Skip to contentSkip to conversation

test: regmap: Increase size of syscon0 memory

PATCH Superseded Mario Six 3 participants 35 messages Aug 3, 2018 – Aug 8, 2018

Subsystems (2)

v4 of 11 Superseded Latest is v11 →

lore
Share & tools
lore

Mario Six

author
#0 lore raw

PATCH v4 1/13 test: regmap: Increase size of syscon0 memory

The upcoming changes to the regmap interface will contain a proper check for plausibility when reading/writing from/to a register map. To still have the current tests pass, increase the size of the memory region for the syscon0 device, since one of the tests reads and writes beyond this range.

---

v3 -> v4:
No changes

v2 -> v3:
New in v3

---
 arch/sandbox/dts/test.dts | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/sandbox/dts/test.dts +1 -1
index e941cea3e5c..47cc961890f 100644
@@ -406,7 +406,7 @@
 
 	syscon at 0 {
 		compatible = "sandbox,syscon0";
-		reg = <0x10 4>;
+		reg = <0x10 16>;
 	};
 
 	syscon at 1 {

--
2.11.0

author
#1 lore raw

PATCH v4 2/13 regmap: Fix documentation

The documentation in regmap.h is not in kernel-doc format. Correct this.

---

v3 -> v4:
No changes

v2 -> v3:
New in v3

---
 include/regmap.h | 48 +++++++++++++++++++++++++++++++++++++++---------
 1 file changed, 39 insertions(+), 9 deletions(-)
include/regmap.h +39 -9
index 6a574eaa412..32f75e06f59 100644
@@ -21,8 +21,8 @@ struct regmap_range {
 /**
  * struct regmap - a way of accessing hardware/bus registers
  *
- * @range_count: Number of ranges available within the map
- * @ranges:	Array of ranges
+ * @range_count:	Number of ranges available within the map
+ * @ranges:		Array of ranges
  */
 struct regmap {
 	int range_count;
@@ -33,7 +33,28 @@ struct regmap {
  * Interface to provide access to registers either through a direct memory
  * bus or through a peripheral bus like I2C, SPI.
  */
+
+/**
+ * regmap_write() - Write a 32-bit value to a regmap
+ *
+ * @map:	Regmap to write to
+ * @offset:	Offset in the regmap to write to
+ * @val:	Data to write to the regmap at the specified offset
+ *
+ * Return: 0 if OK, -ve on error
+ */
 int regmap_write(struct regmap *map, uint offset, uint val);
+
+/**
+ * regmap_read() - Read a 32-bit value from a regmap
+ *
+ * @map:	Regmap to read from
+ * @offset:	Offset in the regmap to read from
+ * @valp:	Pointer to the buffer to receive the data read from the regmap
+ *		at the specified offset
+ *
+ * Return: 0 if OK, -ve on error
+ */
 int regmap_read(struct regmap *map, uint offset, uint *valp);
 
 #define regmap_write32(map, ptr, member, val) \
@@ -49,31 +70,36 @@ int regmap_read(struct regmap *map, uint offset, uint *valp);
  * @offset:	Offset of the memory
  * @mask:	Mask to apply to the read value
  * @val:	Value to apply to the value to write
+ * Return: 0 if OK, -ve on error
  */
 int regmap_update_bits(struct regmap *map, uint offset, uint mask, uint val);
 
 /**
  * regmap_init_mem() - Set up a new register map that uses memory access
  *
- * Use regmap_uninit() to free it.
- *
  * @node:	Device node that uses this map
  * @mapp:	Returns allocated map
+ * Return: 0 if OK, -ve on error
+ *
+ * Use regmap_uninit() to free it.
  */
 int regmap_init_mem(ofnode node, struct regmap **mapp);
 
 /**
- * regmap_init_mem_platdata() - Set up a new memory register map for of-platdata
+ * regmap_init_mem_platdata() - Set up a new memory register map for
+ *				of-platdata
+ *
+ * @dev:	Device that uses this map
+ * @reg:	List of address, size pairs
+ * @count:	Number of pairs (e.g. 1 if the regmap has a single entry)
+ * @mapp:	Returns allocated map
+ * Return: 0 if OK, -ve on error
  *
  * This creates a new regmap with a list of regions passed in, rather than
  * using the device tree. It only supports 32-bit machines.
  *
  * Use regmap_uninit() to free it.
  *
- * @dev:	Device that uses this map
- * @reg:	List of address, size pairs
- * @count:	Number of pairs (e.g. 1 if the regmap has a single entry)
- * @mapp:	Returns allocated map
  */
 int regmap_init_mem_platdata(struct udevice *dev, fdt_val_t *reg, int count,
 			     struct regmap **mapp);
@@ -83,11 +109,15 @@ int regmap_init_mem_platdata(struct udevice *dev, fdt_val_t *reg, int count,
  *
  * @map:	Regmap to query
  * @range_num:	Range to look up
+ * Return: Pointer to the range in question if OK, NULL on error
  */
 void *regmap_get_range(struct regmap *map, unsigned int range_num);
 
 /**
  * regmap_uninit() - free a previously inited regmap
+ *
+ * @map:	Regmap to free
+ * Return: 0 if OK, -ve on error
  */
 int regmap_uninit(struct regmap *map);
 

--
2.11.0

#2 lore raw

PATCH v4 2/13 regmap: Fix documentation

Quoted 11 lines · On Fri, 3 Aug 2018 10:01:07 +0200 Mario Six mario.six at gdsys.cc wrote:

The documentation in regmap.h is not in kernel-doc format. Correct this.

---

v3 -> v4:
No changes

Quoted 6 lines

v2 -> v3:
New in v3

---
 include/regmap.h | 48 +++++++++++++++++++++++++++++++++++++++---------
 1 file changed, 39 insertions(+), 9 deletions(-)

Applied to u-boot-staging/agust at denx.de, thanks!

--
Anatolij

author
#3 lore raw

PATCH v4 3/13 regmap: Add documentation

In reply to Mario Six

Document the regmap_alloc() function.

---

v3 -> v4:
No changes

v2 -> v3:
New in v3

---
 drivers/core/regmap.c | 6 ++++++
 1 file changed, 6 insertions(+)
drivers/core/regmap.c +6 -0
index 8e5c3bcf61b..77f6f520a06 100644
@@ -17,6 +17,12 @@
 
 DECLARE_GLOBAL_DATA_PTR;
 
+/**
+ * regmap_alloc() - Allocate a regmap with a given number of ranges.
+ *
+ * @count: Number of ranges to be allocated for the regmap.
+ * Return: A pointer to the newly allocated regmap, or NULL on error.
+ */
 static struct regmap *regmap_alloc(int count)
 {
 	struct regmap *map;

--
2.11.0

#4 lore raw

PATCH v4 3/13 regmap: Add documentation

Quoted 11 lines · On Fri, 3 Aug 2018 10:01:08 +0200 Mario Six mario.six at gdsys.cc wrote:

Document the regmap_alloc() function.

---

v3 -> v4:
No changes

Quoted 6 lines

v2 -> v3:
New in v3

---
 drivers/core/regmap.c | 6 ++++++
 1 file changed, 6 insertions(+)

Applied to u-boot-staging/agust at denx.de, thanks!

--
Anatolij

author
#5 lore raw

PATCH v4 4/13 regmap: Improve error handling

In reply to Mario Six

ofnode_read_simple_addr_cells may fail and return a negative error code. Check for this when initializing regmaps.

Also check if both_len is zero, since this is perfectly possible, and would lead to a division-by-zero further down the line.

---

v3 -> v4:
No changes

v2 -> v3:
New in v3

---
 drivers/core/regmap.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)
drivers/core/regmap.c +17 -0
index 77f6f520a06..4ebab233490 100644
@@ -67,8 +67,25 @@ int regmap_init_mem(ofnode node, struct regmap **mapp)
 	struct resource r;
 
 	addr_len = ofnode_read_simple_addr_cells(ofnode_get_parent(node));
+	if (addr_len < 0) {
+		debug("%s: Error while reading the addr length (ret = %d)\n",
+		      ofnode_get_name(node), addr_len);
+		return addr_len;
+	}
+
 	size_len = ofnode_read_simple_size_cells(ofnode_get_parent(node));
+	if (size_len < 0) {
+		debug("%s: Error while reading the size length: (ret = %d)\n",
+		      ofnode_get_name(node), size_len);
+		return size_len;
+	}
+
 	both_len = addr_len + size_len;
+	if (!both_len) {
+		debug("%s: Both addr and size length are zero\n",
+		      ofnode_get_name(node));
+		return -EINVAL;
+	}
 
 	len = ofnode_read_size(node, "reg");
 	if (len < 0)

--
2.11.0

#6 lore raw

PATCH v4 4/13 regmap: Improve error handling

Quoted 15 lines · On Fri, 3 Aug 2018 10:01:09 +0200 Mario Six mario.six at gdsys.cc wrote:

ofnode_read_simple_addr_cells may fail and return a negative error code. Check for this when initializing regmaps.

Also check if both_len is zero, since this is perfectly possible, and would lead to a division-by-zero further down the line.

---

v3 -> v4:
No changes

Quoted 6 lines

v2 -> v3:
New in v3

---
 drivers/core/regmap.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

Applied to u-boot-staging/agust at denx.de, thanks!

--
Anatolij

author
#7 lore raw

PATCH v4 5/13 regmap: Introduce init_range

In reply to Mario Six

Both fdtdec_get_addr_size_fixed and of_address_to_resource can fail with an error, which is not currently checked during regmap initialization.

Since the indentation depth is already quite deep, extract a new 'init_range' method to do the initialization.

---

v3 -> v4:
* Introduced else clause in of_live_active() if statement to make the distinction between live and non-live cases clearer

v2 -> v3:
New in v3

---
 drivers/core/regmap.c | 68 ++++++++++++++++++++++++++++++++++++++++++---------
 1 file changed, 56 insertions(+), 12 deletions(-)
drivers/core/regmap.c +56 -12
index 4ebab233490..25c1ae5d7df 100644
@@ -56,6 +56,58 @@ int regmap_init_mem_platdata(struct udevice *dev, fdt_val_t *reg, int count,
 	return 0;
 }
 #else
+/**
+ * init_range() - Initialize a single range of a regmap
+ * @node:     Device node that will use the map in question
+ * @range:    Pointer to a regmap_range structure that will be initialized
+ * @addr_len: The length of the addr parts of the reg property
+ * @size_len: The length of the size parts of the reg property
+ * @index:    The index of the range to initialize
+ *
+ * This function will read the necessary 'reg' information from the device tree
+ * (the 'addr' part, and the 'length' part), and initialize the range in
+ * quesion.
+ *
+ * Return: 0 if OK, -ve on error
+ */
+static int init_range(ofnode node, struct regmap_range *range, int addr_len,
+		      int size_len, int index)
+{
+	fdt_size_t sz;
+	struct resource r;
+
+	if (of_live_active()) {
+		int ret;
+
+		ret = of_address_to_resource(ofnode_to_np(node),
+					     index, &r);
+		if (ret) {
+			debug("%s: Could not read resource of range %d (ret = %d)\n",
+			      ofnode_get_name(node), index, ret);
+			return ret;
+		}
+
+		range->start = r.start;
+		range->size = r.end - r.start + 1;
+	} else {
+		int offset = ofnode_to_offset(node);
+
+		range->start = fdtdec_get_addr_size_fixed(gd->fdt_blob, offset,
+							  "reg", index,
+							  addr_len, size_len,
+							  &sz, true);
+		if (range->start == FDT_ADDR_T_NONE) {
+			debug("%s: Could not read start of range %d\n",
+			      ofnode_get_name(node), index);
+			return -EINVAL;
+		}
+
+		range->size = sz;
+	}
+
+	return 0;
+}
+
 int regmap_init_mem(ofnode node, struct regmap **mapp)
 {
 	struct regmap_range *range;
@@ -64,7 +116,6 @@ int regmap_init_mem(ofnode node, struct regmap **mapp)
 	int addr_len, size_len, both_len;
 	int len;
 	int index;
-	struct resource r;
 
 	addr_len = ofnode_read_simple_addr_cells(ofnode_get_parent(node));
 	if (addr_len < 0) {
@@ -101,17 +152,10 @@ int regmap_init_mem(ofnode node, struct regmap **mapp)
 
 	for (range = map->ranges, index = 0; count > 0;
 	     count--, range++, index++) {
-		fdt_size_t sz;
-		if (of_live_active()) {
-			of_address_to_resource(ofnode_to_np(node), index, &r);
-			range->start = r.start;
-			range->size = r.end - r.start + 1;
-		} else {
-			range->start = fdtdec_get_addr_size_fixed(gd->fdt_blob,
-					ofnode_to_offset(node), "reg", index,
-					addr_len, size_len, &sz, true);
-			range->size = sz;
-		}
+		int ret = init_range(node, range, addr_len, size_len, index);
+
+		if (ret)
+			return ret;
 	}
 
 	*mapp = map;

--
2.11.0

#8 lore raw

PATCH v4 5/13 regmap: Introduce init_range

Quoted 16 lines · On Fri, 3 Aug 2018 10:01:10 +0200 Mario Six mario.six at gdsys.cc wrote:

Both fdtdec_get_addr_size_fixed and of_address_to_resource can fail with an error, which is not currently checked during regmap initialization.

Since the indentation depth is already quite deep, extract a new 'init_range' method to do the initialization.

---

v3 -> v4:
* Introduced else clause in of_live_active() if statement to make the distinction between live and non-live cases clearer

Quoted 6 lines

v2 -> v3:
New in v3

---
 drivers/core/regmap.c | 68 ++++++++++++++++++++++++++++++++++++++++++---------
 1 file changed, 56 insertions(+), 12 deletions(-)

Applied to u-boot-staging/agust at denx.de, thanks!

--
Anatolij

author
#9 lore raw

PATCH v4 6/13 regmap: Add error output

In reply to Mario Six

Add some debug output in cases where the initialization of a regmap fails.

---

v3 -> v4:
No changes

v2 -> v3:
New in v3

---
 drivers/core/regmap.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)
drivers/core/regmap.c +8 -2
index 25c1ae5d7df..154426269d9 100644
@@ -139,12 +139,18 @@ int regmap_init_mem(ofnode node, struct regmap **mapp)
 	}
 
 	len = ofnode_read_size(node, "reg");
-	if (len < 0)
+	if (len < 0) {
+		debug("%s: Error while reading reg size (ret = %d)\n",
+		      ofnode_get_name(node), len);
 		return len;
+	}
 	len /= sizeof(fdt32_t);
 	count = len / both_len;
-	if (!count)
+	if (!count) {
+		debug("%s: Not enough data in reg property\n",
+		      ofnode_get_name(node));
 		return -EINVAL;
+	}
 
 	map = regmap_alloc(count);
 	if (!map)

--
2.11.0

#10 lore raw

PATCH v4 6/13 regmap: Add error output

Quoted 12 lines · On Fri, 3 Aug 2018 10:01:11 +0200 Mario Six mario.six at gdsys.cc wrote:

Add some debug output in cases where the initialization of a regmap fails.

---

v3 -> v4:
No changes

Quoted 6 lines

v2 -> v3:
New in v3

---
 drivers/core/regmap.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

Applied to u-boot-staging/agust at denx.de, thanks!

--
Anatolij

author
#11 lore raw

PATCH v4 7/13 regmap: Add raw read/write functions

In reply to Mario Six

The regmap functions currently assume that all register map accesses have a data width of 32 bits, but there are maps that have different widths.

To rectify this, implement the regmap_raw_read and regmap_raw_write functions from the Linux kernel API that specify the width of a desired read or write operation on a regmap.

Implement the regmap_read and regmap_write functions using these raw functions in a backwards-compatible manner.

---

v3 -> v4:
* Switched 'ranges[0] + offset' to 'ranges[0].start + offset'
* Explained the difference between the raw and non-raw read/write functions better in the docs

v2 -> v3:
* Implement the "raw" functions from Linux instead of adding a size parameter to the regmap_{read,write} functions
* Fixed style violation
* Improved error handling

v1 -> v2:
New in v2

---
 drivers/core/regmap.c | 54 ++++++++++++++++++++++++++++++++++++++++++-------
 include/regmap.h      | 56 +++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 103 insertions(+), 7 deletions(-)
drivers/core/regmap.c +47 -7
index 154426269d9..87c4c950a63 100644
@@ -188,22 +188,62 @@ int regmap_uninit(struct regmap *map)
 	return 0;
 }
 
+int regmap_raw_read(struct regmap *map, uint offset, void *valp, size_t val_len)
+{
+	void *ptr;
+
+	ptr = map_physmem(map->ranges[0].start + offset, val_len, MAP_NOCACHE);
+
+	switch (val_len) {
+	case REGMAP_SIZE_8:
+		*((u8 *)valp) = in_8((u8 *)ptr);
+		break;
+	case REGMAP_SIZE_16:
+		*((u16 *)valp) = in_le16((u16 *)ptr);
+		break;
+	case REGMAP_SIZE_32:
+		*((u32 *)valp) = in_le32((u32 *)ptr);
+		break;
+	default:
+		debug("%s: regmap size %u unknown\n", __func__, val_len);
+		return -EINVAL;
+	}
+	return 0;
+}
+
 int regmap_read(struct regmap *map, uint offset, uint *valp)
 {
-	u32 *ptr = map_physmem(map->ranges[0].start + offset, 4, MAP_NOCACHE);
+	return regmap_raw_read(map, offset, valp, REGMAP_SIZE_32);
+}
 
-	*valp = le32_to_cpu(readl(ptr));
+int regmap_raw_write(struct regmap *map, uint offset, const void *val,
+		     size_t val_len)
+{
+	void *ptr;
+
+	ptr = map_physmem(map->ranges[0].start + offset, val_len, MAP_NOCACHE);
+
+	switch (val_len) {
+	case REGMAP_SIZE_8:
+		out_8((u8 *)ptr, *((u8 *)val));
+		break;
+	case REGMAP_SIZE_16:
+		out_le16((u16 *)ptr, *((u16 *)val));
+		break;
+	case REGMAP_SIZE_32:
+		out_le32((u32 *)ptr, *((u32 *)val));
+		break;
+	default:
+		debug("%s: regmap size %u unknown\n", __func__, val_len);
+		return -EINVAL;
+	}
 
 	return 0;
 }
 
 int regmap_write(struct regmap *map, uint offset, uint val)
 {
-	u32 *ptr = map_physmem(map->ranges[0].start + offset, 4, MAP_NOCACHE);
-
-	writel(cpu_to_le32(val), ptr);
-
-	return 0;
+	return regmap_raw_write(map, offset, &val, REGMAP_SIZE_32);
 }
 
 int regmap_update_bits(struct regmap *map, uint offset, uint mask, uint val)
include/regmap.h +56 -0

78 lines not on this page · raw message

--
2.11.0

#12 lore raw

PATCH v4 7/13 regmap: Add raw read/write functions

Quoted 27 lines · On Fri, 3 Aug 2018 10:01:12 +0200 Mario Six mario.six at gdsys.cc wrote:

The regmap functions currently assume that all register map accesses have a data width of 32 bits, but there are maps that have different widths.

To rectify this, implement the regmap_raw_read and regmap_raw_write functions from the Linux kernel API that specify the width of a desired read or write operation on a regmap.

Implement the regmap_read and regmap_write functions using these raw functions in a backwards-compatible manner.

---

v3 -> v4:
* Switched 'ranges[0] + offset' to 'ranges[0].start + offset'
* Explained the difference between the raw and non-raw read/write functions better in the docs

v2 -> v3:
* Implement the "raw" functions from Linux instead of adding a size parameter to the regmap_{read,write} functions
* Fixed style violation
* Improved error handling

Quoted 7 lines

v1 -> v2:
New in v2

---
 drivers/core/regmap.c | 54 ++++++++++++++++++++++++++++++++++++++++++-------
 include/regmap.h      | 56 +++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 103 insertions(+), 7 deletions(-)

Applied to u-boot-staging/agust at denx.de, thanks!

--
Anatolij

#13 lore raw

PATCH v4 7/13 regmap: Add raw read/write functions

In reply to Mario Six

Hi Mario,

On Fri, 3 Aug 2018 10:01:12 +0200
Mario Six mario.six at gdsys.cc wrote:
...

Quoted 5 lines
drivers/core/regmap.c +5 -0
+int regmap_raw_read(struct regmap *map, uint offset, void *valp, size_t val_len)
+{
+	void *ptr;
+
+	ptr = map_physmem(map->ranges[0].start + offset, val_len, MAP_NOCACHE);

drivers/core/regmap.c int regmap_uninit(struct regmap *map) · line 205

Quoted 10 lines
drivers/core/regmap.c +10 -0
+
+	switch (val_len) {
+	case REGMAP_SIZE_8:
+		*((u8 *)valp) = in_8((u8 *)ptr);
+		break;
+	case REGMAP_SIZE_16:
+		*((u16 *)valp) = in_le16((u16 *)ptr);
+		break;
+	case REGMAP_SIZE_32:
+		*((u32 *)valp) = in_le32((u32 *)ptr);

this breaks building for MIPS:

https://travis-ci.org/vdsao/u-boot-video/jobs/412722192#L869

--
Anatolij

author
#14 lore raw

PATCH v4 7/13 regmap: Add raw read/write functions

Hi Anatolij,

Quoted 24 lines · Anatolij Gustschin

Hi Mario,

On Fri, 3 Aug 2018 10:01:12 +0200
Mario Six mario.six at gdsys.cc wrote:
...

Quoted 5 lines
drivers/core/regmap.c +5 -0
+int regmap_raw_read(struct regmap *map, uint offset, void *valp, size_t val_len)
+{
+     void *ptr;
+
+     ptr = map_physmem(map->ranges[0].start + offset, val_len, MAP_NOCACHE);

drivers/core/regmap.c int regmap_uninit(struct regmap *map) · line 205

Quoted 10 lines
drivers/core/regmap.c +10 -0
+
+     switch (val_len) {
+     case REGMAP_SIZE_8:
+             *((u8 *)valp) = in_8((u8 *)ptr);
+             break;
+     case REGMAP_SIZE_16:
+             *((u16 *)valp) = in_le16((u16 *)ptr);
+             break;
+     case REGMAP_SIZE_32:
+             *((u32 *)valp) = in_le32((u32 *)ptr);

this breaks building for MIPS:

Quoted 2 lines

https://travis-ci.org/vdsao/u-boot-video/jobs/412722192#L869

I see. MIPS apparently doesn't implement any of the in/out functions at all. But since the necessary __raw_read/__raw_write functions exist, it's just a matter of a few lines to add them. I'll add a patch that will fix this.

There also seems to be a warning about a format specifier on MIPS, I'll fix that as well.

To make things less confusing, I'll import your v5 of "misc: Sort Makefile entries" and send a full v6 series.

Quoted 1 line

Anatolij

Best regards,
Mario

#15 lore raw

PATCH v4 7/13 regmap: Add raw read/write functions

Hi Mario,

On Tue, 7 Aug 2018 09:36:33 +0200
Mario Six mario.six at gdsys.cc wrote:
...

Quoted 7 lines
Quoted 4 lines

this breaks building for MIPS:

https://travis-ci.org/vdsao/u-boot-video/jobs/412722192#L869

I see. MIPS apparently doesn't implement any of the in/out functions at all. But since the necessary __raw_read/__raw_write functions exist, it's just a matter of a few lines to add them. I'll add a patch that will fix this.

OK, thanks!

Quoted 2 lines

There also seems to be a warning about a format specifier on MIPS, I'll fix that as well.

Yes, for build-testing I applied the patch with changes for "%zu" format specifier to fix warnings when building sandbox:

https://github.com/vdsao/u-boot-video/commit/0968a2bf28b7c5c0dd4535b5925b77cdbf5d2557

Not sure if it is enough.

Quoted 2 lines

To make things less confusing, I'll import your v5 of "misc: Sort Makefile entries" and send a full v6 series.

Thanks!

--
Anatolij

#16 lore raw

PATCH v4 7/13 regmap: Add raw read/write functions

In reply to Mario Six

2018-08-07 9:36 GMT+02:00 Mario Six <[email protected]>:

Quoted 26 lines

Hi Anatolij,

Quoted 24 lines · Anatolij Gustschin

Hi Mario,

On Fri, 3 Aug 2018 10:01:12 +0200
Mario Six mario.six at gdsys.cc wrote:
...

Quoted 5 lines
drivers/core/regmap.c +5 -0
+int regmap_raw_read(struct regmap *map, uint offset, void *valp, size_t val_len)
+{
+     void *ptr;
+
+     ptr = map_physmem(map->ranges[0].start + offset, val_len, MAP_NOCACHE);

drivers/core/regmap.c int regmap_uninit(struct regmap *map) · line 205

Quoted 10 lines
drivers/core/regmap.c +10 -0
+
+     switch (val_len) {
+     case REGMAP_SIZE_8:
+             *((u8 *)valp) = in_8((u8 *)ptr);
+             break;
+     case REGMAP_SIZE_16:
+             *((u16 *)valp) = in_le16((u16 *)ptr);
+             break;
+     case REGMAP_SIZE_32:
+             *((u32 *)valp) = in_le32((u32 *)ptr);

this breaks building for MIPS:

Quoted 5 lines
Quoted 2 lines

https://travis-ci.org/vdsao/u-boot-video/jobs/412722192#L869

I see. MIPS apparently doesn't implement any of the in/out functions at all. But since the necessary __raw_read/__raw_write functions exist, it's just a matter of a few lines to add them. I'll add a patch that will fix this.

why don't you directly use readb(), readw(), readl() etc.? What about 64 Bit registers (readq())?

Quoted 15 lines

There also seems to be a warning about a format specifier on MIPS, I'll fix that as well.

To make things less confusing, I'll import your v5 of "misc: Sort Makefile entries" and send a full v6 series.

Quoted 1 line

Anatolij

Best regards,
Mario
_______________________________________________
U-Boot mailing list
U-Boot at lists.denx.de
https://lists.denx.de/listinfo/u-boot

Signature · 1 line
- Daniel
author
#17 lore raw

PATCH v4 7/13 regmap: Add raw read/write functions

Quoted 28 lines · Daniel Schwierzeck

2018-08-07 9:36 GMT+02:00 Mario Six <[email protected]>:

Quoted 26 lines

Hi Anatolij,

Quoted 24 lines · Anatolij Gustschin

Hi Mario,

On Fri, 3 Aug 2018 10:01:12 +0200
Mario Six mario.six at gdsys.cc wrote:
...

Quoted 5 lines
drivers/core/regmap.c +5 -0
+int regmap_raw_read(struct regmap *map, uint offset, void *valp, size_t val_len)
+{
+     void *ptr;
+
+     ptr = map_physmem(map->ranges[0].start + offset, val_len, MAP_NOCACHE);

drivers/core/regmap.c int regmap_uninit(struct regmap *map) · line 205

Quoted 10 lines
drivers/core/regmap.c +10 -0
+
+     switch (val_len) {
+     case REGMAP_SIZE_8:
+             *((u8 *)valp) = in_8((u8 *)ptr);
+             break;
+     case REGMAP_SIZE_16:
+             *((u16 *)valp) = in_le16((u16 *)ptr);
+             break;
+     case REGMAP_SIZE_32:
+             *((u32 *)valp) = in_le32((u32 *)ptr);

this breaks building for MIPS:

Quoted 9 lines
Quoted 5 lines
Quoted 2 lines

https://travis-ci.org/vdsao/u-boot-video/jobs/412722192#L869

I see. MIPS apparently doesn't implement any of the in/out functions at all. But since the necessary __raw_read/__raw_write functions exist, it's just a matter of a few lines to add them. I'll add a patch that will fix this.

why don't you directly use readb(), readw(), readl() etc.? What about 64 Bit registers (readq())?

readb/readw/etc. are not endianness-safe, so the regmap functions would return different values on a big-endian and a little-endian system for the same IC, and since the vast majority of IC registers are little-endian, having LE as the default seems reasonable. That does make it a bit more complicated to implement drivers for big-endian register spaces (wrapping values in le*_to_cpu), but for everything else the default will work as expected.

The 64-bit support is a good point. It could be added with a few lines, so if it's needed I could.

Quoted 7 lines
Quoted 7 lines

There also seems to be a warning about a format specifier on MIPS, I'll fix that as well.

To make things less confusing, I'll import your v5 of "misc: Sort Makefile entries" and send a full v6 series.

Quoted 7 lines
Quoted 5 lines
Quoted 1 line

Anatolij

Best regards,
Mario

- Daniel

Best regards,
Mario

author
#18 lore raw

PATCH v4 8/13 regmap: Support reading from specific range

In reply to Mario Six

It is useful to be able to treat the different ranges of a regmap separately to be able to use distinct offset for them, but this is currently not implemented in the regmap API.

To preserve backwards compatibility, add regmap_read_range and regmap_write_range functions that take an additional parameter 'range_num' that identifies the range to operate on.

---

v3 -> v4:
No changes

v2 -> v3:
* Renamed the functions to regmap_{write,read}_range
* Added function comments
* Fixed style violations
* Improved error handling

v1 -> v2:
New in v2

---
 drivers/core/regmap.c | 49 ++++++++++++++++++++++++++++++++++++++++++++-----
 include/regmap.h      | 31 +++++++++++++++++++++++++++++++
 2 files changed, 75 insertions(+), 5 deletions(-)
drivers/core/regmap.c +44 -5
index 87c4c950a63..86b35edc050 100644
@@ -188,11 +188,25 @@ int regmap_uninit(struct regmap *map)
 	return 0;
 }
 
-int regmap_raw_read(struct regmap *map, uint offset, void *valp, size_t val_len)
+int regmap_raw_read_range(struct regmap *map, uint range_num, uint offset,
+			  void *valp, size_t val_len)
 {
+	struct regmap_range *range;
 	void *ptr;
 
-	ptr = map_physmem(map->ranges[0].start + offset, val_len, MAP_NOCACHE);
+	if (range_num >= map->range_count) {
+		debug("%s: range index %d larger than range count\n",
+		      __func__, range_num);
+		return -ERANGE;
+	}
+	range = &map->ranges[range_num];
+
+	ptr = map_physmem(range->start + offset, val_len, MAP_NOCACHE);
+
+	if (offset + val_len > range->size) {
+		debug("%s: offset/size combination invalid\n", __func__);
+		return -ERANGE;
+	}
 
 	switch (val_len) {
 	case REGMAP_SIZE_8:
@@ -208,20 +222,39 @@ int regmap_raw_read(struct regmap *map, uint offset, void *valp, size_t val_len)
 		debug("%s: regmap size %u unknown\n", __func__, val_len);
 		return -EINVAL;
 	}
+
 	return 0;
 }
 
+int regmap_raw_read(struct regmap *map, uint offset, void *valp, size_t val_len)
+{
+	return regmap_raw_read_range(map, 0, offset, valp, val_len);
+}
+
 int regmap_read(struct regmap *map, uint offset, uint *valp)
 {
 	return regmap_raw_read(map, offset, valp, REGMAP_SIZE_32);
 }
 
-int regmap_raw_write(struct regmap *map, uint offset, const void *val,
-		     size_t val_len)
+int regmap_raw_write_range(struct regmap *map, uint range_num, uint offset,
+			   const void *val, size_t val_len)
 {
+	struct regmap_range *range;
 	void *ptr;
 
-	ptr = map_physmem(map->ranges[0].start + offset, val_len, MAP_NOCACHE);
+	if (range_num >= map->range_count) {
+		debug("%s: range index %d larger than range count\n",
+		      __func__, range_num);
+		return -ERANGE;
+	}
+	range = &map->ranges[range_num];
+
+	ptr = map_physmem(range->start + offset, val_len, MAP_NOCACHE);
+
+	if (offset + val_len > range->size) {
+		debug("%s: offset/size combination invalid\n", __func__);
+		return -ERANGE;
+	}
 
 	switch (val_len) {
 	case REGMAP_SIZE_8:
@@ -241,6 +274,12 @@ int regmap_raw_write(struct regmap *map, uint offset, const void *val,
 	return 0;
 }
 
+int regmap_raw_write(struct regmap *map, uint offset, const void *val,
+		     size_t val_len)
+{
+	return regmap_raw_write_range(map, 0, offset, val, val_len);
+}
+
 int regmap_write(struct regmap *map, uint offset, uint val)
 {
 	return regmap_raw_write(map, offset, &val, REGMAP_SIZE_32);
include/regmap.h +31 -0
index 46ae4ed0a03..df672df9fe1 100644
@@ -113,6 +113,37 @@ int regmap_raw_write(struct regmap *map, uint offset, const void *val,
 int regmap_raw_read(struct regmap *map, uint offset, void *valp,
 		    size_t val_len);
 
+/**
+ * regmap_raw_write_range() - Write a value of specified length to a range of a
+ *			      regmap
+ *
+ * @map:	Regmap to write to
+ * @range_num:	Number of the range in the regmap to write to
+ * @offset:	Offset in the regmap to write to
+ * @val:	Value to write to the regmap at the specified offset
+ * @val_len:	Length of the data to be written to the regmap
+ *
+ * Return: 0 if OK, -ve on error
+ */
+int regmap_raw_write_range(struct regmap *map, uint range_num, uint offset,
+			   const void *val, size_t val_len);
+
+/**
+ * regmap_raw_read_range() - Read a value of specified length from a range of a
+ *			     regmap
+ *
+ * @map:	Regmap to read from
+ * @range_num:	Number of the range in the regmap to write to
+ * @offset:	Offset in the regmap to read from
+ * @valp:	Pointer to the buffer to receive the data read from the regmap
+ *		at the specified offset
+ * @val_len:	Length of the data to be read from the regmap
+ *
+ * Return: 0 if OK, -ve on error
+ */
+int regmap_raw_read_range(struct regmap *map, uint range_num, uint offset,
+			  void *valp, size_t val_len);
+
 #define regmap_write32(map, ptr, member, val) \
 	regmap_write(map, (uint32_t *)(ptr)->member - (uint32_t *)(ptr), val)
 

--
2.11.0

#19 lore raw

PATCH v4 8/13 regmap: Support reading from specific range

On Fri, 3 Aug 2018 10:01:13 +0200
Mario Six mario.six at gdsys.cc wrote:
...

Quoted 4 lines
---
 drivers/core/regmap.c | 49 ++++++++++++++++++++++++++++++++++++++++++++-----
 include/regmap.h      | 31 +++++++++++++++++++++++++++++++
 2 files changed, 75 insertions(+), 5 deletions(-)

Applied to u-boot-staging/agust at denx.de, thanks!

--
Anatolij

#20 lore raw

PATCH v4 8/13 regmap: Support reading from specific range

In reply to Mario Six

On Fri, 3 Aug 2018 10:01:13 +0200
Mario Six mario.six at gdsys.cc wrote:
...

Quoted 4 lines
---
 drivers/core/regmap.c | 49 ++++++++++++++++++++++++++++++++++++++++++++-----
 include/regmap.h      | 31 +++++++++++++++++++++++++++++++
 2 files changed, 75 insertions(+), 5 deletions(-)

Applied to u-boot-staging/agust at denx.de, thanks!

--
Anatolij

author
#21 lore raw

PATCH v4 9/13 regmap: Define regmap_{get,set}

In reply to Mario Six

It would be convenient if one could use the regmap API in conjunction with register maps defined as structs (i.e. structs that directly mirror the memory layout of the registers in question). A similar approach was planned with the regmap_write32/regmap_read32 macros, but was never used.

Hence, implement regmap_set/regmap_range_set and regmap_get/regmap_range_get macros, which, given a register map, a struct describing the layout of the register map, and a member name automatically produce regmap_read/regmap_write calls that access the specified member in the register map.

---

v3 -> v4:
No changes

v2 -> v3:
* Fixed style violations
* Added documentation

v1 -> v2:
New in v2

---
 include/regmap.h | 54 ++++++++++++++++++++++++++++++++++++++++++++++++++----
 1 file changed, 50 insertions(+), 4 deletions(-)
include/regmap.h +50 -4
index df672df9fe1..85c4a8e502f 100644
@@ -144,11 +144,57 @@ int regmap_raw_write_range(struct regmap *map, uint range_num, uint offset,
 int regmap_raw_read_range(struct regmap *map, uint range_num, uint offset,
 			  void *valp, size_t val_len);
 
-#define regmap_write32(map, ptr, member, val) \
-	regmap_write(map, (uint32_t *)(ptr)->member - (uint32_t *)(ptr), val)
+/**
+ * regmap_range_set() - Set a value in a regmap range described by a struct
+ * @map:    Regmap in which a value should be set
+ * @range:  Range of the regmap in which a value should be set
+ * @type:   Structure type that describes the memory layout of the regmap range
+ * @member: Member of the describing structure that should be set in the regmap
+ *          range
+ * @val:    Value which should be written to the regmap range
+ */
+#define regmap_range_set(map, range, type, member, val) \
+	do { \
+		typeof(((type *)0)->member) __tmp = val; \
+		regmap_raw_write_range(map, range, offsetof(type, member), \
+				       &__tmp, sizeof(((type *)0)->member)); \
+	} while (0)
+
+/**
+ * regmap_set() - Set a value in a regmap described by a struct
+ * @map:    Regmap in which a value should be set
+ * @type:   Structure type that describes the memory layout of the regmap
+ * @member: Member of the describing structure that should be set in the regmap
+ * @val:    Value which should be written to the regmap
+ */
+#define regmap_set(map, type, member, val) \
+	regmap_range_set(map, 0, type, member, val)
 
-#define regmap_read32(map, ptr, member, valp) \
-	regmap_read(map, (uint32_t *)(ptr)->member - (uint32_t *)(ptr), valp)
+/**
+ * regmap_range_get() - Get a value from a regmap range described by a struct
+ * @map:    Regmap from which a value should be read
+ * @range:  Range of the regmap from which a value should be read
+ * @type:   Structure type that describes the memory layout of the regmap
+ *          range
+ * @member: Member of the describing structure that should be read in the
+ *          regmap range
+ * @valp:   Variable that receives the value read from the regmap range
+ */
+#define regmap_range_get(map, range, type, member, valp) \
+	regmap_raw_read_range(map, range, offsetof(type, member), \
+			      (void *)valp, sizeof(((type *)0)->member))
+
+/**
+ * regmap_get() - Get a value from a regmap described by a struct
+ * @map:    Regmap from which a value should be read
+ * @type:   Structure type that describes the memory layout of the regmap
+ *          range
+ * @member: Member of the describing structure that should be read in the
+ *          regmap
+ * @valp:   Variable that receives the value read from the regmap
+ */
+#define regmap_get(map, type, member, valp) \
+	regmap_range_get(map, 0, type, member, valp)
 
 /**
  * regmap_update_bits() - Perform a read/modify/write using a mask

--
2.11.0

author
#22 lore raw

PATCH v4 10/13 test: regmap: Add test for regmap_{set, get}

In reply to Mario Six

Add test for regmap_{set,get} functions.

---

v3 -> v4:
No changes

v2 -> v3:
New in v3

---
 test/dm/regmap.c | 28 ++++++++++++++++++++++++++++
 1 file changed, 28 insertions(+)
test/dm/regmap.c +28 -0
index d4b86b3b03c..152b3a4b800 100644
@@ -116,3 +116,31 @@ static int dm_test_regmap_rw(struct unit_test_state *uts)
 }
 
 DM_TEST(dm_test_regmap_rw, DM_TESTF_SCAN_PDATA | DM_TESTF_SCAN_FDT);
+
+/* Get/Set test */
+static int dm_test_regmap_getset(struct unit_test_state *uts)
+{
+	struct udevice *dev;
+	struct regmap *map;
+	uint reg;
+	struct layout {
+		u32 val0;
+		u32 val1;
+		u32 val2;
+		u32 val3;
+	};
+
+	ut_assertok(uclass_get_device(UCLASS_SYSCON, 0, &dev));
+	map = syscon_get_regmap(dev);
+	ut_assertok_ptr(map);
+
+	regmap_set(map, struct layout, val0, 0xcacafafa);
+	regmap_set(map, struct layout, val3, 0x55aa2211);
+
+	ut_assertok(regmap_get(map, struct layout, val0, &reg));
+	ut_assertok(regmap_get(map, struct layout, val3, &reg));
+
+	return 0;
+}
+
+DM_TEST(dm_test_regmap_getset, DM_TESTF_SCAN_PDATA | DM_TESTF_SCAN_FDT);

--
2.11.0

#23 lore raw

PATCH v4 10/13 test: regmap: Add test for regmap_{set, get}

On Fri, 3 Aug 2018 10:01:15 +0200
Mario Six mario.six at gdsys.cc wrote:
...

Quoted 3 lines
---
 test/dm/regmap.c | 28 ++++++++++++++++++++++++++++
 1 file changed, 28 insertions(+)

Applied to u-boot-staging/agust at denx.de, thanks!

--
Anatolij

author
#24 lore raw

PATCH v4 11/13 misc: Sort Makefile entries

In reply to Mario Six

Makefile entries should be sorted.

---

v3 -> v4:
No changes

v2 -> v3:
New in v3

---
 drivers/misc/Makefile | 60 +++++++++++++++++++++++++++------------------------
 1 file changed, 32 insertions(+), 28 deletions(-)
drivers/misc/Makefile +32 -28
index b0ef97b6b31..3ab110a1aaa 100644
@@ -4,11 +4,7 @@
 # Wolfgang Denk, DENX Software Engineering, wd at denx.de.
 
 obj-$(CONFIG_MISC) += misc-uclass.o
-obj-$(CONFIG_ALI152X) += ali512x.o
-obj-$(CONFIG_ALTERA_SYSID) += altera_sysid.o
-obj-$(CONFIG_ATSHA204A) += atsha204a-i2c.o
-obj-$(CONFIG_DS4510)  += ds4510.o
-obj-$(CONFIG_CBMEM_CONSOLE) += cbmem_console.o
+
 ifndef CONFIG_SPL_BUILD
 obj-$(CONFIG_CROS_EC) += cros_ec.o
 obj-$(CONFIG_CROS_EC_LPC) += cros_ec_lpc.o
@@ -16,43 +12,51 @@ obj-$(CONFIG_CROS_EC_I2C) += cros_ec_i2c.o
 obj-$(CONFIG_CROS_EC_SANDBOX) += cros_ec_sandbox.o
 obj-$(CONFIG_CROS_EC_SPI) += cros_ec_spi.o
 endif
-obj-$(CONFIG_FSL_IIM) += fsl_iim.o
-obj-$(CONFIG_LED_STATUS_GPIO) += gpio_led.o
-obj-$(CONFIG_$(SPL_)I2C_EEPROM) += i2c_eeprom.o
-obj-$(CONFIG_FSL_MC9SDZ60) += mc9sdz60.o
-obj-$(CONFIG_MXC_OCOTP) += mxc_ocotp.o
-obj-$(CONFIG_MXS_OCOTP) += mxs_ocotp.o
-obj-$(CONFIG_NUVOTON_NCT6102D) += nuvoton_nct6102d.o
-obj-$(CONFIG_NS87308) += ns87308.o
-obj-$(CONFIG_$(SPL_)PWRSEQ) += pwrseq-uclass.o
+
 ifdef CONFIG_DM_I2C
 ifndef CONFIG_SPL_BUILD
 obj-$(CONFIG_SANDBOX) += i2c_eeprom_emul.o
 endif
 endif
-obj-$(CONFIG_SMSC_LPC47M) += smsc_lpc47m.o
-obj-$(CONFIG_SMSC_SIO1007) += smsc_sio1007.o
-obj-$(CONFIG_LED_STATUS) += status_led.o
-obj-$(CONFIG_SANDBOX) += swap_case.o
+
 ifdef CONFIG_SPL_OF_PLATDATA
 ifdef CONFIG_SPL_BUILD
 obj-$(CONFIG_SANDBOX) += spltest_sandbox.o
 endif
 endif
-obj-$(CONFIG_SANDBOX) += syscon_sandbox.o misc_sandbox.o
-obj-$(CONFIG_TEGRA_CAR) += tegra_car.o
-obj-$(CONFIG_TEGRA186_BPMP) += tegra186_bpmp.o
-obj-$(CONFIG_TWL4030_LED) += twl4030_led.o
+
+obj-$(CONFIG_ALI152X) += ali512x.o
+obj-$(CONFIG_ALTERA_SYSID) += altera_sysid.o
+obj-$(CONFIG_ATSHA204A) += atsha204a-i2c.o
+obj-$(CONFIG_CBMEM_CONSOLE) += cbmem_console.o
+obj-$(CONFIG_DS4510)  += ds4510.o
+obj-$(CONFIG_FSL_DEVICE_DISABLE) += fsl_devdis.o
 obj-$(CONFIG_FSL_IFC) += fsl_ifc.o
+obj-$(CONFIG_FSL_IIM) += fsl_iim.o
+obj-$(CONFIG_FSL_MC9SDZ60) += mc9sdz60.o
 obj-$(CONFIG_FSL_SEC_MON) += fsl_sec_mon.o
+obj-$(CONFIG_GDSYS_IOEP) += gdsys_ioep.o
+obj-$(CONFIG_GDSYS_RXAUI_CTRL) += gdsys_rxaui_ctrl.o
+obj-$(CONFIG_$(SPL_)I2C_EEPROM) += i2c_eeprom.o
+obj-$(CONFIG_LED_STATUS) += status_led.o
+obj-$(CONFIG_LED_STATUS_GPIO) += gpio_led.o
+obj-$(CONFIG_MPC83XX_SERDES) += mpc83xx_serdes.o
+obj-$(CONFIG_MXC_OCOTP) += mxc_ocotp.o
+obj-$(CONFIG_MXS_OCOTP) += mxs_ocotp.o
+obj-$(CONFIG_NS87308) += ns87308.o
+obj-$(CONFIG_NUVOTON_NCT6102D) += nuvoton_nct6102d.o
 obj-$(CONFIG_PCA9551_LED) += pca9551_led.o
-obj-$(CONFIG_FSL_DEVICE_DISABLE) += fsl_devdis.o
-obj-$(CONFIG_WINBOND_W83627) += winbond_w83627.o
+obj-$(CONFIG_$(SPL_)PWRSEQ) += pwrseq-uclass.o
 obj-$(CONFIG_QFW) += qfw.o
 obj-$(CONFIG_ROCKCHIP_EFUSE) += rockchip-efuse.o
-obj-$(CONFIG_STM32_RCC) += stm32_rcc.o
+obj-$(CONFIG_SANDBOX) += swap_case.o
+obj-$(CONFIG_SANDBOX) += syscon_sandbox.o misc_sandbox.o
+obj-$(CONFIG_SMSC_LPC47M) += smsc_lpc47m.o
+obj-$(CONFIG_SMSC_SIO1007) += smsc_sio1007.o
 obj-$(CONFIG_STM32MP_FUSE) += stm32mp_fuse.o
+obj-$(CONFIG_STM32_RCC) += stm32_rcc.o
 obj-$(CONFIG_SYS_DPAA_QBMAN) += fsl_portals.o
-obj-$(CONFIG_GDSYS_RXAUI_CTRL) += gdsys_rxaui_ctrl.o
-obj-$(CONFIG_MPC83XX_SERDES) += mpc83xx_serdes.o
-obj-$(CONFIG_GDSYS_IOEP) += gdsys_ioep.o
+obj-$(CONFIG_TEGRA186_BPMP) += tegra186_bpmp.o
+obj-$(CONFIG_TEGRA_CAR) += tegra_car.o
+obj-$(CONFIG_TWL4030_LED) += twl4030_led.o
+obj-$(CONFIG_WINBOND_W83627) += winbond_w83627.o

--
2.11.0

#25 lore raw

PATCH v5 11/13 misc: Sort Makefile entries

From: Mario Six <[email protected]>

Makefile entries should be sorted.

---
v4 -> v5:
Drop re-ordered entries that are not in mainline (e.g. CONFIG_GDSYS_IOEP, CONFIG_MPC83XX_SERDES, misc_sandbox.o)

v3 -> v4:
No changes

v2 -> v3:
New in v3

 drivers/misc/Makefile | 56 +++++++++++++++++++++++++++------------------------
 1 file changed, 30 insertions(+), 26 deletions(-)
drivers/misc/Makefile +30 -26
index 4ce9d213f0..677a230ab4 100644
@@ -4,11 +4,7 @@
 # Wolfgang Denk, DENX Software Engineering, wd at denx.de.
 
 obj-$(CONFIG_MISC) += misc-uclass.o
-obj-$(CONFIG_ALI152X) += ali512x.o
-obj-$(CONFIG_ALTERA_SYSID) += altera_sysid.o
-obj-$(CONFIG_ATSHA204A) += atsha204a-i2c.o
-obj-$(CONFIG_DS4510)  += ds4510.o
-obj-$(CONFIG_CBMEM_CONSOLE) += cbmem_console.o
+
 ifndef CONFIG_SPL_BUILD
 obj-$(CONFIG_CROS_EC) += cros_ec.o
 obj-$(CONFIG_CROS_EC_LPC) += cros_ec_lpc.o
@@ -16,41 +12,49 @@ obj-$(CONFIG_CROS_EC_I2C) += cros_ec_i2c.o
 obj-$(CONFIG_CROS_EC_SANDBOX) += cros_ec_sandbox.o
 obj-$(CONFIG_CROS_EC_SPI) += cros_ec_spi.o
 endif
-obj-$(CONFIG_FSL_IIM) += fsl_iim.o
-obj-$(CONFIG_LED_STATUS_GPIO) += gpio_led.o
-obj-$(CONFIG_$(SPL_)I2C_EEPROM) += i2c_eeprom.o
-obj-$(CONFIG_FSL_MC9SDZ60) += mc9sdz60.o
-obj-$(CONFIG_MXC_OCOTP) += mxc_ocotp.o
-obj-$(CONFIG_MXS_OCOTP) += mxs_ocotp.o
-obj-$(CONFIG_NUVOTON_NCT6102D) += nuvoton_nct6102d.o
-obj-$(CONFIG_NS87308) += ns87308.o
-obj-$(CONFIG_$(SPL_)PWRSEQ) += pwrseq-uclass.o
+
 ifdef CONFIG_DM_I2C
 ifndef CONFIG_SPL_BUILD
 obj-$(CONFIG_SANDBOX) += i2c_eeprom_emul.o
 endif
 endif
-obj-$(CONFIG_SMSC_LPC47M) += smsc_lpc47m.o
-obj-$(CONFIG_SMSC_SIO1007) += smsc_sio1007.o
-obj-$(CONFIG_LED_STATUS) += status_led.o
-obj-$(CONFIG_SANDBOX) += swap_case.o
+
 ifdef CONFIG_SPL_OF_PLATDATA
 ifdef CONFIG_SPL_BUILD
 obj-$(CONFIG_SANDBOX) += spltest_sandbox.o
 endif
 endif
-obj-$(CONFIG_SANDBOX) += syscon_sandbox.o
-obj-$(CONFIG_TEGRA_CAR) += tegra_car.o
-obj-$(CONFIG_TEGRA186_BPMP) += tegra186_bpmp.o
-obj-$(CONFIG_TWL4030_LED) += twl4030_led.o
+
+obj-$(CONFIG_ALI152X) += ali512x.o
+obj-$(CONFIG_ALTERA_SYSID) += altera_sysid.o
+obj-$(CONFIG_ATSHA204A) += atsha204a-i2c.o
+obj-$(CONFIG_CBMEM_CONSOLE) += cbmem_console.o
+obj-$(CONFIG_DS4510)  += ds4510.o
+obj-$(CONFIG_FSL_DEVICE_DISABLE) += fsl_devdis.o
 obj-$(CONFIG_FSL_IFC) += fsl_ifc.o
+obj-$(CONFIG_FSL_IIM) += fsl_iim.o
+obj-$(CONFIG_FSL_MC9SDZ60) += mc9sdz60.o
 obj-$(CONFIG_FSL_SEC_MON) += fsl_sec_mon.o
+obj-$(CONFIG_GDSYS_RXAUI_CTRL) += gdsys_rxaui_ctrl.o
+obj-$(CONFIG_$(SPL_)I2C_EEPROM) += i2c_eeprom.o
+obj-$(CONFIG_LED_STATUS) += status_led.o
+obj-$(CONFIG_LED_STATUS_GPIO) += gpio_led.o
+obj-$(CONFIG_MXC_OCOTP) += mxc_ocotp.o
+obj-$(CONFIG_MXS_OCOTP) += mxs_ocotp.o
+obj-$(CONFIG_NS87308) += ns87308.o
+obj-$(CONFIG_NUVOTON_NCT6102D) += nuvoton_nct6102d.o
 obj-$(CONFIG_PCA9551_LED) += pca9551_led.o
-obj-$(CONFIG_FSL_DEVICE_DISABLE) += fsl_devdis.o
-obj-$(CONFIG_WINBOND_W83627) += winbond_w83627.o
+obj-$(CONFIG_$(SPL_)PWRSEQ) += pwrseq-uclass.o
 obj-$(CONFIG_QFW) += qfw.o
 obj-$(CONFIG_ROCKCHIP_EFUSE) += rockchip-efuse.o
-obj-$(CONFIG_STM32_RCC) += stm32_rcc.o
+obj-$(CONFIG_SANDBOX) += swap_case.o
+obj-$(CONFIG_SANDBOX) += syscon_sandbox.o
+obj-$(CONFIG_SMSC_LPC47M) += smsc_lpc47m.o
+obj-$(CONFIG_SMSC_SIO1007) += smsc_sio1007.o
 obj-$(CONFIG_STM32MP_FUSE) += stm32mp_fuse.o
+obj-$(CONFIG_STM32_RCC) += stm32_rcc.o
 obj-$(CONFIG_SYS_DPAA_QBMAN) += fsl_portals.o
-obj-$(CONFIG_GDSYS_RXAUI_CTRL) += gdsys_rxaui_ctrl.o
+obj-$(CONFIG_TEGRA186_BPMP) += tegra186_bpmp.o
+obj-$(CONFIG_TEGRA_CAR) += tegra_car.o
+obj-$(CONFIG_TWL4030_LED) += twl4030_led.o
+obj-$(CONFIG_WINBOND_W83627) += winbond_w83627.o
Signature · 1 line
2.14.1
#26 lore raw

PATCH v5 11/13 misc: Sort Makefile entries

Quoted 13 lines · On Mon, 6 Aug 2018 16:18:47 +0200 Anatolij Gustschin agust at denx.de wrote:

From: Mario Six <[email protected]>

Makefile entries should be sorted.

---
v4 -> v5:
Drop re-ordered entries that are not in mainline (e.g. CONFIG_GDSYS_IOEP, CONFIG_MPC83XX_SERDES, misc_sandbox.o)

Quoted 8 lines

v3 -> v4:
No changes

v2 -> v3:
New in v3

 drivers/misc/Makefile | 56 +++++++++++++++++++++++++++------------------------
 1 file changed, 30 insertions(+), 26 deletions(-)

Applied to u-boot-staging/agust at denx.de, thanks!

--
Anatolij

#27 lore raw

PATCH v4 11/13 misc: Sort Makefile entries

In reply to Mario Six

Hi Mario,

On Fri, 3 Aug 2018 10:01:16 +0200
Mario Six mario.six at gdsys.cc wrote:
...

Quoted 3 lines
---
 drivers/misc/Makefile | 60 +++++++++++++++++++++++++++------------------------
 1 file changed, 32 insertions(+), 28 deletions(-)

this one doesn't apply on git/master, I've sent rebased patch v5.

--
Anatolij

author
#28 lore raw

PATCH v4 12/13 misc: Add gdsys_soc driver

In reply to Mario Six

This patch adds a driver for the bus associated with a IHS FPGA.

---

v3 -> v4:
No changes

v2 -> v3:
* Fixed style violations
* Added bindings file
* Added more debug output in case of errors
* Switched all printfs to debug
* Documented the private data structure
* Formatted documentation as proper kernel-doc
* Expanded Kconfig description

v1 -> v2:
* Switched to correct uclass for IHS FPGA driver (now in MISC uclass)

---
 .../devicetree/bindings/misc/gdsys,soc.txt         | 16 +++++
 drivers/misc/Kconfig                               |  9 +++
 drivers/misc/Makefile                              |  1 +
 drivers/misc/gdsys_soc.c                           | 74 ++++++++++++++++++++++
 drivers/misc/gdsys_soc.h                           | 23 +++++++
 5 files changed, 123 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/misc/gdsys,soc.txt
 create mode 100644 drivers/misc/gdsys_soc.c
 create mode 100644 drivers/misc/gdsys_soc.h
Documentation/devicetree/bindings/misc/gdsys,soc.txt new file +16 -0
new file mode 100644
index 00000000000..278e935b166
@@ -0,0 +1,16 @@
+gdsys soc bus driver
+
+This driver provides a simple interface for the busses associated with gdsys
+IHS FPGAs. The bus itself contains devices whose register maps are contained
+within the FPGA's register space.
+
+Required properties:
+- fpga: A phandle to the controlling IHS FPGA
+
+Example:
+
+FPGA0BUS: fpga0bus {
+	compatible = "gdsys,soc";
+	ranges = <0x0 0xe0600000 0x00004000>;
+	fpga = <&FPGA0>;
+};
drivers/misc/Kconfig +9 -0
index 89854a4941b..535be1491e3 100644
@@ -289,4 +289,13 @@ config GDSYS_IOEP
 	depends on MISC
 	help
 	  Support gdsys FPGA's IO endpoint driver.
+
+config GDSYS_SOC
+	bool "Enable gdsys SOC driver"
+	depends on MISC
+	help
+	  Support for gdsys IHS SOC, a simple bus associated with each gdsys
+	  IHS (Integrated Hardware Systems) FPGA, which holds all devices whose
+	  register maps are contained within the FPGA's register map.
+
 endmenu
drivers/misc/Makefile +1 -0
index 3ab110a1aaa..b1f9ce7ed39 100644
@@ -37,6 +37,7 @@ obj-$(CONFIG_FSL_MC9SDZ60) += mc9sdz60.o
 obj-$(CONFIG_FSL_SEC_MON) += fsl_sec_mon.o
 obj-$(CONFIG_GDSYS_IOEP) += gdsys_ioep.o
 obj-$(CONFIG_GDSYS_RXAUI_CTRL) += gdsys_rxaui_ctrl.o
+obj-$(CONFIG_GDSYS_SOC) += gdsys_soc.o
 obj-$(CONFIG_$(SPL_)I2C_EEPROM) += i2c_eeprom.o
 obj-$(CONFIG_LED_STATUS) += status_led.o
 obj-$(CONFIG_LED_STATUS_GPIO) += gpio_led.o
drivers/misc/gdsys_soc.c new file +74 -0

74 lines not on this page · raw message

drivers/misc/gdsys_soc.h new file +23 -0
new file mode 100644
index 00000000000..088d3b65234
@@ -0,0 +1,23 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+/*
+ * (C) Copyright 2017
+ * Mario Six,  Guntermann & Drunck GmbH, mario.six at gdsys.cc
+ */
+
+#ifndef _GDSYS_SOC_H_
+#define _GDSYS_SOC_H_
+
+/**
+ * gdsys_soc_get_fpga() - Retrieve pointer to parent bus' FPGA device
+ * @child:	The child device on the FPGA bus needing access to the FPGA.
+ * @fpga:	Pointer to the retrieved FPGA device.
+ *
+ * To access their register maps, devices on gdsys soc buses usually have
+ * facilitate the accessor function of the IHS FPGA their parent bus is
+ * attached to. To access the FPGA device from within the bus' children, this
+ * function returns a pointer to it.
+ *
+ * Return: 0 on success, -ve on failure
+ */
+int gdsys_soc_get_fpga(struct udevice *child, struct udevice **fpga);
+#endif /* _GDSYS_SOC_H_ */

--
2.11.0

#29 lore raw

PATCH v4 12/13 misc: Add gdsys_soc driver

On Fri, 3 Aug 2018 10:01:17 +0200
Mario Six mario.six at gdsys.cc wrote:
...

Quoted 10 lines
---
 .../devicetree/bindings/misc/gdsys,soc.txt         | 16 +++++
 drivers/misc/Kconfig                               |  9 +++
 drivers/misc/Makefile                              |  1 +
 drivers/misc/gdsys_soc.c                           | 74 ++++++++++++++++++++++
 drivers/misc/gdsys_soc.h                           | 23 +++++++
 5 files changed, 123 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/misc/gdsys,soc.txt
 create mode 100644 drivers/misc/gdsys_soc.c
 create mode 100644 drivers/misc/gdsys_soc.h

Applied to u-boot-staging/agust at denx.de, thanks!

--
Anatolij

author
#30 lore raw

PATCH v4 13/13 misc: Add IHS FPGA driver

In reply to Mario Six

Add a driver for gdsys IHS (Integrated Hardware Systems) FPGAs, which supports initialization of the FPGA, as well as information gathering.

---

v3 -> v4:
* Switched from 'res' as the name for return variables to 'ret'

v2 -> v3:
* Fixed style violations
* Added full documentation
* Extracted some magic numbers to constants
* Removed unnecessary includes
* Extracted wait_for_fpga_done
* Improved error handling and reporting
* Added device-tree-binding files
* Improved Kconfig entry

v1 -> v2:
New in v2

---
 .../devicetree/bindings/misc/gdsys,iocon_fpga.txt  |  19 +
 .../devicetree/bindings/misc/gdsys,iocpu_fpga.txt  |  19 +
 drivers/misc/Kconfig                               |   9 +
 drivers/misc/Makefile                              |   1 +
 drivers/misc/ihs_fpga.c                            | 867 +++++++++++++++++++++
 drivers/misc/ihs_fpga.h                            |  49 ++
 6 files changed, 964 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/misc/gdsys,iocon_fpga.txt
 create mode 100644 Documentation/devicetree/bindings/misc/gdsys,iocpu_fpga.txt
 create mode 100644 drivers/misc/ihs_fpga.c
 create mode 100644 drivers/misc/ihs_fpga.h
Documentation/devicetree/bindings/misc/gdsys,iocon_fpga.txt new file +19 -0
new file mode 100644
index 00000000000..acd466fdc6d
@@ -0,0 +1,19 @@
+gdsys IHS FPGA for CON devices
+
+The gdsys IHS FPGA is the main FPGA on gdsys CON devices. This driver provides
+support for enabling and starting the FPGA, as well as verifying working bus
+communication.
+
+Required properties:
+- compatible: must be "gdsys,iocon_fpga"
+- reset-gpios: List of GPIOs controlling the FPGA's reset
+- done-gpios: List of GPIOs notifying whether the FPGA's reconfiguration is
+              done
+
+Example:
+
+FPGA0 {
+	compatible = "gdsys,iocon_fpga";
+	reset-gpios = <&PPCPCA 26 0>;
+	done-gpios = <&GPIO_VB0 19 0>;
+};
Documentation/devicetree/bindings/misc/gdsys,iocpu_fpga.txt new file +19 -0
new file mode 100644
index 00000000000..819db22bf7d
@@ -0,0 +1,19 @@
+gdsys IHS FPGA for CPU devices
+
+The gdsys IHS FPGA is the main FPGA on gdsys CPU devices. This driver provides
+support for enabling and starting the FPGA, as well as verifying working bus
+communication.
+
+Required properties:
+- compatible: must be "gdsys,iocpu_fpga"
+- reset-gpios: List of GPIOs controlling the FPGA's reset
+- done-gpios: List of GPIOs notifying whether the FPGA's reconfiguration is
+              done
+
+Example:
+
+FPGA0 {
+	compatible = "gdsys,iocpu_fpga";
+	reset-gpios = <&PPCPCA 26 0>;
+	done-gpios = <&GPIO_VB0 19 0>;
+};
drivers/misc/Kconfig +9 -0
index 535be1491e3..09d3a6d75ea 100644
@@ -298,4 +298,13 @@ config GDSYS_SOC
 	  IHS (Integrated Hardware Systems) FPGA, which holds all devices whose
 	  register maps are contained within the FPGA's register map.
 
+config IHS_FPGA
+	bool "Enable IHS FPGA driver"
+	depends on MISC
+	help
+	  Support IHS (Integrated Hardware Systems) FPGA, the main FPGAs on
+	  gdsys devices, which supply the majority of the functionality offered
+	  by the devices. This driver supports both CON and CPU variants of the
+	  devices, depending on the device tree entry.
+
 endmenu
drivers/misc/Makefile +1 -0
index b1f9ce7ed39..710475c9419 100644
@@ -39,6 +39,7 @@ obj-$(CONFIG_GDSYS_IOEP) += gdsys_ioep.o
 obj-$(CONFIG_GDSYS_RXAUI_CTRL) += gdsys_rxaui_ctrl.o
 obj-$(CONFIG_GDSYS_SOC) += gdsys_soc.o
 obj-$(CONFIG_$(SPL_)I2C_EEPROM) += i2c_eeprom.o
+obj-$(CONFIG_IHS_FPGA) += ihs_fpga.o
 obj-$(CONFIG_LED_STATUS) += status_led.o
 obj-$(CONFIG_LED_STATUS_GPIO) += gpio_led.o
 obj-$(CONFIG_MPC83XX_SERDES) += mpc83xx_serdes.o
drivers/misc/ihs_fpga.c new file +867 -0

867 lines not on this page · raw message

drivers/misc/ihs_fpga.h new file +49 -0

49 lines not on this page · raw message

--
2.11.0

#31 lore raw

PATCH v4 13/13 misc: Add IHS FPGA driver

On Fri, 3 Aug 2018 10:01:18 +0200
Mario Six mario.six at gdsys.cc wrote:
...

Quoted 4 lines
---
 .../devicetree/bindings/misc/gdsys,iocon_fpga.txt  |  19 +
 .../devicetree/bindings/misc/gdsys,iocpu_fpga.txt  |  19 +
 drivers/misc/Kconfig                               |   9 +
Quoted 8 lines
 drivers/misc/Makefile                              |   1 +
 drivers/misc/ihs_fpga.c                            | 867 +++++++++++++++++++++
 drivers/misc/ihs_fpga.h                            |  49 ++
 6 files changed, 964 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/misc/gdsys,iocon_fpga.txt
 create mode 100644 Documentation/devicetree/bindings/misc/gdsys,iocpu_fpga.txt
 create mode 100644 drivers/misc/ihs_fpga.c
 create mode 100644 drivers/misc/ihs_fpga.h

Applied to u-boot-staging/agust at denx.de, thanks!

--
Anatolij

#32 lore raw

PATCH v4 1/13 test: regmap: Increase size of syscon0 memory

In reply to Mario Six
Quoted 15 lines · On Fri, 3 Aug 2018 10:01:06 +0200 Mario Six mario.six at gdsys.cc wrote:

The upcoming changes to the regmap interface will contain a proper check for plausibility when reading/writing from/to a register map. To still have the current tests pass, increase the size of the memory region for the syscon0 device, since one of the tests reads and writes beyond this range.

---

v3 -> v4:
No changes

Quoted 6 lines

v2 -> v3:
New in v3

---
 arch/sandbox/dts/test.dts | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Applied to u-boot-staging/agust at denx.de, thanks!

--
Anatolij

#33 lore raw

PATCH v4 1/13 test: regmap: Increase size of syscon0 memory

In reply to Mario Six

Hi Mario,

On Fri, 3 Aug 2018 10:01:06 +0200
Mario Six mario.six at gdsys.cc wrote:
...

arch/sandbox/dts/test.dts line 410

Quoted 11 lines
arch/sandbox/dts/test.dts +1 -1
index e941cea3e5c..47cc961890f 100644
@@ -406,7 +406,7 @@
 
 	syscon at 0 {
 		compatible = "sandbox,syscon0";
-		reg = <0x10 4>;
+		reg = <0x10 16>;
 	};

Shouldn't this patch also update the test?

test/dm/regmap.c +1 -1
index 152b3a4b80..a8d7e6829e 100644
@@ -25,7 +25,7 @@ static int dm_test_regmap_base(struct unit_test_state *uts)
        ut_assertok_ptr(map);
        ut_asserteq(1, map->range_count);
        ut_asserteq(0x10, map->ranges[0].start);
-       ut_asserteq(4, map->ranges[0].size);
+       ut_asserteq(16, map->ranges[0].size);
        ut_asserteq(0x10, map_to_sysmem(regmap_get_range(map, 0)));
 
        ut_assertok(uclass_get_device(UCLASS_SYSCON, 1, &dev));

Otherwise I get test errors:

https://travis-ci.org/vdsao/u-boot-video/jobs/412722215

--
Anatolij

author
#34 lore raw

PATCH v4 1/13 test: regmap: Increase size of syscon0 memory

Hi Anatolij,

Quoted 34 lines · Anatolij Gustschin

Hi Mario,

On Fri, 3 Aug 2018 10:01:06 +0200
Mario Six mario.six at gdsys.cc wrote:
...

arch/sandbox/dts/test.dts line 410

Quoted 11 lines
arch/sandbox/dts/test.dts +1 -1
index e941cea3e5c..47cc961890f 100644
@@ -406,7 +406,7 @@
 
      syscon at 0 {
              compatible = "sandbox,syscon0";
-             reg = <0x10 4>;
+             reg = <0x10 16>;
      };

Shouldn't this patch also update the test?

test/dm/regmap.c +1 -1
index 152b3a4b80..a8d7e6829e 100644
@@ -25,7 +25,7 @@ static int dm_test_regmap_base(struct unit_test_state *uts)
        ut_assertok_ptr(map);
        ut_asserteq(1, map->range_count);
        ut_asserteq(0x10, map->ranges[0].start);
-       ut_asserteq(4, map->ranges[0].size);
+       ut_asserteq(16, map->ranges[0].size);
        ut_asserteq(0x10, map_to_sysmem(regmap_get_range(map, 0)));
 
        ut_assertok(uclass_get_device(UCLASS_SYSCON, 1, &dev));
Quoted 4 lines

Otherwise I get test errors:

https://travis-ci.org/vdsao/u-boot-video/jobs/412722215

Indeed. I'll fix that in v6 as well. Thanks for catching that!

Quoted 3 lines

--
Anatolij

Best regards,
Mario