Commit graph

97 commits

Author SHA1 Message Date
Lars-Peter Clausen
61cddc57dc regmap: Fix cache defaults initialization from raw cache defaults
Currently registers with a value of 0 are ignored when initializing the register
defaults from raw defaults. This worked in the past, because registers without a
explicit default were assumed to have a default value of 0. This was changed in
commit b03622a8 ("regmap: Ensure rbtree syncs registers set to zero properly").
As a result registers, which have a raw default value of 0 are now assumed to
have no default. This again can result in unnecessary writes when syncing the
cache. It will also result in unnecessary reads for e.g. the first update
operation. In the case where readback is not possible this will even let the
update operation fail, if the register has not been written to before.

So this patch removes the check. Instead it adds a check to ignore raw defaults
for registers which are volatile, since those registers are not cached.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Cc: stable@vger.kernel.org
2012-02-15 08:31:32 -08:00
Mark Brown
421e8d2de3 regmap: Reset cache status when reinitialsing the cache
When we reinitialise the cache make sure that we reset the cache access
flags, ensuring that the reinitialised cache is in the default state
which is what callers would and do expect given the function name.

This is particularly likely to cause issues in systems where there was no
cache previously as those systems have cache bypass enabled, as for the
wm8994 driver where this was noticed.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2012-01-20 13:49:30 +00:00
Mark Brown
681ba97d9f Merge branch 'regmap/irq' into regmap-next 2011-12-05 16:21:05 +00:00
Mark Brown
209a600623 regmap: Add irq_base accessor to regmap_irq
Allows devices to discover their own interrupt without having to remember
it themselves.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-12-05 16:13:31 +00:00
Mark Brown
8569d023a0 Merge branch 'topic/cache' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regmap into regmap-next 2011-12-05 13:18:50 +00:00
Mark Brown
bf31517335 regmap: Allow drivers to reinitialise the register cache at runtime
Sometimes the register map information may change in ways that drivers can
discover at runtime. For example, new revisions of a device may add new
registers. Support runtime discovery by drivers by allowing the register
cache to be reinitialised with a new function regmap_reinit_cache() which
discards the existing cache and creates a new one.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-12-05 13:17:36 +00:00
Mark Brown
bc7ee55633 regmap: Add trace event for successful cache reads
Currently we only trace physical reads, there's no instrumentation if
the read is satisfied from cache.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-30 20:51:09 +00:00
Mark Brown
018690d33e regmap: Allow regmap_update_bits() users to detect changes
Some users of regmap_update_bits() would like to be able to tell their
users if they actually did an update so provide a variant which also
returns a flag indicating if an update took place. We could return a
tristate in the return value of regmap_update_bits() but this makes the
API more cumbersome to use and doesn't fit with the general zero for
success idiom we have.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-29 20:15:50 +00:00
Mark Brown
c86845dc7b Merge remote-tracking branch 'regmap/topic/irq' into regmap-next 2011-11-28 18:52:24 +00:00
Mark Brown
d23511f959 regmap: Report if we actually handled an interrupt in regmap-irq
While the IRQ core doesn't currently support shared threaded interrupts
that's no reason for drivers not to do their bit and report IRQ_NONE when
they don't get an interrupt. This allows the core spurious/wedget interrupt
detection support to do its thing.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-28 18:51:57 +00:00
Mark Brown
cce585ce1e regmap: Fix rbtreee build when not using debugfs
The debugfs functions don't stub themselves out quite so well as might
be desirable so provide functions which do do this stubbing.

Reported-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-22 11:43:01 +00:00
Mark Brown
bad2ab4b6d regmap: Provide debugfs dump of the rbtree cache data
Show the register ranges we have in each rbtree node in debugfs, plus
some statistics on how big each node is and the total number of nodes.
It may also be worth collecting data on the ranges of dirty registers
to see if there's much mileage in trying to coalesce writes on sync.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-21 19:50:10 +00:00
Mark Brown
052d2cd123 regmap: Do debugfs init before cache init
This allows caches to add custom debugfs files.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-21 19:48:26 +00:00
Mark Brown
d91e8db2c3 regmap: Suppress noop writes in regmap_update_bits()
If the new register value is identical to the original one then suppress
the write to the hardware in regmap_update_bits(), saving some I/O cost.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-20 20:56:38 +00:00
Mark Brown
4c69166458 regmap: Remove indexed cache type
There should be no situation where it offers any advantage over rbtree
and there are no current users so remove the code for simplicity.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-20 20:56:34 +00:00
Lars-Peter Clausen
b44d48c1cc regmap: Drop check whether a register is readable in regcache_read
One of the reasons for using a cache is to have a software shadow of a register
which is writable but not readable. This allows us to do a read-modify-write
operation on such a register.

Currently regcache checks whether a register is readable when performing a
cached read and returns an error if it is not. Drop this check, since it will
prevent us from using the cache for registers where read-back is not possible.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-17 16:51:27 +00:00
Lars-Peter Clausen
064d4db11e regmap: Properly round cache_word_size
regcache currently only properly works with val bit sizes of 8 or 16, since
it will, when calculating the cache word size, round down. This causes the
cache storage to be too small to hold the full register value. Fix this by
rounding up instead.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-17 16:51:23 +00:00
Lars-Peter Clausen
7e5ec63ef5 regmap: Add support for 10/14 register formating
This patch adds support for 10 bits register, 14 bits value type register
formating. This is for example used by the Analog Devices AD5380.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-16 17:34:53 +00:00
Lars-Peter Clausen
19254411db regmap: Try cached read before checking if a hardware read is possible
For some register format types we do not provide a parse_val so we can not do a
hardware read. But a cached read is still possible, so try to read from the
cache first, before checking whether a hardware read is possible. Otherwise the
cache becomes pretty useless for these register types.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-16 17:34:53 +00:00
Lars-Peter Clausen
720e4616e8 regmap: Make reg_config reg_defaults const
The reg_defaults field usually points to a static per driver array, which should
not be modified. Make requirement this explicit by making reg_defaults const.
To allow this the regcache_init code needs some minor changes. Previoulsy the
reg_config was not available in regcache_init and regmap->reg_defaults was used
to pass the default register set to regcache_init. Now that the reg_config is
available we can work on it directly.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-16 17:34:53 +00:00
Lars-Peter Clausen
e5e3b8abed regmap: Move initialization of regcache related fields to regcache_init
Move the initialization regcache related fields of the regmap struct to
regcache_init. This allows us to keep regmap and regcache code better
separated.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-16 17:34:53 +00:00
Lars-Peter Clausen
c2b1ecd13c regmap: Do not call regcache_exit from regcache_lzo_init error path
Calling regcache_exit from regcache_lzo_init is first of all a layering
violation and secondly will cause double frees. regcache_exit will free buffers
allocated by the core, but the core will also free the same buffers when the
cacheops init callback returns an error. Thus we end up with a double free.
Fix this by not calling regcache_exit but only free those buffers which, have
been allocated in this function.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Acked-by: Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-15 19:22:59 +00:00
Lars-Peter Clausen
462a185c5c regmap: Do not call regcache_exit from regcache_rbtree_init error path
Calling regcache_exit from regcache_rbtree_init is first of all a layering
violation and secondly will cause double frees. regcache_exit will free buffers
allocated by the core, but the core will also free the same buffers when the
cacheops init callback returns an error. Thus we end up with a double free.
Fix this by not calling regcache_exit but only free those buffers which, have
been allocated in this function.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Acked-by: Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-15 19:22:59 +00:00
Lars-Peter Clausen
bd061c78ca regmap: Fix memory leak in regcache_init error path
Make sure all allocated memory gets freed again in case initializing the cache
failed.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Acked-by: Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-14 21:46:25 +00:00
Lars-Peter Clausen
021cd616de regmap: Fix memory leak in regcache_hw_init error path
Make sure reg_defaults_raw gets freed in case of an error.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Acked-by: Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-14 21:46:09 +00:00
Lars-Peter Clausen
abbb18fb4a regmap: return ERR_PTR instead of NULL in regmap_init
The regmap_init documentation states that it will either return a pointer to a
valid regmap structure or a ERR_PTR in case of an error. Currently it returns a
NULL pointer in case no bus or no config was given. Since NULL is not a
ERR_PTR a caller might assume that it is a pointer to a valid regmap structure,
so return a ERR_PTR(-EINVAL) instead.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-14 21:44:45 +00:00
Lars-Peter Clausen
58072cbfc5 regmap: Fix memory leak in regmap_init error path
If regcache initialization fails regmap_init will currently exit without
freeing work_buf.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-10 17:41:20 +00:00
Lars-Peter Clausen
c48a9d7492 regmap: Support some more block operations on cached devices
Commit 10a08d9f ("regmap: Support some block operations on cached devices")
allowed raw read operations without throwing a warning when using caches if
all registers are volatile. This patch does the same for raw write operations.

This is for example useful when loading a firmware in a predefined volatile
region on a chip where we otherwise want registers to be cached.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-09 00:56:52 +00:00
Lars-Peter Clausen
82cd9965c3 regmap: Add helper function for checking if a register range is volatile
We already have the same code for checking whether a register range is volatile
in two different places. Instead of duplicating it once more  add a small helper
function for checking whether a register range is voltaile.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-09 00:56:51 +00:00
Mark Brown
81bf58eb3c Merge branches 'regmap/irq' and 'regmap/cache' into regmap-next 2011-11-08 14:16:06 +00:00
Mark Brown
50b776fc71 regmap: Rename LZO cache type to compressed
Users probably don't care about the specific compression algorithm and
we might want to use a different algorithm (snappy being the one I'm
thinking of right now) so update the public interface to have a more
generic name.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-08 14:15:00 +00:00
Mark Brown
8ae0d7e8a9 regmap: Track if the register cache is dirty and suppress unneeded syncs
Allow drivers to optimise out the register cache sync if they didn't need
to do one. If the hardware is desynced from the register cache (by power
loss for example) then the driver should call regcache_mark_dirty() to
let the core know about this.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-08 11:38:15 +00:00
Mark Brown
b973aa3624 regmap: Fix typo in kerneldoc for regmap_update_bits()
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-08 11:32:55 +00:00
Mark Brown
7ea7580183 regmap: Fix word wrap in Makefile
80 columns FTW.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-08 11:32:31 +00:00
Mark Brown
82732bdd66 regmap: Prepare LZO cache for variable block sizes
Give regcache_lzo_block_count() a copy of the map so that when we decide
we want to make the LZO cache more controllable we can more easily do so.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
2011-11-08 11:32:05 +00:00
Mark Brown
f8beab2bb6 regmap: Add a reusable irq_chip for regmap based interrupt controllers
There seem to be lots of regmap-using devices with very similar interrupt
controllers with a small bank of interrupt registers and mask registers
with an interrupt per bit. This won't cover everything but it's a good
start.

Each chip supplies a base for the status registers, a base for the mask
registers, an optional base for writing acknowledgements (which may be the
same as the status registers) and an array of bits within each of these
register banks which indicate the interrupt.

There is an assumption that the bit for each interrupt will be the same
in each of the register bank.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-08 11:29:48 +00:00
Paul Gortmaker
1b6bc32f0a drivers/base: Add export.h for EXPORT_SYMBOL/THIS_MODULE as required.
Most of these files were implicitly getting EXPORT_SYMBOL via
device.h which was including module.h, but that path will be broken
soon.

[ with input from Stephen Rothwell <sfr@canb.auug.org.au> ]

Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2011-10-31 19:31:38 -04:00
Linus Torvalds
4e7e2a2008 Merge branch 'for-linus' of git://opensource.wolfsonmicro.com/regmap
* 'for-linus' of git://opensource.wolfsonmicro.com/regmap: (62 commits)
  mfd: Enable rbtree cache for wm831x devices
  regmap: Support some block operations on cached devices
  regmap: Allow caches for devices with no defaults
  regmap: Ensure rbtree syncs registers set to zero properly
  regmap: Allow rbtree to cache zero default values
  regmap: Warn on raw I/O as well as bulk reads that bypass cache
  regmap: Return a sensible error code if we fail to read the cache
  regmap: Use bsearch() to search the register defaults
  regmap: Fix doc comment
  regmap: Optimize the lookup path to use binary search
  regmap: Ensure we scream if we enable cache bypass/only at the same time
  regmap: Implement regcache_cache_bypass helper function
  regmap: Save/restore the bypass state upon syncing
  regmap: Lock the sync path, ensure we use the lockless _regmap_write()
  regmap: Fix apostrophe usage
  regmap: Make _regmap_write() global
  regmap: Fix lock used for regcache_cache_only()
  regmap: Grab the lock in regcache_cache_only()
  regmap: Modify map->cache_bypass directly
  regmap: Fix regcache_sync generic implementation
  ...
2011-10-25 13:57:45 +02:00
Mark Brown
de2d808f4d regmap: Support some block operations on cached devices
Support raw reads if all the registers being read are volatile, the cache
will have no impact for tem.

Support bulk reads either directly (if all the registers are volatile) or
by falling back to iterating over single register reads otherwise.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-10-13 11:41:54 +01:00
Mark Brown
8528bdd450 regmap: Allow caches for devices with no defaults
We only really need the defaults in order to cut down the number of
registers we sync and to satisfy reads while the device is powered off
but not all devices are going to need to do that (always on devices like
PMICs being the prime example) so don't require those devices to supply
a default.  Instead only try to fall back to hardware defaults if the
driver told us to.

Devices using LZO won't be able to instantiate with this, that will require
some updates in the LZO code to handle this case.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
2011-10-10 11:55:32 +01:00
Mark Brown
b03622a80d regmap: Ensure rbtree syncs registers set to zero properly
Simplify the check for registers set at their default value by avoiding
picking a default value in the case where we don't have one. Instead we
only compare the current value to the current value when we looked one
up. This fixes the case where we don't have a default stored but the value
was set to zero when that isn't the chip default.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
2011-10-10 10:26:48 +01:00
Mark Brown
e42c5a9a42 regmap: Allow rbtree to cache zero default values
Ensure that when we start up in cache only mode we can store defaults of
zero, otherwise if the hardware is unavailable we won't be able to read.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
2011-10-10 10:25:00 +01:00
Mark Brown
04e016adca regmap: Warn on raw I/O as well as bulk reads that bypass cache
As with the bulk reads we really should be able to make these play
nicely with the cache but warn for now.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
2011-10-10 10:24:33 +01:00
Mark Brown
6e6ace00a0 regmap: Return a sensible error code if we fail to read the cache
If a register isn't cached then let callers know that so they can fall
back or error handle appropriately.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
2011-10-10 10:24:03 +01:00
Mark Brown
f094fea68f regmap: Use bsearch() to search the register defaults
Rather than open coding a binary search use the standard bsearch() using
the comparison function we're already using for sort() on insert. This
fixes a lockup I was observing due to iterating on min <= max rather
than min < max when we fail to look up.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
2011-10-09 12:36:56 +01:00
Dimitris Papastamos
0eef6b0415 regmap: Fix doc comment
Signed-off-by: Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-10-03 11:50:53 +01:00
Dimitris Papastamos
c08604b8ae regmap: Optimize the lookup path to use binary search
Since there are more lookups than insertions in a typical
scenario, optimize the linear search into a binary search.  For
this to work, we need to keep reg_defaults sorted upon
insertions, for now be lazy and use sort().

Signed-off-by: Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-10-03 11:50:53 +01:00
Dimitris Papastamos
ac77a765cb regmap: Ensure we scream if we enable cache bypass/only at the same time
Signed-off-by: Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-09-30 13:57:47 +01:00
Dimitris Papastamos
6eb0f5e015 regmap: Implement regcache_cache_bypass helper function
Ensure we've got a function so users can enable/disable the
cache bypass option.

Signed-off-by: Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-09-30 13:57:47 +01:00
Dimitris Papastamos
beb1a10f21 regmap: Save/restore the bypass state upon syncing
Signed-off-by: Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-09-30 13:57:47 +01:00