Mirrors/u-boot
2
0
Fork 0
mirror of https://github.com/AsahiLinux/u-boot synced 2024-12-23 03:23:47 +00:00
Code Issues Projects Releases Packages Wiki Activity
u-boot/drivers/clk/clk-uclass.c

749 lines
15 KiB
C
Raw Normal View History

SPDX: Convert all of our single license tags to Linux Kernel style When U-Boot started using SPDX tags we were among the early adopters and there weren't a lot of other examples to borrow from. So we picked the area of the file that usually had a full license text and replaced it with an appropriate SPDX-License-Identifier: entry. Since then, the Linux Kernel has adopted SPDX tags and they place it as the very first line in a file (except where shebangs are used, then it's second line) and with slightly different comment styles than us. In part due to community overlap, in part due to better tag visibility and in part for other minor reasons, switch over to that style. This commit changes all instances where we have a single declared license in the tag as both the before and after are identical in tag contents. There's also a few places where I found we did not have a tag and have introduced one. Signed-off-by: Tom Rini <trini@konsulko.com>
2018-05-06 21:58:06 +00:00
// SPDX-License-Identifier: GPL-2.0+
dm: Add a clock uclass Clocks are an important feature of platforms and have become increasing complex with time. Most modern SoCs have multiple PLLs and dozens of clock dividers which distribute clocks to on-chip peripherals. Some SoC implementations have a clock API which is private to that SoC family, e.g. Tegra and Exynos. This is useful but it would be better to have a common API that can be understood and used throughout U-Boot. Add a simple clock API as a starting point. It supports querying and setting the rate of a clock. Each clock is a device. To reduce memory and processing overhead the concept of peripheral clocks is provided. These do not need to be explicit devices - it is possible to write a driver that can adjust the I2C clock (for example) without an explicit I2C clock device. This can dramatically reduce the number of devices (and associated overhead) in a complex SoC. Clocks are referenced by a number, and it is expected that SoCs will define that numbering themselves via an enum. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-06-23 21:39:15 +00:00
/*
* Copyright (C) 2015 Google, Inc
* Written by Simon Glass <sjg@chromium.org>
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
* Copyright (c) 2016, NVIDIA CORPORATION.
clk: implement clk_set_defaults() Linux uses the properties 'assigned-clocks', 'assigned-clock-parents' and 'assigned-clock-rates' to configure the clock subsystem for use with various peripheral nodes. This implements clk_set_defaults() and hooks it up with the general device probibin in drivers/core/device.c: when a new device is probed, clk_set_defaults() will be called for it and will process the properties mentioned above. Note that this functionality is designed to fail gracefully (i.e. if a clock-driver does not implement set_parent(), we simply accept this and ignore the error) as not to break existing board-support. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address. Series-version: 2 Cover-letter: clk: support assigned-clock, assigned-clock-parents, assigned-clock-rates For various peripherals on Rockchip SoCs (e.g. for the Ethernet GMAC), the parent-clock needs to be set via the DTS. This adds the required plumbing and implements the GMAC case for the RK3399. END
2018-01-08 12:59:18 +00:00
* Copyright (c) 2018, Theobroma Systems Design und Consulting GmbH
dm: Add a clock uclass Clocks are an important feature of platforms and have become increasing complex with time. Most modern SoCs have multiple PLLs and dozens of clock dividers which distribute clocks to on-chip peripherals. Some SoC implementations have a clock API which is private to that SoC family, e.g. Tegra and Exynos. This is useful but it would be better to have a common API that can be understood and used throughout U-Boot. Add a simple clock API as a starting point. It supports querying and setting the rate of a clock. Each clock is a device. To reduce memory and processing overhead the concept of peripheral clocks is provided. These do not need to be explicit devices - it is possible to write a driver that can adjust the I2C clock (for example) without an explicit I2C clock device. This can dramatically reduce the number of devices (and associated overhead) in a complex SoC. Clocks are referenced by a number, and it is expected that SoCs will define that numbering themselves via an enum. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-06-23 21:39:15 +00:00
*/
#include <common.h>
#include <clk.h>
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
#include <clk-uclass.h>
dm: Add a clock uclass Clocks are an important feature of platforms and have become increasing complex with time. Most modern SoCs have multiple PLLs and dozens of clock dividers which distribute clocks to on-chip peripherals. Some SoC implementations have a clock API which is private to that SoC family, e.g. Tegra and Exynos. This is useful but it would be better to have a common API that can be understood and used throughout U-Boot. Add a simple clock API as a starting point. It supports querying and setting the rate of a clock. Each clock is a device. To reduce memory and processing overhead the concept of peripheral clocks is provided. These do not need to be explicit devices - it is possible to write a driver that can adjust the I2C clock (for example) without an explicit I2C clock device. This can dramatically reduce the number of devices (and associated overhead) in a complex SoC. Clocks are referenced by a number, and it is expected that SoCs will define that numbering themselves via an enum. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-06-23 21:39:15 +00:00
#include <dm.h>
dm: clk: Add support for of-platdata Add support for this feature in the core clock code. Signed-off-by: Simon Glass <sjg@chromium.org>
2016-07-04 17:58:03 +00:00
#include <dt-structs.h>
dm: Add a clock uclass Clocks are an important feature of platforms and have become increasing complex with time. Most modern SoCs have multiple PLLs and dozens of clock dividers which distribute clocks to on-chip peripherals. Some SoC implementations have a clock API which is private to that SoC family, e.g. Tegra and Exynos. This is useful but it would be better to have a common API that can be understood and used throughout U-Boot. Add a simple clock API as a starting point. It supports querying and setting the rate of a clock. Each clock is a device. To reduce memory and processing overhead the concept of peripheral clocks is provided. These do not need to be explicit devices - it is possible to write a driver that can adjust the I2C clock (for example) without an explicit I2C clock device. This can dramatically reduce the number of devices (and associated overhead) in a complex SoC. Clocks are referenced by a number, and it is expected that SoCs will define that numbering themselves via an enum. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-06-23 21:39:15 +00:00
#include <errno.h>
common: Drop log.h from common header Move this header out of the common header. Signed-off-by: Simon Glass <sjg@chromium.org>
2020-05-10 17:40:05 +00:00
#include <log.h>
dm: core: Create a new header file for 'compat' features At present dm/device.h includes the linux-compatible features. This requires including linux/compat.h which in turn includes a lot of headers. One of these is malloc.h which we thus end up including in every file in U-Boot. Apart from the inefficiency of this, it is problematic for sandbox which needs to use the system malloc() in some files. Move the compatibility features into a separate header file. Signed-off-by: Simon Glass <sjg@chromium.org>
2020-02-03 14:36:16 +00:00
#include <malloc.h>
dm: core: Require users of devres to include the header At present devres.h is included in all files that include dm.h but few make use of it. Also this pulls in linux/compat which adds several more headers. Drop the automatic inclusion and require files to include devres themselves. This provides a good indication of which files use devres. Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Anatolij Gustschin <agust@denx.de>
2020-02-03 14:36:15 +00:00
#include <dm/devres.h>
#include <dm/read.h>
common: Drop linux/bug.h from common header Move this uncommon header out of the common header. Signed-off-by: Simon Glass <sjg@chromium.org>
2020-05-10 17:40:08 +00:00
#include <linux/bug.h>
dm: clk: Define clk_get_parent() for clk operations This commit adds the clk_get_parent() function, which is responsible for getting the parent's struct clock pointer. U-Boot's DM support for getting parent is different (the parent relationship is in udevice) than the one in Common Clock Framework [CCF] in Linux. To obtain the pointer to struct clk of parent the pdev->uclass_priv field is read via dev_get_clk_ptr() wrapper. Signed-off-by: Lukasz Majewski <lukma@denx.de> Reviewed-by: Peng Fan <peng.fan@nxp.com>
2019-06-24 13:50:42 +00:00
#include <linux/clk-provider.h>
dm: core: Require users of devres to include the header At present devres.h is included in all files that include dm.h but few make use of it. Also this pulls in linux/compat which adds several more headers. Drop the automatic inclusion and require files to include devres themselves. This provides a good indication of which files use devres. Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Anatolij Gustschin <agust@denx.de>
2020-02-03 14:36:15 +00:00
#include <linux/err.h>
dm: Add a clock uclass Clocks are an important feature of platforms and have become increasing complex with time. Most modern SoCs have multiple PLLs and dozens of clock dividers which distribute clocks to on-chip peripherals. Some SoC implementations have a clock API which is private to that SoC family, e.g. Tegra and Exynos. This is useful but it would be better to have a common API that can be understood and used throughout U-Boot. Add a simple clock API as a starting point. It supports querying and setting the rate of a clock. Each clock is a device. To reduce memory and processing overhead the concept of peripheral clocks is provided. These do not need to be explicit devices - it is possible to write a driver that can adjust the I2C clock (for example) without an explicit I2C clock device. This can dramatically reduce the number of devices (and associated overhead) in a complex SoC. Clocks are referenced by a number, and it is expected that SoCs will define that numbering themselves via an enum. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-06-23 21:39:15 +00:00
clk: clk-uclass: Fix style violations checkpatch.pl complains that the clk_ops structures used in clk-uclass.c ought to be const, so we mark them as const. Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Mario Six <mario.six@gdsys.cc>
2018-01-15 10:06:51 +00:00
static inline const struct clk_ops *clk_dev_ops(struct udevice *dev)
dm: Add a clock uclass Clocks are an important feature of platforms and have become increasing complex with time. Most modern SoCs have multiple PLLs and dozens of clock dividers which distribute clocks to on-chip peripherals. Some SoC implementations have a clock API which is private to that SoC family, e.g. Tegra and Exynos. This is useful but it would be better to have a common API that can be understood and used throughout U-Boot. Add a simple clock API as a starting point. It supports querying and setting the rate of a clock. Each clock is a device. To reduce memory and processing overhead the concept of peripheral clocks is provided. These do not need to be explicit devices - it is possible to write a driver that can adjust the I2C clock (for example) without an explicit I2C clock device. This can dramatically reduce the number of devices (and associated overhead) in a complex SoC. Clocks are referenced by a number, and it is expected that SoCs will define that numbering themselves via an enum. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-06-23 21:39:15 +00:00
{
clk: clk-uclass: Fix style violations checkpatch.pl complains that the clk_ops structures used in clk-uclass.c ought to be const, so we mark them as const. Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Mario Six <mario.six@gdsys.cc>
2018-01-15 10:06:51 +00:00
return (const struct clk_ops *)dev->driver->ops;
dm: Add a clock uclass Clocks are an important feature of platforms and have become increasing complex with time. Most modern SoCs have multiple PLLs and dozens of clock dividers which distribute clocks to on-chip peripherals. Some SoC implementations have a clock API which is private to that SoC family, e.g. Tegra and Exynos. This is useful but it would be better to have a common API that can be understood and used throughout U-Boot. Add a simple clock API as a starting point. It supports querying and setting the rate of a clock. Each clock is a device. To reduce memory and processing overhead the concept of peripheral clocks is provided. These do not need to be explicit devices - it is possible to write a driver that can adjust the I2C clock (for example) without an explicit I2C clock device. This can dramatically reduce the number of devices (and associated overhead) in a complex SoC. Clocks are referenced by a number, and it is expected that SoCs will define that numbering themselves via an enum. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-06-23 21:39:15 +00:00
}
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
#if CONFIG_IS_ENABLED(OF_CONTROL)
dm: clk: Add support for of-platdata Add support for this feature in the core clock code. Signed-off-by: Simon Glass <sjg@chromium.org>
2016-07-04 17:58:03 +00:00
# if CONFIG_IS_ENABLED(OF_PLATDATA)
int clk_get_by_index_platdata(struct udevice *dev, int index,
dtoc: Rename the phandle struct Rather than naming the phandle struct according to the number of cells it uses (e.g. struct phandle_2_cell) name it according to the number of arguments it has (e.g. struct phandle_1_arg). This is a more intuitive naming. Signed-off-by: Simon Glass <sjg@chromium.org> Tested-by: Kever Yang <kever.yang@rock-chips.com>
2017-08-29 20:15:56 +00:00
struct phandle_1_arg *cells, struct clk *clk)
dm: clk: Add support for of-platdata Add support for this feature in the core clock code. Signed-off-by: Simon Glass <sjg@chromium.org>
2016-07-04 17:58:03 +00:00
{
int ret;
if (index != 0)
return -ENOSYS;
ret = uclass_get_device(UCLASS_CLK, 0, &clk->dev);
if (ret)
return ret;
dtoc: Put phandle args in an array We want to support more than one phandle argument. It makes sense to use an array for this rather than discrete struct members. Adjust the code to support this. Rename the member to 'arg' instead of 'id'. Signed-off-by: Simon Glass <sjg@chromium.org> Tested-by: Kever Yang <kever.yang@rock-chips.com>
2017-08-29 20:15:58 +00:00
clk->id = cells[0].arg[0];
dm: clk: Add support for of-platdata Add support for this feature in the core clock code. Signed-off-by: Simon Glass <sjg@chromium.org>
2016-07-04 17:58:03 +00:00
return 0;
}
# else
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
static int clk_of_xlate_default(struct clk *clk,
clk: Modify xlate() method for livetree Update the xlate() method to use ofnode_phandle_args instead of the fdtdec variant. This will allow drivers to support a live device tree. Signed-off-by: Simon Glass <sjg@chromium.org>
2017-05-19 02:09:40 +00:00
struct ofnode_phandle_args *args)
dm: Add a clock uclass Clocks are an important feature of platforms and have become increasing complex with time. Most modern SoCs have multiple PLLs and dozens of clock dividers which distribute clocks to on-chip peripherals. Some SoC implementations have a clock API which is private to that SoC family, e.g. Tegra and Exynos. This is useful but it would be better to have a common API that can be understood and used throughout U-Boot. Add a simple clock API as a starting point. It supports querying and setting the rate of a clock. Each clock is a device. To reduce memory and processing overhead the concept of peripheral clocks is provided. These do not need to be explicit devices - it is possible to write a driver that can adjust the I2C clock (for example) without an explicit I2C clock device. This can dramatically reduce the number of devices (and associated overhead) in a complex SoC. Clocks are referenced by a number, and it is expected that SoCs will define that numbering themselves via an enum. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-06-23 21:39:15 +00:00
{
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
debug("%s(clk=%p)\n", __func__, clk);
dm: Add a clock uclass Clocks are an important feature of platforms and have become increasing complex with time. Most modern SoCs have multiple PLLs and dozens of clock dividers which distribute clocks to on-chip peripherals. Some SoC implementations have a clock API which is private to that SoC family, e.g. Tegra and Exynos. This is useful but it would be better to have a common API that can be understood and used throughout U-Boot. Add a simple clock API as a starting point. It supports querying and setting the rate of a clock. Each clock is a device. To reduce memory and processing overhead the concept of peripheral clocks is provided. These do not need to be explicit devices - it is possible to write a driver that can adjust the I2C clock (for example) without an explicit I2C clock device. This can dramatically reduce the number of devices (and associated overhead) in a complex SoC. Clocks are referenced by a number, and it is expected that SoCs will define that numbering themselves via an enum. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-06-23 21:39:15 +00:00
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
if (args->args_count > 1) {
debug("Invaild args_count: %d\n", args->args_count);
return -EINVAL;
}
dm: Add a clock uclass Clocks are an important feature of platforms and have become increasing complex with time. Most modern SoCs have multiple PLLs and dozens of clock dividers which distribute clocks to on-chip peripherals. Some SoC implementations have a clock API which is private to that SoC family, e.g. Tegra and Exynos. This is useful but it would be better to have a common API that can be understood and used throughout U-Boot. Add a simple clock API as a starting point. It supports querying and setting the rate of a clock. Each clock is a device. To reduce memory and processing overhead the concept of peripheral clocks is provided. These do not need to be explicit devices - it is possible to write a driver that can adjust the I2C clock (for example) without an explicit I2C clock device. This can dramatically reduce the number of devices (and associated overhead) in a complex SoC. Clocks are referenced by a number, and it is expected that SoCs will define that numbering themselves via an enum. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-06-23 21:39:15 +00:00
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
if (args->args_count)
clk->id = args->args[0];
else
clk->id = 0;
dm: Add a clock uclass Clocks are an important feature of platforms and have become increasing complex with time. Most modern SoCs have multiple PLLs and dozens of clock dividers which distribute clocks to on-chip peripherals. Some SoC implementations have a clock API which is private to that SoC family, e.g. Tegra and Exynos. This is useful but it would be better to have a common API that can be understood and used throughout U-Boot. Add a simple clock API as a starting point. It supports querying and setting the rate of a clock. Each clock is a device. To reduce memory and processing overhead the concept of peripheral clocks is provided. These do not need to be explicit devices - it is possible to write a driver that can adjust the I2C clock (for example) without an explicit I2C clock device. This can dramatically reduce the number of devices (and associated overhead) in a complex SoC. Clocks are referenced by a number, and it is expected that SoCs will define that numbering themselves via an enum. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-06-23 21:39:15 +00:00
clk: initialize clk->data when using default xlate Right now when using clk_of_xlate_default(), clk->data remains un-initialized because clk_get_bulk() does not initialize memory on allocation of clock structure. This can cause problems when data is used to match if two clocks pointers are exactly the same underlying clocks, for example. Fix it by initializing clk->data to 0. Suggested-by: Lokesh Vutla <lokeshvutla@ti.com> Signed-off-by: Sekhar Nori <nsekhar@ti.com>
2019-07-11 09:00:24 +00:00
clk->data = 0;
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
return 0;
dm: Add a clock uclass Clocks are an important feature of platforms and have become increasing complex with time. Most modern SoCs have multiple PLLs and dozens of clock dividers which distribute clocks to on-chip peripherals. Some SoC implementations have a clock API which is private to that SoC family, e.g. Tegra and Exynos. This is useful but it would be better to have a common API that can be understood and used throughout U-Boot. Add a simple clock API as a starting point. It supports querying and setting the rate of a clock. Each clock is a device. To reduce memory and processing overhead the concept of peripheral clocks is provided. These do not need to be explicit devices - it is possible to write a driver that can adjust the I2C clock (for example) without an explicit I2C clock device. This can dramatically reduce the number of devices (and associated overhead) in a complex SoC. Clocks are referenced by a number, and it is expected that SoCs will define that numbering themselves via an enum. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-06-23 21:39:15 +00:00
}
clk: Get the CLK by index without device Getting a CLK by index with device is not straight forward for some use-cases like handling clock operations for child node in parent driver. So we need to process the child node in parent probe via ofnode and process CLK operation for child without udevice but with ofnode. So add clk_get_by_index_nodev() and move the common code in clk_get_by_index_tail() to use for clk_get_by_index() Cc: Stephen Warren <swarren@nvidia.com> Signed-off-by: Jagan Teki <jagan@amarulasolutions.com> Reviewed-by: Simon Glass <sjg@chromium.org>
2019-02-27 18:56:52 +00:00
static int clk_get_by_index_tail(int ret, ofnode node,
struct ofnode_phandle_args *args,
const char *list_name, int index,
struct clk *clk)
{
struct udevice *dev_clk;
const struct clk_ops *ops;
assert(clk);
clk->dev = NULL;
if (ret)
goto err;
ret = uclass_get_device_by_ofnode(UCLASS_CLK, args->node, &dev_clk);
if (ret) {
debug("%s: uclass_get_device_by_of_offset failed: err=%d\n",
__func__, ret);
return ret;
}
clk->dev = dev_clk;
ops = clk_dev_ops(dev_clk);
if (ops->of_xlate)
ret = ops->of_xlate(clk, args);
else
ret = clk_of_xlate_default(clk, args);
if (ret) {
debug("of_xlate() failed: %d\n", ret);
return ret;
}
return clk_request(dev_clk, clk);
err:
debug("%s: Node '%s', property '%s', failed to request CLK index %d: %d\n",
__func__, ofnode_get_name(node), list_name, index, ret);
return ret;
}
clk: refactor clk_get_by_index() into clk_get_by_indexed_prop() The logic in clk_get_by_index() may be useful for other properties than 'clocks': e.g. 'assigned-clocks' and 'assigned-clock-parents' follows the same model. This commit refactors clk_get_by_index() by introducing an internal function clk_get_by_indexed_prop() that allows to specify the name of the property to process. The original clk_get_by_index() call is simply directed through this helper function with the property name fixed to "clocks". Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address.
2018-01-08 10:18:18 +00:00
static int clk_get_by_indexed_prop(struct udevice *dev, const char *prop_name,
int index, struct clk *clk)
dm: clk: Add support for decoding clocks from the device tree Add a method which can locate a clock for a device, given its index. This uses the normal device tree bindings to return the clock device and the first argument which is normally used as a peripheral ID in U-Boot. Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Masahiro Yamada <yamada.masahiro@socionext.com>
2016-01-21 02:43:02 +00:00
{
int ret;
dm: clk: Update uclass to support livetree Update the clk uclass to support a live device tree. Signed-off-by: Simon Glass <sjg@chromium.org>
2017-05-31 03:47:29 +00:00
struct ofnode_phandle_args args;
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
debug("%s(dev=%p, index=%d, clk=%p)\n", __func__, dev, index, clk);
dm: clk: Add support for decoding clocks from the device tree Add a method which can locate a clock for a device, given its index. This uses the normal device tree bindings to return the clock device and the first argument which is normally used as a peripheral ID in U-Boot. Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Masahiro Yamada <yamada.masahiro@socionext.com>
2016-01-21 02:43:02 +00:00
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
assert(clk);
clk: add clk_release_all() Add clk_release_all() method which Disable/Free an array of clocks that has been previously requested by clk_request/get_by_*() Signed-off-by: Patrice Chotard <patrice.chotard@st.com> Reviewed-by: Simon Glass <sjg@chromium.org>
2017-07-18 09:57:07 +00:00
clk->dev = NULL;
clk: refactor clk_get_by_index() into clk_get_by_indexed_prop() The logic in clk_get_by_index() may be useful for other properties than 'clocks': e.g. 'assigned-clocks' and 'assigned-clock-parents' follows the same model. This commit refactors clk_get_by_index() by introducing an internal function clk_get_by_indexed_prop() that allows to specify the name of the property to process. The original clk_get_by_index() call is simply directed through this helper function with the property name fixed to "clocks". Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address.
2018-01-08 10:18:18 +00:00
ret = dev_read_phandle_with_args(dev, prop_name, "#clock-cells", 0,
clk: clk-uclass: Fix style violations checkpatch.pl complains that the clk_ops structures used in clk-uclass.c ought to be const, so we mark them as const. Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Mario Six <mario.six@gdsys.cc>
2018-01-15 10:06:51 +00:00
index, &args);
dm: clk: Add support for decoding clocks from the device tree Add a method which can locate a clock for a device, given its index. This uses the normal device tree bindings to return the clock device and the first argument which is normally used as a peripheral ID in U-Boot. Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Masahiro Yamada <yamada.masahiro@socionext.com>
2016-01-21 02:43:02 +00:00
if (ret) {
debug("%s: fdtdec_parse_phandle_with_args failed: err=%d\n",
__func__, ret);
return ret;
}
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
clk: Use clk_get_by_index_tail() clk_get_by_index_tail() now handle common clk get by index code so use it from clk_get_by_indexed_prop(). Cc: Stephen Warren <swarren@nvidia.com> Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
2019-02-27 18:56:53 +00:00
return clk_get_by_index_tail(ret, dev_ofnode(dev), &args, "clocks",
clk: Fix clk_get_by_* handling of index clk_get_by_index_nodev only ever fetched clock 1, due to passing a boolean predicate instead of the index. Other clk_get_by_* functions got the clock correctly, but passed a predicate instead of the index to clk_get_by_tail. This could lead to confusing error messages. Signed-off-by: Sean Anderson <seanga2@gmail.com> CC: Lukasz Majewski <lukma@denx.de>
2020-06-24 10:41:08 +00:00
index, clk);
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
}
clk: refactor clk_get_by_index() into clk_get_by_indexed_prop() The logic in clk_get_by_index() may be useful for other properties than 'clocks': e.g. 'assigned-clocks' and 'assigned-clock-parents' follows the same model. This commit refactors clk_get_by_index() by introducing an internal function clk_get_by_indexed_prop() that allows to specify the name of the property to process. The original clk_get_by_index() call is simply directed through this helper function with the property name fixed to "clocks". Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address.
2018-01-08 10:18:18 +00:00
int clk_get_by_index(struct udevice *dev, int index, struct clk *clk)
{
clk: Get the CLK by index without device Getting a CLK by index with device is not straight forward for some use-cases like handling clock operations for child node in parent driver. So we need to process the child node in parent probe via ofnode and process CLK operation for child without udevice but with ofnode. So add clk_get_by_index_nodev() and move the common code in clk_get_by_index_tail() to use for clk_get_by_index() Cc: Stephen Warren <swarren@nvidia.com> Signed-off-by: Jagan Teki <jagan@amarulasolutions.com> Reviewed-by: Simon Glass <sjg@chromium.org>
2019-02-27 18:56:52 +00:00
struct ofnode_phandle_args args;
int ret;
ret = dev_read_phandle_with_args(dev, "clocks", "#clock-cells", 0,
index, &args);
return clk_get_by_index_tail(ret, dev_ofnode(dev), &args, "clocks",
clk: Fix clk_get_by_* handling of index clk_get_by_index_nodev only ever fetched clock 1, due to passing a boolean predicate instead of the index. Other clk_get_by_* functions got the clock correctly, but passed a predicate instead of the index to clk_get_by_tail. This could lead to confusing error messages. Signed-off-by: Sean Anderson <seanga2@gmail.com> CC: Lukasz Majewski <lukma@denx.de>
2020-06-24 10:41:08 +00:00
index, clk);
clk: Get the CLK by index without device Getting a CLK by index with device is not straight forward for some use-cases like handling clock operations for child node in parent driver. So we need to process the child node in parent probe via ofnode and process CLK operation for child without udevice but with ofnode. So add clk_get_by_index_nodev() and move the common code in clk_get_by_index_tail() to use for clk_get_by_index() Cc: Stephen Warren <swarren@nvidia.com> Signed-off-by: Jagan Teki <jagan@amarulasolutions.com> Reviewed-by: Simon Glass <sjg@chromium.org>
2019-02-27 18:56:52 +00:00
}
int clk_get_by_index_nodev(ofnode node, int index, struct clk *clk)
{
struct ofnode_phandle_args args;
int ret;
ret = ofnode_parse_phandle_with_args(node, "clocks", "#clock-cells", 0,
clk: Fix clk_get_by_* handling of index clk_get_by_index_nodev only ever fetched clock 1, due to passing a boolean predicate instead of the index. Other clk_get_by_* functions got the clock correctly, but passed a predicate instead of the index to clk_get_by_tail. This could lead to confusing error messages. Signed-off-by: Sean Anderson <seanga2@gmail.com> CC: Lukasz Majewski <lukma@denx.de>
2020-06-24 10:41:08 +00:00
index, &args);
clk: Get the CLK by index without device Getting a CLK by index with device is not straight forward for some use-cases like handling clock operations for child node in parent driver. So we need to process the child node in parent probe via ofnode and process CLK operation for child without udevice but with ofnode. So add clk_get_by_index_nodev() and move the common code in clk_get_by_index_tail() to use for clk_get_by_index() Cc: Stephen Warren <swarren@nvidia.com> Signed-off-by: Jagan Teki <jagan@amarulasolutions.com> Reviewed-by: Simon Glass <sjg@chromium.org>
2019-02-27 18:56:52 +00:00
return clk_get_by_index_tail(ret, node, &args, "clocks",
clk: Fix clk_get_by_* handling of index clk_get_by_index_nodev only ever fetched clock 1, due to passing a boolean predicate instead of the index. Other clk_get_by_* functions got the clock correctly, but passed a predicate instead of the index to clk_get_by_tail. This could lead to confusing error messages. Signed-off-by: Sean Anderson <seanga2@gmail.com> CC: Lukasz Majewski <lukma@denx.de>
2020-06-24 10:41:08 +00:00
index, clk);
clk: refactor clk_get_by_index() into clk_get_by_indexed_prop() The logic in clk_get_by_index() may be useful for other properties than 'clocks': e.g. 'assigned-clocks' and 'assigned-clock-parents' follows the same model. This commit refactors clk_get_by_index() by introducing an internal function clk_get_by_indexed_prop() that allows to specify the name of the property to process. The original clk_get_by_index() call is simply directed through this helper function with the property name fixed to "clocks". Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address.
2018-01-08 10:18:18 +00:00
}
clk: implement clk_set_defaults() Linux uses the properties 'assigned-clocks', 'assigned-clock-parents' and 'assigned-clock-rates' to configure the clock subsystem for use with various peripheral nodes. This implements clk_set_defaults() and hooks it up with the general device probibin in drivers/core/device.c: when a new device is probed, clk_set_defaults() will be called for it and will process the properties mentioned above. Note that this functionality is designed to fail gracefully (i.e. if a clock-driver does not implement set_parent(), we simply accept this and ignore the error) as not to break existing board-support. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address. Series-version: 2 Cover-letter: clk: support assigned-clock, assigned-clock-parents, assigned-clock-rates For various peripherals on Rockchip SoCs (e.g. for the Ethernet GMAC), the parent-clock needs to be set via the DTS. This adds the required plumbing and implements the GMAC case for the RK3399. END
2018-01-08 12:59:18 +00:00
clk: Add get/enable/disable/release for a bulk of clocks This patch adds a "bulk" API to the clock API in order to get/enable/disable /release a group of clocks associated with a device. This bulk API will avoid adding a copy of the same code to manage a group of clocks in drivers. Signed-off-by: Neil Armstrong <narmstrong@baylibre.com> Reviewed-by: Simon Glass <sjg@chromium.org>
2018-04-03 09:44:18 +00:00
int clk_get_bulk(struct udevice *dev, struct clk_bulk *bulk)
{
int i, ret, err, count;
bulk->count = 0;
count = dev_count_phandle_with_args(dev, "clocks", "#clock-cells");
clk: fix clk_get_bulk when phandle error This fixes the Coverity Defect CID 175347 when dev_count_phandle_with_args() returns a negative value. Fixes: a855be87da49 ("clk: Add get/enable/disable/release for a bulk of clocks") Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
2018-04-17 09:30:31 +00:00
if (count < 1)
return count;
clk: Add get/enable/disable/release for a bulk of clocks This patch adds a "bulk" API to the clock API in order to get/enable/disable /release a group of clocks associated with a device. This bulk API will avoid adding a copy of the same code to manage a group of clocks in drivers. Signed-off-by: Neil Armstrong <narmstrong@baylibre.com> Reviewed-by: Simon Glass <sjg@chromium.org>
2018-04-03 09:44:18 +00:00
bulk->clks = devm_kcalloc(dev, count, sizeof(struct clk), GFP_KERNEL);
if (!bulk->clks)
return -ENOMEM;
for (i = 0; i < count; i++) {
ret = clk_get_by_index(dev, i, &bulk->clks[i]);
if (ret < 0)
goto bulk_get_err;
++bulk->count;
}
return 0;
bulk_get_err:
err = clk_release_all(bulk->clks, bulk->count);
if (err)
debug("%s: could release all clocks for %p\n",
__func__, dev);
return ret;
}
drivers: clk: Fix using assigned-clocks in the node of the clock it sets up This fixes the case where assigned-clocks is used to define a clock defaults inside this same clock's node. This is used sometimes to setup a default parents and/or rate for a clock. example: muxed_clock: muxed_clock { clocks = <&clk_provider 0>, <&clk_provider 1>; #clock-cells = <0>; assigned-clocks = <&muxed_clock>; assigned-clock-parents = <&clk_provider 1>; }; It doesn't work in u-boot because the assigned-clocks are setup *before* the clock is probed. (clk_set_parent() will likely crash or fail if called before the device probe function) Making it work by handling "assigned-clocks" in 2 steps: first before the clk device is probed, and then after the clk device is probed. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:06 +00:00
static int clk_set_default_parents(struct udevice *dev, int stage)
clk: implement clk_set_defaults() Linux uses the properties 'assigned-clocks', 'assigned-clock-parents' and 'assigned-clock-rates' to configure the clock subsystem for use with various peripheral nodes. This implements clk_set_defaults() and hooks it up with the general device probibin in drivers/core/device.c: when a new device is probed, clk_set_defaults() will be called for it and will process the properties mentioned above. Note that this functionality is designed to fail gracefully (i.e. if a clock-driver does not implement set_parent(), we simply accept this and ignore the error) as not to break existing board-support. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address. Series-version: 2 Cover-letter: clk: support assigned-clock, assigned-clock-parents, assigned-clock-rates For various peripherals on Rockchip SoCs (e.g. for the Ethernet GMAC), the parent-clock needs to be set via the DTS. This adds the required plumbing and implements the GMAC case for the RK3399. END
2018-01-08 12:59:18 +00:00
{
struct clk clk, parent_clk;
int index;
int num_parents;
int ret;
num_parents = dev_count_phandle_with_args(dev, "assigned-clock-parents",
"#clock-cells");
if (num_parents < 0) {
debug("%s: could not read assigned-clock-parents for %p\n",
__func__, dev);
return 0;
}
for (index = 0; index < num_parents; index++) {
ret = clk_get_by_indexed_prop(dev, "assigned-clock-parents",
index, &parent_clk);
clk: clk_set_default: accept no-op skip fields The Assigned Clock parents and rates misses the fact that a "0" entry can be passed to skip setting a parent or rate of an assigned clock as described in the Linux clock bindings at [1]. This patch simply skips the clock reparenting if the DT parsing returns -ENOENT and the clock rate setting if "0" is passed as clock rate. [1] https://github.com/torvalds/linux/blob/master/Documentation/devicetree/bindings/clock/clock-bindings.txt#L135 Fixes: f4fcba5c5baa "clk: implement clk_set_defaults()" Signed-off-by: Neil Armstrong <narmstrong@baylibre.com> Reviewed-by: Simon Glass <sjg@chromium.org>
2018-07-26 13:19:32 +00:00
/* If -ENOENT, this is a no-op entry */
if (ret == -ENOENT)
continue;
clk: implement clk_set_defaults() Linux uses the properties 'assigned-clocks', 'assigned-clock-parents' and 'assigned-clock-rates' to configure the clock subsystem for use with various peripheral nodes. This implements clk_set_defaults() and hooks it up with the general device probibin in drivers/core/device.c: when a new device is probed, clk_set_defaults() will be called for it and will process the properties mentioned above. Note that this functionality is designed to fail gracefully (i.e. if a clock-driver does not implement set_parent(), we simply accept this and ignore the error) as not to break existing board-support. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address. Series-version: 2 Cover-letter: clk: support assigned-clock, assigned-clock-parents, assigned-clock-rates For various peripherals on Rockchip SoCs (e.g. for the Ethernet GMAC), the parent-clock needs to be set via the DTS. This adds the required plumbing and implements the GMAC case for the RK3399. END
2018-01-08 12:59:18 +00:00
if (ret) {
debug("%s: could not get parent clock %d for %s\n",
__func__, index, dev_read_name(dev));
return ret;
}
ret = clk_get_by_indexed_prop(dev, "assigned-clocks",
index, &clk);
if (ret) {
debug("%s: could not get assigned clock %d for %s\n",
__func__, index, dev_read_name(dev));
return ret;
}
drivers: clk: Fix using assigned-clocks in the node of the clock it sets up This fixes the case where assigned-clocks is used to define a clock defaults inside this same clock's node. This is used sometimes to setup a default parents and/or rate for a clock. example: muxed_clock: muxed_clock { clocks = <&clk_provider 0>, <&clk_provider 1>; #clock-cells = <0>; assigned-clocks = <&muxed_clock>; assigned-clock-parents = <&clk_provider 1>; }; It doesn't work in u-boot because the assigned-clocks are setup *before* the clock is probed. (clk_set_parent() will likely crash or fail if called before the device probe function) Making it work by handling "assigned-clocks" in 2 steps: first before the clk device is probed, and then after the clk device is probed. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:06 +00:00
/* This is clk provider device trying to reparent itself
* It cannot be done right now but need to wait after the
* device is probed
*/
if (stage == 0 && clk.dev == dev)
continue;
if (stage > 0 && clk.dev != dev)
/* do not setup twice the parent clocks */
continue;
clk: implement clk_set_defaults() Linux uses the properties 'assigned-clocks', 'assigned-clock-parents' and 'assigned-clock-rates' to configure the clock subsystem for use with various peripheral nodes. This implements clk_set_defaults() and hooks it up with the general device probibin in drivers/core/device.c: when a new device is probed, clk_set_defaults() will be called for it and will process the properties mentioned above. Note that this functionality is designed to fail gracefully (i.e. if a clock-driver does not implement set_parent(), we simply accept this and ignore the error) as not to break existing board-support. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address. Series-version: 2 Cover-letter: clk: support assigned-clock, assigned-clock-parents, assigned-clock-rates For various peripherals on Rockchip SoCs (e.g. for the Ethernet GMAC), the parent-clock needs to be set via the DTS. This adds the required plumbing and implements the GMAC case for the RK3399. END
2018-01-08 12:59:18 +00:00
drivers: clk: Fix using assigned-clocks in the node of the clock it sets up This fixes the case where assigned-clocks is used to define a clock defaults inside this same clock's node. This is used sometimes to setup a default parents and/or rate for a clock. example: muxed_clock: muxed_clock { clocks = <&clk_provider 0>, <&clk_provider 1>; #clock-cells = <0>; assigned-clocks = <&muxed_clock>; assigned-clock-parents = <&clk_provider 1>; }; It doesn't work in u-boot because the assigned-clocks are setup *before* the clock is probed. (clk_set_parent() will likely crash or fail if called before the device probe function) Making it work by handling "assigned-clocks" in 2 steps: first before the clk device is probed, and then after the clk device is probed. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:06 +00:00
ret = clk_set_parent(&clk, &parent_clk);
clk: implement clk_set_defaults() Linux uses the properties 'assigned-clocks', 'assigned-clock-parents' and 'assigned-clock-rates' to configure the clock subsystem for use with various peripheral nodes. This implements clk_set_defaults() and hooks it up with the general device probibin in drivers/core/device.c: when a new device is probed, clk_set_defaults() will be called for it and will process the properties mentioned above. Note that this functionality is designed to fail gracefully (i.e. if a clock-driver does not implement set_parent(), we simply accept this and ignore the error) as not to break existing board-support. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address. Series-version: 2 Cover-letter: clk: support assigned-clock, assigned-clock-parents, assigned-clock-rates For various peripherals on Rockchip SoCs (e.g. for the Ethernet GMAC), the parent-clock needs to be set via the DTS. This adds the required plumbing and implements the GMAC case for the RK3399. END
2018-01-08 12:59:18 +00:00
/*
* Not all drivers may support clock-reparenting (as of now).
* Ignore errors due to this.
*/
if (ret == -ENOSYS)
continue;
clk: Fix error check in clk_set_default_parents() According to the documentation in clk.h, clk_set_parent() return the new rate of the clock or a negative error code. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-09-26 13:42:42 +00:00
if (ret < 0) {
clk: implement clk_set_defaults() Linux uses the properties 'assigned-clocks', 'assigned-clock-parents' and 'assigned-clock-rates' to configure the clock subsystem for use with various peripheral nodes. This implements clk_set_defaults() and hooks it up with the general device probibin in drivers/core/device.c: when a new device is probed, clk_set_defaults() will be called for it and will process the properties mentioned above. Note that this functionality is designed to fail gracefully (i.e. if a clock-driver does not implement set_parent(), we simply accept this and ignore the error) as not to break existing board-support. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address. Series-version: 2 Cover-letter: clk: support assigned-clock, assigned-clock-parents, assigned-clock-rates For various peripherals on Rockchip SoCs (e.g. for the Ethernet GMAC), the parent-clock needs to be set via the DTS. This adds the required plumbing and implements the GMAC case for the RK3399. END
2018-01-08 12:59:18 +00:00
debug("%s: failed to reparent clock %d for %s\n",
__func__, index, dev_read_name(dev));
return ret;
}
}
return 0;
}
drivers: clk: Fix using assigned-clocks in the node of the clock it sets up This fixes the case where assigned-clocks is used to define a clock defaults inside this same clock's node. This is used sometimes to setup a default parents and/or rate for a clock. example: muxed_clock: muxed_clock { clocks = <&clk_provider 0>, <&clk_provider 1>; #clock-cells = <0>; assigned-clocks = <&muxed_clock>; assigned-clock-parents = <&clk_provider 1>; }; It doesn't work in u-boot because the assigned-clocks are setup *before* the clock is probed. (clk_set_parent() will likely crash or fail if called before the device probe function) Making it work by handling "assigned-clocks" in 2 steps: first before the clk device is probed, and then after the clk device is probed. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:06 +00:00
static int clk_set_default_rates(struct udevice *dev, int stage)
clk: implement clk_set_defaults() Linux uses the properties 'assigned-clocks', 'assigned-clock-parents' and 'assigned-clock-rates' to configure the clock subsystem for use with various peripheral nodes. This implements clk_set_defaults() and hooks it up with the general device probibin in drivers/core/device.c: when a new device is probed, clk_set_defaults() will be called for it and will process the properties mentioned above. Note that this functionality is designed to fail gracefully (i.e. if a clock-driver does not implement set_parent(), we simply accept this and ignore the error) as not to break existing board-support. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address. Series-version: 2 Cover-letter: clk: support assigned-clock, assigned-clock-parents, assigned-clock-rates For various peripherals on Rockchip SoCs (e.g. for the Ethernet GMAC), the parent-clock needs to be set via the DTS. This adds the required plumbing and implements the GMAC case for the RK3399. END
2018-01-08 12:59:18 +00:00
{
struct clk clk;
int index;
int num_rates;
int size;
int ret = 0;
u32 *rates = NULL;
size = dev_read_size(dev, "assigned-clock-rates");
if (size < 0)
return 0;
num_rates = size / sizeof(u32);
rates = calloc(num_rates, sizeof(u32));
if (!rates)
return -ENOMEM;
ret = dev_read_u32_array(dev, "assigned-clock-rates", rates, num_rates);
if (ret)
goto fail;
for (index = 0; index < num_rates; index++) {
clk: clk_set_default: accept no-op skip fields The Assigned Clock parents and rates misses the fact that a "0" entry can be passed to skip setting a parent or rate of an assigned clock as described in the Linux clock bindings at [1]. This patch simply skips the clock reparenting if the DT parsing returns -ENOENT and the clock rate setting if "0" is passed as clock rate. [1] https://github.com/torvalds/linux/blob/master/Documentation/devicetree/bindings/clock/clock-bindings.txt#L135 Fixes: f4fcba5c5baa "clk: implement clk_set_defaults()" Signed-off-by: Neil Armstrong <narmstrong@baylibre.com> Reviewed-by: Simon Glass <sjg@chromium.org>
2018-07-26 13:19:32 +00:00
/* If 0 is passed, this is a no-op */
if (!rates[index])
continue;
clk: implement clk_set_defaults() Linux uses the properties 'assigned-clocks', 'assigned-clock-parents' and 'assigned-clock-rates' to configure the clock subsystem for use with various peripheral nodes. This implements clk_set_defaults() and hooks it up with the general device probibin in drivers/core/device.c: when a new device is probed, clk_set_defaults() will be called for it and will process the properties mentioned above. Note that this functionality is designed to fail gracefully (i.e. if a clock-driver does not implement set_parent(), we simply accept this and ignore the error) as not to break existing board-support. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address. Series-version: 2 Cover-letter: clk: support assigned-clock, assigned-clock-parents, assigned-clock-rates For various peripherals on Rockchip SoCs (e.g. for the Ethernet GMAC), the parent-clock needs to be set via the DTS. This adds the required plumbing and implements the GMAC case for the RK3399. END
2018-01-08 12:59:18 +00:00
ret = clk_get_by_indexed_prop(dev, "assigned-clocks",
index, &clk);
if (ret) {
debug("%s: could not get assigned clock %d for %s\n",
__func__, index, dev_read_name(dev));
continue;
}
drivers: clk: Fix using assigned-clocks in the node of the clock it sets up This fixes the case where assigned-clocks is used to define a clock defaults inside this same clock's node. This is used sometimes to setup a default parents and/or rate for a clock. example: muxed_clock: muxed_clock { clocks = <&clk_provider 0>, <&clk_provider 1>; #clock-cells = <0>; assigned-clocks = <&muxed_clock>; assigned-clock-parents = <&clk_provider 1>; }; It doesn't work in u-boot because the assigned-clocks are setup *before* the clock is probed. (clk_set_parent() will likely crash or fail if called before the device probe function) Making it work by handling "assigned-clocks" in 2 steps: first before the clk device is probed, and then after the clk device is probed. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:06 +00:00
/* This is clk provider device trying to program itself
* It cannot be done right now but need to wait after the
* device is probed
*/
if (stage == 0 && clk.dev == dev)
continue;
if (stage > 0 && clk.dev != dev)
/* do not setup twice the parent clocks */
continue;
clk: implement clk_set_defaults() Linux uses the properties 'assigned-clocks', 'assigned-clock-parents' and 'assigned-clock-rates' to configure the clock subsystem for use with various peripheral nodes. This implements clk_set_defaults() and hooks it up with the general device probibin in drivers/core/device.c: when a new device is probed, clk_set_defaults() will be called for it and will process the properties mentioned above. Note that this functionality is designed to fail gracefully (i.e. if a clock-driver does not implement set_parent(), we simply accept this and ignore the error) as not to break existing board-support. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address. Series-version: 2 Cover-letter: clk: support assigned-clock, assigned-clock-parents, assigned-clock-rates For various peripherals on Rockchip SoCs (e.g. for the Ethernet GMAC), the parent-clock needs to be set via the DTS. This adds the required plumbing and implements the GMAC case for the RK3399. END
2018-01-08 12:59:18 +00:00
ret = clk_set_rate(&clk, rates[index]);
drivers: clk: Fix using assigned-clocks in the node of the clock it sets up This fixes the case where assigned-clocks is used to define a clock defaults inside this same clock's node. This is used sometimes to setup a default parents and/or rate for a clock. example: muxed_clock: muxed_clock { clocks = <&clk_provider 0>, <&clk_provider 1>; #clock-cells = <0>; assigned-clocks = <&muxed_clock>; assigned-clock-parents = <&clk_provider 1>; }; It doesn't work in u-boot because the assigned-clocks are setup *before* the clock is probed. (clk_set_parent() will likely crash or fail if called before the device probe function) Making it work by handling "assigned-clocks" in 2 steps: first before the clk device is probed, and then after the clk device is probed. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:06 +00:00
clk: implement clk_set_defaults() Linux uses the properties 'assigned-clocks', 'assigned-clock-parents' and 'assigned-clock-rates' to configure the clock subsystem for use with various peripheral nodes. This implements clk_set_defaults() and hooks it up with the general device probibin in drivers/core/device.c: when a new device is probed, clk_set_defaults() will be called for it and will process the properties mentioned above. Note that this functionality is designed to fail gracefully (i.e. if a clock-driver does not implement set_parent(), we simply accept this and ignore the error) as not to break existing board-support. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address. Series-version: 2 Cover-letter: clk: support assigned-clock, assigned-clock-parents, assigned-clock-rates For various peripherals on Rockchip SoCs (e.g. for the Ethernet GMAC), the parent-clock needs to be set via the DTS. This adds the required plumbing and implements the GMAC case for the RK3399. END
2018-01-08 12:59:18 +00:00
if (ret < 0) {
clk: Improve debug message in clk_set_default_rates() It is helpful to print the clock number as well as the index, so that this can be looked up in the binding file. Update the debug() statement to do this. Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
2019-01-21 21:53:19 +00:00
debug("%s: failed to set rate on clock index %d (%ld) for %s\n",
__func__, index, clk.id, dev_read_name(dev));
clk: implement clk_set_defaults() Linux uses the properties 'assigned-clocks', 'assigned-clock-parents' and 'assigned-clock-rates' to configure the clock subsystem for use with various peripheral nodes. This implements clk_set_defaults() and hooks it up with the general device probibin in drivers/core/device.c: when a new device is probed, clk_set_defaults() will be called for it and will process the properties mentioned above. Note that this functionality is designed to fail gracefully (i.e. if a clock-driver does not implement set_parent(), we simply accept this and ignore the error) as not to break existing board-support. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address. Series-version: 2 Cover-letter: clk: support assigned-clock, assigned-clock-parents, assigned-clock-rates For various peripherals on Rockchip SoCs (e.g. for the Ethernet GMAC), the parent-clock needs to be set via the DTS. This adds the required plumbing and implements the GMAC case for the RK3399. END
2018-01-08 12:59:18 +00:00
break;
}
}
fail:
free(rates);
return ret;
}
drivers: clk: Fix using assigned-clocks in the node of the clock it sets up This fixes the case where assigned-clocks is used to define a clock defaults inside this same clock's node. This is used sometimes to setup a default parents and/or rate for a clock. example: muxed_clock: muxed_clock { clocks = <&clk_provider 0>, <&clk_provider 1>; #clock-cells = <0>; assigned-clocks = <&muxed_clock>; assigned-clock-parents = <&clk_provider 1>; }; It doesn't work in u-boot because the assigned-clocks are setup *before* the clock is probed. (clk_set_parent() will likely crash or fail if called before the device probe function) Making it work by handling "assigned-clocks" in 2 steps: first before the clk device is probed, and then after the clk device is probed. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:06 +00:00
int clk_set_defaults(struct udevice *dev, int stage)
clk: implement clk_set_defaults() Linux uses the properties 'assigned-clocks', 'assigned-clock-parents' and 'assigned-clock-rates' to configure the clock subsystem for use with various peripheral nodes. This implements clk_set_defaults() and hooks it up with the general device probibin in drivers/core/device.c: when a new device is probed, clk_set_defaults() will be called for it and will process the properties mentioned above. Note that this functionality is designed to fail gracefully (i.e. if a clock-driver does not implement set_parent(), we simply accept this and ignore the error) as not to break existing board-support. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address. Series-version: 2 Cover-letter: clk: support assigned-clock, assigned-clock-parents, assigned-clock-rates For various peripherals on Rockchip SoCs (e.g. for the Ethernet GMAC), the parent-clock needs to be set via the DTS. This adds the required plumbing and implements the GMAC case for the RK3399. END
2018-01-08 12:59:18 +00:00
{
int ret;
dm: clk: ignore default settings when node not valid When the device not binded with a node, we need ignore the parents and rate settings. Cc: Simon Glass <sjg@chromium.org> Cc: Jagan Teki <jagan@amarulasolutions.com> Cc: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Cc: Neil Armstrong <narmstrong@baylibre.com> Cc: Andreas Dannenberg <dannenberg@ti.com> Signed-off-by: Peng Fan <peng.fan@nxp.com>
2019-07-31 07:01:49 +00:00
if (!dev_of_valid(dev))
return 0;
clk: Allow clock defaults to be set during re-reloc state for SPL only In commit e5e06b65ad65 ("clk: Allow clock defaults to be set also during re-reloc state") the earlier guard against setting clock defaults in pre-reloc state was removed. While it is easy to filter 'assigned-clocks' properties for SPL using CONFIG_OF_SPL_REMOVE_PROPS, no such mechanism exists for the pre-reloc stage of the full U-Boot. With the default defconfig for the RK3399-Q7 (which filter the 'assigned-clocks' property for the DTS used by SPL anyway), this caused a pause during startup of the full U-Boot stage that lasted for almost 10s (due to the CPU not having been clocked up yet). This reintroduces the guard from commit f4fcba5c5baa ("clk: Allow clock defaults to be set also during re-reloc state") and extends it to only apply outside of a TPL/SPL build: i.e. clk_set_defaults will now run in pre-reloc state for SPL, but only after reloc for the full U-Boot. References: commit f4fcba5c5baa ("clk: implement clk_set_defaults()") References: commit e5e06b65ad65 ("clk: Allow clock defaults to be set also during re-reloc state") Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
2018-11-26 19:20:19 +00:00
/* If this not in SPL and pre-reloc state, don't take any action. */
if (!(IS_ENABLED(CONFIG_SPL_BUILD) || (gd->flags & GD_FLG_RELOC)))
return 0;
clk: implement clk_set_defaults() Linux uses the properties 'assigned-clocks', 'assigned-clock-parents' and 'assigned-clock-rates' to configure the clock subsystem for use with various peripheral nodes. This implements clk_set_defaults() and hooks it up with the general device probibin in drivers/core/device.c: when a new device is probed, clk_set_defaults() will be called for it and will process the properties mentioned above. Note that this functionality is designed to fail gracefully (i.e. if a clock-driver does not implement set_parent(), we simply accept this and ignore the error) as not to break existing board-support. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address. Series-version: 2 Cover-letter: clk: support assigned-clock, assigned-clock-parents, assigned-clock-rates For various peripherals on Rockchip SoCs (e.g. for the Ethernet GMAC), the parent-clock needs to be set via the DTS. This adds the required plumbing and implements the GMAC case for the RK3399. END
2018-01-08 12:59:18 +00:00
debug("%s(%s)\n", __func__, dev_read_name(dev));
drivers: clk: Fix using assigned-clocks in the node of the clock it sets up This fixes the case where assigned-clocks is used to define a clock defaults inside this same clock's node. This is used sometimes to setup a default parents and/or rate for a clock. example: muxed_clock: muxed_clock { clocks = <&clk_provider 0>, <&clk_provider 1>; #clock-cells = <0>; assigned-clocks = <&muxed_clock>; assigned-clock-parents = <&clk_provider 1>; }; It doesn't work in u-boot because the assigned-clocks are setup *before* the clock is probed. (clk_set_parent() will likely crash or fail if called before the device probe function) Making it work by handling "assigned-clocks" in 2 steps: first before the clk device is probed, and then after the clk device is probed. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:06 +00:00
ret = clk_set_default_parents(dev, stage);
clk: implement clk_set_defaults() Linux uses the properties 'assigned-clocks', 'assigned-clock-parents' and 'assigned-clock-rates' to configure the clock subsystem for use with various peripheral nodes. This implements clk_set_defaults() and hooks it up with the general device probibin in drivers/core/device.c: when a new device is probed, clk_set_defaults() will be called for it and will process the properties mentioned above. Note that this functionality is designed to fail gracefully (i.e. if a clock-driver does not implement set_parent(), we simply accept this and ignore the error) as not to break existing board-support. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address. Series-version: 2 Cover-letter: clk: support assigned-clock, assigned-clock-parents, assigned-clock-rates For various peripherals on Rockchip SoCs (e.g. for the Ethernet GMAC), the parent-clock needs to be set via the DTS. This adds the required plumbing and implements the GMAC case for the RK3399. END
2018-01-08 12:59:18 +00:00
if (ret)
return ret;
drivers: clk: Fix using assigned-clocks in the node of the clock it sets up This fixes the case where assigned-clocks is used to define a clock defaults inside this same clock's node. This is used sometimes to setup a default parents and/or rate for a clock. example: muxed_clock: muxed_clock { clocks = <&clk_provider 0>, <&clk_provider 1>; #clock-cells = <0>; assigned-clocks = <&muxed_clock>; assigned-clock-parents = <&clk_provider 1>; }; It doesn't work in u-boot because the assigned-clocks are setup *before* the clock is probed. (clk_set_parent() will likely crash or fail if called before the device probe function) Making it work by handling "assigned-clocks" in 2 steps: first before the clk device is probed, and then after the clk device is probed. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:06 +00:00
ret = clk_set_default_rates(dev, stage);
clk: implement clk_set_defaults() Linux uses the properties 'assigned-clocks', 'assigned-clock-parents' and 'assigned-clock-rates' to configure the clock subsystem for use with various peripheral nodes. This implements clk_set_defaults() and hooks it up with the general device probibin in drivers/core/device.c: when a new device is probed, clk_set_defaults() will be called for it and will process the properties mentioned above. Note that this functionality is designed to fail gracefully (i.e. if a clock-driver does not implement set_parent(), we simply accept this and ignore the error) as not to break existing board-support. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address. Series-version: 2 Cover-letter: clk: support assigned-clock, assigned-clock-parents, assigned-clock-rates For various peripherals on Rockchip SoCs (e.g. for the Ethernet GMAC), the parent-clock needs to be set via the DTS. This adds the required plumbing and implements the GMAC case for the RK3399. END
2018-01-08 12:59:18 +00:00
if (ret < 0)
return ret;
return 0;
}
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
int clk_get_by_name(struct udevice *dev, const char *name, struct clk *clk)
{
int index;
debug("%s(dev=%p, name=%s, clk=%p)\n", __func__, dev, name, clk);
clk: add clk_release_all() Add clk_release_all() method which Disable/Free an array of clocks that has been previously requested by clk_request/get_by_*() Signed-off-by: Patrice Chotard <patrice.chotard@st.com> Reviewed-by: Simon Glass <sjg@chromium.org>
2017-07-18 09:57:07 +00:00
clk->dev = NULL;
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
dm: clk: Update uclass to support livetree Update the clk uclass to support a live device tree. Signed-off-by: Simon Glass <sjg@chromium.org>
2017-05-31 03:47:29 +00:00
index = dev_read_stringlist_search(dev, "clock-names", name);
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
if (index < 0) {
libfdt: Bring in upstream stringlist functions These have now landed upstream. The naming is different and in one case the function signature has changed. Update the code to match. This applies the following upstream commits by Thierry Reding <treding@nvidia.com> : 604e61e fdt: Add functions to retrieve strings 8702bd1 fdt: Add a function to get the index of a string 2218387 fdt: Add a function to count strings Signed-off-by: Simon Glass <sjg@chromium.org>
2016-10-02 23:59:28 +00:00
debug("fdt_stringlist_search() failed: %d\n", index);
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
return index;
}
return clk_get_by_index(dev, index, clk);
dm: clk: Add support for decoding clocks from the device tree Add a method which can locate a clock for a device, given its index. This uses the normal device tree bindings to return the clock device and the first argument which is normally used as a peripheral ID in U-Boot. Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Masahiro Yamada <yamada.masahiro@socionext.com>
2016-01-21 02:43:02 +00:00
}
clk: uclass: clk_get_by_name() must not be available if CONFIG_OF_PLATDATA is enabled clk_get_by_name() requires clk_get_by_id() that is not available if CONFIG_OF_PLATDATA is defined, so move clk_get_by_name() into #else condition of #if CONFIG_IS_ENABLED(OF_PLATDATA). Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
2019-12-12 22:53:19 +00:00
# endif /* OF_PLATDATA */
clk: fix compilation errors for poplar platform Move clk_release_all() prototype and definition inside OF_CONTROL flag to avoid following compilation error for poplar platform: aarch64: + poplar +drivers/usb/host/built-in.o: In function `ehci_usb_remove': +drivers/usb/host/ehci-generic.c:159: undefined reference to `clk_release_all' +drivers/usb/host/built-in.o: In function `ehci_usb_probe': +drivers/usb/host/ehci-generic.c:133: undefined reference to `clk_release_all' +make[1]: *** [u-boot] Error 139 +make: *** [sub-make] Error 2 Introduced by 4e542c4 clk: add clk_release_all() Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
2017-07-25 11:24:45 +00:00
clk: add APIs to get (optional) clock by name without a device Sometimes we may need get (optional) clock without a device, that means use ofnode. e.g. when the phy node has subnode, and there is no device created for subnode, in this case, we need these new APIs to get subnode's clock. Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: Ryder Lee <ryder.lee@mediatek.com>
2020-01-09 03:35:07 +00:00
int clk_get_by_name_nodev(ofnode node, const char *name, struct clk *clk)
{
int index;
debug("%s(node=%p, name=%s, clk=%p)\n", __func__,
ofnode_get_name(node), name, clk);
clk->dev = NULL;
index = ofnode_stringlist_search(node, "clock-names", name);
if (index < 0) {
debug("fdt_stringlist_search() failed: %d\n", index);
return index;
}
return clk_get_by_index_nodev(node, index, clk);
}
int clk_get_optional_nodev(ofnode node, const char *name, struct clk *clk)
{
int ret;
ret = clk_get_by_name_nodev(node, name, clk);
if (ret == -ENODATA)
return 0;
return ret;
}
clk: fix compilation errors for poplar platform Move clk_release_all() prototype and definition inside OF_CONTROL flag to avoid following compilation error for poplar platform: aarch64: + poplar +drivers/usb/host/built-in.o: In function `ehci_usb_remove': +drivers/usb/host/ehci-generic.c:159: undefined reference to `clk_release_all' +drivers/usb/host/built-in.o: In function `ehci_usb_probe': +drivers/usb/host/ehci-generic.c:133: undefined reference to `clk_release_all' +make[1]: *** [u-boot] Error 139 +make: *** [sub-make] Error 2 Introduced by 4e542c4 clk: add clk_release_all() Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
2017-07-25 11:24:45 +00:00
int clk_release_all(struct clk *clk, int count)
{
int i, ret;
for (i = 0; i < count; i++) {
debug("%s(clk[%d]=%p)\n", __func__, i, &clk[i]);
/* check if clock has been previously requested */
if (!clk[i].dev)
continue;
ret = clk_disable(&clk[i]);
if (ret && ret != -ENOSYS)
return ret;
ret = clk_free(&clk[i]);
if (ret && ret != -ENOSYS)
return ret;
}
return 0;
}
dm: clk: Add support for of-platdata Add support for this feature in the core clock code. Signed-off-by: Simon Glass <sjg@chromium.org>
2016-07-04 17:58:03 +00:00
#endif /* OF_CONTROL */
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
int clk_request(struct udevice *dev, struct clk *clk)
{
drivers: clk: Handle gracefully NULL pointers Prepare the way for a managed CLK API by handling NULL pointers without crashing nor failing. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:03 +00:00
const struct clk_ops *ops;
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
debug("%s(dev=%p, clk=%p)\n", __func__, dev, clk);
drivers: clk: Handle gracefully NULL pointers Prepare the way for a managed CLK API by handling NULL pointers without crashing nor failing. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:03 +00:00
if (!clk)
return 0;
ops = clk_dev_ops(dev);
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
clk->dev = dev;
if (!ops->request)
return 0;
return ops->request(clk);
}
int clk_free(struct clk *clk)
{
drivers: clk: Handle gracefully NULL pointers Prepare the way for a managed CLK API by handling NULL pointers without crashing nor failing. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:03 +00:00
const struct clk_ops *ops;
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
debug("%s(clk=%p)\n", __func__, clk);
clk: check valid clock by clk_valid() Add valid check for clk->dev, it's useful when get optional clock even when the clk point is valid, but its dev will be NULL. Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: Ryder Lee <ryder.lee@mediatek.com>
2020-01-09 03:35:06 +00:00
if (!clk_valid(clk))
drivers: clk: Handle gracefully NULL pointers Prepare the way for a managed CLK API by handling NULL pointers without crashing nor failing. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:03 +00:00
return 0;
ops = clk_dev_ops(clk->dev);
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
clk: Rename free() to rfree() This function name conflicts with our desire to #define free() to something else on sandbox. Since it deals with resources, rename it to rfree(). Signed-off-by: Simon Glass <sjg@chromium.org>
2020-02-03 14:35:54 +00:00
if (!ops->rfree)
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
return 0;
clk: Rename free() to rfree() This function name conflicts with our desire to #define free() to something else on sandbox. Since it deals with resources, rename it to rfree(). Signed-off-by: Simon Glass <sjg@chromium.org>
2020-02-03 14:35:54 +00:00
return ops->rfree(clk);
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
}
ulong clk_get_rate(struct clk *clk)
{
drivers: clk: Handle gracefully NULL pointers Prepare the way for a managed CLK API by handling NULL pointers without crashing nor failing. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:03 +00:00
const struct clk_ops *ops;
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
debug("%s(clk=%p)\n", __func__, clk);
clk: check valid clock by clk_valid() Add valid check for clk->dev, it's useful when get optional clock even when the clk point is valid, but its dev will be NULL. Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: Ryder Lee <ryder.lee@mediatek.com>
2020-01-09 03:35:06 +00:00
if (!clk_valid(clk))
drivers: clk: Handle gracefully NULL pointers Prepare the way for a managed CLK API by handling NULL pointers without crashing nor failing. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:03 +00:00
return 0;
ops = clk_dev_ops(clk->dev);
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
if (!ops->get_rate)
return -ENOSYS;
return ops->get_rate(clk);
}
dm: clk: Define clk_get_parent() for clk operations This commit adds the clk_get_parent() function, which is responsible for getting the parent's struct clock pointer. U-Boot's DM support for getting parent is different (the parent relationship is in udevice) than the one in Common Clock Framework [CCF] in Linux. To obtain the pointer to struct clk of parent the pdev->uclass_priv field is read via dev_get_clk_ptr() wrapper. Signed-off-by: Lukasz Majewski <lukma@denx.de> Reviewed-by: Peng Fan <peng.fan@nxp.com>
2019-06-24 13:50:42 +00:00
struct clk *clk_get_parent(struct clk *clk)
{
struct udevice *pdev;
struct clk *pclk;
debug("%s(clk=%p)\n", __func__, clk);
clk: check valid clock by clk_valid() Add valid check for clk->dev, it's useful when get optional clock even when the clk point is valid, but its dev will be NULL. Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: Ryder Lee <ryder.lee@mediatek.com>
2020-01-09 03:35:06 +00:00
if (!clk_valid(clk))
drivers: clk: Handle gracefully NULL pointers Prepare the way for a managed CLK API by handling NULL pointers without crashing nor failing. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:03 +00:00
return NULL;
dm: clk: Define clk_get_parent() for clk operations This commit adds the clk_get_parent() function, which is responsible for getting the parent's struct clock pointer. U-Boot's DM support for getting parent is different (the parent relationship is in udevice) than the one in Common Clock Framework [CCF] in Linux. To obtain the pointer to struct clk of parent the pdev->uclass_priv field is read via dev_get_clk_ptr() wrapper. Signed-off-by: Lukasz Majewski <lukma@denx.de> Reviewed-by: Peng Fan <peng.fan@nxp.com>
2019-06-24 13:50:42 +00:00
pdev = dev_get_parent(clk->dev);
pclk = dev_get_clk_ptr(pdev);
if (!pclk)
return ERR_PTR(-ENODEV);
return pclk;
}
dm: clk: Define clk_get_parent_rate() for clk operations This commit adds the clk_get_parent_rate() function, which is responsible for getting the rate of parent clock. Unfortunately, u-boot's DM support for getting parent is different (the parent relationship is in udevice) than the one in Common Clock Framework [CCF] in Linux. To alleviate this problem - the clk_get_parent_rate() function has been introduced to clk-uclass.c. Signed-off-by: Lukasz Majewski <lukma@denx.de> Reviewed-by: Peng Fan <peng.fan@nxp.com>
2019-06-24 13:50:43 +00:00
long long clk_get_parent_rate(struct clk *clk)
{
const struct clk_ops *ops;
struct clk *pclk;
debug("%s(clk=%p)\n", __func__, clk);
clk: check valid clock by clk_valid() Add valid check for clk->dev, it's useful when get optional clock even when the clk point is valid, but its dev will be NULL. Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: Ryder Lee <ryder.lee@mediatek.com>
2020-01-09 03:35:06 +00:00
if (!clk_valid(clk))
drivers: clk: Handle gracefully NULL pointers Prepare the way for a managed CLK API by handling NULL pointers without crashing nor failing. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:03 +00:00
return 0;
dm: clk: Define clk_get_parent_rate() for clk operations This commit adds the clk_get_parent_rate() function, which is responsible for getting the rate of parent clock. Unfortunately, u-boot's DM support for getting parent is different (the parent relationship is in udevice) than the one in Common Clock Framework [CCF] in Linux. To alleviate this problem - the clk_get_parent_rate() function has been introduced to clk-uclass.c. Signed-off-by: Lukasz Majewski <lukma@denx.de> Reviewed-by: Peng Fan <peng.fan@nxp.com>
2019-06-24 13:50:43 +00:00
pclk = clk_get_parent(clk);
if (IS_ERR(pclk))
return -ENODEV;
ops = clk_dev_ops(pclk->dev);
if (!ops->get_rate)
return -ENOSYS;
dm: clk: Extend clk_get_parent_rate() to support CLK_GET_RATE_NOCACHE flag If the CLK_GET_RATE_NOCACHE flag is set - the clk_get_parent_rate() provides recalculated clock value without considering the cache setting. This may be necessary for some clocks tightly coupled with power domains (i.e. imx8), and prevents from reading invalid cached values. Signed-off-by: Lukasz Majewski <lukma@denx.de> Reviewed-by: Peng Fan <peng.fan@nxp.com>
2019-06-24 13:50:46 +00:00
/* Read the 'rate' if not already set or if proper flag set*/
if (!pclk->rate || pclk->flags & CLK_GET_RATE_NOCACHE)
dm: clk: Define clk_get_parent_rate() for clk operations This commit adds the clk_get_parent_rate() function, which is responsible for getting the rate of parent clock. Unfortunately, u-boot's DM support for getting parent is different (the parent relationship is in udevice) than the one in Common Clock Framework [CCF] in Linux. To alleviate this problem - the clk_get_parent_rate() function has been introduced to clk-uclass.c. Signed-off-by: Lukasz Majewski <lukma@denx.de> Reviewed-by: Peng Fan <peng.fan@nxp.com>
2019-06-24 13:50:43 +00:00
pclk->rate = clk_get_rate(pclk);
return pclk->rate;
}
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
ulong clk_set_rate(struct clk *clk, ulong rate)
{
drivers: clk: Handle gracefully NULL pointers Prepare the way for a managed CLK API by handling NULL pointers without crashing nor failing. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:03 +00:00
const struct clk_ops *ops;
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
debug("%s(clk=%p, rate=%lu)\n", __func__, clk, rate);
clk: check valid clock by clk_valid() Add valid check for clk->dev, it's useful when get optional clock even when the clk point is valid, but its dev will be NULL. Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: Ryder Lee <ryder.lee@mediatek.com>
2020-01-09 03:35:06 +00:00
if (!clk_valid(clk))
drivers: clk: Handle gracefully NULL pointers Prepare the way for a managed CLK API by handling NULL pointers without crashing nor failing. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:03 +00:00
return 0;
ops = clk_dev_ops(clk->dev);
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
if (!ops->set_rate)
return -ENOSYS;
return ops->set_rate(clk, rate);
}
clk: add clk_set_parent() Clocks may support multiple parents: this change introduces an optional operation on the clk-uclass to set a clock's parent. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address.
2018-01-08 10:15:08 +00:00
int clk_set_parent(struct clk *clk, struct clk *parent)
{
drivers: clk: Handle gracefully NULL pointers Prepare the way for a managed CLK API by handling NULL pointers without crashing nor failing. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:03 +00:00
const struct clk_ops *ops;
clk: add clk_set_parent() Clocks may support multiple parents: this change introduces an optional operation on the clk-uclass to set a clock's parent. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address.
2018-01-08 10:15:08 +00:00
debug("%s(clk=%p, parent=%p)\n", __func__, clk, parent);
clk: check valid clock by clk_valid() Add valid check for clk->dev, it's useful when get optional clock even when the clk point is valid, but its dev will be NULL. Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: Ryder Lee <ryder.lee@mediatek.com>
2020-01-09 03:35:06 +00:00
if (!clk_valid(clk))
drivers: clk: Handle gracefully NULL pointers Prepare the way for a managed CLK API by handling NULL pointers without crashing nor failing. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:03 +00:00
return 0;
ops = clk_dev_ops(clk->dev);
clk: add clk_set_parent() Clocks may support multiple parents: this change introduces an optional operation on the clk-uclass to set a clock's parent. Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> Tested-by: David Wu <david.wu@rock-chips.com> Series-changes: 2 - Fixed David's email address.
2018-01-08 10:15:08 +00:00
if (!ops->set_parent)
return -ENOSYS;
return ops->set_parent(clk, parent);
}
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
int clk_enable(struct clk *clk)
{
drivers: clk: Handle gracefully NULL pointers Prepare the way for a managed CLK API by handling NULL pointers without crashing nor failing. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:03 +00:00
const struct clk_ops *ops;
clk: prograte clk enable/disable to parent On i.MX8MM, thinking such as clk path OSC->PLL->PLL GATE->CCM ROOT->CCGR GATE->Device Only enabling CCGR GATE is not enough, we also need to enable PLL GATE to make sure the clk path work. So when enabling CCGR GATE, we could prograte to enabling PLL GATE to make life easier. Signed-off-by: Peng Fan <peng.fan@nxp.com>
2019-08-21 13:35:09 +00:00
struct clk *clkp = NULL;
int ret;
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
debug("%s(clk=%p)\n", __func__, clk);
clk: check valid clock by clk_valid() Add valid check for clk->dev, it's useful when get optional clock even when the clk point is valid, but its dev will be NULL. Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: Ryder Lee <ryder.lee@mediatek.com>
2020-01-09 03:35:06 +00:00
if (!clk_valid(clk))
drivers: clk: Handle gracefully NULL pointers Prepare the way for a managed CLK API by handling NULL pointers without crashing nor failing. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:03 +00:00
return 0;
ops = clk_dev_ops(clk->dev);
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
clk: prograte clk enable/disable to parent On i.MX8MM, thinking such as clk path OSC->PLL->PLL GATE->CCM ROOT->CCGR GATE->Device Only enabling CCGR GATE is not enough, we also need to enable PLL GATE to make sure the clk path work. So when enabling CCGR GATE, we could prograte to enabling PLL GATE to make life easier. Signed-off-by: Peng Fan <peng.fan@nxp.com>
2019-08-21 13:35:09 +00:00
if (CONFIG_IS_ENABLED(CLK_CCF)) {
/* Take id 0 as a non-valid clk, such as dummy */
if (clk->id && !clk_get_by_id(clk->id, &clkp)) {
if (clkp->enable_count) {
clkp->enable_count++;
return 0;
}
if (clkp->dev->parent &&
device_get_uclass_id(clkp->dev) == UCLASS_CLK) {
ret = clk_enable(dev_get_clk_ptr(clkp->dev->parent));
if (ret) {
printf("Enable %s failed\n",
clkp->dev->parent->name);
return ret;
}
}
}
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
clk: prograte clk enable/disable to parent On i.MX8MM, thinking such as clk path OSC->PLL->PLL GATE->CCM ROOT->CCGR GATE->Device Only enabling CCGR GATE is not enough, we also need to enable PLL GATE to make sure the clk path work. So when enabling CCGR GATE, we could prograte to enabling PLL GATE to make life easier. Signed-off-by: Peng Fan <peng.fan@nxp.com>
2019-08-21 13:35:09 +00:00
if (ops->enable) {
ret = ops->enable(clk);
if (ret) {
printf("Enable %s failed\n", clk->dev->name);
return ret;
}
}
if (clkp)
clkp->enable_count++;
} else {
if (!ops->enable)
return -ENOSYS;
return ops->enable(clk);
}
return 0;
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
}
clk: Add get/enable/disable/release for a bulk of clocks This patch adds a "bulk" API to the clock API in order to get/enable/disable /release a group of clocks associated with a device. This bulk API will avoid adding a copy of the same code to manage a group of clocks in drivers. Signed-off-by: Neil Armstrong <narmstrong@baylibre.com> Reviewed-by: Simon Glass <sjg@chromium.org>
2018-04-03 09:44:18 +00:00
int clk_enable_bulk(struct clk_bulk *bulk)
{
int i, ret;
for (i = 0; i < bulk->count; i++) {
ret = clk_enable(&bulk->clks[i]);
if (ret < 0 && ret != -ENOSYS)
return ret;
}
return 0;
}
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
int clk_disable(struct clk *clk)
{
drivers: clk: Handle gracefully NULL pointers Prepare the way for a managed CLK API by handling NULL pointers without crashing nor failing. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:03 +00:00
const struct clk_ops *ops;
clk: prograte clk enable/disable to parent On i.MX8MM, thinking such as clk path OSC->PLL->PLL GATE->CCM ROOT->CCGR GATE->Device Only enabling CCGR GATE is not enough, we also need to enable PLL GATE to make sure the clk path work. So when enabling CCGR GATE, we could prograte to enabling PLL GATE to make life easier. Signed-off-by: Peng Fan <peng.fan@nxp.com>
2019-08-21 13:35:09 +00:00
struct clk *clkp = NULL;
int ret;
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
debug("%s(clk=%p)\n", __func__, clk);
clk: check valid clock by clk_valid() Add valid check for clk->dev, it's useful when get optional clock even when the clk point is valid, but its dev will be NULL. Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: Ryder Lee <ryder.lee@mediatek.com>
2020-01-09 03:35:06 +00:00
if (!clk_valid(clk))
drivers: clk: Handle gracefully NULL pointers Prepare the way for a managed CLK API by handling NULL pointers without crashing nor failing. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:03 +00:00
return 0;
ops = clk_dev_ops(clk->dev);
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
clk: prograte clk enable/disable to parent On i.MX8MM, thinking such as clk path OSC->PLL->PLL GATE->CCM ROOT->CCGR GATE->Device Only enabling CCGR GATE is not enough, we also need to enable PLL GATE to make sure the clk path work. So when enabling CCGR GATE, we could prograte to enabling PLL GATE to make life easier. Signed-off-by: Peng Fan <peng.fan@nxp.com>
2019-08-21 13:35:09 +00:00
if (CONFIG_IS_ENABLED(CLK_CCF)) {
if (clk->id && !clk_get_by_id(clk->id, &clkp)) {
if (clkp->enable_count == 0) {
printf("clk %s already disabled\n",
clkp->dev->name);
return 0;
}
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
clk: prograte clk enable/disable to parent On i.MX8MM, thinking such as clk path OSC->PLL->PLL GATE->CCM ROOT->CCGR GATE->Device Only enabling CCGR GATE is not enough, we also need to enable PLL GATE to make sure the clk path work. So when enabling CCGR GATE, we could prograte to enabling PLL GATE to make life easier. Signed-off-by: Peng Fan <peng.fan@nxp.com>
2019-08-21 13:35:09 +00:00
if (--clkp->enable_count > 0)
return 0;
}
if (ops->disable) {
ret = ops->disable(clk);
if (ret)
return ret;
}
if (clkp && clkp->dev->parent &&
device_get_uclass_id(clkp->dev) == UCLASS_CLK) {
ret = clk_disable(dev_get_clk_ptr(clkp->dev->parent));
if (ret) {
printf("Disable %s failed\n",
clkp->dev->parent->name);
return ret;
}
}
} else {
if (!ops->disable)
return -ENOSYS;
return ops->disable(clk);
}
return 0;
clk: convert API to match reset/mailbox style The following changes are made to the clock API: * The concept of "clocks" and "peripheral clocks" are unified; each clock provider now implements a single set of clocks. This provides a simpler conceptual interface to clients, and better aligns with device tree clock bindings. * Clocks are now identified with a single "struct clk", rather than requiring clients to store the clock provider device and clock identity values separately. For simple clock consumers, this isolates clients from internal details of the clock API. * clk.h is split so it only contains the client/consumer API, whereas clk-uclass.h contains the provider API. This aligns with the recently added reset and mailbox APIs. * clk_ops .of_xlate(), .request(), and .free() are added so providers can customize these operations if needed. This also aligns with the recently added reset and mailbox APIs. * clk_disable() is added. * All users of the current clock APIs are updated. * Sandbox clock tests are updated to exercise clock lookup via DT, and clock enable/disable. * rkclk_get_clk() is removed and replaced with standard APIs. Buildman shows no clock-related errors for any board for which buildman can download a toolchain. test/py passes for sandbox (which invokes the dm clk test amongst others). Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-06-17 15:44:00 +00:00
}
dm: clk: Add support for decoding clocks from the device tree Add a method which can locate a clock for a device, given its index. This uses the normal device tree bindings to return the clock device and the first argument which is normally used as a peripheral ID in U-Boot. Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Masahiro Yamada <yamada.masahiro@socionext.com>
2016-01-21 02:43:02 +00:00
clk: Add get/enable/disable/release for a bulk of clocks This patch adds a "bulk" API to the clock API in order to get/enable/disable /release a group of clocks associated with a device. This bulk API will avoid adding a copy of the same code to manage a group of clocks in drivers. Signed-off-by: Neil Armstrong <narmstrong@baylibre.com> Reviewed-by: Simon Glass <sjg@chromium.org>
2018-04-03 09:44:18 +00:00
int clk_disable_bulk(struct clk_bulk *bulk)
{
int i, ret;
for (i = 0; i < bulk->count; i++) {
ret = clk_disable(&bulk->clks[i]);
if (ret < 0 && ret != -ENOSYS)
return ret;
}
return 0;
}
dm: clk: Define clk_get_by_id() for clk operations This commit adds the clk_get_by_id() function, which is responsible for getting the udevice with matching clk->id. Such approach allows re-usage of inherit DM list relationship for the same class (UCLASS_CLK). As a result - we don't need any other external list - it is just enough to look for UCLASS_CLK related udevices. Signed-off-by: Lukasz Majewski <lukma@denx.de> Reviewed-by: Peng Fan <peng.fan@nxp.com>
2019-06-24 13:50:44 +00:00
int clk_get_by_id(ulong id, struct clk **clkp)
{
struct udevice *dev;
struct uclass *uc;
int ret;
ret = uclass_get(UCLASS_CLK, &uc);
if (ret)
return ret;
uclass_foreach_dev(dev, uc) {
struct clk *clk = dev_get_clk_ptr(dev);
if (clk && clk->id == id) {
*clkp = clk;
return 0;
}
}
return -ENOENT;
}
clk: add support for clk_is_match() Add support for clk_is_match() which is required to know if two clock pointers point to the same exact physical clock. Also add a unit test for the new API. Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com> Signed-off-by: Sekhar Nori <nsekhar@ti.com>
2019-08-01 13:42:55 +00:00
bool clk_is_match(const struct clk *p, const struct clk *q)
{
/* trivial case: identical struct clk's or both NULL */
if (p == q)
return true;
drivers: clk: Handle gracefully NULL pointers Prepare the way for a managed CLK API by handling NULL pointers without crashing nor failing. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:03 +00:00
/* trivial case #2: on the clk pointer is NULL */
if (!p || !q)
return false;
clk: add support for clk_is_match() Add support for clk_is_match() which is required to know if two clock pointers point to the same exact physical clock. Also add a unit test for the new API. Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com> Signed-off-by: Sekhar Nori <nsekhar@ti.com>
2019-08-01 13:42:55 +00:00
/* same device, id and data */
if (p->dev == q->dev && p->id == q->id && p->data == q->data)
return true;
return false;
}
drivers: clk: Add a managed API to get clocks from the device-tree Add devm_clk_get(), devm_clk_get_optional() to get clocks from the device-tree. The clocks is automatically released and the data structure freed when the device is unbound. Also add devm_clk_put() to release the clock and free the data structure manually. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:04 +00:00
static void devm_clk_release(struct udevice *dev, void *res)
{
clk_free(res);
}
static int devm_clk_match(struct udevice *dev, void *res, void *data)
{
return res == data;
}
struct clk *devm_clk_get(struct udevice *dev, const char *id)
{
int rc;
struct clk *clk;
clk = devres_alloc(devm_clk_release, sizeof(struct clk), __GFP_ZERO);
if (unlikely(!clk))
return ERR_PTR(-ENOMEM);
rc = clk_get_by_name(dev, id, clk);
if (rc)
return ERR_PTR(rc);
devres_add(dev, clk);
return clk;
}
struct clk *devm_clk_get_optional(struct udevice *dev, const char *id)
{
struct clk *clk = devm_clk_get(dev, id);
clk: fix error check for devm_clk_get_optional() If skip all return error number, it may skip some real error cases, so only skip the error when the clock is not provided in DTS Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: Ryder Lee <ryder.lee@mediatek.com>
2020-01-09 03:35:05 +00:00
if (PTR_ERR(clk) == -ENODATA)
drivers: clk: Add a managed API to get clocks from the device-tree Add devm_clk_get(), devm_clk_get_optional() to get clocks from the device-tree. The clocks is automatically released and the data structure freed when the device is unbound. Also add devm_clk_put() to release the clock and free the data structure manually. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:04 +00:00
return NULL;
return clk;
}
void devm_clk_put(struct udevice *dev, struct clk *clk)
{
int rc;
if (!clk)
return;
rc = devres_release(dev, devm_clk_release, devm_clk_match, clk);
WARN_ON(rc);
}
drivers: clk: Fix using assigned-clocks in the node of the clock it sets up This fixes the case where assigned-clocks is used to define a clock defaults inside this same clock's node. This is used sometimes to setup a default parents and/or rate for a clock. example: muxed_clock: muxed_clock { clocks = <&clk_provider 0>, <&clk_provider 1>; #clock-cells = <0>; assigned-clocks = <&muxed_clock>; assigned-clock-parents = <&clk_provider 1>; }; It doesn't work in u-boot because the assigned-clocks are setup *before* the clock is probed. (clk_set_parent() will likely crash or fail if called before the device probe function) Making it work by handling "assigned-clocks" in 2 steps: first before the clk device is probed, and then after the clk device is probed. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:06 +00:00
int clk_uclass_post_probe(struct udevice *dev)
{
/*
* when a clock provider is probed. Call clk_set_defaults()
* also after the device is probed. This takes care of cases
* where the DT is used to setup default parents and rates
* using assigned-clocks
*/
clk_set_defaults(dev, 1);
return 0;
}
dm: Add a clock uclass Clocks are an important feature of platforms and have become increasing complex with time. Most modern SoCs have multiple PLLs and dozens of clock dividers which distribute clocks to on-chip peripherals. Some SoC implementations have a clock API which is private to that SoC family, e.g. Tegra and Exynos. This is useful but it would be better to have a common API that can be understood and used throughout U-Boot. Add a simple clock API as a starting point. It supports querying and setting the rate of a clock. Each clock is a device. To reduce memory and processing overhead the concept of peripheral clocks is provided. These do not need to be explicit devices - it is possible to write a driver that can adjust the I2C clock (for example) without an explicit I2C clock device. This can dramatically reduce the number of devices (and associated overhead) in a complex SoC. Clocks are referenced by a number, and it is expected that SoCs will define that numbering themselves via an enum. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-06-23 21:39:15 +00:00
UCLASS_DRIVER(clk) = {
.id = UCLASS_CLK,
.name = "clk",
drivers: clk: Fix using assigned-clocks in the node of the clock it sets up This fixes the case where assigned-clocks is used to define a clock defaults inside this same clock's node. This is used sometimes to setup a default parents and/or rate for a clock. example: muxed_clock: muxed_clock { clocks = <&clk_provider 0>, <&clk_provider 1>; #clock-cells = <0>; assigned-clocks = <&muxed_clock>; assigned-clock-parents = <&clk_provider 1>; }; It doesn't work in u-boot because the assigned-clocks are setup *before* the clock is probed. (clk_set_parent() will likely crash or fail if called before the device probe function) Making it work by handling "assigned-clocks" in 2 steps: first before the clk device is probed, and then after the clk device is probed. Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
2019-10-22 12:00:06 +00:00
.post_probe = clk_uclass_post_probe,
dm: Add a clock uclass Clocks are an important feature of platforms and have become increasing complex with time. Most modern SoCs have multiple PLLs and dozens of clock dividers which distribute clocks to on-chip peripherals. Some SoC implementations have a clock API which is private to that SoC family, e.g. Tegra and Exynos. This is useful but it would be better to have a common API that can be understood and used throughout U-Boot. Add a simple clock API as a starting point. It supports querying and setting the rate of a clock. Each clock is a device. To reduce memory and processing overhead the concept of peripheral clocks is provided. These do not need to be explicit devices - it is possible to write a driver that can adjust the I2C clock (for example) without an explicit I2C clock device. This can dramatically reduce the number of devices (and associated overhead) in a complex SoC. Clocks are referenced by a number, and it is expected that SoCs will define that numbering themselves via an enum. Signed-off-by: Simon Glass <sjg@chromium.org>
2015-06-23 21:39:15 +00:00
};
Reference in a new issue Copy permalink