeus@drzeus.cx>