[PATCH net-next 00/17] NXP SJA1105 DSA driver

[PATCH net-next 00/17] NXP SJA1105 DSA driver

[Vladimir Oltean]

This patchset adds a DSA driver for the SPI-managed NXP SJA1105 switch.
Due to the hardware's unfriendliness, most of its state needs to be
shadowed in kernel memory by the driver. To support this and keep a
decent amount of cleanliness in the code, a new generic API for
converting between CPU-accessible ("unpacked") structures and
hardware-accessible ("packed") structures is proposed and used.

Then several small modifications are done to the DSA core, like changing
the order of two calls during initialization, or permitting driver
access to the dp->vlan_filtering property.

These small modifications are done for the greater goal of adding
support for 802.1Q pseudo-switch tagging. The limitations of this type
of tagging are discussed in the commit that adds it, and in the code
comments.

The SJA1105 driver then proceeds to extend this 8021q switch tagging
protocol while adding its own (tag_sja1105). This is done because
SJA1105 needs SPI intervention during transmission of link-local
traffic, which cannot be done from the xmit handler but requires a
deferred worker thread.

The driver is GPL-2.0 licensed. The source code files which are licensed
as BSD-3-Clause are hardware support files and derivative of the
userspace NXP sja1105-tool program, which is BSD-3-Clause licensed.

TODO items:
* Add full support for the P/Q/R/S series. The patches were mostly
  tested on a first-generation T device.
* Add timestamping support and PTP clock manipulation.
* Figure out what the current state of tc-taprio hw offload is, and
  attempt to configure the switch's time-aware scheduler using that.
* Rework link state callbacks to use phylink once the SGMII port
  is supported.

Changes from RFC:
* Removed the packing code for the static configuration tables that were
  not currently used
* Removed the code for unpacking a static configuration structure from
  a memory buffer (not used)
* Completely removed the SGMII stubs, since the configuration is not
  complete anyway.
* Moved some code from the SJA1105 introduction commit into the patch
  that used it.
* Made the code for checking global VLAN filtering generic and made b53
  driver use it.
* Made mt7530 driver use the new generic dp->vlan_filtering
* Fixed check for stringset in .get_sset_count
* Minor cleanup in sja1105_clocking.c
* Fixed a confusing typo in DSA

RFC can be found at:
https://www.mail-archive.com/netdev@vger.kernel.org/msg291717.html

Vladimir Oltean (17):
  lib: Add support for generic packing operations
  net: dsa: Fix pharse -> phase typo
  net: dsa: Store vlan_filtering as a property of dsa_port
  net: dsa: mt7530: Use vlan_filtering property from dsa_port
  net: dsa: Add more convenient functions for installing port VLANs
  net: dsa: Call driver's setup callback after setting up its switchdev
    notifier
  net: dsa: Optional VLAN-based port separation for switches without
    tagging
  net: dsa: Be aware of switches where VLAN filtering is a global
    setting
  net: dsa: b53: Let DSA handle mismatched VLAN filtering settings
  net: dsa: Introduce driver for NXP SJA1105 5-port L2 switch
  net: dsa: sja1105: Add support for FDB and MDB management
  net: dsa: sja1105: Add support for VLAN operations
  net: dsa: sja1105: Add support for ethtool port counters
  net: dsa: sja1105: Add support for traffic through standalone ports
  net: dsa: sja1105: Add support for Spanning Tree Protocol
  Documentation: networking: dsa: Add details about NXP SJA1105 driver
  dt-bindings: net: dsa: Add documentation for NXP SJA1105 driver

 .../devicetree/bindings/net/dsa/sja1105.txt   |  123 ++
 Documentation/networking/dsa/sja1105.txt      |   83 +
 Documentation/packing.txt                     |  150 ++
 MAINTAINERS                                   |   14 +
 drivers/net/dsa/Kconfig                       |    2 +
 drivers/net/dsa/Makefile                      |    1 +
 drivers/net/dsa/b53/b53_common.c              |   25 +-
 drivers/net/dsa/mt7530.c                      |   14 +-
 drivers/net/dsa/mt7530.h                      |    1 -
 drivers/net/dsa/sja1105/Kconfig               |   17 +
 drivers/net/dsa/sja1105/Makefile              |   10 +
 drivers/net/dsa/sja1105/sja1105.h             |  150 ++
 drivers/net/dsa/sja1105/sja1105_clocking.c    |  643 +++++++
 .../net/dsa/sja1105/sja1105_dynamic_config.c  |  518 ++++++
 .../net/dsa/sja1105/sja1105_dynamic_config.h  |   40 +
 drivers/net/dsa/sja1105/sja1105_ethtool.c     |  423 +++++
 drivers/net/dsa/sja1105/sja1105_main.c        | 1547 +++++++++++++++++
 drivers/net/dsa/sja1105/sja1105_spi.c         |  665 +++++++
 .../net/dsa/sja1105/sja1105_static_config.c   | 1074 ++++++++++++
 .../net/dsa/sja1105/sja1105_static_config.h   |  311 ++++
 include/linux/dsa/sja1105.h                   |   52 +
 include/linux/packing.h                       |   49 +
 include/net/dsa.h                             |   11 +
 lib/Makefile                                  |    2 +-
 lib/packing.c                                 |  211 +++
 net/dsa/Kconfig                               |   12 +
 net/dsa/Makefile                              |    2 +
 net/dsa/dsa.c                                 |    6 +
 net/dsa/dsa2.c                                |    8 +-
 net/dsa/dsa_priv.h                            |   15 +
 net/dsa/port.c                                |   75 +-
 net/dsa/slave.c                               |   16 +-
 net/dsa/switch.c                              |    3 +-
 net/dsa/tag_8021q.c                           |  193 ++
 net/dsa/tag_sja1105.c                         |  142 ++
 35 files changed, 6556 insertions(+), 52 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/net/dsa/sja1105.txt
 create mode 100644 Documentation/networking/dsa/sja1105.txt
 create mode 100644 Documentation/packing.txt
 create mode 100644 drivers/net/dsa/sja1105/Kconfig
 create mode 100644 drivers/net/dsa/sja1105/Makefile
 create mode 100644 drivers/net/dsa/sja1105/sja1105.h
 create mode 100644 drivers/net/dsa/sja1105/sja1105_clocking.c
 create mode 100644 drivers/net/dsa/sja1105/sja1105_dynamic_config.c
 create mode 100644 drivers/net/dsa/sja1105/sja1105_dynamic_config.h
 create mode 100644 drivers/net/dsa/sja1105/sja1105_ethtool.c
 create mode 100644 drivers/net/dsa/sja1105/sja1105_main.c
 create mode 100644 drivers/net/dsa/sja1105/sja1105_spi.c
 create mode 100644 drivers/net/dsa/sja1105/sja1105_static_config.c
 create mode 100644 drivers/net/dsa/sja1105/sja1105_static_config.h
 create mode 100644 include/linux/dsa/sja1105.h
 create mode 100644 include/linux/packing.h
 create mode 100644 lib/packing.c
 create mode 100644 net/dsa/tag_8021q.c
 create mode 100644 net/dsa/tag_sja1105.c

-- 
2.17.1

[PATCH net-next 01/17] lib: Add support for generic packing operations

[Vladimir Oltean]

This provides an unified API for accessing register bit fields
regardless of memory layout. The basic unit of data for these API
functions is the u64. The process of transforming an u64 from native CPU
encoding into the peripheral's encoding is called 'pack', and
transforming it from peripheral to native CPU encoding is 'unpack'.

Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
---
 Documentation/packing.txt | 150 +++++++++++++++++++++++++++
 MAINTAINERS               |   8 ++
 include/linux/packing.h   |  49 +++++++++
 lib/Makefile              |   2 +-
 lib/packing.c             | 211 ++++++++++++++++++++++++++++++++++++++
 5 files changed, 419 insertions(+), 1 deletion(-)
 create mode 100644 Documentation/packing.txt
 create mode 100644 include/linux/packing.h
 create mode 100644 lib/packing.c

diff --git a/Documentation/packing.txt b/Documentation/packing.txt
new file mode 100644
index 000000000000..32eba9d23611
--- /dev/null
+++ b/Documentation/packing.txt
@@ -0,0 +1,150 @@
+=============================================
+Generic field packing and unpacking functions
+=============================================
+
+Problem statement
+-----------------
+
+When working with hardware, one has to choose between several approaches of
+interfacing with it.
+One can memory-map a pointer to a carefully crafted struct over the hardware
+device's memory region, and access its fields as struct members (potentially
+declared as bit fields). But writing code this way would make it less portable,
+due to potential endianness mismatches between the CPU and the hardware device.
+Additionally, one has to pay close attention when translating register
+definitions from the hardware documentation into bit field indices for the
+structs. Also, some hardware (typically networking equipment) tends to group
+its register fields in ways that violate any reasonable word boundaries
+(sometimes even 64 bit ones). This creates the inconvenience of having to
+define "high" and "low" portions of register fields within the struct.
+A more robust alternative to struct field definitions would be to extract the
+required fields by shifting the appropriate number of bits. But this would
+still not protect from endianness mismatches, except if all memory accesses
+were performed byte-by-byte. Also the code can easily get cluttered, and the
+high-level idea might get lost among the many bit shifts required.
+Many drivers take the bit-shifting approach and then attempt to reduce the
+clutter with tailored macros, but more often than not these macros take
+shortcuts that still prevent the code from being truly portable.
+
+The solution
+------------
+
+This API deals with 2 basic operations:
+  - Packing a CPU-usable number into a memory buffer (with hardware
+    constraints/quirks)
+  - Unpacking a memory buffer (which has hardware constraints/quirks)
+    into a CPU-usable number.
+
+The API offers an abstraction over said hardware constraints and quirks,
+over CPU endianness and therefore between possible mismatches between
+the two.
+
+The basic unit of these API functions is the u64. From the CPU's
+perspective, bit 63 always means bit offset 7 of byte 7, albeit only
+logically. The question is: where do we lay this bit out in memory?
+
+The following examples cover the memory layout of a packed u64 field.
+The byte offsets in the packed buffer are always implicitly 0, 1, ... 7.
+What the examples show is where the logical bytes and bits sit.
+
+1. Normally (no quirks), we would do it like this:
+
+63 62 61 60 59 58 57 56 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32
+7                       6                       5                        4
+31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10  9  8  7  6  5  4  3  2  1  0
+3                       2                       1                        0
+
+That is, the MSByte (7) of the CPU-usable u64 sits at memory offset 0, and the
+LSByte (0) of the u64 sits at memory offset 7.
+This corresponds to what most folks would regard to as "big endian", where
+bit i corresponds to the number 2^i. This is also referred to in the code
+comments as "logical" notation.
+
+
+2. If QUIRK_MSB_ON_THE_RIGHT is set, we do it like this:
+
+56 57 58 59 60 61 62 63 48 49 50 51 52 53 54 55 40 41 42 43 44 45 46 47 32 33 34 35 36 37 38 39
+7                       6                        5                       4
+24 25 26 27 28 29 30 31 16 17 18 19 20 21 22 23  8  9 10 11 12 13 14 15  0  1  2  3  4  5  6  7
+3                       2                        1                       0
+
+That is, QUIRK_MSB_ON_THE_RIGHT does not affect byte positioning, but
+inverts bit offsets inside a byte.
+
+
+3. If QUIRK_LITTLE_ENDIAN is set, we do it like this:
+
+39 38 37 36 35 34 33 32 47 46 45 44 43 42 41 40 55 54 53 52 51 50 49 48 63 62 61 60 59 58 57 56
+4                       5                       6                       7
+7  6  5  4  3  2  1  0  15 14 13 12 11 10  9  8 23 22 21 20 19 18 17 16 31 30 29 28 27 26 25 24
+0                       1                       2                       3
+
+Therefore, QUIRK_LITTLE_ENDIAN means that inside the memory region, every
+byte from each 4-byte word is placed at its mirrored position compared to
+the boundary of that word.
+
+4. If QUIRK_MSB_ON_THE_RIGHT and QUIRK_LITTLE_ENDIAN are both set, we do it
+   like this:
+
+32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
+4                       5                       6                       7
+0  1  2  3  4  5  6  7  8   9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
+0                       1                       2                       3
+
+
+5. If just QUIRK_LSW32_IS_FIRST is set, we do it like this:
+
+31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10  9  8  7  6  5  4  3  2  1  0
+3                       2                       1                        0
+63 62 61 60 59 58 57 56 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32
+7                       6                       5                        4
+
+In this case the 8 byte memory region is interpreted as follows: first
+4 bytes correspond to the least significant 4-byte word, next 4 bytes to
+the more significant 4-byte word.
+
+
+6. If QUIRK_LSW32_IS_FIRST and QUIRK_MSB_ON_THE_RIGHT are set, we do it like
+   this:
+
+24 25 26 27 28 29 30 31 16 17 18 19 20 21 22 23  8  9 10 11 12 13 14 15  0  1  2  3  4  5  6  7
+3                       2                        1                       0
+56 57 58 59 60 61 62 63 48 49 50 51 52 53 54 55 40 41 42 43 44 45 46 47 32 33 34 35 36 37 38 39
+7                       6                        5                       4
+
+
+7. If QUIRK_LSW32_IS_FIRST and QUIRK_LITTLE_ENDIAN are set, it looks like
+   this:
+
+7  6  5  4  3  2  1  0  15 14 13 12 11 10  9  8 23 22 21 20 19 18 17 16 31 30 29 28 27 26 25 24
+0                       1                       2                       3
+39 38 37 36 35 34 33 32 47 46 45 44 43 42 41 40 55 54 53 52 51 50 49 48 63 62 61 60 59 58 57 56
+4                       5                       6                       7
+
+
+8. If QUIRK_LSW32_IS_FIRST, QUIRK_LITTLE_ENDIAN and QUIRK_MSB_ON_THE_RIGHT
+   are set, it looks like this:
+
+0  1  2  3  4  5  6  7  8   9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
+0                       1                       2                       3
+32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
+4                       5                       6                       7
+
+
+We always think of our offsets as if there were no quirk, and we translate
+them afterwards, before accessing the memory region.
+
+Intended use
+------------
+
+Drivers that opt to use this API first need to identify which of the above 3
+quirk combinations (for a total of 8) match what the hardware documentation
+describes. Then they should wrap the packing() function, creating a new
+xxx_packing() that calls it using the proper QUIRK_* one-hot bits set.
+
+The packing() function returns an int-encoded error code, which protects the
+programmer against incorrect API use.  The errors are not expected to occur
+durring runtime, therefore it is reasonable for xxx_packing() to return void
+and simply swallow those errors. Optionally it can dump stack or print the
+error description.
+
diff --git a/MAINTAINERS b/MAINTAINERS
index f8ff9ae52c21..89315bb1cb83 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -11670,6 +11670,14 @@ L:	linux-i2c@vger.kernel.org
 S:	Orphan
 F:	drivers/i2c/busses/i2c-pasemi.c
 
+PACKING
+M:	Vladimir Oltean <olteanv@gmail.com>
+L:	netdev@vger.kernel.org
+S:	Supported
+F:	lib/packing.c
+F:	include/linux/packing.h
+F:	Documentation/packing.txt
+
 PADATA PARALLEL EXECUTION MECHANISM
 M:	Steffen Klassert <steffen.klassert@secunet.com>
 L:	linux-crypto@vger.kernel.org
diff --git a/include/linux/packing.h b/include/linux/packing.h
new file mode 100644
index 000000000000..cc646e4f5df1
--- /dev/null
+++ b/include/linux/packing.h
@@ -0,0 +1,49 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright (c) 2016-2018, NXP Semiconductors
+ * Copyright (c) 2018-2019, Vladimir Oltean <olteanv@gmail.com>
+ */
+#ifndef _LINUX_PACKING_H
+#define _LINUX_PACKING_H
+
+#include <linux/types.h>
+#include <linux/bitops.h>
+
+#define QUIRK_MSB_ON_THE_RIGHT BIT(0)
+#define QUIRK_LITTLE_ENDIAN    BIT(1)
+#define QUIRK_LSW32_IS_FIRST   BIT(2)
+
+enum packing_op {
+	PACK,
+	UNPACK,
+};
+
+/**
+ * packing - Convert numbers (currently u64) between a packed and an unpacked
+ *	     format. Unpacked means laid out in memory in the CPU's native
+ *	     understanding of integers, while packed means anything else that
+ *	     requires translation.
+ *
+ * @pbuf: Pointer to a buffer holding the packed value.
+ * @uval: Pointer to an u64 holding the unpacked value.
+ * @startbit: The index (in logical notation, compensated for quirks) where
+ *	      the packed value starts within pbuf. Must be larger than, or
+ *	      equal to, endbit.
+ * @endbit: The index (in logical notation, compensated for quirks) where
+ *	    the packed value ends within pbuf. Must be smaller than, or equal
+ *	    to, startbit.
+ * @op: If PACK, then uval will be treated as const pointer and copied (packed)
+ *	into pbuf, between startbit and endbit.
+ *	If UNPACK, then pbuf will be treated as const pointer and the logical
+ *	value between startbit and endbit will be copied (unpacked) to uval.
+ * @quirks: A bit mask of QUIRK_LITTLE_ENDIAN, QUIRK_LSW32_IS_FIRST and
+ *	    QUIRK_MSB_ON_THE_RIGHT.
+ *
+ * Return: 0 on success, EINVAL or ERANGE if called incorrectly. Assuming
+ *	   correct usage, return code may be discarded.
+ *	   If op is PACK, pbuf is modified.
+ *	   If op is UNPACK, uval is modified.
+ */
+int packing(void *pbuf, u64 *uval, int startbit, int endbit, size_t pbuflen,
+	    enum packing_op op, u8 quirks);
+
+#endif
diff --git a/lib/Makefile b/lib/Makefile
index 4e066120a0d6..d5780689f7ef 100644
--- a/lib/Makefile
+++ b/lib/Makefile
@@ -37,7 +37,7 @@ obj-y += bcd.o div64.o sort.o parser.o debug_locks.o random32.o \
 	 bust_spinlocks.o kasprintf.o bitmap.o scatterlist.o \
 	 gcd.o lcm.o list_sort.o uuid.o iov_iter.o clz_ctz.o \
 	 bsearch.o find_bit.o llist.o memweight.o kfifo.o \
-	 percpu-refcount.o rhashtable.o reciprocal_div.o \
+	 packing.o percpu-refcount.o rhashtable.o reciprocal_div.o \
 	 once.o refcount.o usercopy.o errseq.o bucket_locks.o \
 	 generic-radix-tree.o
 obj-$(CONFIG_STRING_SELFTEST) += test_string.o
diff --git a/lib/packing.c b/lib/packing.c
new file mode 100644
index 000000000000..2d0bfd78bfe9
--- /dev/null
+++ b/lib/packing.c
@@ -0,0 +1,211 @@
+// SPDX-License-Identifier: BSD-3-Clause
+/* Copyright (c) 2016-2018, NXP Semiconductors
+ * Copyright (c) 2018-2019, Vladimir Oltean <olteanv@gmail.com>
+ */
+#include <linux/packing.h>
+#include <linux/module.h>
+#include <linux/bitops.h>
+#include <linux/errno.h>
+#include <linux/types.h>
+
+static int get_le_offset(int offset)
+{
+	int closest_multiple_of_4;
+
+	closest_multiple_of_4 = (offset / 4) * 4;
+	offset -= closest_multiple_of_4;
+	return closest_multiple_of_4 + (3 - offset);
+}
+
+static int get_reverse_lsw32_offset(int offset, size_t len)
+{
+	int closest_multiple_of_4;
+	int word_index;
+
+	word_index = offset / 4;
+	closest_multiple_of_4 = word_index * 4;
+	offset -= closest_multiple_of_4;
+	word_index = (len / 4) - word_index - 1;
+	return word_index * 4 + offset;
+}
+
+static u64 bit_reverse(u64 val, unsigned int width)
+{
+	u64 new_val = 0;
+	unsigned int bit;
+	unsigned int i;
+
+	for (i = 0; i < width; i++) {
+		bit = (val & (1 << i)) != 0;
+		new_val |= (bit << (width - i - 1));
+	}
+	return new_val;
+}
+
+static void adjust_for_msb_right_quirk(u64 *to_write, int *box_start_bit,
+				       int *box_end_bit, u8 *box_mask)
+{
+	int box_bit_width = *box_start_bit - *box_end_bit + 1;
+	int new_box_start_bit, new_box_end_bit;
+
+	*to_write >>= *box_end_bit;
+	*to_write = bit_reverse(*to_write, box_bit_width);
+	*to_write <<= *box_end_bit;
+
+	new_box_end_bit   = box_bit_width - *box_start_bit - 1;
+	new_box_start_bit = box_bit_width - *box_end_bit - 1;
+	*box_mask = GENMASK_ULL(new_box_start_bit, new_box_end_bit);
+	*box_start_bit = new_box_start_bit;
+	*box_end_bit   = new_box_end_bit;
+}
+
+/**
+ * packing - Convert numbers (currently u64) between a packed and an unpacked
+ *	     format. Unpacked means laid out in memory in the CPU's native
+ *	     understanding of integers, while packed means anything else that
+ *	     requires translation.
+ *
+ * @pbuf: Pointer to a buffer holding the packed value.
+ * @uval: Pointer to an u64 holding the unpacked value.
+ * @startbit: The index (in logical notation, compensated for quirks) where
+ *	      the packed value starts within pbuf. Must be larger than, or
+ *	      equal to, endbit.
+ * @endbit: The index (in logical notation, compensated for quirks) where
+ *	    the packed value ends within pbuf. Must be smaller than, or equal
+ *	    to, startbit.
+ * @op: If PACK, then uval will be treated as const pointer and copied (packed)
+ *	into pbuf, between startbit and endbit.
+ *	If UNPACK, then pbuf will be treated as const pointer and the logical
+ *	value between startbit and endbit will be copied (unpacked) to uval.
+ * @quirks: A bit mask of QUIRK_LITTLE_ENDIAN, QUIRK_LSW32_IS_FIRST and
+ *	    QUIRK_MSB_ON_THE_RIGHT.
+ *
+ * Return: 0 on success, EINVAL or ERANGE if called incorrectly. Assuming
+ *	   correct usage, return code may be discarded.
+ *	   If op is PACK, pbuf is modified.
+ *	   If op is UNPACK, uval is modified.
+ */
+int packing(void *pbuf, u64 *uval, int startbit, int endbit, size_t pbuflen,
+	    enum packing_op op, u8 quirks)
+{
+	/* Number of bits for storing "uval"
+	 * also width of the field to access in the pbuf
+	 */
+	u64 value_width;
+	/* Logical byte indices corresponding to the
+	 * start and end of the field.
+	 */
+	int plogical_first_u8, plogical_last_u8, box;
+
+	/* startbit is expected to be larger than endbit */
+	if (startbit < endbit)
+		/* Invalid function call */
+		return -EINVAL;
+
+	value_width = startbit - endbit + 1;
+	if (value_width > 64)
+		return -ERANGE;
+
+	/* Check if "uval" fits in "value_width" bits.
+	 * If value_width is 64, the check will fail, but any
+	 * 64-bit uval will surely fit.
+	 */
+	if (op == PACK && value_width < 64 && (*uval >= (1ull << value_width)))
+		/* Cannot store "uval" inside "value_width" bits.
+		 * Truncating "uval" is most certainly not desirable,
+		 * so simply erroring out is appropriate.
+		 */
+		return -ERANGE;
+
+	/* Initialize parameter */
+	if (op == UNPACK)
+		*uval = 0;
+
+	/* Iterate through an idealistic view of the pbuf as an u64 with
+	 * no quirks, u8 by u8 (aligned at u8 boundaries), from high to low
+	 * logical bit significance. "box" denotes the current logical u8.
+	 */
+	plogical_first_u8 = startbit / 8;
+	plogical_last_u8  = endbit / 8;
+
+	for (box = plogical_first_u8; box >= plogical_last_u8; box--) {
+		/* Bit indices into the currently accessed 8-bit box */
+		int box_start_bit, box_end_bit, box_addr;
+		u8  box_mask;
+		/* Corresponding bits from the unpacked u64 parameter */
+		int proj_start_bit, proj_end_bit;
+		u64 proj_mask;
+
+		/* This u8 may need to be accessed in its entirety
+		 * (from bit 7 to bit 0), or not, depending on the
+		 * input arguments startbit and endbit.
+		 */
+		if (box == plogical_first_u8)
+			box_start_bit = startbit % 8;
+		else
+			box_start_bit = 7;
+		if (box == plogical_last_u8)
+			box_end_bit = endbit % 8;
+		else
+			box_end_bit = 0;
+
+		/* We have determined the box bit start and end.
+		 * Now we calculate where this (masked) u8 box would fit
+		 * in the unpacked (CPU-readable) u64 - the u8 box's
+		 * projection onto the unpacked u64. Though the
+		 * box is u8, the projection is u64 because it may fall
+		 * anywhere within the unpacked u64.
+		 */
+		proj_start_bit = ((box * 8) + box_start_bit) - endbit;
+		proj_end_bit   = ((box * 8) + box_end_bit) - endbit;
+		proj_mask = GENMASK_ULL(proj_start_bit, proj_end_bit);
+		box_mask  = GENMASK_ULL(box_start_bit, box_end_bit);
+
+		/* Determine the offset of the u8 box inside the pbuf,
+		 * adjusted for quirks. The adjusted box_addr will be used for
+		 * effective addressing inside the pbuf (so it's not
+		 * logical any longer).
+		 */
+		box_addr = pbuflen - box - 1;
+		if (quirks & QUIRK_LITTLE_ENDIAN)
+			box_addr = get_le_offset(box_addr);
+		if (quirks & QUIRK_LSW32_IS_FIRST)
+			box_addr = get_reverse_lsw32_offset(box_addr,
+							    pbuflen);
+
+		if (op == UNPACK) {
+			u64 pval;
+
+			/* Read from pbuf, write to uval */
+			pval = ((u8 *)pbuf)[box_addr] & box_mask;
+			if (quirks & QUIRK_MSB_ON_THE_RIGHT)
+				adjust_for_msb_right_quirk(&pval,
+							   &box_start_bit,
+							   &box_end_bit,
+							   &box_mask);
+
+			pval >>= box_end_bit;
+			pval <<= proj_end_bit;
+			*uval &= ~proj_mask;
+			*uval |= pval;
+		} else {
+			u64 pval;
+
+			/* Write to pbuf, read from uval */
+			pval = (*uval) & proj_mask;
+			pval >>= proj_end_bit;
+			if (quirks & QUIRK_MSB_ON_THE_RIGHT)
+				adjust_for_msb_right_quirk(&pval,
+							   &box_start_bit,
+							   &box_end_bit,
+							   &box_mask);
+
+			pval <<= box_end_bit;
+			((u8 *)pbuf)[box_addr] &= ~box_mask;
+			((u8 *)pbuf)[box_addr] |= pval;
+		}
+	}
+	return 0;
+}
+EXPORT_SYMBOL(packing);
+
-- 
2.17.1

Re: [PATCH net-next 01/17] lib: Add support for generic packing operations

[Vladimir Oltean]

On 3/31/19 8:42 PM, Vladimir Oltean wrote:
> This provides an unified API for accessing register bit fields
> regardless of memory layout. The basic unit of data for these API
> functions is the u64. The process of transforming an u64 from native CPU
> encoding into the peripheral's encoding is called 'pack', and
> transforming it from peripheral to native CPU encoding is 'unpack'.
> 
> Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
> ---
>   Documentation/packing.txt | 150 +++++++++++++++++++++++++++
>   MAINTAINERS               |   8 ++
>   include/linux/packing.h   |  49 +++++++++
>   lib/Makefile              |   2 +-
>   lib/packing.c             | 211 ++++++++++++++++++++++++++++++++++++++
>   5 files changed, 419 insertions(+), 1 deletion(-)
>   create mode 100644 Documentation/packing.txt
>   create mode 100644 include/linux/packing.h
>   create mode 100644 lib/packing.c
> 
> diff --git a/Documentation/packing.txt b/Documentation/packing.txt
> new file mode 100644
> index 000000000000..32eba9d23611
> --- /dev/null
> +++ b/Documentation/packing.txt
> @@ -0,0 +1,150 @@
> +=============================================
> +Generic field packing and unpacking functions
> +=============================================
> +
> +Problem statement
> +-----------------
> +
> +When working with hardware, one has to choose between several approaches of
> +interfacing with it.
> +One can memory-map a pointer to a carefully crafted struct over the hardware
> +device's memory region, and access its fields as struct members (potentially
> +declared as bit fields). But writing code this way would make it less portable,
> +due to potential endianness mismatches between the CPU and the hardware device.
> +Additionally, one has to pay close attention when translating register
> +definitions from the hardware documentation into bit field indices for the
> +structs. Also, some hardware (typically networking equipment) tends to group
> +its register fields in ways that violate any reasonable word boundaries
> +(sometimes even 64 bit ones). This creates the inconvenience of having to
> +define "high" and "low" portions of register fields within the struct.
> +A more robust alternative to struct field definitions would be to extract the
> +required fields by shifting the appropriate number of bits. But this would
> +still not protect from endianness mismatches, except if all memory accesses
> +were performed byte-by-byte. Also the code can easily get cluttered, and the
> +high-level idea might get lost among the many bit shifts required.
> +Many drivers take the bit-shifting approach and then attempt to reduce the
> +clutter with tailored macros, but more often than not these macros take
> +shortcuts that still prevent the code from being truly portable.
> +
> +The solution
> +------------
> +
> +This API deals with 2 basic operations:
> +  - Packing a CPU-usable number into a memory buffer (with hardware
> +    constraints/quirks)
> +  - Unpacking a memory buffer (which has hardware constraints/quirks)
> +    into a CPU-usable number.
> +
> +The API offers an abstraction over said hardware constraints and quirks,
> +over CPU endianness and therefore between possible mismatches between
> +the two.
> +
> +The basic unit of these API functions is the u64. From the CPU's
> +perspective, bit 63 always means bit offset 7 of byte 7, albeit only
> +logically. The question is: where do we lay this bit out in memory?
> +
> +The following examples cover the memory layout of a packed u64 field.
> +The byte offsets in the packed buffer are always implicitly 0, 1, ... 7.
> +What the examples show is where the logical bytes and bits sit.
> +
> +1. Normally (no quirks), we would do it like this:
> +
> +63 62 61 60 59 58 57 56 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32
> +7                       6                       5                        4
> +31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10  9  8  7  6  5  4  3  2  1  0
> +3                       2                       1                        0
> +
> +That is, the MSByte (7) of the CPU-usable u64 sits at memory offset 0, and the
> +LSByte (0) of the u64 sits at memory offset 7.
> +This corresponds to what most folks would regard to as "big endian", where
> +bit i corresponds to the number 2^i. This is also referred to in the code
> +comments as "logical" notation.
> +
> +
> +2. If QUIRK_MSB_ON_THE_RIGHT is set, we do it like this:
> +
> +56 57 58 59 60 61 62 63 48 49 50 51 52 53 54 55 40 41 42 43 44 45 46 47 32 33 34 35 36 37 38 39
> +7                       6                        5                       4
> +24 25 26 27 28 29 30 31 16 17 18 19 20 21 22 23  8  9 10 11 12 13 14 15  0  1  2  3  4  5  6  7
> +3                       2                        1                       0
> +
> +That is, QUIRK_MSB_ON_THE_RIGHT does not affect byte positioning, but
> +inverts bit offsets inside a byte.
> +
> +
> +3. If QUIRK_LITTLE_ENDIAN is set, we do it like this:
> +
> +39 38 37 36 35 34 33 32 47 46 45 44 43 42 41 40 55 54 53 52 51 50 49 48 63 62 61 60 59 58 57 56
> +4                       5                       6                       7
> +7  6  5  4  3  2  1  0  15 14 13 12 11 10  9  8 23 22 21 20 19 18 17 16 31 30 29 28 27 26 25 24
> +0                       1                       2                       3
> +
> +Therefore, QUIRK_LITTLE_ENDIAN means that inside the memory region, every
> +byte from each 4-byte word is placed at its mirrored position compared to
> +the boundary of that word.
> +
> +4. If QUIRK_MSB_ON_THE_RIGHT and QUIRK_LITTLE_ENDIAN are both set, we do it
> +   like this:
> +
> +32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
> +4                       5                       6                       7
> +0  1  2  3  4  5  6  7  8   9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
> +0                       1                       2                       3
> +
> +
> +5. If just QUIRK_LSW32_IS_FIRST is set, we do it like this:
> +
> +31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10  9  8  7  6  5  4  3  2  1  0
> +3                       2                       1                        0
> +63 62 61 60 59 58 57 56 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32
> +7                       6                       5                        4
> +
> +In this case the 8 byte memory region is interpreted as follows: first
> +4 bytes correspond to the least significant 4-byte word, next 4 bytes to
> +the more significant 4-byte word.
> +
> +
> +6. If QUIRK_LSW32_IS_FIRST and QUIRK_MSB_ON_THE_RIGHT are set, we do it like
> +   this:
> +
> +24 25 26 27 28 29 30 31 16 17 18 19 20 21 22 23  8  9 10 11 12 13 14 15  0  1  2  3  4  5  6  7
> +3                       2                        1                       0
> +56 57 58 59 60 61 62 63 48 49 50 51 52 53 54 55 40 41 42 43 44 45 46 47 32 33 34 35 36 37 38 39
> +7                       6                        5                       4
> +
> +
> +7. If QUIRK_LSW32_IS_FIRST and QUIRK_LITTLE_ENDIAN are set, it looks like
> +   this:
> +
> +7  6  5  4  3  2  1  0  15 14 13 12 11 10  9  8 23 22 21 20 19 18 17 16 31 30 29 28 27 26 25 24
> +0                       1                       2                       3
> +39 38 37 36 35 34 33 32 47 46 45 44 43 42 41 40 55 54 53 52 51 50 49 48 63 62 61 60 59 58 57 56
> +4                       5                       6                       7
> +
> +
> +8. If QUIRK_LSW32_IS_FIRST, QUIRK_LITTLE_ENDIAN and QUIRK_MSB_ON_THE_RIGHT
> +   are set, it looks like this:
> +
> +0  1  2  3  4  5  6  7  8   9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
> +0                       1                       2                       3
> +32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
> +4                       5                       6                       7
> +
> +
> +We always think of our offsets as if there were no quirk, and we translate
> +them afterwards, before accessing the memory region.
> +
> +Intended use
> +------------
> +
> +Drivers that opt to use this API first need to identify which of the above 3
> +quirk combinations (for a total of 8) match what the hardware documentation
> +describes. Then they should wrap the packing() function, creating a new
> +xxx_packing() that calls it using the proper QUIRK_* one-hot bits set.
> +
> +The packing() function returns an int-encoded error code, which protects the
> +programmer against incorrect API use.  The errors are not expected to occur
> +durring runtime, therefore it is reasonable for xxx_packing() to return void
> +and simply swallow those errors. Optionally it can dump stack or print the
> +error description.
> +
> diff --git a/MAINTAINERS b/MAINTAINERS
> index f8ff9ae52c21..89315bb1cb83 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -11670,6 +11670,14 @@ L:	linux-i2c@vger.kernel.org
>   S:	Orphan
>   F:	drivers/i2c/busses/i2c-pasemi.c
>   
> +PACKING
> +M:	Vladimir Oltean <olteanv@gmail.com>
> +L:	netdev@vger.kernel.org
> +S:	Supported
> +F:	lib/packing.c
> +F:	include/linux/packing.h
> +F:	Documentation/packing.txt
> +
>   PADATA PARALLEL EXECUTION MECHANISM
>   M:	Steffen Klassert <steffen.klassert@secunet.com>
>   L:	linux-crypto@vger.kernel.org
> diff --git a/include/linux/packing.h b/include/linux/packing.h
> new file mode 100644
> index 000000000000..cc646e4f5df1
> --- /dev/null
> +++ b/include/linux/packing.h
> @@ -0,0 +1,49 @@
> +/* SPDX-License-Identifier: BSD-3-Clause
> + * Copyright (c) 2016-2018, NXP Semiconductors
> + * Copyright (c) 2018-2019, Vladimir Oltean <olteanv@gmail.com>
> + */
> +#ifndef _LINUX_PACKING_H
> +#define _LINUX_PACKING_H
> +
> +#include <linux/types.h>
> +#include <linux/bitops.h>
> +
> +#define QUIRK_MSB_ON_THE_RIGHT BIT(0)
> +#define QUIRK_LITTLE_ENDIAN    BIT(1)
> +#define QUIRK_LSW32_IS_FIRST   BIT(2)
> +
> +enum packing_op {
> +	PACK,
> +	UNPACK,
> +};
> +
> +/**
> + * packing - Convert numbers (currently u64) between a packed and an unpacked
> + *	     format. Unpacked means laid out in memory in the CPU's native
> + *	     understanding of integers, while packed means anything else that
> + *	     requires translation.
> + *
> + * @pbuf: Pointer to a buffer holding the packed value.
> + * @uval: Pointer to an u64 holding the unpacked value.
> + * @startbit: The index (in logical notation, compensated for quirks) where
> + *	      the packed value starts within pbuf. Must be larger than, or
> + *	      equal to, endbit.
> + * @endbit: The index (in logical notation, compensated for quirks) where
> + *	    the packed value ends within pbuf. Must be smaller than, or equal
> + *	    to, startbit.
> + * @op: If PACK, then uval will be treated as const pointer and copied (packed)
> + *	into pbuf, between startbit and endbit.
> + *	If UNPACK, then pbuf will be treated as const pointer and the logical
> + *	value between startbit and endbit will be copied (unpacked) to uval.
> + * @quirks: A bit mask of QUIRK_LITTLE_ENDIAN, QUIRK_LSW32_IS_FIRST and
> + *	    QUIRK_MSB_ON_THE_RIGHT.
> + *
> + * Return: 0 on success, EINVAL or ERANGE if called incorrectly. Assuming
> + *	   correct usage, return code may be discarded.
> + *	   If op is PACK, pbuf is modified.
> + *	   If op is UNPACK, uval is modified.
> + */
> +int packing(void *pbuf, u64 *uval, int startbit, int endbit, size_t pbuflen,
> +	    enum packing_op op, u8 quirks);
> +
> +#endif
> diff --git a/lib/Makefile b/lib/Makefile
> index 4e066120a0d6..d5780689f7ef 100644
> --- a/lib/Makefile
> +++ b/lib/Makefile
> @@ -37,7 +37,7 @@ obj-y += bcd.o div64.o sort.o parser.o debug_locks.o random32.o \
>   	 bust_spinlocks.o kasprintf.o bitmap.o scatterlist.o \
>   	 gcd.o lcm.o list_sort.o uuid.o iov_iter.o clz_ctz.o \
>   	 bsearch.o find_bit.o llist.o memweight.o kfifo.o \
> -	 percpu-refcount.o rhashtable.o reciprocal_div.o \
> +	 packing.o percpu-refcount.o rhashtable.o reciprocal_div.o \
>   	 once.o refcount.o usercopy.o errseq.o bucket_locks.o \
>   	 generic-radix-tree.o
>   obj-$(CONFIG_STRING_SELFTEST) += test_string.o
> diff --git a/lib/packing.c b/lib/packing.c
> new file mode 100644
> index 000000000000..2d0bfd78bfe9
> --- /dev/null
> +++ b/lib/packing.c
> @@ -0,0 +1,211 @@
> +// SPDX-License-Identifier: BSD-3-Clause
> +/* Copyright (c) 2016-2018, NXP Semiconductors
> + * Copyright (c) 2018-2019, Vladimir Oltean <olteanv@gmail.com>
> + */
> +#include <linux/packing.h>
> +#include <linux/module.h>
> +#include <linux/bitops.h>
> +#include <linux/errno.h>
> +#include <linux/types.h>
> +
> +static int get_le_offset(int offset)
> +{
> +	int closest_multiple_of_4;
> +
> +	closest_multiple_of_4 = (offset / 4) * 4;
> +	offset -= closest_multiple_of_4;
> +	return closest_multiple_of_4 + (3 - offset);
> +}
> +
> +static int get_reverse_lsw32_offset(int offset, size_t len)
> +{
> +	int closest_multiple_of_4;
> +	int word_index;
> +
> +	word_index = offset / 4;
> +	closest_multiple_of_4 = word_index * 4;
> +	offset -= closest_multiple_of_4;
> +	word_index = (len / 4) - word_index - 1;
> +	return word_index * 4 + offset;
> +}
> +
> +static u64 bit_reverse(u64 val, unsigned int width)
> +{
> +	u64 new_val = 0;
> +	unsigned int bit;
> +	unsigned int i;
> +
> +	for (i = 0; i < width; i++) {
> +		bit = (val & (1 << i)) != 0;
> +		new_val |= (bit << (width - i - 1));
> +	}
> +	return new_val;
> +}
> +
> +static void adjust_for_msb_right_quirk(u64 *to_write, int *box_start_bit,
> +				       int *box_end_bit, u8 *box_mask)
> +{
> +	int box_bit_width = *box_start_bit - *box_end_bit + 1;
> +	int new_box_start_bit, new_box_end_bit;
> +
> +	*to_write >>= *box_end_bit;
> +	*to_write = bit_reverse(*to_write, box_bit_width);
> +	*to_write <<= *box_end_bit;
> +
> +	new_box_end_bit   = box_bit_width - *box_start_bit - 1;
> +	new_box_start_bit = box_bit_width - *box_end_bit - 1;
> +	*box_mask = GENMASK_ULL(new_box_start_bit, new_box_end_bit);
> +	*box_start_bit = new_box_start_bit;
> +	*box_end_bit   = new_box_end_bit;
> +}
> +
> +/**
> + * packing - Convert numbers (currently u64) between a packed and an unpacked
> + *	     format. Unpacked means laid out in memory in the CPU's native
> + *	     understanding of integers, while packed means anything else that
> + *	     requires translation.
> + *
> + * @pbuf: Pointer to a buffer holding the packed value.
> + * @uval: Pointer to an u64 holding the unpacked value.
> + * @startbit: The index (in logical notation, compensated for quirks) where
> + *	      the packed value starts within pbuf. Must be larger than, or
> + *	      equal to, endbit.
> + * @endbit: The index (in logical notation, compensated for quirks) where
> + *	    the packed value ends within pbuf. Must be smaller than, or equal
> + *	    to, startbit.
> + * @op: If PACK, then uval will be treated as const pointer and copied (packed)
> + *	into pbuf, between startbit and endbit.
> + *	If UNPACK, then pbuf will be treated as const pointer and the logical
> + *	value between startbit and endbit will be copied (unpacked) to uval.
> + * @quirks: A bit mask of QUIRK_LITTLE_ENDIAN, QUIRK_LSW32_IS_FIRST and
> + *	    QUIRK_MSB_ON_THE_RIGHT.
> + *
> + * Return: 0 on success, EINVAL or ERANGE if called incorrectly. Assuming
> + *	   correct usage, return code may be discarded.
> + *	   If op is PACK, pbuf is modified.
> + *	   If op is UNPACK, uval is modified.
> + */
> +int packing(void *pbuf, u64 *uval, int startbit, int endbit, size_t pbuflen,
> +	    enum packing_op op, u8 quirks)
> +{
> +	/* Number of bits for storing "uval"
> +	 * also width of the field to access in the pbuf
> +	 */
> +	u64 value_width;
> +	/* Logical byte indices corresponding to the
> +	 * start and end of the field.
> +	 */
> +	int plogical_first_u8, plogical_last_u8, box;
> +
> +	/* startbit is expected to be larger than endbit */
> +	if (startbit < endbit)
> +		/* Invalid function call */
> +		return -EINVAL;
> +
> +	value_width = startbit - endbit + 1;
> +	if (value_width > 64)
> +		return -ERANGE;
> +
> +	/* Check if "uval" fits in "value_width" bits.
> +	 * If value_width is 64, the check will fail, but any
> +	 * 64-bit uval will surely fit.
> +	 */
> +	if (op == PACK && value_width < 64 && (*uval >= (1ull << value_width)))
> +		/* Cannot store "uval" inside "value_width" bits.
> +		 * Truncating "uval" is most certainly not desirable,
> +		 * so simply erroring out is appropriate.
> +		 */
> +		return -ERANGE;
> +
> +	/* Initialize parameter */
> +	if (op == UNPACK)
> +		*uval = 0;
> +
> +	/* Iterate through an idealistic view of the pbuf as an u64 with
> +	 * no quirks, u8 by u8 (aligned at u8 boundaries), from high to low
> +	 * logical bit significance. "box" denotes the current logical u8.
> +	 */
> +	plogical_first_u8 = startbit / 8;
> +	plogical_last_u8  = endbit / 8;
> +
> +	for (box = plogical_first_u8; box >= plogical_last_u8; box--) {
> +		/* Bit indices into the currently accessed 8-bit box */
> +		int box_start_bit, box_end_bit, box_addr;
> +		u8  box_mask;
> +		/* Corresponding bits from the unpacked u64 parameter */
> +		int proj_start_bit, proj_end_bit;
> +		u64 proj_mask;
> +
> +		/* This u8 may need to be accessed in its entirety
> +		 * (from bit 7 to bit 0), or not, depending on the
> +		 * input arguments startbit and endbit.
> +		 */
> +		if (box == plogical_first_u8)
> +			box_start_bit = startbit % 8;
> +		else
> +			box_start_bit = 7;
> +		if (box == plogical_last_u8)
> +			box_end_bit = endbit % 8;
> +		else
> +			box_end_bit = 0;
> +
> +		/* We have determined the box bit start and end.
> +		 * Now we calculate where this (masked) u8 box would fit
> +		 * in the unpacked (CPU-readable) u64 - the u8 box's
> +		 * projection onto the unpacked u64. Though the
> +		 * box is u8, the projection is u64 because it may fall
> +		 * anywhere within the unpacked u64.
> +		 */
> +		proj_start_bit = ((box * 8) + box_start_bit) - endbit;
> +		proj_end_bit   = ((box * 8) + box_end_bit) - endbit;
> +		proj_mask = GENMASK_ULL(proj_start_bit, proj_end_bit);
> +		box_mask  = GENMASK_ULL(box_start_bit, box_end_bit);
> +
> +		/* Determine the offset of the u8 box inside the pbuf,
> +		 * adjusted for quirks. The adjusted box_addr will be used for
> +		 * effective addressing inside the pbuf (so it's not
> +		 * logical any longer).
> +		 */
> +		box_addr = pbuflen - box - 1;
> +		if (quirks & QUIRK_LITTLE_ENDIAN)
> +			box_addr = get_le_offset(box_addr);
> +		if (quirks & QUIRK_LSW32_IS_FIRST)
> +			box_addr = get_reverse_lsw32_offset(box_addr,
> +							    pbuflen);
> +
> +		if (op == UNPACK) {
> +			u64 pval;
> +
> +			/* Read from pbuf, write to uval */
> +			pval = ((u8 *)pbuf)[box_addr] & box_mask;
> +			if (quirks & QUIRK_MSB_ON_THE_RIGHT)
> +				adjust_for_msb_right_quirk(&pval,
> +							   &box_start_bit,
> +							   &box_end_bit,
> +							   &box_mask);
> +
> +			pval >>= box_end_bit;
> +			pval <<= proj_end_bit;
> +			*uval &= ~proj_mask;
> +			*uval |= pval;
> +		} else {
> +			u64 pval;
> +
> +			/* Write to pbuf, read from uval */
> +			pval = (*uval) & proj_mask;
> +			pval >>= proj_end_bit;
> +			if (quirks & QUIRK_MSB_ON_THE_RIGHT)
> +				adjust_for_msb_right_quirk(&pval,
> +							   &box_start_bit,
> +							   &box_end_bit,
> +							   &box_mask);
> +
> +			pval <<= box_end_bit;
> +			((u8 *)pbuf)[box_addr] &= ~box_mask;
> +			((u8 *)pbuf)[box_addr] |= pval;
> +		}
> +	}
> +	return 0;
> +}
> +EXPORT_SYMBOL(packing);
> +
> 

Who would be interested in reviewing this? I am slightly reluctant to 
send another driver patchset without making sure the packing API is in a 
shape that is remotely useful to others too.

Thanks,
-Vladimir

[PATCH net-next 02/17] net: dsa: Fix pharse -> phase typo

[Vladimir Oltean]

Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
---
 net/dsa/switch.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/dsa/switch.c b/net/dsa/switch.c
index e1fae969aa73..fde4e9195709 100644
--- a/net/dsa/switch.c
+++ b/net/dsa/switch.c
@@ -196,7 +196,7 @@ static int dsa_port_vlan_check(struct dsa_switch *ds, int port,
 	if (!dp->bridge_dev)
 		return err;
 
-	/* dsa_slave_vlan_rx_{add,kill}_vid() cannot use the prepare pharse and
+	/* dsa_slave_vlan_rx_{add,kill}_vid() cannot use the prepare phase and
 	 * already checks whether there is an overlapping bridge VLAN entry
 	 * with the same VID, so here we only need to check that if we are
 	 * adding a bridge VLAN entry there is not an overlapping VLAN device
-- 
2.17.1

Re: [PATCH net-next 02/17] net: dsa: Fix pharse -> phase typo

[Andrew Lunn]

On Sun, Mar 31, 2019 at 08:42:17PM +0300, Vladimir Oltean wrote:
> Signed-off-by: Vladimir Oltean <olteanv@gmail.com>

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

[PATCH net-next 03/17] net: dsa: Store vlan_filtering as a property of dsa_port

[Vladimir Oltean]

This allows drivers to query the VLAN setting imposed by the bridge
driver directly from DSA, instead of keeping their own state based on
the .port_vlan_filtering callback.

Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
---
 include/net/dsa.h |  1 +
 net/dsa/port.c    | 12 ++++++++----
 2 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/include/net/dsa.h b/include/net/dsa.h
index ae480bba11f5..a16fd577349b 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -142,6 +142,7 @@ struct dsa_port {
 	const struct dsa_port	*cpu_dp;
 	struct device_node	*dn;
 	unsigned int		ageing_time;
+	bool			vlan_filtering;
 	u8			stp_state;
 	struct net_device	*bridge_dev;
 	struct devlink_port	devlink_port;
diff --git a/net/dsa/port.c b/net/dsa/port.c
index caeef4c99dc0..a86fe3be1261 100644
--- a/net/dsa/port.c
+++ b/net/dsa/port.c
@@ -158,15 +158,19 @@ int dsa_port_vlan_filtering(struct dsa_port *dp, bool vlan_filtering,
 			    struct switchdev_trans *trans)
 {
 	struct dsa_switch *ds = dp->ds;
+	int err;
 
 	/* bridge skips -EOPNOTSUPP, so skip the prepare phase */
 	if (switchdev_trans_ph_prepare(trans))
 		return 0;
 
-	if (ds->ops->port_vlan_filtering)
-		return ds->ops->port_vlan_filtering(ds, dp->index,
-						    vlan_filtering);
-
+	if (ds->ops->port_vlan_filtering) {
+		err = ds->ops->port_vlan_filtering(ds, dp->index,
+						   vlan_filtering);
+		if (err)
+			return err;
+		dp->vlan_filtering = vlan_filtering;
+	}
 	return 0;
 }
 
-- 
2.17.1

[PATCH net-next 04/17] net: dsa: mt7530: Use vlan_filtering property from dsa_port

[Vladimir Oltean]

This field was recently introduced, so keeping state inside the driver
is no longer necessary.

Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Suggested-by: Andrew Lunn <andrew@lunn.ch>
---
 drivers/net/dsa/mt7530.c | 14 ++++----------
 drivers/net/dsa/mt7530.h |  1 -
 2 files changed, 4 insertions(+), 11 deletions(-)

diff --git a/drivers/net/dsa/mt7530.c b/drivers/net/dsa/mt7530.c
index 7357b4fc0185..8bb0837792b1 100644
--- a/drivers/net/dsa/mt7530.c
+++ b/drivers/net/dsa/mt7530.c
@@ -828,11 +828,9 @@ mt7530_port_set_vlan_unaware(struct dsa_switch *ds, int port)
 	mt7530_rmw(priv, MT7530_PVC_P(port), VLAN_ATTR_MASK,
 		   VLAN_ATTR(MT7530_VLAN_TRANSPARENT));
 
-	priv->ports[port].vlan_filtering = false;
-
 	for (i = 0; i < MT7530_NUM_PORTS; i++) {
 		if (dsa_is_user_port(ds, i) &&
-		    priv->ports[i].vlan_filtering) {
+		    ds->ports[i].vlan_filtering) {
 			all_user_ports_removed = false;
 			break;
 		}
@@ -891,7 +889,7 @@ mt7530_port_bridge_leave(struct dsa_switch *ds, int port,
 		 * And the other port's port matrix cannot be broken when the
 		 * other port is still a VLAN-aware port.
 		 */
-		if (!priv->ports[i].vlan_filtering &&
+		if (!ds->ports[i].vlan_filtering &&
 		    dsa_is_user_port(ds, i) && i != port) {
 			if (dsa_to_port(ds, i)->bridge_dev != bridge)
 				continue;
@@ -1013,10 +1011,6 @@ static int
 mt7530_port_vlan_filtering(struct dsa_switch *ds, int port,
 			   bool vlan_filtering)
 {
-	struct mt7530_priv *priv = ds->priv;
-
-	priv->ports[port].vlan_filtering = vlan_filtering;
-
 	if (vlan_filtering) {
 		/* The port is being kept as VLAN-unaware port when bridge is
 		 * set up with vlan_filtering not being set, Otherwise, the
@@ -1139,7 +1133,7 @@ mt7530_port_vlan_add(struct dsa_switch *ds, int port,
 	/* The port is kept as VLAN-unaware if bridge with vlan_filtering not
 	 * being set.
 	 */
-	if (!priv->ports[port].vlan_filtering)
+	if (!ds->ports[port].vlan_filtering)
 		return;
 
 	mutex_lock(&priv->reg_mutex);
@@ -1170,7 +1164,7 @@ mt7530_port_vlan_del(struct dsa_switch *ds, int port,
 	/* The port is kept as VLAN-unaware if bridge with vlan_filtering not
 	 * being set.
 	 */
-	if (!priv->ports[port].vlan_filtering)
+	if (!ds->ports[port].vlan_filtering)
 		return 0;
 
 	mutex_lock(&priv->reg_mutex);
diff --git a/drivers/net/dsa/mt7530.h b/drivers/net/dsa/mt7530.h
index a95ed958df5b..1eec7bdc283a 100644
--- a/drivers/net/dsa/mt7530.h
+++ b/drivers/net/dsa/mt7530.h
@@ -410,7 +410,6 @@ struct mt7530_port {
 	bool enable;
 	u32 pm;
 	u16 pvid;
-	bool vlan_filtering;
 };
 
 /* struct mt7530_priv -	This is the main data structure for holding the state
-- 
2.17.1

Re: [PATCH net-next 04/17] net: dsa: mt7530: Use vlan_filtering property from dsa_port

[Andrew Lunn]

On Sun, Mar 31, 2019 at 08:42:19PM +0300, Vladimir Oltean wrote:
> This field was recently introduced, so keeping state inside the driver
> is no longer necessary.
> 
> Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
> Suggested-by: Andrew Lunn <andrew@lunn.ch>

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

[PATCH net-next 05/17] net: dsa: Add more convenient functions for installing port VLANs

[Vladimir Oltean]

This hides the need to perform a two-phase transaction and construct a
switchdev_obj_port_vlan struct.

Call graph (including a function that will be introduced in a follow-up
patch) looks like this now (same for the *_vlan_del function):

dsa_slave_vlan_rx_add_vid   dsa_port_setup_8021q_tagging
            |                        |
            |                        |
            |          +-------------+
            |          |
            v          v
         __dsa_port_vlan_add     dsa_slave_port_obj_add
                  |                         |
                  +-------+         +-------+
                          |         |
                          v         v
                       dsa_port_vlan_add

Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
---
 net/dsa/dsa_priv.h |  2 ++
 net/dsa/port.c     | 31 +++++++++++++++++++++++++++++++
 net/dsa/slave.c    | 16 ++--------------
 3 files changed, 35 insertions(+), 14 deletions(-)

diff --git a/net/dsa/dsa_priv.h b/net/dsa/dsa_priv.h
index 093b7d145eb1..766d015f6f38 100644
--- a/net/dsa/dsa_priv.h
+++ b/net/dsa/dsa_priv.h
@@ -169,6 +169,8 @@ int dsa_port_vlan_add(struct dsa_port *dp,
 		      struct switchdev_trans *trans);
 int dsa_port_vlan_del(struct dsa_port *dp,
 		      const struct switchdev_obj_port_vlan *vlan);
+int __dsa_port_vlan_add(struct dsa_port *dp, u16 vid, u16 flags);
+int __dsa_port_vlan_del(struct dsa_port *dp, u16 vid);
 int dsa_port_link_register_of(struct dsa_port *dp);
 void dsa_port_link_unregister_of(struct dsa_port *dp);
 
diff --git a/net/dsa/port.c b/net/dsa/port.c
index a86fe3be1261..10065a622ada 100644
--- a/net/dsa/port.c
+++ b/net/dsa/port.c
@@ -326,6 +326,37 @@ int dsa_port_vlan_del(struct dsa_port *dp,
 	return 0;
 }
 
+int __dsa_port_vlan_add(struct dsa_port *dp, u16 vid, u16 flags)
+{
+	struct switchdev_obj_port_vlan vlan = {
+		.obj.id = SWITCHDEV_OBJ_ID_PORT_VLAN,
+		.flags = flags,
+		.vid_begin = vid,
+		.vid_end = vid,
+	};
+	struct switchdev_trans trans;
+	int err;
+
+	trans.ph_prepare = true;
+	err = dsa_port_vlan_add(dp, &vlan, &trans);
+	if (err == -EOPNOTSUPP)
+		return 0;
+
+	trans.ph_prepare = false;
+	return dsa_port_vlan_add(dp, &vlan, &trans);
+}
+
+int __dsa_port_vlan_del(struct dsa_port *dp, u16 vid)
+{
+	struct switchdev_obj_port_vlan vlan = {
+		.obj.id = SWITCHDEV_OBJ_ID_PORT_VLAN,
+		.vid_begin = vid,
+		.vid_end = vid,
+	};
+
+	return dsa_port_vlan_del(dp, &vlan);
+}
+
 static struct phy_device *dsa_port_get_phy_device(struct dsa_port *dp)
 {
 	struct device_node *phy_dn;
diff --git a/net/dsa/slave.c b/net/dsa/slave.c
index 093eef6f2599..e644095fdd44 100644
--- a/net/dsa/slave.c
+++ b/net/dsa/slave.c
@@ -987,13 +987,6 @@ static int dsa_slave_vlan_rx_add_vid(struct net_device *dev, __be16 proto,
 				     u16 vid)
 {
 	struct dsa_port *dp = dsa_slave_to_port(dev);
-	struct switchdev_obj_port_vlan vlan = {
-		.vid_begin = vid,
-		.vid_end = vid,
-		/* This API only allows programming tagged, non-PVID VIDs */
-		.flags = 0,
-	};
-	struct switchdev_trans trans;
 	struct bridge_vlan_info info;
 	int ret;
 
@@ -1010,13 +1003,8 @@ static int dsa_slave_vlan_rx_add_vid(struct net_device *dev, __be16 proto,
 			return -EBUSY;
 	}
 
-	trans.ph_prepare = true;
-	ret = dsa_port_vlan_add(dp, &vlan, &trans);
-	if (ret == -EOPNOTSUPP)
-		return 0;
-
-	trans.ph_prepare = false;
-	return dsa_port_vlan_add(dp, &vlan, &trans);
+	/* This API only allows programming tagged, non-PVID VIDs */
+	return __dsa_port_vlan_add(dp, vid, 0);
 }
 
 static int dsa_slave_vlan_rx_kill_vid(struct net_device *dev, __be16 proto,
-- 
2.17.1

Re: [PATCH net-next 05/17] net: dsa: Add more convenient functions for installing port VLANs

[Andrew Lunn]

On Sun, Mar 31, 2019 at 08:42:20PM +0300, Vladimir Oltean wrote:
> This hides the need to perform a two-phase transaction and construct a
> switchdev_obj_port_vlan struct.
> 
> Call graph (including a function that will be introduced in a follow-up
> patch) looks like this now (same for the *_vlan_del function):
> 
> dsa_slave_vlan_rx_add_vid   dsa_port_setup_8021q_tagging
>             |                        |
>             |                        |
>             |          +-------------+
>             |          |
>             v          v
>          __dsa_port_vlan_add     dsa_slave_port_obj_add
>                   |                         |
>                   +-------+         +-------+
>                           |         |
>                           v         v
>                        dsa_port_vlan_add
> 
> Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
> ---
>  net/dsa/dsa_priv.h |  2 ++
>  net/dsa/port.c     | 31 +++++++++++++++++++++++++++++++
>  net/dsa/slave.c    | 16 ++--------------
>  3 files changed, 35 insertions(+), 14 deletions(-)
> 
> diff --git a/net/dsa/dsa_priv.h b/net/dsa/dsa_priv.h
> index 093b7d145eb1..766d015f6f38 100644
> --- a/net/dsa/dsa_priv.h
> +++ b/net/dsa/dsa_priv.h
> @@ -169,6 +169,8 @@ int dsa_port_vlan_add(struct dsa_port *dp,
>  		      struct switchdev_trans *trans);
>  int dsa_port_vlan_del(struct dsa_port *dp,
>  		      const struct switchdev_obj_port_vlan *vlan);
> +int __dsa_port_vlan_add(struct dsa_port *dp, u16 vid, u16 flags);
> +int __dsa_port_vlan_del(struct dsa_port *dp, u16 vid);
>  int dsa_port_link_register_of(struct dsa_port *dp);
>  void dsa_port_link_unregister_of(struct dsa_port *dp);
>  
> diff --git a/net/dsa/port.c b/net/dsa/port.c
> index a86fe3be1261..10065a622ada 100644
> --- a/net/dsa/port.c
> +++ b/net/dsa/port.c
> @@ -326,6 +326,37 @@ int dsa_port_vlan_del(struct dsa_port *dp,
>  	return 0;
>  }
>  
> +int __dsa_port_vlan_add(struct dsa_port *dp, u16 vid, u16 flags)
> +{

Hi Vladimir

Maybe call this dsa_port_vlan_add_trans()?

I don't like the __, especially when the symbol is not static.

  Andrew

[PATCH net-next 06/17] net: dsa: Call driver's setup callback after setting up its switchdev notifier

[Vladimir Oltean]

This allows the driver to perform some manipulations of its own during
setup, using generic code.
One current usage scenario is for the driver to request DSA to set up
802.1Q based switch tagging for its ports.

Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
---
 net/dsa/dsa2.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/net/dsa/dsa2.c b/net/dsa/dsa2.c
index c00ee464afc7..5beceb18b7e2 100644
--- a/net/dsa/dsa2.c
+++ b/net/dsa/dsa2.c
@@ -360,14 +360,14 @@ static int dsa_switch_setup(struct dsa_switch *ds)
 	if (err)
 		return err;
 
-	err = ds->ops->setup(ds);
-	if (err < 0)
-		return err;
-
 	err = dsa_switch_register_notifier(ds);
 	if (err)
 		return err;
 
+	err = ds->ops->setup(ds);
+	if (err < 0)
+		return err;
+
 	if (!ds->slave_mii_bus && ds->ops->phy_read) {
 		ds->slave_mii_bus = devm_mdiobus_alloc(ds->dev);
 		if (!ds->slave_mii_bus)
-- 
2.17.1

Re: [PATCH net-next 06/17] net: dsa: Call driver's setup callback after setting up its switchdev notifier

[Andrew Lunn]

On Sun, Mar 31, 2019 at 08:42:21PM +0300, Vladimir Oltean wrote:
> This allows the driver to perform some manipulations of its own during
> setup, using generic code.
> One current usage scenario is for the driver to request DSA to set up
> 802.1Q based switch tagging for its ports.
> 
> Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
> ---
>  net/dsa/dsa2.c | 8 ++++----
>  1 file changed, 4 insertions(+), 4 deletions(-)
> 
> diff --git a/net/dsa/dsa2.c b/net/dsa/dsa2.c
> index c00ee464afc7..5beceb18b7e2 100644
> --- a/net/dsa/dsa2.c
> +++ b/net/dsa/dsa2.c
> @@ -360,14 +360,14 @@ static int dsa_switch_setup(struct dsa_switch *ds)
>  	if (err)
>  		return err;
>  
> -	err = ds->ops->setup(ds);
> -	if (err < 0)
> -		return err;
> -
>  	err = dsa_switch_register_notifier(ds);
>  	if (err)
>  		return err;

It seems that notifiers are the important thing here? Maybe state that
in the commit message?

I'm also wondering how safe this is in general. If we have not yet
called the driver setup, the switch is potentially not yet ready to
actually handle an requests that come via the notifier. If such
notifiers can only come from the driver itself, it should be
safe. However, if they could come from the rest of the stack, i could
see bad things happening.

    Andrew

Re: [PATCH net-next 06/17] net: dsa: Call driver's setup callback after setting up its switchdev notifier

[Vladimir Oltean]

On 4/3/19 12:03 AM, Andrew Lunn wrote:
> On Sun, Mar 31, 2019 at 08:42:21PM +0300, Vladimir Oltean wrote:
>> This allows the driver to perform some manipulations of its own during
>> setup, using generic code.
>> One current usage scenario is for the driver to request DSA to set up
>> 802.1Q based switch tagging for its ports.
>>
>> Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
>> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
>> ---
>>   net/dsa/dsa2.c | 8 ++++----
>>   1 file changed, 4 insertions(+), 4 deletions(-)
>>
>> diff --git a/net/dsa/dsa2.c b/net/dsa/dsa2.c
>> index c00ee464afc7..5beceb18b7e2 100644
>> --- a/net/dsa/dsa2.c
>> +++ b/net/dsa/dsa2.c
>> @@ -360,14 +360,14 @@ static int dsa_switch_setup(struct dsa_switch *ds)
>>   	if (err)
>>   		return err;
>>   
>> -	err = ds->ops->setup(ds);
>> -	if (err < 0)
>> -		return err;
>> -
>>   	err = dsa_switch_register_notifier(ds);
>>   	if (err)
>>   		return err;
> 
> It seems that notifiers are the important thing here? Maybe state that
> in the commit message?
> 
> I'm also wondering how safe this is in general. If we have not yet
> called the driver setup, the switch is potentially not yet ready to
> actually handle an requests that come via the notifier. If such
> notifiers can only come from the driver itself, it should be
> safe. However, if they could come from the rest of the stack, i could
> see bad things happening.
> 
>      Andrew
> 

Hi Andrew,

Correct me if I'm wrong, but the switchdev notifiers coming from the 
stack are coming through a net device, and the slave net devices are 
only created in dsa_port_setup, which is later than the code I'm 
modifying anyway? Do you have an example of potentially racy situation 
caused by this change?

Thanks,
-Vladimir

[PATCH net-next 07/17] net: dsa: Optional VLAN-based port separation for switches without tagging

[Vladimir Oltean]

This patch provides generic DSA code for using VLAN (802.1Q) tags for
the same purpose as a dedicated switch tag for injection/extraction.
It is based on the discussions and interest that has been so far
expressed in https://www.spinics.net/lists/netdev/msg556125.html.

Unlike all other DSA-supported tagging protocols, CONFIG_NET_DSA_TAG_8021Q
does not offer a complete solution for drivers (nor can it). Instead, it
provides generic code that driver can opt into calling:
- dsa_8021q_xmit: Inserts a VLAN header with the specified contents.
  Currently a few driver are inserting headers that are simply 802.1Q
  with custom fields. Can be called from another tagging protocol's xmit
  function.
- dsa_8021q_rcv: Retrieves the TPID and TCI from a VLAN-tagged skb.
  Removing the VLAN header is left as a decision for the caller to make.
- dsa_port_setup_8021q_tagging: For each user port, installs an Rx VID
  and a Tx VID, for proper untagged traffic identification on ingress
  and steering on egress. Also sets up the VLAN trunk on the upstream
  (CPU or DSA) port. Drivers are intentionally left to call this
  function explicitly, depending on the context and hardware support.
  The expected switch behavior and VLAN semantics should not be violated
  under any conditions. That is, after calling
  dsa_port_setup_8021q_tagging, the hardware should still pass all
  ingress traffic, be it tagged or untagged.

This only works when switch ports are standalone, or when they are added
to a VLAN-unaware bridge. It will probably remain this way for the
reasons below.

When added to a bridge that has vlan_filtering 1, the bridge core will
install its own VLANs and reset the pvids through switchdev. For the
bridge core, switchdev is a write-only pipe. All VLAN-related state is
kept in the bridge core and nothing is read from DSA/switchdev or from
the driver. So the bridge core will break this port separation because
it will install the vlan_default_pvid into all switchdev ports.

Even if we could teach the bridge driver about switchdev preference of a
certain vlan_default_pvid, there would still exist many other challenges.

Firstly, in the DSA rcv callback, a driver would have to perform an
iterative reverse lookup to find the correct switch port. That is
because the port is a bridge slave, so its Rx VID (port PVID) is subject
to user configuration. How would we ensure that the user doesn't reset
the pvid to a different value, or to a non-unique value within this DSA
switch tree?

Finally, not all switch ports are equal in DSA, and that makes it
difficult for the bridge to be completely aware of this anyway.
The CPU port needs to transmit tagged packets (VLAN trunk) in order for
the DSA rcv code to be able to decode source information.
But the bridge code has absolutely no idea which switch port is the CPU
port, if nothing else then just because there is no netdevice registered
by DSA for the CPU port.
Also DSA does not currently allow the user to specify that they want the
CPU port to do VLAN trunking anyway. VLANs are added to the CPU port
using the same flags as they were added on the user port.

So the VLANs installed by dsa_port_setup_8021q_tagging per driver
request should remain private from the bridge's and user's perspective,
and should not alter the hardware's behavior with VLAN-tagged traffic.
If the hardware cannot handle VLAN tag stacking, it should also disable
this port separation when added as slave to a vlan_filtering bridge.
If the hardware does support VLAN tag stacking, it should somehow back
up its private VLAN settings when the bridge tries to override them.

Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
---
 include/net/dsa.h   |   4 +
 net/dsa/Kconfig     |   9 +++
 net/dsa/Makefile    |   1 +
 net/dsa/dsa_priv.h  |  10 +++
 net/dsa/tag_8021q.c | 193 ++++++++++++++++++++++++++++++++++++++++++++
 5 files changed, 217 insertions(+)
 create mode 100644 net/dsa/tag_8021q.c

diff --git a/include/net/dsa.h b/include/net/dsa.h
index a16fd577349b..b22c350c40f0 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -574,5 +574,9 @@ int dsa_port_get_phy_strings(struct dsa_port *dp, uint8_t *data);
 int dsa_port_get_ethtool_phy_stats(struct dsa_port *dp, uint64_t *data);
 int dsa_port_get_phy_sset_count(struct dsa_port *dp);
 void dsa_port_phylink_mac_change(struct dsa_switch *ds, int port, bool up);
+#ifdef CONFIG_NET_DSA_TAG_8021Q
+int dsa_port_setup_8021q_tagging(struct dsa_switch *ds, int index,
+				 bool enabled);
+#endif
 
 #endif
diff --git a/net/dsa/Kconfig b/net/dsa/Kconfig
index fab49132345f..2f3a103d7d1a 100644
--- a/net/dsa/Kconfig
+++ b/net/dsa/Kconfig
@@ -26,6 +26,15 @@ config NET_DSA_LEGACY
 	  This feature is scheduled for removal in 4.17.
 
 # tagging formats
+config NET_DSA_TAG_8021Q
+	bool
+	help
+	  Unlike the other tagging protocols, the 802.1Q config option simply
+	  provides helpers for other tagging implementations that might rely on
+	  VLAN in one way or another. It is not a complete solution.
+
+	  Drivers which use these helpers should select this as dependency.
+
 config NET_DSA_TAG_BRCM
 	bool
 
diff --git a/net/dsa/Makefile b/net/dsa/Makefile
index 6e721f7a2947..d7fc3253d497 100644
--- a/net/dsa/Makefile
+++ b/net/dsa/Makefile
@@ -5,6 +5,7 @@ dsa_core-y += dsa.o dsa2.o master.o port.o slave.o switch.o
 dsa_core-$(CONFIG_NET_DSA_LEGACY) += legacy.o
 
 # tagging formats
+dsa_core-$(CONFIG_NET_DSA_TAG_8021Q) += tag_8021q.o
 dsa_core-$(CONFIG_NET_DSA_TAG_BRCM) += tag_brcm.o
 dsa_core-$(CONFIG_NET_DSA_TAG_BRCM_PREPEND) += tag_brcm.o
 dsa_core-$(CONFIG_NET_DSA_TAG_DSA) += tag_dsa.o
diff --git a/net/dsa/dsa_priv.h b/net/dsa/dsa_priv.h
index 766d015f6f38..4219f8979c8a 100644
--- a/net/dsa/dsa_priv.h
+++ b/net/dsa/dsa_priv.h
@@ -203,6 +203,16 @@ dsa_slave_to_master(const struct net_device *dev)
 int dsa_switch_register_notifier(struct dsa_switch *ds);
 void dsa_switch_unregister_notifier(struct dsa_switch *ds);
 
+/* tag_8021q.c */
+struct sk_buff *dsa_8021q_xmit(struct sk_buff *skb, struct net_device *netdev,
+			       u16 tpid, u16 tci);
+struct sk_buff *dsa_8021q_rcv(struct sk_buff *skb, struct net_device *netdev,
+			      struct packet_type *pt, u16 *tpid, u16 *tci);
+u16 dsa_tagging_tx_vid(struct dsa_switch *ds, int port);
+u16 dsa_tagging_rx_vid(struct dsa_switch *ds, int port);
+int dsa_tagging_rx_switch_id(u16 vid);
+int dsa_tagging_rx_source_port(u16 vid);
+
 /* tag_brcm.c */
 extern const struct dsa_device_ops brcm_netdev_ops;
 extern const struct dsa_device_ops brcm_prepend_netdev_ops;
diff --git a/net/dsa/tag_8021q.c b/net/dsa/tag_8021q.c
new file mode 100644
index 000000000000..1b11b245e2d6
--- /dev/null
+++ b/net/dsa/tag_8021q.c
@@ -0,0 +1,193 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2019, Vladimir Oltean <olteanv@gmail.com>
+ */
+#include <linux/if_bridge.h>
+#include <linux/if_vlan.h>
+
+#include "dsa_priv.h"
+
+#define DSA_TAGGING_VID_RANGE    (DSA_MAX_SWITCHES * DSA_MAX_PORTS)
+#define DSA_TAGGING_VID_BASE     (VLAN_N_VID - 2 * DSA_TAGGING_VID_RANGE - 1)
+#define DSA_TAGGING_RX_VID_BASE  (DSA_TAGGING_VID_BASE)
+#define DSA_TAGGING_TX_VID_BASE  (DSA_TAGGING_VID_BASE + DSA_TAGGING_VID_RANGE)
+
+u16 dsa_tagging_tx_vid(struct dsa_switch *ds, int port)
+{
+	return DSA_TAGGING_TX_VID_BASE + (DSA_MAX_PORTS * ds->index) + port;
+}
+
+u16 dsa_tagging_rx_vid(struct dsa_switch *ds, int port)
+{
+	return DSA_TAGGING_RX_VID_BASE + (DSA_MAX_PORTS * ds->index) + port;
+}
+
+int dsa_tagging_rx_switch_id(u16 vid)
+{
+	return ((vid - DSA_TAGGING_RX_VID_BASE) / DSA_MAX_PORTS);
+}
+
+int dsa_tagging_rx_source_port(u16 vid)
+{
+	return ((vid - DSA_TAGGING_RX_VID_BASE) % DSA_MAX_PORTS);
+}
+
+/* Rx VLAN tagging (left) and Tx VLAN tagging (right) setup shown for a single
+ * front-panel switch port (here swp0).
+ *
+ * Port identification through VLAN (802.1Q) tags has different requirements
+ * for it to work effectively:
+ *  - On Rx (ingress from network): each front-panel port must have a pvid
+ *    that uniquely identifies it, and the egress of this pvid must be tagged
+ *    towards the CPU port, so that software can recover the source port based
+ *    on the VID in the frame. But this would only work for standalone ports;
+ *    if bridged, this VLAN setup would break autonomous forwarding and would
+ *    force all switched traffic to pass through the CPU. So we must also make
+ *    the other front-panel ports members of this VID we're adding, albeit
+ *    we're not making it their PVID (they'll still have their own).
+ *    By the way - just because we're installing the same VID in multiple
+ *    switch ports doesn't mean that they'll start to talk to one another, even
+ *    while not bridged: the final forwarding decision is still an AND between
+ *    the L2 forwarding information (which is limiting forwarding in this case)
+ *    and the VLAN-based restrictions (of which there are none in this case,
+ *    since all ports are members).
+ *  - On Tx (ingress from CPU and towards network) we are faced with a problem.
+ *    If we were to tag traffic (from within DSA) with the port's pvid, all
+ *    would be well, assuming the switch ports were standalone. Frames would
+ *    have no choice but to be directed towards the correct front-panel port.
+ *    But because we also want the Rx VLAN to not break bridging, then
+ *    inevitably that means that we have to give them a choice (of what
+ *    front-panel port to go out on), and therefore we cannot steer traffic
+ *    based on the Rx VID. So what we do is simply install one more VID on the
+ *    front-panel and CPU ports, and profit off of the fact that steering will
+ *    work just by virtue of the fact that there is only one other port that's
+ *    a member of the VID we're tagging the traffic with - the desired one.
+ *
+ * So at the end, each front-panel port will have one Rx VID (also the PVID),
+ * the Rx VID of all other front-panel ports, and one Tx VID. Whereas the CPU
+ * port will have the Rx and Tx VIDs of all front-panel ports, and on top of
+ * that, is also tagged-input and tagged-output (VLAN trunk).
+ *
+ *               CPU port                               CPU port
+ * +-------------+-----+-------------+    +-------------+-----+-------------+
+ * |  Rx VID     |     |             |    |  Tx VID     |     |             |
+ * |  of swp0    |     |             |    |  of swp0    |     |             |
+ * |             +-----+             |    |             +-----+             |
+ * |                ^ T              |    |                | Tagged         |
+ * |                |                |    |                | ingress        |
+ * |    +-------+---+---+-------+    |    |    +-----------+                |
+ * |    |       |       |       |    |    |    | Untagged                   |
+ * |    |     U v     U v     U v    |    |    v egress                     |
+ * | +-----+ +-----+ +-----+ +-----+ |    | +-----+ +-----+ +-----+ +-----+ |
+ * | |     | |     | |     | |     | |    | |     | |     | |     | |     | |
+ * | |PVID | |     | |     | |     | |    | |     | |     | |     | |     | |
+ * +-+-----+-+-----+-+-----+-+-----+-+    +-+-----+-+-----+-+-----+-+-----+-+
+ *   swp0    swp1    swp2    swp3           swp0    swp1    swp2    swp3
+ */
+int dsa_port_setup_8021q_tagging(struct dsa_switch *ds, int port, bool enabled)
+{
+	int upstream = dsa_upstream_port(ds, port);
+	struct dsa_port *dp = &ds->ports[port];
+	struct dsa_port *upstream_dp = &ds->ports[upstream];
+	u16 rx_vid = dsa_tagging_rx_vid(ds, port);
+	u16 tx_vid = dsa_tagging_tx_vid(ds, port);
+	int i, err;
+
+	/* The CPU port is implicitly configured by
+	 * configuring the front-panel ports
+	 */
+	if (!dsa_is_user_port(ds, port))
+		return 0;
+
+	/* Add this user port's Rx VID to the membership list of all others
+	 * (including itself). This is so that bridging will not be hindered.
+	 * L2 forwarding rules still take precedence when there are no VLAN
+	 * restrictions, so there are no concerns about leaking traffic.
+	 */
+	for (i = 0; i < ds->num_ports; i++) {
+		struct dsa_port *other_dp = &ds->ports[i];
+		u16 flags;
+
+		if (i == upstream)
+			/* CPU port needs to see this port's Rx VID
+			 * as tagged egress.
+			 */
+			flags = 0;
+		else if (i == port)
+			/* The Rx VID is pvid on this port */
+			flags = BRIDGE_VLAN_INFO_UNTAGGED |
+				BRIDGE_VLAN_INFO_PVID;
+		else
+			/* The Rx VID is a regular VLAN on all others */
+			flags = BRIDGE_VLAN_INFO_UNTAGGED;
+
+		if (enabled)
+			err = __dsa_port_vlan_add(other_dp, rx_vid, flags);
+		else
+			err = __dsa_port_vlan_del(other_dp, rx_vid);
+		if (err) {
+			dev_err(ds->dev, "Failed to apply Rx VID %d to port %d: %d\n",
+				rx_vid, port, err);
+			return err;
+		}
+	}
+	/* Finally apply the Tx VID on this port and on the CPU port */
+	if (enabled)
+		err = __dsa_port_vlan_add(dp, tx_vid,
+					  BRIDGE_VLAN_INFO_UNTAGGED);
+	else
+		err = __dsa_port_vlan_del(dp, tx_vid);
+	if (err) {
+		dev_err(ds->dev, "Failed to apply Tx VID %d on port %d: %d\n",
+			tx_vid, port, err);
+		return err;
+	}
+	if (enabled)
+		err = __dsa_port_vlan_add(upstream_dp, tx_vid, 0);
+	else
+		err = __dsa_port_vlan_del(upstream_dp, tx_vid);
+	if (err) {
+		dev_err(ds->dev, "Failed to apply Tx VID %d on port %d: %d\n",
+			tx_vid, upstream, err);
+		return err;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(dsa_port_setup_8021q_tagging);
+
+struct sk_buff *dsa_8021q_xmit(struct sk_buff *skb, struct net_device *netdev,
+			       u16 tpid, u16 tci)
+{
+	/* skb->data points at skb_mac_header, which
+	 * is fine for vlan_insert_tag.
+	 */
+	return vlan_insert_tag(skb, tpid, tci);
+}
+EXPORT_SYMBOL_GPL(dsa_8021q_xmit);
+
+struct sk_buff *dsa_8021q_rcv(struct sk_buff *skb, struct net_device *netdev,
+			      struct packet_type *pt, u16 *tpid, u16 *tci)
+{
+	struct vlan_ethhdr *tag;
+
+	if (unlikely(!pskb_may_pull(skb, VLAN_HLEN)))
+		return NULL;
+
+	tag = vlan_eth_hdr(skb);
+	*tpid = ntohs(tag->h_vlan_proto);
+	*tci = ntohs(tag->h_vlan_TCI);
+
+	/* skb->data points in the middle of the VLAN tag,
+	 * after tpid and before tci. This is because so far,
+	 * ETH_HLEN (DMAC, SMAC, EtherType) bytes were pulled.
+	 * There are 2 bytes of VLAN tag left in skb->data, and upper
+	 * layers expect the 'real' EtherType to be consumed as well.
+	 * Coincidentally, a VLAN header is also of the same size as
+	 * the number of bytes that need to be pulled.
+	 */
+	skb_pull_rcsum(skb, VLAN_HLEN);
+
+	return skb;
+}
+EXPORT_SYMBOL_GPL(dsa_8021q_rcv);
+
-- 
2.17.1

[PATCH net-next 08/17] net: dsa: Be aware of switches where VLAN filtering is a global setting

[Vladimir Oltean]

On some switches, the action of whether to parse VLAN frame headers and use
that information for ingress admission is configurable, but not per
port. Such is the case for the Broadcom BCM53xx and the NXP SJA1105
families, for example. In that case, DSA can prevent the bridge core
from trying to apply different VLAN filtering settings on net devices
that belong to the same switch.

Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Suggested-by: Florian Fainelli <f.fainelli@gmail.com>
---
 include/net/dsa.h |  5 +++++
 net/dsa/port.c    | 34 ++++++++++++++++++++++++++++++++++
 net/dsa/switch.c  |  1 +
 3 files changed, 40 insertions(+)

diff --git a/include/net/dsa.h b/include/net/dsa.h
index b22c350c40f0..91375bcf2cfb 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -209,6 +209,11 @@ struct dsa_switch {
 	/* Number of switch port queues */
 	unsigned int		num_tx_queues;
 
+	/* Disallow bridge core from requesting different VLAN awareness
+	 * settings on ports if not hardware-supported
+	 */
+	bool			vlan_filtering_is_global;
+
 	unsigned long		*bitmap;
 	unsigned long		_bitmap;
 
diff --git a/net/dsa/port.c b/net/dsa/port.c
index 10065a622ada..d04e7c4f5efe 100644
--- a/net/dsa/port.c
+++ b/net/dsa/port.c
@@ -154,6 +154,37 @@ void dsa_port_bridge_leave(struct dsa_port *dp, struct net_device *br)
 	dsa_port_set_state_now(dp, BR_STATE_FORWARDING);
 }
 
+static bool dsa_port_can_apply_vlan_filtering(struct dsa_port *dp,
+					      bool vlan_filtering)
+{
+	struct dsa_switch *ds = dp->ds;
+	int i;
+
+	if (!ds->vlan_filtering_is_global)
+		return true;
+
+	/* For cases where enabling/disabling VLAN awareness is global to the
+	 * switch, we need to handle the case where multiple bridges span
+	 * different ports of the same switch device and one of them has a
+	 * different setting than what is being requested.
+	 */
+	for (i = 0; i < ds->num_ports; i++) {
+		struct net_device *other_bridge;
+
+		other_bridge = dsa_to_port(ds, i)->bridge_dev;
+		if (!other_bridge)
+			continue;
+		/* If it's the same bridge, it also has same
+		 * vlan_filtering setting => no need to check
+		 */
+		if (other_bridge == dp->bridge_dev)
+			continue;
+		if (br_vlan_enabled(other_bridge) != vlan_filtering)
+			return false;
+	}
+	return true;
+}
+
 int dsa_port_vlan_filtering(struct dsa_port *dp, bool vlan_filtering,
 			    struct switchdev_trans *trans)
 {
@@ -165,6 +196,9 @@ int dsa_port_vlan_filtering(struct dsa_port *dp, bool vlan_filtering,
 		return 0;
 
 	if (ds->ops->port_vlan_filtering) {
+		if (!dsa_port_can_apply_vlan_filtering(dp, vlan_filtering))
+			return -EINVAL;
+
 		err = ds->ops->port_vlan_filtering(ds, dp->index,
 						   vlan_filtering);
 		if (err)
diff --git a/net/dsa/switch.c b/net/dsa/switch.c
index fde4e9195709..03b8d8928651 100644
--- a/net/dsa/switch.c
+++ b/net/dsa/switch.c
@@ -10,6 +10,7 @@
  * (at your option) any later version.
  */
 
+#include <linux/if_bridge.h>
 #include <linux/netdevice.h>
 #include <linux/notifier.h>
 #include <linux/if_vlan.h>
-- 
2.17.1

[PATCH net-next 09/17] net: dsa: b53: Let DSA handle mismatched VLAN filtering settings

[Vladimir Oltean]

The DSA core is now able to do this check prior to calling the
.port_vlan_filtering callback, so tell it that VLAN filtering is global
for this particular hardware.

Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Suggested-by: Florian Fainelli <f.fainelli@gmail.com>
---
 drivers/net/dsa/b53/b53_common.c | 25 +++++++------------------
 1 file changed, 7 insertions(+), 18 deletions(-)

diff --git a/drivers/net/dsa/b53/b53_common.c b/drivers/net/dsa/b53/b53_common.c
index 0852e5e08177..a779b9c3ab6e 100644
--- a/drivers/net/dsa/b53/b53_common.c
+++ b/drivers/net/dsa/b53/b53_common.c
@@ -966,6 +966,13 @@ static int b53_setup(struct dsa_switch *ds)
 			b53_disable_port(ds, port);
 	}
 
+	/* Let DSA handle the case were multiple bridges span the same switch
+	 * device and different VLAN awareness settings are requested, which
+	 * would be breaking filtering semantics for any of the other bridge
+	 * devices. (not hardware supported)
+	 */
+	ds->vlan_filtering_is_global = true;
+
 	return ret;
 }
 
@@ -1275,26 +1282,8 @@ EXPORT_SYMBOL(b53_phylink_mac_link_up);
 int b53_vlan_filtering(struct dsa_switch *ds, int port, bool vlan_filtering)
 {
 	struct b53_device *dev = ds->priv;
-	struct net_device *bridge_dev;
-	unsigned int i;
 	u16 pvid, new_pvid;
 
-	/* Handle the case were multiple bridges span the same switch device
-	 * and one of them has a different setting than what is being requested
-	 * which would be breaking filtering semantics for any of the other
-	 * bridge devices.
-	 */
-	b53_for_each_port(dev, i) {
-		bridge_dev = dsa_to_port(ds, i)->bridge_dev;
-		if (bridge_dev &&
-		    bridge_dev != dsa_to_port(ds, port)->bridge_dev &&
-		    br_vlan_enabled(bridge_dev) != vlan_filtering) {
-			netdev_err(bridge_dev,
-				   "VLAN filtering is global to the switch!\n");
-			return -EINVAL;
-		}
-	}
-
 	b53_read16(dev, B53_VLAN_PAGE, B53_VLAN_PORT_DEF_TAG(port), &pvid);
 	new_pvid = pvid;
 	if (dev->vlan_filtering_enabled && !vlan_filtering) {
-- 
2.17.1

[PATCH net-next 10/17] net: dsa: Introduce driver for NXP SJA1105 5-port L2 switch

[Vladimir Oltean]

At this moment the following is supported:
* Link state management through phylib
* Autonomous L2 forwarding managed through iproute2 bridge commands. The
  switch ports are initialized in a mode where they can only talk to the
  CPU port. However, IP termination must be done currently through the
  master netdevice.

Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: Georg Waibel <georg.waibel@sensor-technik.de>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
---
 MAINTAINERS                                   |    6 +
 drivers/net/dsa/Kconfig                       |    2 +
 drivers/net/dsa/Makefile                      |    1 +
 drivers/net/dsa/sja1105/Kconfig               |   17 +
 drivers/net/dsa/sja1105/Makefile              |    9 +
 drivers/net/dsa/sja1105/sja1105.h             |  134 +++
 drivers/net/dsa/sja1105/sja1105_clocking.c    |  643 ++++++++++
 .../net/dsa/sja1105/sja1105_dynamic_config.c  |  478 ++++++++
 .../net/dsa/sja1105/sja1105_dynamic_config.h  |   40 +
 drivers/net/dsa/sja1105/sja1105_main.c        |  890 ++++++++++++++
 drivers/net/dsa/sja1105/sja1105_spi.c         |  665 +++++++++++
 .../net/dsa/sja1105/sja1105_static_config.c   | 1036 +++++++++++++++++
 .../net/dsa/sja1105/sja1105_static_config.h   |  308 +++++
 13 files changed, 4229 insertions(+)
 create mode 100644 drivers/net/dsa/sja1105/Kconfig
 create mode 100644 drivers/net/dsa/sja1105/Makefile
 create mode 100644 drivers/net/dsa/sja1105/sja1105.h
 create mode 100644 drivers/net/dsa/sja1105/sja1105_clocking.c
 create mode 100644 drivers/net/dsa/sja1105/sja1105_dynamic_config.c
 create mode 100644 drivers/net/dsa/sja1105/sja1105_dynamic_config.h
 create mode 100644 drivers/net/dsa/sja1105/sja1105_main.c
 create mode 100644 drivers/net/dsa/sja1105/sja1105_spi.c
 create mode 100644 drivers/net/dsa/sja1105/sja1105_static_config.c
 create mode 100644 drivers/net/dsa/sja1105/sja1105_static_config.h

diff --git a/MAINTAINERS b/MAINTAINERS
index 89315bb1cb83..d808520b4fa3 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -11117,6 +11117,12 @@ S:	Maintained
 F:	Documentation/devicetree/bindings/sound/sgtl5000.txt
 F:	sound/soc/codecs/sgtl5000*
 
+NXP SJA1105 ETHERNET SWITCH DRIVER
+M:	Vladimir Oltean <olteanv@gmail.com>
+L:	linux-kernel@vger.kernel.org
+S:	Maintained
+F:	drivers/net/dsa/sja1105
+
 NXP TDA998X DRM DRIVER
 M:	Russell King <linux@armlinux.org.uk>
 S:	Maintained
diff --git a/drivers/net/dsa/Kconfig b/drivers/net/dsa/Kconfig
index 71bb3aebded4..d38e7e00c4e8 100644
--- a/drivers/net/dsa/Kconfig
+++ b/drivers/net/dsa/Kconfig
@@ -51,6 +51,8 @@ source "drivers/net/dsa/microchip/Kconfig"
 
 source "drivers/net/dsa/mv88e6xxx/Kconfig"
 
+source "drivers/net/dsa/sja1105/Kconfig"
+
 config NET_DSA_QCA8K
 	tristate "Qualcomm Atheros QCA8K Ethernet switch family support"
 	depends on NET_DSA
diff --git a/drivers/net/dsa/Makefile b/drivers/net/dsa/Makefile
index 82e5d794c41f..fefb6aaa82ba 100644
--- a/drivers/net/dsa/Makefile
+++ b/drivers/net/dsa/Makefile
@@ -18,3 +18,4 @@ obj-$(CONFIG_NET_DSA_VITESSE_VSC73XX) += vitesse-vsc73xx.o
 obj-y				+= b53/
 obj-y				+= microchip/
 obj-y				+= mv88e6xxx/
+obj-y				+= sja1105/
diff --git a/drivers/net/dsa/sja1105/Kconfig b/drivers/net/dsa/sja1105/Kconfig
new file mode 100644
index 000000000000..e6bb6579a809
--- /dev/null
+++ b/drivers/net/dsa/sja1105/Kconfig
@@ -0,0 +1,17 @@
+config NET_DSA_SJA1105
+tristate "NXP SJA1105 Ethernet switch family support"
+	depends on NET_DSA && SPI
+	select NET_DSA_TAG_SJA1105
+	select NET_DSA_TAG_8021Q
+	help
+	  This is the driver for the NXP SJA1105 automotive Ethernet switch
+	  family. These are 5-port devices and are managed over an SPI
+	  interface. Probing is handled based on OF bindings and so is the
+	  linkage to phylib. The driver supports the following revisions:
+	    - SJA1105E (Gen. 1, No TT-Ethernet)
+	    - SJA1105T (Gen. 1, TT-Ethernet)
+	    - SJA1105P (Gen. 2, No SGMII, No TT-Ethernet)
+	    - SJA1105Q (Gen. 2, No SGMII, TT-Ethernet)
+	    - SJA1105R (Gen. 2, SGMII, No TT-Ethernet)
+	    - SJA1105S (Gen. 2, SGMII, TT-Ethernet)
+
diff --git a/drivers/net/dsa/sja1105/Makefile b/drivers/net/dsa/sja1105/Makefile
new file mode 100644
index 000000000000..ed00840802f4
--- /dev/null
+++ b/drivers/net/dsa/sja1105/Makefile
@@ -0,0 +1,9 @@
+obj-$(CONFIG_NET_DSA_SJA1105) += sja1105.o
+
+sja1105-objs := \
+    sja1105_spi.o \
+    sja1105_main.o \
+    sja1105_clocking.o \
+    sja1105_static_config.o \
+    sja1105_dynamic_config.o \
+
diff --git a/drivers/net/dsa/sja1105/sja1105.h b/drivers/net/dsa/sja1105/sja1105.h
new file mode 100644
index 000000000000..2874423a5cd0
--- /dev/null
+++ b/drivers/net/dsa/sja1105/sja1105.h
@@ -0,0 +1,134 @@
+/* SPDX-License-Identifier: GPL-2.0
+ * Copyright (c) 2018, Sensor-Technik Wiedemann GmbH
+ * Copyright (c) 2018-2019, Vladimir Oltean <olteanv@gmail.com>
+ */
+#ifndef _SJA1105_H
+#define _SJA1105_H
+
+#include <net/dsa.h>
+#include "sja1105_static_config.h"
+
+/* IEEE 802.3 Annex 57A: Slow Protocols PDUs (01:80:C2:xx:xx:xx) */
+#define SJA1105_LINKLOCAL_FILTER_A	0x0180C2000000ull
+#define SJA1105_LINKLOCAL_FILTER_A_MASK	0xFFFFFF000000ull
+/* IEEE 1588 Annex F: Transport of PTP over Ethernet (01:1B:19:xx:xx:xx) */
+#define SJA1105_LINKLOCAL_FILTER_B	0x011B19000000ull
+#define SJA1105_LINKLOCAL_FILTER_B_MASK	0xFFFFFF000000ull
+
+#define SJA1105_NUM_PORTS 5
+#define SJA1105_NUM_TC    8
+#define SJA1105ET_FDB_BIN_SIZE 4
+
+/* Keeps the different addresses between E/T and P/Q/R/S */
+struct sja1105_regs {
+	u64 general_status;
+	u64 rgu;
+	u64 config;
+	u64 rmii_pll1;
+	u64 pad_mii_tx[SJA1105_NUM_PORTS];
+	u64 cgu_idiv[SJA1105_NUM_PORTS];
+	u64 rgmii_pad_mii_tx[SJA1105_NUM_PORTS];
+	u64 mii_tx_clk[SJA1105_NUM_PORTS];
+	u64 mii_rx_clk[SJA1105_NUM_PORTS];
+	u64 mii_ext_tx_clk[SJA1105_NUM_PORTS];
+	u64 mii_ext_rx_clk[SJA1105_NUM_PORTS];
+	u64 rgmii_txc[SJA1105_NUM_PORTS];
+	u64 rmii_ref_clk[SJA1105_NUM_PORTS];
+	u64 rmii_ext_tx_clk[SJA1105_NUM_PORTS];
+	u64 mac[SJA1105_NUM_PORTS];
+	u64 mac_hl1[SJA1105_NUM_PORTS];
+	u64 mac_hl2[SJA1105_NUM_PORTS];
+	u64 qlevel[SJA1105_NUM_PORTS];
+};
+
+struct sja1105_private {
+	const struct sja1105_dynamic_table_ops *dyn_ops;
+	struct sja1105_static_config static_config;
+	struct gpio_desc *reset_gpio;
+	struct spi_device *spidev;
+	struct sja1105_regs *regs;
+	struct dsa_switch *ds;
+	u64 device_id;
+	u64 part_nr; /* Needed for P/R distinction (same switch core) */
+};
+
+#include "sja1105_dynamic_config.h"
+
+struct sja1105_spi_message {
+	u64 access;
+	u64 read_count;
+	u64 address;
+};
+
+enum sja1105_spi_access_mode {
+	SPI_READ = 0,
+	SPI_WRITE = 1,
+};
+
+/* From sja1105-spi.c */
+int
+sja1105_spi_send_packed_buf(const struct sja1105_private *priv,
+			    enum sja1105_spi_access_mode read_or_write,
+			    u64 reg_addr, void *packed_buf, size_t size_bytes);
+int sja1105_spi_send_int(const struct sja1105_private *priv,
+			 enum sja1105_spi_access_mode read_or_write,
+			 u64 reg_addr, u64 *value, u64 size_bytes);
+int
+sja1105_spi_send_long_packed_buf(const struct sja1105_private *priv,
+				 enum sja1105_spi_access_mode read_or_write,
+				 u64 base_addr, void *packed_buf, u64 buf_len);
+int sja1105_static_config_upload(struct sja1105_private *priv);
+int sja1105_device_id_get(struct sja1105_private *priv);
+const char *sja1105_device_id_string_get(u64 device_id, u64 part_nr);
+
+#define SIZE_SPI_MSG_HEADER    4
+#define SIZE_SPI_MSG_MAXLEN    (64 * 4)
+
+/* From sja1105-clocking.c */
+
+typedef enum {
+	XMII_MAC = 0,
+	XMII_PHY = 1,
+} sja1105_mii_mode_t;
+
+typedef enum {
+	XMII_MODE_MII		= 0,
+	XMII_MODE_RMII		= 1,
+	XMII_MODE_RGMII		= 2,
+	XMII_MODE_TRISTATE	= 3,
+} sja1105_phy_interface_t;
+
+typedef enum {
+	SJA1105_SPEED_10MBPS	= 3,
+	SJA1105_SPEED_100MBPS	= 2,
+	SJA1105_SPEED_1000MBPS	= 1,
+	SJA1105_SPEED_AUTO	= 0,
+} sja1105_speed_t;
+
+int sja1105_clocking_setup_port(struct sja1105_private *priv, int port);
+int sja1105_clocking_setup(struct sja1105_private *priv);
+
+/* From sja1105-dynamic-config.c */
+
+int sja1105_dynamic_config_read(struct sja1105_private *priv,
+				enum sja1105_blk_idx blk_idx,
+				int index, void *entry);
+int sja1105_dynamic_config_write(struct sja1105_private *priv,
+				 enum sja1105_blk_idx blk_idx,
+				 int index, void *entry, bool keep);
+void sja1105_dynamic_config_init(struct sja1105_private *priv);
+
+/* Common implementations for the static and dynamic configs */
+size_t sja1105_l2_forwarding_entry_packing(void *buf, void *entry_ptr,
+					   enum packing_op op);
+size_t sja1105pqrs_l2_lookup_entry_packing(void *buf, void *entry_ptr,
+					   enum packing_op op);
+size_t sja1105et_l2_lookup_entry_packing(void *buf, void *entry_ptr,
+					 enum packing_op op);
+size_t sja1105_vlan_lookup_entry_packing(void *buf, void *entry_ptr,
+					 enum packing_op op);
+size_t sja1105pqrs_mac_config_entry_packing(void *buf, void *entry_ptr,
+					    enum packing_op op);
+
+#endif
+
diff --git a/drivers/net/dsa/sja1105/sja1105_clocking.c b/drivers/net/dsa/sja1105/sja1105_clocking.c
new file mode 100644
index 000000000000..62259571e374
--- /dev/null
+++ b/drivers/net/dsa/sja1105/sja1105_clocking.c
@@ -0,0 +1,643 @@
+// SPDX-License-Identifier: BSD-3-Clause
+/* Copyright (c) 2016-2018, NXP Semiconductors
+ * Copyright (c) 2018-2019, Vladimir Oltean <olteanv@gmail.com>
+ */
+#include <linux/packing.h>
+#include "sja1105.h"
+
+struct sja1105_cfg_pad_mii_tx {
+	u64 d32_os;
+	u64 d32_ipud;
+	u64 d10_os;
+	u64 d10_ipud;
+	u64 ctrl_os;
+	u64 ctrl_ipud;
+	u64 clk_os;
+	u64 clk_ih;
+	u64 clk_ipud;
+};
+
+/* UM10944 Table 82.
+ * IDIV_0_C to IDIV_4_C control registers
+ * (addr. 10000Bh to 10000Fh)
+ */
+struct sja1105_cgu_idiv {
+	u64 clksrc;
+	u64 autoblock;
+	u64 idiv;
+	u64 pd;
+};
+
+/* UM10944 Table 80.
+ * PLL_x_S clock status registers 0 and 1
+ * (address 100007h and 100009h)
+ */
+struct sja1105_cgu_pll_status {
+	u64 lock;
+};
+
+/* PLL_1_C control register
+ *
+ * SJA1105 E/T: UM10944 Table 81 (address 10000Ah)
+ * SJA1105 P/Q/R/S: UM11040 Table 116 (address 10000Ah)
+ */
+struct sja1105_cgu_pll_ctrl {
+	u64 pllclksrc;
+	u64 msel;
+	u64 nsel; /* Only for P/Q/R/S series */
+	u64 autoblock;
+	u64 psel;
+	u64 direct;
+	u64 fbsel;
+	u64 p23en; /* Only for P/Q/R/S series */
+	u64 bypass;
+	u64 pd;
+};
+
+#define CLKSRC_MII0_TX_CLK 0x00
+#define CLKSRC_MII0_RX_CLK 0x01
+#define CLKSRC_MII1_TX_CLK 0x02
+#define CLKSRC_MII1_RX_CLK 0x03
+#define CLKSRC_MII2_TX_CLK 0x04
+#define CLKSRC_MII2_RX_CLK 0x05
+#define CLKSRC_MII3_TX_CLK 0x06
+#define CLKSRC_MII3_RX_CLK 0x07
+#define CLKSRC_MII4_TX_CLK 0x08
+#define CLKSRC_MII4_RX_CLK 0x09
+#define CLKSRC_PLL0        0x0B
+#define CLKSRC_PLL1        0x0E
+#define CLKSRC_IDIV0       0x11
+#define CLKSRC_IDIV1       0x12
+#define CLKSRC_IDIV2       0x13
+#define CLKSRC_IDIV3       0x14
+#define CLKSRC_IDIV4       0x15
+
+/* UM10944 Table 83.
+ * MIIx clock control registers 1 to 30
+ * (addresses 100013h to 100035h)
+ */
+struct sja1105_cgu_mii_ctrl {
+	u64 clksrc;
+	u64 autoblock;
+	u64 pd;
+};
+
+static void sja1105_cgu_idiv_packing(void *buf, struct sja1105_cgu_idiv *idiv,
+				     enum packing_op op)
+{
+	const int size = 4;
+
+	if (op == UNPACK)
+		memset(idiv, 0, sizeof(*idiv));
+	else
+		memset(buf, 0, size);
+
+	sja1105_packing(buf, &idiv->clksrc,    28, 24, size, op);
+	sja1105_packing(buf, &idiv->autoblock, 11, 11, size, op);
+	sja1105_packing(buf, &idiv->idiv,       5,  2, size, op);
+	sja1105_packing(buf, &idiv->pd,         0,  0, size, op);
+}
+
+static int sja1105_cgu_idiv_config(struct sja1105_private *priv, int port,
+				   bool enabled, int factor)
+{
+#define BUF_LEN 4
+	struct device *dev = priv->ds->dev;
+	struct sja1105_cgu_idiv idiv;
+	u8 packed_buf[BUF_LEN];
+
+	if (enabled && factor != 1 && factor != 10) {
+		dev_err(dev, "idiv factor must be 1 or 10\n");
+		return -ERANGE;
+	}
+
+	/* Payload for packed_buf */
+	idiv.clksrc    = 0x0A;            /* 25MHz */
+	idiv.autoblock = 1;               /* Block clk automatically */
+	idiv.idiv      = factor - 1;      /* Divide by 1 or 10 */
+	idiv.pd        = enabled ? 0 : 1; /* Power down? */
+	sja1105_cgu_idiv_packing(packed_buf, &idiv, PACK);
+
+	return sja1105_spi_send_packed_buf(priv, SPI_WRITE,
+					   priv->regs->cgu_idiv[port],
+					   packed_buf, BUF_LEN);
+#undef BUF_LEN
+}
+
+static void
+sja1105_cgu_mii_control_packing(void *buf, struct sja1105_cgu_mii_ctrl *cmd,
+				enum packing_op op)
+{
+	const int size = 4;
+
+	if (op == UNPACK)
+		memset(cmd, 0, sizeof(*cmd));
+	else
+		memset(buf, 0, size);
+
+	sja1105_packing(buf, &cmd->clksrc,    28, 24, 4, op);
+	sja1105_packing(buf, &cmd->autoblock, 11, 11, 4, op);
+	sja1105_packing(buf, &cmd->pd,         0,  0, 4, op);
+}
+
+static int sja1105_cgu_mii_tx_clk_config(struct sja1105_private *priv,
+					 int port, sja1105_mii_mode_t mode)
+{
+#define BUF_LEN 4
+	u8 packed_buf[BUF_LEN];
+	struct  sja1105_cgu_mii_ctrl mii_tx_clk;
+	const int mac_clk_sources[] = {
+		CLKSRC_MII0_TX_CLK,
+		CLKSRC_MII1_TX_CLK,
+		CLKSRC_MII2_TX_CLK,
+		CLKSRC_MII3_TX_CLK,
+		CLKSRC_MII4_TX_CLK,
+	};
+	const int phy_clk_sources[] = {
+		CLKSRC_IDIV0,
+		CLKSRC_IDIV1,
+		CLKSRC_IDIV2,
+		CLKSRC_IDIV3,
+		CLKSRC_IDIV4,
+	};
+	int clksrc;
+
+	if (mode == XMII_MAC)
+		clksrc = mac_clk_sources[port];
+	else
+		clksrc = phy_clk_sources[port];
+
+	/* Payload for packed_buf */
+	mii_tx_clk.clksrc    = clksrc;
+	mii_tx_clk.autoblock = 1;  /* Autoblock clk while changing clksrc */
+	mii_tx_clk.pd        = 0;  /* Power Down off => enabled */
+	sja1105_cgu_mii_control_packing(packed_buf, &mii_tx_clk, PACK);
+
+	return sja1105_spi_send_packed_buf(priv, SPI_WRITE,
+					   priv->regs->mii_tx_clk[port],
+					   packed_buf, BUF_LEN);
+#undef BUF_LEN
+}
+
+static int
+sja1105_cgu_mii_rx_clk_config(struct sja1105_private *priv, int port)
+{
+#define BUF_LEN 4
+	u8 packed_buf[BUF_LEN];
+	struct  sja1105_cgu_mii_ctrl mii_rx_clk;
+	const int clk_sources[] = {
+		CLKSRC_MII0_RX_CLK,
+		CLKSRC_MII1_RX_CLK,
+		CLKSRC_MII2_RX_CLK,
+		CLKSRC_MII3_RX_CLK,
+		CLKSRC_MII4_RX_CLK,
+	};
+
+	/* Payload for packed_buf */
+	mii_rx_clk.clksrc    = clk_sources[port];
+	mii_rx_clk.autoblock = 1;  /* Autoblock clk while changing clksrc */
+	mii_rx_clk.pd        = 0;  /* Power Down off => enabled */
+	sja1105_cgu_mii_control_packing(packed_buf, &mii_rx_clk, PACK);
+
+	return sja1105_spi_send_packed_buf(priv, SPI_WRITE,
+					   priv->regs->mii_rx_clk[port],
+					   packed_buf, BUF_LEN);
+#undef BUF_LEN
+}
+
+static int
+sja1105_cgu_mii_ext_tx_clk_config(struct sja1105_private *priv, int port)
+{
+#define BUF_LEN 4
+	u8 packed_buf[BUF_LEN];
+	struct  sja1105_cgu_mii_ctrl mii_ext_tx_clk;
+	const int clk_sources[] = {
+		CLKSRC_IDIV0,
+		CLKSRC_IDIV1,
+		CLKSRC_IDIV2,
+		CLKSRC_IDIV3,
+		CLKSRC_IDIV4,
+	};
+
+	/* Payload for packed_buf */
+	mii_ext_tx_clk.clksrc    = clk_sources[port];
+	mii_ext_tx_clk.autoblock = 1; /* Autoblock clk while changing clksrc */
+	mii_ext_tx_clk.pd        = 0; /* Power Down off => enabled */
+	sja1105_cgu_mii_control_packing(packed_buf, &mii_ext_tx_clk, PACK);
+
+	return sja1105_spi_send_packed_buf(priv, SPI_WRITE,
+					   priv->regs->mii_ext_tx_clk[port],
+					   packed_buf, BUF_LEN);
+#undef BUF_LEN
+}
+
+static int
+sja1105_cgu_mii_ext_rx_clk_config(struct sja1105_private *priv, int port)
+{
+#define BUF_LEN 4
+	u8 packed_buf[BUF_LEN];
+	struct  sja1105_cgu_mii_ctrl mii_ext_rx_clk;
+	const int clk_sources[] = {
+		CLKSRC_IDIV0,
+		CLKSRC_IDIV1,
+		CLKSRC_IDIV2,
+		CLKSRC_IDIV3,
+		CLKSRC_IDIV4,
+	};
+
+	/* Payload for packed_buf */
+	mii_ext_rx_clk.clksrc    = clk_sources[port];
+	mii_ext_rx_clk.autoblock = 1; /* Autoblock clk while changing clksrc */
+	mii_ext_rx_clk.pd        = 0; /* Power Down off => enabled */
+	sja1105_cgu_mii_control_packing(packed_buf, &mii_ext_rx_clk, PACK);
+
+	return sja1105_spi_send_packed_buf(priv, SPI_WRITE,
+					   priv->regs->mii_ext_rx_clk[port],
+					   packed_buf, BUF_LEN);
+#undef BUF_LEN
+}
+
+static int mii_clocking_setup(struct sja1105_private *priv, int port,
+			      sja1105_mii_mode_t mode)
+{
+	struct device *dev = priv->ds->dev;
+	int rc;
+
+	dev_dbg(dev, "Configuring MII-%s clocking\n",
+		(mode == XMII_MAC) ? "MAC" : "PHY");
+	/* If mode is MAC, disable IDIV
+	 * If mode is PHY, enable IDIV and configure for 1/1 divider
+	 */
+	rc = sja1105_cgu_idiv_config(priv, port, (mode == XMII_PHY), 1);
+	if (rc < 0)
+		return rc;
+
+	/* Configure CLKSRC of MII_TX_CLK_n
+	 *   * If mode is MAC, select TX_CLK_n
+	 *   * If mode is PHY, select IDIV_n
+	 */
+	rc = sja1105_cgu_mii_tx_clk_config(priv, port, mode);
+	if (rc < 0)
+		return rc;
+
+	/* Configure CLKSRC of MII_RX_CLK_n
+	 * Select RX_CLK_n
+	 */
+	rc = sja1105_cgu_mii_rx_clk_config(priv, port);
+	if (rc < 0)
+		return rc;
+
+	if (mode == XMII_PHY) {
+		/* Per MII spec, the PHY (which is us) drives the TX_CLK pin */
+
+		/* Configure CLKSRC of EXT_TX_CLK_n
+		 * Select IDIV_n
+		 */
+		rc = sja1105_cgu_mii_ext_tx_clk_config(priv, port);
+		if (rc < 0)
+			return rc;
+
+		/* Configure CLKSRC of EXT_RX_CLK_n
+		 * Select IDIV_n
+		 */
+		rc = sja1105_cgu_mii_ext_rx_clk_config(priv, port);
+		if (rc < 0)
+			return rc;
+	}
+	return 0;
+}
+
+static void
+sja1105_cgu_pll_control_packing(void *buf, struct sja1105_cgu_pll_ctrl *cmd,
+				enum packing_op op)
+{
+	const int size = 4;
+
+	if (op == UNPACK)
+		memset(cmd, 0, sizeof(*cmd));
+	else
+		memset(buf, 0, size);
+
+	sja1105_packing(buf, &cmd->pllclksrc, 28, 24, size, op);
+	sja1105_packing(buf, &cmd->msel,      23, 16, size, op);
+	sja1105_packing(buf, &cmd->autoblock, 11, 11, size, op);
+	sja1105_packing(buf, &cmd->psel,       9,  8, size, op);
+	sja1105_packing(buf, &cmd->direct,     7,  7, size, op);
+	sja1105_packing(buf, &cmd->fbsel,      6,  6, size, op);
+	sja1105_packing(buf, &cmd->bypass,     1,  1, size, op);
+	sja1105_packing(buf, &cmd->pd,         0,  0, size, op);
+	/* P/Q/R/S only, but packing zeroes for E/T doesn't hurt */
+	sja1105_packing(buf, &cmd->nsel,      13, 12, size, op);
+	sja1105_packing(buf, &cmd->p23en,      2,  2, size, op);
+}
+
+static int sja1105_cgu_rgmii_tx_clk_config(struct sja1105_private *priv,
+					   int port, sja1105_speed_t speed)
+{
+#define BUF_LEN 4
+	struct sja1105_cgu_mii_ctrl txc;
+	u8 packed_buf[BUF_LEN];
+	int clksrc;
+
+	if (speed == SJA1105_SPEED_1000MBPS) {
+		clksrc = CLKSRC_PLL0;
+	} else {
+		int clk_sources[] = {CLKSRC_IDIV0, CLKSRC_IDIV1, CLKSRC_IDIV2,
+				     CLKSRC_IDIV3, CLKSRC_IDIV4};
+		clksrc = clk_sources[port];
+	}
+
+	/* RGMII: 125MHz for 1000, 25MHz for 100, 2.5MHz for 10 */
+	txc.clksrc = clksrc;
+	/* Autoblock clk while changing clksrc */
+	txc.autoblock = 1;
+	/* Power Down off => enabled */
+	txc.pd = 0;
+	sja1105_cgu_mii_control_packing(packed_buf, &txc, PACK);
+
+	return sja1105_spi_send_packed_buf(priv, SPI_WRITE,
+					   priv->regs->rgmii_txc[port],
+					   packed_buf, BUF_LEN);
+#undef BUF_LEN
+}
+
+/* AGU */
+static void
+sja1105_cfg_pad_mii_tx_packing(void *buf, struct sja1105_cfg_pad_mii_tx *cmd,
+			       enum packing_op op)
+{
+	const int size = 4;
+
+	if (op == UNPACK)
+		memset(cmd, 0, sizeof(*cmd));
+	else
+		memset(buf, 0, size);
+
+	sja1105_packing(buf, &cmd->d32_os,   28, 27, size, op);
+	sja1105_packing(buf, &cmd->d32_ipud, 25, 24, size, op);
+	sja1105_packing(buf, &cmd->d10_os,   20, 19, size, op);
+	sja1105_packing(buf, &cmd->d10_ipud, 17, 16, size, op);
+	sja1105_packing(buf, &cmd->ctrl_os,  12, 11, size, op);
+	sja1105_packing(buf, &cmd->ctrl_ipud, 9,  8, size, op);
+	sja1105_packing(buf, &cmd->clk_os,    4,  3, size, op);
+	sja1105_packing(buf, &cmd->clk_ih,    2,  2, size, op);
+	sja1105_packing(buf, &cmd->clk_ipud,  1,  0, size, op);
+}
+
+static int sja1105_rgmii_cfg_pad_tx_config(struct sja1105_private *priv,
+					   int port)
+{
+#define BUF_LEN 4
+	u8 packed_buf[BUF_LEN];
+	struct  sja1105_cfg_pad_mii_tx pad_mii_tx;
+
+	/* Payload */
+	pad_mii_tx.d32_os    = 3; /* TXD[3:2] output stage: */
+				  /*          high noise/high speed */
+	pad_mii_tx.d32_ipud  = 2; /* TXD[3:2] input stage: */
+				  /*          plain input (default) */
+	pad_mii_tx.d10_os    = 3; /* TXD[1:0] output stage: */
+				  /*          high noise/high speed */
+	pad_mii_tx.d10_ipud  = 2; /* TXD[1:0] input stage: */
+				  /*          plain input (default) */
+	pad_mii_tx.ctrl_os   = 3; /* TX_CTL / TX_ER output stage */
+	pad_mii_tx.ctrl_ipud = 2; /* TX_CTL / TX_ER input stage (default) */
+	pad_mii_tx.clk_os    = 3; /* TX_CLK output stage */
+	pad_mii_tx.clk_ih    = 0; /* TX_CLK input hysteresis (default) */
+	pad_mii_tx.clk_ipud  = 2; /* TX_CLK input stage (default) */
+	sja1105_cfg_pad_mii_tx_packing(packed_buf, &pad_mii_tx, PACK);
+
+	return sja1105_spi_send_packed_buf(priv, SPI_WRITE,
+					   priv->regs->rgmii_pad_mii_tx[port],
+					   packed_buf, BUF_LEN);
+#undef BUF_LEN
+}
+
+static int rgmii_clocking_setup(struct sja1105_private *priv, int port)
+{
+	struct device *dev = priv->ds->dev;
+	struct sja1105_table *mac;
+	sja1105_speed_t speed;
+	int rc;
+
+	mac = &priv->static_config.tables[BLK_IDX_MAC_CONFIG];
+	speed = ((struct sja1105_mac_config_entry *)mac->entries)[port].speed;
+
+	dev_dbg(dev, "Configuring port %d RGMII at speed %dMbps\n",
+		port, speed);
+
+	switch (speed) {
+	case SJA1105_SPEED_1000MBPS:
+		/* 1000Mbps, IDIV disabled, divide by 1 */
+		rc = sja1105_cgu_idiv_config(priv, port, false, 1);
+		break;
+	case SJA1105_SPEED_100MBPS:
+		/* 100Mbps, IDIV enabled, divide by 1 */
+		rc = sja1105_cgu_idiv_config(priv, port, true, 1);
+		break;
+	case SJA1105_SPEED_10MBPS:
+		/* 10Mbps, IDIV enabled, divide by 10 */
+		rc = sja1105_cgu_idiv_config(priv, port, true, 10);
+		break;
+	case SJA1105_SPEED_AUTO:
+		/* Skip CGU configuration if there is no speed available
+		 * (e.g. link is not established yet)
+		 */
+		dev_dbg(dev, "Speed not available, skipping CGU config\n");
+		return 0;
+	default:
+		rc = -EINVAL;
+	}
+
+	if (rc < 0) {
+		dev_err(dev, "Failed to configure idiv\n");
+		return rc;
+	}
+	rc = sja1105_cgu_rgmii_tx_clk_config(priv, port, speed);
+	if (rc < 0) {
+		dev_err(dev, "Failed to configure RGMII Tx clock\n");
+		return rc;
+	}
+	rc = sja1105_rgmii_cfg_pad_tx_config(priv, port);
+	if (rc < 0) {
+		dev_err(dev, "Failed to configure Tx pad registers\n");
+		return rc;
+	}
+	return 0;
+}
+
+static int sja1105_cgu_rmii_ref_clk_config(struct sja1105_private *priv,
+					   int port)
+{
+#define BUF_LEN 4
+	struct  sja1105_cgu_mii_ctrl ref_clk;
+	u8 packed_buf[BUF_LEN];
+	const int clk_sources[] = {
+		CLKSRC_MII0_TX_CLK,
+		CLKSRC_MII1_TX_CLK,
+		CLKSRC_MII2_TX_CLK,
+		CLKSRC_MII3_TX_CLK,
+		CLKSRC_MII4_TX_CLK,
+	};
+
+	/* Payload for packed_buf */
+	ref_clk.clksrc    = clk_sources[port];
+	ref_clk.autoblock = 1;      /* Autoblock clk while changing clksrc */
+	ref_clk.pd        = 0;      /* Power Down off => enabled */
+	sja1105_cgu_mii_control_packing(packed_buf, &ref_clk, PACK);
+
+	return sja1105_spi_send_packed_buf(priv, SPI_WRITE,
+					   priv->regs->rmii_ref_clk[port],
+					   packed_buf, BUF_LEN);
+#undef BUF_LEN
+}
+
+static int
+sja1105_cgu_rmii_ext_tx_clk_config(struct sja1105_private *priv, int port)
+{
+#define BUF_LEN 4
+	struct sja1105_cgu_mii_ctrl ext_tx_clk;
+	u8 packed_buf[BUF_LEN];
+
+	/* Payload for packed_buf */
+	ext_tx_clk.clksrc    = CLKSRC_PLL1;
+	ext_tx_clk.autoblock = 1;   /* Autoblock clk while changing clksrc */
+	ext_tx_clk.pd        = 0;   /* Power Down off => enabled */
+	sja1105_cgu_mii_control_packing(packed_buf, &ext_tx_clk, PACK);
+
+	return sja1105_spi_send_packed_buf(priv, SPI_WRITE,
+					   priv->regs->rmii_ext_tx_clk[port],
+					   packed_buf, BUF_LEN);
+#undef BUF_LEN
+}
+
+static int sja1105_cgu_rmii_pll_config(struct sja1105_private *priv)
+{
+#define BUF_LEN 4
+	struct device *dev = priv->ds->dev;
+	struct sja1105_cgu_pll_ctrl pll;
+	u8 packed_buf[BUF_LEN];
+	int rc;
+
+	/* PLL1 must be enabled and output 50 Mhz.
+	 * This is done by writing first 0x0A010941 to
+	 * the PLL_1_C register and then deasserting
+	 * power down (PD) 0x0A010940.
+	 */
+
+	/* Step 1: PLL1 setup for 50Mhz */
+	pll.pllclksrc = 0xA;
+	pll.msel      = 0x1;
+	pll.autoblock = 0x1;
+	pll.psel      = 0x1;
+	pll.direct    = 0x0;
+	pll.fbsel     = 0x1;
+	pll.bypass    = 0x0;
+	pll.pd        = 0x1;
+	/* P/Q/R/S only */
+	pll.nsel      = 0x0; /* PLL pre-divider is 1 (nsel + 1) */
+	pll.p23en     = 0x0; /* disable 120 and 240 degree phase PLL outputs */
+
+	sja1105_cgu_pll_control_packing(packed_buf, &pll, PACK);
+	rc = sja1105_spi_send_packed_buf(priv, SPI_WRITE,
+					 priv->regs->rmii_pll1,
+					 packed_buf, BUF_LEN);
+	if (rc < 0) {
+		dev_err(dev, "failed to configure PLL1 for 50MHz\n");
+		return rc;
+	}
+
+	/* Step 2: Enable PLL1 */
+	pll.pd = 0x0;
+
+	sja1105_cgu_pll_control_packing(packed_buf, &pll, PACK);
+	rc = sja1105_spi_send_packed_buf(priv, SPI_WRITE,
+					 priv->regs->rmii_pll1,
+					 packed_buf, BUF_LEN);
+	if (rc < 0) {
+		dev_err(dev, "failed to enable PLL1\n");
+		return rc;
+	}
+	return rc;
+#undef BUF_LEN
+}
+
+static int rmii_clocking_setup(struct sja1105_private *priv, int port,
+			       sja1105_mii_mode_t mode)
+{
+	struct device *dev = priv->ds->dev;
+	int rc;
+
+	dev_dbg(dev, "Configuring RMII-%s clocking\n",
+		(mode == XMII_MAC) ? "MAC" : "PHY");
+	/* AH1601.pdf chapter 2.5.1. Sources */
+	if (mode == XMII_MAC) {
+		/* Configure and enable PLL1 for 50Mhz output */
+		rc = sja1105_cgu_rmii_pll_config(priv);
+		if (rc < 0)
+			return rc;
+	}
+	/* Disable IDIV for this port */
+	rc = sja1105_cgu_idiv_config(priv, port, false, 1);
+	if (rc < 0)
+		return rc;
+	/* Source to sink mappings */
+	rc = sja1105_cgu_rmii_ref_clk_config(priv, port);
+	if (rc < 0)
+		return rc;
+	if (mode == XMII_MAC) {
+		rc = sja1105_cgu_rmii_ext_tx_clk_config(priv, port);
+		if (rc < 0)
+			return rc;
+	}
+	return 0;
+}
+
+int sja1105_clocking_setup_port(struct sja1105_private *priv, int port)
+{
+	struct sja1105_xmii_params_entry *mii;
+	struct device *dev = priv->ds->dev;
+	sja1105_phy_interface_t phy_mode;
+	sja1105_mii_mode_t mode;
+	int rc;
+
+	mii = priv->static_config.tables[BLK_IDX_XMII_PARAMS].entries;
+
+	/* RGMII etc */
+	phy_mode = mii->xmii_mode[port];
+	/* MAC or PHY, for applicable types (not RGMII) */
+	mode = mii->phy_mac[port];
+
+	switch (phy_mode) {
+	case XMII_MODE_MII:
+		rc = mii_clocking_setup(priv, port, mode);
+		break;
+	case XMII_MODE_RMII:
+		rc = rmii_clocking_setup(priv, port, mode);
+		break;
+	case XMII_MODE_RGMII:
+		rc = rgmii_clocking_setup(priv, port);
+		break;
+	default:
+		dev_err(dev, "Invalid interface mode specified: %d\n",
+			phy_mode);
+		return -EINVAL;
+	}
+	if (rc)
+		dev_err(dev, "Clocking setup for port %d failed: %d\n",
+			port, rc);
+	return rc;
+}
+
+int sja1105_clocking_setup(struct sja1105_private *priv)
+{
+	int port, rc;
+
+	for (port = 0; port < SJA1105_NUM_PORTS; port++) {
+		rc = sja1105_clocking_setup_port(priv, port);
+		if (rc < 0)
+			return rc;
+	}
+	return 0;
+}
+
diff --git a/drivers/net/dsa/sja1105/sja1105_dynamic_config.c b/drivers/net/dsa/sja1105/sja1105_dynamic_config.c
new file mode 100644
index 000000000000..4f3756e267ac
--- /dev/null
+++ b/drivers/net/dsa/sja1105/sja1105_dynamic_config.c
@@ -0,0 +1,478 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2018-2019, Vladimir Oltean <olteanv@gmail.com>
+ */
+#include "sja1105.h"
+
+#define SIZE_DYN_CMD                     4
+#define SIZE_MAC_CONFIG_DYN_ENTRY_ET     SIZE_DYN_CMD
+#define SIZE_L2_LOOKUP_DYN_CMD_ET       (SIZE_DYN_CMD + SIZE_L2_LOOKUP_ENTRY_ET)
+#define SIZE_L2_LOOKUP_DYN_CMD_PQRS     (SIZE_DYN_CMD + SIZE_L2_LOOKUP_ENTRY_PQRS)
+#define SIZE_VLAN_LOOKUP_DYN_CMD        (SIZE_DYN_CMD + 4 + SIZE_VLAN_LOOKUP_ENTRY)
+#define SIZE_L2_FORWARDING_DYN_CMD      (SIZE_DYN_CMD + SIZE_L2_FORWARDING_ENTRY)
+#define SIZE_MAC_CONFIG_DYN_CMD_ET      (SIZE_DYN_CMD + SIZE_MAC_CONFIG_DYN_ENTRY_ET)
+#define SIZE_MAC_CONFIG_DYN_CMD_PQRS    (SIZE_DYN_CMD + SIZE_MAC_CONFIG_ENTRY_PQRS)
+#define SIZE_L2_LOOKUP_PARAMS_DYN_CMD_ET SIZE_DYN_CMD
+#define SIZE_GENERAL_PARAMS_DYN_CMD_ET   SIZE_DYN_CMD
+#define MAX_DYN_CMD_SIZE                 SIZE_MAC_CONFIG_DYN_CMD_PQRS
+
+static void
+sja1105pqrs_l2_lookup_cmd_packing(void *buf, struct sja1105_dyn_cmd *cmd,
+				  enum packing_op op)
+{
+	u8 *p = buf + SIZE_L2_LOOKUP_ENTRY_PQRS;
+
+	sja1105_packing(p, &cmd->valid,    31, 31, SIZE_DYN_CMD, op);
+	sja1105_packing(p, &cmd->rdwrset,  30, 30, SIZE_DYN_CMD, op);
+	sja1105_packing(p, &cmd->errors,   29, 29, SIZE_DYN_CMD, op);
+	sja1105_packing(p, &cmd->valident, 27, 27, SIZE_DYN_CMD, op);
+	/* Hack - The hardware takes the 'index' field within
+	 * struct sja1105_l2_lookup_entry as the index on which this command
+	 * will operate. However it will ignore everything else, so 'index'
+	 * is logically part of command but physically part of entry.
+	 * Populate the 'index' entry field from within the command callback,
+	 * such that our API doesn't need to ask for a full-blown entry
+	 * structure when e.g. a delete is requested.
+	 */
+	sja1105_packing(buf, &cmd->index, 29, 20, SIZE_L2_LOOKUP_ENTRY_PQRS, op);
+	/* TODO hostcmd */
+}
+
+static void
+sja1105et_l2_lookup_cmd_packing(void *buf, struct sja1105_dyn_cmd *cmd,
+				enum packing_op op)
+{
+	u8 *p = buf + SIZE_L2_LOOKUP_ENTRY_ET;
+
+	sja1105_packing(p, &cmd->valid,    31, 31, SIZE_DYN_CMD, op);
+	sja1105_packing(p, &cmd->rdwrset,  30, 30, SIZE_DYN_CMD, op);
+	sja1105_packing(p, &cmd->errors,   29, 29, SIZE_DYN_CMD, op);
+	sja1105_packing(p, &cmd->valident, 27, 27, SIZE_DYN_CMD, op);
+	/* Hack - see comments above. */
+	sja1105_packing(buf, &cmd->index, 29, 20, SIZE_L2_LOOKUP_ENTRY_ET, op);
+}
+
+static void
+sja1105et_mgmt_route_cmd_packing(void *buf, struct sja1105_dyn_cmd *cmd,
+				 enum packing_op op)
+{
+	u8 *p = buf + SIZE_L2_LOOKUP_ENTRY_ET;
+	u64 mgmtroute = 1;
+
+	sja1105et_l2_lookup_cmd_packing(buf, cmd, op);
+	if (op == PACK)
+		sja1105_pack(p, &mgmtroute, 26, 26, SIZE_DYN_CMD);
+}
+
+static size_t sja1105et_mgmt_route_entry_packing(void *buf, void *entry_ptr,
+						 enum packing_op op)
+{
+	struct sja1105_mgmt_entry *entry = entry_ptr;
+	const size_t size = SIZE_L2_LOOKUP_ENTRY_ET;
+
+	/* UM10944: To specify if a PTP egress timestamp shall be captured on
+	 * each port upon transmission of the frame, the LSB of VLANID in the
+	 * ENTRY field provided by the host must be set.
+	 * Bit 1 of VLANID then specifies the register where the timestamp for
+	 * this port is stored in.
+	 */
+	sja1105_packing(buf, &entry->tsreg,     85, 85, size, op);
+	sja1105_packing(buf, &entry->takets,    84, 84, size, op);
+	sja1105_packing(buf, &entry->macaddr,   83, 36, size, op);
+	sja1105_packing(buf, &entry->destports, 35, 31, size, op);
+	sja1105_packing(buf, &entry->enfport,   30, 30, size, op);
+	return size;
+}
+
+/* In E/T, entry is at addresses 0x27-0x28. There is a 4 byte gap at 0x29,
+ * and command is at 0x2a. Similarly in P/Q/R/S there is a 1 register gap
+ * between entry (0x2d, 0x2e) and command (0x30).
+ */
+static void
+sja1105_vlan_lookup_cmd_packing(void *buf, struct sja1105_dyn_cmd *cmd,
+				enum packing_op op)
+{
+	u8 *p = buf + SIZE_VLAN_LOOKUP_ENTRY + 4;
+
+	sja1105_packing(p, &cmd->valid,    31, 31, SIZE_DYN_CMD, op);
+	sja1105_packing(p, &cmd->rdwrset,  30, 30, SIZE_DYN_CMD, op);
+	sja1105_packing(p, &cmd->valident, 27, 27, SIZE_DYN_CMD, op);
+	/* Hack - see comments above, applied for 'vlanid' field of
+	 * struct sja1105_vlan_lookup_entry.
+	 */
+	sja1105_packing(buf, &cmd->index, 38, 27, SIZE_VLAN_LOOKUP_ENTRY, op);
+}
+
+static void
+sja1105_l2_forwarding_cmd_packing(void *buf, struct sja1105_dyn_cmd *cmd,
+				  enum packing_op op)
+{
+	u8 *p = buf + SIZE_L2_FORWARDING_ENTRY;
+
+	sja1105_packing(p, &cmd->valid,   31, 31, SIZE_DYN_CMD, op);
+	sja1105_packing(p, &cmd->errors,  30, 30, SIZE_DYN_CMD, op);
+	sja1105_packing(p, &cmd->rdwrset, 29, 29, SIZE_DYN_CMD, op);
+	sja1105_packing(p, &cmd->index,    4,  0, SIZE_DYN_CMD, op);
+}
+
+static void
+sja1105et_mac_config_cmd_packing(void *buf, struct sja1105_dyn_cmd *cmd,
+				 enum packing_op op)
+{
+	/* Yup, user manual definitions are reversed */
+	u8 *reg1 = buf + 4;
+
+	sja1105_packing(reg1, &cmd->valid, 31, 31, SIZE_DYN_CMD, op);
+	sja1105_packing(reg1, &cmd->index, 26, 24, SIZE_DYN_CMD, op);
+}
+
+static size_t sja1105et_mac_config_entry_packing(void *buf, void *entry_ptr,
+						 enum packing_op op)
+{
+	struct sja1105_mac_config_entry *entry = entry_ptr;
+	const int size = SIZE_MAC_CONFIG_DYN_ENTRY_ET;
+	/* Yup, user manual definitions are reversed */
+	u8 *reg1 = buf + 4;
+	u8 *reg2 = buf;
+
+	sja1105_packing(reg1, &entry->speed,     30, 29, size, op);
+	sja1105_packing(reg1, &entry->drpdtag,   23, 23, size, op);
+	sja1105_packing(reg1, &entry->drpuntag,  22, 22, size, op);
+	sja1105_packing(reg1, &entry->retag,     21, 21, size, op);
+	sja1105_packing(reg1, &entry->dyn_learn, 20, 20, size, op);
+	sja1105_packing(reg1, &entry->egress,    19, 19, size, op);
+	sja1105_packing(reg1, &entry->ingress,   18, 18, size, op);
+	sja1105_packing(reg1, &entry->ing_mirr,  17, 17, size, op);
+	sja1105_packing(reg1, &entry->egr_mirr,  16, 16, size, op);
+	sja1105_packing(reg1, &entry->vlanprio,  14, 12, size, op);
+	sja1105_packing(reg1, &entry->vlanid,    11,  0, size, op);
+	sja1105_packing(reg2, &entry->tp_delin,  31, 16, size, op);
+	sja1105_packing(reg2, &entry->tp_delout, 15,  0, size, op);
+	/* MAC configuration table entries which can't be reconfigured:
+	 * top, base, enabled, ifg, maxage, drpnona664
+	 */
+	/* Bogus return value, not used anywhere */
+	return 0;
+}
+
+static void
+sja1105pqrs_mac_config_cmd_packing(void *buf, struct sja1105_dyn_cmd *cmd,
+				   enum packing_op op)
+{
+	u8 *p = buf + SIZE_MAC_CONFIG_ENTRY_PQRS;
+
+	sja1105_packing(p, &cmd->valid,   31, 31, SIZE_DYN_CMD, op);
+	sja1105_packing(p, &cmd->errors,  30, 30, SIZE_DYN_CMD, op);
+	sja1105_packing(p, &cmd->rdwrset, 29, 29, SIZE_DYN_CMD, op);
+	sja1105_packing(p, &cmd->index,    2,  0, SIZE_DYN_CMD, op);
+}
+
+static void
+sja1105et_l2_lookup_params_cmd_packing(void *buf, struct sja1105_dyn_cmd *cmd,
+				       enum packing_op op)
+{
+	sja1105_packing(buf, &cmd->valid, 31, 31,
+			SIZE_L2_LOOKUP_PARAMS_DYN_CMD_ET, op);
+}
+
+static size_t
+sja1105et_l2_lookup_params_entry_packing(void *buf, void *entry_ptr,
+					 enum packing_op op)
+{
+	struct sja1105_l2_lookup_params_entry *entry = entry_ptr;
+
+	sja1105_packing(buf, &entry->poly, 7, 0,
+			SIZE_L2_LOOKUP_PARAMS_DYN_CMD_ET, op);
+	/* Bogus return value, not used anywhere */
+	return 0;
+}
+
+static void
+sja1105et_general_params_cmd_packing(void *buf, struct sja1105_dyn_cmd *cmd,
+				     enum packing_op op)
+{
+	const int size = SIZE_GENERAL_PARAMS_DYN_CMD_ET;
+
+	sja1105_packing(buf, &cmd->valid,  31, 31, size, op);
+	sja1105_packing(buf, &cmd->errors, 30, 30, size, op);
+}
+
+static size_t
+sja1105et_general_params_entry_packing(void *buf, void *entry_ptr,
+				       enum packing_op op)
+{
+	struct sja1105_general_params_entry *entry = entry_ptr;
+	const int size = SIZE_GENERAL_PARAMS_DYN_CMD_ET;
+
+	sja1105_packing(buf, &entry->mirr_port, 2, 0, size, op);
+	/* Bogus return value, not used anywhere */
+	return 0;
+}
+
+#define OP_READ  BIT(0)
+#define OP_WRITE BIT(1)
+#define OP_DEL   BIT(2)
+
+/* SJA1105E/T: First generation */
+static struct sja1105_dynamic_table_ops sja1105et_table_ops[BLK_IDX_MAX_DYN] = {
+	[BLK_IDX_L2_LOOKUP] = {
+		.entry_packing = sja1105et_l2_lookup_entry_packing,
+		.cmd_packing = sja1105et_l2_lookup_cmd_packing,
+		.access = (OP_READ | OP_WRITE | OP_DEL),
+		.max_entry_count = MAX_L2_LOOKUP_COUNT,
+		.packed_size = SIZE_L2_LOOKUP_DYN_CMD_ET,
+		.addr = 0x20,
+	},
+	[BLK_IDX_MGMT_ROUTE] = {
+		.entry_packing = sja1105et_mgmt_route_entry_packing,
+		.cmd_packing = sja1105et_mgmt_route_cmd_packing,
+		.access = (OP_READ | OP_WRITE),
+		.max_entry_count = SJA1105_NUM_PORTS,
+		.packed_size = SIZE_L2_LOOKUP_DYN_CMD_ET,
+		.addr = 0x20,
+	},
+	[BLK_IDX_L2_POLICING] = { 0 },
+	[BLK_IDX_VLAN_LOOKUP] = {
+		.entry_packing = sja1105_vlan_lookup_entry_packing,
+		.cmd_packing = sja1105_vlan_lookup_cmd_packing,
+		.access = (OP_WRITE | OP_DEL),
+		.max_entry_count = MAX_VLAN_LOOKUP_COUNT,
+		.packed_size = SIZE_VLAN_LOOKUP_DYN_CMD,
+		.addr = 0x27,
+	},
+	[BLK_IDX_L2_FORWARDING] = {
+		.entry_packing = sja1105_l2_forwarding_entry_packing,
+		.cmd_packing = sja1105_l2_forwarding_cmd_packing,
+		.max_entry_count = MAX_L2_FORWARDING_COUNT,
+		.access = OP_WRITE,
+		.packed_size = SIZE_L2_FORWARDING_DYN_CMD,
+		.addr = 0x24,
+	},
+	[BLK_IDX_MAC_CONFIG] = {
+		.entry_packing = sja1105et_mac_config_entry_packing,
+		.cmd_packing = sja1105et_mac_config_cmd_packing,
+		.max_entry_count = MAX_MAC_CONFIG_COUNT,
+		.access = OP_WRITE,
+		.packed_size = SIZE_MAC_CONFIG_DYN_CMD_ET,
+		.addr = 0x36,
+	},
+	[BLK_IDX_L2_LOOKUP_PARAMS] = {
+		.entry_packing = sja1105et_l2_lookup_params_entry_packing,
+		.cmd_packing = sja1105et_l2_lookup_params_cmd_packing,
+		.max_entry_count = MAX_L2_LOOKUP_PARAMS_COUNT,
+		.access = OP_WRITE,
+		.packed_size = SIZE_L2_LOOKUP_PARAMS_DYN_CMD_ET,
+		.addr = 0x38,
+	},
+	[BLK_IDX_L2_FORWARDING_PARAMS] = { 0 },
+	[BLK_IDX_GENERAL_PARAMS] = {
+		.entry_packing = sja1105et_general_params_entry_packing,
+		.cmd_packing = sja1105et_general_params_cmd_packing,
+		.max_entry_count = MAX_GENERAL_PARAMS_COUNT,
+		.access = OP_WRITE,
+		.packed_size = SIZE_GENERAL_PARAMS_DYN_CMD_ET,
+		.addr = 0x34,
+	},
+	[BLK_IDX_XMII_PARAMS] = { 0 },
+};
+
+/* SJA1105P/Q/R/S: Second generation: TODO */
+static struct sja1105_dynamic_table_ops sja1105pqrs_table_ops[BLK_IDX_MAX_DYN] = {
+	[BLK_IDX_L2_LOOKUP] = {
+		.entry_packing = sja1105pqrs_l2_lookup_entry_packing,
+		.cmd_packing = sja1105pqrs_l2_lookup_cmd_packing,
+		.access = (OP_READ | OP_WRITE | OP_DEL),
+		.max_entry_count = MAX_L2_LOOKUP_COUNT,
+		.packed_size = SIZE_L2_LOOKUP_DYN_CMD_ET,
+		.addr = 0x24,
+	},
+	[BLK_IDX_L2_POLICING] = { 0 },
+	[BLK_IDX_VLAN_LOOKUP] = {
+		.entry_packing = sja1105_vlan_lookup_entry_packing,
+		.cmd_packing = sja1105_vlan_lookup_cmd_packing,
+		.access = (OP_READ | OP_WRITE | OP_DEL),
+		.max_entry_count = MAX_VLAN_LOOKUP_COUNT,
+		.packed_size = SIZE_VLAN_LOOKUP_DYN_CMD,
+		.addr = 0x2D,
+	},
+	[BLK_IDX_L2_FORWARDING] = {
+		.entry_packing = sja1105_l2_forwarding_entry_packing,
+		.cmd_packing = sja1105_l2_forwarding_cmd_packing,
+		.max_entry_count = MAX_L2_FORWARDING_COUNT,
+		.access = OP_WRITE,
+		.packed_size = SIZE_L2_FORWARDING_DYN_CMD,
+		.addr = 0x2A,
+	},
+	[BLK_IDX_MAC_CONFIG] = {
+		.entry_packing = sja1105pqrs_mac_config_entry_packing,
+		.cmd_packing = sja1105pqrs_mac_config_cmd_packing,
+		.max_entry_count = MAX_MAC_CONFIG_COUNT,
+		.access = (OP_READ | OP_WRITE),
+		.packed_size = SIZE_MAC_CONFIG_DYN_CMD_PQRS,
+		.addr = 0x4B,
+	},
+	[BLK_IDX_L2_LOOKUP_PARAMS] = {
+		.entry_packing = sja1105et_l2_lookup_params_entry_packing,
+		.cmd_packing = sja1105et_l2_lookup_params_cmd_packing,
+		.max_entry_count = MAX_L2_LOOKUP_PARAMS_COUNT,
+		.access = (OP_READ | OP_WRITE),
+		.packed_size = SIZE_L2_LOOKUP_PARAMS_DYN_CMD_ET,
+		.addr = 0x38,
+	},
+	[BLK_IDX_L2_FORWARDING_PARAMS] = { 0 },
+	[BLK_IDX_GENERAL_PARAMS] = {
+		.entry_packing = sja1105et_general_params_entry_packing,
+		.cmd_packing = sja1105et_general_params_cmd_packing,
+		.max_entry_count = MAX_GENERAL_PARAMS_COUNT,
+		.access = OP_WRITE,
+		.packed_size = SIZE_GENERAL_PARAMS_DYN_CMD_ET,
+		.addr = 0x34,
+	},
+	[BLK_IDX_XMII_PARAMS] = { 0 },
+};
+
+int sja1105_dynamic_config_read(struct sja1105_private *priv,
+				enum sja1105_blk_idx blk_idx,
+				int index, void *entry)
+{
+	const struct sja1105_dynamic_table_ops *ops;
+	struct sja1105_dyn_cmd cmd = { 0 };
+	/* SPI payload buffer */
+	u8 packed_buf[MAX_DYN_CMD_SIZE];
+	int retries = 3;
+	int rc;
+
+	if (blk_idx >= BLK_IDX_MAX_DYN)
+		return -ERANGE;
+
+	ops = &priv->dyn_ops[blk_idx];
+
+	if (index >= ops->max_entry_count)
+		return -ERANGE;
+	if (!(ops->access & OP_READ))
+		return -EOPNOTSUPP;
+	if (ops->packed_size > MAX_DYN_CMD_SIZE)
+		return -ERANGE;
+	if (!ops->cmd_packing)
+		return -EOPNOTSUPP;
+	if (!ops->entry_packing)
+		return -EOPNOTSUPP;
+
+	memset(packed_buf, 0, ops->packed_size);
+
+	cmd.valid = true; /* Trigger action on table entry */
+	cmd.rdwrset = SPI_READ; /* Action is read */
+	cmd.index = index;
+	ops->cmd_packing(packed_buf, &cmd, PACK);
+
+	/* Send SPI write operation: read config table entry */
+	rc = sja1105_spi_send_packed_buf(priv, SPI_WRITE, ops->addr,
+					 packed_buf, ops->packed_size);
+	if (rc < 0)
+		return rc;
+
+	/* Loop until we have confirmation that hardware has finished
+	 * processing the command and has cleared the VALID field
+	 */
+	do {
+		memset(packed_buf, 0, ops->packed_size);
+
+		/* Retrieve the read operation's result */
+		rc = sja1105_spi_send_packed_buf(priv, SPI_READ, ops->addr,
+						 packed_buf, ops->packed_size);
+		if (rc < 0)
+			return rc;
+
+		memset(&cmd, 0, sizeof(cmd));
+		ops->cmd_packing(packed_buf, &cmd, UNPACK);
+		/* UM10944: [valident] will always be found cleared
+		 * during a read access with MGMTROUTE set.
+		 * So don't error out in that case.
+		 */
+		if (!cmd.valident && blk_idx != BLK_IDX_MGMT_ROUTE)
+			return -EINVAL;
+		cpu_relax();
+	} while (cmd.valid && --retries);
+
+	if (cmd.valid)
+		return -ETIMEDOUT;
+
+	/* Don't dereference possibly NULL pointer - maybe caller
+	 * only wanted to see whether the entry existed or not.
+	 */
+	if (entry)
+		ops->entry_packing(packed_buf, entry, UNPACK);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(sja1105_dynamic_config_read);
+
+int sja1105_dynamic_config_write(struct sja1105_private *priv,
+				 enum sja1105_blk_idx blk_idx,
+				 int index, void *entry, bool keep)
+{
+	const struct sja1105_dynamic_table_ops *ops;
+	struct sja1105_dyn_cmd cmd = { 0 };
+	/* SPI payload buffer */
+	u8 packed_buf[MAX_DYN_CMD_SIZE];
+	int rc;
+
+	if (blk_idx >= BLK_IDX_MAX_DYN)
+		return -ERANGE;
+
+	ops = &priv->dyn_ops[blk_idx];
+
+	if (index >= ops->max_entry_count)
+		return -ERANGE;
+	if (!(ops->access & OP_WRITE))
+		return -EOPNOTSUPP;
+	if (!keep && !(ops->access & OP_DEL))
+		return -EOPNOTSUPP;
+	if (ops->packed_size > MAX_DYN_CMD_SIZE)
+		return -ERANGE;
+
+	memset(packed_buf, 0, ops->packed_size);
+
+	cmd.valident = keep; /* If false, deletes entry */
+	cmd.valid = true; /* Trigger action on table entry */
+	cmd.rdwrset = SPI_WRITE; /* Action is write */
+	cmd.index = index;
+
+	if (!ops->cmd_packing)
+		return -EOPNOTSUPP;
+	ops->cmd_packing(packed_buf, &cmd, PACK);
+
+	if (!ops->entry_packing)
+		return -EOPNOTSUPP;
+	/* Don't dereference potentially NULL pointer if just
+	 * deleting a table entry is what was requested. For cases
+	 * where 'index' field is physically part of entry structure,
+	 * and needed here, we deal with that in the cmd_packing callback.
+	 */
+	if (keep)
+		ops->entry_packing(packed_buf, entry, PACK);
+
+	/* Send SPI write operation: read config table entry */
+	rc = sja1105_spi_send_packed_buf(priv, SPI_WRITE, ops->addr,
+					 packed_buf, ops->packed_size);
+	if (rc < 0)
+		return rc;
+
+	memset(&cmd, 0, sizeof(cmd));
+	ops->cmd_packing(packed_buf, &cmd, UNPACK);
+	if (cmd.errors)
+		return -EINVAL;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(sja1105_dynamic_config_write);
+
+void sja1105_dynamic_config_init(struct sja1105_private *priv)
+{
+	const struct sja1105_dynamic_table_ops *ops;
+
+	if (IS_ET(priv->device_id))
+		ops = sja1105et_table_ops;
+	else
+		ops = sja1105pqrs_table_ops;
+
+	priv->dyn_ops = ops;
+}
diff --git a/drivers/net/dsa/sja1105/sja1105_dynamic_config.h b/drivers/net/dsa/sja1105/sja1105_dynamic_config.h
new file mode 100644
index 000000000000..afc6d41f0330
--- /dev/null
+++ b/drivers/net/dsa/sja1105/sja1105_dynamic_config.h
@@ -0,0 +1,40 @@
+/* SPDX-License-Identifier: GPL-2.0
+ * Copyright (c) 2019, Vladimir Oltean <olteanv@gmail.com>
+ */
+#ifndef _SJA1105_DYNAMIC_CONFIG_H
+#define _SJA1105_DYNAMIC_CONFIG_H
+
+#include "sja1105.h"
+#include <linux/packing.h>
+
+struct sja1105_dyn_cmd {
+	u64 valid;
+	u64 rdwrset;
+	u64 errors;
+	u64 valident;
+	u64 index;
+};
+
+struct sja1105_dynamic_table_ops {
+	/* This returns size_t just to keep same prototype as the
+	 * static config ops, of which we are reusing some functions.
+	 */
+	size_t (*entry_packing)(void *buf, void *entry_ptr, enum packing_op op);
+	void (*cmd_packing)(void *buf, struct sja1105_dyn_cmd *cmd,
+			    enum packing_op op);
+	size_t max_entry_count;
+	size_t packed_size;
+	u64 addr;
+	u8 access;
+};
+
+struct sja1105_mgmt_entry {
+	u64 tsreg;
+	u64 takets;
+	u64 macaddr;
+	u64 destports;
+	u64 enfport;
+	u64 index;
+};
+
+#endif
diff --git a/drivers/net/dsa/sja1105/sja1105_main.c b/drivers/net/dsa/sja1105/sja1105_main.c
new file mode 100644
index 000000000000..45c233d7fbba
--- /dev/null
+++ b/drivers/net/dsa/sja1105/sja1105_main.c
@@ -0,0 +1,890 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2018, Sensor-Technik Wiedemann GmbH
+ * Copyright (c) 2018-2019, Vladimir Oltean <olteanv@gmail.com>
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/delay.h>
+#include <linux/module.h>
+#include <linux/printk.h>
+#include <linux/spi/spi.h>
+#include <linux/errno.h>
+#include <linux/gpio/consumer.h>
+#include <linux/of.h>
+#include <linux/of_net.h>
+#include <linux/of_mdio.h>
+#include <linux/netdev_features.h>
+#include <linux/netdevice.h>
+#include <linux/if_bridge.h>
+#include <linux/if_ether.h>
+#include "sja1105.h"
+
+static void sja1105_hw_reset(struct gpio_desc *gpio, unsigned int pulse_len,
+			     unsigned int startup_delay)
+{
+	gpiod_set_value_cansleep(gpio, 1);
+	/* Wait for minimum reset pulse length */
+	msleep(pulse_len);
+	gpiod_set_value_cansleep(gpio, 0);
+	/* Wait until chip is ready after reset */
+	msleep(startup_delay);
+}
+
+static void
+sja1105_port_allow_traffic(struct sja1105_l2_forwarding_entry *l2_fwd,
+			   int from, int to, bool allow)
+{
+	if (allow) {
+		l2_fwd[from].bc_domain  |= BIT(to);
+		l2_fwd[from].reach_port |= BIT(to);
+		l2_fwd[from].fl_domain  |= BIT(to);
+	} else {
+		l2_fwd[from].bc_domain  &= ~BIT(to);
+		l2_fwd[from].reach_port &= ~BIT(to);
+		l2_fwd[from].fl_domain  &= ~BIT(to);
+	}
+}
+
+/* Structure used to temporarily transport device tree
+ * settings into sja1105_setup
+ */
+struct sja1105_dt_port {
+	phy_interface_t phy_mode;
+	sja1105_mii_mode_t mode;
+};
+
+static int sja1105_init_mac_settings(struct sja1105_private *priv)
+{
+	struct sja1105_mac_config_entry default_mac = {
+		/* Enable all 8 priority queues on egress.
+		 * Every queue i holds top[i] - base[i] frames.
+		 * Sum of top[i] - base[i] is 511 (max hardware limit).
+		 */
+		.top  = {0x3F, 0x7F, 0xBF, 0xFF, 0x13F, 0x17F, 0x1BF, 0x1FF},
+		.base = {0x0, 0x40, 0x80, 0xC0, 0x100, 0x140, 0x180, 0x1C0},
+		.enabled = {true, true, true, true, true, true, true, true},
+		/* Keep standard IFG of 12 bytes on egress. */
+		.ifg = 0,
+		/* Always put the MAC speed in automatic mode, where it can be
+		 * retrieved from the PHY object through phylib and
+		 * sja1105_adjust_port_config.
+		 */
+		.speed = SJA1105_SPEED_AUTO,
+		/* No static correction for 1-step 1588 events */
+		.tp_delin = 0,
+		.tp_delout = 0,
+		/* Disable aging for critical TTEthernet traffic */
+		.maxage = 0xFF,
+		/* Internal VLAN (pvid) to apply to untagged ingress */
+		.vlanprio = 0,
+		.vlanid = 0,
+		.ing_mirr = false,
+		.egr_mirr = false,
+		/* Don't drop traffic with other EtherType than 800h */
+		.drpnona664 = false,
+		/* Don't drop double-tagged traffic */
+		.drpdtag = false,
+		/* Don't drop VLAN with single outer tag - P/Q/R/S only */
+		.drpsotag = false,
+		/* Don't drop VLAN with single inner tag - P/Q/R/S only */
+		.drpsitag = false,
+		/* Don't drop untagged traffic */
+		.drpuntag = false,
+		/* Don't retag 802.1p (VID 0) traffic with the pvid */
+		.retag = false,
+		/* Enable learning and I/O on user ports by default. */
+		.dyn_learn = true,
+		.egress = false,
+		.ingress = false,
+		.mirrcie = 0,
+		.mirrcetag = 0,
+		.ingmirrvid = 0,
+		.ingmirrpcp = 0,
+		.ingmirrdei = 0,
+	};
+	struct sja1105_mac_config_entry *mac;
+	struct sja1105_table *table;
+	int i;
+
+	table = &priv->static_config.tables[BLK_IDX_MAC_CONFIG];
+
+	/* Discard previous MAC Configuration Table */
+	if (table->entry_count) {
+		kfree(table->entries);
+		table->entry_count = 0;
+	}
+
+	table->entries = kcalloc(SJA1105_NUM_PORTS,
+				 table->ops->unpacked_entry_size, GFP_KERNEL);
+	if (!table->entries)
+		return -ENOMEM;
+
+	/* Override table based on phylib DT bindings */
+	table->entry_count = SJA1105_NUM_PORTS;
+
+	mac = table->entries;
+
+	for (i = 0; i < SJA1105_NUM_PORTS; i++)
+		mac[i] = default_mac;
+
+	return 0;
+}
+
+static int sja1105_init_mii_settings(struct sja1105_private *priv,
+				     struct sja1105_dt_port *ports)
+{
+	struct device *dev = &priv->spidev->dev;
+	struct sja1105_xmii_params_entry *mii;
+	struct sja1105_table *table;
+	int i;
+
+	table = &priv->static_config.tables[BLK_IDX_XMII_PARAMS];
+
+	/* Discard previous xMII Mode Parameters Table */
+	if (table->entry_count) {
+		kfree(table->entries);
+		table->entry_count = 0;
+	}
+
+	table->entries = kcalloc(MAX_XMII_PARAMS_COUNT,
+				 table->ops->unpacked_entry_size, GFP_KERNEL);
+	if (!table->entries)
+		return -ENOMEM;
+
+	/* Override table based on phylib DT bindings */
+	table->entry_count = MAX_XMII_PARAMS_COUNT;
+
+	mii = table->entries;
+
+	for (i = 0; i < SJA1105_NUM_PORTS; i++) {
+		switch (ports[i].phy_mode) {
+		case PHY_INTERFACE_MODE_MII:
+			mii->xmii_mode[i] = XMII_MODE_MII;
+			break;
+		case PHY_INTERFACE_MODE_RMII:
+			mii->xmii_mode[i] = XMII_MODE_RMII;
+			break;
+		case PHY_INTERFACE_MODE_RGMII:
+		case PHY_INTERFACE_MODE_RGMII_ID:
+		case PHY_INTERFACE_MODE_RGMII_RXID:
+		case PHY_INTERFACE_MODE_RGMII_TXID:
+			mii->xmii_mode[i] = XMII_MODE_RGMII;
+			break;
+		default:
+			dev_err(dev, "Unsupported PHY mode %s!\n",
+				phy_modes(ports[i].phy_mode));
+		}
+
+		mii->phy_mac[i] = ports[i].mode;
+	}
+	return 0;
+}
+
+static int sja1105_init_static_fdb(struct sja1105_private *priv)
+{
+	struct sja1105_table *table;
+
+	table = &priv->static_config.tables[BLK_IDX_L2_LOOKUP];
+
+	if (table->entry_count) {
+		kfree(table->entries);
+		table->entry_count = 0;
+	}
+	return 0;
+}
+
+static int sja1105_init_l2_lookup_params(struct sja1105_private *priv)
+{
+	struct sja1105_table *table;
+	struct sja1105_l2_lookup_params_entry default_l2_lookup_params = {
+		/* TODO Learned FDB entries are never forgotten */
+		.maxage = 0,
+		/* All entries within a FDB bin are available for learning */
+		.dyn_tbsz = SJA1105ET_FDB_BIN_SIZE,
+		/* 2^8 + 2^5 + 2^3 + 2^2 + 2^1 + 1 in Koopman notation */
+		.poly = 0x97,
+		/* This selects between Independent VLAN Learning (IVL) and
+		 * Shared VLAN Learning (SVL)
+		 */
+		.shared_learn = false,
+		/* Don't discard management traffic based on ENFPORT -
+		 * we don't perform SMAC port enforcement anyway, so
+		 * what we are setting here doesn't matter.
+		 */
+		.no_enf_hostprt = false,
+		/* Don't learn SMAC for mac_fltres1 and mac_fltres0.
+		 * TODO Maybe correlate with no_linklocal_learn from bridge
+		 * driver?
+		 */
+		.no_mgmt_learn = true,
+	};
+
+	table = &priv->static_config.tables[BLK_IDX_L2_LOOKUP_PARAMS];
+
+	if (table->entry_count) {
+		kfree(table->entries);
+		table->entry_count = 0;
+	}
+
+	table->entries = kcalloc(MAX_L2_LOOKUP_PARAMS_COUNT,
+				 table->ops->unpacked_entry_size, GFP_KERNEL);
+	if (!table->entries)
+		return -ENOMEM;
+
+	table->entry_count = MAX_L2_LOOKUP_PARAMS_COUNT;
+
+	/* This table only has a single entry */
+	((struct sja1105_l2_lookup_params_entry *)table->entries)[0] =
+				default_l2_lookup_params;
+
+	return 0;
+}
+
+static int sja1105_init_static_vlan(struct sja1105_private *priv)
+{
+	struct sja1105_table *table;
+	struct sja1105_vlan_lookup_entry pvid = {
+		.ving_mirr = 0,
+		.vegr_mirr = 0,
+		.vmemb_port = 0,
+		.vlan_bc = 0,
+		.tag_port = 0,
+		.vlanid = 0,
+	};
+	int i;
+
+	table = &priv->static_config.tables[BLK_IDX_VLAN_LOOKUP];
+
+	/* The static VLAN table will only contain the initial pvid of 0.
+	 */
+	if (table->entry_count) {
+		kfree(table->entries);
+		table->entry_count = 0;
+	}
+
+	table->entries = kcalloc(1, table->ops->unpacked_entry_size,
+				 GFP_KERNEL);
+	if (!table->entries)
+		return -ENOMEM;
+
+	table->entry_count = 1;
+
+	/* VLAN ID 0: all DT-defined ports are members; no restrictions on
+	 * forwarding; always transmit priority-tagged frames as untagged.
+	 */
+	for (i = 0; i < SJA1105_NUM_PORTS; i++) {
+		pvid.vmemb_port |= BIT(i);
+		pvid.vlan_bc |= BIT(i);
+		pvid.tag_port &= ~BIT(i);
+	}
+
+	((struct sja1105_vlan_lookup_entry *)table->entries)[0] = pvid;
+	return 0;
+}
+
+static int sja1105_init_l2_forwarding(struct sja1105_private *priv)
+{
+	struct sja1105_l2_forwarding_entry *l2fwd;
+	struct sja1105_table *table;
+	int i, j;
+
+	table = &priv->static_config.tables[BLK_IDX_L2_FORWARDING];
+
+	if (table->entry_count) {
+		kfree(table->entries);
+		table->entry_count = 0;
+	}
+
+	table->entries = kcalloc(MAX_L2_FORWARDING_COUNT,
+				 table->ops->unpacked_entry_size, GFP_KERNEL);
+	if (!table->entries)
+		return -ENOMEM;
+
+	table->entry_count = MAX_L2_FORWARDING_COUNT;
+
+	l2fwd = table->entries;
+
+	/* First 5 entries define the forwarding rules */
+	for (i = 0; i < SJA1105_NUM_PORTS; i++) {
+		unsigned int upstream = dsa_upstream_port(priv->ds, i);
+
+		for (j = 0; j < SJA1105_NUM_TC; j++)
+			l2fwd[i].vlan_pmap[j] = j;
+
+		if (i == upstream)
+			continue;
+
+		sja1105_port_allow_traffic(l2fwd, i, upstream, true);
+		sja1105_port_allow_traffic(l2fwd, upstream, i, true);
+	}
+	/* Next 8 entries define VLAN PCP mapping from ingress to egress.
+	 * Create a one-to-one mapping.
+	 */
+	for (i = 0; i < SJA1105_NUM_TC; i++)
+		for (j = 0; j < SJA1105_NUM_PORTS; j++)
+			l2fwd[SJA1105_NUM_PORTS + i].vlan_pmap[j] = i;
+
+	return 0;
+}
+
+static int sja1105_init_l2_forwarding_params(struct sja1105_private *priv)
+{
+	struct sja1105_l2_forwarding_params_entry default_l2fwd_params = {
+		/* Disallow dynamic reconfiguration of vlan_pmap */
+		.max_dynp = 0,
+		/* Use a single memory partition for all ingress queues */
+		.part_spc = { MAX_FRAME_MEMORY, 0, 0, 0, 0, 0, 0, 0 },
+	};
+	struct sja1105_table *table;
+
+	table = &priv->static_config.tables[BLK_IDX_L2_FORWARDING_PARAMS];
+
+	if (table->entry_count) {
+		kfree(table->entries);
+		table->entry_count = 0;
+	}
+
+	table->entries = kcalloc(MAX_L2_FORWARDING_PARAMS_COUNT,
+				 table->ops->unpacked_entry_size, GFP_KERNEL);
+	if (!table->entries)
+		return -ENOMEM;
+
+	table->entry_count = MAX_L2_FORWARDING_PARAMS_COUNT;
+
+	/* This table only has a single entry */
+	((struct sja1105_l2_forwarding_params_entry *)table->entries)[0] =
+				default_l2fwd_params;
+
+	return 0;
+}
+
+static int sja1105_init_general_params(struct sja1105_private *priv)
+{
+	struct sja1105_general_params_entry default_general_params = {
+		/* Disallow dynamic changing of the mirror port */
+		.mirr_ptacu = 0,
+		.switchid = priv->ds->index,
+		/* Priority queue for link-local frames trapped to CPU */
+		.hostprio = 0,
+		.mac_fltres1 = SJA1105_LINKLOCAL_FILTER_A,
+		.mac_flt1    = SJA1105_LINKLOCAL_FILTER_A_MASK,
+		.incl_srcpt1 = true,
+		.send_meta1  = false,
+		.mac_fltres0 = SJA1105_LINKLOCAL_FILTER_B,
+		.mac_flt0    = SJA1105_LINKLOCAL_FILTER_B_MASK,
+		.incl_srcpt0 = true,
+		.send_meta0  = false,
+		/* The destination for traffic matching mac_fltres1 and
+		 * mac_fltres0 on all ports except host_port. Such traffic
+		 * receieved on host_port itself would be dropped, except
+		 * by installing a temporary 'management route'
+		 */
+		.host_port = dsa_upstream_port(priv->ds, 0),
+		/* Same as host port */
+		.mirr_port = dsa_upstream_port(priv->ds, 0),
+		/* Link-local traffic received on casc_port will be forwarded
+		 * to host_port without embedding the source port and device ID
+		 * info in the destination MAC address (presumably because it
+		 * is a cascaded port and a downstream SJA switch already did
+		 * that). Default to an invalid port (to disable the feature)
+		 * and overwrite this if we find any DSA (cascaded) ports.
+		 */
+		.casc_port = SJA1105_NUM_PORTS,
+		/* No TTEthernet */
+		.vllupformat = 0,
+		.vlmarker = 0,
+		.vlmask = 0,
+		/* Only update correctionField for 1-step PTP (L2 transport) */
+		.ignore2stf = 0,
+		.tpid = ETH_P_8021Q,
+		.tpid2 = ETH_P_8021Q,
+		/* P/Q/R/S only */
+		.queue_ts = 0,
+		.egrmirrvid = 0,
+		.egrmirrpcp = 0,
+		.egrmirrdei = 0,
+		.replay_port = 0,
+	};
+	struct sja1105_table *table;
+	int i;
+
+	for (i = 0; i < SJA1105_NUM_PORTS; i++)
+		if (dsa_is_dsa_port(priv->ds, i))
+			default_general_params.casc_port = i;
+
+	table = &priv->static_config.tables[BLK_IDX_GENERAL_PARAMS];
+
+	if (table->entry_count) {
+		kfree(table->entries);
+		table->entry_count = 0;
+	}
+
+	table->entries = kcalloc(MAX_GENERAL_PARAMS_COUNT,
+				 table->ops->unpacked_entry_size, GFP_KERNEL);
+	if (!table->entries)
+		return -ENOMEM;
+
+	table->entry_count = MAX_GENERAL_PARAMS_COUNT;
+
+	/* This table only has a single entry */
+	((struct sja1105_general_params_entry *)table->entries)[0] =
+				default_general_params;
+
+	return 0;
+}
+
+static inline void
+sja1105_setup_policer(struct sja1105_l2_policing_entry *policing,
+		      int index)
+{
+#define RATE_MBPS(speed) (((speed) * 64000) / 1000)
+	policing[index].sharindx = index;
+	policing[index].smax = 65535; /* Burst size in bytes */
+	policing[index].rate = RATE_MBPS(1000);
+	policing[index].maxlen = ETH_FRAME_LEN + VLAN_HLEN + ETH_FCS_LEN;
+	policing[index].partition = 0;
+#undef RATE_MBPS
+}
+
+static int sja1105_init_l2_policing(struct sja1105_private *priv)
+{
+	struct sja1105_l2_policing_entry *policing;
+	struct sja1105_table *table;
+	int i, j, k;
+
+	table = &priv->static_config.tables[BLK_IDX_L2_POLICING];
+
+	/* Discard previous L2 Policing Table */
+	if (table->entry_count) {
+		kfree(table->entries);
+		table->entry_count = 0;
+	}
+
+	table->entries = kcalloc(MAX_L2_POLICING_COUNT,
+				 table->ops->unpacked_entry_size, GFP_KERNEL);
+	if (!table->entries)
+		return -ENOMEM;
+
+	/* Override table based on phylib DT bindings */
+	table->entry_count = MAX_L2_POLICING_COUNT;
+
+	policing = table->entries;
+
+	/* k sweeps through all unicast policers (0-39).
+	 * bcast sweeps through policers 40-44.
+	 */
+	for (i = 0, k = 0; i < SJA1105_NUM_PORTS; i++) {
+		int bcast = (SJA1105_NUM_PORTS * SJA1105_NUM_TC) + i;
+
+		for (j = 0; j < SJA1105_NUM_TC; j++, k++)
+			sja1105_setup_policer(policing, k);
+
+		/* Set up this port's policer for broadcast traffic */
+		sja1105_setup_policer(policing, bcast);
+	}
+	return 0;
+}
+
+static int sja1105_static_config_load(struct sja1105_private *priv,
+				      struct sja1105_dt_port *ports)
+{
+	int rc;
+
+	sja1105_static_config_free(&priv->static_config);
+	rc = sja1105_static_config_init(&priv->static_config,
+					priv->device_id, priv->part_nr);
+	if (rc)
+		return rc;
+
+	/* Build static configuration */
+	rc = sja1105_init_mac_settings(priv);
+	if (rc < 0)
+		return rc;
+	rc = sja1105_init_mii_settings(priv, ports);
+	if (rc < 0)
+		return rc;
+	rc = sja1105_init_static_fdb(priv);
+	if (rc < 0)
+		return rc;
+	rc = sja1105_init_static_vlan(priv);
+	if (rc < 0)
+		return rc;
+	rc = sja1105_init_l2_lookup_params(priv);
+	if (rc < 0)
+		return rc;
+	rc = sja1105_init_l2_forwarding(priv);
+	if (rc < 0)
+		return rc;
+	rc = sja1105_init_l2_forwarding_params(priv);
+	if (rc < 0)
+		return rc;
+	rc = sja1105_init_l2_policing(priv);
+	if (rc < 0)
+		return rc;
+	rc = sja1105_init_general_params(priv);
+	if (rc < 0)
+		return rc;
+
+	/* Send initial configuration to hardware via SPI */
+	return sja1105_static_config_upload(priv);
+}
+
+static int sja1105_parse_ports_node(struct sja1105_dt_port *ports,
+				    struct device_node *ports_node)
+{
+	struct device_node *child;
+
+	for_each_child_of_node(ports_node, child) {
+		struct device_node *phy_node;
+		int phy_mode;
+		u32 index;
+
+		/* Get switch port number from DT */
+		if (of_property_read_u32(child, "reg", &index) < 0) {
+			pr_err("Port number not defined in device tree (property \"reg\")\n");
+			return -ENODEV;
+		}
+
+		/* Get PHY mode from DT */
+		phy_mode = of_get_phy_mode(child);
+		if (phy_mode < 0) {
+			pr_err("Failed to read phy-mode or phy-interface-type property for port %d\n",
+			       index);
+			return -ENODEV;
+		}
+		ports[index].phy_mode = phy_mode;
+
+		phy_node = of_parse_phandle(child, "phy-handle", 0);
+		if (!phy_node) {
+			if (!of_phy_is_fixed_link(child)) {
+				pr_err("phy-handle or fixed-link properties missing!\n");
+				return -ENODEV;
+			}
+			/* phy-handle is missing, but fixed-link isn't.
+			 * So it's a fixed link. Default to PHY mode.
+			 */
+			ports[index].mode = XMII_PHY;
+		} else {
+			/* phy-handle present => put port in MAC mode */
+			ports[index].mode = XMII_MAC;
+			of_node_put(phy_node);
+		}
+
+		/* The MAC/PHY role can be overridden with explicit bindings */
+		if (of_property_read_bool(child, "sja1105,mac-mode"))
+			ports[index].mode = XMII_MAC;
+		else if (of_property_read_bool(child, "sja1105,phy-mode"))
+			ports[index].mode = XMII_PHY;
+	}
+
+	return 0;
+}
+
+static int sja1105_parse_dt(struct sja1105_private *priv,
+			    struct sja1105_dt_port *ports)
+{
+	struct device *dev = &priv->spidev->dev;
+	struct device_node *switch_node = dev->of_node;
+	struct device_node *ports_node;
+	int rc;
+
+	ports_node = of_get_child_by_name(switch_node, "ports");
+	if (!ports_node) {
+		dev_err(dev, "Incorrect bindings: absent \"ports\" node\n");
+		return -ENODEV;
+	}
+
+	rc = sja1105_parse_ports_node(ports, ports_node);
+	of_node_put(ports_node);
+
+	return rc;
+}
+
+/* Convert back and forth MAC speed from Mbps to SJA1105 encoding */
+static int sja1105_speed[] = {
+	[SJA1105_SPEED_AUTO]     = 0,
+	[SJA1105_SPEED_10MBPS]   = 10,
+	[SJA1105_SPEED_100MBPS]  = 100,
+	[SJA1105_SPEED_1000MBPS] = 1000,
+};
+
+static sja1105_speed_t sja1105_get_speed_cfg(unsigned int speed_mbps)
+{
+	int i;
+
+	for (i = SJA1105_SPEED_AUTO; i <= SJA1105_SPEED_1000MBPS; i++)
+		if (sja1105_speed[i] == speed_mbps)
+			return i;
+	return -EINVAL;
+}
+
+/* Set link speed and enable/disable traffic I/O in the MAC configuration
+ * for a specific port.
+ *
+ * @speed_mbps: If 0, leave the speed unchanged, else adapt MAC to PHY speed.
+ * @enabled: Manage Rx and Tx settings for this port. Overrides the static
+ *	     configuration settings.
+ */
+static int sja1105_adjust_port_config(struct sja1105_private *priv, int port,
+				      int speed_mbps, bool enabled)
+{
+	struct sja1105_xmii_params_entry *mii;
+	struct sja1105_mac_config_entry *mac;
+	struct device *dev = priv->ds->dev;
+	sja1105_phy_interface_t phy_mode;
+	sja1105_speed_t speed;
+	int rc;
+
+	mii = priv->static_config.tables[BLK_IDX_XMII_PARAMS].entries;
+	mac = priv->static_config.tables[BLK_IDX_MAC_CONFIG].entries;
+
+	speed = sja1105_get_speed_cfg(speed_mbps);
+	if (speed_mbps && speed < 0) {
+		dev_err(dev, "Invalid speed %iMbps\n", speed_mbps);
+		return -EINVAL;
+	}
+
+	/* If requested, overwrite SJA1105_SPEED_AUTO from the static MAC
+	 * configuration table, since this will be used for the clocking setup,
+	 * and we no longer need to store it in the static config (already told
+	 * hardware we want auto during upload phase).
+	 */
+	if (speed_mbps)
+		mac[port].speed = speed;
+	else
+		mac[port].speed = SJA1105_SPEED_AUTO;
+
+	/* On P/Q/R/S, one can read from the device via the MAC reconfiguration
+	 * tables. On E/T, MAC reconfig tables are not readable, only writable.
+	 * We have to *know* what the MAC looks like.  For the sake of keeping
+	 * the code common, we'll use the static configuration tables as a
+	 * reasonable approximation for both E/T and P/Q/R/S.
+	 */
+	mac[port].ingress = enabled;
+	mac[port].egress  = enabled;
+
+	/* Write to the dynamic reconfiguration tables */
+	rc = sja1105_dynamic_config_write(priv, BLK_IDX_MAC_CONFIG,
+					  port, &mac[port], true);
+	if (rc < 0) {
+		dev_err(dev, "Failed to write MAC config: %d\n", rc);
+		return rc;
+	}
+
+	/* Reconfigure the PLLs for the RGMII interfaces (required 125 MHz at
+	 * gigabit, 25 MHz at 100 Mbps and 2.5 MHz at 10 Mbps). For MII and
+	 * RMII no change of the clock setup is required. Actually, changing
+	 * the clock setup does interrupt the clock signal for a certain time
+	 * which causes trouble for all PHYs relying on this signal.
+	 */
+	if (!enabled)
+		return 0;
+
+	phy_mode = mii->xmii_mode[port];
+	if (phy_mode != XMII_MODE_RGMII)
+		return 0;
+
+	return sja1105_clocking_setup_port(priv, port);
+}
+
+static void sja1105_adjust_link(struct dsa_switch *ds, int port,
+				struct phy_device *phydev)
+{
+	struct sja1105_private *priv = ds->priv;
+
+	if (!phydev->link)
+		sja1105_adjust_port_config(priv, port, 0, false);
+	else
+		sja1105_adjust_port_config(priv, port, phydev->speed, true);
+}
+
+static int sja1105_bridge_member(struct dsa_switch *ds, int port,
+				 struct net_device *br, bool member)
+{
+	struct sja1105_l2_forwarding_entry *l2_fwd;
+	struct sja1105_private *priv = ds->priv;
+	int i, rc;
+
+	l2_fwd = priv->static_config.tables[BLK_IDX_L2_FORWARDING].entries;
+
+	for (i = 0; i < SJA1105_NUM_PORTS; i++) {
+		/* Add this port to the forwarding matrix of the
+		 * other ports in the same bridge, and viceversa.
+		 */
+		if (!dsa_is_user_port(ds, i))
+			continue;
+		if (i == port)
+			continue;
+		if (dsa_to_port(ds, i)->bridge_dev != br)
+			continue;
+		sja1105_port_allow_traffic(l2_fwd, i, port, member);
+		sja1105_port_allow_traffic(l2_fwd, port, i, member);
+
+		rc = sja1105_dynamic_config_write(priv, BLK_IDX_L2_FORWARDING,
+						  i, &l2_fwd[i], true);
+		if (rc < 0)
+			return rc;
+	}
+
+	return sja1105_dynamic_config_write(priv, BLK_IDX_L2_FORWARDING,
+					    port, &l2_fwd[port], true);
+}
+
+static int sja1105_bridge_join(struct dsa_switch *ds, int port,
+			       struct net_device *br)
+{
+	return sja1105_bridge_member(ds, port, br, true);
+}
+
+static void sja1105_bridge_leave(struct dsa_switch *ds, int port,
+				 struct net_device *br)
+{
+	sja1105_bridge_member(ds, port, br, false);
+}
+
+static enum dsa_tag_protocol
+sja1105_get_tag_protocol(struct dsa_switch *ds, int port)
+{
+	return DSA_TAG_PROTO_NONE;
+}
+
+/* The programming model for the SJA1105 switch is "all-at-once" via static
+ * configuration tables. Some of these can be dynamically modified at runtime,
+ * but not the xMII mode parameters table.
+ * Furthermode, some PHYs may not have crystals for generating their clocks
+ * (e.g. RMII). Instead, their 50MHz clock is supplied via the SJA1105 port's
+ * ref_clk pin. So port clocking needs to be initialized early, before
+ * connecting to PHYs is attempted, otherwise they won't respond through MDIO.
+ * Setting correct PHY link speed does not matter now.
+ * But dsa_slave_phy_setup is called later than sja1105_setup, so the PHY
+ * bindings are not yet parsed by DSA core. We need to parse early so that we
+ * can populate the xMII mode parameters table.
+ */
+static int sja1105_setup(struct dsa_switch *ds)
+{
+	struct sja1105_dt_port ports[SJA1105_NUM_PORTS];
+	struct sja1105_private *priv = ds->priv;
+	int rc;
+
+	rc = sja1105_parse_dt(priv, ports);
+	if (rc < 0) {
+		dev_err(ds->dev, "Failed to parse DT: %d\n", rc);
+		return rc;
+	}
+	/* Create and send configuration down to device */
+	rc = sja1105_static_config_load(priv, ports);
+	if (rc < 0) {
+		dev_err(ds->dev, "Failed to load static config: %d\n", rc);
+		return rc;
+	}
+	/* Configure the CGU (PHY link modes and speeds) */
+	rc = sja1105_clocking_setup(priv);
+	if (rc < 0) {
+		dev_err(ds->dev, "Failed to configure MII clocking: %d\n", rc);
+		return rc;
+	}
+
+	return 0;
+}
+
+static const struct dsa_switch_ops sja1105_switch_ops = {
+	.get_tag_protocol	= sja1105_get_tag_protocol,
+	.setup			= sja1105_setup,
+	.adjust_link		= sja1105_adjust_link,
+	.port_bridge_join	= sja1105_bridge_join,
+	.port_bridge_leave	= sja1105_bridge_leave,
+};
+
+static int sja1105_probe(struct spi_device *spi)
+{
+	struct device *dev = &spi->dev;
+	struct sja1105_private *priv;
+	struct dsa_switch *ds;
+	int rc;
+
+	if (!dev->of_node) {
+		dev_err(dev, "No DTS bindings for SJA1105 driver\n");
+		return -EINVAL;
+	}
+
+	priv = devm_kzalloc(dev, sizeof(struct sja1105_private), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	ds = dsa_switch_alloc(dev, SJA1105_NUM_PORTS);
+	if (!ds)
+		return -ENOMEM;
+
+	ds->ops = &sja1105_switch_ops;
+	ds->priv = priv;
+	priv->ds = ds;
+
+	/* Populate our driver private structure (priv) based on
+	 * the device tree node that was probed (spi)
+	 */
+	priv->spidev = spi;
+	spi_set_drvdata(spi, priv);
+
+	/* Configure the SPI bus */
+	spi->mode = SPI_CPHA;
+	spi->bits_per_word = 8;
+	rc = spi_setup(spi);
+	if (rc < 0) {
+		dev_err(dev, "Could not init SPI\n");
+		return rc;
+	}
+
+	/* Configure the optional reset pin and bring up switch */
+	priv->reset_gpio = devm_gpiod_get(dev, "reset", GPIOD_OUT_HIGH);
+	if (IS_ERR(priv->reset_gpio))
+		dev_dbg(dev, "reset-gpios not defined, ignoring\n");
+	else
+		sja1105_hw_reset(priv->reset_gpio, 1, 1);
+
+	/* Detect hardware device */
+	rc = sja1105_device_id_get(priv);
+	if (rc < 0) {
+		dev_err(dev, "Failed to read device ID: %d\n", rc);
+		return rc;
+	}
+
+	dev_info(dev, "Probed switch chip: %s\n",
+		 sja1105_device_id_string_get(priv->device_id, priv->part_nr));
+
+	sja1105_dynamic_config_init(priv);
+
+	return dsa_register_switch(priv->ds);
+}
+
+static int sja1105_remove(struct spi_device *spi)
+{
+	struct sja1105_private *priv = spi_get_drvdata(spi);
+
+	dsa_unregister_switch(priv->ds);
+	sja1105_static_config_free(&priv->static_config);
+	return 0;
+}
+
+static const struct of_device_id sja1105_dt_ids[] = {
+	{ .compatible = "nxp,sja1105" },
+	{ /* sentinel */ },
+};
+MODULE_DEVICE_TABLE(of, sja1105_dt_ids);
+
+static struct spi_driver sja1105_driver = {
+	.driver = {
+		.name  = "sja1105",
+		.owner = THIS_MODULE,
+		.of_match_table = of_match_ptr(sja1105_dt_ids),
+	},
+	.probe  = sja1105_probe,
+	.remove = sja1105_remove,
+};
+
+module_spi_driver(sja1105_driver);
+
+MODULE_AUTHOR("Vladimir Oltean <olteanv@gmail.com>");
+MODULE_AUTHOR("Georg Waibel <georg.waibel@sensor-technik.de>");
+MODULE_DESCRIPTION("SJA1105 Driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/net/dsa/sja1105/sja1105_spi.c b/drivers/net/dsa/sja1105/sja1105_spi.c
new file mode 100644
index 000000000000..19220ab1f44c
--- /dev/null
+++ b/drivers/net/dsa/sja1105/sja1105_spi.c
@@ -0,0 +1,665 @@
+// SPDX-License-Identifier: BSD-3-Clause
+/* Copyright (c) 2016-2018, NXP Semiconductors
+ * Copyright (c) 2018, Sensor-Technik Wiedemann GmbH
+ * Copyright (c) 2018-2019, Vladimir Oltean <olteanv@gmail.com>
+ */
+#include <linux/spi/spi.h>
+#include <linux/packing.h>
+#include "sja1105.h"
+
+#define SPI_TRANSFER_SIZE_MAX  (SIZE_SPI_MSG_HEADER + SIZE_SPI_MSG_MAXLEN)
+
+static int sja1105_spi_transfer(const struct sja1105_private *priv,
+				const void *tx, void *rx, int size)
+{
+	struct spi_device *spi = priv->spidev;
+	struct spi_transfer transfer = {
+		.tx_buf = tx,
+		.rx_buf = rx,
+		.len = size,
+	};
+	struct spi_message msg;
+	int rc;
+
+	if (size > SPI_TRANSFER_SIZE_MAX) {
+		dev_err(&spi->dev, "SPI message (%d) longer than max of %d\n",
+			size, SPI_TRANSFER_SIZE_MAX);
+		return -EMSGSIZE;
+	}
+
+	spi_message_init(&msg);
+	spi_message_add_tail(&transfer, &msg);
+
+	rc = spi_sync(spi, &msg);
+	if (rc < 0) {
+		dev_err(&spi->dev, "SPI transfer failed: %d\n", rc);
+		return rc;
+	}
+
+	return rc;
+}
+
+static void
+sja1105_spi_message_pack(void *buf, const struct sja1105_spi_message *msg)
+{
+	const int size = SIZE_SPI_MSG_HEADER;
+
+	memset(buf, 0, size);
+
+	sja1105_pack(buf, &msg->access,     31, 31, size);
+	sja1105_pack(buf, &msg->read_count, 30, 25, size);
+	sja1105_pack(buf, &msg->address,    24,  4, size);
+}
+
+/* If read_or_write is:
+ *     * SPI_WRITE: creates and sends an SPI write message at absolute
+ *                  address reg_addr, taking size_bytes from *packed_buf
+ *     * SPI_READ: creates and sends an SPI read message from absolute
+ *                 address reg_addr, writing size_bytes into *packed_buf
+ *
+ * This function should only be called if it is priorly known that
+ * size_bytes is smaller than SIZE_SPI_MSG_MAXLEN. Larger packed buffers
+ * are chunked in smaller pieces by sja1105_spi_send_long_packed_buf below.
+ */
+int
+sja1105_spi_send_packed_buf(const struct sja1105_private *priv,
+			    enum sja1105_spi_access_mode read_or_write,
+			    u64 reg_addr, void *packed_buf, size_t size_bytes)
+{
+	const int msg_len = size_bytes + SIZE_SPI_MSG_HEADER;
+	struct sja1105_spi_message msg;
+	u8 tx_buf[SIZE_SPI_MSG_HEADER + SIZE_SPI_MSG_MAXLEN];
+	u8 rx_buf[SIZE_SPI_MSG_HEADER + SIZE_SPI_MSG_MAXLEN];
+	int rc;
+
+	if (msg_len > SIZE_SPI_MSG_HEADER + SIZE_SPI_MSG_MAXLEN)
+		return -ERANGE;
+
+	memset(rx_buf, 0, msg_len);
+
+	msg.access     = read_or_write;
+	msg.read_count = (read_or_write == SPI_READ) ? (size_bytes / 4) : 0;
+	msg.address    = reg_addr;
+	sja1105_spi_message_pack(tx_buf, &msg);
+
+	if (read_or_write == SPI_READ)
+		memset(tx_buf + SIZE_SPI_MSG_HEADER, 0, size_bytes);
+	else if (read_or_write == SPI_WRITE)
+		memcpy(tx_buf + SIZE_SPI_MSG_HEADER, packed_buf, size_bytes);
+	else
+		return -EINVAL;
+
+	rc = sja1105_spi_transfer(priv, tx_buf, rx_buf, msg_len);
+	if (rc < 0)
+		return rc;
+
+	if (read_or_write == SPI_READ)
+		memcpy(packed_buf, rx_buf + SIZE_SPI_MSG_HEADER, size_bytes);
+
+	return 0;
+}
+
+/* If read_or_write is:
+ *     * SPI_WRITE: creates and sends an SPI write message at absolute
+ *                  address reg_addr, taking size_bytes from *value
+ *     * SPI_READ: creates and sends an SPI read message from absolute
+ *                 address reg_addr, writing size_bytes into *value
+ *
+ * The u64 *value is unpacked, meaning that it's stored in the native
+ * CPU endianness and directly usable by software running on the core.
+ *
+ * This is a wrapper around sja1105_spi_send_packed_buf().
+ */
+int sja1105_spi_send_int(const struct sja1105_private *priv,
+			 enum sja1105_spi_access_mode read_or_write,
+			 u64 reg_addr, u64 *value, u64 size_bytes)
+{
+	u8 packed_buf[SIZE_SPI_MSG_MAXLEN];
+	int rc;
+
+	if (size_bytes > SIZE_SPI_MSG_MAXLEN)
+		return -ERANGE;
+
+	if (read_or_write == SPI_WRITE)
+		sja1105_pack(packed_buf, value, 8 * size_bytes - 1, 0,
+			     size_bytes);
+
+	rc = sja1105_spi_send_packed_buf(priv, read_or_write, reg_addr,
+					 packed_buf, size_bytes);
+
+	if (read_or_write == SPI_READ)
+		sja1105_unpack(packed_buf, value, 8 * size_bytes - 1, 0,
+			       size_bytes);
+
+	return rc;
+}
+
+/* Should be used if a packed_buf larger than SIZE_SPI_MSG_MAXLEN must be
+ * sent/received. Splitting the buffer into chunks and assembling those
+ * into SPI messages is done automatically by this function.
+ */
+int
+sja1105_spi_send_long_packed_buf(const struct sja1105_private *priv,
+				 enum sja1105_spi_access_mode read_or_write,
+				 u64 base_addr, void *packed_buf, u64 buf_len)
+{
+	struct chunk {
+		void *buf_ptr;
+		int len;
+		u64 spi_address;
+	} chunk;
+	int distance_to_end;
+	int rc = 0;
+
+	/* Initialize chunk */
+	chunk.buf_ptr = packed_buf;
+	chunk.spi_address = base_addr;
+	chunk.len = min_t(int, buf_len, SIZE_SPI_MSG_MAXLEN);
+
+	while (chunk.len) {
+		rc = sja1105_spi_send_packed_buf(priv, read_or_write,
+						 chunk.spi_address,
+						 chunk.buf_ptr, chunk.len);
+		if (rc < 0)
+			return rc;
+
+		chunk.buf_ptr += chunk.len;
+		chunk.spi_address += chunk.len / 4;
+		distance_to_end = (uintptr_t)((packed_buf + buf_len) -
+					chunk.buf_ptr);
+		chunk.len = min(distance_to_end, SIZE_SPI_MSG_MAXLEN);
+	}
+
+	return 0;
+}
+
+/* Back-ported structure from UM11040 Table 112.
+ * Reset control register (addr. 100440h)
+ * In the SJA1105 E/T, only warm_rst and cold_rst are
+ * supported (exposed in UM10944 as rst_ctrl), but the bit
+ * offsets of warm_rst and cold_rst are actually reversed.
+ */
+struct sja1105_reset_cmd {
+	u64 switch_rst;
+	u64 cfg_rst;
+	u64 car_rst;
+	u64 otp_rst;
+	u64 warm_rst;
+	u64 cold_rst;
+	u64 por_rst;
+};
+
+static void
+sja1105et_reset_cmd_pack(void *buf, const struct sja1105_reset_cmd *reset)
+{
+	const int size = 4;
+
+	memset(buf, 0, size);
+
+	sja1105_pack(buf, &reset->cold_rst, 3, 3, size);
+	sja1105_pack(buf, &reset->warm_rst, 2, 2, size);
+}
+
+static void
+sja1105pqrs_reset_cmd_pack(void *buf, const struct sja1105_reset_cmd *reset)
+{
+	const int size = 4;
+
+	memset(buf, 0, size);
+
+	sja1105_pack(buf, &reset->switch_rst, 8, 8, size);
+	sja1105_pack(buf, &reset->cfg_rst,    7, 7, size);
+	sja1105_pack(buf, &reset->car_rst,    5, 5, size);
+	sja1105_pack(buf, &reset->otp_rst,    4, 4, size);
+	sja1105_pack(buf, &reset->warm_rst,   3, 3, size);
+	sja1105_pack(buf, &reset->cold_rst,   2, 2, size);
+	sja1105_pack(buf, &reset->por_rst,    1, 1, size);
+}
+
+static int sja1105_reset_cmd_commit(const struct sja1105_private *priv,
+				    const struct sja1105_reset_cmd *reset)
+{
+#define BUF_LEN 4
+	struct device *dev = priv->ds->dev;
+	u8 packed_buf[BUF_LEN];
+
+	if (reset->switch_rst)
+		dev_dbg(dev, "Main reset for all functional modules requested\n");
+	if (reset->cfg_rst)
+		dev_dbg(dev, "Chip configuration reset requested\n");
+	if (reset->car_rst)
+		dev_dbg(dev, "Clock and reset control logic reset requested\n");
+	if (reset->otp_rst)
+		dev_dbg(dev, "OTP read cycle for reading product "
+			"config settings requested\n");
+	if (reset->warm_rst)
+		dev_dbg(dev, "Warm reset requested\n");
+	if (reset->cold_rst)
+		dev_dbg(dev, "Cold reset requested\n");
+	if (reset->por_rst)
+		dev_dbg(dev, "Power-on reset requested\n");
+
+	if ((reset->switch_rst || reset->cfg_rst || reset->car_rst ||
+	     reset->otp_rst || reset->por_rst) && IS_ET(priv->device_id)) {
+		dev_err(dev, "Only warm and cold reset is supported "
+			"for SJA1105 E/T!\n");
+		return -EINVAL;
+	}
+	if (IS_ET(priv->device_id))
+		sja1105et_reset_cmd_pack(packed_buf, reset);
+	else
+		sja1105pqrs_reset_cmd_pack(packed_buf, reset);
+
+	return sja1105_spi_send_packed_buf(priv, SPI_WRITE, priv->regs->rgu,
+					   packed_buf, BUF_LEN);
+#undef BUF_LEN
+}
+
+static int sja1105_cold_reset(const struct sja1105_private *priv)
+{
+	struct sja1105_reset_cmd reset = {0};
+
+	reset.cold_rst = 1;
+	return sja1105_reset_cmd_commit(priv, &reset);
+}
+
+static const char *SJA1105E_DEVICE_ID_STR   = "SJA1105E";
+static const char *SJA1105T_DEVICE_ID_STR   = "SJA1105T";
+static const char *SJA1105P_DEVICE_ID_STR   = "SJA1105P";
+static const char *SJA1105Q_DEVICE_ID_STR   = "SJA1105Q";
+static const char *SJA1105R_DEVICE_ID_STR   = "SJA1105R";
+static const char *SJA1105S_DEVICE_ID_STR   = "SJA1105S";
+static const char *SJA1105PR_DEVICE_ID_STR  = "SJA1105P or SJA1105R";
+static const char *SJA1105QS_DEVICE_ID_STR  = "SJA1105Q or SJA1105S";
+static const char *SJA1105_NO_DEVICE_ID_STR = "None";
+
+const char *sja1105_device_id_string_get(u64 device_id, u64 part_nr)
+{
+	if (device_id == SJA1105E_DEVICE_ID)
+		return SJA1105E_DEVICE_ID_STR;
+	if (device_id == SJA1105T_DEVICE_ID)
+		return SJA1105T_DEVICE_ID_STR;
+	/* P and R have same Device ID, and differ by Part Number.
+	 * Same do Q and S.
+	 */
+	if (IS_P(device_id, part_nr))
+		return SJA1105P_DEVICE_ID_STR;
+	if (IS_Q(device_id, part_nr))
+		return SJA1105Q_DEVICE_ID_STR;
+	if (IS_R(device_id, part_nr))
+		return SJA1105R_DEVICE_ID_STR;
+	if (IS_S(device_id, part_nr))
+		return SJA1105S_DEVICE_ID_STR;
+	/* Fallback: if we don't know/care what the part_nr is, and we
+	 * have a P/R, we can simply pass -1 to part_nr and have this
+	 * function say it's either P or R, instead of reporting it
+	 * as invalid.
+	 */
+	if (device_id == SJA1105PR_DEVICE_ID)
+		return SJA1105PR_DEVICE_ID_STR;
+	if (device_id == SJA1105QS_DEVICE_ID)
+		return SJA1105QS_DEVICE_ID_STR;
+
+	return SJA1105_NO_DEVICE_ID_STR;
+}
+
+struct sja1105_regs sja1105et_regs = {
+	.rgu = 0x100440,
+	.config = 0x020000,
+	.pad_mii_tx = {0x100800, 0x100802, 0x100804, 0x100806, 0x100808},
+	.rmii_pll1 = 0x10000A,
+	.cgu_idiv = {0x10000B, 0x10000C, 0x10000D, 0x10000E, 0x10000F},
+	/* UM10944.pdf, Table 86, ACU Register overview */
+	.rgmii_pad_mii_tx = {0x100800, 0x100802, 0x100804, 0x100806, 0x100808},
+	/* The base address is off-by-1 compared to UM10944,
+	 * because we are skipping device_id from the readout.
+	 */
+	.general_status = 0x1,
+	.mac = {0x200, 0x202, 0x204, 0x206, 0x208},
+	.mac_hl1 = {0x400, 0x410, 0x420, 0x430, 0x440},
+	.mac_hl2 = {0x600, 0x610, 0x620, 0x630, 0x640},
+	/* UM10944.pdf, Table 78, CGU Register overview */
+	.mii_tx_clk = {0x100013, 0x10001A, 0x100021, 0x100028, 0x10002F},
+	.mii_rx_clk = {0x100014, 0x10001B, 0x100022, 0x100029, 0x100030},
+	.mii_ext_tx_clk = {0x100018, 0x10001F, 0x100026, 0x10002D, 0x100034},
+	.mii_ext_rx_clk = {0x100019, 0x100020, 0x100027, 0x10002E, 0x100035},
+	.rgmii_txc = {0x100016, 0x10001D, 0x100024, 0x10002B, 0x100032},
+	.rmii_ref_clk = {0x100015, 0x10001C, 0x100023, 0x10002A, 0x100031},
+	.rmii_ext_tx_clk = {0x100018, 0x10001F, 0x100026, 0x10002D, 0x100034},
+};
+
+struct sja1105_regs sja1105pqrs_regs = {
+	.rgu = 0x100440,
+	.config = 0x020000,
+	.pad_mii_tx = {0x100800, 0x100802, 0x100804, 0x100806, 0x100808},
+	.rmii_pll1 = 0x10000A,
+	.cgu_idiv = {0x10000B, 0x10000C, 0x10000D, 0x10000E, 0x10000F},
+	/* UM10944.pdf, Table 86, ACU Register overview */
+	.rgmii_pad_mii_tx = {0x100800, 0x100802, 0x100804, 0x100806, 0x100808},
+	/* The base address is off-by-1 compared to UM10944,
+	 * because we are skipping device_id from the readout.
+	 */
+	.general_status = 0x1,
+	.mac = {0x200, 0x202, 0x204, 0x206, 0x208},
+	.mac_hl1 = {0x400, 0x410, 0x420, 0x430, 0x440},
+	.mac_hl2 = {0x600, 0x610, 0x620, 0x630, 0x640},
+	/* UM11040.pdf, Table 114 */
+	.mii_tx_clk = {0x100013, 0x100019, 0x10001F, 0x100025, 0x10002B},
+	.mii_rx_clk = {0x100014, 0x10001A, 0x100020, 0x100026, 0x10002C},
+	.mii_ext_tx_clk = {0x100017, 0x10001D, 0x100023, 0x100029, 0x10002F},
+	.mii_ext_rx_clk = {0x100018, 0x10001E, 0x100024, 0x10002A, 0x100030},
+	.rgmii_txc = {0x100016, 0x10001C, 0x100022, 0x100028, 0x10002E},
+	.rmii_ref_clk = {0x100015, 0x10001B, 0x100021, 0x100027, 0x10002D},
+	.rmii_ext_tx_clk = {0x100017, 0x10001D, 0x100023, 0x100029, 0x10002F},
+	.qlevel = {0x604, 0x614, 0x624, 0x634, 0x644},
+};
+
+/* Populates priv structures device_id, part_nr and regs */
+int sja1105_device_id_get(struct sja1105_private *priv)
+{
+#define DEVICE_ID_ADDR	0x0
+#define PROD_ID_ADDR	0x100BC3
+	/* These can't be part of regs, because otherwise we'd have
+	 * a chicken and egg problem
+	 */
+	u64 compatible_device_ids[] = {
+		SJA1105E_DEVICE_ID,
+		SJA1105T_DEVICE_ID,
+		SJA1105PR_DEVICE_ID,
+		SJA1105QS_DEVICE_ID,
+	};
+	struct device *dev = priv->ds->dev;
+	u64 tmp_device_id;
+	u64 tmp_part_nr;
+	unsigned int i;
+	int rc;
+
+	rc = sja1105_spi_send_int(priv, SPI_READ, DEVICE_ID_ADDR,
+				  &tmp_device_id, SIZE_SJA1105_DEVICE_ID);
+	if (rc < 0)
+		return rc;
+
+	priv->device_id = SJA1105_NO_DEVICE_ID;
+	for (i = 0; i < ARRAY_SIZE(compatible_device_ids); i++) {
+		if (tmp_device_id == compatible_device_ids[i]) {
+			priv->device_id = compatible_device_ids[i];
+			break;
+		}
+	}
+	if (priv->device_id == SJA1105_NO_DEVICE_ID) {
+		dev_err(dev, "Unrecognized Device ID 0x%llx\n", tmp_device_id);
+		return -EINVAL;
+	}
+	if (IS_PQRS(priv->device_id)) {
+		rc = sja1105_spi_send_int(priv, SPI_READ, PROD_ID_ADDR,
+					  &tmp_part_nr, 4);
+		if (rc < 0)
+			return -EINVAL;
+
+		sja1105_unpack(&tmp_part_nr, &priv->part_nr, 19, 4, 4);
+	}
+
+	if (IS_ET(priv->device_id))
+		priv->regs = &sja1105et_regs;
+	else if (IS_PQRS(priv->device_id))
+		priv->regs = &sja1105pqrs_regs;
+
+	return 0;
+#undef PROD_ID_ADDR
+#undef DEVICE_ID_ADDR
+}
+
+struct sja1105_general_status {
+	u64 configs;
+	u64 crcchkl;
+	u64 ids;
+	u64 crcchkg;
+	u64 nslot;
+	u64 vlind;
+	u64 vlparind;
+	u64 vlroutes;
+	u64 vlparts;
+	u64 macaddl;
+	u64 portenf;
+	u64 fwds_03h;
+	u64 macfds;
+	u64 enffds;
+	u64 l2busyfds;
+	u64 l2busys;
+	u64 macaddu;
+	u64 macaddhcl;
+	u64 vlanidhc;
+	u64 hashconfs;
+	u64 macaddhcu;
+	u64 wpvlanid;
+	u64 port_07h;
+	u64 vlanbusys;
+	u64 wrongports;
+	u64 vnotfounds;
+	u64 vlid;
+	u64 portvl;
+	u64 vlnotfound;
+	u64 emptys;
+	u64 buffers;
+	u64 buflwmark; /* Only on P/Q/R/S */
+	u64 port_0ah;
+	u64 fwds_0ah;
+	u64 parts;
+	u64 ramparerrl;
+	u64 ramparerru;
+};
+
+static void
+sja1105_general_status_unpack(void *buf, struct sja1105_general_status *status,
+			      u64 device_id)
+{
+	/* So that addition translates to 4 bytes */
+	u32 *p = (u32 *)buf;
+
+	memset(status, 0, sizeof(*status));
+	/* device_id is missing from the buffer, but we don't
+	 * want to diverge from the manual definition of the
+	 * register addresses, so we'll back off one step with
+	 * the register pointer, and never access p[0].
+	 */
+	p--;
+	sja1105_unpack(p + 0x1, &status->configs,   31, 31, 4);
+	sja1105_unpack(p + 0x1, &status->crcchkl,   30, 30, 4);
+	sja1105_unpack(p + 0x1, &status->ids,       29, 29, 4);
+	sja1105_unpack(p + 0x1, &status->crcchkg,   28, 28, 4);
+	sja1105_unpack(p + 0x1, &status->nslot,      3,  0, 4);
+	sja1105_unpack(p + 0x2, &status->vlind,     31, 16, 4);
+	sja1105_unpack(p + 0x2, &status->vlparind,  15,  8, 4);
+	sja1105_unpack(p + 0x2, &status->vlroutes,   1,  1, 4);
+	sja1105_unpack(p + 0x2, &status->vlparts,    0,  0, 4);
+	sja1105_unpack(p + 0x3, &status->macaddl,   31, 16, 4);
+	sja1105_unpack(p + 0x3, &status->portenf,   15,  8, 4);
+	sja1105_unpack(p + 0x3, &status->fwds_03h,   4,  4, 4);
+	sja1105_unpack(p + 0x3, &status->macfds,     3,  3, 4);
+	sja1105_unpack(p + 0x3, &status->enffds,     2,  2, 4);
+	sja1105_unpack(p + 0x3, &status->l2busyfds,  1,  1, 4);
+	sja1105_unpack(p + 0x3, &status->l2busys,    0,  0, 4);
+	sja1105_unpack(p + 0x4, &status->macaddu,   31,  0, 4);
+	sja1105_unpack(p + 0x5, &status->macaddhcl, 31, 16, 4);
+	sja1105_unpack(p + 0x5, &status->vlanidhc,  15,  4, 4);
+	sja1105_unpack(p + 0x5, &status->hashconfs,  0,  0, 4);
+	sja1105_unpack(p + 0x6, &status->macaddhcu, 31,  0, 4);
+	sja1105_unpack(p + 0x7, &status->wpvlanid,  31, 16, 4);
+	sja1105_unpack(p + 0x7, &status->port_07h,  15,  8, 4);
+	sja1105_unpack(p + 0x7, &status->vlanbusys,  4,  4, 4);
+	sja1105_unpack(p + 0x7, &status->wrongports, 3,  3, 4);
+	sja1105_unpack(p + 0x7, &status->vnotfounds, 2,  2, 4);
+	sja1105_unpack(p + 0x8, &status->vlid,      31, 16, 4);
+	sja1105_unpack(p + 0x8, &status->portvl,    15,  8, 4);
+	sja1105_unpack(p + 0x8, &status->vlnotfound, 0,  0, 4);
+	sja1105_unpack(p + 0x9, &status->emptys,    31, 31, 4);
+	sja1105_unpack(p + 0x9, &status->buffers,   30,  0, 4);
+	if (IS_ET(device_id)) {
+		sja1105_unpack(p + 0xA, &status->port_0ah,   15,  8, 4);
+		sja1105_unpack(p + 0xA, &status->fwds_0ah,    1,  1, 4);
+		sja1105_unpack(p + 0xA, &status->parts,       0,  0, 4);
+		sja1105_unpack(p + 0xB, &status->ramparerrl, 20,  0, 4);
+		sja1105_unpack(p + 0xC, &status->ramparerru,  4,  0, 4);
+	} else {
+		sja1105_unpack(p + 0xA, &status->buflwmark,  30,  0, 4);
+		sja1105_unpack(p + 0xB, &status->port_0ah,   15,  8, 4);
+		sja1105_unpack(p + 0xB, &status->fwds_0ah,    1,  1, 4);
+		sja1105_unpack(p + 0xB, &status->parts,       0,  0, 4);
+		sja1105_unpack(p + 0xC, &status->ramparerrl, 22,  0, 4);
+		sja1105_unpack(p + 0xD, &status->ramparerru,  4,  0, 4);
+	}
+}
+
+static int sja1105_general_status_get(struct sja1105_private *priv,
+				      struct sja1105_general_status *status)
+{
+#define SIZE_ET   (0x0C * 4) /* 0x01 to 0x0C */
+#define SIZE_PQRS (0x0D * 4) /* 0x01 to 0x0D */
+#define MAX_SIZE (max(SIZE_ET, SIZE_PQRS))
+	u8 packed_buf[MAX_SIZE];
+	const int size = IS_ET(priv->device_id) ? SIZE_ET : SIZE_PQRS;
+	int rc;
+
+	rc = sja1105_spi_send_packed_buf(priv, SPI_READ,
+					 priv->regs->general_status,
+					 packed_buf, size);
+	if (rc < 0)
+		return rc;
+
+	sja1105_general_status_unpack(packed_buf, status, priv->device_id);
+
+	return 0;
+#undef MAX_SIZE
+#undef SIZE_PQRS
+#undef SIZE_ET
+}
+
+/* Not const because unpacking priv->static_config into buffers and preparing
+ * for upload requires the recalculation of table CRCs and updating the
+ * structures with these.
+ */
+static int
+static_config_buf_prepare_for_upload(struct sja1105_private *priv,
+				     void *config_buf, int buf_len)
+{
+	struct sja1105_static_config *config = &priv->static_config;
+	enum sja1105_static_config_validity valid;
+	struct sja1105_table_header final_header;
+	char *final_header_ptr;
+	int crc_len;
+
+	valid = sja1105_static_config_check_valid(config);
+	if (valid != SJA1105_CONFIG_OK) {
+		dev_err(&priv->spidev->dev,
+			sja1105_static_config_error_msg[valid]);
+		return -EINVAL;
+	}
+
+	if (config->device_id != priv->device_id) {
+		dev_err(&priv->spidev->dev,
+			"The static config is for device id %llx "
+			"but the chip is %s (%llx)\n", config->device_id,
+			sja1105_device_id_string_get(priv->device_id,
+						     priv->part_nr),
+			priv->device_id);
+		return -EINVAL;
+	}
+
+	/* Write Device ID and config tables to config_buf */
+	sja1105_static_config_pack(config_buf, config);
+	/* Recalculate CRC of the last header (right now 0xDEADBEEF).
+	 * Don't include the CRC field itself.
+	 */
+	crc_len = buf_len - 4;
+	/* Read the whole table header */
+	final_header_ptr = config_buf + buf_len - SIZE_TABLE_HEADER;
+	sja1105_table_header_packing(final_header_ptr, &final_header, UNPACK);
+	/* Modify */
+	final_header.crc = sja1105_crc32(config_buf, crc_len);
+	/* Rewrite */
+	sja1105_table_header_packing(final_header_ptr, &final_header, PACK);
+
+	return 0;
+}
+
+int sja1105_static_config_upload(struct sja1105_private *priv)
+{
+#define RETRIES 10
+	struct sja1105_static_config *config = &priv->static_config;
+	struct device *dev = &priv->spidev->dev;
+	struct sja1105_general_status status;
+	int rc, retries = RETRIES;
+	u8 *config_buf;
+	int buf_len;
+
+	buf_len = sja1105_static_config_get_length(config);
+	config_buf = kcalloc(buf_len, sizeof(char), GFP_KERNEL);
+	if (!config_buf)
+		return -ENOMEM;
+
+	rc = static_config_buf_prepare_for_upload(priv, config_buf, buf_len);
+	if (rc < 0) {
+		dev_err(dev, "Invalid config, cannot upload\n");
+		return -EINVAL;
+	}
+	do {
+		/* Put the SJA1105 in programming mode */
+		rc = sja1105_cold_reset(priv);
+		if (rc < 0) {
+			dev_err(dev, "Failed to reset switch, retrying...\n");
+			continue;
+		}
+		/* Wait for the switch to come out of reset */
+		usleep_range(1000, 5000);
+		/* Upload the static config to the device */
+		rc = sja1105_spi_send_long_packed_buf(priv, SPI_WRITE,
+						      priv->regs->config,
+						      config_buf, buf_len);
+		if (rc < 0) {
+			dev_err(dev, "Failed to upload config, retrying...\n");
+			continue;
+		}
+		/* Check that SJA1105 responded well to the config upload */
+		rc = sja1105_general_status_get(priv, &status);
+		if (rc < 0)
+			continue;
+
+		if (status.ids == 1) {
+			dev_err(dev, "Mismatch between hardware and staging area "
+				"device id. Wrote 0x%llx, wants 0x%llx\n",
+				config->device_id, priv->device_id);
+			continue;
+		}
+		if (status.crcchkl == 1) {
+			dev_err(dev, "Switch reported invalid local CRC on "
+				"the uploaded config, retrying...\n");
+			continue;
+		}
+		if (status.crcchkg == 1) {
+			dev_err(dev, "Switch reported invalid global CRC on "
+				"the uploaded config, retrying...\n");
+			continue;
+		}
+		if (status.configs == 0) {
+			dev_err(dev, "Switch reported that configuration is "
+				"invalid, retrying...\n");
+			continue;
+		}
+	} while (--retries && (status.crcchkl == 1 || status.crcchkg == 1 ||
+		 status.configs == 0 || status.ids == 1));
+
+	if (!retries) {
+		rc = -EIO;
+		dev_err(dev, "Failed to upload config to device, giving up\n");
+		goto out;
+	} else if (retries != RETRIES - 1) {
+		dev_info(dev, "Succeeded after %d tried\n", RETRIES - retries);
+	}
+
+	dev_info(dev, "Reset switch and programmed static config\n");
+out:
+	kfree(config_buf);
+	return rc;
+#undef RETRIES
+}
+
diff --git a/drivers/net/dsa/sja1105/sja1105_static_config.c b/drivers/net/dsa/sja1105/sja1105_static_config.c
new file mode 100644
index 000000000000..f5d1866216bd
--- /dev/null
+++ b/drivers/net/dsa/sja1105/sja1105_static_config.c
@@ -0,0 +1,1036 @@
+// SPDX-License-Identifier: BSD-3-Clause
+/* Copyright (c) 2016-2018, NXP Semiconductors
+ * Copyright (c) 2018-2019, Vladimir Oltean <olteanv@gmail.com>
+ */
+#include "sja1105_static_config.h"
+#include <linux/crc32.h>
+#include <linux/slab.h>
+#include <linux/string.h>
+#include <linux/errno.h>
+
+/* Convenience wrappers over the generic packing functions. These take into
+ * account the SJA1105 memory layout quirks and provide some level of
+ * programmer protection against incorrect API use. The errors are not expected
+ * to occur durring runtime, therefore printing and swallowing them here is
+ * appropriate instead of clutterring up higher-level code.
+ */
+void sja1105_pack(void *buf, const u64 *val, int start, int end, size_t len)
+{
+	int rc = packing(buf, (u64 *)val, start, end, len,
+			 PACK, QUIRK_LSW32_IS_FIRST);
+
+	if (likely(!rc))
+		return;
+
+	if (rc == -EINVAL) {
+		pr_err("Start bit (%d) expected to be larger than end (%d)\n",
+		       start, end);
+	} else if (rc == -ERANGE) {
+		if ((start - end + 1) > 64)
+			pr_err("Field %d-%d too large for 64 bits!\n",
+			       start, end);
+		else
+			pr_err("Cannot store %llx inside bits %d-%d (would truncate)\n",
+			       *val, start, end);
+	}
+	dump_stack();
+}
+
+void sja1105_unpack(const void *buf, u64 *val, int start, int end, size_t len)
+{
+	int rc = packing((void *)buf, val, start, end, len,
+			 UNPACK, QUIRK_LSW32_IS_FIRST);
+
+	if (likely(!rc))
+		return;
+
+	if (rc == -EINVAL)
+		pr_err("Start bit (%d) expected to be larger than end (%d)\n",
+		       start, end);
+	else if (rc == -ERANGE)
+		pr_err("Field %d-%d too large for 64 bits!\n",
+		       start, end);
+	dump_stack();
+}
+
+void sja1105_packing(void *buf, u64 *val, int start, int end,
+		     size_t len, enum packing_op op)
+{
+	int rc = packing(buf, val, start, end, len, op, QUIRK_LSW32_IS_FIRST);
+
+	if (likely(!rc))
+		return;
+
+	if (rc == -EINVAL) {
+		pr_err("Start bit (%d) expected to be larger than end (%d)\n",
+		       start, end);
+	} else if (rc == -ERANGE) {
+		if ((start - end + 1) > 64)
+			pr_err("Field %d-%d too large for 64 bits!\n",
+			       start, end);
+		else
+			pr_err("Cannot store %llx inside bits %d-%d (would truncate)\n",
+			       *val, start, end);
+	}
+	dump_stack();
+}
+
+/* Little-endian Ethernet CRC32 of data packed as big-endian u32 words */
+u32 sja1105_crc32(const void *buf, size_t len)
+{
+	unsigned int i;
+	u64 word;
+	u32 crc;
+
+	/* seed */
+	crc = ~0;
+	for (i = 0; i < len; i += 4) {
+		sja1105_unpack((void *)buf + i, &word, 31, 0, 4);
+		crc = crc32_le(crc, (u8 *)&word, 4);
+	}
+	return ~crc;
+}
+
+static size_t sja1105et_general_params_entry_packing(void *buf, void *entry_ptr,
+						     enum packing_op op)
+{
+	const size_t size = SIZE_GENERAL_PARAMS_ENTRY_ET;
+	struct sja1105_general_params_entry *entry;
+
+	entry = (struct sja1105_general_params_entry *)entry_ptr;
+
+	sja1105_packing(buf, &entry->vllupformat, 319, 319, size, op);
+	sja1105_packing(buf, &entry->mirr_ptacu,  318, 318, size, op);
+	sja1105_packing(buf, &entry->switchid,    317, 315, size, op);
+	sja1105_packing(buf, &entry->hostprio,    314, 312, size, op);
+	sja1105_packing(buf, &entry->mac_fltres1, 311, 264, size, op);
+	sja1105_packing(buf, &entry->mac_fltres0, 263, 216, size, op);
+	sja1105_packing(buf, &entry->mac_flt1,    215, 168, size, op);
+	sja1105_packing(buf, &entry->mac_flt0,    167, 120, size, op);
+	sja1105_packing(buf, &entry->incl_srcpt1, 119, 119, size, op);
+	sja1105_packing(buf, &entry->incl_srcpt0, 118, 118, size, op);
+	sja1105_packing(buf, &entry->send_meta1,  117, 117, size, op);
+	sja1105_packing(buf, &entry->send_meta0,  116, 116, size, op);
+	sja1105_packing(buf, &entry->casc_port,   115, 113, size, op);
+	sja1105_packing(buf, &entry->host_port,   112, 110, size, op);
+	sja1105_packing(buf, &entry->mirr_port,   109, 107, size, op);
+	sja1105_packing(buf, &entry->vlmarker,    106,  75, size, op);
+	sja1105_packing(buf, &entry->vlmask,       74,  43, size, op);
+	sja1105_packing(buf, &entry->tpid,         42,  27, size, op);
+	sja1105_packing(buf, &entry->ignore2stf,   26,  26, size, op);
+	sja1105_packing(buf, &entry->tpid2,        25,  10, size, op);
+	return size;
+}
+
+static size_t
+sja1105pqrs_general_params_entry_packing(void *buf, void *entry_ptr,
+					 enum packing_op op)
+{
+	const size_t size = SIZE_GENERAL_PARAMS_ENTRY_PQRS;
+	struct sja1105_general_params_entry *entry;
+
+	entry = (struct sja1105_general_params_entry *)entry_ptr;
+
+	sja1105_packing(buf, &entry->vllupformat, 351, 351, size, op);
+	sja1105_packing(buf, &entry->mirr_ptacu,  350, 350, size, op);
+	sja1105_packing(buf, &entry->switchid,    349, 347, size, op);
+	sja1105_packing(buf, &entry->hostprio,    346, 344, size, op);
+	sja1105_packing(buf, &entry->mac_fltres1, 343, 296, size, op);
+	sja1105_packing(buf, &entry->mac_fltres0, 295, 248, size, op);
+	sja1105_packing(buf, &entry->mac_flt1,    247, 200, size, op);
+	sja1105_packing(buf, &entry->mac_flt0,    199, 152, size, op);
+	sja1105_packing(buf, &entry->incl_srcpt1, 151, 151, size, op);
+	sja1105_packing(buf, &entry->incl_srcpt0, 150, 150, size, op);
+	sja1105_packing(buf, &entry->send_meta1,  149, 149, size, op);
+	sja1105_packing(buf, &entry->send_meta0,  148, 148, size, op);
+	sja1105_packing(buf, &entry->casc_port,   147, 145, size, op);
+	sja1105_packing(buf, &entry->host_port,   144, 142, size, op);
+	sja1105_packing(buf, &entry->mirr_port,   141, 139, size, op);
+	sja1105_packing(buf, &entry->vlmarker,    138, 107, size, op);
+	sja1105_packing(buf, &entry->vlmask,      106,  75, size, op);
+	sja1105_packing(buf, &entry->tpid,         74,  59, size, op);
+	sja1105_packing(buf, &entry->ignore2stf,   58,  58, size, op);
+	sja1105_packing(buf, &entry->tpid2,        57,  42, size, op);
+	sja1105_packing(buf, &entry->queue_ts,     41,  41, size, op);
+	sja1105_packing(buf, &entry->egrmirrvid,   40,  29, size, op);
+	sja1105_packing(buf, &entry->egrmirrpcp,   28,  26, size, op);
+	sja1105_packing(buf, &entry->egrmirrdei,   25,  25, size, op);
+	sja1105_packing(buf, &entry->replay_port,  24,  22, size, op);
+	return size;
+}
+
+static size_t
+sja1105_l2_forwarding_params_entry_packing(void *buf, void *entry_ptr,
+					   enum packing_op op)
+{
+	const size_t size = SIZE_L2_FORWARDING_PARAMS_ENTRY;
+	struct sja1105_l2_forwarding_params_entry *entry;
+	int offset, i;
+
+	entry = (struct sja1105_l2_forwarding_params_entry *)entry_ptr;
+
+	sja1105_packing(buf, &entry->max_dynp, 95, 93, size, op);
+	for (i = 0, offset = 13; i < 8; i++, offset += 10)
+		sja1105_packing(buf, &entry->part_spc[i],
+				offset + 9, offset + 0, size, op);
+	return size;
+}
+
+size_t sja1105_l2_forwarding_entry_packing(void *buf, void *entry_ptr,
+					   enum packing_op op)
+{
+	const size_t size = SIZE_L2_FORWARDING_ENTRY;
+	struct sja1105_l2_forwarding_entry *entry;
+	int offset, i;
+
+	entry = (struct sja1105_l2_forwarding_entry *)entry_ptr;
+
+	sja1105_packing(buf, &entry->bc_domain,  63, 59, size, op);
+	sja1105_packing(buf, &entry->reach_port, 58, 54, size, op);
+	sja1105_packing(buf, &entry->fl_domain,  53, 49, size, op);
+	for (i = 0, offset = 25; i < 8; i++, offset += 3)
+		sja1105_packing(buf, &entry->vlan_pmap[i],
+				offset + 2, offset + 0, size, op);
+	return size;
+}
+
+static size_t
+sja1105et_l2_lookup_params_entry_packing(void *buf, void *entry_ptr,
+					 enum packing_op op)
+{
+	const size_t size = SIZE_L2_LOOKUP_PARAMS_ENTRY_ET;
+	struct sja1105_l2_lookup_params_entry *entry;
+
+	entry = (struct sja1105_l2_lookup_params_entry *)entry_ptr;
+
+	sja1105_packing(buf, &entry->maxage,         31, 17, size, op);
+	sja1105_packing(buf, &entry->dyn_tbsz,       16, 14, size, op);
+	sja1105_packing(buf, &entry->poly,           13,  6, size, op);
+	sja1105_packing(buf, &entry->shared_learn,    5,  5, size, op);
+	sja1105_packing(buf, &entry->no_enf_hostprt,  4,  4, size, op);
+	sja1105_packing(buf, &entry->no_mgmt_learn,   3,  3, size, op);
+	return size;
+}
+
+static size_t
+sja1105pqrs_l2_lookup_params_entry_packing(void *buf, void *entry_ptr,
+					   enum packing_op op)
+{
+	const size_t size = SIZE_L2_LOOKUP_PARAMS_ENTRY_PQRS;
+	struct sja1105_l2_lookup_params_entry *entry;
+	int offset, i;
+
+	entry = (struct sja1105_l2_lookup_params_entry *)entry_ptr;
+
+	sja1105_packing(buf, &entry->drpbc,         127, 123, size, op);
+	sja1105_packing(buf, &entry->drpmc,         122, 118, size, op);
+	sja1105_packing(buf, &entry->drpuni,        117, 113, size, op);
+	for (i = 0, offset = 58; i < 5; i++, offset += 11)
+		sja1105_packing(buf, &entry->maxaddrp[i],
+				offset + 10, offset + 0, size, op);
+	sja1105_packing(buf, &entry->maxage,         57,  43, size, op);
+	sja1105_packing(buf, &entry->start_dynspc,   42,  33, size, op);
+	sja1105_packing(buf, &entry->drpnolearn,     32,  28, size, op);
+	sja1105_packing(buf, &entry->shared_learn,   27,  27, size, op);
+	sja1105_packing(buf, &entry->no_enf_hostprt, 26,  26, size, op);
+	sja1105_packing(buf, &entry->no_mgmt_learn,  25,  25, size, op);
+	sja1105_packing(buf, &entry->use_static,     24,  24, size, op);
+	sja1105_packing(buf, &entry->owr_dyn,        23,  23, size, op);
+	sja1105_packing(buf, &entry->learn_once,     22,  22, size, op);
+	return size;
+}
+
+size_t sja1105et_l2_lookup_entry_packing(void *buf, void *entry_ptr,
+					 enum packing_op op)
+{
+	const size_t size = SIZE_L2_LOOKUP_ENTRY_ET;
+	struct sja1105_l2_lookup_entry *entry;
+
+	entry = (struct sja1105_l2_lookup_entry *)entry_ptr;
+
+	sja1105_packing(buf, &entry->vlanid,    95, 84, size, op);
+	sja1105_packing(buf, &entry->macaddr,   83, 36, size, op);
+	sja1105_packing(buf, &entry->destports, 35, 31, size, op);
+	sja1105_packing(buf, &entry->enfport,   30, 30, size, op);
+	sja1105_packing(buf, &entry->index,     29, 20, size, op);
+	return size;
+}
+
+size_t sja1105pqrs_l2_lookup_entry_packing(void *buf, void *entry_ptr,
+					   enum packing_op op)
+{
+	const size_t size = SIZE_L2_LOOKUP_ENTRY_PQRS;
+	struct sja1105_l2_lookup_entry *entry;
+
+	entry = (struct sja1105_l2_lookup_entry *)entry_ptr;
+
+	/* These are static L2 lookup entries, so the structure
+	 * should match UM11040 Table 16/17 definitions when
+	 * LOCKEDS is 1.
+	 */
+	sja1105_packing(buf, &entry->mirrvlan,     158, 147, size, op);
+	sja1105_packing(buf, &entry->mirr,         145, 145, size, op);
+	sja1105_packing(buf, &entry->retag,        144, 144, size, op);
+	sja1105_packing(buf, &entry->mask_iotag,   143, 143, size, op);
+	sja1105_packing(buf, &entry->mask_vlanid,  142, 131, size, op);
+	sja1105_packing(buf, &entry->mask_macaddr, 130,  83, size, op);
+	sja1105_packing(buf, &entry->iotag,         82,  82, size, op);
+	sja1105_packing(buf, &entry->vlanid,        81,  70, size, op);
+	sja1105_packing(buf, &entry->macaddr,       69,  22, size, op);
+	sja1105_packing(buf, &entry->destports,     21,  17, size, op);
+	sja1105_packing(buf, &entry->enfport,       16,  16, size, op);
+	sja1105_packing(buf, &entry->index,         15,   6, size, op);
+	return size;
+}
+
+static size_t sja1105_l2_policing_entry_packing(void *buf, void *entry_ptr,
+						enum packing_op op)
+{
+	const size_t size = SIZE_L2_POLICING_ENTRY;
+	struct sja1105_l2_policing_entry *entry;
+
+	entry = (struct sja1105_l2_policing_entry *)entry_ptr;
+
+	sja1105_packing(buf, &entry->sharindx,  63, 58, size, op);
+	sja1105_packing(buf, &entry->smax,      57, 42, size, op);
+	sja1105_packing(buf, &entry->rate,      41, 26, size, op);
+	sja1105_packing(buf, &entry->maxlen,    25, 15, size, op);
+	sja1105_packing(buf, &entry->partition, 14, 12, size, op);
+	return size;
+}
+
+static size_t sja1105et_mac_config_entry_packing(void *buf, void *entry_ptr,
+						 enum packing_op op)
+{
+	const size_t size = SIZE_MAC_CONFIG_ENTRY_ET;
+	struct sja1105_mac_config_entry *entry;
+	int offset, i;
+
+	entry = (struct sja1105_mac_config_entry *)entry_ptr;
+
+	for (i = 0, offset = 72; i < 8; i++, offset += 19) {
+		sja1105_packing(buf, &entry->enabled[i],
+				offset +  0, offset +  0, size, op);
+		sja1105_packing(buf, &entry->base[i],
+				offset +  9, offset +  1, size, op);
+		sja1105_packing(buf, &entry->top[i],
+				offset + 18, offset + 10, size, op);
+	}
+	sja1105_packing(buf, &entry->ifg,       71, 67, size, op);
+	sja1105_packing(buf, &entry->speed,     66, 65, size, op);
+	sja1105_packing(buf, &entry->tp_delin,  64, 49, size, op);
+	sja1105_packing(buf, &entry->tp_delout, 48, 33, size, op);
+	sja1105_packing(buf, &entry->maxage,    32, 25, size, op);
+	sja1105_packing(buf, &entry->vlanprio,  24, 22, size, op);
+	sja1105_packing(buf, &entry->vlanid,    21, 10, size, op);
+	sja1105_packing(buf, &entry->ing_mirr,   9,  9, size, op);
+	sja1105_packing(buf, &entry->egr_mirr,   8,  8, size, op);
+	sja1105_packing(buf, &entry->drpnona664, 7,  7, size, op);
+	sja1105_packing(buf, &entry->drpdtag,    6,  6, size, op);
+	sja1105_packing(buf, &entry->drpuntag,   5,  5, size, op);
+	sja1105_packing(buf, &entry->retag,      4,  4, size, op);
+	sja1105_packing(buf, &entry->dyn_learn,  3,  3, size, op);
+	sja1105_packing(buf, &entry->egress,     2,  2, size, op);
+	sja1105_packing(buf, &entry->ingress,    1,  1, size, op);
+	return size;
+}
+
+size_t sja1105pqrs_mac_config_entry_packing(void *buf, void *entry_ptr,
+					    enum packing_op op)
+{
+	const size_t size = SIZE_MAC_CONFIG_ENTRY_PQRS;
+	struct sja1105_mac_config_entry *entry;
+	int offset, i;
+
+	entry = (struct sja1105_mac_config_entry *)entry_ptr;
+
+	for (i = 0, offset = 104; i < 8; i++, offset += 19) {
+		sja1105_packing(buf, &entry->enabled[i],
+				offset +  0, offset +  0, size, op);
+		sja1105_packing(buf, &entry->base[i],
+				offset +  9, offset +  1, size, op);
+		sja1105_packing(buf, &entry->top[i],
+				offset + 18, offset + 10, size, op);
+	}
+	sja1105_packing(buf, &entry->ifg,       103, 99, size, op);
+	sja1105_packing(buf, &entry->speed,      98, 97, size, op);
+	sja1105_packing(buf, &entry->tp_delin,   96, 81, size, op);
+	sja1105_packing(buf, &entry->tp_delout,  80, 65, size, op);
+	sja1105_packing(buf, &entry->maxage,     64, 57, size, op);
+	sja1105_packing(buf, &entry->vlanprio,   56, 54, size, op);
+	sja1105_packing(buf, &entry->vlanid,     53, 42, size, op);
+	sja1105_packing(buf, &entry->ing_mirr,   41, 41, size, op);
+	sja1105_packing(buf, &entry->egr_mirr,   40, 40, size, op);
+	sja1105_packing(buf, &entry->drpnona664, 39, 39, size, op);
+	sja1105_packing(buf, &entry->drpdtag,    38, 38, size, op);
+	sja1105_packing(buf, &entry->drpsotag,   37, 37, size, op);
+	sja1105_packing(buf, &entry->drpsitag,   36, 36, size, op);
+	sja1105_packing(buf, &entry->drpuntag,   35, 35, size, op);
+	sja1105_packing(buf, &entry->retag,      34, 34, size, op);
+	sja1105_packing(buf, &entry->dyn_learn,  33, 33, size, op);
+	sja1105_packing(buf, &entry->egress,     32, 32, size, op);
+	sja1105_packing(buf, &entry->ingress,    31, 31, size, op);
+	sja1105_packing(buf, &entry->mirrcie,    30, 30, size, op);
+	sja1105_packing(buf, &entry->mirrcetag,  29, 29, size, op);
+	sja1105_packing(buf, &entry->ingmirrvid, 28, 17, size, op);
+	sja1105_packing(buf, &entry->ingmirrpcp, 16, 14, size, op);
+	sja1105_packing(buf, &entry->ingmirrdei, 13, 13, size, op);
+	return size;
+}
+
+size_t sja1105_vlan_lookup_entry_packing(void *buf, void *entry_ptr,
+					 enum packing_op op)
+{
+	const size_t size = SIZE_VLAN_LOOKUP_ENTRY;
+	struct sja1105_vlan_lookup_entry *entry;
+
+	entry = (struct sja1105_vlan_lookup_entry *)entry_ptr;
+
+	sja1105_packing(buf, &entry->ving_mirr,  63, 59, size, op);
+	sja1105_packing(buf, &entry->vegr_mirr,  58, 54, size, op);
+	sja1105_packing(buf, &entry->vmemb_port, 53, 49, size, op);
+	sja1105_packing(buf, &entry->vlan_bc,    48, 44, size, op);
+	sja1105_packing(buf, &entry->tag_port,   43, 39, size, op);
+	sja1105_packing(buf, &entry->vlanid,     38, 27, size, op);
+	return size;
+}
+
+static size_t sja1105_xmii_params_entry_packing(void *buf, void *entry_ptr,
+						enum packing_op op)
+{
+	const size_t size = SIZE_XMII_PARAMS_ENTRY;
+	struct sja1105_xmii_params_entry *entry;
+	int offset, i;
+
+	entry = (struct sja1105_xmii_params_entry *)entry_ptr;
+
+	for (i = 0, offset = 17; i < 5; i++, offset += 3) {
+		sja1105_packing(buf, &entry->xmii_mode[i],
+				offset + 1, offset + 0, size, op);
+		sja1105_packing(buf, &entry->phy_mac[i],
+				offset + 2, offset + 2, size, op);
+	}
+	return size;
+}
+
+size_t sja1105_table_header_packing(void *buf, void *entry_ptr,
+				    enum packing_op op)
+{
+	const size_t size = SIZE_TABLE_HEADER;
+	struct sja1105_table_header *entry;
+
+	entry = (struct sja1105_table_header *)entry_ptr;
+
+	sja1105_packing(buf, &entry->block_id, 31, 24, size, op);
+	sja1105_packing(buf, &entry->len,      55, 32, size, op);
+	sja1105_packing(buf, &entry->crc,      95, 64, size, op);
+	return size;
+}
+
+/* WARNING: the *hdr pointer is really non-const, because it is
+ * modifying the CRC of the header for a 2-stage packing operation
+ */
+void
+sja1105_table_header_pack_with_crc(void *buf, struct sja1105_table_header *hdr)
+{
+	/* First pack the table as-is, then calculate the CRC, and
+	 * finally put the proper CRC into the packed buffer
+	 */
+	memset(buf, 0, SIZE_TABLE_HEADER);
+	sja1105_table_header_packing(buf, hdr, PACK);
+	hdr->crc = sja1105_crc32(buf, SIZE_TABLE_HEADER - 4);
+	sja1105_pack(buf + SIZE_TABLE_HEADER - 4, &hdr->crc, 31, 0, 4);
+}
+
+static void sja1105_table_write_crc(u8 *table_start, u8 *crc_ptr)
+{
+	u64 computed_crc;
+	int len_bytes;
+
+	len_bytes = (uintptr_t)(crc_ptr - table_start);
+	computed_crc = sja1105_crc32(table_start, len_bytes);
+	sja1105_pack(crc_ptr, &computed_crc, 31, 0, 4);
+}
+
+/* The block IDs that the switches support are unfortunately sparse, so keep a
+ * mapping table to "block indices" and translate back and forth so that we
+ * don't waste useless memory in struct sja1105_static_config.
+ * Also, since the block id comes from essentially untrusted input (unpacking
+ * the static config from userspace) it has to be sanitized (range-checked)
+ * before blindly indexing kernel memory with the blk_idx.
+ */
+static u64 blk_id_map[BLK_IDX_MAX] = {
+	[BLK_IDX_L2_LOOKUP] = BLKID_L2_LOOKUP,
+	[BLK_IDX_L2_POLICING] = BLKID_L2_POLICING,
+	[BLK_IDX_VLAN_LOOKUP] = BLKID_VLAN_LOOKUP,
+	[BLK_IDX_L2_FORWARDING] = BLKID_L2_FORWARDING,
+	[BLK_IDX_MAC_CONFIG] = BLKID_MAC_CONFIG,
+	[BLK_IDX_L2_LOOKUP_PARAMS] = BLKID_L2_LOOKUP_PARAMS,
+	[BLK_IDX_L2_FORWARDING_PARAMS] = BLKID_L2_FORWARDING_PARAMS,
+	[BLK_IDX_GENERAL_PARAMS] = BLKID_GENERAL_PARAMS,
+	[BLK_IDX_XMII_PARAMS] = BLKID_XMII_PARAMS,
+};
+
+const char *sja1105_static_config_error_msg[] = {
+	[SJA1105_CONFIG_OK] = "",
+	[SJA1105_DEVICE_ID_INVALID] =
+		"Device ID present in the static config is invalid",
+	[SJA1105_MISSING_L2_POLICING_TABLE] =
+		"l2-policing-table needs to have at least one entry",
+	[SJA1105_MISSING_L2_FORWARDING_TABLE] =
+		"l2-forwarding-table is either missing or incomplete",
+	[SJA1105_MISSING_L2_FORWARDING_PARAMS_TABLE] =
+		"l2-forwarding-parameters-table is missing",
+	[SJA1105_MISSING_GENERAL_PARAMS_TABLE] =
+		"general-parameters-table is missing",
+	[SJA1105_MISSING_VLAN_TABLE] =
+		"vlan-lookup-table needs to have at least the default untagged VLAN",
+	[SJA1105_MISSING_XMII_TABLE] =
+		"xmii-table is missing",
+	[SJA1105_MISSING_MAC_TABLE] =
+		"mac-configuration-table needs to contain an entry for each port",
+	[SJA1105_OVERCOMMITTED_FRAME_MEMORY] =
+		"Not allowed to overcommit frame memory. L2 memory partitions "
+		"and VL memory partitions share the same space. The sum of all "
+		"16 memory partitions is not allowed to be larger than 929 "
+		"128-byte blocks (or 910 with retagging). Please adjust "
+		"l2-forwarding-parameters-table.part_spc and/or "
+		"vl-forwarding-parameters-table.partspc.",
+	[SJA1105_UNEXPECTED_END_OF_BUFFER] =
+		"Unexpected end of buffer",
+	[SJA1105_INVALID_DEVICE_ID] =
+		"Invalid device ID present in static config",
+	[SJA1105_INVALID_TABLE_HEADER_CRC] =
+		"One of the table headers has an incorrect CRC",
+	[SJA1105_INVALID_TABLE_HEADER] =
+		"One of the table headers contains an invalid block id",
+	[SJA1105_INCORRECT_TABLE_LENGTH] =
+		"The data length specified in one of the table headers is "
+		"longer than the actual size of the entries that were parsed",
+	[SJA1105_DATA_CRC_INVALID] =
+		"One of the tables has an incorrect CRC over the data area",
+	[SJA1105_EXTRA_BYTES_AT_END_OF_BUFFER] =
+		"Extra bytes found at the end of buffer after parsing it",
+};
+
+static enum sja1105_static_config_validity
+static_config_check_memory_size(const struct sja1105_table *tables)
+{
+	const struct sja1105_l2_forwarding_params_entry *l2_fwd_params;
+	int i, mem = 0;
+
+	l2_fwd_params = tables[BLK_IDX_L2_FORWARDING_PARAMS].entries;
+
+	for (i = 0; i < 8; i++)
+		mem += l2_fwd_params->part_spc[i];
+
+	if (mem > MAX_FRAME_MEMORY)
+		return SJA1105_OVERCOMMITTED_FRAME_MEMORY;
+
+	return SJA1105_CONFIG_OK;
+}
+
+enum sja1105_static_config_validity
+sja1105_static_config_check_valid(const struct sja1105_static_config *config)
+{
+	const struct sja1105_table *tables = config->tables;
+#define IS_FULL(blk_idx) \
+	(tables[blk_idx].entry_count == tables[blk_idx].ops->max_entry_count)
+
+	if (!DEVICE_ID_VALID(config->device_id))
+		return SJA1105_DEVICE_ID_INVALID;
+
+	if (tables[BLK_IDX_L2_POLICING].entry_count == 0)
+		return SJA1105_MISSING_L2_POLICING_TABLE;
+
+	if (tables[BLK_IDX_VLAN_LOOKUP].entry_count == 0)
+		return SJA1105_MISSING_VLAN_TABLE;
+
+	if (!IS_FULL(BLK_IDX_L2_FORWARDING))
+		return SJA1105_MISSING_L2_FORWARDING_TABLE;
+
+	if (!IS_FULL(BLK_IDX_MAC_CONFIG))
+		return SJA1105_MISSING_MAC_TABLE;
+
+	if (!IS_FULL(BLK_IDX_L2_FORWARDING_PARAMS))
+		return SJA1105_MISSING_L2_FORWARDING_PARAMS_TABLE;
+
+	if (!IS_FULL(BLK_IDX_GENERAL_PARAMS))
+		return SJA1105_MISSING_GENERAL_PARAMS_TABLE;
+
+	if (!IS_FULL(BLK_IDX_XMII_PARAMS))
+		return SJA1105_MISSING_XMII_TABLE;
+
+	return static_config_check_memory_size(tables);
+#undef IS_FULL
+}
+
+void
+sja1105_static_config_pack(void *buf, struct sja1105_static_config *config)
+{
+	struct sja1105_table_header header = {0};
+	enum sja1105_blk_idx i;
+	char *p = buf;
+	int j;
+
+	sja1105_pack(p, &config->device_id, 31, 0, 4);
+	p += SIZE_SJA1105_DEVICE_ID;
+
+	for (i = 0; i < BLK_IDX_MAX; i++) {
+		const struct sja1105_table *table;
+		char *table_start;
+
+		table = &config->tables[i];
+		if (!table->entry_count)
+			continue;
+
+		header.block_id = blk_id_map[i];
+		header.len = table->entry_count *
+			     table->ops->packed_entry_size / 4;
+		sja1105_table_header_pack_with_crc(p, &header);
+		p += SIZE_TABLE_HEADER;
+		table_start = p;
+		for (j = 0; j < table->entry_count; j++) {
+			u8 *entry_ptr = table->entries;
+
+			entry_ptr += j * table->ops->unpacked_entry_size;
+			memset(p, 0, table->ops->packed_entry_size);
+			table->ops->packing(p, entry_ptr, PACK);
+			p += table->ops->packed_entry_size;
+		}
+		sja1105_table_write_crc(table_start, p);
+		p += 4;
+	}
+	/* Final header:
+	 * Block ID does not matter
+	 * Length of 0 marks that header is final
+	 * CRC will be replaced on-the-fly on "config upload"
+	 */
+	header.block_id = 0;
+	header.len = 0;
+	header.crc = 0xDEADBEEF;
+	memset(p, 0, SIZE_TABLE_HEADER);
+	sja1105_table_header_packing(p, &header, PACK);
+}
+
+size_t
+sja1105_static_config_get_length(const struct sja1105_static_config *config)
+{
+	unsigned int sum;
+	unsigned int header_count;
+	enum sja1105_blk_idx i;
+
+	/* Ending header */
+	header_count = 1;
+	sum = SIZE_SJA1105_DEVICE_ID;
+
+	/* Tables (headers and entries) */
+	for (i = 0; i < BLK_IDX_MAX; i++) {
+		const struct sja1105_table *table;
+
+		table = &config->tables[i];
+		if (table->entry_count)
+			header_count++;
+
+		sum += table->ops->packed_entry_size * table->entry_count;
+	}
+	/* Headers have an additional CRC at the end */
+	sum += header_count * (SIZE_TABLE_HEADER + 4);
+	/* Last header does not have an extra CRC because there is no data */
+	sum -= 4;
+
+	return sum;
+}
+
+/* Compatibility matrices */
+
+/* SJA1105E: First generation, no TTEthernet */
+static struct sja1105_table_ops sja1105e_table_ops[BLK_IDX_MAX] = {
+	[BLK_IDX_L2_LOOKUP] = {
+		.packing = sja1105et_l2_lookup_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_lookup_entry),
+		.packed_entry_size = SIZE_L2_LOOKUP_ENTRY_ET,
+		.max_entry_count = MAX_L2_LOOKUP_COUNT,
+	},
+	[BLK_IDX_L2_POLICING] = {
+		.packing = sja1105_l2_policing_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_policing_entry),
+		.packed_entry_size = SIZE_L2_POLICING_ENTRY,
+		.max_entry_count = MAX_L2_POLICING_COUNT,
+	},
+	[BLK_IDX_VLAN_LOOKUP] = {
+		.packing = sja1105_vlan_lookup_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_vlan_lookup_entry),
+		.packed_entry_size = SIZE_VLAN_LOOKUP_ENTRY,
+		.max_entry_count = MAX_VLAN_LOOKUP_COUNT,
+	},
+	[BLK_IDX_L2_FORWARDING] = {
+		.packing = sja1105_l2_forwarding_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_forwarding_entry),
+		.packed_entry_size = SIZE_L2_FORWARDING_ENTRY,
+		.max_entry_count = MAX_L2_FORWARDING_COUNT,
+	},
+	[BLK_IDX_MAC_CONFIG] = {
+		.packing = sja1105et_mac_config_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_mac_config_entry),
+		.packed_entry_size = SIZE_MAC_CONFIG_ENTRY_ET,
+		.max_entry_count = MAX_MAC_CONFIG_COUNT,
+	},
+	[BLK_IDX_L2_LOOKUP_PARAMS] = {
+		.packing = sja1105et_l2_lookup_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_lookup_params_entry),
+		.packed_entry_size = SIZE_L2_LOOKUP_PARAMS_ENTRY_ET,
+		.max_entry_count = MAX_L2_LOOKUP_PARAMS_COUNT,
+	},
+	[BLK_IDX_L2_FORWARDING_PARAMS] = {
+		.packing = sja1105_l2_forwarding_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_forwarding_params_entry),
+		.packed_entry_size = SIZE_L2_FORWARDING_PARAMS_ENTRY,
+		.max_entry_count = MAX_L2_FORWARDING_PARAMS_COUNT,
+	},
+	[BLK_IDX_GENERAL_PARAMS] = {
+		.packing = sja1105et_general_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_general_params_entry),
+		.packed_entry_size = SIZE_GENERAL_PARAMS_ENTRY_ET,
+		.max_entry_count = MAX_GENERAL_PARAMS_COUNT,
+	},
+	[BLK_IDX_XMII_PARAMS] = {
+		.packing = sja1105_xmii_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_xmii_params_entry),
+		.packed_entry_size = SIZE_XMII_PARAMS_ENTRY,
+		.max_entry_count = MAX_XMII_PARAMS_COUNT,
+	},
+};
+
+/* SJA1105T: First generation, TTEthernet */
+static struct sja1105_table_ops sja1105t_table_ops[BLK_IDX_MAX] = {
+	[BLK_IDX_L2_LOOKUP] = {
+		.packing = sja1105et_l2_lookup_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_lookup_entry),
+		.packed_entry_size = SIZE_L2_LOOKUP_ENTRY_ET,
+		.max_entry_count = MAX_L2_LOOKUP_COUNT,
+	},
+	[BLK_IDX_L2_POLICING] = {
+		.packing = sja1105_l2_policing_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_policing_entry),
+		.packed_entry_size = SIZE_L2_POLICING_ENTRY,
+		.max_entry_count = MAX_L2_POLICING_COUNT,
+	},
+	[BLK_IDX_VLAN_LOOKUP] = {
+		.packing = sja1105_vlan_lookup_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_vlan_lookup_entry),
+		.packed_entry_size = SIZE_VLAN_LOOKUP_ENTRY,
+		.max_entry_count = MAX_VLAN_LOOKUP_COUNT,
+	},
+	[BLK_IDX_L2_FORWARDING] = {
+		.packing = sja1105_l2_forwarding_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_forwarding_entry),
+		.packed_entry_size = SIZE_L2_FORWARDING_ENTRY,
+		.max_entry_count = MAX_L2_FORWARDING_COUNT,
+	},
+	[BLK_IDX_MAC_CONFIG] = {
+		.packing = sja1105et_mac_config_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_mac_config_entry),
+		.packed_entry_size = SIZE_MAC_CONFIG_ENTRY_ET,
+		.max_entry_count = MAX_MAC_CONFIG_COUNT,
+	},
+	[BLK_IDX_L2_LOOKUP_PARAMS] = {
+		.packing = sja1105et_l2_lookup_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_lookup_params_entry),
+		.packed_entry_size = SIZE_L2_LOOKUP_PARAMS_ENTRY_ET,
+		.max_entry_count = MAX_L2_LOOKUP_PARAMS_COUNT,
+	},
+	[BLK_IDX_L2_FORWARDING_PARAMS] = {
+		.packing = sja1105_l2_forwarding_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_forwarding_params_entry),
+		.packed_entry_size = SIZE_L2_FORWARDING_PARAMS_ENTRY,
+		.max_entry_count = MAX_L2_FORWARDING_PARAMS_COUNT,
+	},
+	[BLK_IDX_GENERAL_PARAMS] = {
+		.packing = sja1105et_general_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_general_params_entry),
+		.packed_entry_size = SIZE_GENERAL_PARAMS_ENTRY_ET,
+		.max_entry_count = MAX_GENERAL_PARAMS_COUNT,
+	},
+	[BLK_IDX_XMII_PARAMS] = {
+		.packing = sja1105_xmii_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_xmii_params_entry),
+		.packed_entry_size = SIZE_XMII_PARAMS_ENTRY,
+		.max_entry_count = MAX_XMII_PARAMS_COUNT,
+	},
+};
+
+/* SJA1105P: Second generation, no TTEthernet, no SGMII */
+static struct sja1105_table_ops sja1105p_table_ops[BLK_IDX_MAX] = {
+	[BLK_IDX_L2_LOOKUP] = {
+		.packing = sja1105pqrs_l2_lookup_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_lookup_entry),
+		.packed_entry_size = SIZE_L2_LOOKUP_ENTRY_PQRS,
+		.max_entry_count = MAX_L2_LOOKUP_COUNT,
+	},
+	[BLK_IDX_L2_POLICING] = {
+		.packing = sja1105_l2_policing_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_policing_entry),
+		.packed_entry_size = SIZE_L2_POLICING_ENTRY,
+		.max_entry_count = MAX_L2_POLICING_COUNT,
+	},
+	[BLK_IDX_VLAN_LOOKUP] = {
+		.packing = sja1105_vlan_lookup_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_vlan_lookup_entry),
+		.packed_entry_size = SIZE_VLAN_LOOKUP_ENTRY,
+		.max_entry_count = MAX_VLAN_LOOKUP_COUNT,
+	},
+	[BLK_IDX_L2_FORWARDING] = {
+		.packing = sja1105_l2_forwarding_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_forwarding_entry),
+		.packed_entry_size = SIZE_L2_FORWARDING_ENTRY,
+		.max_entry_count = MAX_L2_FORWARDING_COUNT,
+	},
+	[BLK_IDX_MAC_CONFIG] = {
+		.packing = sja1105pqrs_mac_config_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_mac_config_entry),
+		.packed_entry_size = SIZE_MAC_CONFIG_ENTRY_PQRS,
+		.max_entry_count = MAX_MAC_CONFIG_COUNT,
+	},
+	[BLK_IDX_L2_LOOKUP_PARAMS] = {
+		.packing = sja1105pqrs_l2_lookup_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_lookup_params_entry),
+		.packed_entry_size = SIZE_L2_LOOKUP_PARAMS_ENTRY_PQRS,
+		.max_entry_count = MAX_L2_LOOKUP_PARAMS_COUNT,
+	},
+	[BLK_IDX_L2_FORWARDING_PARAMS] = {
+		.packing = sja1105_l2_forwarding_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_forwarding_params_entry),
+		.packed_entry_size = SIZE_L2_FORWARDING_PARAMS_ENTRY,
+		.max_entry_count = MAX_L2_FORWARDING_PARAMS_COUNT,
+	},
+	[BLK_IDX_GENERAL_PARAMS] = {
+		.packing = sja1105pqrs_general_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_general_params_entry),
+		.packed_entry_size = SIZE_GENERAL_PARAMS_ENTRY_PQRS,
+		.max_entry_count = MAX_GENERAL_PARAMS_COUNT,
+	},
+	[BLK_IDX_XMII_PARAMS] = {
+		.packing = sja1105_xmii_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_xmii_params_entry),
+		.packed_entry_size = SIZE_XMII_PARAMS_ENTRY,
+		.max_entry_count = MAX_XMII_PARAMS_COUNT,
+	},
+};
+
+/* SJA1105Q: Second generation, TTEthernet, no SGMII */
+static struct sja1105_table_ops sja1105q_table_ops[BLK_IDX_MAX] = {
+	[BLK_IDX_L2_LOOKUP] = {
+		.packing = sja1105pqrs_l2_lookup_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_lookup_entry),
+		.packed_entry_size = SIZE_L2_LOOKUP_ENTRY_PQRS,
+		.max_entry_count = MAX_L2_LOOKUP_COUNT,
+	},
+	[BLK_IDX_L2_POLICING] = {
+		.packing = sja1105_l2_policing_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_policing_entry),
+		.packed_entry_size = SIZE_L2_POLICING_ENTRY,
+		.max_entry_count = MAX_L2_POLICING_COUNT,
+	},
+	[BLK_IDX_VLAN_LOOKUP] = {
+		.packing = sja1105_vlan_lookup_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_vlan_lookup_entry),
+		.packed_entry_size = SIZE_VLAN_LOOKUP_ENTRY,
+		.max_entry_count = MAX_VLAN_LOOKUP_COUNT,
+	},
+	[BLK_IDX_L2_FORWARDING] = {
+		.packing = sja1105_l2_forwarding_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_forwarding_entry),
+		.packed_entry_size = SIZE_L2_FORWARDING_ENTRY,
+		.max_entry_count = MAX_L2_FORWARDING_COUNT,
+	},
+	[BLK_IDX_MAC_CONFIG] = {
+		.packing = sja1105pqrs_mac_config_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_mac_config_entry),
+		.packed_entry_size = SIZE_MAC_CONFIG_ENTRY_PQRS,
+		.max_entry_count = MAX_MAC_CONFIG_COUNT,
+	},
+	[BLK_IDX_L2_LOOKUP_PARAMS] = {
+		.packing = sja1105pqrs_l2_lookup_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_lookup_params_entry),
+		.packed_entry_size = SIZE_L2_LOOKUP_PARAMS_ENTRY_PQRS,
+		.max_entry_count = MAX_L2_LOOKUP_PARAMS_COUNT,
+	},
+	[BLK_IDX_L2_FORWARDING_PARAMS] = {
+		.packing = sja1105_l2_forwarding_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_forwarding_params_entry),
+		.packed_entry_size = SIZE_L2_FORWARDING_PARAMS_ENTRY,
+		.max_entry_count = MAX_L2_FORWARDING_PARAMS_COUNT,
+	},
+	[BLK_IDX_GENERAL_PARAMS] = {
+		.packing = sja1105pqrs_general_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_general_params_entry),
+		.packed_entry_size = SIZE_GENERAL_PARAMS_ENTRY_PQRS,
+		.max_entry_count = MAX_GENERAL_PARAMS_COUNT,
+	},
+	[BLK_IDX_XMII_PARAMS] = {
+		.packing = sja1105_xmii_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_xmii_params_entry),
+		.packed_entry_size = SIZE_XMII_PARAMS_ENTRY,
+		.max_entry_count = MAX_XMII_PARAMS_COUNT,
+	},
+};
+
+/* SJA1105R: Second generation, no TTEthernet, SGMII */
+static struct sja1105_table_ops sja1105r_table_ops[BLK_IDX_MAX] = {
+	[BLK_IDX_L2_LOOKUP] = {
+		.packing = sja1105pqrs_l2_lookup_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_lookup_entry),
+		.packed_entry_size = SIZE_L2_LOOKUP_ENTRY_PQRS,
+		.max_entry_count = MAX_L2_LOOKUP_COUNT,
+	},
+	[BLK_IDX_L2_POLICING] = {
+		.packing = sja1105_l2_policing_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_policing_entry),
+		.packed_entry_size = SIZE_L2_POLICING_ENTRY,
+		.max_entry_count = MAX_L2_POLICING_COUNT,
+	},
+	[BLK_IDX_VLAN_LOOKUP] = {
+		.packing = sja1105_vlan_lookup_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_vlan_lookup_entry),
+		.packed_entry_size = SIZE_VLAN_LOOKUP_ENTRY,
+		.max_entry_count = MAX_VLAN_LOOKUP_COUNT,
+	},
+	[BLK_IDX_L2_FORWARDING] = {
+		.packing = sja1105_l2_forwarding_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_forwarding_entry),
+		.packed_entry_size = SIZE_L2_FORWARDING_ENTRY,
+		.max_entry_count = MAX_L2_FORWARDING_COUNT,
+	},
+	[BLK_IDX_MAC_CONFIG] = {
+		.packing = sja1105pqrs_mac_config_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_mac_config_entry),
+		.packed_entry_size = SIZE_MAC_CONFIG_ENTRY_PQRS,
+		.max_entry_count = MAX_MAC_CONFIG_COUNT,
+	},
+	[BLK_IDX_L2_LOOKUP_PARAMS] = {
+		.packing = sja1105pqrs_l2_lookup_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_lookup_params_entry),
+		.packed_entry_size = SIZE_L2_LOOKUP_PARAMS_ENTRY_PQRS,
+		.max_entry_count = MAX_L2_LOOKUP_PARAMS_COUNT,
+	},
+	[BLK_IDX_L2_FORWARDING_PARAMS] = {
+		.packing = sja1105_l2_forwarding_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_forwarding_params_entry),
+		.packed_entry_size = SIZE_L2_FORWARDING_PARAMS_ENTRY,
+		.max_entry_count = MAX_L2_FORWARDING_PARAMS_COUNT,
+	},
+	[BLK_IDX_GENERAL_PARAMS] = {
+		.packing = sja1105pqrs_general_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_general_params_entry),
+		.packed_entry_size = SIZE_GENERAL_PARAMS_ENTRY_PQRS,
+		.max_entry_count = MAX_GENERAL_PARAMS_COUNT,
+	},
+	[BLK_IDX_XMII_PARAMS] = {
+		.packing = sja1105_xmii_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_xmii_params_entry),
+		.packed_entry_size = SIZE_XMII_PARAMS_ENTRY,
+		.max_entry_count = MAX_XMII_PARAMS_COUNT,
+	},
+};
+
+/* SJA1105S: Second generation, TTEthernet, SGMII */
+static struct sja1105_table_ops sja1105s_table_ops[BLK_IDX_MAX] = {
+	[BLK_IDX_L2_LOOKUP] = {
+		.packing = sja1105pqrs_l2_lookup_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_lookup_entry),
+		.packed_entry_size = SIZE_L2_LOOKUP_ENTRY_PQRS,
+		.max_entry_count = MAX_L2_LOOKUP_COUNT,
+	},
+	[BLK_IDX_L2_POLICING] = {
+		.packing = sja1105_l2_policing_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_policing_entry),
+		.packed_entry_size = SIZE_L2_POLICING_ENTRY,
+		.max_entry_count = MAX_L2_POLICING_COUNT,
+	},
+	[BLK_IDX_VLAN_LOOKUP] = {
+		.packing = sja1105_vlan_lookup_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_vlan_lookup_entry),
+		.packed_entry_size = SIZE_VLAN_LOOKUP_ENTRY,
+		.max_entry_count = MAX_VLAN_LOOKUP_COUNT,
+	},
+	[BLK_IDX_L2_FORWARDING] = {
+		.packing = sja1105_l2_forwarding_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_forwarding_entry),
+		.packed_entry_size = SIZE_L2_FORWARDING_ENTRY,
+		.max_entry_count = MAX_L2_FORWARDING_COUNT,
+	},
+	[BLK_IDX_MAC_CONFIG] = {
+		.packing = sja1105pqrs_mac_config_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_mac_config_entry),
+		.packed_entry_size = SIZE_MAC_CONFIG_ENTRY_PQRS,
+		.max_entry_count = MAX_MAC_CONFIG_COUNT,
+	},
+	[BLK_IDX_L2_LOOKUP_PARAMS] = {
+		.packing = sja1105pqrs_l2_lookup_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_lookup_params_entry),
+		.packed_entry_size = SIZE_L2_LOOKUP_PARAMS_ENTRY_PQRS,
+		.max_entry_count = MAX_L2_LOOKUP_PARAMS_COUNT,
+	},
+	[BLK_IDX_L2_FORWARDING_PARAMS] = {
+		.packing = sja1105_l2_forwarding_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_l2_forwarding_params_entry),
+		.packed_entry_size = SIZE_L2_FORWARDING_PARAMS_ENTRY,
+		.max_entry_count = MAX_L2_FORWARDING_PARAMS_COUNT,
+	},
+	[BLK_IDX_GENERAL_PARAMS] = {
+		.packing = sja1105pqrs_general_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_general_params_entry),
+		.packed_entry_size = SIZE_GENERAL_PARAMS_ENTRY_PQRS,
+		.max_entry_count = MAX_GENERAL_PARAMS_COUNT,
+	},
+	[BLK_IDX_XMII_PARAMS] = {
+		.packing = sja1105_xmii_params_entry_packing,
+		.unpacked_entry_size = sizeof(struct sja1105_xmii_params_entry),
+		.packed_entry_size = SIZE_XMII_PARAMS_ENTRY,
+		.max_entry_count = MAX_XMII_PARAMS_COUNT,
+	},
+};
+
+int sja1105_static_config_init(struct sja1105_static_config *config,
+			       u64 device_id, u64 part_nr)
+{
+	const struct sja1105_table_ops *ops;
+	enum sja1105_blk_idx i;
+
+	memset(config, 0, sizeof(*config));
+
+	if (device_id == SJA1105E_DEVICE_ID)
+		ops = sja1105e_table_ops;
+	else if (device_id == SJA1105T_DEVICE_ID)
+		ops = sja1105t_table_ops;
+	else if (IS_P(device_id, part_nr))
+		ops = sja1105p_table_ops;
+	else if (IS_Q(device_id, part_nr))
+		ops = sja1105q_table_ops;
+	else if (IS_R(device_id, part_nr))
+		ops = sja1105r_table_ops;
+	else if (IS_S(device_id, part_nr))
+		ops = sja1105s_table_ops;
+	else
+		return -EINVAL;
+
+	for (i = 0; i < BLK_IDX_MAX; i++)
+		config->tables[i].ops = &ops[i];
+
+	config->device_id = device_id;
+	return 0;
+}
+
+void sja1105_static_config_free(struct sja1105_static_config *config)
+{
+	enum sja1105_blk_idx i;
+
+	for (i = 0; i < BLK_IDX_MAX; i++) {
+		if (config->tables[i].entry_count) {
+			kfree(config->tables[i].entries);
+			config->tables[i].entry_count = 0;
+		}
+	}
+}
+
diff --git a/drivers/net/dsa/sja1105/sja1105_static_config.h b/drivers/net/dsa/sja1105/sja1105_static_config.h
new file mode 100644
index 000000000000..1c331c9e20b3
--- /dev/null
+++ b/drivers/net/dsa/sja1105/sja1105_static_config.h
@@ -0,0 +1,308 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright (c) 2016-2018, NXP Semiconductors
+ * Copyright (c) 2018-2019, Vladimir Oltean <olteanv@gmail.com>
+ */
+#ifndef _SJA1105_STATIC_CONFIG_H
+#define _SJA1105_STATIC_CONFIG_H
+
+#include <linux/packing.h>
+
+#define SIZE_SJA1105_DEVICE_ID                  4
+#define SIZE_TABLE_HEADER                       12
+#define SIZE_L2_LOOKUP_ENTRY_ET                 12
+#define SIZE_L2_LOOKUP_ENTRY_PQRS               20
+#define SIZE_L2_POLICING_ENTRY                  8
+#define SIZE_VLAN_LOOKUP_ENTRY                  8
+#define SIZE_L2_FORWARDING_ENTRY                8
+#define SIZE_MAC_CONFIG_ENTRY_ET                28
+#define SIZE_MAC_CONFIG_ENTRY_PQRS              32
+#define SIZE_L2_LOOKUP_PARAMS_ENTRY_ET          4
+#define SIZE_L2_LOOKUP_PARAMS_ENTRY_PQRS        16
+#define SIZE_L2_FORWARDING_PARAMS_ENTRY         12
+#define SIZE_GENERAL_PARAMS_ENTRY_ET            40
+#define SIZE_GENERAL_PARAMS_ENTRY_PQRS          44
+#define SIZE_XMII_PARAMS_ENTRY                  4
+
+/* UM10944.pdf Page 11, Table 2. Configuration Blocks */
+#define BLKID_L2_LOOKUP                    0x05
+#define BLKID_L2_POLICING                  0x06
+#define BLKID_VLAN_LOOKUP                  0x07
+#define BLKID_L2_FORWARDING                0x08
+#define BLKID_MAC_CONFIG                   0x09
+#define BLKID_L2_LOOKUP_PARAMS             0x0D
+#define BLKID_L2_FORWARDING_PARAMS         0x0E
+#define BLKID_GENERAL_PARAMS               0x11
+#define BLKID_XMII_PARAMS                  0x4E
+#define BLKID_MAX                          BLKID_XMII_PARAMS
+
+enum sja1105_blk_idx {
+	BLK_IDX_L2_LOOKUP = 0,
+	BLK_IDX_L2_POLICING,
+	BLK_IDX_VLAN_LOOKUP,
+	BLK_IDX_L2_FORWARDING,
+	BLK_IDX_MAC_CONFIG,
+	BLK_IDX_L2_LOOKUP_PARAMS,
+	BLK_IDX_L2_FORWARDING_PARAMS,
+	BLK_IDX_GENERAL_PARAMS,
+	BLK_IDX_XMII_PARAMS,
+	BLK_IDX_MAX,
+	/* Fake block indices that are only valid for dynamic access */
+	BLK_IDX_MGMT_ROUTE,
+	BLK_IDX_MAX_DYN,
+	BLK_IDX_INVAL = -1,
+};
+
+#define MAX_L2_LOOKUP_COUNT                      1024
+#define MAX_L2_POLICING_COUNT                    45
+#define MAX_VLAN_LOOKUP_COUNT                    4096
+#define MAX_L2_FORWARDING_COUNT                  13
+#define MAX_MAC_CONFIG_COUNT                     5
+#define MAX_L2_LOOKUP_PARAMS_COUNT               1
+#define MAX_L2_FORWARDING_PARAMS_COUNT           1
+#define MAX_GENERAL_PARAMS_COUNT                 1
+#define MAX_XMII_PARAMS_COUNT                    1
+
+#define MAX_FRAME_MEMORY                         929
+#define MAX_FRAME_MEMORY_RETAGGING               910
+
+#define SJA1105E_DEVICE_ID         0x9C00000Cull
+#define SJA1105T_DEVICE_ID         0x9E00030Eull
+#define SJA1105PR_DEVICE_ID        0xAF00030Eull
+#define SJA1105QS_DEVICE_ID        0xAE00030Eull
+#define SJA1105_NO_DEVICE_ID       0x00000000ull
+
+#define SJA1105P_PART_NR           0x9A84
+#define SJA1105Q_PART_NR           0x9A85
+#define SJA1105R_PART_NR           0x9A86
+#define SJA1105S_PART_NR           0x9A87
+#define SJA1105_PART_NR_DONT_CARE  0xFFFF
+
+#define IS_PQRS(device_id) \
+	(((device_id) == SJA1105PR_DEVICE_ID) || \
+	 ((device_id) == SJA1105QS_DEVICE_ID))
+#define IS_ET(device_id) \
+	(((device_id) == SJA1105E_DEVICE_ID) || \
+	 ((device_id) == SJA1105T_DEVICE_ID))
+/* P and R have same Device ID, and differ by Part Number */
+#define IS_P(device_id, part_nr) \
+	(((device_id) == SJA1105PR_DEVICE_ID) && \
+	 ((part_nr) == SJA1105P_PART_NR))
+#define IS_R(device_id, part_nr) \
+	(((device_id) == SJA1105PR_DEVICE_ID) && \
+	 ((part_nr) == SJA1105R_PART_NR))
+/* Same do Q and S */
+#define IS_Q(device_id, part_nr) \
+	(((device_id) == SJA1105QS_DEVICE_ID) && \
+	 ((part_nr) == SJA1105Q_PART_NR))
+#define IS_S(device_id, part_nr) \
+	(((device_id) == SJA1105QS_DEVICE_ID) && \
+	 ((part_nr) == SJA1105S_PART_NR))
+#define DEVICE_ID_VALID(device_id) \
+	(IS_ET(device_id) || IS_PQRS(device_id))
+
+#ifdef __KERNEL__
+#include <asm/types.h>
+#include <linux/types.h>
+#endif
+
+struct sja1105_general_params_entry {
+	u64 vllupformat;
+	u64 mirr_ptacu;
+	u64 switchid;
+	u64 hostprio;
+	u64 mac_fltres1;
+	u64 mac_fltres0;
+	u64 mac_flt1;
+	u64 mac_flt0;
+	u64 incl_srcpt1;
+	u64 incl_srcpt0;
+	u64 send_meta1;
+	u64 send_meta0;
+	u64 casc_port;
+	u64 host_port;
+	u64 mirr_port;
+	u64 vlmarker;
+	u64 vlmask;
+	u64 tpid;
+	u64 ignore2stf;
+	u64 tpid2;
+	/* P/Q/R/S only */
+	u64 queue_ts;
+	u64 egrmirrvid;
+	u64 egrmirrpcp;
+	u64 egrmirrdei;
+	u64 replay_port;
+};
+
+struct sja1105_vlan_lookup_entry {
+	u64 ving_mirr;
+	u64 vegr_mirr;
+	u64 vmemb_port;
+	u64 vlan_bc;
+	u64 tag_port;
+	u64 vlanid;
+};
+
+struct sja1105_l2_lookup_entry {
+	u64 mirrvlan;      /* P/Q/R/S only - LOCKEDS=1 */
+	u64 mirr;          /* P/Q/R/S only - LOCKEDS=1 */
+	u64 retag;         /* P/Q/R/S only - LOCKEDS=1 */
+	u64 mask_iotag;    /* P/Q/R/S only */
+	u64 mask_vlanid;   /* P/Q/R/S only */
+	u64 mask_macaddr;  /* P/Q/R/S only */
+	u64 iotag;         /* P/Q/R/S only */
+	u64 vlanid;
+	u64 macaddr;
+	u64 destports;
+	u64 enfport;
+	u64 index;
+};
+
+struct sja1105_l2_lookup_params_entry {
+	u64 drpbc;           /* P/Q/R/S only */
+	u64 drpmc;           /* P/Q/R/S only */
+	u64 drpuni;          /* P/Q/R/S only */
+	u64 maxaddrp[5];     /* P/Q/R/S only */
+	u64 start_dynspc;    /* P/Q/R/S only */
+	u64 drpnolearn;      /* P/Q/R/S only */
+	u64 use_static;      /* P/Q/R/S only */
+	u64 owr_dyn;         /* P/Q/R/S only */
+	u64 learn_once;      /* P/Q/R/S only */
+	u64 maxage;          /* Shared */
+	u64 dyn_tbsz;        /* E/T only */
+	u64 poly;            /* E/T only */
+	u64 shared_learn;    /* Shared */
+	u64 no_enf_hostprt;  /* Shared */
+	u64 no_mgmt_learn;   /* Shared */
+};
+
+struct sja1105_l2_forwarding_entry {
+	u64 bc_domain;
+	u64 reach_port;
+	u64 fl_domain;
+	u64 vlan_pmap[8];
+};
+
+struct sja1105_l2_forwarding_params_entry {
+	u64 max_dynp;
+	u64 part_spc[8];
+};
+
+struct sja1105_l2_policing_entry {
+	u64 sharindx;
+	u64 smax;
+	u64 rate;
+	u64 maxlen;
+	u64 partition;
+};
+
+struct sja1105_mac_config_entry {
+	u64 top[8];
+	u64 base[8];
+	u64 enabled[8];
+	u64 ifg;
+	u64 speed;
+	u64 tp_delin;
+	u64 tp_delout;
+	u64 maxage;
+	u64 vlanprio;
+	u64 vlanid;
+	u64 ing_mirr;
+	u64 egr_mirr;
+	u64 drpnona664;
+	u64 drpdtag;
+	u64 drpsotag;   /* only on P/Q/R/S */
+	u64 drpsitag;   /* only on P/Q/R/S */
+	u64 drpuntag;
+	u64 retag;
+	u64 dyn_learn;
+	u64 egress;
+	u64 ingress;
+	u64 mirrcie;    /* only on P/Q/R/S */
+	u64 mirrcetag;  /* only on P/Q/R/S */
+	u64 ingmirrvid; /* only on P/Q/R/S */
+	u64 ingmirrpcp; /* only on P/Q/R/S */
+	u64 ingmirrdei; /* only on P/Q/R/S */
+};
+
+struct sja1105_xmii_params_entry {
+	u64 phy_mac[5];
+	u64 xmii_mode[5];
+};
+
+struct sja1105_avb_params_entry {
+	u64 l2cbs; /* only on P/Q/R/S */
+	u64 cas_master; /* only on P/Q/R/S */
+	u64 destmeta;
+	u64 srcmeta;
+};
+
+struct sja1105_table_header {
+	u64 block_id;
+	u64 len;
+	u64 crc;
+};
+
+struct sja1105_table_ops {
+	size_t (*packing)(void *buf, void *entry_ptr, enum packing_op op);
+	size_t unpacked_entry_size;
+	size_t packed_entry_size;
+	size_t max_entry_count;
+};
+
+struct sja1105_table {
+	const struct sja1105_table_ops *ops;
+	size_t entry_count;
+	void *entries;
+};
+
+struct sja1105_static_config {
+	u64 device_id;
+	struct sja1105_table tables[BLK_IDX_MAX];
+};
+
+size_t sja1105_table_header_packing(void *buf, void *hdr, enum packing_op op);
+void
+sja1105_table_header_pack_with_crc(void *buf, struct sja1105_table_header *hdr);
+size_t
+sja1105_static_config_get_length(const struct sja1105_static_config *config);
+
+enum sja1105_static_config_validity {
+	SJA1105_CONFIG_OK = 0,
+	SJA1105_DEVICE_ID_INVALID,
+	SJA1105_MISSING_L2_POLICING_TABLE,
+	SJA1105_MISSING_L2_FORWARDING_TABLE,
+	SJA1105_MISSING_L2_FORWARDING_PARAMS_TABLE,
+	SJA1105_MISSING_GENERAL_PARAMS_TABLE,
+	SJA1105_MISSING_VLAN_TABLE,
+	SJA1105_MISSING_XMII_TABLE,
+	SJA1105_MISSING_MAC_TABLE,
+	SJA1105_OVERCOMMITTED_FRAME_MEMORY,
+	SJA1105_UNEXPECTED_END_OF_BUFFER,
+	SJA1105_INVALID_DEVICE_ID,
+	SJA1105_INVALID_TABLE_HEADER_CRC,
+	SJA1105_INVALID_TABLE_HEADER,
+	SJA1105_INCORRECT_TABLE_LENGTH,
+	SJA1105_DATA_CRC_INVALID,
+	SJA1105_EXTRA_BYTES_AT_END_OF_BUFFER,
+};
+
+extern const char *sja1105_static_config_error_msg[];
+
+enum sja1105_static_config_validity
+sja1105_static_config_check_valid(const struct sja1105_static_config *config);
+void
+sja1105_static_config_pack(void *buf, struct sja1105_static_config *config);
+int sja1105_static_config_init(struct sja1105_static_config *config,
+			       u64 device_id, u64 part_nr);
+void sja1105_static_config_free(struct sja1105_static_config *config);
+
+u32 sja1105_crc32(const void *buf, size_t len);
+
+void sja1105_pack(void *buf, const u64 *val, int start, int end, size_t len);
+void sja1105_unpack(const void *buf, u64 *val, int start, int end, size_t len);
+void sja1105_packing(void *buf, u64 *val, int start, int end,
+		     size_t len, enum packing_op op);
+
+#endif
+
-- 
2.17.1

Re: [PATCH net-next 10/17] net: dsa: Introduce driver for NXP SJA1105 5-port L2 switch

[Andrew Lunn]

> +static int sja1105_init_mii_settings(struct sja1105_private *priv,
> +				     struct sja1105_dt_port *ports)
> +{
> +	struct device *dev = &priv->spidev->dev;
> +	struct sja1105_xmii_params_entry *mii;
> +	struct sja1105_table *table;
> +	int i;
> +
> +	table = &priv->static_config.tables[BLK_IDX_XMII_PARAMS];
> +
> +	/* Discard previous xMII Mode Parameters Table */
> +	if (table->entry_count) {
> +		kfree(table->entries);
> +		table->entry_count = 0;
> +	}
> +
> +	table->entries = kcalloc(MAX_XMII_PARAMS_COUNT,
> +				 table->ops->unpacked_entry_size, GFP_KERNEL);
> +	if (!table->entries)
> +		return -ENOMEM;
> +
> +	/* Override table based on phylib DT bindings */
> +	table->entry_count = MAX_XMII_PARAMS_COUNT;
> +
> +	mii = table->entries;
> +
> +	for (i = 0; i < SJA1105_NUM_PORTS; i++) {
> +		switch (ports[i].phy_mode) {
> +		case PHY_INTERFACE_MODE_MII:
> +			mii->xmii_mode[i] = XMII_MODE_MII;
> +			break;
> +		case PHY_INTERFACE_MODE_RMII:
> +			mii->xmii_mode[i] = XMII_MODE_RMII;
> +			break;
> +		case PHY_INTERFACE_MODE_RGMII:
> +		case PHY_INTERFACE_MODE_RGMII_ID:
> +		case PHY_INTERFACE_MODE_RGMII_RXID:
> +		case PHY_INTERFACE_MODE_RGMII_TXID:
> +			mii->xmii_mode[i] = XMII_MODE_RGMII;
> +			break;

https://www.nxp.com/docs/en/data-sheet/SJA1105.pdf

Section 6.2.3 RGMII signaling and encoding

Note that RGMII requires an external delay of between 1.5 ns and 2 ns
on TXC and RXC.

So it sounds like the switch only supports PHY_INTERFACE_MODE_RGMII.

If the port is in MAC mode, you should pass this phy-mode to the PHY
when you connect to it. The PHY can then add the delay if needed.

If however, the port is in PHY mode, and it is asked to do RGMII other
than PHY_INTERFACE_MODE_RGMII, you should report an error. It cannot
do it.

   Andrew

[PATCH net-next 11/17] net: dsa: sja1105: Add support for FDB and MDB management

[Vladimir Oltean]

Currently only the (more difficult) first generation E/T series is
supported. Here the TCAM is only 4-way associative, and to know where
the hardware will search for a FDB entry, we need to perform the same
hash algorithm in order to install the entry in the correct bin.

On P/Q/R/S, the TCAM should be fully associative. However the SPI
command interface is different, and because I don't have access to a
new-generation device at the moment, support for it is TODO.

Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
---
 drivers/net/dsa/sja1105/sja1105.h             |   2 +
 .../net/dsa/sja1105/sja1105_dynamic_config.c  |  40 ++++
 drivers/net/dsa/sja1105/sja1105_main.c        | 193 ++++++++++++++++++
 3 files changed, 235 insertions(+)

diff --git a/drivers/net/dsa/sja1105/sja1105.h b/drivers/net/dsa/sja1105/sja1105.h
index 2874423a5cd0..26d93afa77a1 100644
--- a/drivers/net/dsa/sja1105/sja1105.h
+++ b/drivers/net/dsa/sja1105/sja1105.h
@@ -118,6 +118,8 @@ int sja1105_dynamic_config_write(struct sja1105_private *priv,
 				 int index, void *entry, bool keep);
 void sja1105_dynamic_config_init(struct sja1105_private *priv);
 
+u8 sja1105_fdb_hash(struct sja1105_private *priv, const u8 *addr, u16 vid);
+
 /* Common implementations for the static and dynamic configs */
 size_t sja1105_l2_forwarding_entry_packing(void *buf, void *entry_ptr,
 					   enum packing_op op);
diff --git a/drivers/net/dsa/sja1105/sja1105_dynamic_config.c b/drivers/net/dsa/sja1105/sja1105_dynamic_config.c
index 4f3756e267ac..545988779fe5 100644
--- a/drivers/net/dsa/sja1105/sja1105_dynamic_config.c
+++ b/drivers/net/dsa/sja1105/sja1105_dynamic_config.c
@@ -476,3 +476,43 @@ void sja1105_dynamic_config_init(struct sja1105_private *priv)
 
 	priv->dyn_ops = ops;
 }
+
+static u8 crc8_add(u8 crc, u8 byte, u8 poly)
+{
+	int i;
+
+	for (i = 0; i < 8; i++) {
+		if ((crc ^ byte) & (1 << 7)) {
+			crc <<= 1;
+			crc ^= poly;
+		} else {
+			crc <<= 1;
+		}
+		byte <<= 1;
+	}
+	return crc;
+}
+
+/* CRC8 algorithm with non-reversed input, non-reversed output,
+ * no input xor and no output xor. Code customized for receiving
+ * the SJA1105 E/T FDB keys (vlanid, macaddr) as input. CRC polynomial
+ * is also received as argument in the Koopman notation that the switch
+ * hardware stores it in.
+ */
+u8 sja1105_fdb_hash(struct sja1105_private *priv, const u8 *addr, u16 vid)
+{
+	struct sja1105_l2_lookup_params_entry *l2_lookup_params =
+		priv->static_config.tables[BLK_IDX_L2_LOOKUP_PARAMS].entries;
+	u64 poly_koopman = l2_lookup_params->poly;
+	/* Convert polynomial from Koopman to 'normal' notation */
+	u8 poly = (u8)(1 + (poly_koopman << 1));
+	u64 vlanid = l2_lookup_params->shared_learn ? 0 : vid;
+	u64 input = (vlanid << 48) | ether_addr_to_u64(addr);
+	u8 crc = 0; /* seed */
+	int i;
+
+	/* Mask the eight bytes starting from MSB one at a time */
+	for (i = 56; i >= 0; i -= 8)
+		crc = crc8_add(crc, (input & (0xffull << i)) >> i, poly);
+	return crc;
+}
diff --git a/drivers/net/dsa/sja1105/sja1105_main.c b/drivers/net/dsa/sja1105/sja1105_main.c
index 45c233d7fbba..93d1f607eeab 100644
--- a/drivers/net/dsa/sja1105/sja1105_main.c
+++ b/drivers/net/dsa/sja1105/sja1105_main.c
@@ -187,6 +187,9 @@ static int sja1105_init_static_fdb(struct sja1105_private *priv)
 
 	table = &priv->static_config.tables[BLK_IDX_L2_LOOKUP];
 
+	/* We only populate the FDB table through dynamic
+	 * L2 Address Lookup entries
+	 */
 	if (table->entry_count) {
 		kfree(table->entries);
 		table->entry_count = 0;
@@ -699,6 +702,190 @@ static void sja1105_adjust_link(struct dsa_switch *ds, int port,
 		sja1105_adjust_port_config(priv, port, phydev->speed, true);
 }
 
+#define fdb(bin, index) \
+	((bin) * SJA1105ET_FDB_BIN_SIZE + (index))
+#define is_bin_index_valid(i) \
+	((i) >= 0 && (i) < SJA1105ET_FDB_BIN_SIZE)
+
+static int
+sja1105_is_fdb_entry_in_bin(struct sja1105_private *priv, int bin,
+			    const u8 *addr, u16 vid,
+			    struct sja1105_l2_lookup_entry *fdb_match,
+			    int *last_unused)
+{
+	int index_in_bin;
+
+	for (index_in_bin = 0; index_in_bin < SJA1105ET_FDB_BIN_SIZE;
+	     index_in_bin++) {
+		struct sja1105_l2_lookup_entry l2_lookup = { 0 };
+
+		/* Skip unused entries, optionally marking them
+		 * into the return value
+		 */
+		if (sja1105_dynamic_config_read(priv, BLK_IDX_L2_LOOKUP,
+						fdb(bin, index_in_bin),
+						&l2_lookup)) {
+			if (last_unused)
+				*last_unused = index_in_bin;
+			continue;
+		}
+
+		if (l2_lookup.macaddr == ether_addr_to_u64(addr) &&
+		    l2_lookup.vlanid == vid) {
+			if (fdb_match)
+				*fdb_match = l2_lookup;
+			return index_in_bin;
+		}
+	}
+	/* Return an invalid entry index if not found */
+	return SJA1105ET_FDB_BIN_SIZE;
+}
+
+static int sja1105_fdb_add(struct dsa_switch *ds, int port,
+			   const unsigned char *addr, u16 vid)
+{
+	struct sja1105_l2_lookup_entry l2_lookup = { 0 };
+	struct sja1105_private *priv = ds->priv;
+	struct device *dev = ds->dev;
+	int bin, index_in_bin;
+	int last_unused;
+
+	bin = sja1105_fdb_hash(priv, addr, vid);
+
+	index_in_bin = sja1105_is_fdb_entry_in_bin(priv, bin, addr, vid,
+						   &l2_lookup, &last_unused);
+	if (is_bin_index_valid(index_in_bin)) {
+		/* We have an FDB entry. Is our port in the destination
+		 * mask? If yes, we need to do nothing. If not, we need
+		 * to rewrite the entry by adding this port to it.
+		 */
+		if (l2_lookup.destports & BIT(port))
+			return 0;
+		l2_lookup.destports |= BIT(port);
+	} else {
+		/* We don't have an FDB entry. We construct a new one and
+		 * try to find a place for it within the FDB table.
+		 */
+		l2_lookup.macaddr = ether_addr_to_u64(addr);
+		l2_lookup.destports = BIT(port);
+		l2_lookup.vlanid = vid;
+
+		if (is_bin_index_valid(last_unused)) {
+			index_in_bin = last_unused;
+		} else {
+			/* Bin is full, need to evict somebody.
+			 * Choose victim at random. If you get these messages
+			 * often, you may need to consider changing the
+			 * distribution function:
+			 * static_config[BLK_IDX_L2_LOOKUP_PARAMS].entries->poly
+			 */
+			get_random_bytes(&index_in_bin, sizeof(u8));
+			index_in_bin %= SJA1105ET_FDB_BIN_SIZE;
+			dev_warn(dev, "Warning, FDB bin %d full while adding entry for %pM. Evicting entry %u.\n",
+				 bin, addr, index_in_bin);
+			/* Evict entry */
+			sja1105_dynamic_config_write(priv, BLK_IDX_L2_LOOKUP,
+						     fdb(bin, index_in_bin),
+						     NULL, false);
+		}
+	}
+	l2_lookup.index = fdb(bin, index_in_bin);
+
+	return sja1105_dynamic_config_write(priv, BLK_IDX_L2_LOOKUP,
+				l2_lookup.index, &l2_lookup, true);
+}
+
+static int sja1105_fdb_del(struct dsa_switch *ds, int port,
+			   const unsigned char *addr, u16 vid)
+{
+	struct sja1105_l2_lookup_entry l2_lookup = { 0 };
+	struct sja1105_private *priv = ds->priv;
+	u8 bin, index_in_bin;
+	bool keep;
+
+	bin = sja1105_fdb_hash(priv, addr, vid);
+
+	index_in_bin = sja1105_is_fdb_entry_in_bin(priv, bin, addr, vid,
+						   &l2_lookup, NULL);
+	if (!is_bin_index_valid(index_in_bin))
+		return 0;
+
+	/* We have an FDB entry. Is our port in the destination mask? If yes,
+	 * we need to remove it. If the resulting port mask becomes empty, we
+	 * need to completely evict the FDB entry.
+	 * Otherwise we just write it back.
+	 */
+	if (l2_lookup.destports & BIT(port))
+		l2_lookup.destports &= ~BIT(port);
+	if (l2_lookup.destports)
+		keep = true;
+	else
+		keep = false;
+
+	return sja1105_dynamic_config_write(priv, BLK_IDX_L2_LOOKUP,
+					    fdb(bin, index_in_bin),
+					    &l2_lookup, keep);
+}
+
+static int sja1105_fdb_dump(struct dsa_switch *ds, int port,
+			    dsa_fdb_dump_cb_t *cb, void *data)
+{
+	struct sja1105_private *priv = ds->priv;
+	struct device *dev = ds->dev;
+	int i;
+
+	for (i = 0; i < MAX_L2_LOOKUP_COUNT; i++) {
+		struct sja1105_l2_lookup_entry l2_lookup;
+		u8 macaddr[ETH_ALEN];
+		int rc;
+
+		memset(&l2_lookup, 0, sizeof(l2_lookup));
+		rc = sja1105_dynamic_config_read(priv, BLK_IDX_L2_LOOKUP,
+						 i, &l2_lookup);
+		/* No fdb entry at i, not an issue */
+		if (rc == -EINVAL)
+			continue;
+		if (rc) {
+			dev_err(dev, "Failed to dump FDB: %d\n", rc);
+			return rc;
+		}
+
+		/* FDB dump callback is per port. This means we have to
+		 * disregard a valid entry if it's not for this port, even if
+		 * only to revisit it later. This is inefficient because the
+		 * 1024-sized FDB table needs to be traversed 4 times through
+		 * SPI during a 'bridge fdb show' command.
+		 */
+		if (!(l2_lookup.destports & BIT(port)))
+			continue;
+		u64_to_ether_addr(l2_lookup.macaddr, macaddr);
+		cb(macaddr, l2_lookup.vlanid, false, data);
+	}
+	return 0;
+}
+
+#undef fdb
+#undef is_bin_index_valid
+
+/* This callback needs to be present */
+static int sja1105_mdb_prepare(struct dsa_switch *ds, int port,
+			       const struct switchdev_obj_port_mdb *mdb)
+{
+	return 0;
+}
+
+static void sja1105_mdb_add(struct dsa_switch *ds, int port,
+			    const struct switchdev_obj_port_mdb *mdb)
+{
+	sja1105_fdb_add(ds, port, mdb->addr, mdb->vid);
+}
+
+static int sja1105_mdb_del(struct dsa_switch *ds, int port,
+			   const struct switchdev_obj_port_mdb *mdb)
+{
+	return sja1105_fdb_del(ds, port, mdb->addr, mdb->vid);
+}
+
 static int sja1105_bridge_member(struct dsa_switch *ds, int port,
 				 struct net_device *br, bool member)
 {
@@ -792,8 +979,14 @@ static const struct dsa_switch_ops sja1105_switch_ops = {
 	.get_tag_protocol	= sja1105_get_tag_protocol,
 	.setup			= sja1105_setup,
 	.adjust_link		= sja1105_adjust_link,
+	.port_fdb_dump		= sja1105_fdb_dump,
+	.port_fdb_add		= sja1105_fdb_add,
+	.port_fdb_del		= sja1105_fdb_del,
 	.port_bridge_join	= sja1105_bridge_join,
 	.port_bridge_leave	= sja1105_bridge_leave,
+	.port_mdb_prepare	= sja1105_mdb_prepare,
+	.port_mdb_add		= sja1105_mdb_add,
+	.port_mdb_del		= sja1105_mdb_del,
 };
 
 static int sja1105_probe(struct spi_device *spi)
-- 
2.17.1

[PATCH net-next 12/17] net: dsa: sja1105: Add support for VLAN operations

[Vladimir Oltean]

VLAN filtering cannot be properly disabled in SJA1105. So in order to
emulate the "no VLAN awareness" behavior (not dropping traffic that is
tagged with a VID that isn't configured on the port), we need to hack
another switch feature: programmable TPID (which is 0x8100 for 802.1Q).
We are reprogramming the TPID to a bogus value (ETH_P_EDSA) which leaves
the switch thinking that all traffic is untagged, and therefore accepts
it.

Under a vlan_filtering bridge, the proper TPID of ETH_P_8021Q is
installed again, and the switch starts identifying 802.1Q-tagged
traffic.

Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
---
 drivers/net/dsa/sja1105/sja1105_main.c        | 260 +++++++++++++++++-
 .../net/dsa/sja1105/sja1105_static_config.c   |  38 +++
 .../net/dsa/sja1105/sja1105_static_config.h   |   3 +
 3 files changed, 299 insertions(+), 2 deletions(-)

diff --git a/drivers/net/dsa/sja1105/sja1105_main.c b/drivers/net/dsa/sja1105/sja1105_main.c
index 93d1f607eeab..028ed1ebe358 100644
--- a/drivers/net/dsa/sja1105/sja1105_main.c
+++ b/drivers/net/dsa/sja1105/sja1105_main.c
@@ -260,6 +260,13 @@ static int sja1105_init_static_vlan(struct sja1105_private *priv)
 	table = &priv->static_config.tables[BLK_IDX_VLAN_LOOKUP];
 
 	/* The static VLAN table will only contain the initial pvid of 0.
+	 * All other VLANs are to be configured through dynamic entries,
+	 * and kept in the static configuration table as backing memory.
+	 * The pvid of 0 is sufficient to pass traffic while the ports are
+	 * standalone and when vlan_filtering is disabled. When filtering
+	 * gets enabled, the switchdev core sets up the VLAN ID 1 and sets
+	 * it as the new pvid. Actually 'pvid 1' still comes up in 'bridge
+	 * vlan' even when vlan_filtering is off, but it has no effect.
 	 */
 	if (table->entry_count) {
 		kfree(table->entries);
@@ -400,8 +407,11 @@ static int sja1105_init_general_params(struct sja1105_private *priv)
 		.vlmask = 0,
 		/* Only update correctionField for 1-step PTP (L2 transport) */
 		.ignore2stf = 0,
-		.tpid = ETH_P_8021Q,
-		.tpid2 = ETH_P_8021Q,
+		/* Forcefully disable VLAN filtering by telling
+		 * the switch that VLAN has a different EtherType.
+		 */
+		.tpid = ETH_P_EDSA,
+		.tpid2 = ETH_P_EDSA,
 		/* P/Q/R/S only */
 		.queue_ts = 0,
 		.egrmirrvid = 0,
@@ -930,12 +940,245 @@ static void sja1105_bridge_leave(struct dsa_switch *ds, int port,
 	sja1105_bridge_member(ds, port, br, false);
 }
 
+/* For situations where we need to change a setting at runtime that is only
+ * available through the static configuration, resetting the switch in order
+ * to upload the new static config is unavoidable. Back up the settings we
+ * modify at runtime (currently only MAC) and restore them after uploading,
+ * such that this operation is relatively seamless.
+ */
+static int sja1105_static_config_reload(struct sja1105_private *priv)
+{
+	struct sja1105_mac_config_entry *mac;
+	int speed_mbps[SJA1105_NUM_PORTS];
+	int rc, i;
+
+	mac = priv->static_config.tables[BLK_IDX_MAC_CONFIG].entries;
+
+	/* Back up settings changed by sja1105_adjust_port_config and
+	 * and restore their defaults.
+	 */
+	for (i = 0; i < SJA1105_NUM_PORTS; i++) {
+		speed_mbps[i] = sja1105_speed[mac[i].speed];
+		mac[i].speed = SJA1105_SPEED_AUTO;
+	}
+
+	/* Reset switch and send updated static configuration */
+	rc = sja1105_static_config_upload(priv);
+	if (rc < 0)
+		goto out;
+
+	/* Configure the CGU (PLLs) for MII and RMII PHYs.
+	 * For these interfaces there is no dynamic configuration
+	 * needed, since PLLs have same settings at all speeds.
+	 */
+	rc = sja1105_clocking_setup(priv);
+	if (rc < 0)
+		goto out;
+
+	for (i = 0; i < SJA1105_NUM_PORTS; i++) {
+		bool enabled = (speed_mbps[i] != 0);
+
+		rc = sja1105_adjust_port_config(priv, i, speed_mbps[i],
+						enabled);
+		if (rc < 0)
+			goto out;
+	}
+out:
+	return rc;
+}
+
+#define sja1105_vlan_filtering_enabled(priv) \
+	(((struct sja1105_general_params_entry *) \
+	((struct sja1105_private *)priv)->static_config. \
+	tables[BLK_IDX_GENERAL_PARAMS].entries)->tpid == ETH_P_8021Q)
+
+/* The TPID setting belongs to the General Parameters table,
+ * which can only be partially reconfigured at runtime (and not the TPID).
+ * So a switch reset is required.
+ */
+static int sja1105_change_tpid(struct sja1105_private *priv,
+			       u16 tpid, u16 tpid2)
+{
+	struct sja1105_general_params_entry *general_params;
+	struct sja1105_table *table;
+
+	table = &priv->static_config.tables[BLK_IDX_GENERAL_PARAMS];
+	general_params = table->entries;
+	general_params->tpid = tpid;
+	general_params->tpid2 = tpid2;
+	return sja1105_static_config_reload(priv);
+}
+
+static int sja1105_pvid_apply(struct sja1105_private *priv, int port, u16 pvid)
+{
+	struct sja1105_mac_config_entry *mac;
+
+	mac = priv->static_config.tables[BLK_IDX_MAC_CONFIG].entries;
+
+	mac[port].vlanid = pvid;
+
+	return sja1105_dynamic_config_write(priv, BLK_IDX_MAC_CONFIG, port,
+					   &mac[port], true);
+}
+
+static int sja1105_is_vlan_configured(struct sja1105_private *priv, u16 vid)
+{
+	struct sja1105_vlan_lookup_entry *vlan;
+	int count, i;
+
+	vlan = priv->static_config.tables[BLK_IDX_VLAN_LOOKUP].entries;
+	count = priv->static_config.tables[BLK_IDX_VLAN_LOOKUP].entry_count;
+
+	for (i = 0; i < count; i++)
+		if (vlan[i].vlanid == vid)
+			return i;
+
+	/* Return an invalid entry index if not found */
+	return -1;
+}
+
+static int sja1105_vlan_apply(struct sja1105_private *priv, int port, u16 vid,
+			      bool enabled, bool untagged)
+{
+	struct sja1105_vlan_lookup_entry *vlan;
+	struct sja1105_table *table;
+	bool keep = true;
+	int match, rc;
+
+	table = &priv->static_config.tables[BLK_IDX_VLAN_LOOKUP];
+
+	match = sja1105_is_vlan_configured(priv, vid);
+	if (match < 0) {
+		/* Can't delete a missing entry. */
+		if (!enabled)
+			return 0;
+		rc = sja1105_table_resize(table, table->entry_count + 1);
+		if (rc)
+			return rc;
+		match = table->entry_count - 1;
+	}
+	/* Assign pointer after the resize (it's new memory) */
+	vlan = table->entries;
+	vlan[match].vlanid = vid;
+	if (enabled) {
+		vlan[match].vlan_bc |= BIT(port);
+		vlan[match].vmemb_port |= BIT(port);
+	} else {
+		vlan[match].vlan_bc &= ~BIT(port);
+		vlan[match].vmemb_port &= ~BIT(port);
+	}
+	/* Also unset tag_port if removing this VLAN was requested,
+	 * just so we don't have a confusing bitmap (no practical purpose).
+	 */
+	if (untagged || !enabled)
+		vlan[match].tag_port &= ~BIT(port);
+	else
+		vlan[match].tag_port |= BIT(port);
+	/* If there's no port left as member of this VLAN,
+	 * it's time for it to go.
+	 */
+	if (!vlan[match].vmemb_port)
+		keep = false;
+
+	dev_dbg(priv->ds->dev,
+		"%s: port %d, vid %llu, broadcast domain 0x%llx, "
+		"port members 0x%llx, tagged ports 0x%llx, keep %d\n",
+		__func__, port, vlan[match].vlanid, vlan[match].vlan_bc,
+		vlan[match].vmemb_port, vlan[match].tag_port, keep);
+
+	rc = sja1105_dynamic_config_write(priv, BLK_IDX_VLAN_LOOKUP, vid,
+					  &vlan[match], keep);
+	if (rc < 0)
+		return rc;
+
+	if (!keep)
+		return sja1105_table_delete_entry(table, match);
+
+	return 0;
+}
+
 static enum dsa_tag_protocol
 sja1105_get_tag_protocol(struct dsa_switch *ds, int port)
 {
 	return DSA_TAG_PROTO_NONE;
 }
 
+/* This callback needs to be present */
+static int sja1105_vlan_prepare(struct dsa_switch *ds, int port,
+				const struct switchdev_obj_port_vlan *vlan)
+{
+	return 0;
+}
+
+static int sja1105_vlan_filtering(struct dsa_switch *ds, int port, bool enabled)
+{
+	struct sja1105_private *priv = ds->priv;
+	int rc;
+
+	if (enabled && !sja1105_vlan_filtering_enabled(priv))
+		/* Enable VLAN filtering.
+		 * TODO read these from bridge_dev->protocol.
+		 */
+		rc = sja1105_change_tpid(priv, ETH_P_8021Q, ETH_P_8021AD);
+	else if (!enabled && sja1105_vlan_filtering_enabled(priv))
+		/* Disable VLAN filtering. TODO: Install a TPID
+		 * that also encodes the switch ID (aka ds->index)
+		 * so that stacking switch tags will be supported.
+		 */
+		rc = sja1105_change_tpid(priv, ETH_P_EDSA, ETH_P_EDSA);
+	else
+		return 0;
+	if (rc)
+		dev_err(ds->dev, "Failed to change VLAN Ethertype\n");
+
+	return rc;
+}
+
+static void sja1105_vlan_add(struct dsa_switch *ds, int port,
+			     const struct switchdev_obj_port_vlan *vlan)
+{
+	struct sja1105_private *priv = ds->priv;
+	u16 vid;
+	int rc;
+
+	for (vid = vlan->vid_begin; vid <= vlan->vid_end; vid++) {
+		rc = sja1105_vlan_apply(priv, port, vid, true, vlan->flags &
+					BRIDGE_VLAN_INFO_UNTAGGED);
+		if (rc < 0) {
+			dev_err(ds->dev, "Failed to add VLAN %d to port %d: %d\n",
+				vid, port, rc);
+			return;
+		}
+		if (vlan->flags & BRIDGE_VLAN_INFO_PVID) {
+			rc = sja1105_pvid_apply(ds->priv, port, vid);
+			if (rc < 0) {
+				dev_err(ds->dev, "Failed to set pvid %d on port %d: %d\n",
+					vid, port, rc);
+				return;
+			}
+		}
+	}
+}
+
+static int sja1105_vlan_del(struct dsa_switch *ds, int port,
+			    const struct switchdev_obj_port_vlan *vlan)
+{
+	struct sja1105_private *priv = ds->priv;
+	u16 vid;
+	int rc;
+
+	for (vid = vlan->vid_begin; vid <= vlan->vid_end; vid++) {
+		rc = sja1105_vlan_apply(priv, port, vid, false, vlan->flags &
+					BRIDGE_VLAN_INFO_UNTAGGED);
+		if (rc < 0) {
+			dev_err(ds->dev, "Failed to remove VLAN %d from port %d: %d\n",
+				vid, port, rc);
+			return rc;
+		}
+	}
+	return 0;
+}
+
 /* The programming model for the SJA1105 switch is "all-at-once" via static
  * configuration tables. Some of these can be dynamically modified at runtime,
  * but not the xMII mode parameters table.
@@ -971,6 +1214,15 @@ static int sja1105_setup(struct dsa_switch *ds)
 		dev_err(ds->dev, "Failed to configure MII clocking: %d\n", rc);
 		return rc;
 	}
+	/* On SJA1105, VLAN filtering per se is always enabled in hardware.
+	 * The only thing we can do to disable it is lie about what the 802.1Q
+	 * EtherType is. So it will still try to apply VLAN filtering, but all
+	 * ingress traffic (except frames received with EtherType of
+	 * ETH_P_EDSA, which are invalid) will be internally tagged with a
+	 * distorted VLAN header where the TPID is ETH_P_EDSA, and the VLAN ID
+	 * is the port pvid.
+	 */
+	ds->vlan_filtering_is_global = true;
 
 	return 0;
 }
@@ -984,6 +1236,10 @@ static const struct dsa_switch_ops sja1105_switch_ops = {
 	.port_fdb_del		= sja1105_fdb_del,
 	.port_bridge_join	= sja1105_bridge_join,
 	.port_bridge_leave	= sja1105_bridge_leave,
+	.port_vlan_prepare	= sja1105_vlan_prepare,
+	.port_vlan_filtering	= sja1105_vlan_filtering,
+	.port_vlan_add		= sja1105_vlan_add,
+	.port_vlan_del		= sja1105_vlan_del,
 	.port_mdb_prepare	= sja1105_mdb_prepare,
 	.port_mdb_add		= sja1105_mdb_add,
 	.port_mdb_del		= sja1105_mdb_del,
diff --git a/drivers/net/dsa/sja1105/sja1105_static_config.c b/drivers/net/dsa/sja1105/sja1105_static_config.c
index f5d1866216bd..c477f0520141 100644
--- a/drivers/net/dsa/sja1105/sja1105_static_config.c
+++ b/drivers/net/dsa/sja1105/sja1105_static_config.c
@@ -1034,3 +1034,41 @@ void sja1105_static_config_free(struct sja1105_static_config *config)
 	}
 }
 
+int sja1105_table_delete_entry(struct sja1105_table *table, int i)
+{
+	size_t entry_size = table->ops->unpacked_entry_size;
+	u8 *entries = table->entries;
+
+	if (i > table->entry_count)
+		return -ERANGE;
+
+	memmove(entries + i * entry_size, entries + (i + 1) * entry_size,
+		(table->entry_count - i) * entry_size);
+
+	table->entry_count--;
+
+	return 0;
+}
+
+/* No pointers to table->entries should be kept when this is called. */
+int sja1105_table_resize(struct sja1105_table *table, size_t new_count)
+{
+	size_t entry_size = table->ops->unpacked_entry_size;
+	void *new_entries, *old_entries = table->entries;
+
+	if (new_count > table->ops->max_entry_count)
+		return -ERANGE;
+
+	new_entries = kcalloc(new_count, entry_size, GFP_KERNEL);
+	if (!new_entries)
+		return -ENOMEM;
+
+	memcpy(new_entries, old_entries, min(new_count, table->entry_count) *
+		entry_size);
+
+	table->entries = new_entries;
+	table->entry_count = new_count;
+	kfree(old_entries);
+	return 0;
+}
+
diff --git a/drivers/net/dsa/sja1105/sja1105_static_config.h b/drivers/net/dsa/sja1105/sja1105_static_config.h
index 1c331c9e20b3..2f9db15a454d 100644
--- a/drivers/net/dsa/sja1105/sja1105_static_config.h
+++ b/drivers/net/dsa/sja1105/sja1105_static_config.h
@@ -297,6 +297,9 @@ int sja1105_static_config_init(struct sja1105_static_config *config,
 			       u64 device_id, u64 part_nr);
 void sja1105_static_config_free(struct sja1105_static_config *config);
 
+int sja1105_table_delete_entry(struct sja1105_table *table, int i);
+int sja1105_table_resize(struct sja1105_table *table, size_t new_count);
+
 u32 sja1105_crc32(const void *buf, size_t len);
 
 void sja1105_pack(void *buf, const u64 *val, int start, int end, size_t len);
-- 
2.17.1

[PATCH net-next 13/17] net: dsa: sja1105: Add support for ethtool port counters

[Vladimir Oltean]

Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
---
 drivers/net/dsa/sja1105/Makefile          |   1 +
 drivers/net/dsa/sja1105/sja1105.h         |   6 +
 drivers/net/dsa/sja1105/sja1105_ethtool.c | 423 ++++++++++++++++++++++
 drivers/net/dsa/sja1105/sja1105_main.c    |   3 +
 4 files changed, 433 insertions(+)
 create mode 100644 drivers/net/dsa/sja1105/sja1105_ethtool.c

diff --git a/drivers/net/dsa/sja1105/Makefile b/drivers/net/dsa/sja1105/Makefile
index ed00840802f4..bb4404c79eb2 100644
--- a/drivers/net/dsa/sja1105/Makefile
+++ b/drivers/net/dsa/sja1105/Makefile
@@ -3,6 +3,7 @@ obj-$(CONFIG_NET_DSA_SJA1105) += sja1105.o
 sja1105-objs := \
     sja1105_spi.o \
     sja1105_main.o \
+    sja1105_ethtool.o \
     sja1105_clocking.o \
     sja1105_static_config.o \
     sja1105_dynamic_config.o \
diff --git a/drivers/net/dsa/sja1105/sja1105.h b/drivers/net/dsa/sja1105/sja1105.h
index 26d93afa77a1..dd41d7c70370 100644
--- a/drivers/net/dsa/sja1105/sja1105.h
+++ b/drivers/net/dsa/sja1105/sja1105.h
@@ -108,6 +108,12 @@ typedef enum {
 int sja1105_clocking_setup_port(struct sja1105_private *priv, int port);
 int sja1105_clocking_setup(struct sja1105_private *priv);
 
+/* From sja1105-ethtool.c */
+void sja1105_get_ethtool_stats(struct dsa_switch *ds, int port, u64 *data);
+void sja1105_get_strings(struct dsa_switch *ds, int port,
+			 u32 stringset, u8 *data);
+int sja1105_get_sset_count(struct dsa_switch *ds, int port, int sset);
+
 /* From sja1105-dynamic-config.c */
 
 int sja1105_dynamic_config_read(struct sja1105_private *priv,
diff --git a/drivers/net/dsa/sja1105/sja1105_ethtool.c b/drivers/net/dsa/sja1105/sja1105_ethtool.c
new file mode 100644
index 000000000000..a52ee8107fc2
--- /dev/null
+++ b/drivers/net/dsa/sja1105/sja1105_ethtool.c
@@ -0,0 +1,423 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2018-2019, Vladimir Oltean <olteanv@gmail.com>
+ */
+#include "sja1105.h"
+
+struct sja1105_port_status_mac {
+	u64 n_runt;
+	u64 n_soferr;
+	u64 n_alignerr;
+	u64 n_miierr;
+	u64 typeerr;
+	u64 sizeerr;
+	u64 tctimeout;
+	u64 priorerr;
+	u64 nomaster;
+	u64 memov;
+	u64 memerr;
+	u64 invtyp;
+	u64 intcyov;
+	u64 domerr;
+	u64 pcfbagdrop;
+	u64 spcprior;
+	u64 ageprior;
+	u64 portdrop;
+	u64 lendrop;
+	u64 bagdrop;
+	u64 policeerr;
+	u64 drpnona664err;
+	u64 spcerr;
+	u64 agedrp;
+};
+
+struct sja1105_port_status_hl1 {
+	u64 n_n664err;
+	u64 n_vlanerr;
+	u64 n_unreleased;
+	u64 n_sizeerr;
+	u64 n_crcerr;
+	u64 n_vlnotfound;
+	u64 n_ctpolerr;
+	u64 n_polerr;
+	u64 n_rxfrmsh;
+	u64 n_rxfrm;
+	u64 n_rxbytesh;
+	u64 n_rxbyte;
+	u64 n_txfrmsh;
+	u64 n_txfrm;
+	u64 n_txbytesh;
+	u64 n_txbyte;
+};
+
+struct sja1105_port_status_hl2 {
+	u64 n_qfull;
+	u64 n_part_drop;
+	u64 n_egr_disabled;
+	u64 n_not_reach;
+	u64 qlevel_hwm[8]; /* Only for P/Q/R/S */
+	u64 qlevel[8];     /* Only for P/Q/R/S */
+};
+
+struct sja1105_port_status {
+	struct sja1105_port_status_mac mac;
+	struct sja1105_port_status_hl1 hl1;
+	struct sja1105_port_status_hl2 hl2;
+};
+
+static void
+sja1105_port_status_mac_unpack(void *buf,
+			       struct sja1105_port_status_mac *status)
+{
+	/* Make pointer arithmetic work on 4 bytes */
+	u32 *p = (u32 *)buf;
+
+	sja1105_unpack(p + 0x0, &status->n_runt,       31, 24, 4);
+	sja1105_unpack(p + 0x0, &status->n_soferr,     23, 16, 4);
+	sja1105_unpack(p + 0x0, &status->n_alignerr,   15,  8, 4);
+	sja1105_unpack(p + 0x0, &status->n_miierr,      7,  0, 4);
+	sja1105_unpack(p + 0x1, &status->typeerr,      27, 27, 4);
+	sja1105_unpack(p + 0x1, &status->sizeerr,      26, 26, 4);
+	sja1105_unpack(p + 0x1, &status->tctimeout,    25, 25, 4);
+	sja1105_unpack(p + 0x1, &status->priorerr,     24, 24, 4);
+	sja1105_unpack(p + 0x1, &status->nomaster,     23, 23, 4);
+	sja1105_unpack(p + 0x1, &status->memov,        22, 22, 4);
+	sja1105_unpack(p + 0x1, &status->memerr,       21, 21, 4);
+	sja1105_unpack(p + 0x1, &status->invtyp,       19, 19, 4);
+	sja1105_unpack(p + 0x1, &status->intcyov,      18, 18, 4);
+	sja1105_unpack(p + 0x1, &status->domerr,       17, 17, 4);
+	sja1105_unpack(p + 0x1, &status->pcfbagdrop,   16, 16, 4);
+	sja1105_unpack(p + 0x1, &status->spcprior,     15, 12, 4);
+	sja1105_unpack(p + 0x1, &status->ageprior,     11,  8, 4);
+	sja1105_unpack(p + 0x1, &status->portdrop,      6,  6, 4);
+	sja1105_unpack(p + 0x1, &status->lendrop,       5,  5, 4);
+	sja1105_unpack(p + 0x1, &status->bagdrop,       4,  4, 4);
+	sja1105_unpack(p + 0x1, &status->policeerr,     3,  3, 4);
+	sja1105_unpack(p + 0x1, &status->drpnona664err, 2,  2, 4);
+	sja1105_unpack(p + 0x1, &status->spcerr,        1,  1, 4);
+	sja1105_unpack(p + 0x1, &status->agedrp,        0,  0, 4);
+}
+
+static void
+sja1105_port_status_hl1_unpack(void *buf,
+			       struct sja1105_port_status_hl1 *status)
+{
+	/* Make pointer arithmetic work on 4 bytes */
+	u32 *p = (u32 *)buf;
+
+	sja1105_unpack(p + 0xF, &status->n_n664err,    31,  0, 4);
+	sja1105_unpack(p + 0xE, &status->n_vlanerr,    31,  0, 4);
+	sja1105_unpack(p + 0xD, &status->n_unreleased, 31,  0, 4);
+	sja1105_unpack(p + 0xC, &status->n_sizeerr,    31,  0, 4);
+	sja1105_unpack(p + 0xB, &status->n_crcerr,     31,  0, 4);
+	sja1105_unpack(p + 0xA, &status->n_vlnotfound, 31,  0, 4);
+	sja1105_unpack(p + 0x9, &status->n_ctpolerr,   31,  0, 4);
+	sja1105_unpack(p + 0x8, &status->n_polerr,     31,  0, 4);
+	sja1105_unpack(p + 0x7, &status->n_rxfrmsh,    31,  0, 4);
+	sja1105_unpack(p + 0x6, &status->n_rxfrm,      31,  0, 4);
+	sja1105_unpack(p + 0x5, &status->n_rxbytesh,   31,  0, 4);
+	sja1105_unpack(p + 0x4, &status->n_rxbyte,     31,  0, 4);
+	sja1105_unpack(p + 0x3, &status->n_txfrmsh,    31,  0, 4);
+	sja1105_unpack(p + 0x2, &status->n_txfrm,      31,  0, 4);
+	sja1105_unpack(p + 0x1, &status->n_txbytesh,   31,  0, 4);
+	sja1105_unpack(p + 0x0, &status->n_txbyte,     31,  0, 4);
+	status->n_rxfrm  += status->n_rxfrmsh  << 32;
+	status->n_rxbyte += status->n_rxbytesh << 32;
+	status->n_txfrm  += status->n_txfrmsh  << 32;
+	status->n_txbyte += status->n_txbytesh << 32;
+}
+
+static void
+sja1105_port_status_hl2_unpack(void *buf,
+			       struct sja1105_port_status_hl2 *status)
+{
+	/* Make pointer arithmetic work on 4 bytes */
+	u32 *p = (u32 *)buf;
+
+	sja1105_unpack(p + 0x3, &status->n_qfull,        31,  0, 4);
+	sja1105_unpack(p + 0x2, &status->n_part_drop,    31,  0, 4);
+	sja1105_unpack(p + 0x1, &status->n_egr_disabled, 31,  0, 4);
+	sja1105_unpack(p + 0x0, &status->n_not_reach,    31,  0, 4);
+}
+
+static void
+sja1105pqrs_port_status_qlevel_unpack(void *buf,
+				      struct sja1105_port_status_hl2 *status)
+{
+	/* Make pointer arithmetic work on 4 bytes */
+	u32 *p = (u32 *)buf;
+	int i;
+
+	for (i = 0; i < 8; i++) {
+		sja1105_unpack(p + i, &status->qlevel_hwm[i], 24, 16, 4);
+		sja1105_unpack(p + i, &status->qlevel[i],      8,  0, 4);
+	}
+}
+
+static int sja1105_port_status_get_mac(struct sja1105_private *priv,
+				       struct sja1105_port_status_mac *status,
+				       int port)
+{
+#define SIZE_MAC_AREA (0x02 * 4)
+	u8 packed_buf[SIZE_MAC_AREA];
+	int rc = 0;
+
+	memset(status, 0, sizeof(*status));
+
+	/* MAC area */
+	rc = sja1105_spi_send_packed_buf(priv, SPI_READ, priv->regs->mac[port],
+					 packed_buf, SIZE_MAC_AREA);
+	if (rc < 0)
+		return rc;
+
+	sja1105_port_status_mac_unpack(packed_buf, status);
+
+	return 0;
+#undef SIZE_MAC_AREA
+}
+
+static int sja1105_port_status_get_hl1(struct sja1105_private *priv,
+				       struct sja1105_port_status_hl1 *status,
+				       int port)
+{
+#define SIZE_HL1_AREA (0x10 * 4)
+	u8 packed_buf[SIZE_HL1_AREA];
+	int rc = 0;
+
+	memset(status, 0, sizeof(*status));
+
+	rc = sja1105_spi_send_packed_buf(priv, SPI_READ,
+					 priv->regs->mac_hl1[port],
+					 packed_buf, SIZE_HL1_AREA);
+	if (rc < 0)
+		return rc;
+
+	sja1105_port_status_hl1_unpack(packed_buf, status);
+
+	return 0;
+#undef SIZE_HL1_AREA
+}
+
+static int sja1105_port_status_get_hl2(struct sja1105_private *priv,
+				       struct sja1105_port_status_hl2 *status,
+				       int port)
+{
+#define SIZE_HL2_AREA (0x4 * 4)
+#define SIZE_QLEVEL_AREA (0x8 * 4) /* 0x4 to 0xB */
+	u8 packed_buf[SIZE_QLEVEL_AREA];
+	int rc = 0;
+
+	memset(status, 0, sizeof(*status));
+
+	rc = sja1105_spi_send_packed_buf(priv, SPI_READ,
+					 priv->regs->mac_hl2[port],
+					 packed_buf, SIZE_HL2_AREA);
+	if (rc < 0)
+		return rc;
+
+	sja1105_port_status_hl2_unpack(packed_buf, status);
+
+	if (IS_ET(priv->device_id))
+		/* Code below is strictly P/Q/R/S specific. */
+		return 0;
+
+	rc = sja1105_spi_send_packed_buf(priv, SPI_READ,
+					 priv->regs->qlevel[port],
+					 packed_buf, SIZE_QLEVEL_AREA);
+	if (rc < 0)
+		return rc;
+
+	sja1105pqrs_port_status_qlevel_unpack(packed_buf, status);
+
+	return 0;
+#undef SIZE_QLEVEL_AREA
+#undef SIZE_HL2_AREA
+}
+
+static int sja1105_port_status_get(struct sja1105_private *priv,
+				   struct sja1105_port_status *status,
+				   int port)
+{
+	int rc;
+
+	rc = sja1105_port_status_get_mac(priv, &status->mac, port);
+	if (rc < 0)
+		return rc;
+	rc = sja1105_port_status_get_hl1(priv, &status->hl1, port);
+	if (rc < 0)
+		return rc;
+	rc = sja1105_port_status_get_hl2(priv, &status->hl2, port);
+	if (rc < 0)
+		return rc;
+
+	return 0;
+}
+
+static char sja1105_port_stats[][ETH_GSTRING_LEN] = {
+	/* MAC-Level Diagnostic Counters */
+	"n_runt",
+	"n_soferr",
+	"n_alignerr",
+	"n_miierr",
+	/* MAC-Level Diagnostic Flags */
+	"typeerr",
+	"sizeerr",
+	"tctimeout",
+	"priorerr",
+	"nomaster",
+	"memov",
+	"memerr",
+	"invtyp",
+	"intcyov",
+	"domerr",
+	"pcfbagdrop",
+	"spcprior",
+	"ageprior",
+	"portdrop",
+	"lendrop",
+	"bagdrop",
+	"policeerr",
+	"drpnona664err",
+	"spcerr",
+	"agedrp",
+	/* High-Level Diagnostic Counters */
+	"n_n664err",
+	"n_vlanerr",
+	"n_unreleased",
+	"n_sizeerr",
+	"n_crcerr",
+	"n_vlnotfound",
+	"n_ctpolerr",
+	"n_polerr",
+	"n_rxfrm",
+	"n_rxbyte",
+	"n_txfrm",
+	"n_txbyte",
+	"n_qfull",
+	"n_part_drop",
+	"n_egr_disabled",
+	"n_not_reach",
+};
+
+static char sja1105pqrs_extra_port_stats[][ETH_GSTRING_LEN] = {
+	/* Queue Levels */
+	"qlevel_hwm_0",
+	"qlevel_hwm_1",
+	"qlevel_hwm_2",
+	"qlevel_hwm_3",
+	"qlevel_hwm_4",
+	"qlevel_hwm_5",
+	"qlevel_hwm_6",
+	"qlevel_hwm_7",
+	"qlevel_0",
+	"qlevel_1",
+	"qlevel_2",
+	"qlevel_3",
+	"qlevel_4",
+	"qlevel_5",
+	"qlevel_6",
+	"qlevel_7",
+};
+
+void sja1105_get_ethtool_stats(struct dsa_switch *ds, int port, u64 *data)
+{
+	struct sja1105_private *priv = ds->priv;
+	struct sja1105_port_status status;
+	int rc, i, k = 0;
+
+	rc = sja1105_port_status_get(priv, &status, port);
+	if (rc < 0) {
+		dev_err(ds->dev, "Failed to read port %d counters: %d\n",
+			port, rc);
+		return;
+	}
+	memset(data, 0, ARRAY_SIZE(sja1105_port_stats) * sizeof(u64));
+	data[k++] = status.mac.n_runt;
+	data[k++] = status.mac.n_soferr;
+	data[k++] = status.mac.n_alignerr;
+	data[k++] = status.mac.n_miierr;
+	data[k++] = status.mac.typeerr;
+	data[k++] = status.mac.sizeerr;
+	data[k++] = status.mac.tctimeout;
+	data[k++] = status.mac.priorerr;
+	data[k++] = status.mac.nomaster;
+	data[k++] = status.mac.memov;
+	data[k++] = status.mac.memerr;
+	data[k++] = status.mac.invtyp;
+	data[k++] = status.mac.intcyov;
+	data[k++] = status.mac.domerr;
+	data[k++] = status.mac.pcfbagdrop;
+	data[k++] = status.mac.spcprior;
+	data[k++] = status.mac.ageprior;
+	data[k++] = status.mac.portdrop;
+	data[k++] = status.mac.lendrop;
+	data[k++] = status.mac.bagdrop;
+	data[k++] = status.mac.policeerr;
+	data[k++] = status.mac.drpnona664err;
+	data[k++] = status.mac.spcerr;
+	data[k++] = status.mac.agedrp;
+	data[k++] = status.hl1.n_n664err;
+	data[k++] = status.hl1.n_vlanerr;
+	data[k++] = status.hl1.n_unreleased;
+	data[k++] = status.hl1.n_sizeerr;
+	data[k++] = status.hl1.n_crcerr;
+	data[k++] = status.hl1.n_vlnotfound;
+	data[k++] = status.hl1.n_ctpolerr;
+	data[k++] = status.hl1.n_polerr;
+	data[k++] = status.hl1.n_rxfrm;
+	data[k++] = status.hl1.n_rxbyte;
+	data[k++] = status.hl1.n_txfrm;
+	data[k++] = status.hl1.n_txbyte;
+	data[k++] = status.hl2.n_qfull;
+	data[k++] = status.hl2.n_part_drop;
+	data[k++] = status.hl2.n_egr_disabled;
+	data[k++] = status.hl2.n_not_reach;
+
+	if (!IS_PQRS(priv->device_id))
+		return;
+
+	memset(data + k, 0, ARRAY_SIZE(sja1105pqrs_extra_port_stats) *
+			sizeof(u64));
+	for (i = 0; i < 8; i++) {
+		data[k++] = status.hl2.qlevel_hwm[i];
+		data[k++] = status.hl2.qlevel[i];
+	}
+}
+
+void sja1105_get_strings(struct dsa_switch *ds, int port,
+			 u32 stringset, u8 *data)
+{
+	struct sja1105_private *priv = ds->priv;
+	u8 *p = data;
+	int i;
+
+	switch (stringset) {
+	case ETH_SS_STATS:
+		for (i = 0; i < ARRAY_SIZE(sja1105_port_stats); i++) {
+			strlcpy(p, sja1105_port_stats[i], ETH_GSTRING_LEN);
+			p += ETH_GSTRING_LEN;
+		}
+		if (!IS_PQRS(priv->device_id))
+			return;
+		for (i = 0; i < ARRAY_SIZE(sja1105pqrs_extra_port_stats); i++) {
+			strlcpy(p, sja1105pqrs_extra_port_stats[i],
+				ETH_GSTRING_LEN);
+			p += ETH_GSTRING_LEN;
+		}
+		break;
+	}
+}
+
+int sja1105_get_sset_count(struct dsa_switch *ds, int port, int sset)
+{
+	int count = ARRAY_SIZE(sja1105_port_stats);
+	struct sja1105_private *priv = ds->priv;
+
+	if (sset != ETH_SS_STATS)
+		return -EOPNOTSUPP;
+
+	if (IS_PQRS(priv->device_id))
+		count += ARRAY_SIZE(sja1105pqrs_extra_port_stats);
+
+	return count;
+}
+
diff --git a/drivers/net/dsa/sja1105/sja1105_main.c b/drivers/net/dsa/sja1105/sja1105_main.c
index 028ed1ebe358..066577192511 100644
--- a/drivers/net/dsa/sja1105/sja1105_main.c
+++ b/drivers/net/dsa/sja1105/sja1105_main.c
@@ -1231,6 +1231,9 @@ static const struct dsa_switch_ops sja1105_switch_ops = {
 	.get_tag_protocol	= sja1105_get_tag_protocol,
 	.setup			= sja1105_setup,
 	.adjust_link		= sja1105_adjust_link,
+	.get_strings		= sja1105_get_strings,
+	.get_ethtool_stats	= sja1105_get_ethtool_stats,
+	.get_sset_count		= sja1105_get_sset_count,
 	.port_fdb_dump		= sja1105_fdb_dump,
 	.port_fdb_add		= sja1105_fdb_add,
 	.port_fdb_del		= sja1105_fdb_del,
-- 
2.17.1

[PATCH net-next 14/17] net: dsa: sja1105: Add support for traffic through standalone ports

[Vladimir Oltean]

In order to support this, we are creating a make-shift switch tag out of
a VLAN trunk configured on the CPU port. Termination on switch ports
only works when not under a vlan_filtering bridge. We are making use of
the generic CONFIG_NET_DSA_TAG_8021Q code and leveraging it from our own
CONFIG_NET_DSA_TAG_SJA1105.

There are two types of traffic: regular and link-local.
The link-local traffic received on the CPU port is trapped from the
switch's regular forwarding decisions because it matched one of the two
DMAC filters for management traffic.
On transmission, the switch requires special massaging for these
link-local frames. Due to a weird implementation of the switching IP, by
default it drops link-local frames that originate on the CPU port. It
needs to be told where to forward them to, through an SPI command
("management route") that is valid for only a single frame.
So when we're sending link-local traffic, we need to clone skb's from
DSA and send them in our custom xmit worker that also performs SPI access.

For that purpose, the DSA xmit handler and the xmit worker communicate
through a per-port "skb ring" software structure, with a producer and a
consumer index. At the moment this structure is rather fragile
(ping-flooding to a link-local DMAC would cause most of the frames to
get dropped). I would like to move the management traffic on a separate
netdev queue that I can stop when the skb ring got full and hardware is
busy processing, so that we are not forced to drop traffic.

Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
---
 drivers/net/dsa/sja1105/sja1105.h      |   8 ++
 drivers/net/dsa/sja1105/sja1105_main.c | 117 +++++++++++++++++++-
 include/linux/dsa/sja1105.h            |  52 +++++++++
 include/net/dsa.h                      |   1 +
 net/dsa/Kconfig                        |   3 +
 net/dsa/Makefile                       |   1 +
 net/dsa/dsa.c                          |   6 ++
 net/dsa/dsa_priv.h                     |   3 +
 net/dsa/tag_sja1105.c                  | 142 +++++++++++++++++++++++++
 9 files changed, 332 insertions(+), 1 deletion(-)
 create mode 100644 include/linux/dsa/sja1105.h
 create mode 100644 net/dsa/tag_sja1105.c

diff --git a/drivers/net/dsa/sja1105/sja1105.h b/drivers/net/dsa/sja1105/sja1105.h
index dd41d7c70370..3677f9dd78bb 100644
--- a/drivers/net/dsa/sja1105/sja1105.h
+++ b/drivers/net/dsa/sja1105/sja1105.h
@@ -5,6 +5,7 @@
 #ifndef _SJA1105_H
 #define _SJA1105_H
 
+#include <linux/dsa/sja1105.h>
 #include <net/dsa.h>
 #include "sja1105_static_config.h"
 
@@ -19,6 +20,12 @@
 #define SJA1105_NUM_TC    8
 #define SJA1105ET_FDB_BIN_SIZE 4
 
+struct sja1105_port {
+	struct dsa_port *dp;
+	struct work_struct xmit_work;
+	struct sja1105_skb_ring xmit_ring;
+};
+
 /* Keeps the different addresses between E/T and P/Q/R/S */
 struct sja1105_regs {
 	u64 general_status;
@@ -50,6 +57,7 @@ struct sja1105_private {
 	struct dsa_switch *ds;
 	u64 device_id;
 	u64 part_nr; /* Needed for P/R distinction (same switch core) */
+	struct sja1105_port ports[SJA1105_NUM_PORTS];
 };
 
 #include "sja1105_dynamic_config.h"
diff --git a/drivers/net/dsa/sja1105/sja1105_main.c b/drivers/net/dsa/sja1105/sja1105_main.c
index 066577192511..adf79b1caf55 100644
--- a/drivers/net/dsa/sja1105/sja1105_main.c
+++ b/drivers/net/dsa/sja1105/sja1105_main.c
@@ -1097,6 +1097,21 @@ static int sja1105_vlan_apply(struct sja1105_private *priv, int port, u16 vid,
 	return 0;
 }
 
+static int sja1105_setup_8021q_tagging(struct dsa_switch *ds, bool enabled)
+{
+	int rc, i;
+
+	for (i = 0; i < SJA1105_NUM_PORTS; i++) {
+		rc = dsa_port_setup_8021q_tagging(ds, i, enabled);
+		if (rc < 0) {
+			dev_err(ds->dev, "Failed to setup VLAN tagging for port %d: %d\n",
+				i, rc);
+			return rc;
+		}
+	}
+	return 0;
+}
+
 static enum dsa_tag_protocol
 sja1105_get_tag_protocol(struct dsa_switch *ds, int port)
 {
@@ -1131,7 +1146,11 @@ static int sja1105_vlan_filtering(struct dsa_switch *ds, int port, bool enabled)
 	if (rc)
 		dev_err(ds->dev, "Failed to change VLAN Ethertype\n");
 
-	return rc;
+	/* Switch port identification based on 802.1Q is only passable
+	 * if we are not under a vlan_filtering bridge. So make sure
+	 * the two configurations are mutually exclusive.
+	 */
+	return sja1105_setup_8021q_tagging(ds, !enabled);
 }
 
 static void sja1105_vlan_add(struct dsa_switch *ds, int port,
@@ -1227,6 +1246,100 @@ static int sja1105_setup(struct dsa_switch *ds)
 	return 0;
 }
 
+#include "../../../net/dsa/dsa_priv.h"
+/* Deferred work is unfortunately necessary because setting up the management
+ * route cannot be done from atomit context (SPI transfer takes a sleepable
+ * lock on the bus)
+ */
+static void sja1105_xmit_work_handler(struct work_struct *work)
+{
+	struct sja1105_port *sp = container_of(work, struct sja1105_port,
+						xmit_work);
+	struct sja1105_private *priv = sp->dp->ds->priv;
+	struct net_device *slave = sp->dp->slave;
+	struct net_device *master = dsa_slave_to_master(slave);
+	int port = (uintptr_t)(sp - priv->ports);
+	struct sk_buff *skb;
+	int i, rc;
+
+	while ((i = sja1105_skb_ring_get(&sp->xmit_ring, &skb)) >= 0) {
+		struct sja1105_mgmt_entry mgmt_route = { 0 };
+		struct ethhdr *hdr;
+		int timeout = 10;
+		int skb_len;
+
+		skb_len = skb->len;
+		hdr = eth_hdr(skb);
+
+		mgmt_route.macaddr = ether_addr_to_u64(hdr->h_dest);
+		mgmt_route.destports = BIT(port);
+		mgmt_route.enfport = 1;
+		mgmt_route.tsreg = 0;
+		mgmt_route.takets = true;
+
+		rc = sja1105_dynamic_config_write(priv, BLK_IDX_MGMT_ROUTE,
+						  port, &mgmt_route, true);
+		if (rc < 0) {
+			kfree_skb(skb);
+			slave->stats.tx_dropped++;
+			continue;
+		}
+
+		/* Transfer skb to the host port. */
+		skb->dev = master;
+		dev_queue_xmit(skb);
+
+		/* Wait until the switch has processed the frame */
+		do {
+			rc = sja1105_dynamic_config_read(priv, BLK_IDX_MGMT_ROUTE,
+							 port, &mgmt_route);
+			if (rc < 0) {
+				slave->stats.tx_errors++;
+				dev_err(priv->ds->dev,
+					"xmit: failed to poll for mgmt route\n");
+				continue;
+			}
+
+			/* UM10944: The ENFPORT flag of the respective entry is
+			 * cleared when a match is found. The host can use this
+			 * flag as an acknowledgment.
+			 */
+			usleep_range(1000, 2000);
+		} while (mgmt_route.enfport && --timeout);
+
+		if (!timeout) {
+			dev_err(priv->ds->dev, "xmit timed out\n");
+			slave->stats.tx_errors++;
+			continue;
+		}
+
+		slave->stats.tx_packets++;
+		slave->stats.tx_bytes += skb_len;
+	}
+}
+
+static int sja1105_port_enable(struct dsa_switch *ds, int port,
+			       struct phy_device *phydev)
+{
+	struct sja1105_private *priv = ds->priv;
+	struct sja1105_port *sp = &priv->ports[port];
+
+	sp->dp = &ds->ports[port];
+	INIT_WORK(&sp->xmit_work, sja1105_xmit_work_handler);
+	return 0;
+}
+
+static void sja1105_port_disable(struct dsa_switch *ds, int port)
+{
+	struct sja1105_private *priv = ds->priv;
+	struct sja1105_port *sp = &priv->ports[port];
+	struct sk_buff *skb;
+
+	cancel_work_sync(&sp->xmit_work);
+	while (sja1105_skb_ring_get(&sp->xmit_ring, &skb) >= 0)
+		kfree_skb(skb);
+}
+
 static const struct dsa_switch_ops sja1105_switch_ops = {
 	.get_tag_protocol	= sja1105_get_tag_protocol,
 	.setup			= sja1105_setup,
@@ -1246,6 +1359,8 @@ static const struct dsa_switch_ops sja1105_switch_ops = {
 	.port_mdb_prepare	= sja1105_mdb_prepare,
 	.port_mdb_add		= sja1105_mdb_add,
 	.port_mdb_del		= sja1105_mdb_del,
+	.port_enable		= sja1105_port_enable,
+	.port_disable		= sja1105_port_disable,
 };
 
 static int sja1105_probe(struct spi_device *spi)
diff --git a/include/linux/dsa/sja1105.h b/include/linux/dsa/sja1105.h
new file mode 100644
index 000000000000..d2419951b0c7
--- /dev/null
+++ b/include/linux/dsa/sja1105.h
@@ -0,0 +1,52 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2019, Vladimir Oltean <olteanv@gmail.com>
+ */
+
+/* Included by drivers/net/dsa/sja1105/sja1105.h and net/dsa/tag_sja1105.c */
+
+#ifndef _NET_DSA_SJA1105_H
+#define _NET_DSA_SJA1105_H
+
+#include <linux/skbuff.h>
+#include <net/dsa.h>
+
+#define SJA1105_SKB_RING_SIZE    20
+
+struct sja1105_skb_ring {
+	struct sk_buff *skb[SJA1105_SKB_RING_SIZE];
+	int count;
+	int pi; /* Producer index */
+	int ci; /* Consumer index */
+};
+
+static inline int sja1105_skb_ring_add(struct sja1105_skb_ring *ring,
+				       struct sk_buff *skb)
+{
+	int index;
+
+	if (ring->count == SJA1105_SKB_RING_SIZE)
+		return -1;
+
+	index = ring->pi;
+	ring->skb[index] = skb;
+	ring->pi = (index + 1) % SJA1105_SKB_RING_SIZE;
+	ring->count++;
+	return index;
+}
+
+static inline int sja1105_skb_ring_get(struct sja1105_skb_ring *ring,
+				       struct sk_buff **skb)
+{
+	int index;
+
+	if (ring->count == 0)
+		return -1;
+
+	index = ring->ci;
+	*skb = ring->skb[index];
+	ring->ci = (index + 1) % SJA1105_SKB_RING_SIZE;
+	ring->count--;
+	return index;
+}
+
+#endif /* _NET_DSA_SJA1105_Hk*/
diff --git a/include/net/dsa.h b/include/net/dsa.h
index 91375bcf2cfb..b09cae52d6e4 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -41,6 +41,7 @@ enum dsa_tag_protocol {
 	DSA_TAG_PROTO_KSZ9893,
 	DSA_TAG_PROTO_LAN9303,
 	DSA_TAG_PROTO_MTK,
+	DSA_TAG_PROTO_SJA1105,
 	DSA_TAG_PROTO_QCA,
 	DSA_TAG_PROTO_TRAILER,
 	DSA_TAG_LAST,		/* MUST BE LAST */
diff --git a/net/dsa/Kconfig b/net/dsa/Kconfig
index 2f3a103d7d1a..feaa40c30425 100644
--- a/net/dsa/Kconfig
+++ b/net/dsa/Kconfig
@@ -63,6 +63,9 @@ config NET_DSA_TAG_LAN9303
 config NET_DSA_TAG_MTK
 	bool
 
+config NET_DSA_TAG_SJA1105
+	bool
+
 config NET_DSA_TAG_TRAILER
 	bool
 
diff --git a/net/dsa/Makefile b/net/dsa/Makefile
index d7fc3253d497..8c294cdb895a 100644
--- a/net/dsa/Makefile
+++ b/net/dsa/Makefile
@@ -15,4 +15,5 @@ dsa_core-$(CONFIG_NET_DSA_TAG_KSZ) += tag_ksz.o
 dsa_core-$(CONFIG_NET_DSA_TAG_LAN9303) += tag_lan9303.o
 dsa_core-$(CONFIG_NET_DSA_TAG_MTK) += tag_mtk.o
 dsa_core-$(CONFIG_NET_DSA_TAG_QCA) += tag_qca.o
+dsa_core-$(CONFIG_NET_DSA_TAG_SJA1105) += tag_sja1105.o
 dsa_core-$(CONFIG_NET_DSA_TAG_TRAILER) += tag_trailer.o
diff --git a/net/dsa/dsa.c b/net/dsa/dsa.c
index 36de4f2a3366..4b5d7cd0294a 100644
--- a/net/dsa/dsa.c
+++ b/net/dsa/dsa.c
@@ -65,6 +65,9 @@ const struct dsa_device_ops *dsa_device_ops[DSA_TAG_LAST] = {
 #ifdef CONFIG_NET_DSA_TAG_MTK
 	[DSA_TAG_PROTO_MTK] = &mtk_netdev_ops,
 #endif
+#ifdef CONFIG_NET_DSA_TAG_SJA1105
+	[DSA_TAG_PROTO_SJA1105] = &sja1105_netdev_ops,
+#endif
 #ifdef CONFIG_NET_DSA_TAG_QCA
 	[DSA_TAG_PROTO_QCA] = &qca_netdev_ops,
 #endif
@@ -102,6 +105,9 @@ const char *dsa_tag_protocol_to_str(const struct dsa_device_ops *ops)
 #ifdef CONFIG_NET_DSA_TAG_MTK
 		[DSA_TAG_PROTO_MTK] = "mtk",
 #endif
+#ifdef CONFIG_NET_DSA_TAG_SJA1105
+		[DSA_TAG_PROTO_SJA1105] = "sja1105",
+#endif
 #ifdef CONFIG_NET_DSA_TAG_QCA
 		[DSA_TAG_PROTO_QCA] = "qca",
 #endif
diff --git a/net/dsa/dsa_priv.h b/net/dsa/dsa_priv.h
index 4219f8979c8a..9e092f6e37e5 100644
--- a/net/dsa/dsa_priv.h
+++ b/net/dsa/dsa_priv.h
@@ -236,6 +236,9 @@ extern const struct dsa_device_ops lan9303_netdev_ops;
 /* tag_mtk.c */
 extern const struct dsa_device_ops mtk_netdev_ops;
 
+/* tag_sja1105.c */
+extern const struct dsa_device_ops sja1105_netdev_ops;
+
 /* tag_qca.c */
 extern const struct dsa_device_ops qca_netdev_ops;
 
diff --git a/net/dsa/tag_sja1105.c b/net/dsa/tag_sja1105.c
new file mode 100644
index 000000000000..fc5d37ec4fd7
--- /dev/null
+++ b/net/dsa/tag_sja1105.c
@@ -0,0 +1,142 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2019, Vladimir Oltean <olteanv@gmail.com>
+ */
+#include <linux/etherdevice.h>
+#include <linux/if_vlan.h>
+#include <linux/dsa/sja1105.h>
+#include "../../drivers/net/dsa/sja1105/sja1105.h"
+
+#include "dsa_priv.h"
+
+/* Similar to is_link_local_ether_addr(hdr->h_dest) but also covers PTP */
+static inline bool sja1105_is_link_local(struct sk_buff *skb)
+{
+	struct ethhdr *hdr = eth_hdr(skb);
+	u64 dmac = ether_addr_to_u64(hdr->h_dest);
+
+	if ((dmac & SJA1105_LINKLOCAL_FILTER_A_MASK) ==
+		    SJA1105_LINKLOCAL_FILTER_A)
+		return true;
+	if ((dmac & SJA1105_LINKLOCAL_FILTER_B_MASK) ==
+		    SJA1105_LINKLOCAL_FILTER_B)
+		return true;
+	return false;
+}
+
+static struct sk_buff *sja1105_xmit(struct sk_buff *skb,
+				    struct net_device *netdev)
+{
+	struct dsa_port *dp = dsa_slave_to_port(netdev);
+	struct dsa_switch *ds = dp->ds;
+	struct sja1105_private *priv = ds->priv;
+	struct sja1105_port *sp = &priv->ports[dp->index];
+	struct sk_buff *clone;
+
+	if (likely(!sja1105_is_link_local(skb))) {
+		/* Normal traffic path. */
+		u16 tx_vid = dsa_tagging_tx_vid(ds, dp->index);
+		u8 pcp = skb->priority;
+
+		/* If we are under a vlan_filtering bridge, IP termination on
+		 * switch ports based on 802.1Q tags is simply too brittle to
+		 * be passable. So just defer to the dsa_slave_notag_xmit
+		 * implementation.
+		 */
+		if (dp->vlan_filtering)
+			return skb;
+
+		return dsa_8021q_xmit(skb, netdev, ETH_P_EDSA,
+				     ((pcp << VLAN_PRIO_SHIFT) | tx_vid));
+	}
+
+	/* Code path for transmitting management traffic. This does not rely
+	 * upon switch tagging, but instead SPI-installed management routes.
+	 */
+	clone = skb_clone(skb, GFP_ATOMIC);
+	if (!clone) {
+		dev_err(ds->dev, "xmit: failed to clone skb\n");
+		return NULL;
+	}
+
+	if (sja1105_skb_ring_add(&sp->xmit_ring, clone) < 0) {
+		dev_err(ds->dev, "xmit: skb ring full\n");
+		kfree_skb(clone);
+		return NULL;
+	}
+
+	if (sp->xmit_ring.count == SJA1105_SKB_RING_SIZE)
+		/* TODO setup a dedicated netdev queue for management traffic
+		 * so that we can selectively apply backpressure and not be
+		 * required to stop the entire traffic when the software skb
+		 * ring is full. This requires hooking the ndo_select_queue
+		 * from DSA and matching on mac_fltres.
+		 */
+		dev_err(ds->dev, "xmit: reached maximum skb ring size\n");
+
+	schedule_work(&sp->xmit_work);
+	/* Let DSA free its reference to the skb and we will free
+	 * the clone in the deferred worker
+	 */
+	return NULL;
+}
+
+static struct sk_buff *sja1105_rcv(struct sk_buff *skb,
+				   struct net_device *netdev,
+				   struct packet_type *pt)
+{
+	unsigned int source_port, switch_id;
+	struct ethhdr *hdr = eth_hdr(skb);
+	u16 tpid, vid, tci;
+
+	skb = dsa_8021q_rcv(skb, netdev, pt, &tpid, &tci);
+	if (!skb)
+		return NULL;
+
+	if (tpid != ETH_P_EDSA) {
+		netdev_warn(netdev, "TPID 0x%04x not for tagging\n", tpid);
+		return NULL;
+	}
+
+	skb->priority = (tci & VLAN_PRIO_MASK) >> VLAN_PRIO_SHIFT;
+	vid = tci & VLAN_VID_MASK;
+
+	skb->offload_fwd_mark = 1;
+
+	if (likely(!sja1105_is_link_local(skb))) {
+		/* Normal traffic path. */
+		source_port = dsa_tagging_rx_source_port(vid);
+		switch_id = dsa_tagging_rx_switch_id(vid);
+	} else {
+		/* Management traffic path. Switch embeds the switch ID and
+		 * port ID into bytes of the destination MAC, courtesy of
+		 * the incl_srcpt options.
+		 */
+		source_port = hdr->h_dest[3];
+		switch_id = hdr->h_dest[4];
+		/* Clear the DMAC bytes that were mangled by the switch */
+		hdr->h_dest[3] = 0;
+		hdr->h_dest[4] = 0;
+	}
+
+	skb->dev = dsa_master_find_slave(netdev, switch_id, source_port);
+	if (!skb->dev) {
+		netdev_warn(netdev, "Packet with invalid switch id %u and source port %u\n",
+			    switch_id, source_port);
+		return NULL;
+	}
+
+	/* Delete/overwrite fake VLAN header, DSA expects to not find
+	 * it there, see dsa_switch_rcv: skb_push(skb, ETH_HLEN).
+	 */
+	memmove(skb->data - ETH_HLEN, skb->data - ETH_HLEN - VLAN_HLEN,
+		ETH_HLEN - VLAN_HLEN);
+
+	return skb;
+}
+
+const struct dsa_device_ops sja1105_netdev_ops = {
+	.xmit = sja1105_xmit,
+	.rcv = sja1105_rcv,
+	.overhead = VLAN_HLEN,
+};
+
-- 
2.17.1

[PATCH net-next 15/17] net: dsa: sja1105: Add support for Spanning Tree Protocol

[Vladimir Oltean]

Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
---
 drivers/net/dsa/sja1105/sja1105_main.c | 108 ++++++++++++++++++++++---
 1 file changed, 99 insertions(+), 9 deletions(-)

diff --git a/drivers/net/dsa/sja1105/sja1105_main.c b/drivers/net/dsa/sja1105/sja1105_main.c
index adf79b1caf55..4064abc46f48 100644
--- a/drivers/net/dsa/sja1105/sja1105_main.c
+++ b/drivers/net/dsa/sja1105/sja1105_main.c
@@ -93,8 +93,10 @@ static int sja1105_init_mac_settings(struct sja1105_private *priv)
 		.drpuntag = false,
 		/* Don't retag 802.1p (VID 0) traffic with the pvid */
 		.retag = false,
-		/* Enable learning and I/O on user ports by default. */
-		.dyn_learn = true,
+		/* Disable learning and I/O on user ports by default -
+		 * STP will enable it.
+		 */
+		.dyn_learn = false,
 		.egress = false,
 		.ingress = false,
 		.mirrcie = 0,
@@ -125,8 +127,17 @@ static int sja1105_init_mac_settings(struct sja1105_private *priv)
 
 	mac = table->entries;
 
-	for (i = 0; i < SJA1105_NUM_PORTS; i++)
+	for (i = 0; i < SJA1105_NUM_PORTS; i++) {
 		mac[i] = default_mac;
+		if (i == dsa_upstream_port(priv->ds, i)) {
+			/* STP doesn't get called for CPU port, so we need to
+			 * set the I/O parameters statically.
+			 */
+			mac[i].dyn_learn = true;
+			mac[i].ingress = true;
+			mac[i].egress = true;
+		}
+	}
 
 	return 0;
 }
@@ -636,12 +647,14 @@ static sja1105_speed_t sja1105_get_speed_cfg(unsigned int speed_mbps)
  * for a specific port.
  *
  * @speed_mbps: If 0, leave the speed unchanged, else adapt MAC to PHY speed.
- * @enabled: Manage Rx and Tx settings for this port. Overrides the static
- *	     configuration settings.
+ * @enabled: Manage Rx and Tx settings for this port. If false, overrides the
+ *	     settings from the STP state, but not persistently (does not
+ *	     overwrite the static MAC info for this port).
  */
 static int sja1105_adjust_port_config(struct sja1105_private *priv, int port,
 				      int speed_mbps, bool enabled)
 {
+	struct sja1105_mac_config_entry dyn_mac;
 	struct sja1105_xmii_params_entry *mii;
 	struct sja1105_mac_config_entry *mac;
 	struct device *dev = priv->ds->dev;
@@ -674,12 +687,13 @@ static int sja1105_adjust_port_config(struct sja1105_private *priv, int port,
 	 * the code common, we'll use the static configuration tables as a
 	 * reasonable approximation for both E/T and P/Q/R/S.
 	 */
-	mac[port].ingress = enabled;
-	mac[port].egress  = enabled;
+	dyn_mac = mac[port];
+	dyn_mac.ingress = enabled && mac[port].ingress;
+	dyn_mac.egress  = enabled && mac[port].egress;
 
 	/* Write to the dynamic reconfiguration tables */
 	rc = sja1105_dynamic_config_write(priv, BLK_IDX_MAC_CONFIG,
-					  port, &mac[port], true);
+					  port, &dyn_mac, true);
 	if (rc < 0) {
 		dev_err(dev, "Failed to write MAC config: %d\n", rc);
 		return rc;
@@ -928,6 +942,50 @@ static int sja1105_bridge_member(struct dsa_switch *ds, int port,
 					    port, &l2_fwd[port], true);
 }
 
+static void sja1105_bridge_stp_state_set(struct dsa_switch *ds, int port,
+					 u8 state)
+{
+	struct sja1105_private *priv = ds->priv;
+	struct sja1105_mac_config_entry *mac;
+
+	mac = priv->static_config.tables[BLK_IDX_MAC_CONFIG].entries;
+
+	switch (state) {
+	case BR_STATE_DISABLED:
+	case BR_STATE_BLOCKING:
+		/* From UM10944 description of DRPDTAG (why put this there?):
+		 * "Management traffic flows to the port regardless of the state
+		 * of the INGRESS flag". So BPDUs are still be allowed to pass.
+		 * At the moment no difference between DISABLED and BLOCKING.
+		 */
+		mac[port].ingress   = false;
+		mac[port].egress    = false;
+		mac[port].dyn_learn = false;
+		break;
+	case BR_STATE_LISTENING:
+		mac[port].ingress   = true;
+		mac[port].egress    = false;
+		mac[port].dyn_learn = false;
+		break;
+	case BR_STATE_LEARNING:
+		mac[port].ingress   = true;
+		mac[port].egress    = false;
+		mac[port].dyn_learn = true;
+		break;
+	case BR_STATE_FORWARDING:
+		mac[port].ingress   = true;
+		mac[port].egress    = true;
+		mac[port].dyn_learn = true;
+		break;
+	default:
+		dev_err(ds->dev, "invalid STP state: %d\n", state);
+		return;
+	}
+
+	sja1105_dynamic_config_write(priv, BLK_IDX_MAC_CONFIG, port,
+				     &mac[port], true);
+}
+
 static int sja1105_bridge_join(struct dsa_switch *ds, int port,
 			       struct net_device *br)
 {
@@ -940,6 +998,23 @@ static void sja1105_bridge_leave(struct dsa_switch *ds, int port,
 	sja1105_bridge_member(ds, port, br, false);
 }
 
+static u8 sja1105_stp_state_get(struct sja1105_private *priv, int port)
+{
+	struct sja1105_mac_config_entry *mac;
+
+	mac = priv->static_config.tables[BLK_IDX_MAC_CONFIG].entries;
+
+	if (!mac[port].ingress && !mac[port].egress && !mac[port].dyn_learn)
+		return BR_STATE_BLOCKING;
+	if (mac[port].ingress && !mac[port].egress && !mac[port].dyn_learn)
+		return BR_STATE_LISTENING;
+	if (mac[port].ingress && !mac[port].egress && mac[port].dyn_learn)
+		return BR_STATE_LEARNING;
+	if (mac[port].ingress && mac[port].egress && mac[port].dyn_learn)
+		return BR_STATE_FORWARDING;
+	return -EINVAL;
+}
+
 /* For situations where we need to change a setting at runtime that is only
  * available through the static configuration, resetting the switch in order
  * to upload the new static config is unavoidable. Back up the settings we
@@ -950,16 +1025,27 @@ static int sja1105_static_config_reload(struct sja1105_private *priv)
 {
 	struct sja1105_mac_config_entry *mac;
 	int speed_mbps[SJA1105_NUM_PORTS];
+	u8 stp_state[SJA1105_NUM_PORTS];
 	int rc, i;
 
 	mac = priv->static_config.tables[BLK_IDX_MAC_CONFIG].entries;
 
 	/* Back up settings changed by sja1105_adjust_port_config and
-	 * and restore their defaults.
+	 * sja1105_bridge_stp_state_set and restore their defaults.
 	 */
 	for (i = 0; i < SJA1105_NUM_PORTS; i++) {
 		speed_mbps[i] = sja1105_speed[mac[i].speed];
 		mac[i].speed = SJA1105_SPEED_AUTO;
+		if (i == dsa_upstream_port(priv->ds, i)) {
+			mac[i].ingress = true;
+			mac[i].egress = true;
+			mac[i].dyn_learn = true;
+		} else {
+			stp_state[i] = sja1105_stp_state_get(priv, i);
+			mac[i].ingress = false;
+			mac[i].egress = false;
+			mac[i].dyn_learn = false;
+		}
 	}
 
 	/* Reset switch and send updated static configuration */
@@ -978,6 +1064,9 @@ static int sja1105_static_config_reload(struct sja1105_private *priv)
 	for (i = 0; i < SJA1105_NUM_PORTS; i++) {
 		bool enabled = (speed_mbps[i] != 0);
 
+		if (i != dsa_upstream_port(priv->ds, i))
+			sja1105_bridge_stp_state_set(priv->ds, i, stp_state[i]);
+
 		rc = sja1105_adjust_port_config(priv, i, speed_mbps[i],
 						enabled);
 		if (rc < 0)
@@ -1352,6 +1441,7 @@ static const struct dsa_switch_ops sja1105_switch_ops = {
 	.port_fdb_del		= sja1105_fdb_del,
 	.port_bridge_join	= sja1105_bridge_join,
 	.port_bridge_leave	= sja1105_bridge_leave,
+	.port_stp_state_set	= sja1105_bridge_stp_state_set,
 	.port_vlan_prepare	= sja1105_vlan_prepare,
 	.port_vlan_filtering	= sja1105_vlan_filtering,
 	.port_vlan_add		= sja1105_vlan_add,
-- 
2.17.1

[PATCH net-next 16/17] Documentation: networking: dsa: Add details about NXP SJA1105 driver

[Vladimir Oltean]

Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
---
 Documentation/networking/dsa/sja1105.txt | 83 ++++++++++++++++++++++++
 1 file changed, 83 insertions(+)
 create mode 100644 Documentation/networking/dsa/sja1105.txt

diff --git a/Documentation/networking/dsa/sja1105.txt b/Documentation/networking/dsa/sja1105.txt
new file mode 100644
index 000000000000..b6f2c1bedd02
--- /dev/null
+++ b/Documentation/networking/dsa/sja1105.txt
@@ -0,0 +1,83 @@
+NXP SJA1105 switch driver
+=========================
+
+The NXP SJA1105 is a family of 6 devices:
+* SJA1105E: First generation, no TTEthernet
+* SJA1105T: First generation, TTEthernet
+* SJA1105P: Second generation, no TTEthernet, no SGMII
+* SJA1105Q: Second generation, TTEthernet, no SGMII
+* SJA1105R: Second generation, no TTEthernet, SGMII
+* SJA1105S: Second generation, TTEthernet, SGMII
+
+These are SPI-managed automotive switches, with all ports being gigabit
+capable, and supporting MII/RMII/RGMII and optionally SGMII on one port.
+
+The switches do not have an MDIO bus of their own and do not support
+in-band autonegotiation, so for proper PHY management, the host's MDIO
+bus controller needs to be used.
+
+Being automotive parts, their configuration interface is geared towards
+set-and-forget use, with minimal dynamic interaction at runtime. They
+require a static configuration to be composed by software and packed
+with CRC and table headers, and sent over SPI.
+
+The static configuration is composed of several configuration tables. Each
+table takes a number of entries. Some configuration tables can be (partially)
+reconfigured at runtime, some not. Some tables are mandatory, some not.
+
+Table                        | Mandatory        | Reconfigurable
+-----------------------------+------------------+-----------------------------
+Schedule                     | no               | no
+Schedule entry points        | if Scheduling    | no
+VL Lookup                    | no               | no
+VL Policing                  | if VL Lookup     | no
+VL Forwarding                | if VL Lookup     | no
+L2 Lookup                    | no               | no
+L2 Policing                  | yes              | no
+VLAN Lookup                  | yes              | yes
+L2 Forwarding                | yes              | partially (fully on P/Q/R/S)
+MAC Config                   | yes              | partially (fully on P/Q/R/S)
+Schedule Params              | if Scheduling    | no
+Schedule Entry Points Params | if Scheduling    | no
+VL Forwarding Params         | if VL Forwarding | no
+L2 Lookup Params             | no               | partially (fully on P/Q/R/S)
+L2 Forwarding Params         | yes              | no
+Clock Sync Params            | no               | no
+AVB Params                   | no               | no
+General Params               | yes              | partially
+Retagging                    | no               | yes
+xMII Params                  | yes              | no
+SGMII                        | no               | yes
+
+Also the configuration is write-only (software cannot read it back from the
+switch except for very few exceptions).
+
+So the driver creates the static configuration at probe time, and keeps it at
+all times in memory, as a shadow for the hardware state. When required to
+change a hardware setting, the static configuration is also updated.
+If that changed setting can be transmitted to the switch through the dynamic
+reconfiguration interface, it is; otherwise the switch is reset and
+reprogrammed with the updated static configuration.
+
+The switches do not support switch tagging in hardware. But they do support
+customizing the TPID by which VLAN traffic is identified as such. The switch
+driver is leveraging CONFIG_NET_DSA_TAG_8021Q by requesting that special VLANs
+(with a custom TPID of ETH_P_EDSA instead of ETH_P_8021Q) are installed on its
+ports when not in vlan_filtering mode. This does not interfere with the
+reception and transmission of real 802.1Q-tagged traffic, because the switch
+does no longer parse those packets as VLAN after the TPID change.
+The TPID is restored when vlan_filtering is requested, and IP termination
+becomes no longer possible through the switch netdevices in this mode.
+
+The switches have two programmable filters for link-local destination MACs.
+These are used to trap BPDUs and PTP traffic to the master netdevice, and are
+further used to support STP and 1588 ordinary clock/boundary clock
+functionality.
+
+Among other notable features, the switches have a PTP Hardware Clock that can
+be steered through SPI and used for timestamping on ingress and egress.
+Also, the T, Q and S devices support TTEthernet (an implementation of
+SAE AS6802 from TTTech), which is a set of Ethernet QoS enhancements similar in
+behavior to IEEE TSN. Configuring these features is currently not supported in
+the driver.
+
-- 
2.17.1

Re: [PATCH net-next 16/17] Documentation: networking: dsa: Add details about NXP SJA1105 driver

[Andrew Lunn]

> +The switches do not support switch tagging in hardware. But they do support
> +customizing the TPID by which VLAN traffic is identified as such. The switch
> +driver is leveraging CONFIG_NET_DSA_TAG_8021Q by requesting that special VLANs
> +(with a custom TPID of ETH_P_EDSA instead of ETH_P_8021Q)

Hi Vladimir

Reusing ETH_P_EDSA is possibility going to case problems in tcpdump.
It has code which looks for this Ethertype and uses it to call the
Marvell EDSA header decoder code. This is not going to work for your
switch, where you have a normal VLAN header.

I wounder if we should allocate a new EtherType for this application
of VLANs?

   Andrew

Re: [PATCH net-next 16/17] Documentation: networking: dsa: Add details about NXP SJA1105 driver

[Vladimir Oltean]

On 4/3/19 12:24 AM, Andrew Lunn wrote:
>> +The switches do not support switch tagging in hardware. But they do support
>> +customizing the TPID by which VLAN traffic is identified as such. The switch
>> +driver is leveraging CONFIG_NET_DSA_TAG_8021Q by requesting that special VLANs
>> +(with a custom TPID of ETH_P_EDSA instead of ETH_P_8021Q)
> 
> Hi Vladimir
> 
> Reusing ETH_P_EDSA is possibility going to case problems in tcpdump.
> It has code which looks for this Ethertype and uses it to call the
> Marvell EDSA header decoder code. This is not going to work for your
> switch, where you have a normal VLAN header.
> 
> I wounder if we should allocate a new EtherType for this application
> of VLANs?
> 
>     Andrew
> 

Hi Andrew,

This makes me wonder whether I would be able to use an EtherType range 
to support cascading switch tags on SJA1105.
In theory if I were to consistently lie to all switches within a switch 
tree, I could make all of them see each other's fake VLAN-tagged traffic 
as untagged.
By the way what is the simple ETH_P_DSA used for?

Thank you,
-Vladimir

Re: [PATCH net-next 16/17] Documentation: networking: dsa: Add details about NXP SJA1105 driver

[Florian Fainelli]

On 4/3/2019 3:09 AM, Vladimir Oltean wrote:
> On 4/3/19 12:24 AM, Andrew Lunn wrote:
>>> +The switches do not support switch tagging in hardware. But they do
>>> support
>>> +customizing the TPID by which VLAN traffic is identified as such.
>>> The switch
>>> +driver is leveraging CONFIG_NET_DSA_TAG_8021Q by requesting that
>>> special VLANs
>>> +(with a custom TPID of ETH_P_EDSA instead of ETH_P_8021Q)
>>
>> Hi Vladimir
>>
>> Reusing ETH_P_EDSA is possibility going to case problems in tcpdump.
>> It has code which looks for this Ethertype and uses it to call the
>> Marvell EDSA header decoder code. This is not going to work for your
>> switch, where you have a normal VLAN header.
>>
>> I wounder if we should allocate a new EtherType for this application
>> of VLANs?
>>
>>     Andrew
>>
> 
> Hi Andrew,
> 
> This makes me wonder whether I would be able to use an EtherType range
> to support cascading switch tags on SJA1105.
> In theory if I were to consistently lie to all switches within a switch
> tree, I could make all of them see each other's fake VLAN-tagged traffic
> as untagged.
> By the way what is the simple ETH_P_DSA used for?

ETH_P_DSA is not used since commit
3e8a72d1dae374cf6fc1dba97cec663585845ff9 ("net: dsa: reduce number of
protocol hooks") but since it is in an UAPI exported file we probably
cannot remove it.
-- 
Florian

[PATCH net-next 17/17] dt-bindings: net: dsa: Add documentation for NXP SJA1105 driver

[Vladimir Oltean]

Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
---
 .../devicetree/bindings/net/dsa/sja1105.txt   | 123 ++++++++++++++++++
 1 file changed, 123 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/net/dsa/sja1105.txt

diff --git a/Documentation/devicetree/bindings/net/dsa/sja1105.txt b/Documentation/devicetree/bindings/net/dsa/sja1105.txt
new file mode 100644
index 000000000000..2c82b6fc37e3
--- /dev/null
+++ b/Documentation/devicetree/bindings/net/dsa/sja1105.txt
@@ -0,0 +1,123 @@
+NXP SJA1105 switch driver
+=========================
+
+Required properties:
+
+- compatible: Must be "nxp,sja1105". Device ID identification (one of
+  E/T/P/Q/R/S) is performed by driver at probe time. Swapping pin-compatible
+  parts is possible with no DTS change.
+
+Optional properties:
+
+- sja1105,mac-mode, sja1105,phy-mode: Boolean properties that can be assigned
+  under each port node that is MII or RMII (has no effect for RGMII).  By
+  default (unless otherwise specified) a port is configured as MAC if it is
+  driving a PHY (phy-handle is present) or as PHY if it is PHY-less (fixed-link
+  specified, presumably because it is connected to a MAC).  These properties
+  are required in the case where SJA1105 ports are at both ends of an MII/RMII
+  PHY-less setup. One end would need to have sja1105,mac-mode, while the other
+  sja1105,phy-mode.
+
+See Documentation/devicetree/bindings/net/dsa/dsa.txt for the list of standard
+DSA required and optional properties.
+
+Other observations:
+
+The SJA1105 SPI interface requires a CS-to-CLK time (t2 in UM10944) of at least
+one half of t_CLK. At an SPI frequency of 1MHz, this means a minimum
+cs_sck_delay of 500ns. Ensuring that this SPI timing requirement is observed
+depends on the SPI bus master driver.
+
+Example:
+
+Ethernet switch connected via SPI to the host, CPU port wired to eth0:
+
+arch/arm/boot/dts/ls1021a-tsn.dts:
+
+/* SPI controller of the LS1021 */
+&dspi0 {
+	sja1105@1 {
+		reg = <0x1>;
+		#address-cells = <1>;
+		#size-cells = <0>;
+		compatible = "nxp,sja1105";
+		spi-max-frequency = <4000000>;
+		fsl,spi-cs-sck-delay = <1000>;
+		fsl,spi-sck-cs-delay = <1000>;
+		ports {
+			#address-cells = <1>;
+			#size-cells = <0>;
+			port@0 {
+				/* ETH5 written on chassis */
+				label = "swp5";
+				phy-handle = <&rgmii_phy6>;
+				phy-mode = "rgmii";
+				reg = <0>;
+				/* Implicit "sja1105,mac-mode;" */
+			};
+			port@1 {
+				/* ETH2 written on chassis */
+				label = "swp2";
+				phy-handle = <&rgmii_phy3>;
+				phy-mode = "rgmii";
+				reg = <1>;
+				/* Implicit "sja1105,mac-mode;" */
+			};
+			port@2 {
+				/* ETH3 written on chassis */
+				label = "swp3";
+				phy-handle = <&rgmii_phy4>;
+				phy-mode = "rgmii";
+				reg = <2>;
+				/* Implicit "sja1105,mac-mode;" */
+			};
+			port@3 {
+				/* ETH4 written on chassis */
+				phy-handle = <&rgmii_phy5>;
+				label = "swp4";
+				phy-mode = "rgmii";
+				reg = <3>;
+				/* Implicit "sja1105,mac-mode;" */
+			};
+			port@4 {
+				/* Internal port connected to eth2 */
+				ethernet = <&enet2>;
+				phy-mode = "rgmii";
+				reg = <4>;
+				/* Implicit "sja1105,phy-mode;" */
+				fixed-link {
+					speed = <1000>;
+					full-duplex;
+				};
+			};
+		};
+	};
+};
+
+/* MDIO controller of the LS1021 */
+&mdio0 {
+	/* BCM5464 */
+	rgmii_phy3: ethernet-phy@3 {
+		reg = <0x3>;
+	};
+	rgmii_phy4: ethernet-phy@4 {
+		reg = <0x4>;
+	};
+	rgmii_phy5: ethernet-phy@5 {
+		reg = <0x5>;
+	};
+	rgmii_phy6: ethernet-phy@6 {
+		reg = <0x6>;
+	};
+};
+
+/* Ethernet master port of the LS1021 */
+&enet2 {
+	phy-connection-type = "rgmii";
+	status = "ok";
+	fixed-link {
+		speed = <1000>;
+		full-duplex;
+	};
+};
+
-- 
2.17.1

Re: [PATCH net-next 17/17] dt-bindings: net: dsa: Add documentation for NXP SJA1105 driver

[Andrew Lunn]

> +Optional properties:
> +
> +- sja1105,mac-mode, sja1105,phy-mode: Boolean properties that can be assigned
> +  under each port node that is MII or RMII (has no effect for RGMII).  By
> +  default (unless otherwise specified) a port is configured as MAC if it is
> +  driving a PHY (phy-handle is present) or as PHY if it is PHY-less (fixed-link
> +  specified, presumably because it is connected to a MAC).  These properties
> +  are required in the case where SJA1105 ports are at both ends of an MII/RMII
> +  PHY-less setup. One end would need to have sja1105,mac-mode, while the other
> +  sja1105,phy-mode.

Hi Vladimir

phy-mode has a well known meaning, indicating mii, gmii, sqmii, rmii,
rxaud, etc.

The meaning here is quite different. To maybe avoid confusion, could
we flip the name around?

sja1105,mode-mac and sja1105,mode-phy?

> +			port@4 {
> +				/* Internal port connected to eth2 */
> +				ethernet = <&enet2>;
> +				phy-mode = "rgmii";
> +				reg = <4>;
> +				/* Implicit "sja1105,phy-mode;" */
> +				fixed-link {
> +					speed = <1000>;
> +					full-duplex;
> +				};
> +			};
> +		};
> +	};
> +};


> +&enet2 {
> +	phy-connection-type = "rgmii";

You don't see this used much, phy-mode is preferred. Do you have a
reason for this?

Also, you have the MAC using RGMII and the port using RGMII. Neither
is inserting delays. That implies the delays are added by the track
layout of the PCB. It would be good to add a comment about
this. Anybody copying this using a design without the delays via PCB
are probably going to get it wrong to start with and not realise they
need to change one end to RGMII-ID.

Thanks
     Andrew

Re: [PATCH net-next 17/17] dt-bindings: net: dsa: Add documentation for NXP SJA1105 driver

[Vladimir Oltean]

On 4/3/19 12:38 AM, Andrew Lunn wrote:
>> +Optional properties:
>> +
>> +- sja1105,mac-mode, sja1105,phy-mode: Boolean properties that can be assigned
>> +  under each port node that is MII or RMII (has no effect for RGMII).  By
>> +  default (unless otherwise specified) a port is configured as MAC if it is
>> +  driving a PHY (phy-handle is present) or as PHY if it is PHY-less (fixed-link
>> +  specified, presumably because it is connected to a MAC).  These properties
>> +  are required in the case where SJA1105 ports are at both ends of an MII/RMII
>> +  PHY-less setup. One end would need to have sja1105,mac-mode, while the other
>> +  sja1105,phy-mode.
> 
> Hi Vladimir
> 
> phy-mode has a well known meaning, indicating mii, gmii, sqmii, rmii,
> rxaud, etc.
> 
> The meaning here is quite different. To maybe avoid confusion, could
> we flip the name around?
> 
> sja1105,mode-mac and sja1105,mode-phy?
> 

Hi Andrew,

I agree the clash of words is confusing.
Which one of "sja1105,mode-phy", "sja1105,type-phy", "sja1105,role-phy" 
sounds easier to digest?

>> +			port@4 {
>> +				/* Internal port connected to eth2 */
>> +				ethernet = <&enet2>;
>> +				phy-mode = "rgmii";
>> +				reg = <4>;
>> +				/* Implicit "sja1105,phy-mode;" */
>> +				fixed-link {
>> +					speed = <1000>;
>> +					full-duplex;
>> +				};
>> +			};
>> +		};
>> +	};
>> +};
> 
> 
>> +&enet2 {
>> +	phy-connection-type = "rgmii";
> 
> You don't see this used much, phy-mode is preferred. Do you have a
> reason for this?

No particular reason for using phy-connection-type. It is just a 
copy-pasta from the (not yet submitted) ls1021a-tsn.dts that I'm using 
to test the driver on.
The other LS1021A DTS files were already using this notation, so that's 
how it got there. But the gianfar driver works just as well with 
phy-mode, so maybe I will send another patch for converting the existing 
notations when I submit the new board DTS too.

> 
> Also, you have the MAC using RGMII and the port using RGMII. Neither
> is inserting delays. That implies the delays are added by the track
> layout of the PCB. It would be good to add a comment about
> this. Anybody copying this using a design without the delays via PCB
> are probably going to get it wrong to start with and not realise they
> need to change one end to RGMII-ID.
> 

You've hit the nail on the head with this excellent comment.
So the second-generation switches (P/Q/R/S) have addressed this hardware 
limitation and do provide some tunable delay lines for RGMII.
But on the LS1021A-TSN, which is using a first-generation T chip, the 
RGMII delays for correct sample/hold times are obtained through ~50cm 
copper wires, since the LS1021 cannot apply them internally either.

> https://www.nxp.com/docs/en/data-sheet/SJA1105.pdf
> 
> Section 6.2.3 RGMII signaling and encoding
> 
> Note that RGMII requires an external delay of between 1.5 ns and 2 ns
> on TXC and RXC.
> 
> So it sounds like the switch only supports PHY_INTERFACE_MODE_RGMII.
> 
> If the port is in MAC mode, you should pass this phy-mode to the PHY
> when you connect to it. The PHY can then add the delay if needed.
> 
> If however, the port is in PHY mode, and it is asked to do RGMII other
> than PHY_INTERFACE_MODE_RGMII, you should report an error. It cannot
> do it.

I expect that next week I will be able to get a reworked LS1021A-TSN 
board with a second-generation switch, and the delay wires removed.
Using that, I can at least add support for the tunable delay lines in a 
fashion similar to what you're suggesting.

But since you brought this up, let me point out some complications I see:
* I think the way this works is that phy_attach_direct populates 
phydev->interface based on the MAC's DT bindings. Then the PHY driver is 
supposed to apply internal delays based on that. But this shadows the 
MAC's own ability to add internal delays based on the same DT bindings. 
You told me to pass the internal delay settings to the PHY if 
"sja1105,mode-mac" is (implicitly or explicitly) specified, or otherwise 
("sja1105,mode-phy"), take the internal delay settings and apply them 
myself (if the switch revision supports this). But in the latter case, 
should I mask off the RGMII_*ID modes and convert everything to RGMII 
when doing the of_phy_connect (to ensure that at the other end nobody 
will add the internal delays twice)? I know the theory but I have never 
actually seen a MAC driver apply the internal delay settings. And even 
in my case I'm a bit surprised that there isn't a more generic mechanism 
to specify this and that I have to rely on custom DT bindings.
* The 2 groups of devices (E, T, P, Q) and (R, S) are pin-compatible 
with one another, and in principle hot-swappable. Initially I had 
explicit "compatible" properties for each device model, but with the 
code for device_id detection, this became a bit redundant so I just made 
everything "nxp,sja1105". But with the different RGMII internal delay 
capabilities, and later on with SGMII support (R, S), some differences 
at the DT level will become apparent. For example my T -> Q swap on the 
LS1021A-TSN board will require a DT change for the internal delay 
settings. How should I handle the "compatible" property for this driver?

> Thanks
>       Andrew
> 

Thank you,
-Vladimir