| Message ID | 20230804145342.1600136-2-nuno.sa@analog.com (mailing list archive) |
|---|---|
| State | Changes Requested |
| Headers | show |
| Series | Add converter framework | expand |
On Fri, 4 Aug 2023 16:53:39 +0200 Nuno Sa <nuno.sa@analog.com> wrote: > Signed-off-by: Nuno Sa <nuno.sa@analog.com> Hi Nuno, One general comment is that you could have stripped this back a fair bit for ease of understanding. At this stage we don't care about things like debug or control of test patterns. Bring those in as extra patches. I haven't fully gotten my head around the ordering constraints on removal. Are there other users of the component framework that have similar problems? Also, I don't yet understand how a multiple front end, single backend setup would work. Or indeed single front end, multiple backend... Maybe we don't need those cases, but if we want this to be useful beyond adi-axi we probably at least want an outline of how they work. Jonathan > --- > drivers/iio/addac/converter.c | 547 ++++++++++++++++++++++++++++ > include/linux/iio/addac/converter.h | 485 ++++++++++++++++++++++++ > 2 files changed, 1032 insertions(+) > create mode 100644 drivers/iio/addac/converter.c > create mode 100644 include/linux/iio/addac/converter.h > > diff --git a/drivers/iio/addac/converter.c b/drivers/iio/addac/converter.c > new file mode 100644 > index 000000000000..31ac704255ad > --- /dev/null > +++ b/drivers/iio/addac/converter.c > @@ -0,0 +1,547 @@ > +// SPDX-License-Identifier: GPL-2.0-only > +/* > + * Framework to handle complex IIO aggregate devices > + * > + * A note on some of the design expectations with regards to lifetimes and > + * devices bringup/removal. > + * > + * The Framework is using, under the wood, the component API which makes it to > + * easy treat a bunch of devices as one aggregate device. This means that the > + * complete thing is only brought to live when all the deviced are probed. To do devices > + * this, two callbacks are used that should in fact completely replace .probe() > + * and .remove(). The formers should only be used the minimum needed. Ideally, I don't follow the sentence in the middle of the line above. > + * only to call the functions to add and remove frontend annd backend devices. Spell check... > + * > + * It is advised for frontend and backend drivers to use their .remove() I'd not 'advise' things. I'd say the 'use' them. > + * callbacks (to use devres API during the frontend and backends initialization). > + * See the comment in @converter_frontend_bind(). > + * > + * It is also assumed that converter objects cannot be accessed once one of the > + * devices of the aggregate device is removed (effectively bringing the all the bringing all the devices down > + * devices down). Based on that assumption, these objects are not refcount which recounted > + * means accessing them will likely fail miserably. I hope that doesn't mean there will be no protection. I don't mind if nothing works but breaking the kernel isn't an option. > + * > + * Copyright (C) 2023 Analog Devices Inc. > + */ > + > +#define dev_fmt(fmt) "Converter - " fmt > + > +#include <linux/component.h> > +#include <linux/debugfs.h> > +#include <linux/device.h> > +#include <linux/err.h> > +#include <linux/iio/addc/converter.h> > +#include <linux/iio/iio.h> > +#include <linux/list.h> > +#include <linux/module.h> > +#include <linux/property.h> > +#include <linux/regmap.h> > + > +struct converter_backend { > + struct list_head entry; > + struct device *dev; > + const struct converter_ops *ops; > + const char *name; > + void *drvdata; > + > + struct regmap *regmap; > + unsigned int cached_reg_addr; > +}; > + > +struct converter_frontend { > + struct list_head list; > + const struct frontend_ops *ops; > + struct device *dev; > +}; > +static int converter_bind(struct device *dev, struct device *aggregate, > + void *data) > +{ > + struct converter_frontend *frontend = dev_get_drvdata(aggregate); > + struct converter_backend *conv = dev_get_drvdata(dev); > + int ret; > + > + ret = conv->ops->backend_init(conv, dev); > + if (ret) > + return ret; > + > + list_add_tail(&conv->entry, &frontend->list); > + > + return 0; > +} > + > +static void converter_unbind(struct device *dev, struct device *aggregate, > + void *data) > +{ > + struct converter_backend *conv = dev_get_drvdata(dev); > + > + if (conv->ops->backend_close) > + conv->ops->backend_close(conv); > + > + /* after this point the converter should not be used anymore */ > + converter_set_drvdata(conv, NULL); > +} > + > +static const struct component_ops converter_component_ops = { > + .bind = converter_bind, > + .unbind = converter_unbind, > +}; > + > +static int converter_frontend_bind(struct device *dev) > +{ > + struct converter_frontend *frontend = dev_get_drvdata(dev); > + int ret; > + > + ret = component_bind_all(dev, NULL); > + if (ret) > + return ret; > + /* > + * We open a new group so that we can control when resources are > + * released and still use device managed (devm_) calls. The expectations > + * are that on probe, backend resources are allocated first followed by > + * the frontend resources (where registering the IIO device must happen) > + * Naturally we want the reverse order on the unbind path and that would > + * not be possible without opening our own devres group. > + > + * Note that the component API also opens it's own devres group when > + * calling the .bind() callbacks for both the aggregate device > + * (our frontend) and each of the components (our backends). On the > + * unbind path, the aggregate .unbind() function is called > + * (@converter_frontend_unbind()) which should be responsible to tear > + * down all the components (effectively releasing all the resources > + * allocated on each component devres group) and only then the aggregate > + * devres group is released. Hence, the order we want to maintain for > + * releasing resources would not be satisfied because backend resources > + * would be freed first. With our own group, we can control when > + * releasing the resources and we do it before @component_unbind_all(). > + * > + * This also relies that internally the component API is releasing each > + * of the component's devres group. That is likely not to change, but > + * maybe we should not trust it and also open our own groups for backend > + * devices?! > + * > + * Another very important thing to keep in mind is that this is only > + * valid if frontend and backend driver's are implementing their > + * .remove() callback to call @converter_frontend_del() and > + * @converter_backend_del(). Calling those functions from > + * devm_add_action* and use devm APIs in .frontend_init() and > + * .backend_init() is not going to work. Not perfect but still better > + * than having to tear everything down in .frontend_close() and > + * .backend_close() That last bit is nasty and will be non obvious to driver authors. I wonder if we can come up with some means to make it hard to do. > + */ > + if (!devres_open_group(dev, frontend, GFP_KERNEL)) > + return -ENOMEM; > + > + ret = frontend->ops->frontend_init(frontend, dev); > + if (ret) { > + devres_release_group(dev, frontend); > + return ret; > + } > + > + devres_close_group(dev, NULL); > + return 0; > +} > + > +static void converter_frontend_unbind(struct device *dev) > +{ > + struct converter_frontend *frontend = dev_get_drvdata(dev); > + > + if (frontend->ops->frontend_close) > + frontend->ops->frontend_close(frontend); > + > + devres_release_group(dev, frontend); > + component_unbind_all(dev, NULL); > + list_del_init(&frontend->list); > +} > + > +static const struct component_master_ops frontend_component_ops = { > + .bind = converter_frontend_bind, > + .unbind = converter_frontend_unbind, > +}; > diff --git a/include/linux/iio/addac/converter.h b/include/linux/iio/addac/converter.h > new file mode 100644 > index 000000000000..09d9d491b2b8 > --- /dev/null > +++ b/include/linux/iio/addac/converter.h > @@ -0,0 +1,485 @@ > +/* SPDX-License-Identifier: GPL-2.0-only */ > +#ifndef _CONVERTER_H > +#define _CONVERTER_H > + > +struct converter_frontend; > +struct converter_backend; > +struct iio_dev; > +struct device; > +struct regmap; > + > +enum converter_test_pattern { > + CONVERTER_PRBS_7, > + CONVERTER_PRBS_15, > + CONVERTER_PRBS_23, > + CONVERTER_PRBS_31, > + CONVERTER_RAMP_NIBBLE, > + CONVERTER_RAMP_16, > + /* vendor specific from 32 */ > + CONVERTER_ADI_PRBS_9A = 32, > + CONVERTER_ADI_PRBS_23A, > + CONVERTER_ADI_PRBS_X, > + CONVERTER_TEST_PATTERN_MAX > +}; > + > +enum converter_data_type { > + CONVERTER_TWOS_COMPLEMENT, > + CONVERTER_OFFSET_BINARY, > + CONVERTER_DATA_TYPE_MAX > +}; > + > +enum converter_edge { > + CONVERTER_RISING_EDGE_SAMPLE, > + CONVERTER_FALLING_EDGE_SAMPLE, > + CONVERTER_EDGE_MAX > +}; > + > +struct converter_chan_status { > + bool errors; > +}; > + > +/** > + * struct converter_data_fmt - Backend data format > + * @type: Data type. > + * @sign_extend: Bool to tell if the data is sign extended. > + * @enable: Enable/Disable the data format module. If disabled, > + * not formatting will happen. > + */ > +struct converter_data_fmt { > + enum converter_data_type type; > + bool sign_extend; > + bool enable; > +}; > + > +/** > + * struct converter_test_pattern_xlate - Helper struct for test pattern handling > + * @pattern: Pattern to configure. > + * @reg_val: Register value for the pattern to configure. > + */ > +struct converter_test_pattern_xlate { > + enum converter_test_pattern pattern; > + unsigned int reg_val; > +}; > + > +/** > + * struct converter_ops - Backend supported operations > + * @backend_init: Mandatory function to initialize the backend device. It > + * should be a replacement for .probe() where the latest just say .probe() again as 'the latest' is a fiddly bit of English > + * should only have to care about doing @converter_add(). > + * @backend_close: Optional function to tear down the device. > + * @enable: Enable the backend device. > + * @disable: Disable the backend device. > + * @data_format_set: Configure the data format for a specific channel. > + * @chan_enable: Enable one channel. > + * @chan_disable: Disable one channel. > + * @iodelay_set: Controls the IO delay for all the lanes at the interface > + * (where data is actually transferred between frontend and > + backend) level. > + * @test_pattern_set: Set's a test pattern to be transmitted/received by the > + * backend. Typically useful for debug or interface > + * purposes calibration. > + */ > +struct converter_ops { > + int (*backend_init)(struct converter_backend *conv, struct device *dev); > + void (*backend_close)(struct converter_backend *conv); > + int (*enable)(struct converter_backend *conv); > + void (*disable)(struct converter_backend *conv); > + int (*data_format_set)(struct converter_backend *conv, > + unsigned int chan, > + const struct converter_data_fmt *data); > + int (*chan_enable)(struct converter_backend *conv, unsigned int chan); > + int (*chan_disable)(struct converter_backend *conv, unsigned int chan); > + int (*iodelay_set)(struct converter_backend *conv, > + unsigned int num_lanes, unsigned int delay); > + int (*test_pattern_set)(struct converter_backend *conv, > + unsigned int chan, > + enum converter_test_pattern pattern); > + int (*chan_status)(struct converter_backend *conv, unsigned int chan, > + struct converter_chan_status *status); > + int (*sample_edge_select)(struct converter_backend *conv, > + enum converter_edge edge); > +}; > + > +/** > + * struct frontend_ops - Frontend supported operations > + * @frontend_init: Mandatory function to initialize the frontend device. It > + * should be a replacement for .probe() where the latest As above. > + * should only have to care about doing @frontend_add(). > + * @frontend_close: Optional function to tear down the device. > + */ > +struct frontend_ops { > + int (*frontend_init)(struct converter_frontend *frontend, > + struct device *dev); > + void (*frontend_close)(struct converter_frontend *frontend); > +}; > + > +/** > + * converter_test_pattern_xlate() - Helper macro for translatting test patterns > + * @pattern: Pattern to translate. > + * @xlate: List of @struct converter_test_pattern_xlate pairs. > + * > + * Simple helper to match a supported pattern and get the register value. Should > + * only be called by backend devices. Automatically computes the number of > + * @xlate entries. > + */ > +#define converter_test_pattern_xlate(pattern, xlate) \ > + __converter_test_pattern_xlate(pattern, xlate, ARRAY_SIZE(xlate)); > + > +#if IS_ENABLED(CONFIG_IIO_CONVERTER) Why? I'd expect any driver that uses this framework to be useless without it, so we shouldn't need protections. Handle that with Kconfig select / depends > + > +/** > + * converter_get_drvdata - Get driver private data > + * @conv: Converter device. Comments should be next to implementation (I got this wrong in original IIO code and still haven't pushed them all down). One reason is that people often change the internals without realizing there is a comment on them in a header that also needs updating. Much harder to do that if it's right in front of you in the c file. > + */ > +void *converter_get_drvdata(const struct converter_backend *conv); > + > +/** > + * converter_set_drvdata - Set driver private data > + * @conv: Converter device. > + * @drvdata: Driver private data. > + */ > +void converter_set_drvdata(struct converter_backend *conv, void *drvdata); > + > +/** > + * converter_set_regmap - Add a regmap object to a converter > + * @conv: Converter device. > + * @regmap: Regmap object. > + */ > +void converter_set_regmap(struct converter_backend *conv, > + struct regmap *regmap); > + > +/** > + * __converter_test_pattern_xlate - Helper macro for translatting test patterns > + * @pattern: Pattern to translate. > + * @xlate: List of @struct converter_test_pattern_xlate pairs. > + * @n_matches: Number of entries in @xlate. > + * > + * Simple helper to match a supported pattern and get the register value. Should > + * only be called by backend devices. > + */ > +int __converter_test_pattern_xlate(unsigned int pattern, > + const struct converter_test_pattern_xlate *xlate, > + int n_matches); > + > +/** > + * > + */ > +int converter_add(struct device *dev, const struct converter_ops *ops); > + > +/** > + * converter_del - Remove the converter device > + * @dev: device to remove from the aggregate > + * > + * Removes the converter from the aggregate device. This tears down the frontend > + * and all the converters. > + * > + * Ideally, this should be called from the backend driver .remove() callback. > + * This means that all the converters (and the frontend) will be tear down before > + * running any specific devres cleanup (at the driver core level). What this all > + * means is that we can use devm_ apis in @backend_init() and being sure those > + * resources will be released after the backend resources and before any devm_* > + * used in @probe(). If that is not the case, one should likely not use any > + * devm_ API in @backend_init(). That means .backend_close() should be > + * provided to do all the necessary cleanups. > + */ > +void converter_del(struct device *dev); > + > +/** > + * converter_enable - Enable the device > + * @conv: Converter device. > + * > + * Enables the backend device. > + * > + * RETURNS: > + * 0 on success, negative error number on failure. > + */ > +int converter_enable(struct converter_backend *conv); > + > +/** > + * converter_disable - Disable the device > + * @conv: Converter device. > + * > + * Disables the backend device. > + */ > +void converter_disable(struct converter_backend *conv); > + > +/** > + * converter_test_pattern_set - Set a test pattern > + * @conv: Converter device. > + * @chan: Channel number. > + * @pattern: Pattern to set. > + * > + * Set's a test pattern to be transmitted/received by the backend. Typically > + * useful for debug or interface calibration purposes. A backend driver can > + * call the @converter_test_pattern_xlate() helper to validate the pattern > + * (given an array of @struct converter_test_pattern_xlate). > + * > + * Note that some patterns might be frontend specific. I.e, as far as the > + * backend is concerned the pattern is valid (from a register point of view) but > + * the actual support for the pattern is not implemented in the device for this > + * specific frontend. It's up to the frontend to ask for a proper pattern > + * (as it should know better). > + * > + * RETURNS: > + * 0 on success, negative error number on failure. > + */ > +int converter_test_pattern_set(struct converter_backend *conv, > + unsigned int chan, > + enum converter_test_pattern pattern); > + > +int converter_chan_status_get(struct converter_backend *conv, > + unsigned int chan, > + struct converter_chan_status *status); > + > +/** > + * converter_data_format_set - Configure the data format > + * @conv: Converter device. > + * @chan: Channel number. > + * @data: Data format. > + * > + * Properly configure a channel with respect to the expected data format. A Configure a channel ... We won't do it improperly ;) > + * @struct converter_data_fmt must be passed with the settings. > + * > + * RETURNS: > + * 0 on success, negative error number on failure. > + */ > +int converter_data_format_set(struct converter_backend *conv, > + unsigned int chan, > + const struct converter_data_fmt *data); > + > +int converter_sample_edge_select(struct converter_backend *conv, > + enum converter_edge edge); > + > +static inline int > +converter_sample_on_falling_edge(struct converter_backend *conv) > +{ > + return converter_sample_edge_select(conv, CONVERTER_RISING_EDGE_SAMPLE); > +} > + > +static inline int > +converter_sample_on_rising_edge(struct converter_backend *conv) > +{ > + return converter_sample_edge_select(conv, CONVERTER_FALLING_EDGE_SAMPLE); > +} > + > +/** > + * converter_chan_enable - Enable a backend channel > + * @conv: Converter device. > + * @chan: Channel number. > + * > + * Enables a channel on the backend device. > + * > + * RETURNS: > + * 0 on success, negative error number on failure. > + */ > +int converter_chan_enable(struct converter_backend *conv, unsigned int chan); > + > +/** > + * converter_chan_disable - Disable a backend channel > + * @conv: Converter device. > + * @chan: Channel number. > + * > + * Disables a channel on the backend device. > + * > + * RETURNS: > + * 0 on success, negative error number on failure. > + */ > +int converter_chan_disable(struct converter_backend *conv, unsigned int chan); > + > +/** > + * converter_iodelay_set - Set's the backend data interface IO delay > + * @conv: Converter device. > + * @num_lanes: Number of lanes in the data interface. > + * @delay: Delay to set. > + * > + * Controls the IO delay for all the lanes at the data interface (where data is > + * actually transferred between frontend and backend) level. > + * > + * RETURNS: > + * 0 on success, negative error number on failure. > + */ > +int converter_iodelay_set(struct converter_backend *conv, > + unsigned int num_lanes, unsigned int delay); > + > +/** > + * converter_frontend_del - Remove the frontend device > + * @dev: Device to remove from the aggregate > + * > + * Removes the frontend from the aggregate device. This tears down the frontend > + * and all the converters. > + * > + * Ideally, this should be called from the frontend driver .remove() callback. > + * This means that all the converters (and the frontend) will be tear down torn down > + * before running any specific devres cleanup (at the driver core level). What > + * this all means is that we can use devm_ apis in .frontend_init() and being > + * sure those resources will be released after the backend resources and before > + * any devm_* used in .probe(). If that is not the case, one should likely not > + * use any devm_ API in .frontend_init(). That means .frontend_close() should be > + * provided to do all the necessary cleanups. You can force a driver remove to tear down another driver binding first though it all gets fiddly. Take a look at how device_release_driver() is used. May well not help you here though - I've not thought it through properly. > + */ > +void converter_frontend_del(struct device *dev); > + > +/** > + * converter_frontend_add - Allocate and add a frontend device > + * @dev: Device to allocate frontend for. > + * @ops: Frontend callbacks. > + * > + * This allocates the frontend device and looks for all converters needed > + * so that, when they are available, all of the devices in the aggregate can be > + * initialized. > + * > + * RETURNS: > + * 0 on success, negative error number on failure. > + */ > +int converter_frontend_add(struct device *dev, const struct frontend_ops *ops); > + > +/** > + * converter_get - Get a converter object > + * @frontend: Frontend device. > + * @name: Converter name. > + * > + * Get's a pointer to a converter device. If name is NULL, then it is assumed > + * that only one backend device is bond with the frontend and the first element > + * in the list is retrieved. Should only be called from the .frontend_init() > + * callback. > + * > + * RETURNS: > + * A converter pointer, negative error pointer otherwise. > + */ > +struct converter_backend *__must_check > +converter_get(const struct converter_frontend *frontend, const char *name); > + > +/** > + * converter_add_direct_reg_access - Add debugfs direct register access > + * @conv: Coverter device > + * @indio_dev: IIO device > + * > + * This is analogous to the typical IIO direct register access in debugfs. The > + * extra converter file will be added in the same debugs dir as @indio_dev. > + * Moreover, if @conv->name is NULL, the file will be called > + * converter_direct_reg_access. Otherwise, will be > + * @conv->name_converter_direct_reg_access. > + */ > +void converter_add_direct_reg_access(struct converter_backend *conv, > + struct iio_dev *indio_dev); > + ... /
On Wed, 2023-08-30 at 18:02 +0100, Jonathan Cameron wrote: > On Fri, 4 Aug 2023 16:53:39 +0200 > Nuno Sa <nuno.sa@analog.com> wrote: > > > Signed-off-by: Nuno Sa <nuno.sa@analog.com> > > Hi Nuno, > Hi Jonathan, Thanks for the initial review... > > One general comment is that you could have stripped this back a fair bit > for ease of understanding. At this stage we don't care about things > like debug or control of test patterns. Bring those in as extra patches. > Agreed... As I mentioned (I think) in the cover, I made the RFC bigger than needed to kind of showcase how we can properly configure the hdl core to support things (interface calibration) that were very hard to do with the current implementation. I'll make sure to add the minimum needed API to accommodate what we have right now. > I haven't fully gotten my head around the ordering constraints on removal. > Are there other users of the component framework that have similar problems? > My understanding on the component API is that one should do all the tear down in the .unbind() callback. As usual, I can see some drivers not really doing that. > Also, I don't yet understand how a multiple front end, single backend setup > would work. Or indeed single front end, multiple backend... Maybe we don't > need those cases, but if we want this to be useful beyond adi-axi we > probably at least want an outline of how they work. > Indeed we can have multiple (and we have it out of tree) backends on one frontend. Think on an ADC/DAC with fairly complex data path with more than one channel/interface (CMOS, LVDS, etc). Typically, in those case, each of the interface will be connected to an instance of the hdl core (the backend). > Jonathan > > > --- > > drivers/iio/addac/converter.c | 547 ++++++++++++++++++++++++++++ > > include/linux/iio/addac/converter.h | 485 ++++++++++++++++++++++++ > > 2 files changed, 1032 insertions(+) > > create mode 100644 drivers/iio/addac/converter.c > > create mode 100644 include/linux/iio/addac/converter.h > > > > diff --git a/drivers/iio/addac/converter.c b/drivers/iio/addac/converter.c > > new file mode 100644 > > index 000000000000..31ac704255ad > > --- /dev/null > > +++ b/drivers/iio/addac/converter.c > > @@ -0,0 +1,547 @@ > > +// SPDX-License-Identifier: GPL-2.0-only > > +/* > > + * Framework to handle complex IIO aggregate devices > > + * > > + * A note on some of the design expectations with regards to lifetimes and > > + * devices bringup/removal. > > + * > > + * The Framework is using, under the wood, the component API which makes it to > > + * easy treat a bunch of devices as one aggregate device. This means that the > > + * complete thing is only brought to live when all the deviced are probed. To do > > devices > > > + * this, two callbacks are used that should in fact completely replace .probe() > > + * and .remove(). The formers should only be used the minimum needed. Ideally, > > I don't follow the sentence in the middle of the line above. > > > + * only to call the functions to add and remove frontend annd backend devices. > Spell check... > > > + * > > + * It is advised for frontend and backend drivers to use their .remove() > > I'd not 'advise' things. I'd say the 'use' them. > > > + * callbacks (to use devres API during the frontend and backends > > initialization). > > + * See the comment in @converter_frontend_bind(). > > + * > > + * It is also assumed that converter objects cannot be accessed once one of the > > + * devices of the aggregate device is removed (effectively bringing the all the > > bringing all the devices down > > > + * devices down). Based on that assumption, these objects are not refcount which > > recounted > > > + * means accessing them will likely fail miserably. > > I hope that doesn't mean there will be no protection. I don't mind if nothing > works > but breaking the kernel isn't an option. > Hmm, well, you'll have a use after free... But one will have to be creative to use one of these objects after releasing the device from the driver (on the unbind path). And here we don't have any interaction with chardevs, etc which might keep references to devices even after unbind. The only place where I can see someone doing it wrong is from a frontend driver if for some reason (that I cannot think of now) we need to keep references/use 'struct converter' after .frontend_close() is called. In that case and if the backend driver was the one being removed/unbind, the converter object will effectively be freed (as it was allocated with devres) and we are left with a possible use after free. But that would be a very strange usecase to be missed in review (I hope :)). We can always refcount the converters (not sure if we need to do it for frontend devices). Sure, drivers can still screw up but at least in that case, the framework is not to blame :). > > + * > > + * Copyright (C) 2023 Analog Devices Inc. > > + */ > > + > > +#define dev_fmt(fmt) "Converter - " fmt > > + > > +#include <linux/component.h> > > +#include <linux/debugfs.h> > > +#include <linux/device.h> > > +#include <linux/err.h> > > +#include <linux/iio/addc/converter.h> > > +#include <linux/iio/iio.h> > > +#include <linux/list.h> > > +#include <linux/module.h> > > +#include <linux/property.h> > > +#include <linux/regmap.h> > > + > > +struct converter_backend { > > + struct list_head entry; > > + struct device *dev; > > + const struct converter_ops *ops; > > + const char *name; > > + void *drvdata; > > + > > + struct regmap *regmap; > > + unsigned int cached_reg_addr; > > +}; > > + > > +struct converter_frontend { > > + struct list_head list; > > + const struct frontend_ops *ops; > > + struct device *dev; > > +}; > > > > > +static int converter_bind(struct device *dev, struct device *aggregate, > > + void *data) > > +{ > > + struct converter_frontend *frontend = dev_get_drvdata(aggregate); > > + struct converter_backend *conv = dev_get_drvdata(dev); > > + int ret; > > + > > + ret = conv->ops->backend_init(conv, dev); > > + if (ret) > > + return ret; > > + > > + list_add_tail(&conv->entry, &frontend->list); > > + > > + return 0; > > +} > > + > > +static void converter_unbind(struct device *dev, struct device *aggregate, > > + void *data) > > +{ > > + struct converter_backend *conv = dev_get_drvdata(dev); > > + > > + if (conv->ops->backend_close) > > + conv->ops->backend_close(conv); > > + > > + /* after this point the converter should not be used anymore */ > > + converter_set_drvdata(conv, NULL); > > +} > > + > > +static const struct component_ops converter_component_ops = { > > + .bind = converter_bind, > > + .unbind = converter_unbind, > > +}; > > + > > +static int converter_frontend_bind(struct device *dev) > > +{ > > + struct converter_frontend *frontend = dev_get_drvdata(dev); > > + int ret; > > + > > + ret = component_bind_all(dev, NULL); > > + if (ret) > > + return ret; > > + /* > > + * We open a new group so that we can control when resources are > > + * released and still use device managed (devm_) calls. The expectations > > + * are that on probe, backend resources are allocated first followed by > > + * the frontend resources (where registering the IIO device must happen) > > + * Naturally we want the reverse order on the unbind path and that would > > + * not be possible without opening our own devres group. > > + > > + * Note that the component API also opens it's own devres group when > > + * calling the .bind() callbacks for both the aggregate device > > + * (our frontend) and each of the components (our backends). On the > > + * unbind path, the aggregate .unbind() function is called > > + * (@converter_frontend_unbind()) which should be responsible to tear > > + * down all the components (effectively releasing all the resources > > + * allocated on each component devres group) and only then the aggregate > > + * devres group is released. Hence, the order we want to maintain for > > + * releasing resources would not be satisfied because backend resources > > + * would be freed first. With our own group, we can control when > > + * releasing the resources and we do it before @component_unbind_all(). > > + * > > + * This also relies that internally the component API is releasing each > > + * of the component's devres group. That is likely not to change, but > > + * maybe we should not trust it and also open our own groups for backend > > + * devices?! > > + * > > + * Another very important thing to keep in mind is that this is only > > + * valid if frontend and backend driver's are implementing their > > + * .remove() callback to call @converter_frontend_del() and > > + * @converter_backend_del(). Calling those functions from > > + * devm_add_action* and use devm APIs in .frontend_init() and > > + * .backend_init() is not going to work. Not perfect but still better > > + * than having to tear everything down in .frontend_close() and > > + * .backend_close() > > That last bit is nasty and will be non obvious to driver authors. > > I wonder if we can come up with some means to make it hard to do. > Yeah, I agree. The alternative is to always bring everything down in .frontend_close() and .backend_close(). But that can also be prone to subtle bugs because it's easy to mess up the ordering when not using devres. So, at this point, I cannot really think on a perfect solution rather than keeping some rules like (assuming we keep the logic we have now): * Using devres on frontend|backend_init() only when .remove() is provided on the driver. * No mixes of devres and .frontend|backend_close() But yeah, would be nice if we could come up with something to make it more obvious to driver authors. We might be able to detect that converter_backend_del() and converter_frontend_del() are under devres while no .frontend|backend_close() is being given. I guess that could be a valid indicator of likely misusage. Or even better (but I'm not sure it's doable with the current devres API), detecting that converter_backend_del() or converter_frontend_del() are under devres while more resources are also allocated in our specific opened groups. That would always be a problem (I think) because the only way for the _del() functions to be under devres is if someone added them (from .probe) with devm_add_action() which means that tearing down the aggregate will happen after some resources (which were allocated in the _init() function) are already freed (as even with new groups, devres will remove things on the reverse order). And that would defenitely be problematic. And, in fact, is the whole reason why I have the .del() functions on .remove() (so, tearing down the aggregate device is the first thing to happen and resources are freed in the reverse order they were allocated). Other thought would be some generic helper macros to use in these type of drivers so a .remove() callback is always added to remove the components. Anyways, even the above might be tricky enough to not include it. I'll give it some thought. > > + */ > > + if (!devres_open_group(dev, frontend, GFP_KERNEL)) > > + return -ENOMEM; > > + > > + ret = frontend->ops->frontend_init(frontend, dev); > > + if (ret) { > > + devres_release_group(dev, frontend); > > + return ret; > > + } > > + > > + devres_close_group(dev, NULL); > > + return 0; > > +} > > + > > +static void converter_frontend_unbind(struct device *dev) > > +{ > > + struct converter_frontend *frontend = dev_get_drvdata(dev); > > + > > + if (frontend->ops->frontend_close) > > + frontend->ops->frontend_close(frontend); > > + > > + devres_release_group(dev, frontend); > > + component_unbind_all(dev, NULL); > > + list_del_init(&frontend->list); > > +} > > + > > +static const struct component_master_ops frontend_component_ops = { > > + .bind = converter_frontend_bind, > > + .unbind = converter_frontend_unbind, > > +}; > > > > diff --git a/include/linux/iio/addac/converter.h > > b/include/linux/iio/addac/converter.h > > new file mode 100644 > > index 000000000000..09d9d491b2b8 > > --- /dev/null > > +++ b/include/linux/iio/addac/converter.h > > @@ -0,0 +1,485 @@ > > +/* SPDX-License-Identifier: GPL-2.0-only */ > > +#ifndef _CONVERTER_H > > +#define _CONVERTER_H > > + > > +struct converter_frontend; > > +struct converter_backend; > > +struct iio_dev; > > +struct device; > > +struct regmap; > > + > > +enum converter_test_pattern { > > + CONVERTER_PRBS_7, > > + CONVERTER_PRBS_15, > > + CONVERTER_PRBS_23, > > + CONVERTER_PRBS_31, > > + CONVERTER_RAMP_NIBBLE, > > + CONVERTER_RAMP_16, > > + /* vendor specific from 32 */ > > + CONVERTER_ADI_PRBS_9A = 32, > > + CONVERTER_ADI_PRBS_23A, > > + CONVERTER_ADI_PRBS_X, > > + CONVERTER_TEST_PATTERN_MAX > > +}; > > + > > +enum converter_data_type { > > + CONVERTER_TWOS_COMPLEMENT, > > + CONVERTER_OFFSET_BINARY, > > + CONVERTER_DATA_TYPE_MAX > > +}; > > + > > +enum converter_edge { > > + CONVERTER_RISING_EDGE_SAMPLE, > > + CONVERTER_FALLING_EDGE_SAMPLE, > > + CONVERTER_EDGE_MAX > > +}; > > + > > +struct converter_chan_status { > > + bool errors; > > +}; > > + > > +/** > > + * struct converter_data_fmt - Backend data format > > + * @type: Data type. > > + * @sign_extend: Bool to tell if the data is sign extended. > > + * @enable: Enable/Disable the data format module. If disabled, > > + * not formatting will happen. > > + */ > > +struct converter_data_fmt { > > + enum converter_data_type type; > > + bool sign_extend; > > + bool enable; > > +}; > > + > > +/** > > + * struct converter_test_pattern_xlate - Helper struct for test pattern handling > > + * @pattern: Pattern to configure. > > + * @reg_val: Register value for the pattern to configure. > > + */ > > +struct converter_test_pattern_xlate { > > + enum converter_test_pattern pattern; > > + unsigned int reg_val; > > +}; > > + > > +/** > > + * struct converter_ops - Backend supported operations > > + * @backend_init: Mandatory function to initialize the backend device. It > > + * should be a replacement for .probe() where the latest > > just say .probe() again as 'the latest' is a fiddly bit of English > > > + * should only have to care about doing @converter_add(). > > + * @backend_close: Optional function to tear down the device. > > + * @enable: Enable the backend device. > > + * @disable: Disable the backend device. > > + * @data_format_set: Configure the data format for a specific channel. > > + * @chan_enable: Enable one channel. > > + * @chan_disable: Disable one channel. > > + * @iodelay_set: Controls the IO delay for all the lanes at the interface > > + * (where data is actually transferred between frontend and > > + backend) level. > > + * @test_pattern_set: Set's a test pattern to be transmitted/received by the > > + * backend. Typically useful for debug or interface > > + * purposes calibration. > > + */ > > +struct converter_ops { > > + int (*backend_init)(struct converter_backend *conv, struct device *dev); > > + void (*backend_close)(struct converter_backend *conv); > > + int (*enable)(struct converter_backend *conv); > > + void (*disable)(struct converter_backend *conv); > > + int (*data_format_set)(struct converter_backend *conv, > > + unsigned int chan, > > + const struct converter_data_fmt *data); > > + int (*chan_enable)(struct converter_backend *conv, unsigned int chan); > > + int (*chan_disable)(struct converter_backend *conv, unsigned int chan); > > + int (*iodelay_set)(struct converter_backend *conv, > > + unsigned int num_lanes, unsigned int delay); > > + int (*test_pattern_set)(struct converter_backend *conv, > > + unsigned int chan, > > + enum converter_test_pattern pattern); > > + int (*chan_status)(struct converter_backend *conv, unsigned int chan, > > + struct converter_chan_status *status); > > + int (*sample_edge_select)(struct converter_backend *conv, > > + enum converter_edge edge); > > +}; > > + > > +/** > > + * struct frontend_ops - Frontend supported operations > > + * @frontend_init: Mandatory function to initialize the frontend device. It > > + * should be a replacement for .probe() where the latest > > As above. > > > > + * should only have to care about doing @frontend_add(). > > + * @frontend_close: Optional function to tear down the device. > > + */ > > +struct frontend_ops { > > + int (*frontend_init)(struct converter_frontend *frontend, > > + struct device *dev); > > + void (*frontend_close)(struct converter_frontend *frontend); > > +}; > > + > > +/** > > + * converter_test_pattern_xlate() - Helper macro for translatting test patterns > > + * @pattern: Pattern to translate. > > + * @xlate: List of @struct converter_test_pattern_xlate pairs. > > + * > > + * Simple helper to match a supported pattern and get the register value. Should > > + * only be called by backend devices. Automatically computes the number of > > + * @xlate entries. > > + */ > > +#define converter_test_pattern_xlate(pattern, xlate) \ > > + __converter_test_pattern_xlate(pattern, xlate, ARRAY_SIZE(xlate)); > > + > > +#if IS_ENABLED(CONFIG_IIO_CONVERTER) > > Why? I'd expect any driver that uses this framework to be useless without > it, so we shouldn't need protections. Handle that with Kconfig select / depends > Well, we do have cases of frontends that might be used in standalone mode (I mean, with no backend device) or with the backend connected. But alright, I will keep things simple for now and let's take care if such case ever get's upstream (hopefully they eventual do :)). > > + > > +/** > > + * converter_get_drvdata - Get driver private data > > + * @conv: Converter device. > > Comments should be next to implementation (I got this wrong in original IIO > code and still haven't pushed them all down). One reason is that people often > change the internals without realizing there is a comment on them in a header > that also needs updating. Much harder to do that if it's right in front of > you in the c file. > Ohh I didn't realised that but makes sense... > > + */ > > +void *converter_get_drvdata(const struct converter_backend *conv); > > + > > +/** > > + * converter_set_drvdata - Set driver private data > > + * @conv: Converter device. > > + * @drvdata: Driver private data. > > + */ > > +void converter_set_drvdata(struct converter_backend *conv, void *drvdata); > > + > > +/** > > + * converter_set_regmap - Add a regmap object to a converter > > + * @conv: Converter device. > > + * @regmap: Regmap object. > > + */ > > +void converter_set_regmap(struct converter_backend *conv, > > + struct regmap *regmap); > > + > > +/** > > + * __converter_test_pattern_xlate - Helper macro for translatting test patterns > > + * @pattern: Pattern to translate. > > + * @xlate: List of @struct converter_test_pattern_xlate pairs. > > + * @n_matches: Number of entries in @xlate. > > + * > > + * Simple helper to match a supported pattern and get the register value. Should > > + * only be called by backend devices. > > + */ > > +int __converter_test_pattern_xlate(unsigned int pattern, > > + const struct converter_test_pattern_xlate > > *xlate, > > + int n_matches); > > + > > +/** > > + * > > + */ > > +int converter_add(struct device *dev, const struct converter_ops *ops); > > + > > +/** > > + * converter_del - Remove the converter device > > + * @dev: device to remove from the aggregate > > + * > > + * Removes the converter from the aggregate device. This tears down the frontend > > + * and all the converters. > > + * > > + * Ideally, this should be called from the backend driver .remove() callback. > > + * This means that all the converters (and the frontend) will be tear down > > before > > + * running any specific devres cleanup (at the driver core level). What this all > > + * means is that we can use devm_ apis in @backend_init() and being sure those > > + * resources will be released after the backend resources and before any devm_* > > + * used in @probe(). If that is not the case, one should likely not use any > > + * devm_ API in @backend_init(). That means .backend_close() should be > > + * provided to do all the necessary cleanups. > > + */ > > +void converter_del(struct device *dev); > > + > > +/** > > + * converter_enable - Enable the device > > + * @conv: Converter device. > > + * > > + * Enables the backend device. > > + * > > + * RETURNS: > > + * 0 on success, negative error number on failure. > > + */ > > +int converter_enable(struct converter_backend *conv); > > + > > +/** > > + * converter_disable - Disable the device > > + * @conv: Converter device. > > + * > > + * Disables the backend device. > > + */ > > +void converter_disable(struct converter_backend *conv); > > + > > +/** > > + * converter_test_pattern_set - Set a test pattern > > + * @conv: Converter device. > > + * @chan: Channel number. > > + * @pattern: Pattern to set. > > + * > > + * Set's a test pattern to be transmitted/received by the backend. Typically > > + * useful for debug or interface calibration purposes. A backend driver can > > + * call the @converter_test_pattern_xlate() helper to validate the pattern > > + * (given an array of @struct converter_test_pattern_xlate). > > + * > > + * Note that some patterns might be frontend specific. I.e, as far as the > > + * backend is concerned the pattern is valid (from a register point of view) but > > + * the actual support for the pattern is not implemented in the device for this > > + * specific frontend. It's up to the frontend to ask for a proper pattern > > + * (as it should know better). > > + * > > + * RETURNS: > > + * 0 on success, negative error number on failure. > > + */ > > +int converter_test_pattern_set(struct converter_backend *conv, > > + unsigned int chan, > > + enum converter_test_pattern pattern); > > + > > +int converter_chan_status_get(struct converter_backend *conv, > > + unsigned int chan, > > + struct converter_chan_status *status); > > + > > +/** > > + * converter_data_format_set - Configure the data format > > + * @conv: Converter device. > > + * @chan: Channel number. > > + * @data: Data format. > > + * > > + * Properly configure a channel with respect to the expected data format. A > Configure a channel ... > > We won't do it improperly ;) > > > + * @struct converter_data_fmt must be passed with the settings. > > + * > > + * RETURNS: > > + * 0 on success, negative error number on failure. > > + */ > > +int converter_data_format_set(struct converter_backend *conv, > > + unsigned int chan, > > + const struct converter_data_fmt *data); > > + > > +int converter_sample_edge_select(struct converter_backend *conv, > > + enum converter_edge edge); > > + > > +static inline int > > +converter_sample_on_falling_edge(struct converter_backend *conv) > > +{ > > + return converter_sample_edge_select(conv, CONVERTER_RISING_EDGE_SAMPLE); > > +} > > + > > +static inline int > > +converter_sample_on_rising_edge(struct converter_backend *conv) > > +{ > > + return converter_sample_edge_select(conv, CONVERTER_FALLING_EDGE_SAMPLE); > > +} > > + > > +/** > > + * converter_chan_enable - Enable a backend channel > > + * @conv: Converter device. > > + * @chan: Channel number. > > + * > > + * Enables a channel on the backend device. > > + * > > + * RETURNS: > > + * 0 on success, negative error number on failure. > > + */ > > +int converter_chan_enable(struct converter_backend *conv, unsigned int chan); > > + > > +/** > > + * converter_chan_disable - Disable a backend channel > > + * @conv: Converter device. > > + * @chan: Channel number. > > + * > > + * Disables a channel on the backend device. > > + * > > + * RETURNS: > > + * 0 on success, negative error number on failure. > > + */ > > +int converter_chan_disable(struct converter_backend *conv, unsigned int chan); > > + > > +/** > > + * converter_iodelay_set - Set's the backend data interface IO delay > > + * @conv: Converter device. > > + * @num_lanes: Number of lanes in the data interface. > > + * @delay: Delay to set. > > + * > > + * Controls the IO delay for all the lanes at the data interface (where data is > > + * actually transferred between frontend and backend) level. > > + * > > + * RETURNS: > > + * 0 on success, negative error number on failure. > > + */ > > +int converter_iodelay_set(struct converter_backend *conv, > > + unsigned int num_lanes, unsigned int delay); > > + > > +/** > > + * converter_frontend_del - Remove the frontend device > > + * @dev: Device to remove from the aggregate > > + * > > + * Removes the frontend from the aggregate device. This tears down the frontend > > + * and all the converters. > > + * > > + * Ideally, this should be called from the frontend driver .remove() callback. > > + * This means that all the converters (and the frontend) will be tear down > torn down > > + * before running any specific devres cleanup (at the driver core level). What > > + * this all means is that we can use devm_ apis in .frontend_init() and being > > + * sure those resources will be released after the backend resources and before > > + * any devm_* used in .probe(). If that is not the case, one should likely not > > + * use any devm_ API in .frontend_init(). That means .frontend_close() should be > > + * provided to do all the necessary cleanups. > > You can force a driver remove to tear down another driver binding first though it > all > gets fiddly. Take a look at how device_release_driver() is used. May well not > help you here though - I've not thought it through properly. > Yeah, I know but I don't think it would help us here or even be correct and I think the problem would be the same. I mean, we would still need to call converter_frontend_del() or converter_backend_del() from the respective .remove() callbacks. BTW, thanks for all the English fixes and help :) - Nuno Sá >
On Thu, 31 Aug 2023 11:32:54 +0200 Nuno Sá <noname.nuno@gmail.com> wrote: > On Wed, 2023-08-30 at 18:02 +0100, Jonathan Cameron wrote: > > On Fri, 4 Aug 2023 16:53:39 +0200 > > Nuno Sa <nuno.sa@analog.com> wrote: > > > > > Signed-off-by: Nuno Sa <nuno.sa@analog.com> > > > > Hi Nuno, > > > > Hi Jonathan, > > Thanks for the initial review... > > > > > One general comment is that you could have stripped this back a fair bit > > for ease of understanding. At this stage we don't care about things > > like debug or control of test patterns. Bring those in as extra patches. > > > > Agreed... As I mentioned (I think) in the cover, I made the RFC bigger than needed to > kind of showcase how we can properly configure the hdl core to support things > (interface calibration) that were very hard to do with the current implementation. > I'll make sure to add the minimum needed API to accommodate what we have right now. > > > I haven't fully gotten my head around the ordering constraints on removal. > > Are there other users of the component framework that have similar problems? > > > > My understanding on the component API is that one should do all the tear down in the > .unbind() callback. As usual, I can see some drivers not really doing that. > > > Also, I don't yet understand how a multiple front end, single backend setup > > would work. Or indeed single front end, multiple backend... Maybe we don't > > need those cases, but if we want this to be useful beyond adi-axi we > > probably at least want an outline of how they work. > > > > Indeed we can have multiple (and we have it out of tree) backends on one frontend. > Think on an ADC/DAC with fairly complex data path with more than one > channel/interface (CMOS, LVDS, etc). Typically, in those case, each of the interface > will be connected to an instance of the hdl core (the backend). That might work out for your case, but not the stm32 one where I think we can end up with interleaved data from two front ends in the same buffer... > > > Jonathan > > > > > --- > > > drivers/iio/addac/converter.c | 547 ++++++++++++++++++++++++++++ > > > include/linux/iio/addac/converter.h | 485 ++++++++++++++++++++++++ > > > 2 files changed, 1032 insertions(+) > > > create mode 100644 drivers/iio/addac/converter.c > > > create mode 100644 include/linux/iio/addac/converter.h > > > > > > diff --git a/drivers/iio/addac/converter.c b/drivers/iio/addac/converter.c > > > new file mode 100644 > > > index 000000000000..31ac704255ad > > > --- /dev/null > > > +++ b/drivers/iio/addac/converter.c > > > @@ -0,0 +1,547 @@ > > > +// SPDX-License-Identifier: GPL-2.0-only > > > +/* > > > + * Framework to handle complex IIO aggregate devices > > > + * > > > + * A note on some of the design expectations with regards to lifetimes and > > > + * devices bringup/removal. > > > + * > > > + * The Framework is using, under the wood, the component API which makes it to > > > + * easy treat a bunch of devices as one aggregate device. This means that the > > > + * complete thing is only brought to live when all the deviced are probed. To do > > > > devices > > > > > + * this, two callbacks are used that should in fact completely replace .probe() > > > + * and .remove(). The formers should only be used the minimum needed. Ideally, > > > > I don't follow the sentence in the middle of the line above. > > > > > + * only to call the functions to add and remove frontend annd backend devices. > > Spell check... > > > > > + * > > > + * It is advised for frontend and backend drivers to use their .remove() > > > > I'd not 'advise' things. I'd say the 'use' them. > > > > > + * callbacks (to use devres API during the frontend and backends > > > initialization). > > > + * See the comment in @converter_frontend_bind(). > > > + * > > > + * It is also assumed that converter objects cannot be accessed once one of the > > > + * devices of the aggregate device is removed (effectively bringing the all the > > > > bringing all the devices down > > > > > + * devices down). Based on that assumption, these objects are not refcount which > > > > recounted > > > > > + * means accessing them will likely fail miserably. > > > > I hope that doesn't mean there will be no protection. I don't mind if nothing > > works > > but breaking the kernel isn't an option. > > > > Hmm, well, you'll have a use after free... But one will have to be creative to use > one of these objects after releasing the device from the driver (on the unbind path). > And here we don't have any interaction with chardevs, etc which might keep references > to devices even after unbind. > > The only place where I can see someone doing it wrong is from a frontend driver if > for some reason (that I cannot think of now) we need to keep references/use 'struct > converter' after .frontend_close() is called. In that case and if the backend driver > was the one being removed/unbind, the converter object will effectively be freed (as > it was allocated with devres) and we are left with a possible use after free. But > that would be a very strange usecase to be missed in review (I hope :)). > > We can always refcount the converters (not sure if we need to do it for frontend > devices). Sure, drivers can still screw up but at least in that case, the framework > is not to blame :). If the rules are clearly stated (with some reasoning) I don't think we need to care about saying what happens if you break them. People will always shoot themselves in the foot, but as long as it is reasonably fiddly to do that's fine by me :) ... > > > +static int converter_frontend_bind(struct device *dev) > > > +{ > > > + struct converter_frontend *frontend = dev_get_drvdata(dev); > > > + int ret; > > > + > > > + ret = component_bind_all(dev, NULL); > > > + if (ret) > > > + return ret; > > > + /* > > > + * We open a new group so that we can control when resources are > > > + * released and still use device managed (devm_) calls. The expectations > > > + * are that on probe, backend resources are allocated first followed by > > > + * the frontend resources (where registering the IIO device must happen) > > > + * Naturally we want the reverse order on the unbind path and that would > > > + * not be possible without opening our own devres group. > > > + > > > + * Note that the component API also opens it's own devres group when > > > + * calling the .bind() callbacks for both the aggregate device > > > + * (our frontend) and each of the components (our backends). On the > > > + * unbind path, the aggregate .unbind() function is called > > > + * (@converter_frontend_unbind()) which should be responsible to tear > > > + * down all the components (effectively releasing all the resources > > > + * allocated on each component devres group) and only then the aggregate > > > + * devres group is released. Hence, the order we want to maintain for > > > + * releasing resources would not be satisfied because backend resources > > > + * would be freed first. With our own group, we can control when > > > + * releasing the resources and we do it before @component_unbind_all(). > > > + * > > > + * This also relies that internally the component API is releasing each > > > + * of the component's devres group. That is likely not to change, but > > > + * maybe we should not trust it and also open our own groups for backend > > > + * devices?! > > > + * > > > + * Another very important thing to keep in mind is that this is only > > > + * valid if frontend and backend driver's are implementing their > > > + * .remove() callback to call @converter_frontend_del() and > > > + * @converter_backend_del(). Calling those functions from > > > + * devm_add_action* and use devm APIs in .frontend_init() and > > > + * .backend_init() is not going to work. Not perfect but still better > > > + * than having to tear everything down in .frontend_close() and > > > + * .backend_close() > > > > That last bit is nasty and will be non obvious to driver authors. > > > > I wonder if we can come up with some means to make it hard to do. > > > > Yeah, I agree. The alternative is to always bring everything down in > .frontend_close() and .backend_close(). But that can also be prone to subtle bugs > because it's easy to mess up the ordering when not using devres. > > So, at this point, I cannot really think on a perfect solution rather than keeping > some rules like (assuming we keep the logic we have now): > > * Using devres on frontend|backend_init() only when .remove() is provided on the > driver. > * No mixes of devres and .frontend|backend_close() > > But yeah, would be nice if we could come up with something to make it more obvious to > driver authors. > > We might be able to detect that converter_backend_del() and converter_frontend_del() > are under devres while no .frontend|backend_close() is being given. I guess that > could be a valid indicator of likely misusage. > > Or even better (but I'm not sure it's doable with the current devres API), detecting > that converter_backend_del() or converter_frontend_del() are under devres while more > resources are also allocated in our specific opened groups. That would always be a > problem (I think) because the only way for the _del() functions to be under devres is > if someone added them (from .probe) with devm_add_action() which means that tearing > down the aggregate will happen after some resources (which were allocated in the > _init() function) are already freed (as even with new groups, devres will remove > things on the reverse order). And that would defenitely be problematic. And, in fact, > is the whole reason why I have the .del() functions on .remove() (so, tearing down > the aggregate device is the first thing to happen and resources are freed in the > reverse order they were allocated). > I couldn't work out how to do anything easily and would need some experiments. Maybe some 'hidden' devres callbacks and a state flag somewhere. If we register that very late we can perhaps detect that we entered devres cleanup before calling expected manual cleanup. I'm thinking have the setup path register a flag checking callback and the cleanup path set a flag (devres now safe). Then we can at least make it scream if we end up doing things in wrong way. > Other thought would be some generic helper macros to use in these type of drivers so > a .remove() callback is always added to remove the components. I wondered if that could work but it's an ugly macro because needs to deal with different bus types. > > Anyways, even the above might be tricky enough to not include it. I'll give it some > thought. Agreed - seems fiddly but maybe there is a neat trick to it. Or indeed another subsystem already doing something. > > > +#define converter_test_pattern_xlate(pattern, xlate) \ > > > + __converter_test_pattern_xlate(pattern, xlate, ARRAY_SIZE(xlate)); > > > + > > > +#if IS_ENABLED(CONFIG_IIO_CONVERTER) > > > > Why? I'd expect any driver that uses this framework to be useless without > > it, so we shouldn't need protections. Handle that with Kconfig select / depends > > > > Well, we do have cases of frontends that might be used in standalone mode (I mean, > with no backend device) or with the backend connected. But alright, I will keep > things simple for now and let's take care if such case ever get's upstream (hopefully > they eventual do :)). Yeah. Introduce the stubs when you need them ;) ... > > > + > > > +/** > > > + * converter_frontend_del - Remove the frontend device > > > + * @dev: Device to remove from the aggregate > > > + * > > > + * Removes the frontend from the aggregate device. This tears down the frontend > > > + * and all the converters. > > > + * > > > + * Ideally, this should be called from the frontend driver .remove() callback. > > > + * This means that all the converters (and the frontend) will be tear down > > torn down > > > + * before running any specific devres cleanup (at the driver core level). What > > > + * this all means is that we can use devm_ apis in .frontend_init() and being > > > + * sure those resources will be released after the backend resources and before > > > + * any devm_* used in .probe(). If that is not the case, one should likely not > > > + * use any devm_ API in .frontend_init(). That means .frontend_close() should be > > > + * provided to do all the necessary cleanups. > > > > You can force a driver remove to tear down another driver binding first though it > > all > > gets fiddly. Take a look at how device_release_driver() is used. May well not > > help you here though - I've not thought it through properly. > > > > Yeah, I know but I don't think it would help us here or even be correct and I think > the problem would be the same. I mean, we would still need to call > converter_frontend_del() or converter_backend_del() from the respective .remove() > callbacks. > Agreed. I'm also not sure it helps and low on time at the moment to experiment around this. Jonathan
On Sun, 2023-09-03 at 11:56 +0100, Jonathan Cameron wrote: > On Thu, 31 Aug 2023 11:32:54 +0200 > Nuno Sá <noname.nuno@gmail.com> wrote: > > > On Wed, 2023-08-30 at 18:02 +0100, Jonathan Cameron wrote: > > > On Fri, 4 Aug 2023 16:53:39 +0200 > > > Nuno Sa <nuno.sa@analog.com> wrote: > > > > > > > Signed-off-by: Nuno Sa <nuno.sa@analog.com> > > > > > > Hi Nuno, > > > > > > > Hi Jonathan, > > > > Thanks for the initial review... > > > > > > > > One general comment is that you could have stripped this back a fair bit > > > for ease of understanding. At this stage we don't care about things > > > like debug or control of test patterns. Bring those in as extra patches. > > > > > > > Agreed... As I mentioned (I think) in the cover, I made the RFC bigger than > > needed to > > kind of showcase how we can properly configure the hdl core to support > > things > > (interface calibration) that were very hard to do with the current > > implementation. > > I'll make sure to add the minimum needed API to accommodate what we have > > right now. > > > > > I haven't fully gotten my head around the ordering constraints on removal. > > > Are there other users of the component framework that have similar > > > problems? > > > > > > > My understanding on the component API is that one should do all the tear > > down in the > > .unbind() callback. As usual, I can see some drivers not really doing that. > > > > > Also, I don't yet understand how a multiple front end, single backend > > > setup > > > would work. Or indeed single front end, multiple backend... Maybe we > > > don't > > > need those cases, but if we want this to be useful beyond adi-axi we > > > probably at least want an outline of how they work. > > > > > > > Indeed we can have multiple (and we have it out of tree) backends on one > > frontend. > > Think on an ADC/DAC with fairly complex data path with more than one > > channel/interface (CMOS, LVDS, etc). Typically, in those case, each of the > > interface > > will be connected to an instance of the hdl core (the backend). > > That might work out for your case, but not the stm32 one where I think we can > end > up with interleaved data from two front ends in the same buffer... > Not sure I'm following this one. But wouldn't that be something specific for each system (through devicetree)? I haven't tried but I think the same backend could be used in different frontend devices (using the component API). That is not really a usecase for me but definitely something that could be supported (if we need to start doing things like keep enable/disable counters and so on) if it is a usecase for stm32. > > > > > Jonathan > > > > > > > --- > > > > drivers/iio/addac/converter.c | 547 ++++++++++++++++++++++++++++ > > > > include/linux/iio/addac/converter.h | 485 ++++++++++++++++++++++++ > > > > 2 files changed, 1032 insertions(+) > > > > create mode 100644 drivers/iio/addac/converter.c > > > > create mode 100644 include/linux/iio/addac/converter.h > > > > > > > > diff --git a/drivers/iio/addac/converter.c > > > > b/drivers/iio/addac/converter.c > > > > new file mode 100644 > > > > index 000000000000..31ac704255ad > > > > --- /dev/null > > > > +++ b/drivers/iio/addac/converter.c > > > > @@ -0,0 +1,547 @@ > > > > +// SPDX-License-Identifier: GPL-2.0-only > > > > +/* > > > > + * Framework to handle complex IIO aggregate devices > > > > + * > > > > + * A note on some of the design expectations with regards to lifetimes > > > > and > > > > + * devices bringup/removal. > > > > + * > > > > + * The Framework is using, under the wood, the component API which > > > > makes it to > > > > + * easy treat a bunch of devices as one aggregate device. This means > > > > that the > > > > + * complete thing is only brought to live when all the deviced are > > > > probed. To do > > > > > > devices > > > > > > > + * this, two callbacks are used that should in fact completely replace > > > > .probe() > > > > + * and .remove(). The formers should only be used the minimum needed. > > > > Ideally, > > > > > > I don't follow the sentence in the middle of the line above. > > > > > > > + * only to call the functions to add and remove frontend annd backend > > > > devices. > > > Spell check... > > > > > > > + * > > > > + * It is advised for frontend and backend drivers to use their > > > > .remove() > > > > > > I'd not 'advise' things. I'd say the 'use' them. > > > > > > > + * callbacks (to use devres API during the frontend and backends > > > > initialization). > > > > + * See the comment in @converter_frontend_bind(). > > > > + * > > > > + * It is also assumed that converter objects cannot be accessed once > > > > one of the > > > > + * devices of the aggregate device is removed (effectively bringing the > > > > all the > > > > > > bringing all the devices down > > > > > > > + * devices down). Based on that assumption, these objects are not > > > > refcount which > > > > > > recounted > > > > > > > + * means accessing them will likely fail miserably. > > > > > > I hope that doesn't mean there will be no protection. I don't mind if > > > nothing > > > works > > > but breaking the kernel isn't an option. > > > > > > > Hmm, well, you'll have a use after free... But one will have to be creative > > to use > > one of these objects after releasing the device from the driver (on the > > unbind path). > > And here we don't have any interaction with chardevs, etc which might keep > > references > > to devices even after unbind. > > > > The only place where I can see someone doing it wrong is from a frontend > > driver if > > for some reason (that I cannot think of now) we need to keep references/use > > 'struct > > converter' after .frontend_close() is called. In that case and if the > > backend driver > > was the one being removed/unbind, the converter object will effectively be > > freed (as > > it was allocated with devres) and we are left with a possible use after > > free. But > > that would be a very strange usecase to be missed in review (I hope :)). > > > > We can always refcount the converters (not sure if we need to do it for > > frontend > > devices). Sure, drivers can still screw up but at least in that case, the > > framework > > is not to blame :). > > If the rules are clearly stated (with some reasoning) I don't think we need > to care about saying what happens if you break them. People will always shoot > themselves in the foot, but as long as it is reasonably fiddly to do that's > fine by me :) > > ... > > > > > +static int converter_frontend_bind(struct device *dev) > > > > +{ > > > > + struct converter_frontend *frontend = dev_get_drvdata(dev); > > > > + int ret; > > > > + > > > > + ret = component_bind_all(dev, NULL); > > > > + if (ret) > > > > + return ret; > > > > + /* > > > > + * We open a new group so that we can control when resources are > > > > + * released and still use device managed (devm_) calls. The > > > > expectations > > > > + * are that on probe, backend resources are allocated first > > > > followed by > > > > + * the frontend resources (where registering the IIO device must > > > > happen) > > > > + * Naturally we want the reverse order on the unbind path and > > > > that would > > > > + * not be possible without opening our own devres group. > > > > + > > > > + * Note that the component API also opens it's own devres group > > > > when > > > > + * calling the .bind() callbacks for both the aggregate device > > > > + * (our frontend) and each of the components (our backends). On > > > > the > > > > + * unbind path, the aggregate .unbind() function is called > > > > + * (@converter_frontend_unbind()) which should be responsible to > > > > tear > > > > + * down all the components (effectively releasing all the > > > > resources > > > > + * allocated on each component devres group) and only then the > > > > aggregate > > > > + * devres group is released. Hence, the order we want to > > > > maintain for > > > > + * releasing resources would not be satisfied because backend > > > > resources > > > > + * would be freed first. With our own group, we can control when > > > > + * releasing the resources and we do it before > > > > @component_unbind_all(). > > > > + * > > > > + * This also relies that internally the component API is > > > > releasing each > > > > + * of the component's devres group. That is likely not to > > > > change, but > > > > + * maybe we should not trust it and also open our own groups for > > > > backend > > > > + * devices?! > > > > + * > > > > + * Another very important thing to keep in mind is that this is > > > > only > > > > + * valid if frontend and backend driver's are implementing their > > > > + * .remove() callback to call @converter_frontend_del() and > > > > + * @converter_backend_del(). Calling those functions from > > > > + * devm_add_action* and use devm APIs in .frontend_init() and > > > > + * .backend_init() is not going to work. Not perfect but still > > > > better > > > > + * than having to tear everything down in .frontend_close() and > > > > + * .backend_close() > > > > > > That last bit is nasty and will be non obvious to driver authors. > > > > > > I wonder if we can come up with some means to make it hard to do. > > > > > > > Yeah, I agree. The alternative is to always bring everything down in > > .frontend_close() and .backend_close(). But that can also be prone to subtle > > bugs > > because it's easy to mess up the ordering when not using devres. > > > > So, at this point, I cannot really think on a perfect solution rather than > > keeping > > some rules like (assuming we keep the logic we have now): > > > > * Using devres on frontend|backend_init() only when .remove() is provided on > > the > > driver. > > * No mixes of devres and .frontend|backend_close() > > > > But yeah, would be nice if we could come up with something to make it more > > obvious to > > driver authors. > > > > > We might be able to detect that converter_backend_del() and > > converter_frontend_del() > > are under devres while no .frontend|backend_close() is being given. I guess > > that > > could be a valid indicator of likely misusage. > > > > Or even better (but I'm not sure it's doable with the current devres API), > > detecting > > that converter_backend_del() or converter_frontend_del() are under devres > > while more > > resources are also allocated in our specific opened groups. That would > > always be a > > problem (I think) because the only way for the _del() functions to be under > > devres is > > if someone added them (from .probe) with devm_add_action() which means that > > tearing > > down the aggregate will happen after some resources (which were allocated in > > the > > _init() function) are already freed (as even with new groups, devres will > > remove > > things on the reverse order). And that would defenitely be problematic. And, > > in fact, > > is the whole reason why I have the .del() functions on .remove() (so, > > tearing down > > the aggregate device is the first thing to happen and resources are freed in > > the > > reverse order they were allocated). > > > This would actually be very messy and hard to do properly. Concurrency between the aggregate probing (at the component level) and unbinding (at driver core level) would be very tricky if doable at all. On top that, we do have devres_find() but no way to tell if a devres group has resources or not. It would be easy to add one new API but likely not worth it just for this usecase. I also thought about an helper macro to wrap every devm_ call but your below suggestion is way better and transparent to users. > I couldn't work out how to do anything easily and would need some experiments. > Maybe some 'hidden' devres callbacks and a state flag somewhere. If we > register > that very late we can perhaps detect that we entered devres cleanup before > calling > expected manual cleanup. I'm thinking have the setup path register a flag > checking > callback and the cleanup path set a flag (devres now safe). Then we can at > least > make it scream if we end up doing things in wrong way. > Hmm, that might actually be a good idea and something to try. It likely means having a mutex (I was happy not to have one for now :) but likely inevitable somewhere down the road) for the flag but it might work. At least we'll be able to dump a WARN or something if we suspect something is wrong. > > Other thought would be some generic helper macros to use in these type of > > drivers so > > a .remove() callback is always added to remove the components. > I wondered if that could work but it's an ugly macro because needs to deal > with > different bus types. > > We could have a macro per bus type. I'm not really seeing anything different than platform, spi and i2c. But even like this, it could easily start to get ugly because of accepted parameters (and any deviation would again mean a different version per bus type). Another thing that crossed my mind was wrappers to module_spi_driver() and friends. Maybe that could work. Anyways, as you said in one of your replies, this is in kernel interface that we can easily add/change. So, I will defer this to a later point in time. - Nuno Sá
On Mon, 04 Sep 2023 16:14:17 +0200 Nuno Sá <noname.nuno@gmail.com> wrote: > On Sun, 2023-09-03 at 11:56 +0100, Jonathan Cameron wrote: > > On Thu, 31 Aug 2023 11:32:54 +0200 > > Nuno Sá <noname.nuno@gmail.com> wrote: > > > > > On Wed, 2023-08-30 at 18:02 +0100, Jonathan Cameron wrote: > > > > On Fri, 4 Aug 2023 16:53:39 +0200 > > > > Nuno Sa <nuno.sa@analog.com> wrote: > > > > > > > > > Signed-off-by: Nuno Sa <nuno.sa@analog.com> > > > > > > > > Hi Nuno, > > > > > > > > > > Hi Jonathan, > > > > > > Thanks for the initial review... > > > > > > > > > > > One general comment is that you could have stripped this back a fair bit > > > > for ease of understanding. At this stage we don't care about things > > > > like debug or control of test patterns. Bring those in as extra patches. > > > > > > > > > > Agreed... As I mentioned (I think) in the cover, I made the RFC bigger than > > > needed to > > > kind of showcase how we can properly configure the hdl core to support > > > things > > > (interface calibration) that were very hard to do with the current > > > implementation. > > > I'll make sure to add the minimum needed API to accommodate what we have > > > right now. > > > > > > > I haven't fully gotten my head around the ordering constraints on removal. > > > > Are there other users of the component framework that have similar > > > > problems? > > > > > > > > > > My understanding on the component API is that one should do all the tear > > > down in the > > > .unbind() callback. As usual, I can see some drivers not really doing that. > > > > > > > Also, I don't yet understand how a multiple front end, single backend > > > > setup > > > > would work. Or indeed single front end, multiple backend... Maybe we > > > > don't > > > > need those cases, but if we want this to be useful beyond adi-axi we > > > > probably at least want an outline of how they work. > > > > > > > > > > Indeed we can have multiple (and we have it out of tree) backends on one > > > frontend. > > > Think on an ADC/DAC with fairly complex data path with more than one > > > channel/interface (CMOS, LVDS, etc). Typically, in those case, each of the > > > interface > > > will be connected to an instance of the hdl core (the backend). > > > > That might work out for your case, but not the stm32 one where I think we can > > end > > up with interleaved data from two front ends in the same buffer... > > > > Not sure I'm following this one. But wouldn't that be something specific for > each system (through devicetree)? I haven't tried but I think the same backend > could be used in different frontend devices (using the component API). That is > not really a usecase for me but definitely something that could be supported (if > we need to start doing things like keep enable/disable counters and so on) if it > is a usecase for stm32. If we are going to support both usecases, we just need to figure out what composite devices with N-M backend - frontend look like and make sure that doesn't cause problems. I'd expect the separation between backend instances might reflect data storage on capture but then again that might end up like the many IIO devices for many buffers mess we had before the multiple buffer support was added. Might need enable /disable counters as you say - I'm not quite sure without trying it! > > > > > + * than having to tear everything down in .frontend_close() and > > > > > + * .backend_close() > > > > > > > > That last bit is nasty and will be non obvious to driver authors. > > > > > > > > I wonder if we can come up with some means to make it hard to do. > > > > > > > > > > Yeah, I agree. The alternative is to always bring everything down in > > > .frontend_close() and .backend_close(). But that can also be prone to subtle > > > bugs > > > because it's easy to mess up the ordering when not using devres. > > > > > > So, at this point, I cannot really think on a perfect solution rather than > > > keeping > > > some rules like (assuming we keep the logic we have now): > > > > > > * Using devres on frontend|backend_init() only when .remove() is provided on > > > the > > > driver. > > > * No mixes of devres and .frontend|backend_close() > > > > > > But yeah, would be nice if we could come up with something to make it more > > > obvious to > > > driver authors. > > > > > > > > We might be able to detect that converter_backend_del() and > > > converter_frontend_del() > > > are under devres while no .frontend|backend_close() is being given. I guess > > > that > > > could be a valid indicator of likely misusage. > > > > > > Or even better (but I'm not sure it's doable with the current devres API), > > > detecting > > > that converter_backend_del() or converter_frontend_del() are under devres > > > while more > > > resources are also allocated in our specific opened groups. That would > > > always be a > > > problem (I think) because the only way for the _del() functions to be under > > > devres is > > > if someone added them (from .probe) with devm_add_action() which means that > > > tearing > > > down the aggregate will happen after some resources (which were allocated in > > > the > > > _init() function) are already freed (as even with new groups, devres will > > > remove > > > things on the reverse order). And that would defenitely be problematic. And, > > > in fact, > > > is the whole reason why I have the .del() functions on .remove() (so, > > > tearing down > > > the aggregate device is the first thing to happen and resources are freed in > > > the > > > reverse order they were allocated). > > > > > > > This would actually be very messy and hard to do properly. Concurrency between > the aggregate probing (at the component level) and unbinding (at driver core > level) would be very tricky if doable at all. On top that, we do have > devres_find() but no way to tell if a devres group has resources or not. It > would be easy to add one new API but likely not worth it just for this usecase. > > I also thought about an helper macro to wrap every devm_ call but your below > suggestion is way better and transparent to users. > > > I couldn't work out how to do anything easily and would need some experiments. > > Maybe some 'hidden' devres callbacks and a state flag somewhere. If we > > register > > that very late we can perhaps detect that we entered devres cleanup before > > calling > > expected manual cleanup. I'm thinking have the setup path register a flag > > checking > > callback and the cleanup path set a flag (devres now safe). Then we can at > > least > > make it scream if we end up doing things in wrong way. > > > > Hmm, that might actually be a good idea and something to try. It likely means > having a mutex (I was happy not to have one for now :) but likely inevitable > somewhere down the road) for the flag but it might work. At least we'll be able > to dump a WARN or something if we suspect something is wrong. Careful ordering an atomic might work. > > > > Other thought would be some generic helper macros to use in these type of > > > drivers so > > > a .remove() callback is always added to remove the components. > > I wondered if that could work but it's an ugly macro because needs to deal > > with > > different bus types. > > > > > > We could have a macro per bus type. I'm not really seeing anything different > than platform, spi and i2c. But even like this, it could easily start to get > ugly because of accepted parameters (and any deviation would again mean a > different version per bus type). Another thing that crossed my mind was wrappers > to module_spi_driver() and friends. Maybe that could work. Anyways, as you said > in one of your replies, this is in kernel interface that we can easily > add/change. So, I will defer this to a later point in time. Makes sense J > > - Nuno Sá >
Hi Nuno, Jonathan, On 9/4/23 17:31, Jonathan Cameron wrote: > On Mon, 04 Sep 2023 16:14:17 +0200 > Nuno Sá <noname.nuno@gmail.com> wrote: > >> On Sun, 2023-09-03 at 11:56 +0100, Jonathan Cameron wrote: >>> On Thu, 31 Aug 2023 11:32:54 +0200 >>> Nuno Sá <noname.nuno@gmail.com> wrote: >>> >>>> On Wed, 2023-08-30 at 18:02 +0100, Jonathan Cameron wrote: >>>>> On Fri, 4 Aug 2023 16:53:39 +0200 >>>>> Nuno Sa <nuno.sa@analog.com> wrote: >>>>> >>>>>> Signed-off-by: Nuno Sa <nuno.sa@analog.com> >>>>> >>>>> Hi Nuno, >>>>> >>>> >>>> Hi Jonathan, >>>> >>>> Thanks for the initial review... >>>> >>>>> >>>>> One general comment is that you could have stripped this back a fair bit >>>>> for ease of understanding. At this stage we don't care about things >>>>> like debug or control of test patterns. Bring those in as extra patches. >>>>> >>>> >>>> Agreed... As I mentioned (I think) in the cover, I made the RFC bigger than >>>> needed to >>>> kind of showcase how we can properly configure the hdl core to support >>>> things >>>> (interface calibration) that were very hard to do with the current >>>> implementation. >>>> I'll make sure to add the minimum needed API to accommodate what we have >>>> right now. >>>> >>>>> I haven't fully gotten my head around the ordering constraints on removal. >>>>> Are there other users of the component framework that have similar >>>>> problems? >>>>> >>>> >>>> My understanding on the component API is that one should do all the tear >>>> down in the >>>> .unbind() callback. As usual, I can see some drivers not really doing that. >>>> >>>>> Also, I don't yet understand how a multiple front end, single backend >>>>> setup >>>>> would work. Or indeed single front end, multiple backend... Maybe we >>>>> don't >>>>> need those cases, but if we want this to be useful beyond adi-axi we >>>>> probably at least want an outline of how they work. >>>>> >>>> >>>> Indeed we can have multiple (and we have it out of tree) backends on one >>>> frontend. >>>> Think on an ADC/DAC with fairly complex data path with more than one >>>> channel/interface (CMOS, LVDS, etc). Typically, in those case, each of the >>>> interface >>>> will be connected to an instance of the hdl core (the backend). >>> >>> That might work out for your case, but not the stm32 one where I think we can >>> end >>> up with interleaved data from two front ends in the same buffer... >>> >> >> Not sure I'm following this one. But wouldn't that be something specific for >> each system (through devicetree)? I haven't tried but I think the same backend >> could be used in different frontend devices (using the component API). That is >> not really a usecase for me but definitely something that could be supported (if >> we need to start doing things like keep enable/disable counters and so on) if it >> is a usecase for stm32. > > If we are going to support both usecases, we just need to figure out what composite > devices with N-M backend - frontend look like and make sure that doesn't > cause problems. I'd expect the separation between backend instances might > reflect data storage on capture but then again that might end up like the many > IIO devices for many buffers mess we had before the multiple buffer support > was added. > The stm32 dfsdm interleaved use case is not a problem as it is possible to associate several backends to a frontend. I did some experiments based on converter framework, and did not identified blocking points regarding dfsdm use cases. Some limitations where discussed in [1], about generic bindings support. The preferred solution was to extend converter_frontend_add_matches() to parse also child nodes. I have added converters_get_from_fwnode() API and adapted converter_frontend_add_matches() to test this approach. With this changes and an additional api to support channel attributes read, the framework fulfills all the needs for dfsdm. So, I feel comfortable to drop my previous "backend framework" proposal, and move to the current proposal. If we go further in converter framework adaption, I will push these updates. [1]: https://lore.kernel.org/lkml/60e913e81ce67192ed0449fe7b718434fd360d97.camel@gmail.com/ BRs Olivier > Might need enable /disable counters as you say - I'm not quite sure without > trying it! >>>>>> + * than having to tear everything down in .frontend_close() and >>>>>> + * .backend_close() >>>>> >>>>> That last bit is nasty and will be non obvious to driver authors. >>>>> >>>>> I wonder if we can come up with some means to make it hard to do. >>>>> >>>> >>>> Yeah, I agree. The alternative is to always bring everything down in >>>> .frontend_close() and .backend_close(). But that can also be prone to subtle >>>> bugs >>>> because it's easy to mess up the ordering when not using devres. >>>> >>>> So, at this point, I cannot really think on a perfect solution rather than >>>> keeping >>>> some rules like (assuming we keep the logic we have now): >>>> >>>> * Using devres on frontend|backend_init() only when .remove() is provided on >>>> the >>>> driver. >>>> * No mixes of devres and .frontend|backend_close() >>>> >>>> But yeah, would be nice if we could come up with something to make it more >>>> obvious to >>>> driver authors. >>> >>>> >>>> We might be able to detect that converter_backend_del() and >>>> converter_frontend_del() >>>> are under devres while no .frontend|backend_close() is being given. I guess >>>> that >>>> could be a valid indicator of likely misusage. >>>> >>>> Or even better (but I'm not sure it's doable with the current devres API), >>>> detecting >>>> that converter_backend_del() or converter_frontend_del() are under devres >>>> while more >>>> resources are also allocated in our specific opened groups. That would >>>> always be a >>>> problem (I think) because the only way for the _del() functions to be under >>>> devres is >>>> if someone added them (from .probe) with devm_add_action() which means that >>>> tearing >>>> down the aggregate will happen after some resources (which were allocated in >>>> the >>>> _init() function) are already freed (as even with new groups, devres will >>>> remove >>>> things on the reverse order). And that would defenitely be problematic. And, >>>> in fact, >>>> is the whole reason why I have the .del() functions on .remove() (so, >>>> tearing down >>>> the aggregate device is the first thing to happen and resources are freed in >>>> the >>>> reverse order they were allocated). >>>> >>> >> >> This would actually be very messy and hard to do properly. Concurrency between >> the aggregate probing (at the component level) and unbinding (at driver core >> level) would be very tricky if doable at all. On top that, we do have >> devres_find() but no way to tell if a devres group has resources or not. It >> would be easy to add one new API but likely not worth it just for this usecase. >> >> I also thought about an helper macro to wrap every devm_ call but your below >> suggestion is way better and transparent to users. >> >>> I couldn't work out how to do anything easily and would need some experiments. >>> Maybe some 'hidden' devres callbacks and a state flag somewhere. If we >>> register >>> that very late we can perhaps detect that we entered devres cleanup before >>> calling >>> expected manual cleanup. I'm thinking have the setup path register a flag >>> checking >>> callback and the cleanup path set a flag (devres now safe). Then we can at >>> least >>> make it scream if we end up doing things in wrong way. >>> >> >> Hmm, that might actually be a good idea and something to try. It likely means >> having a mutex (I was happy not to have one for now :) but likely inevitable >> somewhere down the road) for the flag but it might work. At least we'll be able >> to dump a WARN or something if we suspect something is wrong. > > Careful ordering an atomic might work. > >> >>>> Other thought would be some generic helper macros to use in these type of >>>> drivers so >>>> a .remove() callback is always added to remove the components. >>> I wondered if that could work but it's an ugly macro because needs to deal >>> with >>> different bus types. >>> >>> >> >> We could have a macro per bus type. I'm not really seeing anything different >> than platform, spi and i2c. But even like this, it could easily start to get >> ugly because of accepted parameters (and any deviation would again mean a >> different version per bus type). Another thing that crossed my mind was wrappers >> to module_spi_driver() and friends. Maybe that could work. Anyways, as you said >> in one of your replies, this is in kernel interface that we can easily >> add/change. So, I will defer this to a later point in time. > > Makes sense > > J >> >> - Nuno Sá >> >
On Mon, 2023-11-13 at 18:20 +0100, Olivier MOYSAN wrote: Ho Olivier, > Hi Nuno, Jonathan, > > On 9/4/23 17:31, Jonathan Cameron wrote: > > On Mon, 04 Sep 2023 16:14:17 +0200 > > Nuno Sá <noname.nuno@gmail.com> wrote: > > > > > On Sun, 2023-09-03 at 11:56 +0100, Jonathan Cameron wrote: > > > > On Thu, 31 Aug 2023 11:32:54 +0200 > > > > Nuno Sá <noname.nuno@gmail.com> wrote: > > > > > > > > > On Wed, 2023-08-30 at 18:02 +0100, Jonathan Cameron wrote: > > > > > > On Fri, 4 Aug 2023 16:53:39 +0200 > > > > > > Nuno Sa <nuno.sa@analog.com> wrote: > > > > > > > > > > > > > Signed-off-by: Nuno Sa <nuno.sa@analog.com> > > > > > > > > > > > > Hi Nuno, > > > > > > > > > > > > > > > > Hi Jonathan, > > > > > > > > > > Thanks for the initial review... > > > > > > > > > > > > > > > > > One general comment is that you could have stripped this back a fair > > > > > > bit > > > > > > for ease of understanding. At this stage we don't care about things > > > > > > like debug or control of test patterns. Bring those in as extra > > > > > > patches. > > > > > > > > > > > > > > > > Agreed... As I mentioned (I think) in the cover, I made the RFC bigger > > > > > than > > > > > needed to > > > > > kind of showcase how we can properly configure the hdl core to support > > > > > things > > > > > (interface calibration) that were very hard to do with the current > > > > > implementation. > > > > > I'll make sure to add the minimum needed API to accommodate what we > > > > > have > > > > > right now. > > > > > > > > > > > I haven't fully gotten my head around the ordering constraints on > > > > > > removal. > > > > > > Are there other users of the component framework that have similar > > > > > > problems? > > > > > > > > > > > > > > > > My understanding on the component API is that one should do all the > > > > > tear > > > > > down in the > > > > > .unbind() callback. As usual, I can see some drivers not really doing > > > > > that. > > > > > > > > > > > Also, I don't yet understand how a multiple front end, single > > > > > > backend > > > > > > setup > > > > > > would work. Or indeed single front end, multiple backend... Maybe > > > > > > we > > > > > > don't > > > > > > need those cases, but if we want this to be useful beyond adi-axi we > > > > > > probably at least want an outline of how they work. > > > > > > > > > > > > > > > > Indeed we can have multiple (and we have it out of tree) backends on > > > > > one > > > > > frontend. > > > > > Think on an ADC/DAC with fairly complex data path with more than one > > > > > channel/interface (CMOS, LVDS, etc). Typically, in those case, each of > > > > > the > > > > > interface > > > > > will be connected to an instance of the hdl core (the backend). > > > > > > > > That might work out for your case, but not the stm32 one where I think > > > > we can > > > > end > > > > up with interleaved data from two front ends in the same buffer... > > > > > > > > > > Not sure I'm following this one. But wouldn't that be something specific > > > for > > > each system (through devicetree)? I haven't tried but I think the same > > > backend > > > could be used in different frontend devices (using the component API). > > > That is > > > not really a usecase for me but definitely something that could be > > > supported (if > > > we need to start doing things like keep enable/disable counters and so on) > > > if it > > > is a usecase for stm32. > > > > If we are going to support both usecases, we just need to figure out what > > composite > > devices with N-M backend - frontend look like and make sure that doesn't > > cause problems. I'd expect the separation between backend instances might > > reflect data storage on capture but then again that might end up like the > > many > > IIO devices for many buffers mess we had before the multiple buffer support > > was added. > > > > The stm32 dfsdm interleaved use case is not a problem as it is possible > to associate several backends to a frontend. > I did some experiments based on converter framework, and did not > identified blocking points regarding dfsdm use cases. > > Some limitations where discussed in [1], about generic bindings support. > The preferred solution was to extend converter_frontend_add_matches() to > parse also child nodes. I have added converters_get_from_fwnode() API > and adapted converter_frontend_add_matches() to test this approach. > With this changes and an additional api to support channel attributes > read, the framework fulfills all the needs for dfsdm. > > So, I feel comfortable to drop my previous "backend framework" proposal, > and move to the current proposal. > > If we go further in converter framework adaption, I will push these updates. > I hope you didn't had too much trouble with those patches. The reason I'm saying this is because, after some thought, I'm strongly considering in moving to normal OF/ACPI lookup. 3 mains reasons for it: 1) That "hack/rule" for a driver to provide a .remove() callback (in order to devm_*) is really non obvious and might even be prune to subtle bugs (that I'm not seeing now :)). But my main argument is that it can become hard to maintain it (depending on how much people starts to use the framework). 2) From the discussion we had about the limitations you pointed in your link, I started to realize that it might get harder to scale the framework. Yes, we found a fairly easy way of doing it but still took more code to do it when compared to a typical lookup. 3) This is the most important together with 1). You mentioned something like cascaded designs and I just found an usercase in ADI out of tree drivers. We have a design where we have something like: ------------------------------------------ | FPGA | -------------- | ------------- ------------------- | |DAC Frontend| -> | |DAC Backend| -> |DAC Interpolation| | -------------- | ------------- ------------------- | | | ------------------------------------------ In the above design we kind of have a cascaded thing where the DAC backend is both a frontend and a backend and the Intrerpolation stuff just serves as backend to the DAC core. So, ideally the DAC frontend should not have to know a thing about the interpolation... I realized that having this with the component framework is far from straight because we would need two components/aggregate devices to accomplish this (DAC Front + DAC Back) and (DAC Back + DAC Interp) and I think we would need some extra care regarding one of the components going away (not sure though). One way to make it simple would be to not "respect" the HW configuration and just have one aggregate device with 1 Frontend + 2 Backends and so the frontend would need to "know" about the interpolation core. Again, I think that with OF/ACPI this setup with be fairly straight to get and "respect". Anyways, all the above makes me feel that component might not be the best choice (even though I was eager to use it :)) I'll also get to work on this again and I might just use an industrialio- backend.c file in the base dir as you had in your RFC. From a quick look on your series, I'm no sure how much code I will reuse but we can see later if a Co- authored-by tag makes sense or not. Let me know if there's something you don't agree or if there's any concern on your side. - Nuno Sá > > >
Hi Nuno, On 11/14/23 10:03, Nuno Sá wrote: > On Mon, 2023-11-13 at 18:20 +0100, Olivier MOYSAN wrote: > > Ho Olivier, > >> Hi Nuno, Jonathan, >> >> On 9/4/23 17:31, Jonathan Cameron wrote: >>> On Mon, 04 Sep 2023 16:14:17 +0200 >>> Nuno Sá <noname.nuno@gmail.com> wrote: >>> >>>> On Sun, 2023-09-03 at 11:56 +0100, Jonathan Cameron wrote: >>>>> On Thu, 31 Aug 2023 11:32:54 +0200 >>>>> Nuno Sá <noname.nuno@gmail.com> wrote: >>>>> >>>>>> On Wed, 2023-08-30 at 18:02 +0100, Jonathan Cameron wrote: >>>>>>> On Fri, 4 Aug 2023 16:53:39 +0200 >>>>>>> Nuno Sa <nuno.sa@analog.com> wrote: >>>>>>> >>>>>>>> Signed-off-by: Nuno Sa <nuno.sa@analog.com> >>>>>>> >>>>>>> Hi Nuno, >>>>>>> >>>>>> >>>>>> Hi Jonathan, >>>>>> >>>>>> Thanks for the initial review... >>>>>> >>>>>>> >>>>>>> One general comment is that you could have stripped this back a fair >>>>>>> bit >>>>>>> for ease of understanding. At this stage we don't care about things >>>>>>> like debug or control of test patterns. Bring those in as extra >>>>>>> patches. >>>>>>> >>>>>> >>>>>> Agreed... As I mentioned (I think) in the cover, I made the RFC bigger >>>>>> than >>>>>> needed to >>>>>> kind of showcase how we can properly configure the hdl core to support >>>>>> things >>>>>> (interface calibration) that were very hard to do with the current >>>>>> implementation. >>>>>> I'll make sure to add the minimum needed API to accommodate what we >>>>>> have >>>>>> right now. >>>>>> >>>>>>> I haven't fully gotten my head around the ordering constraints on >>>>>>> removal. >>>>>>> Are there other users of the component framework that have similar >>>>>>> problems? >>>>>>> >>>>>> >>>>>> My understanding on the component API is that one should do all the >>>>>> tear >>>>>> down in the >>>>>> .unbind() callback. As usual, I can see some drivers not really doing >>>>>> that. >>>>>> >>>>>>> Also, I don't yet understand how a multiple front end, single >>>>>>> backend >>>>>>> setup >>>>>>> would work. Or indeed single front end, multiple backend... Maybe >>>>>>> we >>>>>>> don't >>>>>>> need those cases, but if we want this to be useful beyond adi-axi we >>>>>>> probably at least want an outline of how they work. >>>>>>> >>>>>> >>>>>> Indeed we can have multiple (and we have it out of tree) backends on >>>>>> one >>>>>> frontend. >>>>>> Think on an ADC/DAC with fairly complex data path with more than one >>>>>> channel/interface (CMOS, LVDS, etc). Typically, in those case, each of >>>>>> the >>>>>> interface >>>>>> will be connected to an instance of the hdl core (the backend). >>>>> >>>>> That might work out for your case, but not the stm32 one where I think >>>>> we can >>>>> end >>>>> up with interleaved data from two front ends in the same buffer... >>>>> >>>> >>>> Not sure I'm following this one. But wouldn't that be something specific >>>> for >>>> each system (through devicetree)? I haven't tried but I think the same >>>> backend >>>> could be used in different frontend devices (using the component API). >>>> That is >>>> not really a usecase for me but definitely something that could be >>>> supported (if >>>> we need to start doing things like keep enable/disable counters and so on) >>>> if it >>>> is a usecase for stm32. >>> >>> If we are going to support both usecases, we just need to figure out what >>> composite >>> devices with N-M backend - frontend look like and make sure that doesn't >>> cause problems. I'd expect the separation between backend instances might >>> reflect data storage on capture but then again that might end up like the >>> many >>> IIO devices for many buffers mess we had before the multiple buffer support >>> was added. >>> >> >> The stm32 dfsdm interleaved use case is not a problem as it is possible >> to associate several backends to a frontend. >> I did some experiments based on converter framework, and did not >> identified blocking points regarding dfsdm use cases. >> >> Some limitations where discussed in [1], about generic bindings support. >> The preferred solution was to extend converter_frontend_add_matches() to >> parse also child nodes. I have added converters_get_from_fwnode() API >> and adapted converter_frontend_add_matches() to test this approach. >> With this changes and an additional api to support channel attributes >> read, the framework fulfills all the needs for dfsdm. >> >> So, I feel comfortable to drop my previous "backend framework" proposal, >> and move to the current proposal. >> >> If we go further in converter framework adaption, I will push these updates. >> > > I hope you didn't had too much trouble with those patches. The reason I'm saying > this is because, after some thought, I'm strongly considering in moving to > normal OF/ACPI lookup. 3 mains reasons for it: > No problem, this was an opportunity to discover the component framework. The converter API is quite simple to use once you understand the basics of component concepts. It does the job for stm dfsdm, but I agree that for the long-term, a component-based solution is probably less scalable. > 1) That "hack/rule" for a driver to provide a .remove() callback (in order to > devm_*) is really non obvious and might even be prune to subtle bugs (that I'm > not seeing now :)). But my main argument is that it can become hard to maintain > it (depending on how much people starts to use the framework). > > 2) From the discussion we had about the limitations you pointed in your link, I > started to realize that it might get harder to scale the framework. Yes, we > found a fairly easy way of doing it but still took more code to do it when > compared to a typical lookup. > > 3) This is the most important together with 1). You mentioned something like > cascaded designs and I just found an usercase in ADI out of tree drivers. We > have a design where we have something like: > > ------------------------------------------ > | FPGA | > -------------- | ------------- ------------------- | > |DAC Frontend| -> | |DAC Backend| -> |DAC Interpolation| | > -------------- | ------------- ------------------- | > | | > ------------------------------------------ > > In the above design we kind of have a cascaded thing where the DAC backend is > both a frontend and a backend and the Intrerpolation stuff just serves as > backend to the DAC core. So, ideally the DAC frontend should not have to know a > thing about the interpolation... I realized that having this with the component > framework is far from straight because we would need two components/aggregate > devices to accomplish this (DAC Front + DAC Back) and (DAC Back + DAC Interp) > and I think we would need some extra care regarding one of the components going > away (not sure though). One way to make it simple would be to not "respect" the > HW configuration and just have one aggregate device with 1 Frontend + 2 Backends > and so the frontend would need to "know" about the interpolation core. Again, I > think that with OF/ACPI this setup with be fairly straight to get and "respect". > > Anyways, all the above makes me feel that component might not be the best choice > (even though I was eager to use it :)) > > I'll also get to work on this again and I might just use an industrialio- > backend.c file in the base dir as you had in your RFC. From a quick look on your > series, I'm no sure how much code I will reuse but we can see later if a Co- > authored-by tag makes sense or not. > The APIs in my RFC were kept minimalistic to assess the proposal's relevance. As you pointed out, there is room for improvement, particularly with regards to object releasing. If you are on the way to rework your framework, feel free to re-use some parts of this RFC. At the end, what really matters is to converge towards a solution :) Regards Olivier > Let me know if there's something you don't agree or if there's any concern on > your side. > > - Nuno Sá >>>>
diff --git a/drivers/iio/addac/converter.c b/drivers/iio/addac/converter.c new file mode 100644 index 000000000000..31ac704255ad --- /dev/null +++ b/drivers/iio/addac/converter.c @@ -0,0 +1,547 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Framework to handle complex IIO aggregate devices + * + * A note on some of the design expectations with regards to lifetimes and + * devices bringup/removal. + * + * The Framework is using, under the wood, the component API which makes it to + * easy treat a bunch of devices as one aggregate device. This means that the + * complete thing is only brought to live when all the deviced are probed. To do + * this, two callbacks are used that should in fact completely replace .probe() + * and .remove(). The formers should only be used the minimum needed. Ideally, + * only to call the functions to add and remove frontend annd backend devices. + * + * It is advised for frontend and backend drivers to use their .remove() + * callbacks (to use devres API during the frontend and backends initialization). + * See the comment in @converter_frontend_bind(). + * + * It is also assumed that converter objects cannot be accessed once one of the + * devices of the aggregate device is removed (effectively bringing the all the + * devices down). Based on that assumption, these objects are not refcount which + * means accessing them will likely fail miserably. + * + * Copyright (C) 2023 Analog Devices Inc. + */ + +#define dev_fmt(fmt) "Converter - " fmt + +#include <linux/component.h> +#include <linux/debugfs.h> +#include <linux/device.h> +#include <linux/err.h> +#include <linux/iio/addc/converter.h> +#include <linux/iio/iio.h> +#include <linux/list.h> +#include <linux/module.h> +#include <linux/property.h> +#include <linux/regmap.h> + +struct converter_backend { + struct list_head entry; + struct device *dev; + const struct converter_ops *ops; + const char *name; + void *drvdata; + + struct regmap *regmap; + unsigned int cached_reg_addr; +}; + +struct converter_frontend { + struct list_head list; + const struct frontend_ops *ops; + struct device *dev; +}; + +static ssize_t converter_debugfs_read_reg(struct file *file, + char __user *userbuf, + size_t count, loff_t *ppos) +{ + struct converter_backend *conv = file->private_data; + unsigned int val = 0; + char read_buf[20]; + int ret, len; + + ret = regmap_read(conv->regmap, conv->cached_reg_addr, &val); + if (ret) { + dev_err(conv->dev, "%s: read failed\n", __func__); + return ret; + } + + len = scnprintf(read_buf, sizeof(read_buf), "0x%X\n", val); + + return simple_read_from_buffer(userbuf, count, ppos, read_buf, len); +} + +static ssize_t converter_debugfs_write_reg(struct file *file, + const char __user *userbuf, + size_t count, loff_t *ppos) +{ + struct converter_backend *conv = file->private_data; + unsigned int val; + char buf[80]; + ssize_t rc; + int ret; + + rc = simple_write_to_buffer(buf, sizeof(buf), ppos, userbuf, count); + if (rc < 0) + return rc; + + ret = sscanf(buf, "%i %i", &conv->cached_reg_addr, &val); + + switch (ret) { + case 1: + break; + case 2: + ret = regmap_write(conv->regmap, conv->cached_reg_addr, val); + if (ret) { + dev_err(conv->dev, "%s: write failed\n", __func__); + return ret; + } + break; + default: + return -EINVAL; + } + + return count; +} + +static const struct file_operations converter_debugfs_reg_fops = { + .open = simple_open, + .read = converter_debugfs_read_reg, + .write = converter_debugfs_write_reg, +}; + +static void __converter_add_direct_reg_access(struct converter_backend *conv, + struct iio_dev *indio_dev) +{ + struct dentry *d = iio_get_debugfs_dentry(indio_dev); + const char *name = conv->name; + char file_name[64]; + + if (!conv->regmap) + return; + if (!d) + return; + + if (!conv->name) + name = "converter"; + + snprintf(file_name, sizeof(file_name), "%s_direct_reg_access", name); + + debugfs_create_file(file_name, 0644, d, conv, + &converter_debugfs_reg_fops); +} + +void converter_add_direct_reg_access(struct converter_backend *conv, + struct iio_dev *indio_dev) +{ + if (IS_ENABLED(CONFIG_DEBUG_FS)) + __converter_add_direct_reg_access(conv, indio_dev); +} +EXPORT_SYMBOL_NS_GPL(converter_add_direct_reg_access, IIO_CONVERTER); + +static int converter_bind(struct device *dev, struct device *aggregate, + void *data) +{ + struct converter_frontend *frontend = dev_get_drvdata(aggregate); + struct converter_backend *conv = dev_get_drvdata(dev); + int ret; + + ret = conv->ops->backend_init(conv, dev); + if (ret) + return ret; + + list_add_tail(&conv->entry, &frontend->list); + + return 0; +} + +static void converter_unbind(struct device *dev, struct device *aggregate, + void *data) +{ + struct converter_backend *conv = dev_get_drvdata(dev); + + if (conv->ops->backend_close) + conv->ops->backend_close(conv); + + /* after this point the converter should not be used anymore */ + converter_set_drvdata(conv, NULL); +} + +static const struct component_ops converter_component_ops = { + .bind = converter_bind, + .unbind = converter_unbind, +}; + +static int converter_frontend_bind(struct device *dev) +{ + struct converter_frontend *frontend = dev_get_drvdata(dev); + int ret; + + ret = component_bind_all(dev, NULL); + if (ret) + return ret; + /* + * We open a new group so that we can control when resources are + * released and still use device managed (devm_) calls. The expectations + * are that on probe, backend resources are allocated first followed by + * the frontend resources (where registering the IIO device must happen) + * Naturally we want the reverse order on the unbind path and that would + * not be possible without opening our own devres group. + + * Note that the component API also opens it's own devres group when + * calling the .bind() callbacks for both the aggregate device + * (our frontend) and each of the components (our backends). On the + * unbind path, the aggregate .unbind() function is called + * (@converter_frontend_unbind()) which should be responsible to tear + * down all the components (effectively releasing all the resources + * allocated on each component devres group) and only then the aggregate + * devres group is released. Hence, the order we want to maintain for + * releasing resources would not be satisfied because backend resources + * would be freed first. With our own group, we can control when + * releasing the resources and we do it before @component_unbind_all(). + * + * This also relies that internally the component API is releasing each + * of the component's devres group. That is likely not to change, but + * maybe we should not trust it and also open our own groups for backend + * devices?! + * + * Another very important thing to keep in mind is that this is only + * valid if frontend and backend driver's are implementing their + * .remove() callback to call @converter_frontend_del() and + * @converter_backend_del(). Calling those functions from + * devm_add_action* and use devm APIs in .frontend_init() and + * .backend_init() is not going to work. Not perfect but still better + * than having to tear everything down in .frontend_close() and + * .backend_close() + */ + if (!devres_open_group(dev, frontend, GFP_KERNEL)) + return -ENOMEM; + + ret = frontend->ops->frontend_init(frontend, dev); + if (ret) { + devres_release_group(dev, frontend); + return ret; + } + + devres_close_group(dev, NULL); + return 0; +} + +static void converter_frontend_unbind(struct device *dev) +{ + struct converter_frontend *frontend = dev_get_drvdata(dev); + + if (frontend->ops->frontend_close) + frontend->ops->frontend_close(frontend); + + devres_release_group(dev, frontend); + component_unbind_all(dev, NULL); + list_del_init(&frontend->list); +} + +static const struct component_master_ops frontend_component_ops = { + .bind = converter_frontend_bind, + .unbind = converter_frontend_unbind, +}; + +struct converter_backend *converter_get(const struct converter_frontend *frontend, + const char *name) +{ + struct converter_backend *iter, *conv = NULL; + struct device *dev = frontend->dev; + struct fwnode_handle *fwnode; + int index = 0; + + if (list_empty(&frontend->list)) { + dev_err(dev, "Backend list is empty...\n"); + return ERR_PTR(-ENODEV); + } + + /* if no name given, we assume only one converter_backend exists */ + if (!name) + return list_first_entry(&frontend->list, + struct converter_backend, entry); + + index = device_property_match_string(frontend->dev, "converter-names", + name); + if (index < 0) + return ERR_PTR(index); + + fwnode = fwnode_find_reference(dev_fwnode(dev), "converters", index); + if (IS_ERR(fwnode)) + return ERR_CAST(fwnode); + + list_for_each_entry(iter, &frontend->list, entry) { + if (device_match_fwnode(iter->dev, fwnode)) { + conv = iter; + break; + } + } + + fwnode_handle_put(fwnode); + + if (!conv) { + dev_err(dev, "Converter (%s) not found in the list\n", name); + return ERR_PTR(-ENODEV); + } + + /* See if we can add device_property_string_read_index() */ + conv->name = kstrdup_const(name, GFP_KERNEL); + if (!conv->name) + return ERR_PTR(-ENOMEM); + + return conv; +} +EXPORT_SYMBOL_NS_GPL(converter_get, IIO_CONVERTER); + +static int converter_frontend_add_matches(struct converter_frontend *frontend, + struct component_match **match) +{ + struct device *dev = frontend->dev; + struct fwnode_handle *fwnode; + int index = 0; + + do { + fwnode = fwnode_find_reference(dev_fwnode(dev), "converters", + index); + if (IS_ERR(fwnode)) + break; + + component_match_add_release(dev, match, + component_release_fwnode, + component_compare_fwnode, fwnode); + index++; + } while (true); + + /* no devices?! */ + if (!index) { + dev_err(dev, "No converters. Make sure the \"converters\" property is given!\n"); + return -ENODEV; + } + + if (PTR_ERR(fwnode) != -ENOENT) + return PTR_ERR(fwnode); + + return 0; +} + +int converter_test_pattern_set(struct converter_backend *conv, + unsigned int chan, + enum converter_test_pattern pattern) +{ + if (pattern >= CONVERTER_TEST_PATTERN_MAX) + return -EINVAL; + if (!conv->ops->test_pattern_set) + return -ENOTSUPP; + + return conv->ops->test_pattern_set(conv, chan, pattern); +} +EXPORT_SYMBOL_NS_GPL(converter_test_pattern_set, IIO_CONVERTER); + +int converter_chan_status_get(struct converter_backend *conv, + unsigned int chan, + struct converter_chan_status *status) +{ + if (!conv->ops->chan_status) + return -ENOTSUPP; + + return conv->ops->chan_status(conv, chan, status); +} +EXPORT_SYMBOL_NS_GPL(converter_chan_status_get, IIO_CONVERTER); + +int converter_iodelay_set(struct converter_backend *conv, + unsigned int num_lanes, unsigned int delay) +{ + if (!num_lanes) + return -EINVAL; + if (!conv->ops->iodelay_set) + return -ENOTSUPP; + + return conv->ops->iodelay_set(conv, num_lanes, delay); +} +EXPORT_SYMBOL_NS_GPL(converter_iodelay_set, IIO_CONVERTER); + +int converter_data_format_set(struct converter_backend *conv, + unsigned int chan, + const struct converter_data_fmt *data) +{ + if (data->type >= CONVERTER_DATA_TYPE_MAX) + return -EINVAL; + if (!conv->ops->data_format_set) + return -ENOTSUPP; + + return conv->ops->data_format_set(conv, chan, data); +} +EXPORT_SYMBOL_NS_GPL(converter_data_format_set, IIO_CONVERTER); + +int converter_sample_edge_select(struct converter_backend *conv, + enum converter_edge edge) +{ + if (edge >= CONVERTER_EDGE_MAX) + return -EINVAL; + if (conv->ops->sample_edge_select) + return -ENOTSUPP; + + return conv->ops->sample_edge_select(conv, edge); +} +EXPORT_SYMBOL_NS_GPL(converter_sample_edge_select, IIO_CONVERTER); + +int converter_chan_enable(struct converter_backend *conv, unsigned int chan) +{ + if (!conv->ops->chan_enable) + return -ENOTSUPP; + + return conv->ops->chan_enable(conv, chan); +} +EXPORT_SYMBOL_NS_GPL(converter_chan_enable, IIO_CONVERTER); + +int converter_chan_disable(struct converter_backend *conv, unsigned int chan) +{ + if (!conv->ops->disable) + return -ENOTSUPP; + + return conv->ops->chan_disable(conv, chan); +} +EXPORT_SYMBOL_NS_GPL(converter_chan_disable, IIO_CONVERTER); + +int converter_enable(struct converter_backend *conv) +{ + if (!conv->ops->enable) + return -ENOTSUPP; + + return conv->ops->enable(conv); +} +EXPORT_SYMBOL_NS_GPL(converter_enable, IIO_CONVERTER); + +void converter_disable(struct converter_backend *conv) +{ + if (!conv->ops->disable) + return; + + conv->ops->disable(conv); +} +EXPORT_SYMBOL_NS_GPL(converter_disable, IIO_CONVERTER); + +int __converter_test_pattern_xlate(unsigned int pattern, + const struct converter_test_pattern_xlate *xlate, + int n_matches) +{ + unsigned int p = n_matches; + + while (p--) { + if (pattern == xlate[p].pattern) + return xlate[p].reg_val; + } + + return -EINVAL; +} +EXPORT_SYMBOL_NS_GPL(__converter_test_pattern_xlate, IIO_CONVERTER); + +void converter_set_regmap(struct converter_backend *conv, + struct regmap *regmap) +{ + conv->regmap = regmap; +} +EXPORT_SYMBOL_NS_GPL(converter_set_regmap, IIO_CONVERTER); + +void converter_set_drvdata(struct converter_backend *conv, void *drvdata) +{ + conv->drvdata = drvdata; +} +EXPORT_SYMBOL_NS_GPL(converter_set_drvdata, IIO_CONVERTER); + +void *converter_get_drvdata(const struct converter_backend *conv) +{ + WARN_ON(!conv->drvdata); + return conv->drvdata; +} +EXPORT_SYMBOL_NS_GPL(converter_get_drvdata, IIO_CONVERTER); + +void converter_del(struct device *dev) +{ + component_del(dev, &converter_component_ops); +} +EXPORT_SYMBOL_NS_GPL(converter_del, IIO_CONVERTER); + +static void converter_free(void *conv) +{ + struct converter_backend *__conv = conv; + + if (__conv->name) + kfree_const(__conv->name); + + kfree(__conv); +} + +int converter_add(struct device *dev, const struct converter_ops *ops) +{ + struct converter_backend *conv; + int ret; + + if (!ops || !ops->backend_init) + return -EINVAL; + + conv = kzalloc(sizeof(*conv), GFP_KERNEL); + if (!conv) + return -ENOMEM; + + /* + * The expectation is that everything goes up and down in + * .converter_bind() and .converter_unbind() respectively. Hence, it's + * not expected for converter objects to be accessed after unbind(). As + * soon as that does not stand anymore, we need to + * drop devm_add_action_or_reset() and properly refcount the objects. + */ + ret = devm_add_action_or_reset(dev, converter_free, conv); + if (ret) + return ret; + + conv->ops = ops; + dev_set_drvdata(dev, conv); + conv->dev = dev; + + return component_add(dev, &converter_component_ops); +} +EXPORT_SYMBOL_NS_GPL(converter_add, IIO_CONVERTER); + +void converter_frontend_del(struct device *dev) +{ + component_master_del(dev, &frontend_component_ops); +} +EXPORT_SYMBOL_NS_GPL(converter_frontend_del, IIO_CONVERTER); + +int converter_frontend_add(struct device *dev, const struct frontend_ops *ops) +{ + struct converter_frontend *frontend; + struct component_match *match; + int ret; + + if (!ops || !ops->frontend_init) { + dev_err(dev, "Mandatory ops missing\n"); + return -EINVAL; + } + + frontend = devm_kzalloc(dev, sizeof(*frontend), GFP_KERNEL); + if (!frontend) + return -ENOMEM; + + frontend->ops = ops; + frontend->dev = dev; + INIT_LIST_HEAD(&frontend->list); + dev_set_drvdata(dev, frontend); + + ret = converter_frontend_add_matches(frontend, &match); + if (ret) + return ret; + + return component_master_add_with_match(dev, &frontend_component_ops, + match); +} +EXPORT_SYMBOL_NS_GPL(converter_frontend_add, IIO_CONVERTER); + +MODULE_AUTHOR("Nuno Sa <nuno.sa@analog.com>"); +MODULE_DESCRIPTION("Framework to handle complex IIO aggregate devices"); +MODULE_LICENSE("GPL v2"); diff --git a/include/linux/iio/addac/converter.h b/include/linux/iio/addac/converter.h new file mode 100644 index 000000000000..09d9d491b2b8 --- /dev/null +++ b/include/linux/iio/addac/converter.h @@ -0,0 +1,485 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +#ifndef _CONVERTER_H +#define _CONVERTER_H + +struct converter_frontend; +struct converter_backend; +struct iio_dev; +struct device; +struct regmap; + +enum converter_test_pattern { + CONVERTER_PRBS_7, + CONVERTER_PRBS_15, + CONVERTER_PRBS_23, + CONVERTER_PRBS_31, + CONVERTER_RAMP_NIBBLE, + CONVERTER_RAMP_16, + /* vendor specific from 32 */ + CONVERTER_ADI_PRBS_9A = 32, + CONVERTER_ADI_PRBS_23A, + CONVERTER_ADI_PRBS_X, + CONVERTER_TEST_PATTERN_MAX +}; + +enum converter_data_type { + CONVERTER_TWOS_COMPLEMENT, + CONVERTER_OFFSET_BINARY, + CONVERTER_DATA_TYPE_MAX +}; + +enum converter_edge { + CONVERTER_RISING_EDGE_SAMPLE, + CONVERTER_FALLING_EDGE_SAMPLE, + CONVERTER_EDGE_MAX +}; + +struct converter_chan_status { + bool errors; +}; + +/** + * struct converter_data_fmt - Backend data format + * @type: Data type. + * @sign_extend: Bool to tell if the data is sign extended. + * @enable: Enable/Disable the data format module. If disabled, + * not formatting will happen. + */ +struct converter_data_fmt { + enum converter_data_type type; + bool sign_extend; + bool enable; +}; + +/** + * struct converter_test_pattern_xlate - Helper struct for test pattern handling + * @pattern: Pattern to configure. + * @reg_val: Register value for the pattern to configure. + */ +struct converter_test_pattern_xlate { + enum converter_test_pattern pattern; + unsigned int reg_val; +}; + +/** + * struct converter_ops - Backend supported operations + * @backend_init: Mandatory function to initialize the backend device. It + * should be a replacement for .probe() where the latest + * should only have to care about doing @converter_add(). + * @backend_close: Optional function to tear down the device. + * @enable: Enable the backend device. + * @disable: Disable the backend device. + * @data_format_set: Configure the data format for a specific channel. + * @chan_enable: Enable one channel. + * @chan_disable: Disable one channel. + * @iodelay_set: Controls the IO delay for all the lanes at the interface + * (where data is actually transferred between frontend and + backend) level. + * @test_pattern_set: Set's a test pattern to be transmitted/received by the + * backend. Typically useful for debug or interface + * purposes calibration. + */ +struct converter_ops { + int (*backend_init)(struct converter_backend *conv, struct device *dev); + void (*backend_close)(struct converter_backend *conv); + int (*enable)(struct converter_backend *conv); + void (*disable)(struct converter_backend *conv); + int (*data_format_set)(struct converter_backend *conv, + unsigned int chan, + const struct converter_data_fmt *data); + int (*chan_enable)(struct converter_backend *conv, unsigned int chan); + int (*chan_disable)(struct converter_backend *conv, unsigned int chan); + int (*iodelay_set)(struct converter_backend *conv, + unsigned int num_lanes, unsigned int delay); + int (*test_pattern_set)(struct converter_backend *conv, + unsigned int chan, + enum converter_test_pattern pattern); + int (*chan_status)(struct converter_backend *conv, unsigned int chan, + struct converter_chan_status *status); + int (*sample_edge_select)(struct converter_backend *conv, + enum converter_edge edge); +}; + +/** + * struct frontend_ops - Frontend supported operations + * @frontend_init: Mandatory function to initialize the frontend device. It + * should be a replacement for .probe() where the latest + * should only have to care about doing @frontend_add(). + * @frontend_close: Optional function to tear down the device. + */ +struct frontend_ops { + int (*frontend_init)(struct converter_frontend *frontend, + struct device *dev); + void (*frontend_close)(struct converter_frontend *frontend); +}; + +/** + * converter_test_pattern_xlate() - Helper macro for translatting test patterns + * @pattern: Pattern to translate. + * @xlate: List of @struct converter_test_pattern_xlate pairs. + * + * Simple helper to match a supported pattern and get the register value. Should + * only be called by backend devices. Automatically computes the number of + * @xlate entries. + */ +#define converter_test_pattern_xlate(pattern, xlate) \ + __converter_test_pattern_xlate(pattern, xlate, ARRAY_SIZE(xlate)); + +#if IS_ENABLED(CONFIG_IIO_CONVERTER) + +/** + * converter_get_drvdata - Get driver private data + * @conv: Converter device. + */ +void *converter_get_drvdata(const struct converter_backend *conv); + +/** + * converter_set_drvdata - Set driver private data + * @conv: Converter device. + * @drvdata: Driver private data. + */ +void converter_set_drvdata(struct converter_backend *conv, void *drvdata); + +/** + * converter_set_regmap - Add a regmap object to a converter + * @conv: Converter device. + * @regmap: Regmap object. + */ +void converter_set_regmap(struct converter_backend *conv, + struct regmap *regmap); + +/** + * __converter_test_pattern_xlate - Helper macro for translatting test patterns + * @pattern: Pattern to translate. + * @xlate: List of @struct converter_test_pattern_xlate pairs. + * @n_matches: Number of entries in @xlate. + * + * Simple helper to match a supported pattern and get the register value. Should + * only be called by backend devices. + */ +int __converter_test_pattern_xlate(unsigned int pattern, + const struct converter_test_pattern_xlate *xlate, + int n_matches); + +/** + * + */ +int converter_add(struct device *dev, const struct converter_ops *ops); + +/** + * converter_del - Remove the converter device + * @dev: device to remove from the aggregate + * + * Removes the converter from the aggregate device. This tears down the frontend + * and all the converters. + * + * Ideally, this should be called from the backend driver .remove() callback. + * This means that all the converters (and the frontend) will be tear down before + * running any specific devres cleanup (at the driver core level). What this all + * means is that we can use devm_ apis in @backend_init() and being sure those + * resources will be released after the backend resources and before any devm_* + * used in @probe(). If that is not the case, one should likely not use any + * devm_ API in @backend_init(). That means .backend_close() should be + * provided to do all the necessary cleanups. + */ +void converter_del(struct device *dev); + +/** + * converter_enable - Enable the device + * @conv: Converter device. + * + * Enables the backend device. + * + * RETURNS: + * 0 on success, negative error number on failure. + */ +int converter_enable(struct converter_backend *conv); + +/** + * converter_disable - Disable the device + * @conv: Converter device. + * + * Disables the backend device. + */ +void converter_disable(struct converter_backend *conv); + +/** + * converter_test_pattern_set - Set a test pattern + * @conv: Converter device. + * @chan: Channel number. + * @pattern: Pattern to set. + * + * Set's a test pattern to be transmitted/received by the backend. Typically + * useful for debug or interface calibration purposes. A backend driver can + * call the @converter_test_pattern_xlate() helper to validate the pattern + * (given an array of @struct converter_test_pattern_xlate). + * + * Note that some patterns might be frontend specific. I.e, as far as the + * backend is concerned the pattern is valid (from a register point of view) but + * the actual support for the pattern is not implemented in the device for this + * specific frontend. It's up to the frontend to ask for a proper pattern + * (as it should know better). + * + * RETURNS: + * 0 on success, negative error number on failure. + */ +int converter_test_pattern_set(struct converter_backend *conv, + unsigned int chan, + enum converter_test_pattern pattern); + +int converter_chan_status_get(struct converter_backend *conv, + unsigned int chan, + struct converter_chan_status *status); + +/** + * converter_data_format_set - Configure the data format + * @conv: Converter device. + * @chan: Channel number. + * @data: Data format. + * + * Properly configure a channel with respect to the expected data format. A + * @struct converter_data_fmt must be passed with the settings. + * + * RETURNS: + * 0 on success, negative error number on failure. + */ +int converter_data_format_set(struct converter_backend *conv, + unsigned int chan, + const struct converter_data_fmt *data); + +int converter_sample_edge_select(struct converter_backend *conv, + enum converter_edge edge); + +static inline int +converter_sample_on_falling_edge(struct converter_backend *conv) +{ + return converter_sample_edge_select(conv, CONVERTER_RISING_EDGE_SAMPLE); +} + +static inline int +converter_sample_on_rising_edge(struct converter_backend *conv) +{ + return converter_sample_edge_select(conv, CONVERTER_FALLING_EDGE_SAMPLE); +} + +/** + * converter_chan_enable - Enable a backend channel + * @conv: Converter device. + * @chan: Channel number. + * + * Enables a channel on the backend device. + * + * RETURNS: + * 0 on success, negative error number on failure. + */ +int converter_chan_enable(struct converter_backend *conv, unsigned int chan); + +/** + * converter_chan_disable - Disable a backend channel + * @conv: Converter device. + * @chan: Channel number. + * + * Disables a channel on the backend device. + * + * RETURNS: + * 0 on success, negative error number on failure. + */ +int converter_chan_disable(struct converter_backend *conv, unsigned int chan); + +/** + * converter_iodelay_set - Set's the backend data interface IO delay + * @conv: Converter device. + * @num_lanes: Number of lanes in the data interface. + * @delay: Delay to set. + * + * Controls the IO delay for all the lanes at the data interface (where data is + * actually transferred between frontend and backend) level. + * + * RETURNS: + * 0 on success, negative error number on failure. + */ +int converter_iodelay_set(struct converter_backend *conv, + unsigned int num_lanes, unsigned int delay); + +/** + * converter_frontend_del - Remove the frontend device + * @dev: Device to remove from the aggregate + * + * Removes the frontend from the aggregate device. This tears down the frontend + * and all the converters. + * + * Ideally, this should be called from the frontend driver .remove() callback. + * This means that all the converters (and the frontend) will be tear down + * before running any specific devres cleanup (at the driver core level). What + * this all means is that we can use devm_ apis in .frontend_init() and being + * sure those resources will be released after the backend resources and before + * any devm_* used in .probe(). If that is not the case, one should likely not + * use any devm_ API in .frontend_init(). That means .frontend_close() should be + * provided to do all the necessary cleanups. + */ +void converter_frontend_del(struct device *dev); + +/** + * converter_frontend_add - Allocate and add a frontend device + * @dev: Device to allocate frontend for. + * @ops: Frontend callbacks. + * + * This allocates the frontend device and looks for all converters needed + * so that, when they are available, all of the devices in the aggregate can be + * initialized. + * + * RETURNS: + * 0 on success, negative error number on failure. + */ +int converter_frontend_add(struct device *dev, const struct frontend_ops *ops); + +/** + * converter_get - Get a converter object + * @frontend: Frontend device. + * @name: Converter name. + * + * Get's a pointer to a converter device. If name is NULL, then it is assumed + * that only one backend device is bond with the frontend and the first element + * in the list is retrieved. Should only be called from the .frontend_init() + * callback. + * + * RETURNS: + * A converter pointer, negative error pointer otherwise. + */ +struct converter_backend *__must_check +converter_get(const struct converter_frontend *frontend, const char *name); + +/** + * converter_add_direct_reg_access - Add debugfs direct register access + * @conv: Coverter device + * @indio_dev: IIO device + * + * This is analogous to the typical IIO direct register access in debugfs. The + * extra converter file will be added in the same debugs dir as @indio_dev. + * Moreover, if @conv->name is NULL, the file will be called + * converter_direct_reg_access. Otherwise, will be + * @conv->name_converter_direct_reg_access. + */ +void converter_add_direct_reg_access(struct converter_backend *conv, + struct iio_dev *indio_dev); + +#else + +static inline void *converter_get_drvdata(const struct converter_backend *conv) +{ + WARN_ONCE(1, "converter API is disabled"); + return NULL; +} + +static inline void converter_set_drvdata(struct converter_backend *conv, + void *drvdata) +{ + WARN_ONCE(1, "converter API is disabled"); +} + +static inline void converter_set_regmap(struct converter_backend *conv, + struct regmap *regmap) +{ + WARN_ONCE(1, "converter API is disabled"); +} + +static inline int +__converter_test_pattern_xlate(unsigned int pattern, + const struct converter_test_pattern_xlate *xlate, + int n_matches) +{ + WARN_ONCE(1, "converter API is disabled"); + return -ENOTSUPP; +} + +static inline struct converter_backend *__must_check +converter_get(const struct converter_frontend *frontend, const char *name) +{ + WARN_ONCE(1, "converter API is disabled"); + return ERR_PTR(-ENOTSUPP); +} + +static inline int converter_add(struct device *dev, + const struct converter_ops *ops) +{ + WARN_ONCE(1, "converter API is disabled"); + return -ENOTSUPP; +} + +static inline void converter_del(struct device *dev) +{ + WARN_ONCE(1, "converter API is disabled"); +} + +static inline int converter_enable(struct converter_backend *conv) +{ + WARN_ONCE(1, "converter API is disabled"); + return -ENOTSUPP; +} + +static inline void converter_disable(struct converter_backend *conv) +{ + WARN_ONCE(1, "converter API is disabled"); +} + +static inline int +converter_test_pattern_set(struct converter_backend *conv, + unsigned int chan, + enum converter_test_pattern pattern) +{ + WARN_ONCE(1, "converter API is disabled"); + return -ENOTSUPP; +} + +static inline int +converter_data_format_set(struct converter_backend *conv, + unsigned int chan, + const struct converter_data_fmt *data) +{ + WARN_ONCE(1, "converter API is disabled"); + return -ENOTSUPP; +} + +static inline int converter_chan_enable(struct converter_backend *conv, + unsigned int chan) +{ + WARN_ONCE(1, "converter API is disabled"); + return -ENOTSUPP; +} + +static inline int converter_chan_disable(struct converter_backend *conv, + unsigned int chan) +{ + WARN_ONCE(1, "converter API is disabled"); + return -ENOTSUPP; +} + +static inline int converter_iodelay_set(struct converter_backend *conv, + unsigned int num_lanes, + unsigned int val) +{ + WARN_ONCE(1, "converter API is disabled"); + return -ENOTSUPP; +} + +static inline void +converter_add_direct_reg_access(struct converter_backend *conv, + struct iio_dev *indio_dev) +{ + WARN_ONCE(1, "converter API is disabled"); +} + +static inline int converter_frontend_add(struct device *dev, + const struct frontend_ops *ops) +{ + WARN_ONCE(1, "converter API is disabled"); + return -ENOTSUPP; +} + +static inline void converter_frontend_del(struct device *dev) +{ + WARN_ONCE(1, "converter API is disabled"); +} + +#endif +#endif
Signed-off-by: Nuno Sa <nuno.sa@analog.com> --- drivers/iio/addac/converter.c | 547 ++++++++++++++++++++++++++++ include/linux/iio/addac/converter.h | 485 ++++++++++++++++++++++++ 2 files changed, 1032 insertions(+) create mode 100644 drivers/iio/addac/converter.c create mode 100644 include/linux/iio/addac/converter.h