Skip to contentSkip to conversation

x86: Add support for sound

PATCH New Simon Glass 3 participants 51 messages Feb 17, 2019 – Feb 20, 2019

Subsystems (5)
Device tree bindings x86 PCI ASoC +1 more subsystems
lore
Share & tools
lore
#0 lore raw

v2 cover x86: Add support for sound

At present no x86 boards support sound. This series adds several options:

- legacy beep
- Intel HDA
- I2S with codec

These are enabled on coreboot, link and samus, respectively.

A number of related patches are included to make the above work.

Changes in v2:
- Add a PCI ID for the HD audio class
- Add a link to the HD audio PDF fie
- Add a note about undocumented IOMMU values
- Add a note as to why the default log level must be set to LOGL_DEBUG
- Add an example to the commit message for clarity
- Add comments next to verbs in the DT mentioning the HDA spec
- Add new patch to fix up Kconfig log level names
- Add the log-level names into the Kconfig help
- Adjust code to use #defines instead of open-coded values
- Also remove this feature from the TODO
- Change the driver name to i8254_drv
- Correct the default log level to 6 (LOGL_INFO)
- Drop useless 'set' comment
- Fix capital letter in Ivybridge in Kconfig help
- Fix comment about @return value in ioctl() method and function
- Fix ordering of PCI IDs
- Fix use of C++ comments in two places
- Move comment about silencing on error so that it is clearer
- Rename PPC_PORTB to SYSCTL_PORTB
- Rename sound driver to bd82x6x-hda
- Update in codec_init() to use defines for codec access
- Use "sandbox-pch" for the compatible string, for consistency
- Use PIT_TICK_RATE instead of creating a new define
- Use a more generic "intel,hd-audio" as the compatible string
- Use defines for all azalia verbs in binding file
- User lower-case hex consistently in the DT

Simon Glass (34):
log: Fix up Kconfig log level names
log: Add a Kconfig option to set the default log level
log: Add documentation for convenience functions
log: Allow #define LOG_DEBUG to enable logging in a file
dm: syscon: Don't require a regmap for PCI devices
i2c: designware: Add error checking on init
pci: Don't export pci_hose_config_device()
pci: Fix comment in struct pci_child_platdata
gpio: Show inactive GPIOs when explicitly requested
gpio: Use case-insentive matching on the GPIO name
pci: Add IDs for Intel high-definition audio
sandbox: Add a note about the growing state_info struct
x86: Adjust I/O macros to work on 64-bit machines
x86: samus: Increase the pre-reloc memory
x86: link: Increase malloc size and decrease code size
x86: sandbox: pch: Add a CONFIG option for PCH
sandbox: pch: Add a test for the PCH uclass
pch: Add ioctl support
x86: ivybridge: Add a way to get the HDA config setting
sound: Mark sound_setup() as optional
sound: Add uclass operations for beeping
sound: Add support for Intel HDA
sandbox: sound: Silence sound for testing
sound: x86: link: Add sound support
sound: x86: Add beeping support in i8254
sound: Add a driver for the i8254 beep
x86: coreboot: Enable the beeper sound driver
x86: broadwell: Add support for the ADSP
x86: broadwell: Don't bother probing the PCH for pinctrl
x86: broadwell: Add support for serial I/O devices
x86: sound: Add support for broadwell I2S
sound: Add a driver for RealTek RT5677
x86: sound: Add sound support for samus (broadwell)
x86: Add sound support for samus

 arch/Kconfig                                  |    2 +
 arch/sandbox/dts/test.dts                     |    5 +
 arch/sandbox/include/asm/state.h              |    9 +
 arch/sandbox/include/asm/test.h               |   24 +
 arch/x86/cpu/broadwell/Makefile               |    1 +
 arch/x86/cpu/broadwell/adsp.c                 |  156 ++
 arch/x86/cpu/broadwell/pch.c                  |  110 ++
 arch/x86/cpu/broadwell/pinctrl_broadwell.c    |    3 +-
 arch/x86/cpu/ivybridge/Kconfig                |    1 +
 arch/x86/cpu/ivybridge/bd82x6x.c              |   27 +-
 arch/x86/cpu/ivybridge/northbridge.c          |   32 +
 arch/x86/dts/chromebook_link.dts              |   96 ++
 arch/x86/dts/chromebook_samus.dts             |   49 +-
 arch/x86/include/asm/arch-broadwell/adsp.h    |   46 +
 arch/x86/include/asm/arch-broadwell/pch.h     |    3 +
 arch/x86/include/asm/arch-broadwell/rcb.h     |    2 +
 .../x86/include/asm/arch-broadwell/serialio.h |   82 +
 .../include/asm/arch-ivybridge/sandybridge.h  |    3 +
 arch/x86/include/asm/i8254.h                  |   18 +
 arch/x86/include/asm/io.h                     |   16 +-
 arch/x86/lib/i8254.c                          |   35 +-
 cmd/gpio.c                                    |   10 +-
 common/Kconfig                                |   86 +-
 common/log.c                                  |    2 +-
 configs/chromebook_link_defconfig             |    5 +-
 configs/chromebook_samus_defconfig            |   11 +-
 configs/coreboot_defconfig                    |    3 +
 doc/README.log                                |   51 +-
 doc/device-tree-bindings/sound/intel-hda.txt  |   25 +
 drivers/Kconfig                               |    2 +
 drivers/Makefile                              |    3 +-
 drivers/core/syscon-uclass.c                  |    4 +
 drivers/i2c/designware_i2c.c                  |   29 +-
 drivers/pch/Kconfig                           |    9 +
 drivers/pch/Makefile                          |    1 +
 drivers/pch/pch-uclass.c                      |   10 +
 drivers/pch/sandbox_pch.c                     |   86 +
 drivers/pci/pci.c                             |    7 +-
 drivers/sound/Kconfig                         |   40 +
 drivers/sound/Makefile                        |    5 +
 drivers/sound/broadwell_i2s.c                 |  306 ++++
 drivers/sound/broadwell_i2s.h                 |  301 ++++
 drivers/sound/broadwell_sound.c               |   65 +
 drivers/sound/hda_codec.c                     |  556 +++++++
 drivers/sound/i8254_beep.c                    |   38 +
 drivers/sound/ivybridge_sound.c               |  137 ++
 drivers/sound/rt5677.c                        |  334 ++++
 drivers/sound/rt5677.h                        | 1428 +++++++++++++++++
 drivers/sound/sandbox.c                       |   71 +-
 drivers/sound/sound-uclass.c                  |   36 +-
 include/dt-bindings/sound/azalia.h            |   44 +
 include/hda_codec.h                           |  103 ++
 include/log.h                                 |    9 +-
 include/pch.h                                 |   51 +-
 include/pci.h                                 |    8 +-
 include/pci_ids.h                             |    5 +
 include/sound.h                               |   46 +-
 test/dm/Makefile                              |    1 +
 test/dm/pch.c                                 |   55 +
 test/dm/sound.c                               |   21 +
 60 files changed, 4623 insertions(+), 101 deletions(-)
 create mode 100644 arch/x86/cpu/broadwell/adsp.c
 create mode 100644 arch/x86/include/asm/arch-broadwell/adsp.h
 create mode 100644 arch/x86/include/asm/arch-broadwell/serialio.h
 create mode 100644 doc/device-tree-bindings/sound/intel-hda.txt
 create mode 100644 drivers/pch/Kconfig
 create mode 100644 drivers/pch/sandbox_pch.c
 create mode 100644 drivers/sound/broadwell_i2s.c
 create mode 100644 drivers/sound/broadwell_i2s.h
 create mode 100644 drivers/sound/broadwell_sound.c
 create mode 100644 drivers/sound/hda_codec.c
 create mode 100644 drivers/sound/i8254_beep.c
 create mode 100644 drivers/sound/ivybridge_sound.c
 create mode 100644 drivers/sound/rt5677.c
 create mode 100644 drivers/sound/rt5677.h
 create mode 100644 include/dt-bindings/sound/azalia.h
 create mode 100644 include/hda_codec.h
 create mode 100644 test/dm/pch.c
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#1 lore raw

PATCH v2 1/34 log: Fix up Kconfig log level names

The log level numbers in the Kconfig are not actually correct. Fix them and also add a missing space in the header-file comment.

---

Changes in v2:
- Add new patch to fix up Kconfig log level names

 common/Kconfig | 66 ++++++++++++++++++++++++++++----------------------
 include/log.h  |  2 +-
 2 files changed, 38 insertions(+), 30 deletions(-)
common/Kconfig +37 -29
index 0a14bdedaa6..849d8ff90ad 100644
@@ -286,14 +286,16 @@ config LOGLEVEL
 	  All Messages with a loglevel smaller than the console loglevel will
 	  be compiled in. The loglevels are defined as follows:
 
-	  0 (KERN_EMERG)          system is unusable
-	  1 (KERN_ALERT)          action must be taken immediately
-	  2 (KERN_CRIT)           critical conditions
-	  3 (KERN_ERR)            error conditions
-	  4 (KERN_WARNING)        warning conditions
-	  5 (KERN_NOTICE)         normal but significant condition
-	  6 (KERN_INFO)           informational
-	  7 (KERN_DEBUG)          debug-level messages
+	    0 - emergency
+	    1 - alert
+	    2 - critical
+	    3 - error
+	    4 - warning
+	    5 - note
+	    6 - info
+	    7 - debug
+	    8 - debug content
+	    9 - debug hardware I/O
 
 config SPL_LOGLEVEL
 	int
@@ -482,14 +484,16 @@ config LOG_MAX_LEVEL
 	  higher than this will be ignored. If possible log statements below
 	  this level will be discarded at build time. Levels:
 
-	    0 - panic
-	    1 - critical
-	    2 - error
-	    3 - warning
-	    4 - note
-	    5 - info
-	    6 - detail
+	    0 - emergency
+	    1 - alert
+	    2 - critical
+	    3 - error
+	    4 - warning
+	    5 - note
+	    6 - info
 	    7 - debug
+	    8 - debug content
+	    9 - debug hardware I/O
 
 config SPL_LOG_MAX_LEVEL
 	int "Maximum log level to record in SPL"
@@ -500,14 +504,16 @@ config SPL_LOG_MAX_LEVEL
 	  higher than this will be ignored. If possible log statements below
 	  this level will be discarded at build time. Levels:
 
-	    0 - panic
-	    1 - critical
-	    2 - error
-	    3 - warning
-	    4 - note
-	    5 - info
-	    6 - detail
+	    0 - emergency
+	    1 - alert
+	    2 - critical
+	    3 - error
+	    4 - warning
+	    5 - note
+	    6 - info
 	    7 - debug
+	    8 - debug content
+	    9 - debug hardware I/O
 
 config TPL_LOG_MAX_LEVEL
 	int "Maximum log level to record in TPL"
@@ -518,14 +524,16 @@ config TPL_LOG_MAX_LEVEL
 	  higher than this will be ignored. If possible log statements below
 	  this level will be discarded at build time. Levels:
 
-	    0 - panic
-	    1 - critical
-	    2 - error
-	    3 - warning
-	    4 - note
-	    5 - info
-	    6 - detail
+	    0 - emergency
+	    1 - alert
+	    2 - critical
+	    3 - error
+	    4 - warning
+	    5 - note
+	    6 - info
 	    7 - debug
+	    8 - debug content
+	    9 - debug hardware I/O
 
 config LOG_CONSOLE
 	bool "Allow log output to the console"
include/log.h +1 -1
index d7f64710061..33e99ab7030 100644
@@ -14,7 +14,7 @@
 
 /** Log levels supported, ranging from most to least important */
 enum log_level_t {
-	LOGL_EMERG = 0,		/*U-Boot is unstable */
+	LOGL_EMERG = 0,		/* U-Boot is unstable */
 	LOGL_ALERT,		/* Action must be taken immediately */
 	LOGL_CRIT,		/* Critical conditions */
 	LOGL_ERR,		/* Error that prevents something from working */
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#2 lore raw

PATCH v2 1/34 log: Fix up Kconfig log level names

Quoted 5 lines · Simon Glass

The log level numbers in the Kconfig are not actually correct. Fix them and also add a missing space in the header-file comment.

Quoted 9 lines

---

Changes in v2:
- Add new patch to fix up Kconfig log level names

 common/Kconfig | 66 ++++++++++++++++++++++++++++----------------------
 include/log.h  |  2 +-
 2 files changed, 38 insertions(+), 30 deletions(-)
#3 lore raw

PATCH v2 2/34 log: Add a Kconfig option to set the default log level

In reply to Simon Glass

At present the default log level is set to LOGL_INFO on start-up. Allow this to be controlled from Kconfig.

---

Changes in v2:
- Add the log-level names into the Kconfig help
- Correct the default log level to 6 (LOGL_INFO)

 common/Kconfig | 20 ++++++++++++++++++++
 common/log.c   |  2 +-
 2 files changed, 21 insertions(+), 1 deletion(-)
common/Kconfig +20 -0
index 849d8ff90ad..e100c229496 100644
@@ -535,6 +535,26 @@ config TPL_LOG_MAX_LEVEL
 	    8 - debug content
 	    9 - debug hardware I/O
 
+config LOG_DEFAULT_LEVEL
+	int "Default logging level to display"
+	default 6
+	help
+	  This is the default logging level set when U-Boot starts. It can
+	  be adjusted later using the 'log level' command. Note that setting
+	  this to a value abnove LOG_MAX_LEVEL will be ineffective, since the
+	  higher levels are not compiled in to U-Boot.
+
+	    0 - emergency
+	    1 - alert
+	    2 - critical
+	    3 - error
+	    4 - warning
+	    5 - note
+	    6 - info
+	    7 - debug
+	    8 - debug content
+	    9 - debug hardware I/O
+
 config LOG_CONSOLE
 	bool "Allow log output to the console"
 	depends on LOG
common/log.c +1 -1
index ec14644516c..ffb3cd69332 100644
@@ -316,7 +316,7 @@ int log_init(void)
 	}
 	gd->flags |= GD_FLG_LOG_READY;
 	if (!gd->default_log_level)
-		gd->default_log_level = LOGL_INFO;
+		gd->default_log_level = CONFIG_LOG_DEFAULT_LEVEL;
 	gd->log_fmt = LOGF_DEFAULT;
 
 	return 0;
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#4 lore raw

PATCH v2 3/34 log: Add documentation for convenience functions

In reply to Simon Glass

The log_debug(), etc. function are documented only in the header file at present. Add a section to README.log also.

---

Changes in v2:
- Also remove this feature from the TODO

 doc/README.log | 36 ++++++++++++++++++++++++++----------
 1 file changed, 26 insertions(+), 10 deletions(-)
doc/README.log +26 -10
index 2ee1b753bc7..75350ecd41d 100644
@@ -69,6 +69,32 @@ If CONFIG_LOG is not set, then no logging will be available.
 The above have SPL versions also, e.g. CONFIG_SPL_MAX_LOG_LEVEL.
 
 
+Convenience functions
+---------------------
+
+A number of convenience functions are available to shorten the code needed
+for logging:
+
+	log_err(_fmt...)
+	log_warning(_fmt...)
+	log_notice(_fmt...)
+	log_info(_fmt...)
+	log_debug(_fmt...)
+	log_content(_fmt...)
+	log_io(_fmt...)
+
+With these the log level is implicit in the name. The category is set by
+LOG_CATEGORY, which you can only define once per file, above all #includes:
+
+	#define LOG_CATEGORY LOGC_ALLOC
+
+or
+
+	#define LOG_CATEGORY UCLASS_SPI
+
+Remember that all uclasses IDs are log categories too.
+
+
 Log commands
 ------------
 
@@ -187,16 +213,6 @@ Convenience functions to support setting the category:
    log_core(level, format_string, ...) - category LOGC_CORE
    log_dt(level, format_string, ...) - category LOGC_DT
 
-Convenience functions to support a category defined for a single file, for
-example:
-
-   #define LOG_CATEGORY   UCLASS_USB
-
-all of these can use LOG_CATEGORY as the category, and a log level
-corresponding to the function name:
-
-   logc(level, format_string, ...)
-
 More logging destinations:
 
    device - goes to a device (e.g. serial)
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#5 lore raw

PATCH v2 4/34 log: Allow #define LOG_DEBUG to enable logging in a file

In reply to Simon Glass

At present it is possible to '#define DEBUG' at the top of a file which causes all debug() statements in that file to become active. There is currently no equivalent with logging, but this is a useful function.

Add a LOG_DEBUG define along with documentation.

---

Changes in v2:
- Add a note as to why the default log level must be set to LOGL_DEBUG

 doc/README.log | 15 +++++++++++++++
 include/log.h  |  7 ++++++-
 2 files changed, 21 insertions(+), 1 deletion(-)
doc/README.log +15 -0
index 75350ecd41d..19856d43da2 100644
@@ -69,6 +69,21 @@ If CONFIG_LOG is not set, then no logging will be available.
 The above have SPL versions also, e.g. CONFIG_SPL_MAX_LOG_LEVEL.
 
 
+Temporary logging within a single file
+--------------------------------------
+
+Sometimes it is useful to turn on logging just in one file. You can use this:
+
+   #define LOG_DEBUG
+
+to enable building in of all logging statements in a single file. Put it at
+the top of the file, before any #includes.
+
+To actually get U-Boot to output this you need to also set the default logging
+level - e.g. set CONFIG_LOG_DEFAULT_LEVEL to 7 (LOGL_DEBUG) or more. Otherwise
+debug output is suppressed and will not be generated.
+
+
 Convenience functions
 ---------------------
 
include/log.h +6 -1
index 33e99ab7030..7566ba7f2db 100644
@@ -111,11 +111,16 @@ int _log(enum log_category_t cat, enum log_level_t level, const char *file,
 #endif
 
 #if CONFIG_IS_ENABLED(LOG)
+#ifdef LOG_DEBUG
+#define _LOG_DEBUG	1
+#else
+#define _LOG_DEBUG	0
+#endif
 
 /* Emit a log record if the level is less that the maximum */
 #define log(_cat, _level, _fmt, _args...) ({ \
 	int _l = _level; \
-	if (CONFIG_IS_ENABLED(LOG) && _l <= _LOG_MAX_LEVEL) \
+	if (CONFIG_IS_ENABLED(LOG) && (_l <= _LOG_MAX_LEVEL || _LOG_DEBUG)) \
 		_log((enum log_category_t)(_cat), _l, __FILE__, __LINE__, \
 		      __func__, \
 		      pr_fmt(_fmt), ##_args); \
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#6 lore raw

PATCH v2 4/34 log: Allow #define LOG_DEBUG to enable logging in a file

Quoted 8 lines · Simon Glass

At present it is possible to '#define DEBUG' at the top of a file which causes all debug() statements in that file to become active. There is currently no equivalent with logging, but this is a useful function.

Add a LOG_DEBUG define along with documentation.

Quoted 9 lines

---

Changes in v2:
- Add a note as to why the default log level must be set to LOGL_DEBUG

 doc/README.log | 15 +++++++++++++++
 include/log.h  |  7 ++++++-
 2 files changed, 21 insertions(+), 1 deletion(-)
#7 lore raw

PATCH v2 5/34 dm: syscon: Don't require a regmap for PCI devices

In reply to Simon Glass

At present it is not possible to use the syscon devices for PCI devices since a regmap is required. Since PCI uses a 3-cell address the conversion of the 'reg' property to an address always fails. In any case, the regmap is not useful with PCI since devices are accessed through the PCI bus which regmap does not support.

Add a special case for PCI syscon devices, so that they don't set up a regmap.

---

Changes in v2: None

 drivers/core/syscon-uclass.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/core/syscon-uclass.c +4 -0
index 3cf9dd9dbe4..afac6d6e37e 100644
@@ -37,6 +37,10 @@ static int syscon_pre_probe(struct udevice *dev)
 {
 	struct syscon_uc_info *priv = dev_get_uclass_priv(dev);
 
+	/* Special case for PCI devices, which don't have a regmap */
+	if (device_get_uclass_id(dev->parent) == UCLASS_PCI)
+		return 0;
+
 	/*
 	 * With OF_PLATDATA we really have no way of knowing the format of
 	 * the device-specific platform data. So we assume that it starts with
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#8 lore raw

PATCH v2 6/34 i2c: designware: Add error checking on init

In reply to Simon Glass

At present this driver does not check whether it is able to actually communicate with the I2C controller. It prints a timeout message but still considers the probe to be successful.

To fix this, add some checking that the init succeeds.

---

Changes in v2: None

 drivers/i2c/designware_i2c.c | 29 +++++++++++++++++++----------
 1 file changed, 19 insertions(+), 10 deletions(-)
drivers/i2c/designware_i2c.c +19 -10
index dbc3326b5a9..63e40823f12 100644
@@ -38,14 +38,16 @@ struct dw_i2c {
 };
 
 #ifdef CONFIG_SYS_I2C_DW_ENABLE_STATUS_UNSUPPORTED
-static void dw_i2c_enable(struct i2c_regs *i2c_base, bool enable)
+static int  dw_i2c_enable(struct i2c_regs *i2c_base, bool enable)
 {
 	u32 ena = enable ? IC_ENABLE_0B : 0;
 
 	writel(ena, &i2c_base->ic_enable);
+
+	return 0;
 }
 #else
-static void dw_i2c_enable(struct i2c_regs *i2c_base, bool enable)
+static int dw_i2c_enable(struct i2c_regs *i2c_base, bool enable)
 {
 	u32 ena = enable ? IC_ENABLE_0B : 0;
 	int timeout = 100;
@@ -53,7 +55,7 @@ static void dw_i2c_enable(struct i2c_regs *i2c_base, bool enable)
 	do {
 		writel(ena, &i2c_base->ic_enable);
 		if ((readl(&i2c_base->ic_enable_status) & IC_ENABLE_0B) == ena)
-			return;
+			return 0;
 
 		/*
 		 * Wait 10 times the signaling period of the highest I2C
@@ -62,8 +64,9 @@ static void dw_i2c_enable(struct i2c_regs *i2c_base, bool enable)
 		 */
 		udelay(25);
 	} while (timeout--);
-
 	printf("timeout in %sabling I2C adapter\n", enable ? "en" : "dis");
+
+	return -ETIMEDOUT;
 }
 #endif
 
@@ -370,10 +373,14 @@ static int __dw_i2c_write(struct i2c_regs *i2c_base, u8 dev, uint addr,
  *
  * Initialization function.
  */
-static void __dw_i2c_init(struct i2c_regs *i2c_base, int speed, int slaveaddr)
+static int __dw_i2c_init(struct i2c_regs *i2c_base, int speed, int slaveaddr)
 {
+	int ret;
+
 	/* Disable i2c */
-	dw_i2c_enable(i2c_base, false);
+	ret = dw_i2c_enable(i2c_base, false);
+	if (ret)
+		return ret;
 
 	writel(IC_CON_SD | IC_CON_RE | IC_CON_SPD_FS | IC_CON_MM,
 	       &i2c_base->ic_con);
@@ -386,7 +393,11 @@ static void __dw_i2c_init(struct i2c_regs *i2c_base, int speed, int slaveaddr)
 #endif
 
 	/* Enable i2c */
-	dw_i2c_enable(i2c_base, true);
+	ret = dw_i2c_enable(i2c_base, true);
+	if (ret)
+		return ret;
+
+	return 0;
 }
 
 #ifndef CONFIG_DM_I2C
@@ -558,9 +569,7 @@ static int designware_i2c_probe(struct udevice *bus)
 	if (&priv->reset_ctl)
 		reset_deassert(&priv->reset_ctl);
 
-	__dw_i2c_init(priv->regs, 0, 0);
-
-	return 0;
+	return __dw_i2c_init(priv->regs, 0, 0);
 }
 
 static int designware_i2c_bind(struct udevice *dev)
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#9 lore raw

PATCH v2 6/34 i2c: designware: Add error checking on init

Hello Simon,

Am 17.02.2019 um 04:24 schrieb Simon Glass:

Quoted 6 lines

At present this driver does not check whether it is able to actually communicate with the I2C controller. It prints a timeout message but still considers the probe to be successful.

To fix this, add some checking that the init succeeds.

Quoted 8 lines

---

Changes in v2: None

  drivers/i2c/designware_i2c.c | 29 +++++++++++++++++++----------
  1 file changed, 19 insertions(+), 10 deletions(-)

bye,
Heiko

Signature · 3 lines
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-52   Fax: +49-8142-66989-80   Email: hs at denx.de
#10 lore raw

PATCH v2 7/34 pci: Don't export pci_hose_config_device()

In reply to Simon Glass

This function is not used outside this file so make it static.

---

Changes in v2: None

 drivers/pci/pci.c | 7 ++-----
 include/pci.h     | 6 ------
 2 files changed, 2 insertions(+), 11 deletions(-)
drivers/pci/pci.c +2 -5
index 221054593ef..e2195726c8a 100644
@@ -184,11 +184,8 @@ pci_dev_t pci_find_devices(struct pci_device_id *ids, int index)
 	return -1;
 }
 
-int pci_hose_config_device(struct pci_controller *hose,
-			   pci_dev_t dev,
-			   unsigned long io,
-			   pci_addr_t mem,
-			   unsigned long command)
+static int pci_hose_config_device(struct pci_controller *hose, pci_dev_t dev,
+				  ulong io, pci_addr_t mem, ulong command)
 {
 	u32 bar_response;
 	unsigned int old_command;
include/pci.h +0 -6
index 041f8e37476..dd5f051989d 100644
@@ -735,12 +735,6 @@ extern pci_dev_t pci_find_device (unsigned int vendor, unsigned int device, int
 extern pci_dev_t pci_find_devices (struct pci_device_id *ids, int index);
 pci_dev_t pci_find_class(unsigned int find_class, int index);
 
-extern int pci_hose_config_device(struct pci_controller *hose,
-				  pci_dev_t dev,
-				  unsigned long io,
-				  pci_addr_t mem,
-				  unsigned long command);
-
 extern int pci_hose_find_capability(struct pci_controller *hose, pci_dev_t dev,
 				    int cap);
 extern int pci_hose_find_cap_start(struct pci_controller *hose, pci_dev_t dev,
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#11 lore raw

PATCH v2 8/34 pci: Fix comment in struct pci_child_platdata

In reply to Simon Glass

This is platdata, not private data, so the comment is currently incorrect. Fix it to avoid confusion.

---

Changes in v2: None

 include/pci.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
include/pci.h +1 -1
index dd5f051989d..936cfe975cb 100644
@@ -822,7 +822,7 @@ struct udevice;
  *
  * Every device on a PCI bus has this per-child data.
  *
- * It can be accessed using dev_get_parent_priv(dev) if dev->parent is a
+ * It can be accessed using dev_get_parent_platdata(dev) if dev->parent is a
  * PCI bus (i.e. UCLASS_PCI)
  *
  * @devfn:	Encoded device and function index - see PCI_DEVFN()
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#12 lore raw

PATCH v2 9/34 gpio: Show inactive GPIOs when explicitly requested

In reply to Simon Glass

At present the gpio command only shows GPIOs which are marked as in use. This makes sense with 'gpio status' since we already have the '-a' flag to indicate that all GPIOs should be shown. But when a particular GPIO is requested, it seems better to always display it. At present the request is simply ignored.

For example if GPIO a10 is not in use, then:

Quoted 1 line

gpio a10

shows nothing, not even the function being used for that GPIO. With this change, it shows the pin status:

Quoted 1 line

gpio a10

a10: input: 0 [ ]

Add an extra parameter for this to avoid changing the existing flag parameter.

---

Changes in v2:
- Add an example to the commit message for clarity

 cmd/gpio.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
cmd/gpio.c +4 -4
index 4ac1f1e4180..539e07ee077 100644
@@ -34,7 +34,7 @@ enum {
 };
 
 static void gpio_get_description(struct udevice *dev, const char *bank_name,
-				 int offset, int *flagsp)
+				 int offset, int *flagsp, bool show_all)
 {
 	char buf[80];
 	int ret;
@@ -42,7 +42,7 @@ static void gpio_get_description(struct udevice *dev, const char *bank_name,
 	ret = gpio_get_function(dev, offset, NULL);
 	if (ret < 0)
 		goto err;
-	if (!(*flagsp & FLAG_SHOW_ALL) && ret == GPIOF_UNUSED)
+	if (!show_all && !(*flagsp & FLAG_SHOW_ALL) && ret == GPIOF_UNUSED)
 		return;
 	if ((*flagsp & FLAG_SHOW_BANK) && bank_name) {
 		if (*flagsp & FLAG_SHOW_NEWLINE) {
@@ -98,11 +98,11 @@ static int do_gpio_status(bool all, const char *gpio_name)
 			if (gpio_name && *p) {
 				offset = simple_strtoul(p, NULL, 10);
 				gpio_get_description(dev, bank_name, offset,
-						     &flags);
+						     &flags, true);
 			} else {
 				for (offset = 0; offset < num_bits; offset++) {
 					gpio_get_description(dev, bank_name,
-							     offset, &flags);
+						     offset, &flags, false);
 				}
 			}
 		}
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#13 lore raw

PATCH v2 9/34 gpio: Show inactive GPIOs when explicitly requested

Hi Simon,

Quoted 10 lines · Simon Glass

At present the gpio command only shows GPIOs which are marked as in use. This makes sense with 'gpio status' since we already have the '-a' flag to indicate that all GPIOs should be shown. But when a particular GPIO is requested, it seems better to always display it. At present the request is simply ignored.

For example if GPIO a10 is not in use, then:

Quoted 1 line

gpio a10

I think this should be "gpio status a10".

Quoted 5 lines

shows nothing, not even the function being used for that GPIO. With this change, it shows the pin status:

Quoted 1 line

gpio a10

ditto

Quoted 5 lines
   a10: input: 0 [ ]

Add an extra parameter for this to avoid changing the existing flag parameter.

Quoted 9 lines

---

Changes in v2:
- Add an example to the commit message for clarity

 cmd/gpio.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

I can fix that when applying if you don't mind.

Regards,
Bin

#14 lore raw

PATCH v2 9/34 gpio: Show inactive GPIOs when explicitly requested

Quoted 36 lines · Bin Meng

Hi Simon,

Quoted 10 lines · Simon Glass

At present the gpio command only shows GPIOs which are marked as in use. This makes sense with 'gpio status' since we already have the '-a' flag to indicate that all GPIOs should be shown. But when a particular GPIO is requested, it seems better to always display it. At present the request is simply ignored.

For example if GPIO a10 is not in use, then:

Quoted 1 line

gpio a10

I think this should be "gpio status a10".

Quoted 5 lines

shows nothing, not even the function being used for that GPIO. With this change, it shows the pin status:

Quoted 1 line

gpio a10

ditto

Quoted 11 lines
   a10: input: 0 [ ]

Add an extra parameter for this to avoid changing the existing flag parameter.

---

Changes in v2:
- Add an example to the commit message for clarity

Quoted 8 lines
Quoted 3 lines
 cmd/gpio.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

I can fix that when applying if you don't mind.

Yes please.

Quoted 3 lines

Regards,
Bin

#15 lore raw

PATCH v2 10/34 gpio: Use case-insentive matching on the GPIO name

In reply to Simon Glass

Allow the 'gpio' command to match GPIO bank names regardless of the case of each. While these are generally in upper case, it is useful to be able to provide lower case with the command.

---

Changes in v2: None

 cmd/gpio.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
cmd/gpio.c +1 -1
index 539e07ee077..53366f36e76 100644
@@ -90,7 +90,7 @@ static int do_gpio_status(bool all, const char *gpio_name)
 		banklen = bank_name ? strlen(bank_name) : 0;
 
 		if (!gpio_name || !bank_name ||
-		    !strncmp(gpio_name, bank_name, banklen)) {
+		    !strncasecmp(gpio_name, bank_name, banklen)) {
 			const char *p = NULL;
 			int offset;
 
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#16 lore raw

PATCH v2 11/34 pci: Add IDs for Intel high-definition audio

In reply to Simon Glass

Add a few IDs for common HDA blocks and the ADSP used on samus.

---

Changes in v2:
- Add a PCI ID for the HD audio class
- Fix ordering of PCI IDs

 include/pci_ids.h | 5 +++++
 1 file changed, 5 insertions(+)
include/pci_ids.h +5 -0
index fdda679cc05..bd59578ccb9 100644
@@ -41,6 +41,7 @@
 #define PCI_CLASS_MULTIMEDIA_VIDEO	0x0400
 #define PCI_CLASS_MULTIMEDIA_AUDIO	0x0401
 #define PCI_CLASS_MULTIMEDIA_PHONE	0x0402
+#define PCI_CLASS_MULTIMEDIA_HD_AUDIO	0x0403
 #define PCI_CLASS_MULTIMEDIA_OTHER	0x0480
 
 #define PCI_BASE_CLASS_MEMORY		0x05
@@ -1363,6 +1364,7 @@
 #define PCI_DEVICE_ID_CREATIVE_EMU10K1	0x0002
 #define PCI_DEVICE_ID_CREATIVE_20K1	0x0005
 #define PCI_DEVICE_ID_CREATIVE_20K2	0x000b
+#define PCI_DEVICE_ID_CREATIVE_CA01322	0x0011
 #define PCI_SUBDEVICE_ID_CREATIVE_SB0760	0x0024
 #define PCI_SUBDEVICE_ID_CREATIVE_SB08801	0x0041
 #define PCI_SUBDEVICE_ID_CREATIVE_SB08802	0x0042
@@ -2827,6 +2829,7 @@
 #define PCI_DEVICE_ID_INTEL_ICH7_19	0x27dd
 #define PCI_DEVICE_ID_INTEL_ICH7_20	0x27de
 #define PCI_DEVICE_ID_INTEL_ICH7_21	0x27df
+#define PCI_DEVICE_ID_INTEL_COUGARPOINT_HDMI	0x2806
 #define PCI_DEVICE_ID_INTEL_ICH8_0	0x2810
 #define PCI_DEVICE_ID_INTEL_ICH8_1	0x2811
 #define PCI_DEVICE_ID_INTEL_ICH8_2	0x2812
@@ -3025,6 +3028,8 @@
 #define PCI_DEVICE_ID_INTEL_LYNXPOINT_AHCI	0x9c03
 #define PCI_DEVICE_ID_INTEL_LYNXPOINT_LPC	0x9c45
 #define PCI_DEVICE_ID_INTEL_WILDCATPOINT_AHCI	0x9c83
+#define PCI_DEVICE_ID_INTEL_WILDCATPOINT_HDA	0x9ca0
+#define PCI_DEVICE_ID_INTEL_WILDCATPOINT_ADSP	0x9cb6
 #define PCI_DEVICE_ID_INTEL_WILDCATPOINT_LPC	0x9cc3
 #define PCI_DEVICE_ID_INTEL_S21152BB	0xb152
 
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#17 lore raw

PATCH v2 11/34 pci: Add IDs for Intel high-definition audio

Quoted 4 lines · Simon Glass

Add a few IDs for common HDA blocks and the ADSP used on samus.

Quoted 9 lines

---

Changes in v2:
- Add a PCI ID for the HD audio class
- Fix ordering of PCI IDs

 include/pci_ids.h | 5 +++++
 1 file changed, 5 insertions(+)
#18 lore raw

PATCH v2 12/34 sandbox: Add a note about the growing state_info struct

In reply to Simon Glass

This struct is getting larger and in some cases is being used for things which would be better put into a driver. For example hwspinlock is not used outside of sandbox_hwspinlock.c.

Add a note to encourage people to put things elsewhere.

---

Changes in v2: None

 arch/sandbox/include/asm/state.h | 9 +++++++++
 1 file changed, 9 insertions(+)
arch/sandbox/include/asm/state.h +9 -0
index c724827f6cf..2d773d3fa6b 100644
@@ -102,6 +102,15 @@ struct sandbox_state {
 	ulong next_tag;			/* Next address tag to allocate */
 	struct list_head mapmem_head;	/* struct sandbox_mapmem_entry */
 	bool hwspinlock;		/* Hardware Spinlock status */
+
+	/*
+	 * This struct is getting large.
+	 *
+	 * Consider putting test data in driver-private structs, like
+	 * sandbox_pch.c.
+	 *
+	 * If you add new members, please put them above this comment.
+	 */
 };
 
 /* Minimum space we guarantee in the state FDT when calling read/write*/
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#19 lore raw

PATCH v2 13/34 x86: Adjust I/O macros to work on 64-bit machines

In reply to Simon Glass

At present these macros give warnings on 64-bit machines and do not correctly do 32-bit accesses. Update them to use linux types.

---

Changes in v2: None

 arch/x86/include/asm/io.h | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)
arch/x86/include/asm/io.h +8 -8
index 81def0afd30..cf6c33cbe3a 100644
@@ -57,19 +57,19 @@
  * memory location directly.
  */
 
-#define readb(addr) (*(volatile unsigned char *) (addr))
-#define readw(addr) (*(volatile unsigned short *) (addr))
-#define readl(addr) (*(volatile unsigned int *) (addr))
-#define readq(addr) (*(volatile unsigned long long *) (addr))
+#define readb(addr) (*(volatile u8 *)(uintptr_t)(addr))
+#define readw(addr) (*(volatile u16 *)(uintptr_t)(addr))
+#define readl(addr) (*(volatile u32 *)(uintptr_t)(addr))
+#define readq(addr) (*(volatile u64 *)(uintptr_t)(addr))
 #define __raw_readb readb
 #define __raw_readw readw
 #define __raw_readl readl
 #define __raw_readq readq
 
-#define writeb(b,addr) (*(volatile unsigned char *) (addr) = (b))
-#define writew(b,addr) (*(volatile unsigned short *) (addr) = (b))
-#define writel(b,addr) (*(volatile unsigned int *) (addr) = (b))
-#define writeq(b,addr) (*(volatile unsigned long long *) (addr) = (b))
+#define writeb(b, addr) (*(volatile u8 *)(addr) = (b))
+#define writew(b, addr) (*(volatile u16 *)(addr) = (b))
+#define writel(b, addr) (*(volatile u32 *)(addr) = (b))
+#define writeq(b, addr) (*(volatile u64 *)(addr) = (b))
 #define __raw_writeb writeb
 #define __raw_writew writew
 #define __raw_writel writel
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#20 lore raw

PATCH v2 14/34 x86: samus: Increase the pre-reloc memory

In reply to Simon Glass

This is too small now, likely due to the pre_reloc_only fixes. Increase it.

---

Changes in v2: None

 configs/chromebook_samus_defconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
configs/chromebook_samus_defconfig +1 -1
index e5483c59205..745cb3db0a6 100644
@@ -1,6 +1,6 @@
 CONFIG_X86=y
 CONFIG_SYS_TEXT_BASE=0xFFE00000
-CONFIG_SYS_MALLOC_F_LEN=0x1a00
+CONFIG_SYS_MALLOC_F_LEN=0x1c00
 CONFIG_DEBUG_UART_BOARD_INIT=y
 CONFIG_DEBUG_UART_BASE=0x3f8
 CONFIG_DEBUG_UART_CLOCK=1843200
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#21 lore raw

PATCH v2 15/34 x86: link: Increase malloc size and decrease code size

In reply to Simon Glass

At present link does not boot since it has outgrown its pre-relocation malloc() size and its assigned code area. Increase the former and drop EFI loader support, which adds about 45KB!

---

Changes in v2: None

 configs/chromebook_link_defconfig | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
configs/chromebook_link_defconfig +2 -1
index b90aae610b2..c402568c17e 100644
@@ -1,6 +1,6 @@
 CONFIG_X86=y
 CONFIG_SYS_TEXT_BASE=0xFFF00000
-CONFIG_SYS_MALLOC_F_LEN=0x2000
+CONFIG_SYS_MALLOC_F_LEN=0x2400
 CONFIG_DEBUG_UART_BOARD_INIT=y
 CONFIG_DEBUG_UART_BASE=0x3f8
 CONFIG_DEBUG_UART_CLOCK=1843200
@@ -63,3 +63,4 @@ CONFIG_VIDEO_IVYBRIDGE_IGD=y
 CONFIG_CONSOLE_SCROLL_LINES=5
 CONFIG_CMD_DHRYSTONE=y
 CONFIG_TPM=y
+# CONFIG_EFI_LOADER is not set
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#22 lore raw

PATCH v2 16/34 x86: sandbox: pch: Add a CONFIG option for PCH

In reply to Simon Glass

At present this uclass is selected only on x86. In order to add a test for it, it must also support sandbox. Create a new CONFIG_PCH option and enable it on x86 and sandbox.

---

Changes in v2: None

 arch/Kconfig        | 2 ++
 drivers/Kconfig     | 2 ++
 drivers/Makefile    | 3 ++-
 drivers/pch/Kconfig | 9 +++++++++
 4 files changed, 15 insertions(+), 1 deletion(-)
 create mode 100644 drivers/pch/Kconfig
arch/Kconfig +2 -0
index 35e2712fce9..d9afe269f20 100644
@@ -117,6 +117,7 @@ config SANDBOX
 	imply VIRTIO_BLK
 	imply VIRTIO_NET
 	imply DM_SOUND
+	imply PCH
 
 config SH
 	bool "SuperH architecture"
@@ -160,6 +161,7 @@ config X86
 	imply USB_ETHER_ASIX
 	imply USB_ETHER_SMSC95XX
 	imply USB_HOST_ETHER
+	imply PCH
 
 config XTENSA
 	bool "Xtensa architecture"
drivers/Kconfig +2 -0
index e9fbadd13d5..f24351ac4f1 100644
@@ -64,6 +64,8 @@ source "drivers/nvme/Kconfig"
 
 source "drivers/pci/Kconfig"
 
+source "drivers/pch/Kconfig"
+
 source "drivers/pcmcia/Kconfig"
 
 source "drivers/phy/Kconfig"
drivers/Makefile +2 -1
index eca023ac04d..a7bba3ed564 100644
@@ -86,7 +86,8 @@ obj-y += misc/
 obj-$(CONFIG_MMC) += mmc/
 obj-$(CONFIG_NVME) += nvme/
 obj-y += pcmcia/
-obj-$(CONFIG_X86) += pch/
+obj-y += dfu/
+obj-$(CONFIG_PCH) += pch/
 obj-y += phy/allwinner/
 obj-y += phy/marvell/
 obj-y += rtc/
drivers/pch/Kconfig new file +9 -0
new file mode 100644
index 00000000000..18f006de24c
@@ -0,0 +1,9 @@
+config PCH
+	bool "Enable Platform-controller Hub (PCH) support"
+	depends on X86 || SANDBOX
+	help
+	  Most x86 chips include a PCH which is responsible for handling
+	  parts of the system not handled by that CPU. It supersedes the
+	  northbridge / southbridge architecture that was previously used. The
+	  PCH allows for higher performance since the memory functions are
+	  handled in the CPU.
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#23 lore raw

PATCH v2 17/34 sandbox: pch: Add a test for the PCH uclass

In reply to Simon Glass

This uclass currently has no tests. Add a sandbox driver and some simple tests to provide basic coverage.

---

Changes in v2:
- Use "sandbox-pch" for the compatible string, for consistency

 arch/sandbox/dts/test.dts       |  4 ++
 arch/sandbox/include/asm/test.h |  8 ++++
 drivers/pch/Makefile            |  1 +
 drivers/pch/sandbox_pch.c       | 69 +++++++++++++++++++++++++++++++++
 test/dm/Makefile                |  1 +
 test/dm/pch.c                   | 36 +++++++++++++++++
 6 files changed, 119 insertions(+)
 create mode 100644 drivers/pch/sandbox_pch.c
 create mode 100644 test/dm/pch.c
arch/sandbox/dts/test.dts +4 -0
index 1d011ded7cc..f9a86f2596f 100644
@@ -407,6 +407,10 @@
 		compatible = "sandbox,mmc";
 	};
 
+	pch {
+		compatible = "sandbox-pch";
+	};
+
 	pci0: pci-controller0 {
 		compatible = "sandbox,pci";
 		device_type = "pci";
arch/sandbox/include/asm/test.h +8 -0
index 74f96188220..ce00ba51e47 100644
@@ -161,4 +161,12 @@ int sandbox_get_setup_called(struct udevice *dev);
  */
 int sandbox_get_sound_sum(struct udevice *dev);
 
+/**
+ * sandbox_get_pch_spi_protect() - Get the PCI SPI protection status
+ *
+ * @dev: Device to check
+ * @return 0 if not protected, 1 if protected
+ */
+int sandbox_get_pch_spi_protect(struct udevice *dev);
+
 #endif
drivers/pch/Makefile +1 -0
index 696cdffedb6..8ea6b7852ac 100644
@@ -3,3 +3,4 @@
 obj-y += pch-uclass.o
 obj-y += pch7.o
 obj-y += pch9.o
+obj-$(CONFIG_SANDBOX) += sandbox_pch.o
drivers/pch/sandbox_pch.c new file +69 -0

69 lines not on this page · raw message

test/dm/Makefile +1 -0
index 1b089960cbb..49857c50929 100644
@@ -30,6 +30,7 @@ obj-y += ofnode.o
 obj-$(CONFIG_OSD) += osd.o
 obj-$(CONFIG_DM_VIDEO) += panel.o
 obj-$(CONFIG_DM_PCI) += pci.o
+obj-$(CONFIG_PCH) += pch.o
 obj-$(CONFIG_PHY) += phy.o
 obj-$(CONFIG_POWER_DOMAIN) += power-domain.o
 obj-$(CONFIG_DM_PWM) += pwm.o
test/dm/pch.c new file +36 -0
new file mode 100644
index 00000000000..f184445342b
@@ -0,0 +1,36 @@
+// SPDX-License-Identifier: GPL-2.0+
+/*
+ * Copyright 2018 Google LLC
+ */
+
+#include <common.h>
+#include <dm.h>
+#include <pch.h>
+#include <asm/test.h>
+#include <dm/test.h>
+#include <test/ut.h>
+
+/* Test that sandbox PCH works correctly */
+static int dm_test_pch_base(struct unit_test_state *uts)
+{
+	struct udevice *dev;
+	u32 gbase, iobase;
+	ulong sbase;
+
+	ut_assertok(uclass_first_device_err(UCLASS_PCH, &dev));
+	ut_assertok(pch_get_spi_base(dev, &sbase));
+	ut_asserteq(0x10, sbase);
+
+	ut_asserteq(0, sandbox_get_pch_spi_protect(dev));
+	ut_assertok(pch_set_spi_protect(dev, true));
+	ut_asserteq(1, sandbox_get_pch_spi_protect(dev));
+
+	ut_assertok(pch_get_gpio_base(dev, &gbase));
+	ut_asserteq(0x20, gbase);
+
+	ut_assertok(pch_get_io_base(dev, &iobase));
+	ut_asserteq(0x30, iobase);
+
+	return 0;
+}
+DM_TEST(dm_test_pch_base, DM_TESTF_SCAN_PDATA | DM_TESTF_SCAN_FDT);
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#24 lore raw

PATCH v2 17/34 sandbox: pch: Add a test for the PCH uclass

Hi Simon,

Quoted 10 lines · Simon Glass

This uclass currently has no tests. Add a sandbox driver and some simple tests to provide basic coverage.

---

Changes in v2:
- Use "sandbox-pch" for the compatible string, for consistency

I think it should be "sandbox,pch" instead. If you agree, I can fix this when applying.

Quoted 11 lines
 arch/sandbox/dts/test.dts       |  4 ++
 arch/sandbox/include/asm/test.h |  8 ++++
 drivers/pch/Makefile            |  1 +
 drivers/pch/sandbox_pch.c       | 69 +++++++++++++++++++++++++++++++++
 test/dm/Makefile                |  1 +
 test/dm/pch.c                   | 36 +++++++++++++++++
 6 files changed, 119 insertions(+)
 create mode 100644 drivers/pch/sandbox_pch.c
 create mode 100644 test/dm/pch.c

Regards,
Bin

#25 lore raw

PATCH v2 17/34 sandbox: pch: Add a test for the PCH uclass

Quoted 12 lines · Bin Meng

Hi Simon,

Quoted 8 lines · Simon Glass

This uclass currently has no tests. Add a sandbox driver and some simple tests to provide basic coverage.

---

Quoted 5 lines
Quoted 2 lines

Changes in v2:
- Use "sandbox-pch" for the compatible string, for consistency

I think it should be "sandbox,pch" instead. If you agree, I can fix this when applying.

That's fine with me, it is more consistent with what we have.

Quoted 15 lines
Quoted 12 lines

 arch/sandbox/dts/test.dts       |  4 ++
 arch/sandbox/include/asm/test.h |  8 ++++
 drivers/pch/Makefile            |  1 +
 drivers/pch/sandbox_pch.c       | 69 +++++++++++++++++++++++++++++++++
 test/dm/Makefile                |  1 +
 test/dm/pch.c                   | 36 +++++++++++++++++
 6 files changed, 119 insertions(+)
 create mode 100644 drivers/pch/sandbox_pch.c
 create mode 100644 test/dm/pch.c

Regards,
Bin

#26 lore raw

PATCH v2 18/34 pch: Add ioctl support

In reply to Simon Glass

At present the PCH has 4 operations and these are reasonably widely used in the drivers. But sometimes we want to add rarely used operations, and each of these currently adds to the size of the PCH operations table.

Add an ioctl() method which can be easily expanded without any more impact on the operations table.

---

Changes in v2:
- Fix comment about @return value in ioctl() method and function

 drivers/pch/pch-uclass.c  | 10 ++++++++
 drivers/pch/sandbox_pch.c | 17 ++++++++++++++
 include/pch.h             | 48 ++++++++++++++++++++++++++++++++++++++-
 test/dm/pch.c             | 19 ++++++++++++++++
 4 files changed, 93 insertions(+), 1 deletion(-)
drivers/pch/pch-uclass.c +10 -0
index 831b283d7b4..caf8b72803c 100644
@@ -51,6 +51,16 @@ int pch_get_io_base(struct udevice *dev, u32 *iobasep)
 	return ops->get_io_base(dev, iobasep);
 }
 
+int pch_ioctl(struct udevice *dev, ulong req, void *data, int size)
+{
+	struct pch_ops *ops = pch_get_ops(dev);
+
+	if (!ops->ioctl)
+		return -ENOSYS;
+
+	return ops->ioctl(dev, req, data, size);
+}
+
 UCLASS_DRIVER(pch) = {
 	.id		= UCLASS_PCH,
 	.name		= "pch",
drivers/pch/sandbox_pch.c +17 -0
index 004de901ac7..01ff3bed351 100644
@@ -48,11 +48,28 @@ static int sandbox_pch_get_io_base(struct udevice *dev, u32 *iobasep)
 	return 0;
 }
 
+int sandbox_pch_ioctl(struct udevice *dev, enum pch_req_t req, void *data,
+		      int size)
+{
+	switch (req) {
+	case PCH_REQ_TEST1:
+		return -ENOSYS;
+	case PCH_REQ_TEST2:
+		return *(char *)data;
+	case PCH_REQ_TEST3:
+		*(char *)data = 'x';
+		return 1;
+	default:
+		return -ENOSYS;
+	}
+}
+
 static const struct pch_ops sandbox_pch_ops = {
 	.get_spi_base	= sandbox_pch_get_spi_base,
 	.set_spi_protect = sandbox_pch_set_spi_protect,
 	.get_gpio_base	= sandbox_pch_get_gpio_base,
 	.get_io_base = sandbox_pch_get_io_base,
+	.ioctl		= sandbox_pch_ioctl,
 };
 
 static const struct udevice_id sandbox_pch_ids[] = {
include/pch.h +47 -1

64 lines not on this page · raw message

test/dm/pch.c +19 -0
index f184445342b..54e33d187b6 100644
@@ -34,3 +34,22 @@ static int dm_test_pch_base(struct unit_test_state *uts)
 	return 0;
 }
 DM_TEST(dm_test_pch_base, DM_TESTF_SCAN_PDATA | DM_TESTF_SCAN_FDT);
+
+/* Test sandbox PCH ioctl */
+static int dm_test_pch_ioctl(struct unit_test_state *uts)
+{
+	struct udevice *dev;
+	char data;
+
+	ut_assertok(uclass_first_device_err(UCLASS_PCH, &dev));
+
+	ut_asserteq(-ENOSYS, pch_ioctl(dev, PCH_REQ_TEST1, NULL, 0));
+
+	ut_asserteq('a', pch_ioctl(dev, PCH_REQ_TEST2, "a", 1));
+
+	ut_asserteq(1, pch_ioctl(dev, PCH_REQ_TEST3, &data, 1));
+	ut_asserteq('x', data);
+
+	return 0;
+}
+DM_TEST(dm_test_pch_ioctl, DM_TESTF_SCAN_PDATA | DM_TESTF_SCAN_FDT);
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#27 lore raw

PATCH v2 19/34 x86: ivybridge: Add a way to get the HDA config setting

In reply to Simon Glass

Add a way check to whether HD audio is enabled. Use ioctl() to avoid adding too many unusual operations to PCH.

---

Changes in v2:
- Adjust code to use #defines instead of open-coded values

 arch/x86/cpu/ivybridge/bd82x6x.c | 27 +++++++++++++++++++++++++--
 include/pch.h                    |  3 +++
 2 files changed, 28 insertions(+), 2 deletions(-)
arch/x86/cpu/ivybridge/bd82x6x.c +25 -2
index a78bb025440..ed9bce64168 100644
@@ -20,8 +20,12 @@
 
 DECLARE_GLOBAL_DATA_PTR;
 
-#define GPIO_BASE	0x48
-#define BIOS_CTRL	0xdc
+#define GPIO_BASE		0x48
+#define BIOS_CTRL		0xdc
+
+#define RCBA_AUDIO_CONFIG	0x2030
+#define RCBA_AUDIO_CONFIG_HDA	BIT(31)
+#define RCBA_AUDIO_CONFIG_MASK	0xfe
 
 #ifndef CONFIG_HAVE_FSP
 static int pch_revision_id = -1;
@@ -212,10 +216,29 @@ static int bd82x6x_get_gpio_base(struct udevice *dev, u32 *gbasep)
 	return 0;
 }
 
+static int bd82x6x_ioctl(struct udevice *dev, enum pch_req_t req, void *data,
+			 int size)
+{
+	u32 rcba, val;
+
+	switch (req) {
+	case PCH_REQ_HDA_CONFIG:
+		dm_pci_read_config32(dev, PCH_RCBA, &rcba);
+		val = readl(rcba + RCBA_AUDIO_CONFIG);
+		if (!(val & RCBA_AUDIO_CONFIG_HDA))
+			return -ENOENT;
+
+		return val & RCBA_AUDIO_CONFIG_MASK;
+	default:
+		return -ENOSYS;
+	}
+}
+
 static const struct pch_ops bd82x6x_pch_ops = {
 	.get_spi_base	= bd82x6x_pch_get_spi_base,
 	.set_spi_protect = bd82x6x_set_spi_protect,
 	.get_gpio_base	= bd82x6x_get_gpio_base,
+	.ioctl		= bd82x6x_ioctl,
 };
 
 static const struct udevice_id bd82x6x_ids[] = {
include/pch.h +3 -0
index b8b62d74acb..046a5fde3ab 100644
@@ -13,6 +13,9 @@
 
 /* All the supported PCH ioctls */
 enum pch_req_t {
+	/* Returns HDA config info if Azalia V1CTL enabled, -ENOENT if not */
+	PCH_REQ_HDA_CONFIG,
+
 	PCH_REQ_TEST1,		/* Test requests for sandbox driver */
 	PCH_REQ_TEST2,
 	PCH_REQ_TEST3,
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#28 lore raw

PATCH v2 20/34 sound: Mark sound_setup() as optional

In reply to Simon Glass

This method in the sound API is optional since some drivers can do this when probing or as part of SoC init. Mark it as such.

---

Changes in v2: None

 drivers/sound/sound-uclass.c | 2 +-
 include/sound.h              | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)
drivers/sound/sound-uclass.c +1 -1
index 2b836268896..067660623b2 100644
@@ -16,7 +16,7 @@ int sound_setup(struct udevice *dev)
 	struct sound_ops *ops = sound_get_ops(dev);
 
 	if (!ops->setup)
-		return -ENOSYS;
+		return 0;
 
 	return ops->setup(dev);
 }
include/sound.h +1 -1
index b7959cc2607..7d528c479ef 100644
@@ -54,7 +54,7 @@ void sound_create_square_wave(uint sample_rate, unsigned short *data, int size,
 /* Operations for sound */
 struct sound_ops {
 	/**
-	 * setup() - Set up to play a sound
+	 * setup() - Set up to play a sound (optional)
 	 */
 	int (*setup)(struct udevice *dev);
 
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#29 lore raw

PATCH v2 21/34 sound: Add uclass operations for beeping

In reply to Simon Glass

Some audio codecs such as Intel HDA do not need to use digital data to play sounds, but instead have a way to emit beeps. Add this interface as an option. If the beep interface is not supported, then the sound uclass falls back to the I2S interface.

---

Changes in v2: None

 arch/sandbox/include/asm/test.h | 16 ++++++++++++
 drivers/sound/sandbox.c         | 46 ++++++++++++++++++++++++++++++---
 drivers/sound/sound-uclass.c    | 34 +++++++++++++++++++++++-
 include/sound.h                 | 44 +++++++++++++++++++++++++++++++
 test/dm/sound.c                 | 21 +++++++++++++++
 5 files changed, 157 insertions(+), 4 deletions(-)
arch/sandbox/include/asm/test.h +16 -0
index ce00ba51e47..fc52f47f821 100644
@@ -161,6 +161,22 @@ int sandbox_get_setup_called(struct udevice *dev);
  */
 int sandbox_get_sound_sum(struct udevice *dev);
 
+/**
+ * sandbox_set_allow_beep() - Set whether the 'beep' interface is supported
+ *
+ * @dev: Device to update
+ * @allow: true to allow the start_beep() method, false to disallow it
+ */
+void sandbox_set_allow_beep(struct udevice *dev, bool allow);
+
+/**
+ * sandbox_get_beep_frequency() - Get the frequency of the current beep
+ *
+ * @dev: Device to check
+ * @return frequency of beep, if there is an active beep, else 0
+ */
+int sandbox_get_beep_frequency(struct udevice *dev);
+
 /**
  * sandbox_get_pch_spi_protect() - Get the PCI SPI protection status
  *
drivers/sound/sandbox.c +43 -3

70 lines not on this page · raw message

drivers/sound/sound-uclass.c +33 -1
index 067660623b2..d49f29bcd5b 100644
@@ -31,10 +31,30 @@ int sound_play(struct udevice *dev, void *data, uint data_size)
 	return ops->play(dev, data, data_size);
 }
 
+int sound_start_beep(struct udevice *dev, int frequency_hz)
+{
+	struct sound_ops *ops = sound_get_ops(dev);
+
+	if (!ops->start_beep)
+		return -ENOSYS;
+
+	return ops->start_beep(dev, frequency_hz);
+}
+
+int sound_stop_beep(struct udevice *dev)
+{
+	struct sound_ops *ops = sound_get_ops(dev);
+
+	if (!ops->stop_beep)
+		return -ENOSYS;
+
+	return ops->stop_beep(dev);
+}
+
 int sound_beep(struct udevice *dev, int msecs, int frequency_hz)
 {
 	struct sound_uc_priv *uc_priv = dev_get_uclass_priv(dev);
-	struct i2s_uc_priv *i2s_uc_priv = dev_get_uclass_priv(uc_priv->i2s);
+	struct i2s_uc_priv *i2s_uc_priv;
 	unsigned short *data;
 	uint data_size;
 	int ret;
@@ -43,7 +63,19 @@ int sound_beep(struct udevice *dev, int msecs, int frequency_hz)
 	if (ret && ret != -EALREADY)
 		return ret;
 
+	/* Try using the beep interface if available */
+	ret = sound_start_beep(dev, frequency_hz);
+	if (ret != -ENOSYS) {
+		if (ret)
+			return ret;
+		mdelay(msecs);
+		ret = sound_stop_beep(dev);
+
+		return ret;
+	}
+
 	/* Buffer length computation */
+	i2s_uc_priv = dev_get_uclass_priv(uc_priv->i2s);
 	data_size = i2s_uc_priv->samplingrate * i2s_uc_priv->channels;
 	data_size *= (i2s_uc_priv->bitspersample / SOUND_BITS_IN_BYTE);
 	data = malloc(data_size);
include/sound.h +44 -0

56 lines not on this page · raw message

test/dm/sound.c +21 -0
index 7d0b36e7a56..3767abbd1c7 100644
@@ -32,3 +32,24 @@ static int dm_test_sound(struct unit_test_state *uts)
 	return 0;
 }
 DM_TEST(dm_test_sound, DM_TESTF_SCAN_PDATA | DM_TESTF_SCAN_FDT);
+
+/* Test of the 'start beep' operations */
+static int dm_test_sound_beep(struct unit_test_state *uts)
+{
+	struct udevice *dev;
+
+	/* check probe success */
+	ut_assertok(uclass_first_device_err(UCLASS_SOUND, &dev));
+	ut_asserteq(-ENOSYS, sound_start_beep(dev, 100));
+	ut_asserteq(0, sandbox_get_beep_frequency(dev));
+
+	sandbox_set_allow_beep(dev, true);
+	ut_asserteq(0, sound_start_beep(dev, 100));
+	ut_asserteq(100, sandbox_get_beep_frequency(dev));
+
+	ut_asserteq(0, sound_stop_beep(dev));
+	ut_asserteq(0, sandbox_get_beep_frequency(dev));
+
+	return 0;
+}
+DM_TEST(dm_test_sound_beep, DM_TESTF_SCAN_PDATA | DM_TESTF_SCAN_FDT);
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#30 lore raw

PATCH v2 22/34 sound: Add support for Intel HDA

In reply to Simon Glass

The Intel High-definition Audio is a newer-generation audio system which provides for transfer of a large number of audio stream, each containing up to 16 channels.

Add support for HDA as a library which can be used by other drivers. U-Boot currently uses only two channels (stereo).

---

Changes in v2:
- Add a link to the HD audio PDF fie
- Update in codec_init() to use defines for codec access
- Use a more generic "intel,hd-audio" as the compatible string
- Use defines for all azalia verbs in binding file

 doc/device-tree-bindings/sound/intel-hda.txt |  25 +
 drivers/sound/Kconfig                        |   9 +
 drivers/sound/Makefile                       |   1 +
 drivers/sound/hda_codec.c                    | 556 +++++++++++++++++++
 include/dt-bindings/sound/azalia.h           |  44 ++
 include/hda_codec.h                          | 103 ++++
 6 files changed, 738 insertions(+)
 create mode 100644 doc/device-tree-bindings/sound/intel-hda.txt
 create mode 100644 drivers/sound/hda_codec.c
 create mode 100644 include/dt-bindings/sound/azalia.h
 create mode 100644 include/hda_codec.h
doc/device-tree-bindings/sound/intel-hda.txt new file +25 -0
new file mode 100644
index 00000000000..fb2ce550063
@@ -0,0 +1,25 @@
+* Intel High-definition Audio
+
+Configuration is set using 'verbs' which are blocks of 16 bytes of data each
+with a different purpose, a little like a simple instruction set.
+
+Top-level node
+--------------
+
+Required properties:
+- compatible: "intel,hd-audio"
+- beep-verbs: list of verbs to send for a beep
+
+Optional properties
+- intel,beep-nid: Node ID to use for beep (will be detected if not provided)
+
+Required subnodes:
+- codecs: Contains a list of codec nodes
+
+
+* Codec nodes
+
+Required properties:
+- vendor-id: 16-bit vendor ID for audio codec
+- device-id: 16-bit device ID for audio codec
+- verbs: List of verbs, each 4 cells in length
drivers/sound/Kconfig +9 -0
index 40f4f7598aa..0dca6ffe88c 100644
@@ -40,6 +40,15 @@ config I2S_SAMSUNG
 	  option provides an implementation for sound_init() and
 	  sound_play().
 
+config SOUND_INTEL_HDA
+	bool "Intel HDA audio codec"
+	depends on SOUND
+	help
+	  Most Intel chips have an HDA (High-definition audio) codec which can
+	  be used by U-Boot to play simple beeps. This is also sometimes called
+	  Azalia which was the development code-name. It requires setup
+	  information in the device tree (see intel-hda.txt).
+
 config SOUND_MAX98088
 	bool "Support Maxim max98088 audio codec"
 	depends on I2S
drivers/sound/Makefile +1 -0
index 170e06ad538..2e21a0d0674 100644
@@ -15,3 +15,4 @@ obj-$(CONFIG_SOUND_WM8994)	+= wm8994.o
 obj-$(CONFIG_SOUND_MAX98088)	+= max98088.o maxim_codec.o
 obj-$(CONFIG_SOUND_MAX98090)	+= max98090.o maxim_codec.o
 obj-$(CONFIG_SOUND_MAX98095)	+= max98095.o maxim_codec.o
+obj-$(CONFIG_SOUND_INTEL_HDA)	+= hda_codec.o
drivers/sound/hda_codec.c new file +556 -0

556 lines not on this page · raw message

include/dt-bindings/sound/azalia.h new file +44 -0

44 lines not on this page · raw message

include/hda_codec.h new file +103 -0

103 lines not on this page · raw message

Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#31 lore raw

PATCH v2 22/34 sound: Add support for Intel HDA

Quoted 17 lines · Simon Glass

The Intel High-definition Audio is a newer-generation audio system which provides for transfer of a large number of audio stream, each containing up to 16 channels.

Add support for HDA as a library which can be used by other drivers. U-Boot currently uses only two channels (stereo).

---

Changes in v2:
- Add a link to the HD audio PDF fie
- Update in codec_init() to use defines for codec access
- Use a more generic "intel,hd-audio" as the compatible string
- Use defines for all azalia verbs in binding file

Quoted 12 lines
 doc/device-tree-bindings/sound/intel-hda.txt |  25 +
 drivers/sound/Kconfig                        |   9 +
 drivers/sound/Makefile                       |   1 +
 drivers/sound/hda_codec.c                    | 556 +++++++++++++++++++
 include/dt-bindings/sound/azalia.h           |  44 ++
 include/hda_codec.h                          | 103 ++++
 6 files changed, 738 insertions(+)
 create mode 100644 doc/device-tree-bindings/sound/intel-hda.txt
 create mode 100644 drivers/sound/hda_codec.c
 create mode 100644 include/dt-bindings/sound/azalia.h
 create mode 100644 include/hda_codec.h
#32 lore raw

PATCH v2 23/34 sandbox: sound: Silence sound for testing

In reply to Simon Glass

When testing the sound system we don't need the hear the beeps. The testing works by checking the data that would be emitted. Add a device-tree property to silence the sound, and enable it for testing.

---

Changes in v2:
- Move comment about silencing on error so that it is clearer

 arch/sandbox/dts/test.dts |  1 +
 drivers/sound/sandbox.c   | 25 ++++++++++++++++++++++---
 2 files changed, 23 insertions(+), 3 deletions(-)
arch/sandbox/dts/test.dts +1 -0
index f9a86f2596f..e541d6da37e 100644
@@ -389,6 +389,7 @@
 	i2s: i2s {
 		compatible = "sandbox,i2s";
 		#sound-dai-cells = <1>;
+		sandbox,silent;	/* Don't emit sounds while testing */
 	};
 
 	misc-test {
drivers/sound/sandbox.c +22 -3
index 600523160f0..363c687bafd 100644
@@ -3,6 +3,8 @@
  * Copyright (c) 2013 Google, Inc
  */
 
+#define LOG_CATEGORY UCLASS_SOUND
+
 #include <common.h>
 #include <audio_codec.h>
 #include <dm.h>
@@ -20,6 +22,7 @@ struct sandbox_codec_priv {
 
 struct sandbox_i2s_priv {
 	int sum;	/* Use to sum the provided audio data */
+	bool silent;	/* Sound is silent, don't use SDL */
 };
 
 struct sandbox_sound_priv {
@@ -101,12 +104,21 @@ static int sandbox_i2s_tx_data(struct udevice *dev, void *data,
 	for (i = 0; i < data_size; i++)
 		priv->sum += ((uint8_t *)data)[i];
 
-	return sandbox_sdl_sound_play(data, data_size);
+	if (!priv->silent) {
+		int ret;
+
+		ret = sandbox_sdl_sound_play(data, data_size);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
 }
 
 static int sandbox_i2s_probe(struct udevice *dev)
 {
 	struct i2s_uc_priv *uc_priv = dev_get_uclass_priv(dev);
+	struct sandbox_i2s_priv *priv = dev_get_priv(dev);
 
 	/* Use hard-coded values here */
 	uc_priv->rfs = 256;
@@ -117,8 +129,15 @@ static int sandbox_i2s_probe(struct udevice *dev)
 	uc_priv->channels = 2;
 	uc_priv->id = 1;
 
-	/* Ignore any error here - we'll just have no sound */
-	sandbox_sdl_sound_init(uc_priv->samplingrate, uc_priv->channels);
+	priv->silent = dev_read_bool(dev, "sandbox,silent");
+
+	if (priv->silent) {
+		log_warning("Sound is silenced\n");
+	} else if (sandbox_sdl_sound_init(uc_priv->samplingrate,
+					  uc_priv->channels)) {
+		/* Ignore any error here - we'll just have no sound */
+		priv->silent = true;
+	}
 
 	return 0;
 }
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#33 lore raw

PATCH v2 23/34 sandbox: sound: Silence sound for testing

Quoted 6 lines · Simon Glass

When testing the sound system we don't need the hear the beeps. The testing works by checking the data that would be emitted. Add a device-tree property to silence the sound, and enable it for testing.

Quoted 9 lines

---

Changes in v2:
- Move comment about silencing on error so that it is clearer

 arch/sandbox/dts/test.dts |  1 +
 drivers/sound/sandbox.c   | 25 ++++++++++++++++++++++---
 2 files changed, 23 insertions(+), 3 deletions(-)
#34 lore raw

PATCH v2 24/34 sound: x86: link: Add sound support

In reply to Simon Glass

Add sound support for link, using the HDA codec implementation.

---

Changes in v2:
- Add a note about undocumented IOMMU values
- Add comments next to verbs in the DT mentioning the HDA spec
- Fix capital letter in Ivybridge in Kconfig help
- Fix use of C++ comments in two places
- Rename sound driver to bd82x6x-hda
- Use a more generic "intel,hd-audio" as the compatible string
- User lower-case hex consistently in the DT

 arch/x86/cpu/ivybridge/Kconfig                |   1 +
 arch/x86/cpu/ivybridge/northbridge.c          |  32 ++++
 arch/x86/dts/chromebook_link.dts              |  96 ++++++++++++
 .../include/asm/arch-ivybridge/sandybridge.h  |   3 +
 configs/chromebook_link_defconfig             |   2 +
 drivers/sound/Kconfig                         |  10 ++
 drivers/sound/Makefile                        |   1 +
 drivers/sound/ivybridge_sound.c               | 137 ++++++++++++++++++
 8 files changed, 282 insertions(+)
 create mode 100644 drivers/sound/ivybridge_sound.c
arch/x86/cpu/ivybridge/Kconfig +1 -0
index 5f0e60837ca..2f423937864 100644
@@ -21,6 +21,7 @@ config NORTHBRIDGE_INTEL_IVYBRIDGE
 	imply USB_EHCI_HCD
 	imply USB_XHCI_HCD
 	imply VIDEO_VESA
+	imply SOUND_IVYBRIDGE
 
 if NORTHBRIDGE_INTEL_IVYBRIDGE
 
arch/x86/cpu/ivybridge/northbridge.c +32 -0
index 39bab7bdf34..a809b823b3b 100644
@@ -177,6 +177,35 @@ static void sandybridge_setup_northbridge_bars(struct udevice *dev)
 	dm_pci_write_config8(dev, PAM6, 0x33);
 }
 
+/**
+ * sandybridge_init_iommu() - Set up IOMMU so that azalia can be used
+ *
+ * It is not obvious where these values come from. They may be undocumented.
+ */
+static void sandybridge_init_iommu(struct udevice *dev)
+{
+	u32 capid0_a;
+
+	dm_pci_read_config32(dev, 0xe4, &capid0_a);
+	if (capid0_a & (1 << 23)) {
+		log_debug("capid0_a not needed\n");
+		return;
+	}
+
+	/* setup BARs */
+	writel(IOMMU_BASE1 >> 32, MCHBAR_REG(0x5404));
+	writel(IOMMU_BASE1 | 1, MCHBAR_REG(0x5400));
+	writel(IOMMU_BASE2 >> 32, MCHBAR_REG(0x5414));
+	writel(IOMMU_BASE2 | 1, MCHBAR_REG(0x5410));
+
+	/* lock policies */
+	writel(0x80000000, IOMMU_BASE1 + 0xff0);
+
+	/* Enable azalia sound */
+	writel(0x20000000, IOMMU_BASE2 + 0xff0);
+	writel(0xa0000000, IOMMU_BASE2 + 0xff0);
+}
+
 static int bd82x6x_northbridge_early_init(struct udevice *dev)
 {
 	const int chipset_type = SANDYBRIDGE_MOBILE;
@@ -197,6 +226,9 @@ static int bd82x6x_northbridge_early_init(struct udevice *dev)
 
 	sandybridge_setup_northbridge_bars(dev);
 
+	/* Setup IOMMU BARs */
+	sandybridge_init_iommu(dev);
+
 	/* Device Enable */
 	dm_pci_write_config32(dev, DEVEN, DEVEN_HOST | DEVEN_IGD);
 
arch/x86/dts/chromebook_link.dts +96 -0

111 lines not on this page · raw message

arch/x86/include/asm/arch-ivybridge/sandybridge.h +3 -0
index a96c951c85e..a3a507f2b36 100644
@@ -43,6 +43,9 @@
 /* 4 KB per PCIe device */
 #define DEFAULT_PCIEXBAR	CONFIG_PCIE_ECAM_BASE
 
+#define IOMMU_BASE1		0xfed90000ULL
+#define IOMMU_BASE2		0xfed91000ULL
+
 /* Device 0:0.0 PCI configuration space (Host Bridge) */
 #define EPBAR		0x40
 #define MCHBAR		0x48
configs/chromebook_link_defconfig +2 -0
index c402568c17e..6058dfaf0f2 100644
@@ -33,6 +33,7 @@ CONFIG_CMD_DHCP=y
 # CONFIG_CMD_NFS is not set
 CONFIG_CMD_PING=y
 CONFIG_CMD_TIME=y
+CONFIG_CMD_SOUND=y
 CONFIG_CMD_BOOTSTAGE=y
 CONFIG_CMD_TPM=y
 CONFIG_CMD_TPM_TEST=y
@@ -53,6 +54,7 @@ CONFIG_SYS_I2C_INTEL=y
 CONFIG_CROS_EC=y
 CONFIG_CROS_EC_LPC=y
 CONFIG_SYS_NS16550=y
+CONFIG_SOUND=y
 CONFIG_SPI=y
 CONFIG_TPM_TIS_LPC=y
 CONFIG_USB_STORAGE=y
drivers/sound/Kconfig +10 -0
index 0dca6ffe88c..4b893f035b0 100644
@@ -49,6 +49,16 @@ config SOUND_INTEL_HDA
 	  Azalia which was the development code-name. It requires setup
 	  information in the device tree (see intel-hda.txt).
 
+config SOUND_IVYBRIDGE
+	bool "Intel Ivybridge sound support"
+	depends on SOUND
+	select SOUND_INTEL_HDA
+	help
+	  Enable sound output on supported Intel Ivybridge-based boards. This
+	  driver uses Intel's High-definition Audio (HDA) architecture,
+	  sometimes called Azalia. The audio codec is detected using a
+	  semi-automatic mechanism.
+
 config SOUND_MAX98088
 	bool "Support Maxim max98088 audio codec"
 	depends on I2S
drivers/sound/Makefile +1 -0
index 2e21a0d0674..771c1165167 100644
@@ -16,3 +16,4 @@ obj-$(CONFIG_SOUND_MAX98088)	+= max98088.o maxim_codec.o
 obj-$(CONFIG_SOUND_MAX98090)	+= max98090.o maxim_codec.o
 obj-$(CONFIG_SOUND_MAX98095)	+= max98095.o maxim_codec.o
 obj-$(CONFIG_SOUND_INTEL_HDA)	+= hda_codec.o
+obj-$(CONFIG_SOUND_IVYBRIDGE)	+= ivybridge_sound.o
drivers/sound/ivybridge_sound.c new file +137 -0

137 lines not on this page · raw message

Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#35 lore raw

PATCH v2 24/34 sound: x86: link: Add sound support

Quoted 15 lines · Simon Glass

Add sound support for link, using the HDA codec implementation.

---

Changes in v2:
- Add a note about undocumented IOMMU values
- Add comments next to verbs in the DT mentioning the HDA spec
- Fix capital letter in Ivybridge in Kconfig help
- Fix use of C++ comments in two places
- Rename sound driver to bd82x6x-hda
- Use a more generic "intel,hd-audio" as the compatible string
- User lower-case hex consistently in the DT

Quoted 11 lines
 arch/x86/cpu/ivybridge/Kconfig                |   1 +
 arch/x86/cpu/ivybridge/northbridge.c          |  32 ++++
 arch/x86/dts/chromebook_link.dts              |  96 ++++++++++++
 .../include/asm/arch-ivybridge/sandybridge.h  |   3 +
 configs/chromebook_link_defconfig             |   2 +
 drivers/sound/Kconfig                         |  10 ++
 drivers/sound/Makefile                        |   1 +
 drivers/sound/ivybridge_sound.c               | 137 ++++++++++++++++++
 8 files changed, 282 insertions(+)
 create mode 100644 drivers/sound/ivybridge_sound.c
#36 lore raw

PATCH v2 25/34 sound: x86: Add beeping support in i8254

In reply to Simon Glass

Adjust the code to allow beeping at different frequencies, using a calculated value for timer 2.

---

Changes in v2:
- Rename PPC_PORTB to SYSCTL_PORTB
- Use PIT_TICK_RATE instead of creating a new define

 arch/x86/include/asm/i8254.h | 18 ++++++++++++++++++
 arch/x86/lib/i8254.c         | 35 +++++++++++++++++++++++++++++++----
 2 files changed, 49 insertions(+), 4 deletions(-)
arch/x86/include/asm/i8254.h +18 -0
index 65c97614d3e..d769daf85dc 100644
@@ -35,4 +35,22 @@
 /* The clock frequency of the i8253/i8254 PIT */
 #define PIT_TICK_RATE	1193182
 
+/**
+ * i8254_enable_beep() - Start a beep using the PCAT timer
+ *
+ * This starts beeping using the legacy i8254 timer. The beep may be silenced
+ * after a delay with i8254_disable_beep().
+ *
+ * @frequency_hz: Frequency of beep in Hz
+ * @return 0 if OK, -EINVAL if frequency_hz is 0
+ */
+int i8254_enable_beep(uint frequency_hz);
+
+/**
+ * i8254_disable_beep() - Disable the bepper
+ *
+ * This stops any existing beep
+ */
+void i8254_disable_beep(void);
+
 #endif /* _ASMI386_I8954_H_ */
arch/x86/lib/i8254.c +31 -4
index 1f1012488cb..d0227954b16 100644
@@ -8,8 +8,20 @@
 #include <asm/io.h>
 #include <asm/i8254.h>
 
-#define TIMER1_VALUE	18	/* 15.6us */
-#define TIMER2_VALUE	0x0a8e	/* 440Hz */
+#define TIMER1_VALUE		18	/* 15.6us */
+#define BEEP_FREQUENCY_HZ	440
+#define SYSCTL_PORTB		0x61
+#define PORTB_BEEP_ENABLE	0x3
+
+static void i8254_set_beep_freq(uint frequency_hz)
+{
+	uint countdown;
+
+	countdown = PIT_TICK_RATE / frequency_hz;
+
+	outb(countdown & 0xff, PIT_BASE + PIT_T2);
+	outb((countdown >> 8) & 0xff, PIT_BASE + PIT_T2);
+}
 
 int i8254_init(void)
 {
@@ -29,8 +41,23 @@ int i8254_init(void)
 	 */
 	outb(PIT_CMD_CTR2 | PIT_CMD_BOTH | PIT_CMD_MODE3,
 	     PIT_BASE + PIT_COMMAND);
-	outb(TIMER2_VALUE & 0xff, PIT_BASE + PIT_T2);
-	outb(TIMER2_VALUE >> 8, PIT_BASE + PIT_T2);
+	i8254_set_beep_freq(BEEP_FREQUENCY_HZ);
+
+	return 0;
+}
+
+int i8254_enable_beep(uint frequency_hz)
+{
+	if (!frequency_hz)
+		return -EINVAL;
+
+	i8254_set_beep_freq(frequency_hz);
+	setio_8(SYSCTL_PORTB, PORTB_BEEP_ENABLE);
 
 	return 0;
 }
+
+void i8254_disable_beep(void)
+{
+	clrio_8(SYSCTL_PORTB, PORTB_BEEP_ENABLE);
+}
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#37 lore raw

PATCH v2 25/34 sound: x86: Add beeping support in i8254

Quoted 7 lines · Simon Glass

Adjust the code to allow beeping at different frequencies, using a calculated value for timer 2.

---

Quoted 8 lines

Changes in v2:
- Rename PPC_PORTB to SYSCTL_PORTB
- Use PIT_TICK_RATE instead of creating a new define

 arch/x86/include/asm/i8254.h | 18 ++++++++++++++++++
 arch/x86/lib/i8254.c         | 35 +++++++++++++++++++++++++++++++----
 2 files changed, 49 insertions(+), 4 deletions(-)
#38 lore raw

PATCH v2 26/34 sound: Add a driver for the i8254 beep

In reply to Simon Glass

Add a sound driver which can output simple beeps using this legacy timer.

---

Changes in v2:
- Change the driver name to i8254_drv

 drivers/sound/Kconfig      | 12 ++++++++++++
 drivers/sound/Makefile     |  1 +
 drivers/sound/i8254_beep.c | 38 ++++++++++++++++++++++++++++++++++++++
 3 files changed, 51 insertions(+)
 create mode 100644 drivers/sound/i8254_beep.c
drivers/sound/Kconfig +12 -0
index 4b893f035b0..c1847df07fa 100644
@@ -40,6 +40,18 @@ config I2S_SAMSUNG
 	  option provides an implementation for sound_init() and
 	  sound_play().
 
+config SOUND_I8254
+	bool "Intel i8254 timer / beeper"
+	depends on SOUND
+	help
+	  This enables support for a beeper that uses the i8254 timer chip.
+	  This can emit beeps at a fixed frequency. It is possible to control
+	  the length of the beeps, by turning a beep on, waiting for a period
+	  of time, then turning it off.
+
+	  This is quite an old feature, called PIT (Programmable Interval
+	  Timer), but is nonetheless still available on modern x86 machines.
+
 config SOUND_INTEL_HDA
 	bool "Intel HDA audio codec"
 	depends on SOUND
drivers/sound/Makefile +1 -0
index 771c1165167..2a6fd7420c2 100644
@@ -16,4 +16,5 @@ obj-$(CONFIG_SOUND_MAX98088)	+= max98088.o maxim_codec.o
 obj-$(CONFIG_SOUND_MAX98090)	+= max98090.o maxim_codec.o
 obj-$(CONFIG_SOUND_MAX98095)	+= max98095.o maxim_codec.o
 obj-$(CONFIG_SOUND_INTEL_HDA)	+= hda_codec.o
+obj-$(CONFIG_SOUND_I8254)	+= i8254_beep.o
 obj-$(CONFIG_SOUND_IVYBRIDGE)	+= ivybridge_sound.o
drivers/sound/i8254_beep.c new file +38 -0
new file mode 100644
index 00000000000..5572dc4d265
@@ -0,0 +1,38 @@
+// SPDX-License-Identifier: GPL-2.0+
+/*
+ * Copyright 2018 Google LLC
+ */
+
+#include <common.h>
+#include <dm.h>
+#include <sound.h>
+#include <asm/i8254.h>
+
+int i8254_start_beep(struct udevice *dev, int frequency_hz)
+{
+	return i8254_enable_beep(frequency_hz);
+}
+
+int i8254_stop_beep(struct udevice *dev)
+{
+	i8254_disable_beep();
+
+	return 0;
+}
+
+static const struct sound_ops i8254_ops = {
+	.start_beep	= i8254_start_beep,
+	.stop_beep	= i8254_stop_beep,
+};
+
+static const struct udevice_id i8254_ids[] = {
+	{ .compatible = "i8254,beeper" },
+	{ }
+};
+
+U_BOOT_DRIVER(i8254_drv) = {
+	.name		= "i8254_drv",
+	.id		= UCLASS_SOUND,
+	.of_match	= i8254_ids,
+	.ops		= &i8254_ops,
+};
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#39 lore raw

PATCH v2 26/34 sound: Add a driver for the i8254 beep

Quoted 6 lines · Simon Glass

Add a sound driver which can output simple beeps using this legacy timer.

---

Quoted 9 lines

Changes in v2:
- Change the driver name to i8254_drv

 drivers/sound/Kconfig      | 12 ++++++++++++
 drivers/sound/Makefile     |  1 +
 drivers/sound/i8254_beep.c | 38 ++++++++++++++++++++++++++++++++++++++
 3 files changed, 51 insertions(+)
 create mode 100644 drivers/sound/i8254_beep.c
#40 lore raw

PATCH v2 27/34 x86: coreboot: Enable the beeper sound driver

In reply to Simon Glass

Use the i8254 sound driver to support creating simple beeps.

---

Changes in v2: None

 configs/coreboot_defconfig | 3 +++
 1 file changed, 3 insertions(+)
configs/coreboot_defconfig +3 -0
index ad0b35f2046..2795fe97ca2 100644
@@ -23,6 +23,7 @@ CONFIG_CMD_DHCP=y
 # CONFIG_CMD_NFS is not set
 CONFIG_CMD_PING=y
 CONFIG_CMD_TIME=y
+CONFIG_CMD_SOUND=y
 CONFIG_CMD_EXT2=y
 CONFIG_CMD_EXT4=y
 CONFIG_CMD_EXT4_WRITE=y
@@ -35,4 +36,6 @@ CONFIG_DEFAULT_DEVICE_TREE="coreboot"
 CONFIG_REGMAP=y
 CONFIG_SYSCON=y
 # CONFIG_PCI_PNP is not set
+CONFIG_SOUND=y
+CONFIG_SOUND_I8254=y
 CONFIG_CONSOLE_SCROLL_LINES=5
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#41 lore raw

PATCH v2 28/34 x86: broadwell: Add support for the ADSP

In reply to Simon Glass

The Application Digital Signal Processor is used for sound processing with broadwell. Add a driver to support this.

---

Changes in v2:
- Drop useless 'set' comment

 arch/x86/cpu/broadwell/Makefile            |   1 +
 arch/x86/cpu/broadwell/adsp.c              | 156 +++++++++++++++++++++
 arch/x86/include/asm/arch-broadwell/adsp.h |  46 ++++++
 arch/x86/include/asm/arch-broadwell/pch.h  |   3 +
 arch/x86/include/asm/arch-broadwell/rcb.h  |   2 +
 5 files changed, 208 insertions(+)
 create mode 100644 arch/x86/cpu/broadwell/adsp.c
 create mode 100644 arch/x86/include/asm/arch-broadwell/adsp.h
arch/x86/cpu/broadwell/Makefile +1 -0
index a032861e57c..d3785aabdf2 100644
@@ -2,6 +2,7 @@
 #
 # Copyright (c) 2016 Google, Inc
 
+obj-y += adsp.o
 obj-y += cpu.o
 obj-y += iobp.o
 obj-y += lpc.o
arch/x86/cpu/broadwell/adsp.c new file +156 -0

156 lines not on this page · raw message

arch/x86/include/asm/arch-broadwell/adsp.h new file +46 -0

46 lines not on this page · raw message

arch/x86/include/asm/arch-broadwell/pch.h +3 -0
index 23ccd684842..23153a040fa 100644
@@ -109,6 +109,9 @@
 #define SATA_DTLE_EDGE_SHIFT	16
 
 /* Power Management */
+#define PCH_PCS			0x84
+#define  PCH_PCS_PS_D3HOT	3
+
 #define GEN_PMCON_1		0xa0
 #define  SMI_LOCK		(1 << 4)
 #define GEN_PMCON_2		0xa2
arch/x86/include/asm/arch-broadwell/rcb.h +2 -0
index e7340c1e5a6..b7ce8746c85 100644
@@ -6,6 +6,8 @@
 #ifndef __asm_arch_rcba_h
 #define __asm_arch_rcba_h
 
+#define ACPIIRQEN	0x31e0	/* 32bit */
+
 #define PMSYNC_CONFIG	0x33c4	/* 32bit */
 #define PMSYNC_CONFIG2	0x33cc	/* 32bit */
 
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#42 lore raw

PATCH v2 28/34 x86: broadwell: Add support for the ADSP

Quoted 10 lines · Simon Glass

The Application Digital Signal Processor is used for sound processing with broadwell. Add a driver to support this.

---

Changes in v2:
- Drop useless 'set' comment

Quoted 9 lines
 arch/x86/cpu/broadwell/Makefile            |   1 +
 arch/x86/cpu/broadwell/adsp.c              | 156 +++++++++++++++++++++
 arch/x86/include/asm/arch-broadwell/adsp.h |  46 ++++++
 arch/x86/include/asm/arch-broadwell/pch.h  |   3 +
 arch/x86/include/asm/arch-broadwell/rcb.h  |   2 +
 5 files changed, 208 insertions(+)
 create mode 100644 arch/x86/cpu/broadwell/adsp.c
 create mode 100644 arch/x86/include/asm/arch-broadwell/adsp.h
#43 lore raw

PATCH v2 29/34 x86: broadwell: Don't bother probing the PCH for pinctrl

In reply to Simon Glass

At present the pinctrl probes the PCH but since it only uses it to obtain a PCI address, this is no necessary. Avoiding this fixes one of the two co-dependent loops in broadwell.

This driver really should be a proper pinctrl driver, but for now it remains a syscon device.

---

Changes in v2: None

 arch/x86/cpu/broadwell/pinctrl_broadwell.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
arch/x86/cpu/broadwell/pinctrl_broadwell.c +2 -1
index 914ecfb3144..aa83abbf855 100644
@@ -16,6 +16,7 @@
 #include <asm/arch/gpio.h>
 #include <dt-bindings/gpio/x86-gpio.h>
 #include <dm/pinctrl.h>
+#include <dm/uclass-internal.h>
 
 DECLARE_GLOBAL_DATA_PTR;
 
@@ -214,7 +215,7 @@ static int broadwell_pinctrl_probe(struct udevice *dev)
 	u32 gpiobase;
 	int ret;
 
-	ret = uclass_first_device(UCLASS_PCH, &pch);
+	ret = uclass_find_first_device(UCLASS_PCH, &pch);
 	if (ret)
 		return ret;
 	if (!pch)
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#44 lore raw

PATCH v2 30/34 x86: broadwell: Add support for serial I/O devices

In reply to Simon Glass

Add support for initing the I2C device and ADSP on broadwell. These are needed for sound to work.

---

Changes in v2: None

 arch/x86/cpu/broadwell/pch.c                  | 110 ++++++++++++++++++
 .../x86/include/asm/arch-broadwell/serialio.h |  82 +++++++++++++
 2 files changed, 192 insertions(+)
 create mode 100644 arch/x86/include/asm/arch-broadwell/serialio.h
arch/x86/cpu/broadwell/pch.c +110 -0
index 82506ba35e3..73d3d3b5154 100644
@@ -20,7 +20,9 @@
 #include <asm/arch/pch.h>
 #include <asm/arch/pm.h>
 #include <asm/arch/rcb.h>
+#include <asm/arch/serialio.h>
 #include <asm/arch/spi.h>
+#include <dm/uclass-internal.h>
 
 #define BIOS_CTRL	0xdc
 
@@ -456,6 +458,111 @@ static void systemagent_init(void)
 	cpu_set_power_limits(28);
 }
 
+/* Enable LTR Auto Mode for D21:F1-F6 */
+static void serialio_d21_ltr(u32 bar0)
+{
+	/* 1. Program BAR0 + 808h[2] = 0b */
+	clrbits_le32(bar0 + SIO_REG_PPR_GEN, SIO_REG_PPR_GEN_LTR_MODE_MASK);
+
+	/* 2. Program BAR0 + 804h[1:0] = 00b */
+	clrbits_le32(bar0 + SIO_REG_PPR_RST, SIO_REG_PPR_RST_ASSERT);
+
+	/* 3. Program BAR0 + 804h[1:0] = 11b */
+	setbits_le32(bar0 + SIO_REG_PPR_RST, SIO_REG_PPR_RST_ASSERT);
+
+	/* 4. Program BAR0 + 814h[31:0] = 00000000h */
+	writel(0, bar0 + SIO_REG_AUTO_LTR);
+}
+
+/* Select I2C voltage of 1.8V or 3.3V */
+static void serialio_i2c_voltage_sel(u32 bar0, uint voltage)
+{
+	clrsetbits_le32(bar0 + SIO_REG_PPR_GEN, SIO_REG_PPR_GEN_VOLTAGE_MASK,
+			SIO_REG_PPR_GEN_VOLTAGE(voltage));
+}
+
+/* Put Serial IO D21:F0-F6 device into desired mode */
+static void serialio_d21_mode(int sio_index, int int_pin, bool acpi_mode)
+{
+	u32 portctrl = SIO_IOBP_PORTCTRL_PM_CAP_PRSNT;
+
+	/* Snoop select 1 */
+	portctrl |= SIO_IOBP_PORTCTRL_SNOOP_SELECT(1);
+
+	/* Set interrupt pin */
+	portctrl |= SIO_IOBP_PORTCTRL_INT_PIN(int_pin);
+
+	if (acpi_mode) {
+		/* Enable ACPI interrupt mode */
+		portctrl |= SIO_IOBP_PORTCTRL_ACPI_IRQ_EN;
+	}
+
+	pch_iobp_update(SIO_IOBP_PORTCTRLX(sio_index), 0, portctrl);
+}
+
+/* Init sequence to be run once, done as part of D21:F0 (SDMA) init */
+static void serialio_init_once(bool acpi_mode)
+{
+	if (acpi_mode) {
+		/* Enable ACPI IRQ for IRQ13, IRQ7, IRQ6, IRQ5 in RCBA */
+		setbits_le32(RCB_REG(ACPIIRQEN),
+			     1 << 13 | 1 << 7 | 1 << 6 | 1 << 5);
+	}
+
+	/* Program IOBP CB000154h[12,9:8,4:0] = 1001100011111b */
+	pch_iobp_update(SIO_IOBP_GPIODF, ~0x0000131f, 0x0000131f);
+
+	/* Program IOBP CB000180h[5:0] = 111111b (undefined register) */
+	pch_iobp_update(0xcb000180, ~0x0000003f, 0x0000003f);
+}
+
+/**
+ * pch_serialio_init() - set up serial I/O devices
+ *
+ * @return 0 if OK, -ve on error
+ */
+static int pch_serialio_init(void)
+{
+	struct udevice *dev, *hda;
+	bool acpi_mode = true;
+	u32 bar0, bar1;
+	int ret;
+
+	ret = uclass_find_first_device(UCLASS_I2C, &dev);
+	if (ret)
+		return ret;
+	bar0 = dm_pci_read_bar32(dev, 0);
+	if (!bar0)
+		return -EINVAL;
+	bar1 = dm_pci_read_bar32(dev, 1);
+	if (!bar1)
+		return -EINVAL;
+
+	serialio_init_once(acpi_mode);
+	serialio_d21_mode(SIO_ID_SDMA, SIO_PIN_INTB, acpi_mode);
+
+	serialio_d21_ltr(bar0);
+	serialio_i2c_voltage_sel(bar0, 1); /* Select 1.8V always */
+	serialio_d21_mode(SIO_ID_I2C0, SIO_PIN_INTC, acpi_mode);
+	setbits_le32(bar1 + PCH_PCS, PCH_PCS_PS_D3HOT);
+
+	clrbits_le32(bar1 + PCH_PCS, PCH_PCS_PS_D3HOT);
+
+	setbits_le32(bar0 + SIO_REG_PPR_CLOCK, SIO_REG_PPR_CLOCK_EN);
+
+	/* Manually find the High-definition audio, to turn it off */
+	ret = dm_pci_bus_find_bdf(PCI_BDF(0, 0x1b, 0), &hda);
+	if (ret)
+		return -ENOENT;
+	dm_pci_clrset_config8(hda, 0x43, 0, 0x6f);
+
+	/* Route I/O buffers to ADSP function */
+	dm_pci_clrset_config8(hda, 0x42, 0, 1 << 7 | 1 << 6);
+	log_debug("HDA disabled, I/O buffers routed to ADSP\n");
+
+	return 0;
+}
+
 static int broadwell_pch_init(struct udevice *dev)
 {
 	int ret;
@@ -482,6 +589,9 @@ static int broadwell_pch_init(struct udevice *dev)
 		return ret;
 	pch_pm_init(dev);
 	pch_cg_init(dev);
+	ret = pch_serialio_init();
+	if (ret)
+		return ret;
 	systemagent_init();
 
 	return 0;
arch/x86/include/asm/arch-broadwell/serialio.h new file +82 -0

82 lines not on this page · raw message

Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#45 lore raw

PATCH v2 31/34 x86: sound: Add support for broadwell I2S

In reply to Simon Glass

I2S is used to send digital audio data to an audio codec. Add support for this on broadwell.

---

Changes in v2: None

 drivers/sound/Makefile        |   1 +
 drivers/sound/broadwell_i2s.c | 306 ++++++++++++++++++++++++++++++++++
 drivers/sound/broadwell_i2s.h | 301 +++++++++++++++++++++++++++++++++
 3 files changed, 608 insertions(+)
 create mode 100644 drivers/sound/broadwell_i2s.c
 create mode 100644 drivers/sound/broadwell_i2s.h
drivers/sound/Makefile +1 -0
index 2a6fd7420c2..07b590156c9 100644
@@ -17,4 +17,5 @@ obj-$(CONFIG_SOUND_MAX98090)	+= max98090.o maxim_codec.o
 obj-$(CONFIG_SOUND_MAX98095)	+= max98095.o maxim_codec.o
 obj-$(CONFIG_SOUND_INTEL_HDA)	+= hda_codec.o
 obj-$(CONFIG_SOUND_I8254)	+= i8254_beep.o
+obj-$(CONFIG_INTEL_BROADWELL)	+= broadwell_i2s.o
 obj-$(CONFIG_SOUND_IVYBRIDGE)	+= ivybridge_sound.o
drivers/sound/broadwell_i2s.c new file +306 -0

306 lines not on this page · raw message

drivers/sound/broadwell_i2s.h new file +301 -0

301 lines not on this page · raw message

Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#46 lore raw

PATCH v2 32/34 sound: Add a driver for RealTek RT5677

In reply to Simon Glass

This audio codec is used on samus. Add a driver for it.

---

Changes in v2: None

 drivers/sound/Kconfig  |    9 +
 drivers/sound/Makefile |    1 +
 drivers/sound/rt5677.c |  334 ++++++++++
 drivers/sound/rt5677.h | 1428 ++++++++++++++++++++++++++++++++++++++++
 4 files changed, 1772 insertions(+)
 create mode 100644 drivers/sound/rt5677.c
 create mode 100644 drivers/sound/rt5677.h
drivers/sound/Kconfig +9 -0
index c1847df07fa..6e9dcefcb90 100644
@@ -95,6 +95,15 @@ config SOUND_MAX98095
 	  audio data and I2C for codec control. At present it only works
 	  with the Samsung I2S driver.
 
+config SOUND_RT5677
+	bool "Support Realtek RT5677 audio codec"
+	depends on SOUND
+	help
+	  Enable the Realtek RT5677 audio codec. This is an I2S device used on
+	  some Chromebooks from around 2015 ('auron'). It is configured using
+	  an I2C interface and supports multiple sound inputs and outputs,
+	  including digital microphones.
+
 config SOUND_SANDBOX
 	bool "Support sandbox emulated audio codec"
 	depends on SANDBOX && SOUND
drivers/sound/Makefile +1 -0
index 07b590156c9..fe3ccd6e5c6 100644
@@ -17,5 +17,6 @@ obj-$(CONFIG_SOUND_MAX98090)	+= max98090.o maxim_codec.o
 obj-$(CONFIG_SOUND_MAX98095)	+= max98095.o maxim_codec.o
 obj-$(CONFIG_SOUND_INTEL_HDA)	+= hda_codec.o
 obj-$(CONFIG_SOUND_I8254)	+= i8254_beep.o
+obj-$(CONFIG_SOUND_RT5677)	+= rt5677.o
 obj-$(CONFIG_INTEL_BROADWELL)	+= broadwell_i2s.o
 obj-$(CONFIG_SOUND_IVYBRIDGE)	+= ivybridge_sound.o
drivers/sound/rt5677.c new file +334 -0

334 lines not on this page · raw message

drivers/sound/rt5677.h new file +1428 -0

1,428 lines not on this page · raw message

Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#47 lore raw

PATCH v2 33/34 x86: sound: Add sound support for samus (broadwell)

In reply to Simon Glass

Add a sound driver for samus which ties together the audio codec and I2S controller.

For now broadwell_sound is commented out in the makefile since we cannot compile it without sound support enabled. The next commit fixes this.

---

Changes in v2: None

 drivers/sound/Makefile          |  2 +-
 drivers/sound/broadwell_sound.c | 65 +++++++++++++++++++++++++++++++++
 2 files changed, 66 insertions(+), 1 deletion(-)
 create mode 100644 drivers/sound/broadwell_sound.c
drivers/sound/Makefile +1 -1
index fe3ccd6e5c6..452a5a17616 100644
@@ -18,5 +18,5 @@ obj-$(CONFIG_SOUND_MAX98095)	+= max98095.o maxim_codec.o
 obj-$(CONFIG_SOUND_INTEL_HDA)	+= hda_codec.o
 obj-$(CONFIG_SOUND_I8254)	+= i8254_beep.o
 obj-$(CONFIG_SOUND_RT5677)	+= rt5677.o
-obj-$(CONFIG_INTEL_BROADWELL)	+= broadwell_i2s.o
+obj-$(CONFIG_INTEL_BROADWELL)	+= broadwell_i2s.o #broadwell_sound.o
 obj-$(CONFIG_SOUND_IVYBRIDGE)	+= ivybridge_sound.o
drivers/sound/broadwell_sound.c new file +65 -0

65 lines not on this page · raw message

Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#48 lore raw

PATCH v2 33/34 x86: sound: Add sound support for samus (broadwell)

Quoted 8 lines · Simon Glass

Add a sound driver for samus which ties together the audio codec and I2S controller.

For now broadwell_sound is commented out in the makefile since we cannot compile it without sound support enabled. The next commit fixes this.

Quoted 9 lines

---

Changes in v2: None

 drivers/sound/Makefile          |  2 +-
 drivers/sound/broadwell_sound.c | 65 +++++++++++++++++++++++++++++++++
 2 files changed, 66 insertions(+), 1 deletion(-)
 create mode 100644 drivers/sound/broadwell_sound.c
#49 lore raw

PATCH v2 34/34 x86: Add sound support for samus

In reply to Simon Glass

Enable sound on samus using the broadwell I2S and an RT5677 audio codec.

---

Changes in v2: None

 arch/x86/dts/chromebook_samus.dts  | 49 +++++++++++++++++++++++++++---
 configs/chromebook_samus_defconfig |  9 ++++++
 drivers/sound/Makefile             |  2 +-
 3 files changed, 55 insertions(+), 5 deletions(-)
arch/x86/dts/chromebook_samus.dts +45 -4
index b58936b4ac9..ad687ce2cc3 100644
@@ -494,10 +494,20 @@
 			intel,pre-graphics-delay = <200>;
 		};
 
-		me at 16,0 {
-			reg = <0x0000b000 0 0 0 0>;
-			compatible = "intel,me";
-			u-boot,dm-pre-reloc;
+		adsp at 13,0 {
+			reg = <0x00009800 0 0 0 0>;
+			compatible = "intel,wildcatpoint-adsp";
+			intel,adsp-d3-pg-enable = <0>;
+			intel,adsp-sram-pg-enable = <0>;
+			intel,sio-acpi-mode;
+			#address-cells = <1>;
+			#size-cells = <0>;
+
+			i2s: shim {
+				compatible = "intel,broadwell-i2s";
+				#sound-dai-cells = <1>;
+				reg = <0xfb000 0xfc000 0xfd000>;
+			};
 		};
 
 		usb_1: usb at 14,0 {
@@ -505,6 +515,25 @@
 			compatible = "xhci-pci";
 		};
 
+		i2c0: i2c at 15,1 {
+			reg = <0x0000a900 0 0 0 0>;
+			compatible = "snps,designware-i2c";
+			#address-cells = <1>;
+			#size-cells = <0>;
+
+			rt5677: rt5677 at 2c {
+				compatible = "realtek,rt5677";
+				#sound-dai-cells = <1>;
+				reg = <0x2c>;
+			};
+		};
+
+		me at 16,0 {
+			reg = <0x0000b000 0 0 0 0>;
+			compatible = "intel,me";
+			u-boot,dm-pre-reloc;
+		};
+
 		usb_0: usb at 1d,0 {
 			status = "disabled";
 			reg = <0x0000e800 0 0 0 0>;
@@ -626,4 +655,16 @@
 		};
 	};
 
+	sound {
+		compatible = "google,samus-sound";
+		codec-enable-gpio = <&gpio_b 11 GPIO_ACTIVE_HIGH>;
+		cpu {
+			sound-dai = <&i2s 0>;
+		};
+
+		codec {
+			sound-dai = <&rt5677 0>;
+		};
+	};
+
 };
configs/chromebook_samus_defconfig +9 -0
index 745cb3db0a6..7e2453f2440 100644
@@ -17,6 +17,7 @@ CONFIG_BOOTSTAGE_REPORT=y
 CONFIG_USE_BOOTARGS=y
 CONFIG_BOOTARGS="root=/dev/sdb3 init=/sbin/init rootwait ro"
 CONFIG_SYS_CONSOLE_INFO_QUIET=y
+CONFIG_LOG_DEFAULT_LEVEL=7
 CONFIG_MISC_INIT_R=y
 CONFIG_DISPLAY_BOARDINFO_LATE=y
 CONFIG_LAST_STAGE_INIT=y
@@ -24,7 +25,9 @@ CONFIG_HUSH_PARSER=y
 CONFIG_CMD_CPU=y
 # CONFIG_CMD_FLASH is not set
 CONFIG_CMD_GPIO=y
+CONFIG_CMD_I2C=y
 CONFIG_CMD_PART=y
+CONFIG_CMD_SATA=y
 CONFIG_CMD_SF=y
 CONFIG_CMD_SPI=y
 CONFIG_CMD_USB=y
@@ -33,6 +36,7 @@ CONFIG_CMD_DHCP=y
 # CONFIG_CMD_NFS is not set
 CONFIG_CMD_PING=y
 CONFIG_CMD_TIME=y
+CONFIG_CMD_SOUND=y
 CONFIG_CMD_BOOTSTAGE=y
 CONFIG_CMD_TPM=y
 CONFIG_CMD_TPM_TEST=y
@@ -48,9 +52,14 @@ CONFIG_DEFAULT_DEVICE_TREE="chromebook_samus"
 CONFIG_REGMAP=y
 CONFIG_SYSCON=y
 CONFIG_CPU=y
+CONFIG_DM_I2C=y
+CONFIG_SYS_I2C_DW=y
 CONFIG_CROS_EC=y
 CONFIG_CROS_EC_LPC=y
 CONFIG_SYS_NS16550=y
+CONFIG_SOUND=y
+CONFIG_SOUND_I8254=y
+CONFIG_SOUND_RT5677=y
 CONFIG_SPI=y
 CONFIG_TPM_TIS_LPC=y
 CONFIG_USB_STORAGE=y
drivers/sound/Makefile +1 -1
index 452a5a17616..e155041ff5f 100644
@@ -18,5 +18,5 @@ obj-$(CONFIG_SOUND_MAX98095)	+= max98095.o maxim_codec.o
 obj-$(CONFIG_SOUND_INTEL_HDA)	+= hda_codec.o
 obj-$(CONFIG_SOUND_I8254)	+= i8254_beep.o
 obj-$(CONFIG_SOUND_RT5677)	+= rt5677.o
-obj-$(CONFIG_INTEL_BROADWELL)	+= broadwell_i2s.o #broadwell_sound.o
+obj-$(CONFIG_INTEL_BROADWELL)	+= broadwell_i2s.o broadwell_sound.o
 obj-$(CONFIG_SOUND_IVYBRIDGE)	+= ivybridge_sound.o
Signature · 1 line
2.21.0.rc0.258.g878e2cd30e-goog
#50 lore raw

v2 cover x86: Add support for sound

In reply to Simon Glass
Quoted 66 lines · Simon Glass

At present no x86 boards support sound. This series adds several options:

- legacy beep
- Intel HDA
- I2S with codec

These are enabled on coreboot, link and samus, respectively.

A number of related patches are included to make the above work.

Changes in v2:
- Add a PCI ID for the HD audio class
- Add a link to the HD audio PDF fie
- Add a note about undocumented IOMMU values
- Add a note as to why the default log level must be set to LOGL_DEBUG
- Add an example to the commit message for clarity
- Add comments next to verbs in the DT mentioning the HDA spec
- Add new patch to fix up Kconfig log level names
- Add the log-level names into the Kconfig help
- Adjust code to use #defines instead of open-coded values
- Also remove this feature from the TODO
- Change the driver name to i8254_drv
- Correct the default log level to 6 (LOGL_INFO)
- Drop useless 'set' comment
- Fix capital letter in Ivybridge in Kconfig help
- Fix comment about @return value in ioctl() method and function
- Fix ordering of PCI IDs
- Fix use of C++ comments in two places
- Move comment about silencing on error so that it is clearer
- Rename PPC_PORTB to SYSCTL_PORTB
- Rename sound driver to bd82x6x-hda
- Update in codec_init() to use defines for codec access
- Use "sandbox-pch" for the compatible string, for consistency
- Use PIT_TICK_RATE instead of creating a new define
- Use a more generic "intel,hd-audio" as the compatible string
- Use defines for all azalia verbs in binding file
- User lower-case hex consistently in the DT

Simon Glass (34):
log: Fix up Kconfig log level names
log: Add a Kconfig option to set the default log level
log: Add documentation for convenience functions
log: Allow #define LOG_DEBUG to enable logging in a file
dm: syscon: Don't require a regmap for PCI devices
i2c: designware: Add error checking on init
pci: Don't export pci_hose_config_device()
pci: Fix comment in struct pci_child_platdata
gpio: Show inactive GPIOs when explicitly requested
gpio: Use case-insentive matching on the GPIO name
pci: Add IDs for Intel high-definition audio
sandbox: Add a note about the growing state_info struct
x86: Adjust I/O macros to work on 64-bit machines
x86: samus: Increase the pre-reloc memory
x86: link: Increase malloc size and decrease code size
x86: sandbox: pch: Add a CONFIG option for PCH
sandbox: pch: Add a test for the PCH uclass
pch: Add ioctl support
x86: ivybridge: Add a way to get the HDA config setting
sound: Mark sound_setup() as optional
sound: Add uclass operations for beeping
sound: Add support for Intel HDA
sandbox: sound: Silence sound for testing
sound: x86: link: Add sound support
sound: x86: Add beeping support in i8254
sound: Add a driver for the i8254 beep

Quoted 9 lines

x86: coreboot: Enable the beeper sound driver
x86: broadwell: Add support for the ADSP
x86: broadwell: Don't bother probing the PCH for pinctrl
x86: broadwell: Add support for serial I/O devices
x86: sound: Add support for broadwell I2S
sound: Add a driver for RealTek RT5677
x86: sound: Add sound support for samus (broadwell)
x86: Add sound support for samus

Series applied to u-boot-x86, thanks!