Skip to contentSkip to conversation

Mailbox: Provide support STi based platforms

PATCH RESEND Superseded Lee Jones 3 participants 10 messages Oct 16, 2015 – Oct 17, 2015

Subsystems (2)
Device tree bindings Mailbox API
lore
Share & tools
lore

Lee Jones

author
#0 lore raw

cover Mailbox: Provide support STi based platforms

Hi Jassi,

[Resending the updated patch-set this time]

This should be it. Exciting times!

ST's platforms currently support a maximum of 5 Mailboxes, one for each of the supported co-processors situated on the platform. Each Mailbox is divided up into 4 instances which consist of 32 channels. Messages are passed between the application and co-processors using shared memory areas.

Also included in the set is an example Client which should be generic enough to use as a testing environment for all Mailbox IPs which use shared memory or the API to pass messages. With small changes, it should also be able to test more extravagant implementations.

v3 -> v4:
- More protection in MBOX_BASE() MACRO
- Remove unused tx-only marker

v2 => v3:
- Controller
- Shared (DT <=> Device Driver) defines removed
- Return actual error codes instead of NULL in xlate()
- Not passing 'direction' configuration via DT anymore
- Removed all accompanying configuration checking code
- Test F/W
- Allow Rx-only and Tx-only controllers to be tested
- Cater for controllers requiring pre-Tx 'signal' to be sent
- Supply 'wr-' message debugfs file
- When written to, will Tx message through Mailbox Framework
- When read from, will print out a hexdump of Rx'ed message

v1 => v2:
- New MACRO() to obtain base address for a given instance
- Move locking into the structure it protects
- Stop checking for 'ready' state when sending data
- Don't clear channel data (that belongs to the API)
- #define register offsets instead of providing via pdata
- Register driver with module_platform_driver()

Lee Jones (5):
mailbox: dt: Supply bindings for ST's Mailbox IP
mailbox: Add support for ST's Mailbox IP
ARM: STi: stih407-family: Add nodes for Mailbox
mailbox: Add generic mechanism for testing Mailbox Controllers
ARM: STi: DT: STiH407: Enable Mailbox testing facility

 .../devicetree/bindings/mailbox/sti-mailbox.txt    |  51 ++
 arch/arm/boot/dts/stih407-family.dtsi              |  39 ++
 drivers/mailbox/Kconfig                            |  14 +
 drivers/mailbox/Makefile                           |   4 +
 drivers/mailbox/mailbox-sti.c                      | 513 +++++++++++++++++++++
 drivers/mailbox/mailbox-test.c                     | 361 +++++++++++++++
 6 files changed, 982 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/mailbox/sti-mailbox.txt
 create mode 100644 drivers/mailbox/mailbox-sti.c
 create mode 100644 drivers/mailbox/mailbox-test.c
Signature · 1 line
1.9.1
author
#1 lore raw

PATCH v4 1/5 mailbox: dt: Supply bindings for ST's Mailbox IP

Signed-off-by Lee Jones <[email protected]>
---
 .../devicetree/bindings/mailbox/sti-mailbox.txt    | 51 ++++++++++++++++++++++
 1 file changed, 51 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/mailbox/sti-mailbox.txt
Documentation/devicetree/bindings/mailbox/sti-mailbox.txt new file +51 -0
new file mode 100644
index 0000000..b61eec9
@@ -0,0 +1,51 @@
+ST Microelectronics Mailbox Driver
+
+Each ST Mailbox IP currently consists of 4 instances of 32 channels.  Messages
+are passed between Application and Remote processors using shared memory.
+
+Controller
+----------
+
+Required properties:
+- compatible		: Should be "st,stih407-mailbox"
+- reg			: Offset and length of the device's register set
+- mbox-name		: Name of the mailbox
+- #mbox-cells:		: Must be 2
+			  <&phandle instance channel direction>
+			    phandle   : Label name of controller
+			    instance  : Instance number
+			    channel   : Channel number
+
+Optional properties
+- interrupts		: Contains the IRQ line for a Rx mailbox
+
+Example:
+
+mailbox0: mailbox at 0  {
+	compatible	= "st,stih407-mailbox";
+	reg		= <0x08f00000 0x1000>;
+	interrupts	= <GIC_SPI 1 IRQ_TYPE_NONE>;
+	#mbox-cells	= <2>;
+	mbox-name	= "a9";
+};
+
+Client
+------
+
+Required properties:
+- compatible		: Many (See the client docs)
+- reg			: Shared (between Application and Remote) memory address
+- mboxes		: Standard property to specify a Mailbox (See ./mailbox.txt)
+			  Cells must match 'mbox-cells' (See Controller docs above)
+
+Optional properties
+- mbox-names		: Name given to channels seen in the 'mboxes' property.
+
+Example:
+
+mailbox_test {
+	compatible	= "mailbox_test";
+	reg		= <0x[shared_memory_address], [shared_memory_size]>;
+	mboxes		= <&mailbox2 0 1>, <&mailbox0 2 1>;
+	mbox-names	= "tx",	"rx";
+};
Signature · 1 line
1.9.1
author
#2 lore raw

PATCH v4 2/5 mailbox: Add support for ST's Mailbox IP

In reply to Lee Jones

ST's platforms currently support a maximum of 5 Mailboxes, one for each of the supported co-processors situated on the platform. Each Mailbox is divided up into 4 instances which consist of 32 channels. Messages are passed between the application and co-processors using shared memory areas. It is the Client's responsibility to manage these areas.

Signed-off-by Lee Jones <[email protected]>
---
 drivers/mailbox/Kconfig       |   7 +
 drivers/mailbox/Makefile      |   2 +
 drivers/mailbox/mailbox-sti.c | 513 ++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 522 insertions(+)
 create mode 100644 drivers/mailbox/mailbox-sti.c
drivers/mailbox/Kconfig +7 -0
index bbec500..7b53386 100644
@@ -71,4 +71,11 @@ config BCM2835_MBOX
 	  the services of the Videocore. Say Y here if you want to use the
 	  BCM2835 Mailbox.
 
+config STI_MBOX
+	tristate "STI Mailbox framework support"
+	depends on ARCH_STI && OF
+	help
+	  Mailbox implementation for STMicroelectonics family chips with
+	  hardware for interprocessor communication.
+
 endif
drivers/mailbox/Makefile +2 -0
index 8e6d822..7cb4766 100644
@@ -13,3 +13,5 @@ obj-$(CONFIG_PCC)		+= pcc.o
 obj-$(CONFIG_ALTERA_MBOX)	+= mailbox-altera.o
 
 obj-$(CONFIG_BCM2835_MBOX)	+= bcm2835-mailbox.o
+
+obj-$(CONFIG_STI_MBOX)		+= mailbox-sti.o
drivers/mailbox/mailbox-sti.c new file +513 -0

513 lines not on this page · raw message

Signature · 1 line
1.9.1
author
#3 lore raw

PATCH v4 3/5 ARM: STi: stih407-family: Add nodes for Mailbox

In reply to Lee Jones

This patch supplies the Mailbox Controller nodes. In order to request channels, these nodes will be referenced by Mailbox Client nodes.

Signed-off-by Lee Jones <[email protected]>
---
 arch/arm/boot/dts/stih407-family.dtsi | 33 +++++++++++++++++++++++++++++++++
 1 file changed, 33 insertions(+)
arch/arm/boot/dts/stih407-family.dtsi +33 -0
index ae05277..0e5d3ea 100644
@@ -610,5 +610,38 @@
 			clocks		= <&clk_sysin>;
 			st,pwm-num-chan = <4>;
 		};
+
+		mailbox0: mailbox at 0  {
+			compatible	= "st,stih407-mailbox";
+			reg		= <0x8f00000 0x1000>;
+			interrupts	= <GIC_SPI 1 IRQ_TYPE_NONE>;
+			#mbox-cells	= <2>;
+			mbox-name	= "a9";
+			status		= "okay";
+		};
+
+		mailbox1: mailbox at 1 {
+			compatible	= "st,stih407-mailbox";
+			reg		= <0x8f01000 0x1000>;
+			#mbox-cells	= <2>;
+			mbox-name	= "st231_gp_1";
+			status		= "okay";
+		};
+
+		mailbox2: mailbox at 2 {
+			compatible	= "st,stih407-mailbox";
+			reg		= <0x8f02000 0x1000>;
+			#mbox-cells	= <2>;
+			mbox-name	= "st231_gp_0";
+			status		= "okay";
+		};
+
+		mailbox3: mailbox at 3 {
+			compatible	= "st,stih407-mailbox";
+			reg		= <0x8f03000 0x1000>;
+			#mbox-cells	= <2>;
+			mbox-name	= "st231_audio_video";
+			status		= "okay";
+		};
 	};
 };
Signature · 1 line
1.9.1
author
#4 lore raw

PATCH v4 4/5 mailbox: Add generic mechanism for testing Mailbox Controllers

In reply to Lee Jones

This particular Client implementation uses shared memory in order to pass messages between Mailbox users; however, it can be easily hacked to support any type of Controller.

Signed-off-by Lee Jones <[email protected]>
---
 drivers/mailbox/Kconfig        |   7 +
 drivers/mailbox/Makefile       |   2 +
 drivers/mailbox/mailbox-test.c | 361 +++++++++++++++++++++++++++++++++++++++++
 3 files changed, 370 insertions(+)
 create mode 100644 drivers/mailbox/mailbox-test.c
drivers/mailbox/Kconfig +7 -0
index 7b53386..546d05f 100644
@@ -78,4 +78,11 @@ config STI_MBOX
 	  Mailbox implementation for STMicroelectonics family chips with
 	  hardware for interprocessor communication.
 
+config MAILBOX_TEST
+	tristate "Mailbox Test Client"
+	depends on OF
+	help
+	  Test client to help with testing new Controller driver
+	  implementations.
+
 endif
drivers/mailbox/Makefile +2 -0
index 7cb4766..92435ef 100644
@@ -2,6 +2,8 @@
 
 obj-$(CONFIG_MAILBOX)		+= mailbox.o
 
+obj-$(CONFIG_MAILBOX_TEST)	+= mailbox-test.o
+
 obj-$(CONFIG_ARM_MHU)	+= arm_mhu.o
 
 obj-$(CONFIG_PL320_MBOX)	+= pl320-ipc.o
drivers/mailbox/mailbox-test.c new file +361 -0

361 lines not on this page · raw message

Signature · 1 line
1.9.1
build bot
#5 lore raw

Re: [PATCH v4 4/5] mailbox: Add generic mechanism for testing Mailbox Controllers

reproduce:

Hi Lee,

[auto build test WARNING on v4.3-rc5 -- if it's inappropriate base, please suggest rules for selecting the more suitable base]

url:    https://github.com/0day-ci/linux/commits/Lee-Jones/Mailbox-Provide-support-STi-based-platforms/20151016-152650
config: x86_64-allmodconfig (attached as .config)
reproduce:
        # save the attached .config to linux build tree
        make ARCH=x86_64

All warnings (new ones prefixed by >>):

drivers/mailbox/mailbox-test.c:225:36: sparse: incorrect type in argument 6 (different address spaces)
drivers/mailbox/mailbox-test.c:225:36: expected void const *buf
drivers/mailbox/mailbox-test.c:225:36: got void [noderef] <asn:2>*mmio
drivers/mailbox/mailbox-test.c:226:17: sparse: incorrect type in argument 2 (different address spaces)
drivers/mailbox/mailbox-test.c:226:17: expected void const *from
drivers/mailbox/mailbox-test.c:226:17: got void [noderef] <asn:2>*mmio
drivers/mailbox/mailbox-test.c:243:25: sparse: incorrect type in argument 1 (different address spaces)
drivers/mailbox/mailbox-test.c:243:25: expected void *to
drivers/mailbox/mailbox-test.c:243:25: got void [noderef] <asn:2>*mmio
drivers/mailbox/mailbox-test.c:245:25: sparse: incorrect type in argument 1 (different address spaces)
drivers/mailbox/mailbox-test.c:245:25: expected void *to
drivers/mailbox/mailbox-test.c:245:25: got void [noderef] <asn:2>*mmio
drivers/mailbox/mailbox-test.c: In function 'mbox_test_signal_write':

>> drivers/mailbox/mailbox-test.c:56:4: warning: format '%d' expects argument of type 'int', but argument 3 has type 'size_t {aka long unsigned int}' [-Wformat=]

"Signal length %d greater than max allowed %d\n",
^
drivers/mailbox/mailbox-test.c: In function 'mbox_test_message_write':
drivers/mailbox/mailbox-test.c:95:4: warning: format '%d' expects argument of type 'int', but argument 3 has type 'size_t {aka long unsigned int}' [-Wformat=]
"Message length %d greater than max allowed %d\n",
^

sparse warnings: (new ones prefixed by >>)

>> drivers/mailbox/mailbox-test.c:225:36: sparse: incorrect type in argument 6 (different address spaces)
   drivers/mailbox/mailbox-test.c:225:36:    expected void const *buf
   drivers/mailbox/mailbox-test.c:225:36:    got void [noderef] <asn:2>*mmio
>> drivers/mailbox/mailbox-test.c:226:17: sparse: incorrect type in argument 2 (different address spaces)
   drivers/mailbox/mailbox-test.c:226:17:    expected void const *from
   drivers/mailbox/mailbox-test.c:226:17:    got void [noderef] <asn:2>*mmio
>> drivers/mailbox/mailbox-test.c:243:25: sparse: incorrect type in argument 1 (different address spaces)

drivers/mailbox/mailbox-test.c:243:25: expected void *to
drivers/mailbox/mailbox-test.c:243:25: got void [noderef] <asn:2>*mmio
drivers/mailbox/mailbox-test.c:245:25: sparse: incorrect type in argument 1 (different address spaces)
drivers/mailbox/mailbox-test.c:245:25: expected void *to
drivers/mailbox/mailbox-test.c:245:25: got void [noderef] <asn:2>*mmio
drivers/mailbox/mailbox-test.c: In function 'mbox_test_signal_write':
drivers/mailbox/mailbox-test.c:56:4: warning: format '%d' expects argument of type 'int', but argument 3 has type 'size_t {aka long unsigned int}' [-Wformat=]
"Signal length %d greater than max allowed %d\n",
^
drivers/mailbox/mailbox-test.c: In function 'mbox_test_message_write':
drivers/mailbox/mailbox-test.c:95:4: warning: format '%d' expects argument of type 'int', but argument 3 has type 'size_t {aka long unsigned int}' [-Wformat=]
"Message length %d greater than max allowed %d\n",
^

vim +56 drivers/mailbox/mailbox-test.c

    50			dev_err(tdev->dev, "Channel cannot do Tx\n");
    51			return -EINVAL;
    52		}
    53	
    54		if (count > MBOX_MAX_SIG_LEN) {
    55			dev_err(tdev->dev,
  > 56				"Signal length %d greater than max allowed %d\n",
    57				count, MBOX_MAX_SIG_LEN);
    58			return -EINVAL;
    59		}
    60	
    61		tdev->signal = kzalloc(MBOX_MAX_SIG_LEN, GFP_KERNEL);
    62		if (!tdev->signal)
    63			return -ENOMEM;
    64	
    65		ret = copy_from_user(tdev->signal, userbuf, count);
    66		if (ret) {
    67			kfree(tdev->signal);
    68			return -EFAULT;
    69		}
    70	
    71		return ret < 0 ? ret : count;
    72	}
    73	
    74	static const struct file_operations mbox_test_signal_ops = {
    75		.write	= mbox_test_signal_write,
    76		.open	= simple_open,
    77		.llseek	= generic_file_llseek,
    78	};
    79	
    80	static ssize_t mbox_test_message_write(struct file *filp,
    81					       const char __user *userbuf,
    82					       size_t count, loff_t *ppos)
    83	{
    84		struct mbox_test_device *tdev = filp->private_data;
    85		void *data;
    86		int ret;
    87	
    88		if (!tdev->tx_channel) {
    89			dev_err(tdev->dev, "Channel cannot do Tx\n");
    90			return -EINVAL;
    91		}
    92	
    93		if (count > MBOX_MAX_MSG_LEN) {
    94			dev_err(tdev->dev,
    95				"Message length %d greater than max allowed %d\n",
    96				count, MBOX_MAX_MSG_LEN);
    97			return -EINVAL;
    98		}
    99	
   100		tdev->message = kzalloc(MBOX_MAX_MSG_LEN, GFP_KERNEL);
   101		if (!tdev->message)
   102			return -ENOMEM;
   103	
   104		ret = copy_from_user(tdev->message, userbuf, count);
   105		if (ret) {
   106			ret = -EFAULT;
   107			goto out;
   108		}
   109	
   110		/*
   111		 * A separate signal is only of use if there is
   112		 * MMIO to subsequently pass the message through
   113		 */
   114		if (tdev->mmio && tdev->signal) {
   115			print_hex_dump(KERN_INFO, "Client: Sending: Signal: ", DUMP_PREFIX_ADDRESS,
   116				       MBOX_BYTES_PER_LINE, 1, tdev->signal, MBOX_MAX_SIG_LEN, true);
   117	
   118			data = tdev->signal;
   119		} else
   120			data = tdev->message;
   121	
   122		print_hex_dump(KERN_INFO, "Client: Sending: Message: ", DUMP_PREFIX_ADDRESS,
   123			       MBOX_BYTES_PER_LINE, 1, tdev->message, MBOX_MAX_MSG_LEN, true);
   124	
   125		ret = mbox_send_message(tdev->tx_channel, data);
   126		if (ret < 0)
   127			dev_err(tdev->dev, "Failed to send message via mailbox\n");
   128	
   129	out:
   130		kfree(tdev->signal);
   131		kfree(tdev->message);
   132	
   133		return ret < 0 ? ret : count;
   134	}
   135	
   136	static ssize_t mbox_test_message_read(struct file *filp, char __user *userbuf,
   137					      size_t count, loff_t *ppos)
   138	{
   139		struct mbox_test_device *tdev = filp->private_data;
   140		unsigned long flags;
   141		char *touser, *ptr;
   142		int l = 0;
   143		int ret;
   144	
   145		touser = kzalloc(MBOX_HEXDUMP_MAX_LEN, GFP_KERNEL);
   146		if (!touser)
   147			return -ENOMEM;
   148	
   149		if (!tdev->rx_channel) {
   150			ret = snprintf(touser, 20, "<NO RX CAPABILITY>\n");
   151			ret = simple_read_from_buffer(userbuf, count, ppos,
   152						      touser, ret);
   153			goto out;
   154		}
   155	
   156		if (tdev->rx_buffer[0] == '\0') {
   157			ret = snprintf(touser, 9, "<EMPTY>\n");
   158			ret = simple_read_from_buffer(userbuf, count, ppos,
   159						      touser, ret);
   160			goto out;
   161		}
   162	
   163		spin_lock_irqsave(&tdev->lock, flags);
   164	
   165		ptr = tdev->rx_buffer;
   166		while (l < MBOX_HEXDUMP_MAX_LEN) {
   167			hex_dump_to_buffer(ptr,
   168					   MBOX_BYTES_PER_LINE,
   169					   MBOX_BYTES_PER_LINE, 1, touser + l,
   170					   MBOX_HEXDUMP_LINE_LEN, true);
   171	
   172			ptr += MBOX_BYTES_PER_LINE;
   173			l += MBOX_HEXDUMP_LINE_LEN;
   174			*(touser + (l - 1)) = '\n';
   175		}
   176		*(touser + l) = '\0';
   177	
   178		memset(tdev->rx_buffer, 0, MBOX_MAX_MSG_LEN);
   179	
   180		spin_unlock_irqrestore(&tdev->lock, flags);
   181	
   182		ret = simple_read_from_buffer(userbuf, count, ppos, touser, MBOX_HEXDUMP_MAX_LEN);
   183	out:
   184		kfree(touser);
   185		return ret;
   186	}
   187	
   188	static const struct file_operations mbox_test_message_ops = {
   189		.write	= mbox_test_message_write,
   190		.read	= mbox_test_message_read,
   191		.open	= simple_open,
   192		.llseek	= generic_file_llseek,
   193	};
   194	
   195	static int mbox_test_add_debugfs(struct platform_device *pdev,
   196					 struct mbox_test_device *tdev)
   197	{
   198		if (!debugfs_initialized())
   199			return 0;
   200	
   201		root_debugfs_dir = debugfs_create_dir("mailbox", NULL);
   202		if (!root_debugfs_dir) {
   203			dev_err(&pdev->dev, "Failed to create Mailbox debugfs\n");
   204			return -EINVAL;
   205		}
   206	
   207		debugfs_create_file("message", 0600, root_debugfs_dir,
   208				    tdev, &mbox_test_message_ops);
   209	
   210		debugfs_create_file("signal", 0200, root_debugfs_dir,
   211				    tdev, &mbox_test_signal_ops);
   212	
   213		return 0;
   214	}
   215	
   216	static void mbox_test_receive_message(struct mbox_client *client, void *message)
   217	{
   218		struct mbox_test_device *tdev = dev_get_drvdata(client->dev);
   219		unsigned long flags;
   220	
   221		spin_lock_irqsave(&tdev->lock, flags);
   222		if (tdev->mmio) {
   223			print_hex_dump(KERN_INFO, "Client: Received [MMIO]: ",
   224				       DUMP_PREFIX_ADDRESS, MBOX_BYTES_PER_LINE, 1,
 > 225				       tdev->mmio, MBOX_MAX_MSG_LEN, true);
 > 226			memcpy(tdev->rx_buffer, tdev->mmio, MBOX_MAX_MSG_LEN);
   227	
   228		} else if (message) {
   229			print_hex_dump(KERN_INFO, "Client: Received [API]: ",
   230				       DUMP_PREFIX_ADDRESS, MBOX_BYTES_PER_LINE, 1,
   231				       message, MBOX_MAX_MSG_LEN, true);
   232			memcpy(tdev->rx_buffer, message, MBOX_MAX_MSG_LEN);
   233		}
   234		spin_unlock_irqrestore(&tdev->lock, flags);
   235	}
   236	
   237	static void mbox_test_prepare_message(struct mbox_client *client, void *message)
   238	{
   239		struct mbox_test_device *tdev = dev_get_drvdata(client->dev);
   240	
   241		if (tdev->mmio) {
   242			if (tdev->signal)
 > 243				memcpy(tdev->mmio, tdev->message, MBOX_MAX_MSG_LEN);

244 else
245 memcpy(tdev->mmio, message, MBOX_MAX_MSG_LEN);
246 }

---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all Intel Corporation
-------------- next part --------------
A non-text attachment was scrubbed...
Name: .config.gz
Type: application/octet-stream
Size: 50056 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20151016/cde8a178/attachment-0001.obj>

author
#6 lore raw

PATCH 1/1 mailbox: Fix a couple of trivial static checker issues

mailbox: Fix a couple of trivial static checker issues

This patch deals with a few spelling, white space and type warnings reported by Intel's Kbuild Test Robot.

Reported-by kbuild test robot <[email protected]>
Signed-off-by Lee Jones <[email protected]>
---
 drivers/mailbox/mailbox-sti.c  |  2 +-
 drivers/mailbox/mailbox-test.c | 10 +++++-----
 2 files changed, 6 insertions(+), 6 deletions(-)
drivers/mailbox/mailbox-sti.c +1 -1
index f11d4e8..4835817 100644
@@ -64,7 +64,7 @@ struct sti_mbox_device {
 };
 
 /**
- * STi Mailbox platform specfic configuration
+ * STi Mailbox platform specific configuration
  *
  * @num_inst:	Maximum number of instances in one HW Mailbox
  * @num_chan:	Maximum number of channel per instance
drivers/mailbox/mailbox-test.c +5 -5
index cac1ba2..196144c 100644
@@ -22,7 +22,7 @@
 #define MBOX_MAX_SIG_LEN	8
 #define MBOX_MAX_MSG_LEN	128
 #define MBOX_BYTES_PER_LINE	16
-#define MBOX_HEXDUMP_LINE_LEN 	((MBOX_BYTES_PER_LINE * 4) + 2)
+#define MBOX_HEXDUMP_LINE_LEN	((MBOX_BYTES_PER_LINE * 4) + 2)
 #define MBOX_HEXDUMP_MAX_LEN	(MBOX_HEXDUMP_LINE_LEN *		\
 				 (MBOX_MAX_MSG_LEN / MBOX_BYTES_PER_LINE))
 

@@ -30,7 +30,7 @@ static struct dentry *root_debugfs_dir;
 
 struct mbox_test_device {
 	struct device		*dev;
-	void __iomem		*mmio;
+	void			*mmio;
 	struct mbox_chan	*tx_channel;
 	struct mbox_chan	*rx_channel;
 	char			*rx_buffer;

@@ -53,7 +53,7 @@ static ssize_t mbox_test_signal_write(struct file *filp,
 
 	if (count > MBOX_MAX_SIG_LEN) {
 		dev_err(tdev->dev,
-			"Signal length %d greater than max allowed %d\n",
+			"Signal length %lu greater than max allowed %d\n",
 			count, MBOX_MAX_SIG_LEN);
 		return -EINVAL;
 	}

@@ -92,7 +92,7 @@ static ssize_t mbox_test_message_write(struct file *filp,
 
 	if (count > MBOX_MAX_MSG_LEN) {
 		dev_err(tdev->dev,
-			"Message length %d greater than max allowed %d\n",
+			"Message length %lu greater than max allowed %d\n",
 			count, MBOX_MAX_MSG_LEN);
 		return -EINVAL;
 	}

@@ -303,7 +303,7 @@ static int mbox_test_probe(struct platform_device *pdev)
 	tdev->tx_channel = mbox_test_request_channel(pdev, "tx");
 	tdev->rx_channel = mbox_test_request_channel(pdev, "rx");
 
-	if (!tdev->tx_channel && !tdev->tx_channel)
+	if (!tdev->tx_channel && !tdev->rx_channel)
 		return -EPROBE_DEFER;
 
 	tdev->dev = &pdev->dev;
author
#7 lore raw

PATCH v2 1/1 mailbox: Fix a couple of trivial static checker issues

mailbox: Fix a couple of trivial static checker issues

This patch deals with a few spelling, white space and type warnings reported by Intel's Kbuild Test Robot.

Reported-by kbuild test robot <[email protected]>
Signed-off-by Lee Jones <[email protected]>

---
v1 => v2:

Looks like I took the AKA in this too literally:

format '%d' expects argument of type 'int',
but argument 3 has type 'size_t {aka long unsigned int}'

After some research I found a new printk format specifier suitable for size_t.

 drivers/mailbox/mailbox-sti.c  |  2 +-
 drivers/mailbox/mailbox-test.c | 10 +++++-----
 2 files changed, 6 insertions(+), 6 deletions(-)
drivers/mailbox/mailbox-sti.c +1 -1
index f11d4e8..4835817 100644
@@ -64,7 +64,7 @@ struct sti_mbox_device {
 };
 
 /**
- * STi Mailbox platform specfic configuration
+ * STi Mailbox platform specific configuration
  *
  * @num_inst:	Maximum number of instances in one HW Mailbox
  * @num_chan:	Maximum number of channel per instance
drivers/mailbox/mailbox-test.c +5 -5
index cac1ba2..22f2e40 100644
@@ -22,7 +22,7 @@
 #define MBOX_MAX_SIG_LEN	8
 #define MBOX_MAX_MSG_LEN	128
 #define MBOX_BYTES_PER_LINE	16
-#define MBOX_HEXDUMP_LINE_LEN 	((MBOX_BYTES_PER_LINE * 4) + 2)
+#define MBOX_HEXDUMP_LINE_LEN	((MBOX_BYTES_PER_LINE * 4) + 2)
 #define MBOX_HEXDUMP_MAX_LEN	(MBOX_HEXDUMP_LINE_LEN *		\
 				 (MBOX_MAX_MSG_LEN / MBOX_BYTES_PER_LINE))
 

@@ -30,7 +30,7 @@ static struct dentry *root_debugfs_dir;
 
 struct mbox_test_device {
 	struct device		*dev;
-	void __iomem		*mmio;
+	void			*mmio;
 	struct mbox_chan	*tx_channel;
 	struct mbox_chan	*rx_channel;
 	char			*rx_buffer;

@@ -53,7 +53,7 @@ static ssize_t mbox_test_signal_write(struct file *filp,
 
 	if (count > MBOX_MAX_SIG_LEN) {
 		dev_err(tdev->dev,
-			"Signal length %d greater than max allowed %d\n",
+			"Signal length %zd greater than max allowed %d\n",
 			count, MBOX_MAX_SIG_LEN);
 		return -EINVAL;
 	}

@@ -92,7 +92,7 @@ static ssize_t mbox_test_message_write(struct file *filp,
 
 	if (count > MBOX_MAX_MSG_LEN) {
 		dev_err(tdev->dev,
-			"Message length %d greater than max allowed %d\n",
+			"Message length %zd greater than max allowed %d\n",
 			count, MBOX_MAX_MSG_LEN);
 		return -EINVAL;
 	}

@@ -303,7 +303,7 @@ static int mbox_test_probe(struct platform_device *pdev)
 	tdev->tx_channel = mbox_test_request_channel(pdev, "tx");
 	tdev->rx_channel = mbox_test_request_channel(pdev, "rx");
 
-	if (!tdev->tx_channel && !tdev->tx_channel)
+	if (!tdev->tx_channel && !tdev->rx_channel)
 		return -EPROBE_DEFER;
 
 	tdev->dev = &pdev->dev;
author
#8 lore raw

PATCH v4 5/5 ARM: STi: DT: STiH407: Enable Mailbox testing facility

In reply to Lee Jones

This patch supplies a Client node to enable the Mailbox testing facility. It will be used to send and receive messages from any given co-processor in order to test the STi Mailbox Controller driver.

Signed-off-by Lee Jones <[email protected]>
---
 arch/arm/boot/dts/stih407-family.dtsi | 6 ++++++
 1 file changed, 6 insertions(+)
arch/arm/boot/dts/stih407-family.dtsi +6 -0
index 0e5d3ea..e6c8243 100644
@@ -611,6 +611,12 @@
 			st,pwm-num-chan = <4>;
 		};
 
+		mailbox_test {
+			compatible = "mailbox_test";
+			mboxes = <&mailbox2 1 31>, <&mailbox0 1 31>;
+			mbox-names = "tx", "rx";
+		};
+
 		mailbox0: mailbox at 0  {
 			compatible	= "st,stih407-mailbox";
 			reg		= <0x8f00000 0x1000>;
Signature · 1 line
1.9.1
#9 lore raw
In reply to Lee Jones
Quoted 42 lines · Lee Jones

Hi Jassi,

[Resending the updated patch-set this time]

This should be it. Exciting times!

ST's platforms currently support a maximum of 5 Mailboxes, one for each of the supported co-processors situated on the platform. Each Mailbox is divided up into 4 instances which consist of 32 channels. Messages are passed between the application and co-processors using shared memory areas.

Also included in the set is an example Client which should be generic enough to use as a testing environment for all Mailbox IPs which use shared memory or the API to pass messages. With small changes, it should also be able to test more extravagant implementations.

v3 -> v4:
- More protection in MBOX_BASE() MACRO
- Remove unused tx-only marker

v2 => v3:
- Controller
- Shared (DT <=> Device Driver) defines removed
- Return actual error codes instead of NULL in xlate()
- Not passing 'direction' configuration via DT anymore
- Removed all accompanying configuration checking code
- Test F/W
- Allow Rx-only and Tx-only controllers to be tested
- Cater for controllers requiring pre-Tx 'signal' to be sent
- Supply 'wr-' message debugfs file
- When written to, will Tx message through Mailbox Framework
- When read from, will print out a hexdump of Rx'ed message

v1 => v2:
- New MACRO() to obtain base address for a given instance
- Move locking into the structure it protects
- Stop checking for 'ready' state when sending data
- Don't clear channel data (that belongs to the API)
- #define register offsets instead of providing via pdata
- Register driver with module_platform_driver()

Quoted 7 lines

Lee Jones (5):
mailbox: dt: Supply bindings for ST's Mailbox IP
mailbox: Add support for ST's Mailbox IP
ARM: STi: stih407-family: Add nodes for Mailbox
mailbox: Add generic mechanism for testing Mailbox Controllers
ARM: STi: DT: STiH407: Enable Mailbox testing facility

Picked patches 1,2 & 4. Thanks for overhauling the old drivers, now we have fewer lines and simpler but more complete driver.

Thanks.