aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAge
* pinctrl: API changes to support multiple states per deviceStephen Warren2012-03-05
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The API model is changed from: p = pinctrl_get(dev, "state1"); pinctrl_enable(p); ... pinctrl_disable(p); pinctrl_put(p); p = pinctrl_get(dev, "state2"); pinctrl_enable(p); ... pinctrl_disable(p); pinctrl_put(p); to this: p = pinctrl_get(dev); s1 = pinctrl_lookup_state(p, "state1"); s2 = pinctrl_lookup_state(p, "state2"); pinctrl_select_state(p, s1); ... pinctrl_select_state(p, s2); ... pinctrl_put(p); This allows devices to directly transition between states without disabling the pin controller programming and put()/get()ing the configuration data each time. This model will also better suit pinconf programming, which doesn't have a concept of "disable". The special-case hogging feature of pin controllers is re-written to use the regular APIs instead of special-case code. Hence, the pinmux-hogs debugfs file is removed; see the top-level pinctrl-handles files for equivalent data. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Dong Aisheng <dong.aisheng@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: add usecount to pins for muxingStephen Warren2012-03-05
| | | | | | | | | | | | Multiple mapping table entries could reference the same pin, and hence "own" it. This would be unusual now that pinctrl_get() represents a single state for a client device, but in the future when it represents all known states for a device, this is quite likely. Implement reference counting for pin ownership to handle this. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Dong Aisheng <dong.aisheng@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: refactor struct pinctrl handling in core.c vs pinmux.cStephen Warren2012-03-05
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This change separates two aspects of struct pinctrl: a) The data representation of the parsed mapping table, into: 1) The top-level struct pinctrl object, a single entity returned by pinctrl_get(). 2) The parsed version of each mapping table entry, struct pinctrl_setting, of which there is one per mapping table entry. b) The code that handles this; the code for (1) above is in core.c, and the code to parse/execute each entry in (2) above is in pinmux.c, while the iteration over multiple settings is lifted to core.c. This will allow the following future changes: 1) pinctrl_get() API rework, so that struct pinctrl represents all states for the device, and the device can select between them without calling put()/get() again. 2) To support that, a struct pinctrl_state object will be inserted into the data model between the struct pinctrl and struct pinctrl_setting. 3) The mapping table will be extended to allow specification of pin config settings too. To support this, struct pinctrl_setting will be enhanced to store either mux settings or config settings, and functions will be added to pinconf.c to parse/execute pin configuration settings. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Linus Walleij <linus.walleij@linaro.org> Acked-by: Dong Aisheng <dong.aisheng@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: fix and simplify lockingStephen Warren2012-03-05
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | There are many problems with the current pinctrl locking: struct pinctrl_dev's gpio_ranges_lock isn't effective; pinctrl_match_gpio_range() only holds this lock while searching for a gpio range, but the found range is return and manipulated after releading the lock. This could allow pinctrl_remove_gpio_range() for that range while it is in use, and the caller may very well delete the range after removing it, causing pinctrl code to touch the now-free range object. Solving this requires the introduction of a higher-level lock, at least a lock per pin controller, which both gpio range registration and pinctrl_get()/put() will acquire. There is missing locking on HW programming; pin controllers may pack the configuration for different pins/groups/config options/... into one register, and hence have to read-modify-write the register. This needs to be protected, but currently isn't. Related, a future change will add a "complete" op to the pin controller drivers, the idea being that each state's programming will be programmed into the pinctrl driver followed by the "complete" call, which may e.g. flush a register cache to HW. For this to work, it must not be possible to interleave the pinctrl driver calls for different devices. As above, solving this requires the introduction of a higher-level lock, at least a lock per pin controller, which will be held for the duration of any pinctrl_enable()/disable() call. However, each pinctrl mapping table entry may affect a different pin controller if necessary. Hence, with a per-pin-controller lock, almost any pinctrl API may need to acquire multiple locks, one per controller. To avoid deadlock, these would need to be acquired in the same order in all cases. This is extremely difficult to implement in the case of pinctrl_get(), which doesn't know which pin controllers to lock until it has parsed the entire mapping table, since it contains somewhat arbitrary data. The simplest solution here is to introduce a single lock that covers all pin controllers at once. This will be acquired by all pinctrl APIs. This then makes struct pinctrl's mutex irrelevant, since that single lock will always be held whenever this mutex is currently held. Signed-off-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: fix the pin descriptor kerneldocLinus Walleij2012-03-02
| | | | | | | The introduction of the owner field on the pin descriptor was not properly documented so fix this up. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: assume map table entries can't have a NULL name fieldStephen Warren2012-03-02
| | | | | | | | | | | | | | | | | | pinctrl_register_mappings() already requires that every mapping table entry have a non-NULL name field. Logically, this makes sense too; drivers should always request a specific named state so they know what they're getting. Relying on getting the first mentioned state in the mapping table is error-prone, and a nasty special case to implement, given that a given the mapping table may define multiple states for a device. Remove a small part of the documentation that talked about optionally requesting a specific state; it's mandatory now. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Dong Aisheng <dong.aisheng@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: introduce PINCTRL_STATE_DEFAULT, define hogs as that stateStephen Warren2012-03-02
| | | | | | | | | | | | | | | | | | | | This provides a single centralized name for the default state. Update PIN_MAP_* macros to use this state name, instead of requiring the user to pass a state name in. With this change, hog entries in the mapping table are defined as those with state name PINCTRL_STATE_DEFAULT, i.e. all entries have the same name. This interacts badly with the nested iteration over mapping table entries in pinctrl_hog_maps() and pinctrl_hog_map() which would now attempt to claim each hog mapping table entry multiple times. Replacing the custom hog code with a simple pinctrl_get()/pinctrl_enable(). Update documentation and mapping tables to use this. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Dong Aisheng <dong.aisheng@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: enhance pinctrl_get() to handle multiple functionsStephen Warren2012-03-02
| | | | | | | | | | | | | At present, pinctrl_get() assumes that all matching mapping table entries have the same "function" value, albeit potentially applied to different pins/groups. This change removes this restriction; pinctrl_get() can now handle a set of mapping tables where different functions are applied to the various pins/groups. Signed-off-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: move pinctrl-maps debugfs file to top-levelStephen Warren2012-02-29
| | | | | | | | | | | | | | The debugfs file pinctrl-maps is a system-wide file, not specific to any pin controller, so place it in the top-level directory. Also, move the code implementing the file to keep the order of all the functions matching the order they're created in pinctrl_init_*debugfs(). The only non-obvious change here is no private data is passed to debugfs_create_file() or single_open(). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Dong Aisheng <dong.aisheng@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: re-order struct pinctrl_mapStephen Warren2012-02-29
| | | | | | | | | | | | | | | | | | The lookup key in struct pinctrl_map is (.dev_name, .name). Re-order the struct definition to put the lookup key fields first, and the result values afterwards. To me at least, this slightly better reflects the lookup process. Update the documentation in a similar fashion. Note: PIN_MAP*() macros aren't updated; I plan to update this once later when enhancing the mapping table format to support pin config to reduce churn. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Dong Aisheng <dong.aisheng@linaro.org> [Rebased for cherry-picking] Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: make the pinmux-pins more helpfulLinus Walleij2012-02-29
| | | | | | | | | | | | | | | | The debugfs file pinmux-pins used to tell which function was enabled but now states simply which device owns the pin. Being owned by the pinctrl driver itself means just that it's hogged so be a bit more helpful by printing that. ChangeLog v1->v2: - Preserve the self-referential owner field, just clarify that when the pin controller states itself as owner this means that it's hogged. Acked-by: Dong Aisheng <dong.aisheng@linaro.org> Acked-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: remove pin and hogs locks from struct pinctrl_devStephen Warren2012-02-24
| | | | | | | | | | | | | | struct pinctrl_dev's pin_desc_tree_lock and pinctrl_hogs_lock aren't useful; the data they protect is read-only except when registering or unregistering a pinctrl_dev, and at those times, it doesn't make sense to protect one part of the structure independently from the rest. Move pinctrl_init_device_debugfs() to the end of pinctrl_register() so that debugfs can't access the struct pinctrl_dev until it's fully initialized, i.e. after the hogs are set up. Signed-off-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: allocate sizeof(*p) instead of sizeof(struct foo)Stephen Warren2012-02-24
| | | | | | | | This hopefully makes it harder to take the sizeof the wrong type. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Dong Aisheng <dong.aisheng@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: use dev_*() instead of pr_*(), add some msgs, minor cleanupsStephen Warren2012-02-24
| | | | | | | | | | | | | e.g. dev_err instead of pr_err prints messages in a slightly more standardized format. Also, add a few more error messages to track down errors. Also, some small cleanups of messages. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Dong Aisheng <dong.aisheng@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: disallow map table entries with NULL dev_name fieldStephen Warren2012-02-24
| | | | | | | | | | | | | | Hog entries are mapping table entries with .ctrl_dev_name == .dev_name. All other mapping table entries need .dev_name set so that they will match some pinctrl_get() call. All extant PIN_MAP*() macros set .dev_name. So, there is no reason to allow mapping table entries without .dev_name set. Update the code and documentation to disallow this. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Dong Aisheng <dong.aisheng@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: fix pinconf_groups_show() to emit newlineStephen Warren2012-02-22
| | | | | | | | pinconf_groups_show() wrote all debug information on one line. Fix it to match pinconf_pins_show() and be legible. Signed-off-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: record a pin owner, not mux function, when requesting pinsStephen Warren2012-02-22
| | | | | | | | | | | | | | | | | When pins are requested/acquired/got, some device becomes the owner of their mux setting. At this point, it isn't certain which mux function will be selected for the pin, since this may vary between each of the device's states in the pinctrl mapping table. As such, we should record the owning device, not what we think the initial mux setting will be, when requesting pins. This doesn't make a lot of difference right now since pinctrl_get gets only one single device/state combination, but this will make a difference when pinctrl_get gets all states, and pinctrl_select_state can switch between states. Signed-off-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: error if mapping table's control dev can't be foundStephen Warren2012-02-22
| | | | | | | | | | This is a serious error, and the pin control system will not function correctly if it ends up not programing the mapping table entries into the HW. Instead of just ignoring this, error out. Signed-off-by: Stephen Warren <swarren@nvidia.com> [rebased to fit the applied patch series, cast error to pointer] Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: downgrade pinctrl_get warning when no maps are foundStephen Warren2012-02-22
| | | | | | | | | | | | | | This may be perfectly legitimate. An IP block may get re-used across SoCs. Not all of those SoCs may need pinmux settings for the IP block, e.g. if one SoC dedicates pins to that function but another doesn't. The driver won't know this, and will always attempt to set up the pinmux. The mapping table defines whether any HW programming is actually needed. Acked-by: Shawn Guo <shawn.guo@linaro.org> Signed-off-by: Stephen Warren <swarren@nvidia.com> [rebased to fit the applied patch series] Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: assume map table entries can't have a NULL ctrl_dev_name fieldStephen Warren2012-02-22
| | | | | | | These are already disallowed. Clean up some code that doesn't assume this. Signed-off-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: spawn U300 pinctrl from the COH901 GPIOLinus Walleij2012-02-22
| | | | | | | | | | | This solves the riddle on how the U300 pin controller shall be able to reference the struct gpio_chip even though these are two separate drivers: spawn the pinctrl child from the GPIO driver and pass in the struct gpio_chip as platform data. In the process we rename the U300 "pinmux-u300" to "pinctrl-u300" so as not to confuse. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: core.c/h cleanupsStephen Warren2012-02-22
| | | | | | | | | * Make all functions internal to core.c static. Remove any of these from core.h. * Add any missing EXPORT_SYMBOL_GPL(). Signed-off-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: Re-order pinconf.[ch] to match each-otherStephen Warren2012-02-22
| | | | | | | | | | Modify the two files so that the order of function prototypes in the header matches the order of implementations in the .c file. Don't prototype a couple of internal functions. Signed-off-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: Re-order pinmux.[ch] to match each-otherStephen Warren2012-02-22
| | | | | | | | Modify the two files so that the order of function prototypes in the header matches the order of implementations in the .c file. Signed-off-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: Store mapping table as a list of chunksStephen Warren2012-02-22
| | | | | | | | | | | | | | | | | | | | Instead of storing a single array of mapping table entries, which requires realloc()ing that array each time it's extended and copying the new data, simply store a list of pointers to the individual chunks. This also removes the need to copy the mapping table at all; a pointer is maintained to the original table, this saving memory. A macro for_each_maps() is introduced to hide the additional complexity of iterating over the map entries. This change will also simplify removing chunks of entries from the mapping table. This isn't important right now, but will be in the future, when mapping table entries are dynamically added when parsing them from the device tree, and removed when drivers no longer need to interact with pinctrl. Signed-off-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: use list_add_tail instead of list_addStephen Warren2012-02-22
| | | | | | | | This mostly makes debugfs files print things in the order that they were added or acquired, which just feels a little more consistent. Signed-off-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: pinctrl_register_mappings() shouldn't be __initStephen Warren2012-02-22
| | | | | | | | | | It may be common for pinctrl_register_mappings() to be used from __init context, but there's no reason that additional mappings shouldn't be added at a later point, e.g. if loading modules that add pin controllers and their mapping tables. Signed-off-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: make "hog" mapping table entries workStephen Warren2012-02-22
| | | | | | | | | | | | Commit 77a5988 "pinctrl: changes hog mechanism to be self-referential" modified the way "hog" entries were represented in the mapping table. However, the new representation failed some error checks in pinctrl_hog_map(). Remove the now-bogus error-check, and fix the code to solve the issue the error-check used to avoid. Acked-by: Dong Aisheng <dong.aisheng@linaro.org> Signed-off-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* serial/sirf: fixup for changes to pin controlLinus Walleij2012-02-22
| | | | | | | | | | | | We changed the signature of the pin multiplexing functions to handle any pin business, so fix up the Sirf driver to call this new interface and rename some variables to make the semantics understandable. Cc: linux-serial@vger.kernel.org Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Acked-by: Barry Song <Baohua.Song@csr.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: changes hog mechanism to be self-referentialLinus Walleij2012-02-10
| | | | | | | | | | | Instead of a specific boolean field to indicate if a map entry shall be hogged, treat self-reference as an indication of desired hogging. This drops one field off the map struct and has a nice Douglas R. Hofstadter-feel to it. Acked-by: Dong Aisheng <dong.aisheng@linaro.org> Acked-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: factor pin control handles over to the coreLinus Walleij2012-02-10
| | | | | | | | | | | | This moves the per-devices struct pinctrl handles and device map over from the pinmux part of the subsystem to the core pinctrl part. This makes the device handles core infrastructure with the goal of using these handles also for pin configuration, so that device drivers (or boards etc) will need one and only one handle to the pin control core. Acked-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: move generic functions to the pinctrl_ namespaceLinus Walleij2012-02-10
| | | | | | | | | | | | | | Since we want to use the former pinmux handles and mapping tables for generic control involving both muxing and configuration we begin refactoring by renaming them from pinmux_* to pinctrl_*. ChangeLog v1->v2: - Also rename the PINMUX_* macros in machine.h to PIN_ as indicated in the documentation so as to reflect the generic nature of these mapping entries from now on. Acked-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: break out a pinctrl consumer headerLinus Walleij2012-02-10
| | | | | | | | | | This breaks out a <linux/pinctrl/consumer.h> header to be used by all pinmux and pinconfig alike, so drivers needing services from pinctrl does not need to include different headers. This is similar to the approach taken by the regulator API. Acked-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: enable pinmux for mmp seriesHaojian Zhuang2012-02-01
| | | | | | | | Support PXA168/PXA910/MMP2 pinmux. Now only support function switch. Signed-off-by: Haojian Zhuang <haojian.zhuang@marvell.com> [Rebase and fix some whitespace issues] Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: delete raw device pointers in pinmux mapsLinus Walleij2012-02-01
| | | | | | | | | | | | After discussion with Mark Brown in an unrelated thread about ADC lookups, it came to my knowledge that the ability to pass a struct device * in the regulator consumers is just a historical artifact, and not really recommended. Since there are no in-kernel users of these pointers, we just kill them right now, before someone starts to use them. Reviewed-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* pinctrl: restore pin namingLinus Walleij2012-02-01
| | | | | | | | | Commit ca53c5f1ca5c936777caca46b7c716a40682ce83 ("pinctrl: conjure names for unnamed pins") made pins lose their identity and only get autogenerated names. Acked-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* Linux 3.3-rc2Linus Torvalds2012-01-31
|
* Merge tag 'for-linus' of git://github.com/prasad-joshi/logfs_upstreamLinus Torvalds2012-01-31
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | There are few important bug fixes for LogFS * tag 'for-linus' of git://github.com/prasad-joshi/logfs_upstream: Logfs: Allow NULL block_isbad() methods logfs: Grow inode in delete path logfs: Free areas before calling generic_shutdown_super() logfs: remove useless BUG_ON MAINTAINERS: Add Prasad Joshi in LogFS maintiners logfs: Propagate page parameter to __logfs_write_inode logfs: set superblock shutdown flag after generic sb shutdown logfs: take write mutex lock during fsync and sync logfs: Prevent memory corruption logfs: update page reference count for pined pages Fix up conflict in fs/logfs/dev_mtd.c due to semantic change in what "mtd->block_isbad" means in commit f2933e86ad93: "Logfs: Allow NULL block_isbad() methods" clashing with the abstraction changes in the commits 7086c19d0742: "mtd: introduce mtd_block_isbad interface" and d58b27ed58a3: "logfs: do not use 'mtd->block_isbad' directly". This resolution takes the semantics from commit f2933e86ad93, and just makes mtd_block_isbad() return zero (false) if the 'block_isbad' function is NULL. But that also means that now "mtd_can_have_bb()" always returns 0. Now, "mtd_block_markbad()" will obviously return an error if the low-level driver doesn't support bad blocks, so this is somewhat non-symmetric, but it actually makes sense if a NULL "block_isbad" function is considered to mean "I assume that all my blocks are always good".
| * Logfs: Allow NULL block_isbad() methodsJoern Engel2012-01-28
| | | | | | | | | | | | | | Not all mtd drivers define block_isbad(). Let's assume no bad blocks instead of refusing to mount. Signed-off-by: Joern Engel <joern@logfs.org>
| * logfs: Grow inode in delete pathJoern Engel2012-01-28
| | | | | | | | | | | | | | | | Can be necessary if an inode gets deleted (through -ENOSPC) before being written. Might be better to move this into logfs_write_rec(), but for now go with the stupid&safe patch. Signed-off-by: Joern Engel <joern@logfs.org>
| * logfs: Free areas before calling generic_shutdown_super()Joern Engel2012-01-28
| | | | | | | | | | | | Or hit an assertion in map_invalidatepage() instead. Signed-off-by: Joern Engel <joern@logfs.org>
| * logfs: remove useless BUG_ONJoern Engel2012-01-28
| | | | | | | | | | | | It prevents write sizes >4k. Signed-off-by: Joern Engel <joern@logfs.org>
| * MAINTAINERS: Add Prasad Joshi in LogFS maintinersPrasad Joshi2012-01-28
| | | | | | | | | | Acked-by: Joern Engel <joern@logfs.org> Signed-off-by: Prasad Joshi <prasadjoshi.linux@gmail.com>
| * logfs: Propagate page parameter to __logfs_write_inodePrasad Joshi2012-01-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | During GC LogFS has to rewrite each valid block to a separate segment. Rewrite operation reads data from an old segment and writes it to a newly allocated segment. Since every write operation changes data block pointers maintained in inode, inode should also be rewritten. In GC path to avoid AB-BA deadlock LogFS marks a page with PG_pre_locked in addition to locking the page (PG_locked). The page lock is ignored iff the page is pre-locked. LogFS uses a special file called segment file. The segment file maintains an 8 bytes entry for every segment. It keeps track of erase count, level etc. for every segment. Bad things happen with a segment belonging to the segment file is GCed ------------[ cut here ]------------ kernel BUG at /home/prasad/logfs/readwrite.c:297! invalid opcode: 0000 [#1] SMP Modules linked in: logfs joydev usbhid hid psmouse e1000 i2c_piix4 serio_raw [last unloaded: logfs] Pid: 20161, comm: mount Not tainted 3.1.0-rc3+ #3 innotek GmbH VirtualBox EIP: 0060:[<f809132a>] EFLAGS: 00010292 CPU: 0 EIP is at logfs_lock_write_page+0x6a/0x70 [logfs] EAX: 00000027 EBX: f73f5b20 ECX: c16007c8 EDX: 00000094 ESI: 00000000 EDI: e59be6e4 EBP: c7337b28 ESP: c7337b18 DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068 Process mount (pid: 20161, ti=c7336000 task=eb323f70 task.ti=c7336000) Stack: f8099a3d c7337b24 f73f5b20 00001002 c7337b50 f8091f6d f8099a4d f80994e4 00000003 00000000 c7337b68 00000000 c67e4400 00001000 c7337b80 f80935e5 00000000 00000000 00000000 00000000 e1fcf000 0000000f e59be618 c70bf900 Call Trace: [<f8091f6d>] logfs_get_write_page.clone.16+0xdd/0x100 [logfs] [<f80935e5>] logfs_mod_segment_entry+0x55/0x110 [logfs] [<f809460d>] logfs_get_segment_entry+0x1d/0x20 [logfs] [<f8091060>] ? logfs_cleanup_journal+0x50/0x50 [logfs] [<f809521b>] ostore_get_erase_count+0x1b/0x40 [logfs] [<f80965b8>] logfs_open_area+0xc8/0x150 [logfs] [<c141a7ec>] ? kmemleak_alloc+0x2c/0x60 [<f809668e>] __logfs_segment_write.clone.16+0x4e/0x1b0 [logfs] [<c10dd563>] ? mempool_kmalloc+0x13/0x20 [<c10dd563>] ? mempool_kmalloc+0x13/0x20 [<f809696f>] logfs_segment_write+0x17f/0x1d0 [logfs] [<f8092e8c>] logfs_write_i0+0x11c/0x180 [logfs] [<f8092f35>] logfs_write_direct+0x45/0x90 [logfs] [<f80934cd>] __logfs_write_buf+0xbd/0xf0 [logfs] [<c102900e>] ? kmap_atomic_prot+0x4e/0xe0 [<f809424b>] logfs_write_buf+0x3b/0x60 [logfs] [<f80947a9>] __logfs_write_inode+0xa9/0x110 [logfs] [<f8094cb0>] logfs_rewrite_block+0xc0/0x110 [logfs] [<f8095300>] ? get_mapping_page+0x10/0x60 [logfs] [<f8095aa0>] ? logfs_load_object_aliases+0x2e0/0x2f0 [logfs] [<f808e57d>] logfs_gc_segment+0x2ad/0x310 [logfs] [<f808e62a>] __logfs_gc_once+0x4a/0x80 [logfs] [<f808ed43>] logfs_gc_pass+0x683/0x6a0 [logfs] [<f8097a89>] logfs_mount+0x5a9/0x680 [logfs] [<c1126b21>] mount_fs+0x21/0xd0 [<c10f6f6f>] ? __alloc_percpu+0xf/0x20 [<c113da41>] ? alloc_vfsmnt+0xb1/0x130 [<c113db4b>] vfs_kern_mount+0x4b/0xa0 [<c113e06e>] do_kern_mount+0x3e/0xe0 [<c113f60d>] do_mount+0x34d/0x670 [<c10f2749>] ? strndup_user+0x49/0x70 [<c113fcab>] sys_mount+0x6b/0xa0 [<c142d87c>] syscall_call+0x7/0xb Code: f8 e8 8b 93 39 c9 8b 45 f8 3e 0f ba 28 00 19 d2 85 d2 74 ca eb d0 0f 0b 8d 45 fc 89 44 24 04 c7 04 24 3d 9a 09 f8 e8 09 92 39 c9 <0f> 0b 8d 74 26 00 55 89 e5 3e 8d 74 26 00 8b 10 80 e6 01 74 09 EIP: [<f809132a>] logfs_lock_write_page+0x6a/0x70 [logfs] SS:ESP 0068:c7337b18 ---[ end trace 96e67d5b3aa3d6ca ]--- The patch passes locked page to __logfs_write_inode. It calls function logfs_get_wblocks() to pre-lock the page. This ensures any further attempts to lock the page are ignored (esp from get_erase_count). Acked-by: Joern Engel <joern@logfs.org> Signed-off-by: Prasad Joshi <prasadjoshi.linux@gmail.com>
| * logfs: set superblock shutdown flag after generic sb shutdownPrasad Joshi2012-01-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | While unmounting the file system LogFS calls generic_shutdown_super. The function does file system independent superblock shutdown. However, it might result in call file system specific inode eviction. LogFS marks FS shutting down by setting bit LOGFS_SB_FLAG_SHUTDOWN in super->s_flags. Since, inode eviction might call truncate on inode, following BUG is observed when file system is unmounted: ------------[ cut here ]------------ kernel BUG at /home/prasad/logfs/segment.c:362! invalid opcode: 0000 [#1] PREEMPT SMP CPU 3 Modules linked in: logfs binfmt_misc ppdev virtio_blk parport_pc lp parport psmouse floppy virtio_pci serio_raw virtio_ring virtio Pid: 1933, comm: umount Not tainted 3.0.0+ #4 Bochs Bochs RIP: 0010:[<ffffffffa008c841>] [<ffffffffa008c841>] logfs_segment_write+0x211/0x230 [logfs] RSP: 0018:ffff880062d7b9e8 EFLAGS: 00010202 RAX: 000000000000000e RBX: ffff88006eca9000 RCX: 0000000000000000 RDX: ffff88006fd87c40 RSI: ffffea00014ff468 RDI: ffff88007b68e000 RBP: ffff880062d7ba48 R08: 8000000020451430 R09: 0000000000000000 R10: dead000000100100 R11: 0000000000000000 R12: ffff88006fd87c40 R13: ffffea00014ff468 R14: ffff88005ad0a460 R15: 0000000000000000 FS: 00007f25d50ea760(0000) GS:ffff88007fd80000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 0000000000d05e48 CR3: 0000000062c72000 CR4: 00000000000006e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process umount (pid: 1933, threadinfo ffff880062d7a000, task ffff880070b44500) Stack: ffff880062d7ba38 ffff88005ad0a508 0000000000001000 0000000000000000 8000000020451430 ffffea00014ff468 ffff880062d7ba48 ffff88005ad0a460 ffff880062d7bad8 ffffea00014ff468 ffff88006fd87c40 0000000000000000 Call Trace: [<ffffffffa0088fee>] logfs_write_i0+0x12e/0x190 [logfs] [<ffffffffa0089360>] __logfs_write_rec+0x140/0x220 [logfs] [<ffffffffa0089312>] __logfs_write_rec+0xf2/0x220 [logfs] [<ffffffffa00894a4>] logfs_write_rec+0x64/0xd0 [logfs] [<ffffffffa0089616>] __logfs_write_buf+0x106/0x110 [logfs] [<ffffffffa008a19e>] logfs_write_buf+0x4e/0x80 [logfs] [<ffffffffa008a6b8>] __logfs_write_inode+0x98/0x110 [logfs] [<ffffffffa008a7c4>] logfs_truncate+0x54/0x290 [logfs] [<ffffffffa008abfc>] logfs_evict_inode+0xdc/0x190 [logfs] [<ffffffff8115eef5>] evict+0x85/0x170 [<ffffffff8115f126>] iput+0xe6/0x1b0 [<ffffffff8115b4a8>] shrink_dcache_for_umount_subtree+0x218/0x280 [<ffffffff8115ce91>] shrink_dcache_for_umount+0x51/0x90 [<ffffffff8114796c>] generic_shutdown_super+0x2c/0x100 [<ffffffffa008cc47>] logfs_kill_sb+0x57/0xf0 [logfs] [<ffffffff81147de5>] deactivate_locked_super+0x45/0x70 [<ffffffff811487ea>] deactivate_super+0x4a/0x70 [<ffffffff81163934>] mntput_no_expire+0xa4/0xf0 [<ffffffff8116469f>] sys_umount+0x6f/0x380 [<ffffffff814dd46b>] system_call_fastpath+0x16/0x1b Code: 55 c8 49 8d b6 a8 00 00 00 45 89 f9 45 89 e8 4c 89 e1 4c 89 55 b8 c7 04 24 00 00 00 00 e8 68 fc ff ff 4c 8b 55 b8 e9 3c ff ff ff <0f> 0b 0f 0b c7 45 c0 00 00 00 00 e9 44 fe ff ff 66 66 66 66 66 RIP [<ffffffffa008c841>] logfs_segment_write+0x211/0x230 [logfs] RSP <ffff880062d7b9e8> ---[ end trace fe6b040cea952290 ]--- Therefore, move super->s_flags setting after the fs-indenpendent work has been finished. Reviewed-by: Joern Engel <joern@logfs.org> Signed-off-by: Prasad Joshi <prasadjoshi.linux@gmail.com>
| * logfs: take write mutex lock during fsync and syncPrasad Joshi2012-01-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | LogFS uses super->s_write_mutex while writing data to disk. Taking the same mutex lock in sync and fsync code path solves the following BUG: ------------[ cut here ]------------ kernel BUG at /home/prasad/logfs/dev_bdev.c:134! Pid: 2387, comm: flush-253:16 Not tainted 3.0.0+ #4 Bochs Bochs RIP: 0010:[<ffffffffa007deed>] [<ffffffffa007deed>] bdev_writeseg+0x25d/0x270 [logfs] Call Trace: [<ffffffffa007c381>] logfs_open_area+0x91/0x150 [logfs] [<ffffffff8128dcb2>] ? find_level.clone.9+0x62/0x100 [<ffffffffa007c49c>] __logfs_segment_write.clone.20+0x5c/0x190 [logfs] [<ffffffff810ef005>] ? mempool_kmalloc+0x15/0x20 [<ffffffff810ef383>] ? mempool_alloc+0x53/0x130 [<ffffffffa007c7a4>] logfs_segment_write+0x1d4/0x230 [logfs] [<ffffffffa0078f8e>] logfs_write_i0+0x12e/0x190 [logfs] [<ffffffffa0079300>] __logfs_write_rec+0x140/0x220 [logfs] [<ffffffffa0079444>] logfs_write_rec+0x64/0xd0 [logfs] [<ffffffffa00795b6>] __logfs_write_buf+0x106/0x110 [logfs] [<ffffffffa007a13e>] logfs_write_buf+0x4e/0x80 [logfs] [<ffffffffa0073e33>] __logfs_writepage+0x23/0x80 [logfs] [<ffffffffa007410c>] logfs_writepage+0xdc/0x110 [logfs] [<ffffffff810f5ba7>] __writepage+0x17/0x40 [<ffffffff810f6208>] write_cache_pages+0x208/0x4f0 [<ffffffff810f5b90>] ? set_page_dirty+0x70/0x70 [<ffffffff810f653a>] generic_writepages+0x4a/0x70 [<ffffffff810f75d1>] do_writepages+0x21/0x40 [<ffffffff8116b9d1>] writeback_single_inode+0x101/0x250 [<ffffffff8116bdbd>] writeback_sb_inodes+0xed/0x1c0 [<ffffffff8116c5fb>] writeback_inodes_wb+0x7b/0x1e0 [<ffffffff8116cc23>] wb_writeback+0x4c3/0x530 [<ffffffff814d984d>] ? sub_preempt_count+0x9d/0xd0 [<ffffffff8116cd6b>] wb_do_writeback+0xdb/0x290 [<ffffffff814d984d>] ? sub_preempt_count+0x9d/0xd0 [<ffffffff814d6208>] ? _raw_spin_unlock_irqrestore+0x18/0x40 [<ffffffff8105aa5a>] ? del_timer+0x8a/0x120 [<ffffffff8116cfac>] bdi_writeback_thread+0x8c/0x2e0 [<ffffffff8116cf20>] ? wb_do_writeback+0x290/0x290 [<ffffffff8106d2e6>] kthread+0x96/0xa0 [<ffffffff814de514>] kernel_thread_helper+0x4/0x10 [<ffffffff8106d250>] ? kthread_worker_fn+0x190/0x190 [<ffffffff814de510>] ? gs_change+0xb/0xb RIP [<ffffffffa007deed>] bdev_writeseg+0x25d/0x270 [logfs] ---[ end trace 0211ad60a57657c4 ]--- Reviewed-by: Joern Engel <joern@logfs.org> Signed-off-by: Prasad Joshi <prasadjoshi.linux@gmail.com>
| * logfs: Prevent memory corruptionJoern Engel2012-01-28
| | | | | | | | | | | | | | | | | | | | This is a bad one. I wonder whether we were so far protected by no_free_segments(sb) usually being smaller than LOGFS_NO_AREAS. Found by Dan Carpenter <dan.carpenter@oracle.com> using smatch. Signed-off-by: Joern Engel <joern@logfs.org> Signed-off-by: Prasad Joshi <prasadjoshi.linux@gmail.com>
| * logfs: update page reference count for pined pagesPrasad Joshi2012-01-28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | LogFS sets PG_private flag to indicate a pined page. We assumed that marking a page as private is enough to ensure its existence. But instead it is necessary to hold a reference count to the page. The change resolves the following BUG BUG: Bad page state in process flush-253:16 pfn:6a6d0 page flags: 0x100000000000808(uptodate|private) Suggested-and-Acked-by: Joern Engel <joern@logfs.org> Signed-off-by: Prasad Joshi <prasadjoshi.linux@gmail.com>
* | Merge branch 'hwmon-for-linus' of ↵Linus Torvalds2012-01-30
|\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging * 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging: hwmon: (w83627ehf) Disable setting DC mode for pwm2, pwm3 on NCT6776F hwmon: (sht15) fix bad error code MAINTAINERS: Drop maintainer for MAX1668 hwmon driver MAINTAINERS: Add hwmon entries for Wolfson hwmon: (f71805f) Fix clamping of temperature limits
| * | hwmon: (w83627ehf) Disable setting DC mode for pwm2, pwm3 on NCT6776FGuenter Roeck2012-01-29
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | NCT6776F only supports pwm mode for pwm2 and pwm3. Return error if an attempt is made to set those pwm channels to DC mode. Signed-off-by: Guenter Roeck <linux@roeck-us.net> Acked-by: Jean Delvare <khali@linux-fr.org> Cc: stable@vger.kernel.org # 3.0+ Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>
generated by cgit v1.2.2 (git 2.25.0) at 2024-12-19 11:50:35 -0500