Regmap #4
Merged
Regmap #4
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The macro "REGMAP_ALLOW_WRITE_DEBUGFS" can be used to enable write support on the registers file in the debugfs. The mode of the file is fixed to 0400 so it is not possible to write the file ever. This patch fixes the mode by setting it to the correct value depending on the macro. Signed-off-by: Markus Pargmann <[email protected]> Signed-off-by: Mark Brown <[email protected]> [[email protected]: resolve trivial merge conflicts] Signed-off-by: Phani Kumar Uppalapati <[email protected]> Signed-off-by: Pranav Vashi <[email protected]>
Add support for devices with 24 data bits. Signed-off-by: Marc Reilly <[email protected]> Signed-off-by: Mark Brown <[email protected]>
This change combines any padding bits into the register address bits when determining register format handlers to use the next byte-divisible register size. A reg_shift member is introduced to the regmap struct to enable fixup of the reg format. Format handlers now take an extra parameter specifying the number of bits to shift the value by. Signed-off-by: Marc Reilly <[email protected]> Signed-off-by: Mark Brown <[email protected]>
The only context needed by I2C and SPI bus definitions is the device itself; this can be converted to an i2c_client or spi_device in order to perform IO on the device. However, other bus types may need more context in order to perform IO. Enable this by having regmap_init accept a bus_context parameter, and pass this to all bus callbacks. The existing callbacks simply pass the struct device here. Future bus types may pass something else. Signed-off-by: Stephen Warren <[email protected]> Signed-off-by: Mark Brown <[email protected]>
Some bus types have very fast IO. For these, acquiring a mutex for every IO operation is a significant overhead. Allow busses to indicate their IO is fast, and enhance regmap to use a spinlock for those busses. [Currently limited to native endian registers -- broonie] Signed-off-by: Stephen Warren <[email protected]> Signed-off-by: Mark Brown <[email protected]>
This is a basic memory-mapped-IO bus for regmap. It has the following features and limitations: * Registers themselves may be 8, 16, 32, or 64-bit. 64-bit is only supported on 64-bit platforms. * Register offsets are limited to precisely 32-bit. * IO is performed using readl/writel, with no provision for using the __raw_readl or readl_relaxed variants. Signed-off-by: Stephen Warren <[email protected]> Signed-off-by: Mark Brown <[email protected]>
Some of the error conditions detected by regmap_mmio_*() are pure internal errors, rather than user-/client-triggerable conditions. Convert these to BUG(). Signed-off-by: Stephen Warren <[email protected]> Signed-off-by: Mark Brown <[email protected]>
These error checks are implemented in regmap core. Remove the duplicate code from regmap-mmio.c Signed-off-by: Stephen Warren <[email protected]> Signed-off-by: Mark Brown <[email protected]>
val_len should be a multiple of val_bytes. If it's not, error out early. Signed-off-by: Stephen Warren <[email protected]> Signed-off-by: Mark Brown <[email protected]>
Some devices have multiple separate register regions. Logically, one regmap would be created per region. One issue that prevents this is that each instance will attempt to create the same debugfs files. Avoid this by allowing regmaps to be named, and use the name to construct the debugfs directory name. Signed-off-by: Stephen Warren <[email protected]> Signed-off-by: Mark Brown <[email protected]>
Commit 79c64d5 "regmap: allow regmap instances to be named" changed the prototype of regmap_debugfs_init, but didn't update the dummy inline used when !CONFIG_DEBUGFS. Fix this. Signed-off-by: Stephen Warren <[email protected]> Signed-off-by: Mark Brown <[email protected]>
regmap_config.reg_stride is introduced. All extant register addresses are a multiple of this value. Users of serial-oriented regmap busses will typically set this to 1. Users of the MMIO regmap bus will typically set this based on the value size of their registers, in bytes, so 4 for a 32-bit register. Throughout the regmap code, actual register addresses are used. Wherever the register address is used to index some array of values, the address is divided by the stride to determine the index, or vice-versa. Error- checking is added to all entry-points for register address data to ensure that register addresses actually satisfy the specified stride. The MMIO bus ensures that the specified stride is large enough for the register size. Signed-off-by: Stephen Warren <[email protected]> Signed-off-by: Mark Brown <[email protected]>
Commit f01ee60fffa4 ("regmap: implement register striding") caused the
compile errors below. Fix them.
drivers/base/regmap/regmap-irq.c: In function 'regmap_irq_sync_unlock':
drivers/base/regmap/regmap-irq.c:62:12: error: 'map' undeclared (first use in this function)
drivers/base/regmap/regmap-irq.c:62:12: note: each undeclared identifier is reported only once for each function it appears in
drivers/base/regmap/regmap-irq.c: In function 'regmap_irq_enable':
drivers/base/regmap/regmap-irq.c:77:37: error: 'map' undeclared (first use in this function)
drivers/base/regmap/regmap-irq.c: In function 'regmap_irq_disable':
drivers/base/regmap/regmap-irq.c:85:37: error: 'map' undeclared (first use in this function)
Signed-off-by: Stephen Warren <[email protected]>
Signed-off-by: Mark Brown <[email protected]>
If we don't have a cached value for a register and we can cache it then when we do a read a value we should add it to the cache to save rereading it later on. Do this for single register reads, for block reads the code would be a little more complex and this covers most practical usage. Signed-off-by: Mark Brown <[email protected]>
Some devices does not support bulk read and write operations, for them we have series of single write and read operations. Signed-off-by: Anthony Olech <[email protected]> Signed-off-by: Ashish Jangam <[email protected]> [Fixed coding style, don't check use_single_rw before assign --broonie ] Signed-off-by: Mark Brown <[email protected]>
Set the use_single_rw flag for devices that use format_write() since format_write() doesn't support any form of block operation. Signed-off-by: Mark Brown <[email protected]>
Use devres to implement dev_get_regmap(). This should mean that in almost all cases devices wishing to take advantage of framework features based on regmap shouldn't need to explicitly pass the regmap into the framework. This simplifies device setup a bit. Signed-off-by: Mark Brown <[email protected]>
It's needed for freeing and for obtaining the IRQ base later on. Signed-off-by: Mark Brown <[email protected]>
This gets us up to date with the recommended current kernel infrastructure and should transparently give us device tree interrupt bindings for any devices using the framework. If an explicit IRQ mapping is passed in then a legacy interrupt range is created, otherwise a simple linear mapping is used. Previously a mapping was mandatory so existing drivers should not be affected. A function regmap_irq_get_virq() is provided to allow drivers to map individual IRQs which should be used in preference to the existing regmap_irq_chip_get_base() which is only valid if a legacy IRQ range is provided. Signed-off-by: Mark Brown <[email protected]>
In some chips the IRQ status registers are not contiguous in the register map but spaced at even spaces. This is an easy case to handle with minor changes. It is assume for this purpose that the stride for status is equal to the stride for mask/ack registers as well. Signed-off-by: Graeme Gregory <[email protected]> Signed-off-by: Mark Brown <[email protected]>
Signed-off-by: Mark Brown <[email protected]>
Ensure that we can't get randconfig breakage by doing the IRQ_DOMAIN select automatically. Don't just do the select from REGMAP_IRQ to ensure that the select actually gets noticed. Reported-by: Randy Dunlap <[email protected]> Signed-off-by: Mark Brown <[email protected]>
Since there are uses for I2C_M_NOSTART which are much more sensible and standard than most of the protocol mangling functionality (the main one being gather writes to devices where something like a register address needs to be inserted before a block of data) create a new I2C_FUNC_NOSTART for this feature and update all the users to use it. Also strengthen the disrecommendation of the protocol mangling while we're at it. In the case of regmap-i2c we remove the requirement for mangling as I2C_M_NOSTART is the only mangling feature which is being used. Signed-off-by: Mark Brown <[email protected]> Acked-by: Wolfram Sang <[email protected]> Signed-off-by: Jean Delvare <[email protected]>
Add a field to struct regmap_bus that allows bus drivers to request that register addresses and values be formatted with a specific endianness. The default endianness is unchanged from current operation: Big. Implement native endian formatting/parsing for 16- and 32-bit values. This will be enough to support regmap-mmio.c. Signed-off-by: Stephen Warren <[email protected]> Signed-off-by: Mark Brown <[email protected]>
This will avoid the regmap core converting all addresses and values into big endian, only for the mmio bus driver to have to convert them back to native endian. Signed-off-by: Stephen Warren <[email protected]> Signed-off-by: Mark Brown <[email protected]>
If debugfs isn't cleaned up, stale files will be left in the filesystem which will cause an OOPS when accessed the first time, and hang the accessing application when accessed again, presumably due to some lock being left held. Signed-off-by: Stephen Warren <[email protected]> Signed-off-by: Mark Brown <[email protected]>
The word to be transmitted/received via regmap is composed by the following parts: config->reg_bits config->val_bits config->pad_bits ,so the total size should be calculated by summing up the number of bits of each element and using a DIV_ROUND_UP to return the number of bytes. Signed-off-by: Fabio Estevam <[email protected]> Signed-off-by: Mark Brown <[email protected]>
We should never be modifying it and it lets drivers declare it const. Signed-off-by: Mark Brown <[email protected]>
If the driver supplied an empty entry in the array of IRQs then return an error rather than trying to do the mapping. This is intended for use with handling chip variants and similar situations. Signed-off-by: Mark Brown <[email protected]>
Allow chips to provide a bank of registers for controlling the wake state in a similar fashion to the masks and propagate the wake count to the parent interrupt controller. Signed-off-by: Mark Brown <[email protected]>
…entry Provide a feel of how much overhead the rbtree cache adds to the game. [Slightly reworded output in debugfs -- broonie] Signed-off-by: Dimitris Papastamos <[email protected]> Signed-off-by: Mark Brown <[email protected]>
Display the name for the chip rather than just the primary IRQ so it is clearer what exactly has failed. Signed-off-by: Mark Brown <[email protected]>
_regmap_raw_write() contains code to call regcache_write() to write values to the cache. That code calls memcpy() to copy the value data to the start of the work_buf. However, at least when _regmap_raw_write() is called from _regmap_bus_raw_write(), the value data is in the work_buf, and this memcpy() operation may over-write part of that value data, depending on the value of reg_bytes + pad_bytes. At least when using reg_bytes==1 and pad_bytes==0, corruption of the value data does occur. To solve this, remove the memcpy() operation, and modify the subsequent .parse_val() call to parse the original value buffer directly. At least in the case of 8-bit register address and 16-bit values, and writes of single registers at a time, this memcpy-then-parse combination used to cancel each-other out; for a work-buffer containing xx 89 03, the memcpy changed it to 89 03 03, and the parse_val changed it back to 89 89 03, thus leaving the value uncorrupted. This appears completely accidental though. Since commit 8a819ff "regmap: core: Split out in place value parsing", .parse_val only returns the parsed value, and does not modify the buffer, and hence does not (accidentally) undo the corruption caused by memcpy(). This caused bogus values to get written to HW, thus preventing e.g. audio playback on systems with a WM8903 CODEC. This patch fixes that. Signed-off-by: Stephen Warren <[email protected]> Signed-off-by: Mark Brown <[email protected]>
Don't grind to a screaming halt, just generate a warning. Signed-off-by: Mark Brown <[email protected]>
This allows the cache to sync values directly to the device when stored in native format and also allows asynchronous I/O. Signed-off-by: Mark Brown <[email protected]>
This patch aims to bring down the average number of nodes in the rbtree cache and increase the average number of registers per node. This should improve general lookup and traversal times. This is achieved by setting the minimum size of a block within the rbnode to the size of the rbnode itself. This will essentially cache possibly non-existent registers so to combat this scenario, we keep a separate bitmap in memory which keeps track of which register exists. The memory overhead of this change is likely in the order of ~5-10%, possibly less depending on the register file layout. On my test system with a bitmap of ~4300 bits and a relatively sparse register layout, the memory requirements for the entire cache did not increase (the cutting down of nodes which was about 50% of the original number compensated the situation). A second patch that can be built on top of this can look at the ratio `sizeof(*rbnode) / map->cache_word_size' in order to suitably adjust the block length of each block. Signed-off-by: Dimitris Papastamos <[email protected]> Signed-off-by: Mark Brown <[email protected]>
Provide a helper to do the size based index into a block of registers and use it when reading a value. Signed-off-by: Mark Brown <[email protected]>
Mainly useful internally but exported since this is a public API that's being checked for. Signed-off-by: Mark Brown <[email protected]>
This will bring no meaningful benefit by itself, it is done as a separate commit to aid bisection if there are problems with the following commits adding support for coalescing adjacent writes. Signed-off-by: Mark Brown <[email protected]>
Let's only write once... Signed-off-by: Mark Brown <[email protected]>
The idea of maintaining a bitmap of present registers is something that can usefully be used by other cache types that maintain blocks of cached registers so move the code out of the rbtree cache and into the generic regcache code. Refactor the interface slightly as we go to wrap the set bit and enlarge bitmap operations (since we never do one without the other) and make it more robust for reads of uncached registers by bounds checking before we look at the bitmap. Signed-off-by: Mark Brown <[email protected]> Reviewed-by: Dimitris Papastamos <[email protected]>
The idea of holding blocks of registers in device format is shared between at least rbtree and lzo cache formats so split out the loop that does the sync from the rbtree code so optimisations on it can be reused. Signed-off-by: Mark Brown <[email protected]> Reviewed-by: Dimitris Papastamos <[email protected]>
For code clarity after implementing block writes split out the raw and non-raw I/O sync implementations. Signed-off-by: Mark Brown <[email protected]> Reviewed-by: Dimitris Papastamos <[email protected]>
When syncing blocks of data using raw writes combine the writes into a single block write, saving us bus overhead for setup, addressing and teardown. Currently the block write is done unconditionally as it is expected that hardware which has a register format which can support raw writes will support auto incrementing writes, this decision may need to be revised in future. Signed-off-by: Mark Brown <[email protected]> Reviewed-by: Dimitris Papastamos <[email protected]>
regcache_sync_block_raw is used only in this file. Hence make it static. Silences the following warning: drivers/base/regmap/regcache.c:608:5: warning: symbol 'regcache_sync_block_raw' was not declared. Should it be static? Signed-off-by: Sachin Kamat <[email protected]> Signed-off-by: Mark Brown <[email protected]>
Fix format specifier in dev_dbg and suppress the following warning drivers/base/regmap/regcache.c: In function ‘regcache_sync_block_raw_flush’: drivers/base/regmap/regcache.c:593:2: warning: format ‘%d’ expects argument of type ‘int’, but argument 4 has type ‘size_t’ [-Wformat] Signed-off-by: Stratos Karafotis <[email protected]> Signed-off-by: Mark Brown <[email protected]>
This reverts commit bc8ce4 (regmap: don't corrupt work buffer in _regmap_raw_write()) since it turns out that it can cause issues when taken in isolation from the other changes in -next that lead to its discovery. On the basis that nobody noticed the problems for quite some time without that subsequent work let's drop it from v3.9. Signed-off-by: Mark Brown <[email protected]>
_regmap_raw_write() contains code to call regcache_write() to write values to the cache. That code calls memcpy() to copy the value data to the start of the work_buf. However, at least when _regmap_raw_write() is called from _regmap_bus_raw_write(), the value data is in the work_buf, and this memcpy() operation may over-write part of that value data, depending on the value of reg_bytes + pad_bytes. At least when using reg_bytes==1 and pad_bytes==0, corruption of the value data does occur. To solve this, remove the memcpy() operation, and modify the subsequent .parse_val() call to parse the original value buffer directly. At least in the case of 8-bit register address and 16-bit values, and writes of single registers at a time, this memcpy-then-parse combination used to cancel each-other out; for a work-buffer containing xx 89 03, the memcpy changed it to 89 03 03, and the parse_val changed it back to 89 89 03, thus leaving the value uncorrupted. This appears completely accidental though. Since commit 8a819ff "regmap: core: Split out in place value parsing", .parse_val only returns the parsed value, and does not modify the buffer, and hence does not (accidentally) undo the corruption caused by memcpy(). This caused bogus values to get written to HW, thus preventing e.g. audio playback on systems with a WM8903 CODEC. This patch fixes that. Signed-off-by: Stephen Warren <[email protected]> Signed-off-by: Mark Brown <[email protected]>
Signed-off-by: Dimitris Papastamos <[email protected]> Signed-off-by: Mark Brown <[email protected]>
The parameter passed to the regmap lock/unlock callbacks needs to be
map->lock_arg, regcache passes just map. This works fine in the case that no
custom locking callbacks are used since in this case map->lock_arg equals map,
but will break when custom locking callbacks are used. The issue was introduced
in commit 0d4529c5("regmap: make lock/unlock functions customizable") and is
fixed by this patch.
Signed-off-by: Lars-Peter Clausen <[email protected]>
Signed-off-by: Mark Brown <[email protected]>
A node starting before the minimum register is no reason to reject it, since its end could be in range. The check for the end already exists two lines lower, so we can just remove the incorrect check. Signed-off-by: Maarten ter Huurne <[email protected]> Signed-off-by: Mark Brown <[email protected]>
commit f2e055e7c9c6084bfbaa68701e52562acf96419e upstream.
Commit f8bd822cb ("regmap: cache: Factor out block sync") made
regcache_rbtree_sync() call regmap_async_complete(), which in turn does
not check for map->bus before dereferencing it.
This causes a NULL pointer dereference on bus-less maps.
Signed-off-by: Daniel Mack <[email protected]>
Signed-off-by: Mark Brown <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
commit 2d49b5987561e480bdbd8692b27fc5f49a1e2f0b upstream.
regcache_sync_block_raw_flush() expects the address of the register after last
register that needs to be synced as its parameter. But the last call to
regcache_sync_block_raw_flush() in regcache_sync_block_raw() passes the address
of the last register in the block. This effectively always skips over the last
register in a block, even if it needs to be synced. In order to fix it increase
the address by one register.
The issue was introduced in commit 75a5f89 ("regmap: cache: Write consecutive
registers in a single block write").
Signed-off-by: Lars-Peter Clausen <[email protected]>
Signed-off-by: Mark Brown <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
commit 4e67fb5f5e336250db944921e3c68057d6203034 upstream. Avoid overlapping register regions by making the initial blklen of a new node 1. If a register write occurs to a yet uncached register, that is lower than but near an existing node's base_reg, a new node is created and it's blklen is set to an arbitrary value (sizeof(*rbnode)). That may cause this node to overlap with another node. Those nodes should be merged, but this merge doesn't happen yet, so this patch at least makes the initial blklen small enough to avoid hitting the wrong node, which may otherwise lead to severe breakage. Signed-off-by: David Jander <[email protected]> Signed-off-by: Mark Brown <[email protected]> Signed-off-by: Zhouping Liu <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]>
commit 2c98e0c1cc6b8e86f1978286c3d4e0769ee9d733 upstream. If 'map->dev' is NULL and there will lead dev_name() to be NULL pointer dereference. So before dev_name(), we need to have check of the map->dev pionter. We also should make sure that the 'name' pointer shouldn't be NULL for debugfs_create_dir(). So here using one default "dummy" debugfs name when the 'name' pointer and 'map->dev' are both NULL. Signed-off-by: Xiubo Li <[email protected]> Signed-off-by: Mark Brown <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]> Signed-off-by: Pranav Vashi <[email protected]>
commit 5336be8416a71b5568d2cf54a2f2066abe9f2a53 upstream. If LOG_DEVICE is defined and map->dev is NULL it will lead to NULL pointer dereference. This patch fixes this issue by adding check for dev->NULL in all such places in regmap.c Signed-off-by: Pankaj Dubey <[email protected]> Signed-off-by: Mark Brown <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]> Signed-off-by: Pranav Vashi <[email protected]>
commit 15b8d2c41fe5839582029f65c5f7004db451cc2b upstream. In big endian mode regmap_bulk_read gives incorrect data for byte reads. This is because memcpy of a single byte from an address after full word read gives different results when endianness differs. ie. we get little-end in LE and big-end in BE. Signed-off-by: Arun Chandran <[email protected]> Signed-off-by: Mark Brown <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]> Signed-off-by: Pranav Vashi <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.