I2C IRQ Probe Improvements

PATCH Superseded Charles Keepax 4 participants 19 messages Jun 11, 2019 – Jun 13, 2019

Subsystems (4)

ACPI I2C subsystem host drivers I2C I2C ACPI support

lore

Share & tools

lore

Charles Keepax

author

2019-06-11 12:30 UTC #0 lore raw

v4 cover I2C IRQ Probe Improvements

This series attempts to align as much IRQ handling into the probe path as possible. Note that I don't have a great setup for testing these patches so they are mostly just build tested and need careful review and testing before any of them are merged.

The series brings the ACPI path inline with the way the device tree path handles the IRQ entirely at probe time. However, it still leaves any IRQ specified through the board_info as being handled at device time. In that case we need to cache something from the board_info until probe time, which leaves any alternative solution with something basically the same as the current handling although perhaps caching more stuff.

Thanks,
Charles

See previous discussions:
- https://lkml.org/lkml/2019/2/15/989
- https://www.spinics.net/lists/linux-i2c/msg39541.html

Charles Keepax (7):
i2c: core: Allow whole core to use i2c_dev_irq_from_resources
i2c: acpi: Use available IRQ helper functions
i2c: acpi: Factor out getting the IRQ from ACPI
i2c: core: Make i2c_acpi_get_irq available to the rest of the I2C core
i2c: core: Move ACPI IRQ handling to probe time
i2c: core: Move ACPI gpio IRQ handling into i2c_acpi_get_irq
i2c: core: Tidy up handling of init_irq

 drivers/i2c/i2c-core-acpi.c | 58 ++++++++++++++++++++++++++++++++-------------
 drivers/i2c/i2c-core-base.c | 11 +++++----
 drivers/i2c/i2c-core.h      |  9 +++++++
 3 files changed, 56 insertions(+), 22 deletions(-)

Signature · 1 line

2.11.0

Charles Keepax

author

2019-06-11 12:30 UTC #1 lore raw

PATCH v4 1/7 i2c: core: Allow whole core to use i2c_dev_irq_from_resources

Remove the static from i2c_dev_irq_from _resources so that other parts of the core code can use this helper function.

Reviewed-by Mika Westerberg <[email protected]>

Signed-off-by Charles Keepax <[email protected]>

---

No changes since v3.

Thanks,
Charles

 drivers/i2c/i2c-core-base.c | 4 ++--
 drivers/i2c/i2c-core.h      | 2 ++
 2 files changed, 4 insertions(+), 2 deletions(-)

drivers/i2c/i2c-core-base.c +2 -2

index 9e43508d4567d..67344586de913 100644
@@ -679,8 +679,8 @@ static void i2c_dev_set_name(struct i2c_adapter *adap,
 		     i2c_encode_flags_to_addr(client));
 }
 
-static int i2c_dev_irq_from_resources(const struct resource *resources,
-				      unsigned int num_resources)
+int i2c_dev_irq_from_resources(const struct resource *resources,
+			       unsigned int num_resources)
 {
 	struct irq_data *irqd;
 	int i;

drivers/i2c/i2c-core.h +2 -0

index 851c11b4c0f3a..2a3b28bf826b1 100644
@@ -19,6 +19,8 @@ extern struct list_head	__i2c_board_list;
 extern int		__i2c_first_dynamic_bus_num;
 
 int i2c_check_7bit_addr_validity_strict(unsigned short addr);
+int i2c_dev_irq_from_resources(const struct resource *resources,
+			       unsigned int num_resources);
 
 /*
  * We only allow atomic transfers for very late communication, e.g. to send

Signature · 1 line

2.11.0

Charles Keepax

author

2019-06-11 12:30 UTC #2 lore raw

PATCH v4 2/7 i2c: acpi: Use available IRQ helper functions

In reply to Charles Keepax

Use the available IRQ helper functions, most of the functions have additional helpful side affects like configuring the trigger type of the IRQ.

Reviewed-by Mika Westerberg <[email protected]>

Signed-off-by Charles Keepax <[email protected]>

---

No changes since v3.

Apologies Andy but I really can't see a sensible way to split this patch up. I have tried moving the resource_list_for_each_loop into a function in one patch then switching to using acpi_dev_resource_interrupt and i2c_dev_irq_from_resource, but the function is basically a completely different function ie. all the arguments are different and the purpose is different. So it makes no sense, and it is very ping pong. I tried doing the switch to i2c_dev_irq_from_resources as a separate patch which isn't so bad, but feels pointless as it is such a small change, let me know if you really want me to send that version.

Basically the switch to the helper function and the use of acpi_dev_resource_interrupt are intrinsically linked, so there isn't really a sensible way to split those up. If you still want some changes I am afraid you will have to be more specific in what you are looking for.

Thanks,
Charles

 drivers/i2c/i2c-core-acpi.c | 24 ++++++++++++++++--------
 1 file changed, 16 insertions(+), 8 deletions(-)

drivers/i2c/i2c-core-acpi.c +16 -8

index f1d648962b223..47d5b1c5ec9e0 100644
@@ -133,14 +133,25 @@ static int i2c_acpi_do_lookup(struct acpi_device *adev,
 	return 0;
 }
 
+static int i2c_acpi_add_resource(struct acpi_resource *ares, void *data)
+{
+	int *irq = data;
+	struct resource r;
+
+	if (*irq <= 0 && acpi_dev_resource_interrupt(ares, 0, &r))
+		*irq = i2c_dev_irq_from_resources(&r, 1);
+
+	return 1; /* No need to add resource to the list */
+}
+
 static int i2c_acpi_get_info(struct acpi_device *adev,
 			     struct i2c_board_info *info,
 			     struct i2c_adapter *adapter,
 			     acpi_handle *adapter_handle)
 {
 	struct list_head resource_list;
-	struct resource_entry *entry;
 	struct i2c_acpi_lookup lookup;
+	int irq = -ENOENT;
 	int ret;
 
 	memset(&lookup, 0, sizeof(lookup));

@@ -172,16 +183,13 @@ static int i2c_acpi_get_info(struct acpi_device *adev,
 
 	/* Then fill IRQ number if any */
 	INIT_LIST_HEAD(&resource_list);
-	ret = acpi_dev_get_resources(adev, &resource_list, NULL, NULL);
+	ret = acpi_dev_get_resources(adev, &resource_list,
+				     i2c_acpi_add_resource, &irq);
 	if (ret < 0)
 		return -EINVAL;
 
-	resource_list_for_each_entry(entry, &resource_list) {
-		if (resource_type(entry->res) == IORESOURCE_IRQ) {
-			info->irq = entry->res->start;
-			break;
-		}
-	}
+	if (irq > 0)
+		info->irq = irq;
 
 	acpi_dev_free_resource_list(&resource_list);

Signature · 1 line

2.11.0

Andy Shevchenko

2019-06-11 12:47 UTC #3 lore raw

Re: [PATCH v4 2/7] i2c: acpi: Use available IRQ helper functions

Quoted 21 lines · Charles Keepax

Use the available IRQ helper functions, most of the functions have additional helpful side affects like configuring the trigger type of the IRQ.
Reviewed-by Mika Westerberg <[email protected]>
Signed-off-by Charles Keepax <[email protected]>
---
No changes since v3.
Apologies Andy but I really can't see a sensible way to split this patch up. I have tried moving the resource_list_for_each_loop into a function in one patch then switching to using acpi_dev_resource_interrupt and i2c_dev_irq_from_resource, but the function is basically a completely different function ie. all the arguments are different and the purpose is different. So it makes no sense, and it is very ping pong. I tried doing the switch to i2c_dev_irq_from_resources as a separate patch which isn't so bad, but feels pointless as it is such a small change, let me know if you really want me to send that version.

Quoted 5 lines

Basically the switch to the helper function and the use of acpi_dev_resource_interrupt are intrinsically linked, so there isn't really a sensible way to split those up. If you still want some changes I am afraid you will have to be more specific in what you are looking for.

It seems I missed few details, thus the current approach works for me, thanks!

Quoted 62 lines

Thanks,
Charles

 drivers/i2c/i2c-core-acpi.c | 24 ++++++++++++++++--------
 1 file changed, 16 insertions(+), 8 deletions(-)

drivers/i2c/i2c-core-acpi.c +16 -8

index f1d648962b223..47d5b1c5ec9e0 100644
@@ -133,14 +133,25 @@ static int i2c_acpi_do_lookup(struct acpi_device *adev,
 	return 0;
 }
 
+static int i2c_acpi_add_resource(struct acpi_resource *ares, void *data)
+{
+	int *irq = data;
+	struct resource r;
+
+	if (*irq <= 0 && acpi_dev_resource_interrupt(ares, 0, &r))
+		*irq = i2c_dev_irq_from_resources(&r, 1);
+
+	return 1; /* No need to add resource to the list */
+}
+
 static int i2c_acpi_get_info(struct acpi_device *adev,
 			     struct i2c_board_info *info,
 			     struct i2c_adapter *adapter,
 			     acpi_handle *adapter_handle)
 {
 	struct list_head resource_list;
-	struct resource_entry *entry;
 	struct i2c_acpi_lookup lookup;
+	int irq = -ENOENT;
 	int ret;
 
 	memset(&lookup, 0, sizeof(lookup));

@@ -172,16 +183,13 @@ static int i2c_acpi_get_info(struct acpi_device *adev,
 
 	/* Then fill IRQ number if any */
 	INIT_LIST_HEAD(&resource_list);
-	ret = acpi_dev_get_resources(adev, &resource_list, NULL, NULL);
+	ret = acpi_dev_get_resources(adev, &resource_list,
+				     i2c_acpi_add_resource, &irq);
 	if (ret < 0)
 		return -EINVAL;
 
-	resource_list_for_each_entry(entry, &resource_list) {
-		if (resource_type(entry->res) == IORESOURCE_IRQ) {
-			info->irq = entry->res->start;
-			break;
-		}
-	}
+	if (irq > 0)
+		info->irq = irq;
 
 	acpi_dev_free_resource_list(&resource_list);

Signature · 1 line

2.11.0

Signature · 2 lines

With Best Regards,
Andy Shevchenko

Charles Keepax

author

2019-06-11 12:30 UTC #4 lore raw

PATCH v4 3/7 i2c: acpi: Factor out getting the IRQ from ACPI

In reply to Charles Keepax

In preparation for future refactoring factor out the fetch of the IRQ into its own helper function.

Reviewed-by Mika Westerberg <[email protected]>

Signed-off-by Charles Keepax <[email protected]>

---

No changes since v3.

Thanks,
Charles

 drivers/i2c/i2c-core-acpi.c | 33 +++++++++++++++++++++------------
 1 file changed, 21 insertions(+), 12 deletions(-)

drivers/i2c/i2c-core-acpi.c +21 ~~-12~~

index 47d5b1c5ec9e0..7d4d66ba752d4 100644
@@ -144,14 +144,30 @@ static int i2c_acpi_add_resource(struct acpi_resource *ares, void *data)
 	return 1; /* No need to add resource to the list */
 }
 
+static int i2c_acpi_get_irq(struct acpi_device *adev)
+{
+	struct list_head resource_list;
+	int irq = -ENOENT;
+	int ret;
+
+	INIT_LIST_HEAD(&resource_list);
+
+	ret = acpi_dev_get_resources(adev, &resource_list,
+				     i2c_acpi_add_resource, &irq);
+	if (ret < 0)
+		return -EINVAL;
+
+	acpi_dev_free_resource_list(&resource_list);
+
+	return irq;
+}
+
 static int i2c_acpi_get_info(struct acpi_device *adev,
 			     struct i2c_board_info *info,
 			     struct i2c_adapter *adapter,
 			     acpi_handle *adapter_handle)
 {
-	struct list_head resource_list;
 	struct i2c_acpi_lookup lookup;
-	int irq = -ENOENT;
 	int ret;
 
 	memset(&lookup, 0, sizeof(lookup));

@@ -182,16 +198,9 @@ static int i2c_acpi_get_info(struct acpi_device *adev,
 		*adapter_handle = lookup.adapter_handle;
 
 	/* Then fill IRQ number if any */
-	INIT_LIST_HEAD(&resource_list);
-	ret = acpi_dev_get_resources(adev, &resource_list,
-				     i2c_acpi_add_resource, &irq);
-	if (ret < 0)
-		return -EINVAL;
-
-	if (irq > 0)
-		info->irq = irq;
-
-	acpi_dev_free_resource_list(&resource_list);
+	ret = i2c_acpi_get_irq(adev);
+	if (ret > 0)
+		info->irq = ret;
 
 	acpi_set_modalias(adev, dev_name(&adev->dev), info->type,
 			  sizeof(info->type));

Signature · 1 line

2.11.0

Charles Keepax

author

2019-06-11 12:30 UTC #5 lore raw

PATCH v4 4/7 i2c: core: Make i2c_acpi_get_irq available to the rest of the I2C core

In reply to Charles Keepax

In preparation for more refactoring make i2c_acpi_get_irq available outside i2c-core-acpi.c.

Signed-off-by Charles Keepax <[email protected]>

---

Changes since v3:
- Move the change to use the helper function from i2c-core-base into its own patch.

Thanks,
Charles

 drivers/i2c/i2c-core-acpi.c | 15 +++++++++++++--
 drivers/i2c/i2c-core.h      |  7 +++++++
 2 files changed, 20 insertions(+), 2 deletions(-)

drivers/i2c/i2c-core-acpi.c +13 -2

index 7d4d66ba752d4..35966cc337dde 100644
@@ -144,8 +144,17 @@ static int i2c_acpi_add_resource(struct acpi_resource *ares, void *data)
 	return 1; /* No need to add resource to the list */
 }
 
-static int i2c_acpi_get_irq(struct acpi_device *adev)
+/**
+ * i2c_acpi_get_irq - get device IRQ number from ACPI
+ * @client: Pointer to the I2C client device
+ *
+ * Find the IRQ number used by a specific client device.
+ *
+ * Return: The IRQ number or an error code.
+ */
+int i2c_acpi_get_irq(struct i2c_client *client)
 {
+	struct acpi_device *adev = ACPI_COMPANION(&client->adapter->dev);
 	struct list_head resource_list;
 	int irq = -ENOENT;
 	int ret;

@@ -162,6 +171,8 @@ static int i2c_acpi_get_irq(struct acpi_device *adev)
 	return irq;
 }
 
+static struct i2c_client *i2c_acpi_find_client_by_adev(struct acpi_device *adev);
+
 static int i2c_acpi_get_info(struct acpi_device *adev,
 			     struct i2c_board_info *info,
 			     struct i2c_adapter *adapter,

@@ -198,7 +209,7 @@ static int i2c_acpi_get_info(struct acpi_device *adev,
 		*adapter_handle = lookup.adapter_handle;
 
 	/* Then fill IRQ number if any */
-	ret = i2c_acpi_get_irq(adev);
+	ret = i2c_acpi_get_irq(i2c_acpi_find_client_by_adev(adev));
 	if (ret > 0)
 		info->irq = ret;

drivers/i2c/i2c-core.h +7 -0

index 2a3b28bf826b1..517d98be68d25 100644
@@ -63,6 +63,8 @@ const struct acpi_device_id *
 i2c_acpi_match_device(const struct acpi_device_id *matches,
 		      struct i2c_client *client);
 void i2c_acpi_register_devices(struct i2c_adapter *adap);
+
+int i2c_acpi_get_irq(struct i2c_client *client);
 #else /* CONFIG_ACPI */
 static inline void i2c_acpi_register_devices(struct i2c_adapter *adap) { }
 static inline const struct acpi_device_id *

@@ -71,6 +73,11 @@ i2c_acpi_match_device(const struct acpi_device_id *matches,
 {
 	return NULL;
 }
+
+static inline int i2c_acpi_get_irq(struct i2c_client *client)
+{
+	return 0;
+}
 #endif /* CONFIG_ACPI */
 extern struct notifier_block i2c_acpi_notifier;

Signature · 1 line

2.11.0

Mika Westerberg

Reviewed-by on v4 4/7

2019-06-12 07:20 UTC #6 lore raw

Re: [PATCH v4 4/7] i2c: core: Make i2c_acpi_get_irq available to the rest of the I2C core

Quoted 4 lines · Charles Keepax

In preparation for more refactoring make i2c_acpi_get_irq available outside i2c-core-acpi.c.
Signed-off-by Charles Keepax <[email protected]>

Reviewed-by Mika Westerberg <[email protected]>

Benjamin Tissoires

2019-06-12 15:11 UTC #7 lore raw

Re: [PATCH v4 4/7] i2c: core: Make i2c_acpi_get_irq available to the rest of the I2C core

In reply to Charles Keepax

Quoted 6 lines · Charles Keepax

In preparation for more refactoring make i2c_acpi_get_irq available outside i2c-core-acpi.c.
Signed-off-by Charles Keepax <[email protected]>
---

So, my bisect fails here: when I recompile the vmlinuz image, it fails after loading the initrd. There is no boot up message, so not sure, but the problem with the non working touchpad is either this one or the next one.

Cheers,
Benjamin

Quoted 80 lines

Changes since v3:
- Move the change to use the helper function from i2c-core-base into its own patch.

Thanks,
Charles

 drivers/i2c/i2c-core-acpi.c | 15 +++++++++++++--
 drivers/i2c/i2c-core.h      |  7 +++++++
 2 files changed, 20 insertions(+), 2 deletions(-)

drivers/i2c/i2c-core-acpi.c +13 -2

index 7d4d66ba752d4..35966cc337dde 100644
@@ -144,8 +144,17 @@ static int i2c_acpi_add_resource(struct acpi_resource *ares, void *data)
        return 1; /* No need to add resource to the list */
 }
 
-static int i2c_acpi_get_irq(struct acpi_device *adev)
+/**
+ * i2c_acpi_get_irq - get device IRQ number from ACPI
+ * @client: Pointer to the I2C client device
+ *
+ * Find the IRQ number used by a specific client device.
+ *
+ * Return: The IRQ number or an error code.
+ */
+int i2c_acpi_get_irq(struct i2c_client *client)
 {
+       struct acpi_device *adev = ACPI_COMPANION(&client->adapter->dev);
        struct list_head resource_list;
        int irq = -ENOENT;
        int ret;

@@ -162,6 +171,8 @@ static int i2c_acpi_get_irq(struct acpi_device *adev)
        return irq;
 }
 
+static struct i2c_client *i2c_acpi_find_client_by_adev(struct acpi_device *adev);
+
 static int i2c_acpi_get_info(struct acpi_device *adev,
                             struct i2c_board_info *info,
                             struct i2c_adapter *adapter,

@@ -198,7 +209,7 @@ static int i2c_acpi_get_info(struct acpi_device *adev,
                *adapter_handle = lookup.adapter_handle;
 
        /* Then fill IRQ number if any */
-       ret = i2c_acpi_get_irq(adev);
+       ret = i2c_acpi_get_irq(i2c_acpi_find_client_by_adev(adev));
        if (ret > 0)
                info->irq = ret;

drivers/i2c/i2c-core.h +7 -1

index 2a3b28bf826b1..517d98be68d25 100644
@@ -63,6 +63,8 @@ const struct acpi_device_id *
 i2c_acpi_match_device(const struct acpi_device_id *matches,
                      struct i2c_client *client);
 void i2c_acpi_register_devices(struct i2c_adapter *adap);
+
+int i2c_acpi_get_irq(struct i2c_client *client);
 #else /* CONFIG_ACPI */
 static inline void i2c_acpi_register_devices(struct i2c_adapter *adap) { }
 static inline const struct acpi_device_id *

@@ -71,6 +73,11 @@ i2c_acpi_match_device(const struct acpi_device_id *matches,
 {
        return NULL;
 }
+
+static inline int i2c_acpi_get_irq(struct i2c_client *client)
+{
+       return 0;
+}
 #endif /* CONFIG_ACPI */
 extern struct notifier_block i2c_acpi_notifier;
 
--

2.11.0

Mika Westerberg

2019-06-12 15:27 UTC #8 lore raw

Re: [PATCH v4 4/7] i2c: core: Make i2c_acpi_get_irq available to the rest of the I2C core

In reply to Charles Keepax

Quoted 26 lines · Charles Keepax

In preparation for more refactoring make i2c_acpi_get_irq available outside i2c-core-acpi.c.
Signed-off-by Charles Keepax <[email protected]>
---
Changes since v3:
- Move the change to use the helper function from i2c-core-base into its own patch.
Thanks,
Charles
 drivers/i2c/i2c-core-acpi.c | 15 +++++++++++++--
 drivers/i2c/i2c-core.h      |  7 +++++++
 2 files changed, 20 insertions(+), 2 deletions(-)
drivers/i2c/i2c-core-acpi.c +1 -1
index 7d4d66ba752d4..35966cc337dde 100644
@@ -144,8 +144,17 @@ static int i2c_acpi_add_resource(struct acpi_resource *ares, void *data)
 	return 1; /* No need to add resource to the list */
 }
 
-static int i2c_acpi_get_irq(struct acpi_device *adev)
+/**

drivers/i2c/i2c-core-acpi.c static int i2c_acpi_add_resource(struct acpi_resource *ares, void *data) · line 157

Quoted 10 lines

drivers/i2c/i2c-core-acpi.c +9 -0

+ * i2c_acpi_get_irq - get device IRQ number from ACPI
+ * @client: Pointer to the I2C client device
+ *
+ * Find the IRQ number used by a specific client device.
+ *
+ * Return: The IRQ number or an error code.
+ */
+int i2c_acpi_get_irq(struct i2c_client *client)
 {
+	struct acpi_device *adev = ACPI_COMPANION(&client->adapter->dev);

Is this adev checked for being NULL somewhere below before it is being dereferenced?

It could explain the issue Benjamin is seeing.

Quoted 3 lines

drivers/i2c/i2c-core-acpi.c +0 -0

 	struct list_head resource_list;
 	int irq = -ENOENT;
 	int ret;

Charles Keepax

author

2019-06-13 08:48 UTC #9 lore raw

Re: [PATCH v4 4/7] i2c: core: Make i2c_acpi_get_irq available to the rest of the I2C core

Quoted 41 lines · Mika Westerberg

Quoted 26 lines · Charles Keepax
In preparation for more refactoring make i2c_acpi_get_irq available outside i2c-core-acpi.c.
Signed-off-by Charles Keepax <[email protected]>
---
Changes since v3:
- Move the change to use the helper function from i2c-core-base into its own patch.
Thanks,
Charles
 drivers/i2c/i2c-core-acpi.c | 15 +++++++++++++--
 drivers/i2c/i2c-core.h      |  7 +++++++
 2 files changed, 20 insertions(+), 2 deletions(-)
drivers/i2c/i2c-core-acpi.c +1 -1
index 7d4d66ba752d4..35966cc337dde 100644
@@ -144,8 +144,17 @@ static int i2c_acpi_add_resource(struct acpi_resource *ares, void *data)
 	return 1; /* No need to add resource to the list */
 }
 
-static int i2c_acpi_get_irq(struct acpi_device *adev)
+/**
drivers/i2c/i2c-core-acpi.c static int i2c_acpi_add_resource(struct acpi_resource *ares, void *data) · line 157
Quoted 10 lines
drivers/i2c/i2c-core-acpi.c +9 -0
+ * i2c_acpi_get_irq - get device IRQ number from ACPI
+ * @client: Pointer to the I2C client device
+ *
+ * Find the IRQ number used by a specific client device.
+ *
+ * Return: The IRQ number or an error code.
+ */
+int i2c_acpi_get_irq(struct i2c_client *client)
 {
+	struct acpi_device *adev = ACPI_COMPANION(&client->adapter->dev);
Is this adev checked for being NULL somewhere below before it is being dereferenced?

Quoted 2 lines

It could explain the issue Benjamin is seeing.

Yeah could be that or just for some reason this isn't returning the same adev as we previously had. I will do some digging see if I can find any likely culprits.

Thanks,
Charles

Benjamin Tissoires

2019-06-13 09:32 UTC #10 lore raw

Re: [PATCH v4 4/7] i2c: core: Make i2c_acpi_get_irq available to the rest of the I2C core

Quoted 43 lines · Charles Keepax

Quoted 41 lines · Mika Westerberg
Quoted 26 lines · Charles Keepax
In preparation for more refactoring make i2c_acpi_get_irq available outside i2c-core-acpi.c.
Signed-off-by Charles Keepax <[email protected]>
---
Changes since v3:
- Move the change to use the helper function from i2c-core-base into its own patch.
Thanks,
Charles
 drivers/i2c/i2c-core-acpi.c | 15 +++++++++++++--
 drivers/i2c/i2c-core.h      |  7 +++++++
 2 files changed, 20 insertions(+), 2 deletions(-)
drivers/i2c/i2c-core-acpi.c +1 -1
index 7d4d66ba752d4..35966cc337dde 100644
@@ -144,8 +144,17 @@ static int i2c_acpi_add_resource(struct acpi_resource *ares, void *data)
    return 1; /* No need to add resource to the list */
 }
 
-static int i2c_acpi_get_irq(struct acpi_device *adev)
+/**
drivers/i2c/i2c-core-acpi.c static int i2c_acpi_add_resource(struct acpi_resource *ares, void *data) · line 157
Quoted 10 lines
drivers/i2c/i2c-core-acpi.c +9 -0
+ * i2c_acpi_get_irq - get device IRQ number from ACPI
+ * @client: Pointer to the I2C client device
+ *
+ * Find the IRQ number used by a specific client device.
+ *
+ * Return: The IRQ number or an error code.
+ */
+int i2c_acpi_get_irq(struct i2c_client *client)
 {
+   struct acpi_device *adev = ACPI_COMPANION(&client->adapter->dev);
Is this adev checked for being NULL somewhere below before it is being dereferenced?

Quoted 6 lines

Quoted 2 lines
It could explain the issue Benjamin is seeing.
Yeah could be that or just for some reason this isn't returning the same adev as we previously had. I will do some digging see if I can find any likely culprits.

That was almost the culprit: client is NULL here. So the call of i2c_acpi_find_client_by_adev(adev) fails.

I guess this explains why the next commit is also not working :)

Cheers,
Benjamin

Charles Keepax

author

2019-06-13 15:06 UTC #11 lore raw

Re: [PATCH v4 4/7] i2c: core: Make i2c_acpi_get_irq available to the rest of the I2C core

Quoted 48 lines · Benjamin Tissoires

Quoted 46 lines · Charles Keepax
Quoted 43 lines · Mika Westerberg
Quoted 26 lines · Charles Keepax
In preparation for more refactoring make i2c_acpi_get_irq available outside i2c-core-acpi.c.
Signed-off-by Charles Keepax <[email protected]>
---
Changes since v3:
- Move the change to use the helper function from i2c-core-base into its own patch.
Thanks,
Charles
 drivers/i2c/i2c-core-acpi.c | 15 +++++++++++++--
 drivers/i2c/i2c-core.h      |  7 +++++++
 2 files changed, 20 insertions(+), 2 deletions(-)
drivers/i2c/i2c-core-acpi.c +1 -1
index 7d4d66ba752d4..35966cc337dde 100644
@@ -144,8 +144,17 @@ static int i2c_acpi_add_resource(struct acpi_resource *ares, void *data)
    return 1; /* No need to add resource to the list */
 }
 
-static int i2c_acpi_get_irq(struct acpi_device *adev)
+/**
drivers/i2c/i2c-core-acpi.c static int i2c_acpi_add_resource(struct acpi_resource *ares, void *data) · line 157
Quoted 10 lines
drivers/i2c/i2c-core-acpi.c +9 -0
+ * i2c_acpi_get_irq - get device IRQ number from ACPI
+ * @client: Pointer to the I2C client device
+ *
+ * Find the IRQ number used by a specific client device.
+ *
+ * Return: The IRQ number or an error code.
+ */
+int i2c_acpi_get_irq(struct i2c_client *client)
 {
+   struct acpi_device *adev = ACPI_COMPANION(&client->adapter->dev);
Is this adev checked for being NULL somewhere below before it is being dereferenced?
It could explain the issue Benjamin is seeing.

Quoted 9 lines

Quoted 3 lines
Yeah could be that or just for some reason this isn't returning the same adev as we previously had. I will do some digging see if I can find any likely culprits.
That was almost the culprit: client is NULL here. So the call of i2c_acpi_find_client_by_adev(adev) fails.
I guess this explains why the next commit is also not working :)

Ah.. yeah because at this stage this code is probably running before that client has been created I think, this crept in on that last splitting of the patch into two. Let me see if I can come up with a new solution and do a new version. Thank you very much for the help going through this.

Thanks,
Charles

Charles Keepax

author

2019-06-11 12:30 UTC #12 lore raw

PATCH v4 5/7 i2c: core: Move ACPI IRQ handling to probe time

In reply to Charles Keepax

Bring the ACPI path in sync with the device tree path and handle all the IRQ fetching at probe time. This leaves the only IRQ handling at device registration time being that which is passed directly through the board info as either a resource or an actual IRQ number.

Reviewed-by Mika Westerberg <[email protected]>

Signed-off-by Charles Keepax <[email protected]>

---

New since v3, but this patch and the previous one leave the code in the same state as this patch did on v3.

Thanks,
Charles

 drivers/i2c/i2c-core-acpi.c | 7 -------
 drivers/i2c/i2c-core-base.c | 5 ++++-
 2 files changed, 4 insertions(+), 8 deletions(-)

drivers/i2c/i2c-core-acpi.c +0 -7

index 35966cc337dde..c70228f5349ef 100644
@@ -171,8 +171,6 @@ int i2c_acpi_get_irq(struct i2c_client *client)
 	return irq;
 }
 
-static struct i2c_client *i2c_acpi_find_client_by_adev(struct acpi_device *adev);
-
 static int i2c_acpi_get_info(struct acpi_device *adev,
 			     struct i2c_board_info *info,
 			     struct i2c_adapter *adapter,

@@ -208,11 +206,6 @@ static int i2c_acpi_get_info(struct acpi_device *adev,
 	if (adapter_handle)
 		*adapter_handle = lookup.adapter_handle;
 
-	/* Then fill IRQ number if any */
-	ret = i2c_acpi_get_irq(i2c_acpi_find_client_by_adev(adev));
-	if (ret > 0)
-		info->irq = ret;
-
 	acpi_set_modalias(adev, dev_name(&adev->dev), info->type,
 			  sizeof(info->type));

drivers/i2c/i2c-core-base.c +4 -1

index 67344586de913..9520bcfc83182 100644
@@ -327,7 +327,10 @@ static int i2c_device_probe(struct device *dev)
 			if (irq == -EINVAL || irq == -ENODATA)
 				irq = of_irq_get(dev->of_node, 0);
 		} else if (ACPI_COMPANION(dev)) {
-			irq = acpi_dev_gpio_irq_get(ACPI_COMPANION(dev), 0);
+			irq = i2c_acpi_get_irq(client);
+
+			if (irq == -ENOENT)
+				irq = acpi_dev_gpio_irq_get(ACPI_COMPANION(dev), 0);
 		}
 		if (irq == -EPROBE_DEFER)
 			return irq;

Signature · 1 line

2.11.0

Charles Keepax

author

2019-06-11 12:31 UTC #13 lore raw

PATCH v4 6/7 i2c: core: Move ACPI gpio IRQ handling into i2c_acpi_get_irq

In reply to Charles Keepax

It makes sense to contain all the ACPI IRQ handling in a single helper function.

Reviewed-by Mika Westerberg <[email protected]>

Signed-off-by Charles Keepax <[email protected]>

---

No change since v3.

Thanks,
Charles

 drivers/i2c/i2c-core-acpi.c | 3 +++
 drivers/i2c/i2c-core-base.c | 3 ---
 2 files changed, 3 insertions(+), 3 deletions(-)

drivers/i2c/i2c-core-acpi.c +3 -0

index c70228f5349ef..b3ccd2dda0fc1 100644
@@ -168,6 +168,9 @@ int i2c_acpi_get_irq(struct i2c_client *client)
 
 	acpi_dev_free_resource_list(&resource_list);
 
+	if (irq == -ENOENT)
+		irq = acpi_dev_gpio_irq_get(adev, 0);
+
 	return irq;
 }

drivers/i2c/i2c-core-base.c +0 -3

index 9520bcfc83182..6be2e3a4784fa 100644
@@ -328,9 +328,6 @@ static int i2c_device_probe(struct device *dev)
 				irq = of_irq_get(dev->of_node, 0);
 		} else if (ACPI_COMPANION(dev)) {
 			irq = i2c_acpi_get_irq(client);
-
-			if (irq == -ENOENT)
-				irq = acpi_dev_gpio_irq_get(ACPI_COMPANION(dev), 0);
 		}
 		if (irq == -EPROBE_DEFER)
 			return irq;

Signature · 1 line

2.11.0

Charles Keepax

author

2019-06-11 12:31 UTC #14 lore raw

PATCH v4 7/7 i2c: core: Tidy up handling of init_irq

In reply to Charles Keepax

Only set init_irq during i2c_device_new and only handle client->irq on the probe/remove paths.

Reviewed-by Mika Westerberg <[email protected]>

Signed-off-by Charles Keepax <[email protected]>

---

No change since v3.

Thanks,
Charles

 drivers/i2c/i2c-core-base.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

drivers/i2c/i2c-core-base.c +3 -2

index 6be2e3a4784fa..9b52427dc7c94 100644
@@ -314,6 +314,8 @@ static int i2c_device_probe(struct device *dev)
 
 	driver = to_i2c_driver(dev->driver);
 
+	client->irq = client->init_irq;
+
 	if (!client->irq && !driver->disable_i2c_core_irq_mapping) {
 		int irq = -ENOENT;

@@ -424,7 +426,7 @@ static int i2c_device_remove(struct device *dev)
 	dev_pm_clear_wake_irq(&client->dev);
 	device_init_wakeup(&client->dev, false);
 
-	client->irq = client->init_irq;
+	client->irq = 0;
 	if (client->flags & I2C_CLIENT_HOST_NOTIFY)
 		pm_runtime_put(&client->adapter->dev);

@@ -741,7 +743,6 @@ i2c_new_client_device(struct i2c_adapter *adap, struct i2c_board_info const *inf
 	if (!client->init_irq)
 		client->init_irq = i2c_dev_irq_from_resources(info->resources,
 							 info->num_resources);
-	client->irq = client->init_irq;
 
 	strlcpy(client->name, info->type, sizeof(client->name));

Signature · 1 line

2.11.0

Andy Shevchenko

Reviewed-by on the v4 series

2019-06-11 12:51 UTC #15 lore raw

In reply to Charles Keepax

Quoted 6 lines · Charles Keepax

This series attempts to align as much IRQ handling into the probe path as possible. Note that I don't have a great setup for testing these patches so they are mostly just build tested and need careful review and testing before any of them are merged.

Quoted 7 lines

The series brings the ACPI path inline with the way the device tree path handles the IRQ entirely at probe time. However, it still leaves any IRQ specified through the board_info as being handled at device time. In that case we need to cache something from the board_info until probe time, which leaves any alternative solution with something basically the same as the current handling although perhaps caching more stuff.

Thank you!
This one looks good.

Reviewed-by Andy Shevchenko <[email protected]>

Quoted 25 lines

Thanks,
Charles
See previous discussions:
- https://lkml.org/lkml/2019/2/15/989
- https://www.spinics.net/lists/linux-i2c/msg39541.html
Charles Keepax (7):
i2c: core: Allow whole core to use i2c_dev_irq_from_resources
i2c: acpi: Use available IRQ helper functions
i2c: acpi: Factor out getting the IRQ from ACPI
i2c: core: Make i2c_acpi_get_irq available to the rest of the I2C core
i2c: core: Move ACPI IRQ handling to probe time
i2c: core: Move ACPI gpio IRQ handling into i2c_acpi_get_irq
i2c: core: Tidy up handling of init_irq
 drivers/i2c/i2c-core-acpi.c | 58 ++++++++++++++++++++++++++++++++-------------
 drivers/i2c/i2c-core-base.c | 11 +++++----
 drivers/i2c/i2c-core.h      |  9 +++++++
 3 files changed, 56 insertions(+), 22 deletions(-)
Signature · 1 line
2.11.0

Signature · 2 lines

With Best Regards,
Andy Shevchenko

Benjamin Tissoires

2019-06-11 15:16 UTC #16 lore raw

In reply to Charles Keepax

Quoted 7 lines · Charles Keepax

This series attempts to align as much IRQ handling into the probe path as possible. Note that I don't have a great setup for testing these patches so they are mostly just build tested and need careful review and testing before any of them are merged.

Quoted 7 lines

The series brings the ACPI path inline with the way the device tree path handles the IRQ entirely at probe time. However, it still leaves any IRQ specified through the board_info as being handled at device time. In that case we need to cache something from the board_info until probe time, which leaves any alternative solution with something basically the same as the current handling although perhaps caching more stuff.

Hmm, I still haven't pinpointed the issue, but I wanted to give a test of the series and I have:

[    5.511806] i2c_hid i2c-DLL075B:01: HID over i2c has not been
provided an Int IRQ
[    5.511825] i2c_hid: probe of i2c-DLL075B:01 failed with error -22

So it seems that there is something wrong happening when fetching the IRQ and providing it to i2c-hid.

That was on a Dell XPS 9360.

Bisecting is starting.

Cheers,
Benjamin

Quoted 25 lines

Thanks,
Charles
See previous discussions:
- https://lkml.org/lkml/2019/2/15/989
- https://www.spinics.net/lists/linux-i2c/msg39541.html
Charles Keepax (7):
i2c: core: Allow whole core to use i2c_dev_irq_from_resources
i2c: acpi: Use available IRQ helper functions
i2c: acpi: Factor out getting the IRQ from ACPI
i2c: core: Make i2c_acpi_get_irq available to the rest of the I2C core
i2c: core: Move ACPI IRQ handling to probe time
i2c: core: Move ACPI gpio IRQ handling into i2c_acpi_get_irq
i2c: core: Tidy up handling of init_irq
 drivers/i2c/i2c-core-acpi.c | 58 ++++++++++++++++++++++++++++++++-------------
 drivers/i2c/i2c-core-base.c | 11 +++++----
 drivers/i2c/i2c-core.h      |  9 +++++++
 3 files changed, 56 insertions(+), 22 deletions(-)
--
2.11.0

Charles Keepax

author

2019-06-11 15:28 UTC #17 lore raw

Quoted 23 lines · Benjamin Tissoires

Quoted 7 lines · Charles Keepax
This series attempts to align as much IRQ handling into the probe path as possible. Note that I don't have a great setup for testing these patches so they are mostly just build tested and need careful review and testing before any of them are merged.
Quoted 7 lines
The series brings the ACPI path inline with the way the device tree path handles the IRQ entirely at probe time. However, it still leaves any IRQ specified through the board_info as being handled at device time. In that case we need to cache something from the board_info until probe time, which leaves any alternative solution with something basically the same as the current handling although perhaps caching more stuff.
Hmm, I still haven't pinpointed the issue, but I wanted to give a test of the series and I have:
[    5.511806] i2c_hid i2c-DLL075B:01: HID over i2c has not been
provided an Int IRQ
[    5.511825] i2c_hid: probe of i2c-DLL075B:01 failed with error -22

Quoted 7 lines

So it seems that there is something wrong happening when fetching the IRQ and providing it to i2c-hid.
That was on a Dell XPS 9360.
Bisecting is starting.

I have a sneaking suspision, does this diff fix it:

(quoted diff) +0 -0

b/drivers/i2c/i2c-core-acpi.c
index 57be6342ba508..a90b05a269c36 100644

drivers/i2c/i2c-core-acpi.c +1 -1

@@ -169,7 +169,7 @@ int i2c_acpi_get_irq(struct i2c_client *client)
        acpi_dev_free_resource_list(&resource_list);
 
        if (irq == -ENOENT)
-           irq = acpi_dev_gpio_irq_get(adev, 0);
+         irq = acpi_dev_gpio_irq_get(ACPI_COMPANION(&client->dev), 0);
 
        return irq;
 }

There was some earlier discussion about which device was suitable for this call.

Thanks,
Charles

Benjamin Tissoires

2019-06-12 15:13 UTC #18 lore raw

Quoted 42 lines · Charles Keepax

Quoted 23 lines · Benjamin Tissoires
Quoted 7 lines · Charles Keepax
This series attempts to align as much IRQ handling into the probe path as possible. Note that I don't have a great setup for testing these patches so they are mostly just build tested and need careful review and testing before any of them are merged.
Quoted 7 lines
The series brings the ACPI path inline with the way the device tree path handles the IRQ entirely at probe time. However, it still leaves any IRQ specified through the board_info as being handled at device time. In that case we need to cache something from the board_info until probe time, which leaves any alternative solution with something basically the same as the current handling although perhaps caching more stuff.
Hmm, I still haven't pinpointed the issue, but I wanted to give a test of the series and I have:
[    5.511806] i2c_hid i2c-DLL075B:01: HID over i2c has not been
provided an Int IRQ
[    5.511825] i2c_hid: probe of i2c-DLL075B:01 failed with error -22
Quoted 7 lines
So it seems that there is something wrong happening when fetching the IRQ and providing it to i2c-hid.
That was on a Dell XPS 9360.
Bisecting is starting.
I have a sneaking suspision, does this diff fix it:
drivers/i2c/i2c-core-acpi.c +0 -0
index 57be6342ba508..a90b05a269c36 100644

drivers/i2c/i2c-core-acpi.c int i2c_acpi_get_irq(struct i2c_client *client) · line 175

Quoted 8 lines

drivers/i2c/i2c-core-acpi.c +1 -1

@@ -169,7 +169,7 @@ int i2c_acpi_get_irq(struct i2c_client *client)
        acpi_dev_free_resource_list(&resource_list);
 
        if (irq == -ENOENT)
-           irq = acpi_dev_gpio_irq_get(adev, 0);
+         irq = acpi_dev_gpio_irq_get(ACPI_COMPANION(&client->dev), 0);
 
        return irq;
 }

Unfortunately, this doesn't solve the issue.

The problem is either in 4/7 or 5/7 (4/7 doesn't boot AFAICT).

(chasing multiple rabbits at the same time, so hard to get to the bottom of it)

Cheers,
Benjamin

Quoted 5 lines

There was some earlier discussion about which device was suitable for this call.
Thanks,
Charles