Skip to contentSkip to conversation

MIPS: Ingenic: Add XBurst / XBurst II infrastructure and improve overall Ingenic SoC support

PATCH In review H. Nikolaus Schaller 4 participants 23 messages Nov 12 – Nov 16

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

v2 cover MIPS: Ingenic: Add XBurst / XBurst II infrastructure and improve overall Ingenic SoC support

PATCH V2 2026-11-12 09:10:31:
The series updates the Ingenic platform support to allow clean transitions between bootloader-provided Device Trees and built-in fallbacks. It provides the foundational Kconfig, PRID, and CPU-probe blocks required for newer Ingenic/Lumissil hardware generations (X1600, X2000, X2600) and adds JZ4730 for completeness.

Changes to V1
- allow user selection of BUILTIN_DTB (suggested by Paul Cercueil <[email protected]>)
- added the CPU_XBURTS2 type here to solve an issue reported by kernel test robot
- removed PRID_IMP_XBURST2_R2 and merged into single commit
- reviewed all PRID_IMP with data sheets and programming manuals
- added full x2600 and jz4730 definition for completeness
- added platform makefile and load address for X1600, X2000, X2600

PATCH V1 2026-10-25 20:20:55:
The series updates the Ingenic platform support to allow clean transitions between bootloader-provided Device Trees and built-in fallbacks. It provides the foundational Kconfig, PRID, and CPU-probe blocks required for newer Ingenic/Lumissil hardware generations (X1600, X2000, X2600).

The first patch corrects an inverted Kconfig condition regarding appended and built-in DTBs. This fix is necessary to allow configurations where the kernel relies entirely on a Device Tree provided dynamically by the bootloader instead of forcing a built-in fallback.

The remaining patches implement CPU-probe initialization, correct PRID mappings (including vendor-kernel values for the Ingenic A1 SoC), platform load addresses, Kconfig updates, and generic board match updates.

Structure of the series:
1. MIPS: Ingenic: Fix inverted MIPS_NO_APPENDED_DTB logic in Kconfig
2. MIPS: Ingenic: Add and update PRIDs for XBurst II R0 and R2
3. MIPS: Ingenic: Expand Kconfig for X1600, X2000, and X2600 SoCs
4. MIPS: Ingenic: Add cpu-probe setup for XBurst and XBurst II R2
5. MIPS: generic: Add MACH_INGENIC type and compatible for X1600
6. MIPS: Ingenic: Add Platform makefile for X1600 and X2000

Testing:
The series has been compile-tested for generic Ingenic targets. PRID values for jz4780, x1600, X2000, X2600 were verified on real hardware (CI20, LX16, LX20 and LX26 boards respectively).

Please review and consider for inclusion.

H. Nikolaus Schaller (8):
MIPS: Kconfig: Allow user selection of BUILTIN_DTB with description
MIPS: Ingenic: Configure CONFIG_BUILTIN_DTB=y for Ingenic based
devices
MIPS: Ingenic: Do not force BUILTIN_DTB in Kconfig
MIPS: Ingenic: Update Ingenic XBurst implementation IDs
MIPS: generic: Add MACH_INGENIC type and compatible for X1600 and

    X2600
  MIPS: Ingenic: Expand Kconfig for X1600, X2000, and X2600 SoCs
  MIPS: Ingenic: Expand Kconfig by MACH_JZ4730
  MIPS: Ingenic: Add Platform makefile and load addresses for X1600,
    X2000, X2600

周琰杰 (Zhou Yanjie) (1):
MIPS: Ingenic: Add a new CPU_XBURST2 type

 arch/mips/Kconfig                      | 25 +++++++++++++--
 arch/mips/configs/ci20_defconfig       |  1 +
 arch/mips/configs/cu1000-neo_defconfig |  1 +
 arch/mips/configs/cu1830-neo_defconfig |  1 +
 arch/mips/configs/gcw0_defconfig       |  1 +
 arch/mips/configs/qi_lb60_defconfig    |  1 +
 arch/mips/configs/rs90_defconfig       |  1 +
 arch/mips/generic/board-ingenic.c      |  8 ++++-
 arch/mips/include/asm/bootinfo.h       |  2 ++
 arch/mips/include/asm/cpu-type.h       |  1 +
 arch/mips/include/asm/cpu.h            | 13 ++++----
 arch/mips/ingenic/Kconfig              | 43 ++++++++++++++++++++++++++
 arch/mips/ingenic/Platform             |  3 ++
 arch/mips/kernel/cpu-probe.c           |  7 +++++
 14 files changed, 99 insertions(+), 9 deletions(-)
 create mode 100644 arch/mips/ingenic/Platform
Signature · 1 line
2.56.0
#1 lore raw

PATCH v2 1/9 MIPS: Kconfig: Allow user selection of BUILTIN_DTB with description

Make CONFIG_BUILTIN_DTB visible in Kconfig by adding a prompt and a detailed help text.

Previously, BUILTIN_DTB was a silent option forcefully selected by certain machine types (such as MACH_INGENIC) whenever appended DTB support was disabled. This prevented modern setups from turning off the built-in DTB fallback when the bootloader is fully capable of passing a dynamic DTB.

By adding a prompt, users can now explicitly toggle this behavior. The added help text clarifies the distinction between MIPS_NO_APPENDED_DTB (which controls image appending) and BUILTIN_DTB (which controls the compiled-in fallback data and code path via get_fdt()), as well as its relationship to BUILTIN_DTB_NAME.

---

Notes:
Help text and commit message generated by public Google AI.

 arch/mips/Kconfig | 24 +++++++++++++++++++++++-
 1 file changed, 23 insertions(+), 1 deletion(-)
arch/mips/Kconfig +23 -1
index dac3ec5120a34..92bca520cd5ae 100644
@@ -2924,7 +2924,29 @@ config UHI_BOOT
 	bool
 
 config BUILTIN_DTB
-	bool
+	bool "Enable builtin DTB fallback support"
+	depends on USE_OF
+	help
+	  Compiles a Device Tree Blob (DTB) directly into the kernel image
+	  and enables the kernel to use it as a fallback during boot.
+
+	  This option:
+	  a) Links the DTB specified by BUILTIN_DTB_NAME into the
+	     kernel binary.
+	  b) Instructs the boot code (get_fdt()) to return this internal
+	     DTB if no dynamic DTB from the bootloader or appended DTB
+	     image data is detected.
+
+	  Note on MIPS boot strategy:
+	  MIPS kernels check for a DTB in the following order inside get_fdt():
+	  1. An appended DTB image (if CONFIG_MIPS_*_APPENDED_DTB is set).
+	  2. The built-in DTB fallback controlled by this option.
+
+	  If your bootloader provides a valid DTB dynamically and you do not
+	  rely on appended or built-in DTBs, you should enable
+	  MIPS_NO_APPENDED_DTB and disable BUILTIN_DTB.
+
+	  If unsure, say Y.
 
 choice
 	prompt "Kernel appended dtb support"
Signature · 1 line
2.56.0
#2 lore raw

PATCH v2 2/9 MIPS: Ingenic: Configure CONFIG_BUILTIN_DTB=y for Ingenic based devices

In reply to H. Nikolaus Schaller

This enables the new CONFIG_BUILTIN_DTB user variable so that we do not need to rely on a select CONFIG_BUILTIN_DTB if ..." any more

---
 arch/mips/configs/ci20_defconfig       | 1 +
 arch/mips/configs/cu1000-neo_defconfig | 1 +
 arch/mips/configs/cu1830-neo_defconfig | 1 +
 arch/mips/configs/gcw0_defconfig       | 1 +
 arch/mips/configs/qi_lb60_defconfig    | 1 +
 arch/mips/configs/rs90_defconfig       | 1 +
 6 files changed, 6 insertions(+)
arch/mips/configs/ci20_defconfig +1 -0
index 90536cab417ee..b45a6c7b54ea8 100644
@@ -23,6 +23,7 @@ CONFIG_MACH_INGENIC_SOC=y
 CONFIG_JZ4780_CI20=y
 CONFIG_HIGHMEM=y
 CONFIG_HZ_100=y
+CONFIG_BUILTIN_DTB=y
 # CONFIG_SUSPEND is not set
 # CONFIG_SECCOMP is not set
 CONFIG_MODULES=y
arch/mips/configs/cu1000-neo_defconfig +1 -0
index ed2a620e4f86a..a505de6e9e367 100644
@@ -22,6 +22,7 @@ CONFIG_MACH_INGENIC_SOC=y
 CONFIG_X1000_CU1000_NEO=y
 CONFIG_HIGHMEM=y
 CONFIG_HZ_100=y
+CONFIG_BUILTIN_DTB=y
 # CONFIG_SECCOMP is not set
 # CONFIG_SUSPEND is not set
 CONFIG_MODULES=y
arch/mips/configs/cu1830-neo_defconfig +1 -0
index 9b4aeeef58ceb..1fa7a49543063 100644
@@ -22,6 +22,7 @@ CONFIG_MACH_INGENIC_SOC=y
 CONFIG_X1830_CU1830_NEO=y
 CONFIG_HIGHMEM=y
 CONFIG_HZ_100=y
+CONFIG_BUILTIN_DTB=y
 # CONFIG_SECCOMP is not set
 # CONFIG_SUSPEND is not set
 CONFIG_MODULES=y
arch/mips/configs/gcw0_defconfig +1 -0
index 6d737810b470f..36ef053376185 100644
@@ -7,6 +7,7 @@ CONFIG_PROFILING=y
 CONFIG_MACH_INGENIC_SOC=y
 CONFIG_JZ4770_GCW0=y
 CONFIG_HIGHMEM=y
+CONFIG_BUILTIN_DTB=y
 # CONFIG_SECCOMP is not set
 CONFIG_MIPS_RAW_APPENDED_DTB=y
 CONFIG_MIPS_CMDLINE_DTB_EXTEND=y
arch/mips/configs/qi_lb60_defconfig +1 -0
index a1bb0792f6eb1..18fe70487817c 100644
@@ -9,6 +9,7 @@ CONFIG_EXPERT=y
 CONFIG_MACH_INGENIC_SOC=y
 CONFIG_JZ4740_QI_LB60=y
 CONFIG_HZ_100=y
+CONFIG_BUILTIN_DTB=y
 # CONFIG_SECCOMP is not set
 CONFIG_MODULES=y
 CONFIG_MODULE_UNLOAD=y
arch/mips/configs/rs90_defconfig +1 -0
index 8382d535e6dc1..0c57d829d1720 100644
@@ -22,6 +22,7 @@ CONFIG_MACH_INGENIC_SOC=y
 CONFIG_JZ4740_RS90=y
 CONFIG_PAGE_SIZE_16KB=y
 CONFIG_HZ_100=y
+CONFIG_BUILTIN_DTB=y
 # CONFIG_SECCOMP is not set
 CONFIG_MIPS_CMDLINE_DTB_EXTEND=y
 # CONFIG_SUSPEND is not set
Signature · 1 line
2.56.0
#3 lore raw

PATCH v2 3/9 MIPS: Ingenic: Do not force BUILTIN_DTB in Kconfig

In reply to H. Nikolaus Schaller

Remove the unconditional selection of BUILTIN_DTB for MACH_INGENIC when MIPS_NO_APPENDED_DTB is enabled.

The current Kconfig logic selects BUILTIN_DTB when MIPS_NO_APPENDED_DTB is set, although this configuration is intended to prevent the kernel from appending a DTB to the kernel image.

While the kernel gives precedence to a DTB supplied by the boot loader, selecting BUILTIN_DTB still triggers additional build steps and may add unnecessary code to the resulting uImage.

Note that BUILTIN_DTB controls the build process, not necessarily whether a DTB is actually embedded in the kernel image. For example, CONFIG_BUILTIN_DTB=y with CONFIG_BUILTIN_DTB_NAME="" still triggers the generation of .builtin-dtbs.S, but does not embed a DTB.

The goal is to allow a configuration in which no .builtin-dtbs.S is generated and no DTB is appended to or built into the zImage/uImage.

Fixes f0f4a753079c ("MIPS: generic: Add support for Ingenic SoCs")
Fixes 15205fc05482 ("MIPS: ingenic: Add support for appended devicetree")
---
 arch/mips/Kconfig | 1 -
 1 file changed, 1 deletion(-)
arch/mips/Kconfig +0 -1
index 92bca520cd5ae..2135a69f9962c 100644
@@ -127,7 +127,6 @@ config MACH_INGENIC
 	select GPIOLIB
 	select COMMON_CLK
 	select GENERIC_IRQ_CHIP
-	select BUILTIN_DTB if MIPS_NO_APPENDED_DTB
 	select USE_OF
 	select CPU_SUPPORTS_CPUFREQ
 	select MIPS_EXTERNAL_TIMER
Signature · 1 line
2.56.0
maintainer
#4 lore raw

Re: [PATCH v2 3/9] MIPS: Ingenic: Do not force BUILTIN_DTB in Kconfig

Hi Nikolaus,

Le jeudi 12 novembre 2026 à 09:10 +0100, H. Nikolaus Schaller a écrit :

Quoted 16 lines

Remove the unconditional selection of BUILTIN_DTB for MACH_INGENIC when MIPS_NO_APPENDED_DTB is enabled.

The current Kconfig logic selects BUILTIN_DTB when MIPS_NO_APPENDED_DTB is set, although this configuration is intended to prevent the kernel from appending a DTB to the kernel image.

While the kernel gives precedence to a DTB supplied by the boot loader, selecting BUILTIN_DTB still triggers additional build steps and may add unnecessary code to the resulting uImage.

Note that BUILTIN_DTB controls the build process, not necessarily whether a DTB is actually embedded in the kernel image. For example, CONFIG_BUILTIN_DTB=y with CONFIG_BUILTIN_DTB_NAME="" still triggers the generation of .builtin-dtbs.S, but does not embed a DTB.

Quoted 5 lines

The goal is to allow a configuration in which no .builtin-dtbs.S is generated and no DTB is appended to or built into the zImage/uImage.

Fixes f0f4a753079c ("MIPS: generic: Add support for Ingenic SoCs")
Fixes 15205fc05482 ("MIPS: ingenic: Add support for appended devicetree")

I don't think you can have two Fixes: tags...

Cheers,
-Paul

Quoted 17 lines

---
 arch/mips/Kconfig | 1 -

 1 file changed, 1 deletion(-)
arch/mips/Kconfig +0 -0
index 92bca520cd5ae..2135a69f9962c 100644
@@ -127,7 +127,6 @@ config MACH_INGENIC

  select GPIOLIB
  select COMMON_CLK
  select GENERIC_IRQ_CHIP

arch/mips/Kconfig +0 -1
-	select BUILTIN_DTB if MIPS_NO_APPENDED_DTB

  select USE_OF
  select CPU_SUPPORTS_CPUFREQ
  select MIPS_EXTERNAL_TIMER

#5 lore raw

Re: [PATCH v2 3/9] MIPS: Ingenic: Do not force BUILTIN_DTB in Kconfig

Quoted 23 lines

Am 12.11.2026 um 14:57 schrieb Paul Cercueil <[email protected]>:

Hi Nikolaus,

Quoted 16 lines · H. Nikolaus Schaller

Remove the unconditional selection of BUILTIN_DTB for MACH_INGENIC when MIPS_NO_APPENDED_DTB is enabled.

The current Kconfig logic selects BUILTIN_DTB when MIPS_NO_APPENDED_DTB is set, although this configuration is intended to prevent the kernel from appending a DTB to the kernel image.

While the kernel gives precedence to a DTB supplied by the boot loader, selecting BUILTIN_DTB still triggers additional build steps and may add unnecessary code to the resulting uImage.

Note that BUILTIN_DTB controls the build process, not necessarily whether a DTB is actually embedded in the kernel image. For example, CONFIG_BUILTIN_DTB=y with CONFIG_BUILTIN_DTB_NAME="" still triggers the generation of .builtin-dtbs.S, but does not embed a DTB.

Quoted 7 lines
Quoted 5 lines

The goal is to allow a configuration in which no .builtin-dtbs.S is generated and no DTB is appended to or built into the zImage/uImage.

Fixes f0f4a753079c ("MIPS: generic: Add support for Ingenic SoCs")
Fixes 15205fc05482 ("MIPS: ingenic: Add support for appended devicetree")

I don't think you can have two Fixes: tags...

Oops. Must be the result of mechanically squashing commits...

Will be fixed for V3.

Quoted 3 lines

Cheers,
-Paul

Same to you :)
Nikolaus

Quoted 18 lines
Quoted 18 lines
---
 arch/mips/Kconfig | 1 -
 1 file changed, 1 deletion(-)
arch/mips/Kconfig +0 -1
index 92bca520cd5ae..2135a69f9962c 100644
@@ -127,7 +127,6 @@ config MACH_INGENIC
  select GPIOLIB
  select COMMON_CLK
  select GENERIC_IRQ_CHIP
- select BUILTIN_DTB if MIPS_NO_APPENDED_DTB
  select USE_OF
  select CPU_SUPPORTS_CPUFREQ
  select MIPS_EXTERNAL_TIMER
#6 lore raw

PATCH v2 4/9 MIPS: Ingenic: Add a new CPU_XBURST2 type

In reply to H. Nikolaus Schaller

From: 周琰杰 (Zhou Yanjie) <[email protected]>

Add a new CPU_XBURST2 type for the MIPS32r5 compatible XBurst®2 CPU.

---

Notes:
This patch solves a build error in

"MIPS: Ingenic: Update Ingenic XBurst implementation IDs"

reported by kernel test robot.

 arch/mips/include/asm/cpu-type.h | 1 +
 arch/mips/include/asm/cpu.h      | 4 ++--
 2 files changed, 3 insertions(+), 2 deletions(-)
arch/mips/include/asm/cpu-type.h +1 -0
index fd37a44a2f199..25a181fd0eb7a 100644
@@ -56,6 +56,7 @@ static inline int __pure __get_cpu_type(const int cpu_type)
 #ifdef CONFIG_SYS_HAS_CPU_MIPS32_R5
 	case CPU_M5150:
 	case CPU_P5600:
+	case CPU_XBURST2:
 #endif
 
 #if defined(CONFIG_SYS_HAS_CPU_MIPS32_R2) || \
arch/mips/include/asm/cpu.h +2 -2
index 0fd9f9bbd21f1..add26f08e8029 100644
@@ -312,8 +312,8 @@ enum cpu_type_enum {
 	 */
 	CPU_4KC, CPU_4KEC, CPU_4KSC, CPU_24K, CPU_34K, CPU_1004K, CPU_74K,
 	CPU_ALCHEMY, CPU_PR4450, CPU_BMIPS32, CPU_BMIPS3300, CPU_BMIPS4350,
-	CPU_BMIPS4380, CPU_BMIPS5000, CPU_XBURST, CPU_LOONGSON32, CPU_M14KC,
-	CPU_M14KEC, CPU_INTERAPTIV, CPU_P5600, CPU_PROAPTIV, CPU_1074K,
+	CPU_BMIPS4380, CPU_BMIPS5000, CPU_XBURST, CPU_XBURST2, CPU_LOONGSON32,
+	CPU_M14KC, CPU_M14KEC, CPU_INTERAPTIV, CPU_P5600, CPU_PROAPTIV, CPU_1074K,
 	CPU_M5150, CPU_I6400, CPU_P6600, CPU_M6250,
 
 	/*
Signature · 1 line
2.56.0
#7 lore raw

PATCH v2 5/9 MIPS: Ingenic: Update Ingenic XBurst implementation IDs

In reply to H. Nikolaus Schaller

Add PRID implementation constants for Ingenic XBurst and XBurst II processors and handle the corresponding implementation IDs in cpu_probe_ingenic().

The PRID table is updated based on Programming Manuals and hardware tests:
- jz4780: 0x3ee1024f (confirmed on CI20 board)

- x1000:  0x2ed1024f (unconfirmed)
- x1600:  0x00d00000 (confirmed on LX16 board)
- X1830:  0x00d00100 (unconfirmed)
- X2000:  0x00132000 (read out from LX20 board)
- X2600:  0x001320xx (confirmed 0x00132000 on LX26 board)

Additionally, clean up extra spaces in the comments for PRID_IMP_XBURST_REV2 and PRID_IMP_XBURST2.

Distinguish XBurst1 implementations with different SIMD ISA revisions and identify XBurst II processors using PRID_IMP_XBURST2. The XBurst II classification follows the Ingenic vendor kernel.

Keep the existing writecombine handling unchanged, since no independent evidence justifies changing the cache attribute for XBurst II.

Public references:
- Ingenic Community, public SoC datasheets and programming manuals:
https://github.com/Ingenic-community/datasheets/tree/main/X
Includes the X1600 programming manual (X1600_PM_20220117.pdf) and documentation for other X-series SoCs, including X1000, X1830 and X2000.
- Ingenic, XBurst1 CPU Core Programming Manual:
https://cdn.hackaday.io/files/1855627939084608/XBurst1_Core_PM.pdf
Documents the XBurst1 core architecture and CP0 configuration.
- Ingenic, XBurst Instruction Set Architecture Programming Manual:
https://www.setphaserstostun.org/ci20/X1000_M200_XBurst_ISA_MXU_PM.pdf
Documents the original XBurst SIMD instruction set.
- Ingenic, XBurst2 CPU Core Programming Manual:
https://www.manualslib.com/manual/2646351/Ingenic-Xburst-2-Cpu-Core.html
Documents the XBurst2 core architecture.
- Ingenic, MIPS-LINUX toolchain R6.0.0 release notes:
https://www.ingenic.com.cn/news-detail/nid-343.html
Distinguishes MSA, MXA and MXU/MXU2 support in the toolchain.
- Ingenic, MIPS-LINUX toolchain R6.2.1 release notes:
https://www.ingenic.com.cn/news-detail/nid-348.html
Documents MXU3 support and identifies X2000 as an MSA target.

The public references document the available CPU, SoC and ISA information. The PRID implementation handling follows the Ingenic vendor kernel.

AI assistance: ChatGPT was used to help review the PRID definitions, compare the Ingenic vendor implementation with the upstream code, and draft the commit message with public references. The code changes, hardware observations, and final patch were reviewed by the authors.

---

Notes:
Changes in V2:
- removed the PRID_IMP_XBURST2_R2 definition and hints for an x2100 as there is no public reference
- merged into single commit changing the PRID definitions and code
- reviewed details and rewrote commit message

Open Issue:
Olivier Louvignes <[email protected]> pointed out that he has simply reused the definition of the X1830 for the X1600. This sets c->writecombine differently (_CACHE_CACHABLE_WA) from _CACHE_UNCACHED_ACCELERATED as we do here.
Both settings appear to work but it needs expert comment which is the right/better one.

 arch/mips/include/asm/cpu.h  | 9 +++++----
 arch/mips/kernel/cpu-probe.c | 7 +++++++
 2 files changed, 12 insertions(+), 4 deletions(-)
arch/mips/include/asm/cpu.h +5 -4
index add26f08e8029..79fd9a0747cf2 100644
@@ -47,7 +47,7 @@
 #define PRID_COMP_CAVIUM	0x0d0000
 #define PRID_COMP_LOONGSON	0x140000
 #define PRID_COMP_INGENIC_13	0x130000	/* X2000, X2100 */
-#define PRID_COMP_INGENIC_D0	0xd00000	/* JZ4730, JZ4740, JZ4750, JZ4755, JZ4760, X1830 */
+#define PRID_COMP_INGENIC_D0	0xd00000	/* JZ4730, JZ4740, JZ4750, JZ4755, JZ4760, X1830, X1600 */
 #define PRID_COMP_INGENIC_D1	0xd10000	/* JZ4770, JZ4775, X1000 */
 #define PRID_COMP_INGENIC_E1	0xe10000	/* JZ4780 */
 
@@ -186,9 +186,10 @@
  * These are the PRID's for when 23:16 == PRID_COMP_INGENIC_*
  */
 
-#define PRID_IMP_XBURST_REV1	0x0200	/* XBurst®1 with MXU1.0/MXU1.1 SIMD ISA	*/
-#define PRID_IMP_XBURST_REV2	0x0100	/* XBurst®1 with MXU2.0 SIMD ISA		*/
-#define PRID_IMP_XBURST2		0x2000	/* XBurst®2 with MXU2.1 SIMD ISA		*/
+#define PRID_IMP_XBURST		0x0000	/* XBurst1 */
+#define PRID_IMP_XBURST_REV1	0x0200	/* XBurst1 with MXU1.0/MXU1.1 SIMD ISA */
+#define PRID_IMP_XBURST_REV2	0x0100	/* XBurst1 with MXU2.0 SIMD ISA */
+#define PRID_IMP_XBURST2	0x2000	/* XBurst2 with MXU2.1 SIMD ISA */
 
 /*
  * These are the PRID's for when 23:16 == PRID_COMP_NETLOGIC
arch/mips/kernel/cpu-probe.c +7 -0
index 489612ed9d498..69b68d00b66f7 100644
@@ -1768,6 +1768,13 @@ static inline void cpu_probe_ingenic(struct cpuinfo_mips *c, unsigned int cpu)
 
 	switch (c->processor_id & PRID_IMP_MASK) {
 
+	/* XBurst®1 with ISA */
+	case PRID_IMP_XBURST:
+		c->cputype = CPU_XBURST;
+		c->writecombine = _CACHE_UNCACHED_ACCELERATED;
+		__cpu_name[cpu] = "Ingenic XBurst";
+		break;
+
 	/* XBurst®1 with MXU1.0/MXU1.1 SIMD ISA */
 	case PRID_IMP_XBURST_REV1:
 
Signature · 1 line
2.56.0
maintainer
#8 lore raw

Re: [PATCH v2 5/9] MIPS: Ingenic: Update Ingenic XBurst implementation IDs

Hi Nikolaus,

Le jeudi 12 novembre 2026 à 09:10 +0100, H. Nikolaus Schaller a écrit :

Quoted 120 lines

Add PRID implementation constants for Ingenic XBurst and XBurst II processors and handle the corresponding implementation IDs in cpu_probe_ingenic().

The PRID table is updated based on Programming Manuals and hardware tests:
- jz4780: 0x3ee1024f (confirmed on CI20 board)
- x1000:  0x2ed1024f (unconfirmed)
- x1600:  0x00d00000 (confirmed on LX16 board)
- X1830:  0x00d00100 (unconfirmed)
- X2000:  0x00132000 (read out from LX20 board)
- X2600:  0x001320xx (confirmed 0x00132000 on LX26 board)

Additionally, clean up extra spaces in the comments for PRID_IMP_XBURST_REV2 and PRID_IMP_XBURST2.

Distinguish XBurst1 implementations with different SIMD ISA revisions and identify XBurst II processors using PRID_IMP_XBURST2. The XBurst II classification follows the Ingenic vendor kernel.

Keep the existing writecombine handling unchanged, since no independent evidence justifies changing the cache attribute for XBurst II.

Public references:
- Ingenic Community, public SoC datasheets and programming manuals:
  https://github.com/Ingenic-community/datasheets/tree/main/X
  Includes the X1600 programming manual (X1600_PM_20220117.pdf) and documentation for other X-series SoCs, including X1000, X1830 and X2000.
- Ingenic, XBurst1 CPU Core Programming Manual:
  https://cdn.hackaday.io/files/1855627939084608/XBurst1_Core_PM.pdf
  Documents the XBurst1 core architecture and CP0 configuration.
- Ingenic, XBurst Instruction Set Architecture Programming Manual:
  https://www.setphaserstostun.org/ci20/X1000_M200_XBurst_ISA_MXU_PM.pdf
  Documents the original XBurst SIMD instruction set.
- Ingenic, XBurst2 CPU Core Programming Manual:
  https://www.manualslib.com/manual/2646351/Ingenic-Xburst-2-Cpu-Core.html
  Documents the XBurst2 core architecture.
- Ingenic, MIPS-LINUX toolchain R6.0.0 release notes:
  https://www.ingenic.com.cn/news-detail/nid-343.html
  Distinguishes MSA, MXA and MXU/MXU2 support in the toolchain.
- Ingenic, MIPS-LINUX toolchain R6.2.1 release notes:
  https://www.ingenic.com.cn/news-detail/nid-348.html
  Documents MXU3 support and identifies X2000 as an MSA target.

The public references document the available CPU, SoC and ISA information. The PRID implementation handling follows the Ingenic vendor kernel.

AI assistance: ChatGPT was used to help review the PRID definitions, compare the Ingenic vendor implementation with the upstream code, and draft the commit message with public references. The code changes, hardware observations, and final patch were reviewed by the authors.

---

Notes:
    Changes in V2:
    - removed the PRID_IMP_XBURST2_R2 definition and hints for an x2100 as there is no public reference
    - merged into single commit changing the PRID definitions and code
    - reviewed details and rewrote commit message

    Open Issue:
    Olivier Louvignes <[email protected]> pointed out that he has simply reused the definition of the X1830 for the X1600. This sets c->writecombine differently (_CACHE_CACHABLE_WA) from _CACHE_UNCACHED_ACCELERATED as we do here.
    Both settings appear to work but it needs expert comment which is the right/better one.

 arch/mips/include/asm/cpu.h  | 9 +++++----
 arch/mips/kernel/cpu-probe.c | 7 +++++++

 2 files changed, 12 insertions(+), 4 deletions(-)
arch/mips/include/asm/cpu.h +0 -1
index add26f08e8029..79fd9a0747cf2 100644
@@ -47,7 +47,7 @@
  #define PRID_COMP_CAVIUM	0x0d0000
  #define PRID_COMP_LOONGSON	0x140000
  #define PRID_COMP_INGENIC_13	0x130000	/* X2000, X2100 */
-#define PRID_COMP_INGENIC_D0	0xd00000	/* JZ4730, JZ4740, JZ4750,

JZ4755, JZ4760, X1830 */
+#define PRID_COMP_INGENIC_D0 0xd00000 /* JZ4730, JZ4740, JZ4750,
JZ4755, JZ4760, X1830, X1600 */
 #define PRID_COMP_INGENIC_D1 0xd10000 /* JZ4770, JZ4775, X1000 */
 #define PRID_COMP_INGENIC_E1 0xe10000 /* JZ4780 */

arch/mips/include/asm/cpu.h +0 -0
@@ -186,9 +186,10 @@
   * These are the PRID's for when 23:16 == PRID_COMP_INGENIC_*
   */


-#define PRID_IMP_XBURST_REV1 0x0200 /* XBurst®1 with MXU1.0/MXU1.1 SIMD
ISA */
-#define PRID_IMP_XBURST_REV2 0x0100 /* XBurst®1 with MXU2.0 SIMD
ISA */
-#define PRID_IMP_XBURST2 0x2000 /* XBurst®2 with MXU2.1 SIMD
ISA */

arch/mips/include/asm/cpu.h +2 -0
+#define PRID_IMP_XBURST		0x0000	/* XBurst1 */
+#define PRID_IMP_XBURST_REV1	0x0200	/* XBurst1 with MXU1.0/MXU1.1 SIMD

ISA */

arch/mips/include/asm/cpu.h +2 -0
+#define PRID_IMP_XBURST_REV2	0x0100	/* XBurst1 with MXU2.0 SIMD ISA */
+#define PRID_IMP_XBURST2	0x2000	/* XBurst2 with MXU2.1 SIMD ISA */


 /*
  * These are the PRID's for when 23:16 == PRID_COMP_NETLOGIC

arch/mips/kernel/cpu-probe.c +0 -0
index 489612ed9d498..69b68d00b66f7 100644
@@ -1768,6 +1768,13 @@ static inline void cpu_probe_ingenic(struct
 cpuinfo_mips *c, unsigned int cpu)


  switch (c->processor_id & PRID_IMP_MASK) {

arch/mips/kernel/cpu-probe.c static inline void cpu_probe_ingenic(struct cpuinfo_mips *c, unsigned int cpu) · line 1776

Quoted 6 lines
arch/mips/kernel/cpu-probe.c +6 -0
+	/* XBurst®1 with ISA */
+	case PRID_IMP_XBURST:
+		c->cputype = CPU_XBURST;
+		c->writecombine = _CACHE_UNCACHED_ACCELERATED;
+		__cpu_name[cpu] = "Ingenic XBurst";
+		break;

What SoCs actually use PRID_IMP_XBURST? The JZ4730?

If it's really used somewhere, you'll also need to add a compatible string for it in Documentation/devicetree/bindings/mips/cpus.yaml. I guess just "ingenic,xburst".

Cheers,
-Paul

#9 lore raw

Re: [PATCH v2 5/9] MIPS: Ingenic: Update Ingenic XBurst implementation IDs

Hi Paul,

Quoted 125 lines

Am 12.11.2026 um 15:03 schrieb Paul Cercueil <[email protected]>:

Hi Nikolaus,

Quoted 120 lines · H. Nikolaus Schaller

Add PRID implementation constants for Ingenic XBurst and XBurst II processors and handle the corresponding implementation IDs in cpu_probe_ingenic().

The PRID table is updated based on Programming Manuals and hardware tests:
- jz4780: 0x3ee1024f (confirmed on CI20 board)

- x1000:  0x2ed1024f (unconfirmed)
- x1600:  0x00d00000 (confirmed on LX16 board)
- X1830:  0x00d00100 (unconfirmed)
- X2000:  0x00132000 (read out from LX20 board)
- X2600:  0x001320xx (confirmed 0x00132000 on LX26 board)

Additionally, clean up extra spaces in the comments for PRID_IMP_XBURST_REV2 and PRID_IMP_XBURST2.

Distinguish XBurst1 implementations with different SIMD ISA revisions and identify XBurst II processors using PRID_IMP_XBURST2. The XBurst II classification follows the Ingenic vendor kernel.

Keep the existing writecombine handling unchanged, since no independent evidence justifies changing the cache attribute for XBurst II.

Public references:
- Ingenic Community, public SoC datasheets and programming manuals:
https://github.com/Ingenic-community/datasheets/tree/main/X
Includes the X1600 programming manual (X1600_PM_20220117.pdf) and documentation for other X-series SoCs, including X1000, X1830 and X2000.
- Ingenic, XBurst1 CPU Core Programming Manual:
https://cdn.hackaday.io/files/1855627939084608/XBurst1_Core_PM.pdf
Documents the XBurst1 core architecture and CP0 configuration.
- Ingenic, XBurst Instruction Set Architecture Programming Manual:
https://www.setphaserstostun.org/ci20/X1000_M200_XBurst_ISA_MXU_PM.pdf
Documents the original XBurst SIMD instruction set.
- Ingenic, XBurst2 CPU Core Programming Manual:
https://www.manualslib.com/manual/2646351/Ingenic-Xburst-2-Cpu-Core.html
Documents the XBurst2 core architecture.
- Ingenic, MIPS-LINUX toolchain R6.0.0 release notes:
https://www.ingenic.com.cn/news-detail/nid-343.html
Distinguishes MSA, MXA and MXU/MXU2 support in the toolchain.
- Ingenic, MIPS-LINUX toolchain R6.2.1 release notes:
https://www.ingenic.com.cn/news-detail/nid-348.html
Documents MXU3 support and identifies X2000 as an MSA target.

The public references document the available CPU, SoC and ISA information. The PRID implementation handling follows the Ingenic vendor kernel.

AI assistance: ChatGPT was used to help review the PRID definitions, compare the Ingenic vendor implementation with the upstream code, and draft the commit message with public references. The code changes, hardware observations, and final patch were reviewed by the authors.

---

Notes:
Changes in V2:
- removed the PRID_IMP_XBURST2_R2 definition and hints for an x2100 as there is no public reference
- merged into single commit changing the PRID definitions and code
- reviewed details and rewrote commit message

Open Issue:
Olivier Louvignes <[email protected]> pointed out that he has simply reused the definition of the X1830 for the X1600. This sets c->writecombine differently (_CACHE_CACHABLE_WA) from _CACHE_UNCACHED_ACCELERATED as we do here.
Both settings appear to work but it needs expert comment which is the right/better one.

 arch/mips/include/asm/cpu.h  | 9 +++++----
 arch/mips/kernel/cpu-probe.c | 7 +++++++
 2 files changed, 12 insertions(+), 4 deletions(-)
arch/mips/include/asm/cpu.h +0 -1
index add26f08e8029..79fd9a0747cf2 100644
@@ -47,7 +47,7 @@
 #define PRID_COMP_CAVIUM 0x0d0000
 #define PRID_COMP_LOONGSON 0x140000
 #define PRID_COMP_INGENIC_13 0x130000 /* X2000, X2100 */
-#define PRID_COMP_INGENIC_D0 0xd00000 /* JZ4730, JZ4740, JZ4750,

JZ4755, JZ4760, X1830 */
+#define PRID_COMP_INGENIC_D0 0xd00000 /* JZ4730, JZ4740, JZ4750, JZ4755, JZ4760, X1830, X1600 */
#define PRID_COMP_INGENIC_D1 0xd10000 /* JZ4770, JZ4775, X1000 */ #define PRID_COMP_INGENIC_E1 0xe10000 /* JZ4780 */

arch/mips/include/asm/cpu.h +0 -1
@@ -186,9 +186,10 @@
  * These are the PRID's for when 23:16 == PRID_COMP_INGENIC_*
  */
 
-#define PRID_IMP_XBURST_REV1 0x0200 /* XBurst®1 with MXU1.0/MXU1.1 SIMD

ISA */
-#define PRID_IMP_XBURST_REV2 0x0100 /* XBurst®1 with MXU2.0 SIMD ISA */
-#define PRID_IMP_XBURST2 0x2000 /* XBurst®2 with MXU2.1 SIMD ISA */

arch/mips/include/asm/cpu.h +2 -0
+#define PRID_IMP_XBURST 0x0000 /* XBurst1 */
+#define PRID_IMP_XBURST_REV1 0x0200 /* XBurst1 with MXU1.0/MXU1.1 SIMD

ISA */

arch/mips/include/asm/cpu.h +2 -0
+#define PRID_IMP_XBURST_REV2 0x0100 /* XBurst1 with MXU2.0 SIMD ISA */
+#define PRID_IMP_XBURST2 0x2000 /* XBurst2 with MXU2.1 SIMD ISA */
 
 /*
  * These are the PRID's for when 23:16 == PRID_COMP_NETLOGIC
arch/mips/kernel/cpu-probe.c +0 -0
index 489612ed9d498..69b68d00b66f7 100644
@@ -1768,6 +1768,13 @@ static inline void cpu_probe_ingenic(struct
 cpuinfo_mips *c, unsigned int cpu)
 
  switch (c->processor_id & PRID_IMP_MASK) {
 
Quoted 8 lines

arch/mips/kernel/cpu-probe.c static inline void cpu_probe_ingenic(struct cpuinfo_mips *c, unsigned int cpu) · line 1776

Quoted 6 lines
arch/mips/kernel/cpu-probe.c +6 -0
+ /* XBurst®1 with ISA */
+ case PRID_IMP_XBURST:
+ c->cputype = CPU_XBURST;
+ c->writecombine = _CACHE_UNCACHED_ACCELERATED;
+ __cpu_name[cpu] = "Ingenic XBurst";
+ break;

What SoCs actually use PRID_IMP_XBURST? The JZ4730?

The x1600 has a PRID of 0x00d00000.

It is a little confusing how these numbers have been assigned over time.

+#define PRID_IMP_XBURST 0x0000 /* XBurst1 */
+#define PRID_IMP_XBURST_REV1 0x0200 /* XBurst1 with MXU1.0/MXU1.1 SIMD ISA */ +#define PRID_IMP_XBURST_REV2 0x0100 /* XBurst1 with MXU2.0 SIMD ISA */

Obviously the x1600 is newer than X1830 (likely PRID_IMP_XBURST_REV2) or jz4780 and x1000 being PRID_IMP_XBURST_REV1.

So it may or may not be a PRID_IMP_XBURST_REV3? But we don't know.

Quoted 4 lines

If it's really used somewhere, you'll also need to add a compatible string for it in Documentation/devicetree/bindings/mips/cpus.yaml. I guess just "ingenic,xburst".

Ok, will check.

BR and thanks,
Nikolaus

maintainer
#10 lore raw

Re: [PATCH v2 5/9] MIPS: Ingenic: Update Ingenic XBurst implementation IDs

Le jeudi 12 novembre 2026 à 16:37 +0100, H. Nikolaus Schaller a écrit :

Quoted 144 lines

Hi Paul,

Quoted 125 lines

Am 12.11.2026 um 15:03 schrieb Paul Cercueil <[email protected]>:

Hi Nikolaus,

Quoted 120 lines · H. Nikolaus Schaller

Add PRID implementation constants for Ingenic XBurst and XBurst II processors and handle the corresponding implementation IDs in cpu_probe_ingenic().

The PRID table is updated based on Programming Manuals and hardware tests:
- jz4780: 0x3ee1024f (confirmed on CI20 board)
- x1000:  0x2ed1024f (unconfirmed)
- x1600:  0x00d00000 (confirmed on LX16 board)
- X1830:  0x00d00100 (unconfirmed)
- X2000:  0x00132000 (read out from LX20 board)
- X2600:  0x001320xx (confirmed 0x00132000 on LX26 board)

Additionally, clean up extra spaces in the comments for PRID_IMP_XBURST_REV2 and PRID_IMP_XBURST2.

Distinguish XBurst1 implementations with different SIMD ISA revisions and identify XBurst II processors using PRID_IMP_XBURST2. The XBurst II classification follows the Ingenic vendor kernel.

Keep the existing writecombine handling unchanged, since no independent evidence justifies changing the cache attribute for XBurst II.

Public references:
- Ingenic Community, public SoC datasheets and programming manuals:
  https://github.com/Ingenic-community/datasheets/tree/main/X
  Includes the X1600 programming manual (X1600_PM_20220117.pdf) and documentation for other X-series SoCs, including X1000, X1830 and X2000.
- Ingenic, XBurst1 CPU Core Programming Manual:
  https://cdn.hackaday.io/files/1855627939084608/XBurst1_Core_PM.pdf
  Documents the XBurst1 core architecture and CP0 configuration.
- Ingenic, XBurst Instruction Set Architecture Programming Manual:
  https://www.setphaserstostun.org/ci20/X1000_M200_XBurst_ISA_MXU_PM.pdf
  Documents the original XBurst SIMD instruction set.
- Ingenic, XBurst2 CPU Core Programming Manual:
  https://www.manualslib.com/manual/2646351/Ingenic-Xburst-2-Cpu-Core.html
  Documents the XBurst2 core architecture.
- Ingenic, MIPS-LINUX toolchain R6.0.0 release notes:
  https://www.ingenic.com.cn/news-detail/nid-343.html
  Distinguishes MSA, MXA and MXU/MXU2 support in the toolchain.
- Ingenic, MIPS-LINUX toolchain R6.2.1 release notes:
  https://www.ingenic.com.cn/news-detail/nid-348.html
  Documents MXU3 support and identifies X2000 as an MSA target.

The public references document the available CPU, SoC and ISA information. The PRID implementation handling follows the Ingenic vendor kernel.

AI assistance: ChatGPT was used to help review the PRID definitions, compare the Ingenic vendor implementation with the upstream code, and draft the commit message with public references. The code changes, hardware observations, and final patch were reviewed by the authors.

---

Notes:
    Changes in V2:
    - removed the PRID_IMP_XBURST2_R2 definition and hints for an x2100 as there is no public reference
    - merged into single commit changing the PRID definitions and code
    - reviewed details and rewrote commit message

    Open Issue:
    Olivier Louvignes <[email protected]> pointed out that he has simply reused the definition of the X1830 for the X1600. This sets c->writecombine differently (_CACHE_CACHABLE_WA) from _CACHE_UNCACHED_ACCELERATED as we do here.
    Both settings appear to work but it needs expert comment which is the right/better one.

 arch/mips/include/asm/cpu.h  | 9 +++++----
 arch/mips/kernel/cpu-probe.c | 7 +++++++

 2 files changed, 12 insertions(+), 4 deletions(-)
arch/mips/include/asm/cpu.h +0 -1
index add26f08e8029..79fd9a0747cf2 100644
@@ -47,7 +47,7 @@
  #define PRID_COMP_CAVIUM 0x0d0000
  #define PRID_COMP_LOONGSON 0x140000
  #define PRID_COMP_INGENIC_13 0x130000 /* X2000, X2100 */
-#define PRID_COMP_INGENIC_D0 0xd00000 /* JZ4730, JZ4740, JZ4750,

JZ4755, JZ4760, X1830 */
+#define PRID_COMP_INGENIC_D0 0xd00000 /* JZ4730, JZ4740, JZ4750, JZ4755, JZ4760, X1830, X1600 */
 #define PRID_COMP_INGENIC_D1 0xd10000 /* JZ4770, JZ4775, X1000 */ #define PRID_COMP_INGENIC_E1 0xe10000 /* JZ4780 */

arch/mips/include/asm/cpu.h +0 -0
@@ -186,9 +186,10 @@
   * These are the PRID's for when 23:16 == PRID_COMP_INGENIC_*
   */


-#define PRID_IMP_XBURST_REV1 0x0200 /* XBurst®1 with MXU1.0/MXU1.1 SIMD ISA */
-#define PRID_IMP_XBURST_REV2 0x0100 /* XBurst®1 with MXU2.0 SIMD ISA */
-#define PRID_IMP_XBURST2 0x2000 /* XBurst®2 with MXU2.1 SIMD ISA */

arch/mips/include/asm/cpu.h +2 -0
+#define PRID_IMP_XBURST 0x0000 /* XBurst1 */
+#define PRID_IMP_XBURST_REV1 0x0200 /* XBurst1 with MXU1.0/MXU1.1 SIMD

ISA */

arch/mips/include/asm/cpu.h +2 -0
+#define PRID_IMP_XBURST_REV2 0x0100 /* XBurst1 with MXU2.0 SIMD ISA */
+#define PRID_IMP_XBURST2 0x2000 /* XBurst2 with MXU2.1 SIMD ISA */


 /*
  * These are the PRID's for when 23:16 == PRID_COMP_NETLOGIC

arch/mips/kernel/cpu-probe.c +0 -0
index 489612ed9d498..69b68d00b66f7 100644
@@ -1768,6 +1768,13 @@ static inline void cpu_probe_ingenic(struct
 cpuinfo_mips *c, unsigned int cpu)


  switch (c->processor_id & PRID_IMP_MASK) {

Quoted 8 lines
Quoted 6 lines

+ /* XBurst®1 with ISA */
+ case PRID_IMP_XBURST:
+ c->cputype = CPU_XBURST;
+ c->writecombine = _CACHE_UNCACHED_ACCELERATED;
+ __cpu_name[cpu] = "Ingenic XBurst";
+ break;

What SoCs actually use PRID_IMP_XBURST? The JZ4730?

The x1600 has a PRID of 0x00d00000.

It is a little confusing how these numbers have been assigned over time.

arch/mips/include/asm/cpu.h +3 -0
+#define PRID_IMP_XBURST 0x0000 /* XBurst1 */
+#define PRID_IMP_XBURST_REV1 0x0200 /* XBurst1 with MXU1.0/MXU1.1 SIMD ISA */
+#define PRID_IMP_XBURST_REV2 0x0100 /* XBurst1 with MXU2.0 SIMD ISA */
Quoted 1 line

Obviously the x1600 is newer than X1830 (likely PRID_IMP_XBURST_REV2)

Obviously? ;)

I thought that their SoC versions were ordered, with the only exception being the JZ4725B being newer than the JZ4740 as a sort-of beta JZ4750.

Quoted 3 lines

or jz4780 and x1000 being PRID_IMP_XBURST_REV1.

So it may or may not be a PRID_IMP_XBURST_REV3? But we don't know.

Can you check if the X1600 supports MXU? A IMP value of 0 might mean a Xburst CPU with no XMU support, maybe?

Quoted 7 lines
Quoted 5 lines

If it's really used somewhere, you'll also need to add a compatible string for
it in Documentation/devicetree/bindings/mips/cpus.yaml. I guess just "ingenic,xburst".

Ok, will check.

"ingenic,xburst" if it does not support MXU, I should clarify.

Cheers,
-Paul

#11 lore raw

Re: [PATCH v2 5/9] MIPS: Ingenic: Update Ingenic XBurst implementation IDs

Hi Paul,

Quoted 143 lines

Am 12.11.2026 um 16:49 schrieb Paul Cercueil <[email protected]>:

Quoted 140 lines · H. Nikolaus Schaller

Hi Paul,

Quoted 125 lines

Am 12.11.2026 um 15:03 schrieb Paul Cercueil <[email protected]>:

Hi Nikolaus,

Quoted 120 lines · H. Nikolaus Schaller

Add PRID implementation constants for Ingenic XBurst and XBurst II processors and handle the corresponding implementation IDs in cpu_probe_ingenic().

The PRID table is updated based on Programming Manuals and hardware tests:
- jz4780: 0x3ee1024f (confirmed on CI20 board)

- x1000:  0x2ed1024f (unconfirmed)
- x1600:  0x00d00000 (confirmed on LX16 board)
- X1830:  0x00d00100 (unconfirmed)
- X2000:  0x00132000 (read out from LX20 board)
- X2600:  0x001320xx (confirmed 0x00132000 on LX26 board)

Additionally, clean up extra spaces in the comments for PRID_IMP_XBURST_REV2 and PRID_IMP_XBURST2.

Distinguish XBurst1 implementations with different SIMD ISA revisions and identify XBurst II processors using PRID_IMP_XBURST2. The XBurst II classification follows the Ingenic vendor kernel.

Keep the existing writecombine handling unchanged, since no independent evidence justifies changing the cache attribute for XBurst II.

Public references:
- Ingenic Community, public SoC datasheets and programming manuals:
https://github.com/Ingenic-community/datasheets/tree/main/X
Includes the X1600 programming manual (X1600_PM_20220117.pdf) and documentation for other X-series SoCs, including X1000, X1830 and X2000.
- Ingenic, XBurst1 CPU Core Programming Manual:
https://cdn.hackaday.io/files/1855627939084608/XBurst1_Core_PM.pdf
Documents the XBurst1 core architecture and CP0 configuration.
- Ingenic, XBurst Instruction Set Architecture Programming Manual:
https://www.setphaserstostun.org/ci20/X1000_M200_XBurst_ISA_MXU_PM.pdf
Documents the original XBurst SIMD instruction set.
- Ingenic, XBurst2 CPU Core Programming Manual:
https://www.manualslib.com/manual/2646351/Ingenic-Xburst-2-Cpu-Core.html
Documents the XBurst2 core architecture.
- Ingenic, MIPS-LINUX toolchain R6.0.0 release notes:
https://www.ingenic.com.cn/news-detail/nid-343.html
Distinguishes MSA, MXA and MXU/MXU2 support in the toolchain.
- Ingenic, MIPS-LINUX toolchain R6.2.1 release notes:
https://www.ingenic.com.cn/news-detail/nid-348.html
Documents MXU3 support and identifies X2000 as an MSA target.

The public references document the available CPU, SoC and ISA information. The PRID implementation handling follows the Ingenic vendor kernel.

AI assistance: ChatGPT was used to help review the PRID definitions, compare the Ingenic vendor implementation with the upstream code, and draft the commit message with public references. The code changes, hardware observations, and final patch were reviewed by the authors.

---

Notes:
Changes in V2:
- removed the PRID_IMP_XBURST2_R2 definition and hints for an x2100 as there is no public reference
- merged into single commit changing the PRID definitions and code
- reviewed details and rewrote commit message

Open Issue:
Olivier Louvignes <[email protected]> pointed out that he has simply reused the definition of the X1830 for the X1600. This sets c->writecombine differently (_CACHE_CACHABLE_WA) from _CACHE_UNCACHED_ACCELERATED as we do here.
Both settings appear to work but it needs expert comment which is the right/better one.

 arch/mips/include/asm/cpu.h  | 9 +++++----
 arch/mips/kernel/cpu-probe.c | 7 +++++++
 2 files changed, 12 insertions(+), 4 deletions(-)
arch/mips/include/asm/cpu.h +0 -1
index add26f08e8029..79fd9a0747cf2 100644
@@ -47,7 +47,7 @@
 #define PRID_COMP_CAVIUM 0x0d0000
 #define PRID_COMP_LOONGSON 0x140000
 #define PRID_COMP_INGENIC_13 0x130000 /* X2000, X2100 */
-#define PRID_COMP_INGENIC_D0 0xd00000 /* JZ4730, JZ4740, JZ4750,

JZ4755, JZ4760, X1830 */
+#define PRID_COMP_INGENIC_D0 0xd00000 /* JZ4730, JZ4740, JZ4750, JZ4755, JZ4760, X1830, X1600 */
#define PRID_COMP_INGENIC_D1 0xd10000 /* JZ4770, JZ4775, X1000 */ #define PRID_COMP_INGENIC_E1 0xe10000 /* JZ4780 */

arch/mips/include/asm/cpu.h +0 -1
@@ -186,9 +186,10 @@
  * These are the PRID's for when 23:16 == PRID_COMP_INGENIC_*
  */
 
-#define PRID_IMP_XBURST_REV1 0x0200 /* XBurst®1 with MXU1.0/MXU1.1 SIMD

ISA */
-#define PRID_IMP_XBURST_REV2 0x0100 /* XBurst®1 with MXU2.0 SIMD ISA */
-#define PRID_IMP_XBURST2 0x2000 /* XBurst®2 with MXU2.1 SIMD ISA */

arch/mips/include/asm/cpu.h +2 -0
+#define PRID_IMP_XBURST 0x0000 /* XBurst1 */
+#define PRID_IMP_XBURST_REV1 0x0200 /* XBurst1 with MXU1.0/MXU1.1 SIMD

ISA */

arch/mips/include/asm/cpu.h +2 -0
+#define PRID_IMP_XBURST_REV2 0x0100 /* XBurst1 with MXU2.0 SIMD ISA */
+#define PRID_IMP_XBURST2 0x2000 /* XBurst2 with MXU2.1 SIMD ISA */
 
 /*
  * These are the PRID's for when 23:16 == PRID_COMP_NETLOGIC
arch/mips/kernel/cpu-probe.c +0 -0
index 489612ed9d498..69b68d00b66f7 100644
@@ -1768,6 +1768,13 @@ static inline void cpu_probe_ingenic(struct
 cpuinfo_mips *c, unsigned int cpu)
 
  switch (c->processor_id & PRID_IMP_MASK) {
 
Quoted 8 lines

arch/mips/kernel/cpu-probe.c static inline void cpu_probe_ingenic(struct cpuinfo_mips *c, unsigned int cpu) · line 1776

Quoted 6 lines
arch/mips/kernel/cpu-probe.c +6 -0
+ /* XBurst®1 with ISA */
+ case PRID_IMP_XBURST:
+ c->cputype = CPU_XBURST;
+ c->writecombine = _CACHE_UNCACHED_ACCELERATED;
+ __cpu_name[cpu] = "Ingenic XBurst";
+ break;

What SoCs actually use PRID_IMP_XBURST? The JZ4730?

The x1600 has a PRID of 0x00d00000.

It is a little confusing how these numbers have been assigned over time.

Quoted 7 lines
Quoted 5 lines
arch/mips/include/asm/cpu.h +3 -0
+#define PRID_IMP_XBURST 0x0000 /* XBurst1 */
+#define PRID_IMP_XBURST_REV1 0x0200 /* XBurst1 with MXU1.0/MXU1.1 SIMD ISA */
+#define PRID_IMP_XBURST_REV2 0x0100 /* XBurst1 with MXU2.0 SIMD ISA */

Obviously the x1600 is newer than X1830 (likely PRID_IMP_XBURST_REV2)

Obviously? ;)

Well,
if REV1 is encoded as 0x02, and REV2 as 0x01, then 0x00 must be REV3 :)

Quoted 10 lines

I thought that their SoC versions were ordered, with the only exception being the JZ4725B being newer than the JZ4740 as a sort-of beta JZ4750.

Quoted 3 lines

or jz4780 and x1000 being PRID_IMP_XBURST_REV1.

So it may or may not be a PRID_IMP_XBURST_REV3? But we don't know.

Can you check if the X1600 supports MXU? A IMP value of 0 might mean a Xburst CPU with no XMU support, maybe?

The Programing Manual of the X1600 only mentions MXU as missing in the auxiliary MCU.

X1600: mentions only as missing in the auxiliary MCU
X1000: Enhanced MXU implements XBurst SIMD instruction set release I and release II
X1830: The MXU2.0(SIMD128) instruction set has been implemented by XBurst® engine, and is one part of the CPU
X2000: mentions only as missing in the auxiliary MCU
but Ingenic SIMD instruction set architecture: 128bit MXA
X2600: same
JZ4780: Media Extension Unit (MXU) Ingenic SIMD instruction set II fully pipelined

So MXU is only available in X1000 and X1830. If the data sheet is right.

Is there some runtime check? Hardware capabilities register to look into? The PRID values are known but not their exact meaning.

But if we take these as MXU-Flags we would have:

- jz4780: 0x02 (confirmed on CI20 board)

- x1000:  0x02 (unconfirmed)
- x1600:  0x00 (confirmed on LX16 board)
- X1830:  0x01 (unconfirmed)
- X2000:  0x20 (confirmed and read out from LX20 board)
- X2600:  0x20 (confirmed 0x00132000 on LX26 board)

So the lower 4 bits could indeed encode the MXU revision and the upper 4 bits the XBurst revision.

x1600: 0x00 none XBURST
jz4780/x1000: 0x02 MXU2.0 XBURST_REV1 (REV has been swapped?)
x1830: 0x01 MXU1.0 XBURST_REV2 (REV has been swapped?)
x2000/2600: 0x20 MXA/XBURST2 XBURST2

IMHO, in that case defining PRID_IMP_XBURST = 0x0000 would be correct. Maybe the _REV is misleading and should be _MXU_REV?

But that is quite some speculation.

Any thoughts?

BR,
Nikolaus

Quoted 14 lines
Quoted 8 lines
Quoted 6 lines

If it's really used somewhere, you'll also need to add a compatible string for
it in Documentation/devicetree/bindings/mips/cpus.yaml. I guess just "ingenic,xburst".

Ok, will check.

"ingenic,xburst" if it does not support MXU, I should clarify.

Cheers,
-Paul

#12 lore raw

Re: [PATCH v2 5/9] MIPS: Ingenic: Update Ingenic XBurst implementation IDs

Hi Nikolaus,

Quoted 2 lines · H. Nikolaus Schaller

So the lower 4 bits could indeed encode the MXU revision and the upper 4 bits the XBurst revision.

[...]

Quoted 6 lines

IMHO, in that case defining PRID_IMP_XBURST = 0x0000 would be correct. Maybe the _REV is misleading and should be _MXU_REV?

But that is quite some speculation.

Any thoughts?

The core manuals define that byte. XBurst1 CPU Core PM, 2.3.1.8 (PRId), bits 15:8:

15:13 Generation: 0 - XBurst1, 1 - XBurst2
12:8 uA, micro architecture version ("please reference to SOC
document")

The XBurst2 CPU Core PM, 3.3.1.7, has the same layout, with uA 0 for the X2000. Decoded:

- X1600 0x00d00000: XBurst1, uA 0
- X1830 0x00d00100: XBurst1, uA 1 (PRID_IMP_XBURST_REV2 today)
- JZ4780 and X1000 (0x...024f): XBurst1, uA 2 (PRID_IMP_XBURST_REV1)
- X2000 and X2600 0x00132000: XBurst2, uA 0 (PRID_IMP_XBURST2)

So the upper bits are the core generation, as you guessed, but the lower ones number the micro-architecture rather than the MXU: the core manual leaves their meaning to the SoC documents, and the X1600 PM only gives the value. This is the reading Maciej gave on my v1, where he suggested PRID_IMP_XBURST_REV0 for 0x0000 (REV1 and REV2 being swapped with respect to their values):
https://lore.kernel.org/all/[email protected]/

I have asked Thomas to drop my cpu-probe patch in favour of this one.

Thanks,
Olivier

#13 lore raw

Re: [PATCH v2 5/9] MIPS: Ingenic: Update Ingenic XBurst implementation IDs

In reply to H. Nikolaus Schaller

Hi Nikolaus,

Quoted 1 line · H. Nikolaus Schaller

Is there some runtime check? Hardware capabilities register to look into?

The MXU2 manual says MXU instructions trap if Config1.C2 isn't set, so that bit on the LX16 should tell you.

Hunter

#14 lore raw

PATCH v2 6/9 MIPS: generic: Add MACH_INGENIC type and compatible for X1600 and X2600

In reply to H. Nikolaus Schaller

Add support for the Ingenic/Lumissil X1600 and X2600 SoC by introducing new machine types.

Update the system type string mapping and add the "ingenic,x1600" and "ingenic,x2600" compatible strings to the DT match table in the generic Ingenic board support.

Note that we do not separate between e.g. X1600, X1600E, X1600H since they differ only in built-in memory size which can be handled by the boot loader.

---

Notes:
Note that there is no bindings document for any of the compatible values so we don't add them.

 arch/mips/generic/board-ingenic.c | 8 +++++++-
 arch/mips/include/asm/bootinfo.h  | 2 ++
 2 files changed, 9 insertions(+), 1 deletion(-)
arch/mips/generic/board-ingenic.c +7 -1
index 1f4906875e7bc..5840293913977 100644
@@ -24,6 +24,8 @@
 static __init char *ingenic_get_system_type(unsigned long machtype)
 {
 	switch (machtype) {
+	case MACH_INGENIC_X2600:
+		return "X2600";
 	case MACH_INGENIC_X2100:
 		return "X2100";
 	case MACH_INGENIC_X2000H:
@@ -34,6 +36,8 @@ static __init char *ingenic_get_system_type(unsigned long machtype)
 		return "X2000";
 	case MACH_INGENIC_X1830:
 		return "X1830";
+	case MACH_INGENIC_X1600:
+		return "X1600";
 	case MACH_INGENIC_X1000E:
 		return "X1000E";
 	case MACH_INGENIC_X1000:
@@ -146,11 +150,13 @@ static const struct of_device_id ingenic_of_match[] __initconst = {
 	{ .compatible = "ingenic,jz4780", .data = (void *)MACH_INGENIC_JZ4780 },
 	{ .compatible = "ingenic,x1000", .data = (void *)MACH_INGENIC_X1000 },
 	{ .compatible = "ingenic,x1000e", .data = (void *)MACH_INGENIC_X1000E },
+	{ .compatible = "ingenic,x1600", .data = (void *)MACH_INGENIC_X1600 },
 	{ .compatible = "ingenic,x1830", .data = (void *)MACH_INGENIC_X1830 },
 	{ .compatible = "ingenic,x2000", .data = (void *)MACH_INGENIC_X2000 },
 	{ .compatible = "ingenic,x2000e", .data = (void *)MACH_INGENIC_X2000E },
 	{ .compatible = "ingenic,x2000h", .data = (void *)MACH_INGENIC_X2000H },
 	{ .compatible = "ingenic,x2100", .data = (void *)MACH_INGENIC_X2100 },
+	{ .compatible = "ingenic,x2600", .data = (void *)MACH_INGENIC_X2600 },
 	{}
 };
 
@@ -188,7 +194,7 @@ static const struct platform_suspend_ops ingenic_pm_ops = {
 
 static int __init ingenic_pm_init(void)
 {
-	if (boot_cpu_type() == CPU_XBURST) {
+	if (boot_cpu_type() == CPU_XBURST || boot_cpu_type() == CPU_XBURST2) {
 		if (IS_ENABLED(CONFIG_PM_SLEEP))
 			suspend_set_ops(&ingenic_pm_ops);
 		_machine_halt = ingenic_halt;
arch/mips/include/asm/bootinfo.h +2 -0
index 2128ba9033911..f6ae983577f6f 100644
@@ -81,11 +81,13 @@ enum ingenic_machine_type {
 	MACH_INGENIC_JZ4780,
 	MACH_INGENIC_X1000,
 	MACH_INGENIC_X1000E,
+	MACH_INGENIC_X1600,
 	MACH_INGENIC_X1830,
 	MACH_INGENIC_X2000,
 	MACH_INGENIC_X2000E,
 	MACH_INGENIC_X2000H,
 	MACH_INGENIC_X2100,
+	MACH_INGENIC_X2600,
 };
 
 extern char *system_type;
Signature · 1 line
2.56.0
maintainer
#15 lore raw

Re: [PATCH v2 6/9] MIPS: generic: Add MACH_INGENIC type and compatible for X1600 and X2600

Hi Nikolaus,

Le jeudi 12 novembre 2026 à 09:10 +0100, H. Nikolaus Schaller a écrit :

Quoted 67 lines

Add support for the Ingenic/Lumissil X1600 and X2600 SoC by introducing new machine types.

Update the system type string mapping and add the "ingenic,x1600" and "ingenic,x2600" compatible strings to the DT match table in the generic Ingenic board support.

Note that we do not separate between e.g. X1600, X1600E, X1600H since they differ only in built-in memory size which can be handled by the boot loader.

---

Notes:
    Note that there is no bindings document for any of the compatible values so we don't add them.

 arch/mips/generic/board-ingenic.c | 8 +++++++-
 arch/mips/include/asm/bootinfo.h  | 2 ++

 2 files changed, 9 insertions(+), 1 deletion(-)
arch/mips/generic/board- renamed from arch/mips/generic/board-ingenic.c +0 -0

ingenic.c
index 1f4906875e7bc..5840293913977 100644

arch/mips/generic/board-ingenic.c +2 -0
@@ -24,6 +24,8 @@
  static __init char *ingenic_get_system_type(unsigned long machtype)
  {
  	switch (machtype) {
+	case MACH_INGENIC_X2600:
+		return "X2600";

  case MACH_INGENIC_X2100:
  return "X2100";
  case MACH_INGENIC_X2000H:

arch/mips/generic/board-ingenic.c +0 -0
@@ -34,6 +36,8 @@ static __init char *ingenic_get_system_type(unsigned long
 machtype)
  		return "X2000";

  case MACH_INGENIC_X1830:
  return "X1830";

arch/mips/generic/board-ingenic.c +2 -0
+	case MACH_INGENIC_X1600:
+		return "X1600";

  case MACH_INGENIC_X1000E:
  return "X1000E";
  case MACH_INGENIC_X1000:

arch/mips/generic/board-ingenic.c +0 -0
@@ -146,11 +150,13 @@ static const struct of_device_id ingenic_of_match[]
 __initconst = {

  { .compatible = "ingenic,jz4780", .data = (void *)MACH_INGENIC_JZ4780 },
  { .compatible = "ingenic,x1000", .data = (void *)MACH_INGENIC_X1000 },
  { .compatible = "ingenic,x1000e", .data = (void *)MACH_INGENIC_X1000E },

arch/mips/generic/board-ingenic.c +1 -0
+	{ .compatible = "ingenic,x1600", .data = (void *)MACH_INGENIC_X1600

},
  { .compatible = "ingenic,x1830", .data = (void *)MACH_INGENIC_X1830 },
  { .compatible = "ingenic,x2000", .data = (void *)MACH_INGENIC_X2000 },
  { .compatible = "ingenic,x2000e", .data = (void *)MACH_INGENIC_X2000E },
  { .compatible = "ingenic,x2000h", .data = (void *)MACH_INGENIC_X2000H },
  { .compatible = "ingenic,x2100", .data = (void *)MACH_INGENIC_X2100 },

arch/mips/generic/board-ingenic.c +1 -0
+	{ .compatible = "ingenic,x2600", .data = (void *)MACH_INGENIC_X2600
Quoted 1 line

},

I was going to say that if you add compatibles here you'll also need to add them to the bindings documentation, but it seems we don't have one?

We have Documentation/devicetree/bindings/mips/ingenic/devices.yaml for boards but most SoCs listed here don't have an accompanying board.

Given that it's not your mess I'm OK with the patch, but we should really fix this at some point.

Cheers,
-Paul

Quoted 33 lines

  {}
 };

arch/mips/generic/board-ingenic.c +0 -0
@@ -188,7 +194,7 @@ static const struct platform_suspend_ops ingenic_pm_ops =
 {


 static int __init ingenic_pm_init(void)
 {

arch/mips/generic/board-ingenic.c +1 -1
-	if (boot_cpu_type() == CPU_XBURST) {
+	if (boot_cpu_type() == CPU_XBURST || boot_cpu_type() == CPU_XBURST2)

{
  if (IS_ENABLED(CONFIG_PM_SLEEP)) suspend_set_ops(&ingenic_pm_ops); _machine_halt = ingenic_halt;

arch/mips/include/asm/bootinfo.h +2 -0
index 2128ba9033911..f6ae983577f6f 100644
@@ -81,11 +81,13 @@ enum ingenic_machine_type {
  	MACH_INGENIC_JZ4780,
  	MACH_INGENIC_X1000,
  	MACH_INGENIC_X1000E,
+	MACH_INGENIC_X1600,
  	MACH_INGENIC_X1830,
  	MACH_INGENIC_X2000,
  	MACH_INGENIC_X2000E,
  	MACH_INGENIC_X2000H,
  	MACH_INGENIC_X2100,
+	MACH_INGENIC_X2600,
  };


 extern char *system_type;

#16 lore raw

Re: [PATCH v2 6/9] MIPS: generic: Add MACH_INGENIC type and compatible for X1600 and X2600

Hi Paul,

Quoted 68 lines

Am 12.11.2026 um 15:09 schrieb Paul Cercueil <[email protected]>:

Hi Nikolaus,

Quoted 63 lines · H. Nikolaus Schaller

Add support for the Ingenic/Lumissil X1600 and X2600 SoC by introducing new machine types.

Update the system type string mapping and add the "ingenic,x1600" and "ingenic,x2600" compatible strings to the DT match table in the generic Ingenic board support.

Note that we do not separate between e.g. X1600, X1600E, X1600H since they differ only in built-in memory size which can be handled by the boot loader.

---

Notes:
Note that there is no bindings document for any of the compatible values so we don't add them.

 arch/mips/generic/board-ingenic.c | 8 +++++++-
 arch/mips/include/asm/bootinfo.h  | 2 ++
 2 files changed, 9 insertions(+), 1 deletion(-)
arch/mips/generic/board- renamed from arch/mips/generic/board-ingenic.c +0 -0

ingenic.c
index 1f4906875e7bc..5840293913977 100644

arch/mips/generic/board-ingenic.c +4 -0
@@ -24,6 +24,8 @@
 static __init char *ingenic_get_system_type(unsigned long machtype)
 {
  switch (machtype) {
+ case MACH_INGENIC_X2600:
+ return "X2600";
  case MACH_INGENIC_X2100:
  return "X2100";
  case MACH_INGENIC_X2000H:
@@ -34,6 +36,8 @@ static __init char *ingenic_get_system_type(unsigned long
 machtype)
  return "X2000";
  case MACH_INGENIC_X1830:
  return "X1830";
+ case MACH_INGENIC_X1600:
+ return "X1600";
  case MACH_INGENIC_X1000E:
  return "X1000E";

case MACH_INGENIC_X1000:

arch/mips/generic/board-ingenic.c +1 -0
@@ -146,11 +150,13 @@ static const struct of_device_id ingenic_of_match[]
 __initconst = {
  { .compatible = "ingenic,jz4780", .data = (void *)MACH_INGENIC_JZ4780
 },
  { .compatible = "ingenic,x1000", .data = (void *)MACH_INGENIC_X1000
 },
  { .compatible = "ingenic,x1000e", .data = (void *)MACH_INGENIC_X1000E
 },
+ { .compatible = "ingenic,x1600", .data = (void *)MACH_INGENIC_X1600
 },
  { .compatible = "ingenic,x1830", .data = (void *)MACH_INGENIC_X1830
 },
  { .compatible = "ingenic,x2000", .data = (void *)MACH_INGENIC_X2000

},
{ .compatible = "ingenic,x2000e", .data = (void *)MACH_INGENIC_X2000E
},
{ .compatible = "ingenic,x2000h", .data = (void *)MACH_INGENIC_X2000H

Quoted 8 lines
Quoted 5 lines

},
{ .compatible = "ingenic,x2100", .data = (void *)MACH_INGENIC_X2100
},
+ { .compatible = "ingenic,x2600", .data = (void *)MACH_INGENIC_X2600 },

I was going to say that if you add compatibles here you'll also need to add them to the bindings documentation, but it seems we don't have one?

Indeed... Checkpatch even reported about.

I added that to the Notes section of the commit that we have no bindings doc.

Quoted 6 lines

We have Documentation/devicetree/bindings/mips/ingenic/devices.yaml for boards but most SoCs listed here don't have an accompanying board.

Given that it's not your mess I'm OK with the patch, but we should really fix this at some point.

Agreed. Should not be too complex, I guess.

BR and thanks,
Nikolaus

Quoted 37 lines

Cheers,
-Paul

Quoted 33 lines

{}
};

arch/mips/generic/board-ingenic.c +1 -1
@@ -188,7 +194,7 @@ static const struct platform_suspend_ops ingenic_pm_ops =
 {
 
 static int __init ingenic_pm_init(void)
 {
- if (boot_cpu_type() == CPU_XBURST) {
+ if (boot_cpu_type() == CPU_XBURST || boot_cpu_type() == CPU_XBURST2)
 {
  if (IS_ENABLED(CONFIG_PM_SLEEP))
  suspend_set_ops(&ingenic_pm_ops);
  _machine_halt = ingenic_halt;
arch/mips/include/asm/bootinfo.h +2 -0
index 2128ba9033911..f6ae983577f6f 100644
@@ -81,11 +81,13 @@ enum ingenic_machine_type {
  MACH_INGENIC_JZ4780,
  MACH_INGENIC_X1000,
  MACH_INGENIC_X1000E,
+ MACH_INGENIC_X1600,
  MACH_INGENIC_X1830,
  MACH_INGENIC_X2000,
  MACH_INGENIC_X2000E,
  MACH_INGENIC_X2000H,
  MACH_INGENIC_X2100,
+ MACH_INGENIC_X2600,
 };
 
 extern char *system_type;
#17 lore raw

PATCH v2 7/9 MIPS: Ingenic: Expand Kconfig for X1600, X2000, and X2600 SoCs

In reply to H. Nikolaus Schaller

Expand MACH_INGENIC_GENERIC to select all existing and newly added machine variants (including JZ4750/55/60/75, X1500, X1600, X2000, X2100, and X2600).

Introduce individual Kconfig blocks for the following SoCs:
- MACH_X1600: Selects MIPS32_R2 and HIGHMEM.
- MACH_X2000 / MACH_X2600: Selects MIPS32_R5, SMP support, broadcast clockevents, O32 FP64 support, and weak ordering optimizations.

In parts developed by Reimu NotMoe <[email protected]>

---
 arch/mips/ingenic/Kconfig | 38 ++++++++++++++++++++++++++++++++++++++
 1 file changed, 38 insertions(+)
arch/mips/ingenic/Kconfig +38 -0
index f595b339a4b84..5d351b42bbc9b 100644
@@ -5,10 +5,19 @@ config MACH_INGENIC_GENERIC
 	select MACH_INGENIC
 	select MACH_JZ4740
 	select MACH_JZ4725B
+	select MACH_JZ4750
+	select MACH_JZ4755
+	select MACH_JZ4760
 	select MACH_JZ4770
+	select MACH_JZ4775
 	select MACH_JZ4780
 	select MACH_X1000
+	select MACH_X1500
+	select MACH_X1600
 	select MACH_X1830
+	select MACH_X2000
+	select MACH_X2100
+	select MACH_X2600
 
 choice
 	prompt "Machine type"
@@ -62,8 +71,10 @@ config MACH_JZ4770
 config MACH_JZ4780
 	bool
 	select MIPS_CPU_SCACHE
+	select GENERIC_CLOCKEVENTS_BROADCAST if SMP
 	select SYS_HAS_CPU_MIPS32_R2
 	select SYS_SUPPORTS_HIGHMEM
+	select SYS_SUPPORTS_SMP
 
 config MACH_X1000
 	bool
@@ -71,8 +82,35 @@ config MACH_X1000
 	select SYS_HAS_CPU_MIPS32_R2
 	select SYS_SUPPORTS_HIGHMEM
 
+config MACH_X1600
+	bool
+	select SYS_HAS_CPU_MIPS32_R2
+	select SYS_SUPPORTS_HIGHMEM
+
 config MACH_X1830
 	bool
 	select MIPS_CPU_SCACHE
 	select SYS_HAS_CPU_MIPS32_R2
 	select SYS_SUPPORTS_HIGHMEM
+
+config MACH_X2000
+	bool
+	select MIPS_CPU_SCACHE
+	select GENERIC_CLOCKEVENTS_BROADCAST if SMP
+	select SYS_HAS_CPU_MIPS32_R5
+	select SYS_SUPPORTS_HIGHMEM
+	select SYS_SUPPORTS_SMP
+	select MIPS_O32_FP64_SUPPORT
+	select WEAK_ORDERING
+	select WEAK_REORDERING_BEYOND_LLSC
+
+config MACH_X2600
+	bool
+	select MIPS_CPU_SCACHE
+	select GENERIC_CLOCKEVENTS_BROADCAST if SMP
+	select SYS_HAS_CPU_MIPS32_R5
+	select SYS_SUPPORTS_HIGHMEM
+	select SYS_SUPPORTS_SMP
+	select MIPS_O32_FP64_SUPPORT
+	select WEAK_ORDERING
+	select WEAK_REORDERING_BEYOND_LLSC
Signature · 1 line
2.56.0
maintainer
#18 lore raw

Re: [PATCH v2 7/9] MIPS: Ingenic: Expand Kconfig for X1600, X2000, and X2600 SoCs

Le jeudi 12 novembre 2026 à 09:10 +0100, H. Nikolaus Schaller a écrit :

Quoted 10 lines

Expand MACH_INGENIC_GENERIC to select all existing and newly added machine variants (including JZ4750/55/60/75, X1500, X1600, X2000, X2100, and X2600).

Introduce individual Kconfig blocks for the following SoCs:
- MACH_X1600: Selects MIPS32_R2 and HIGHMEM.
- MACH_X2000 / MACH_X2600: Selects MIPS32_R5, SMP support, broadcast clockevents, O32 FP64 support, and weak ordering optimizations.

In parts developed by Reimu NotMoe <[email protected]>

Co-developed-by? (+ his Sob)

Quoted 20 lines

---
 arch/mips/ingenic/Kconfig | 38 ++++++++++++++++++++++++++++++++++++++

 1 file changed, 38 insertions(+)
arch/mips/ingenic/Kconfig +0 -0
index f595b339a4b84..5d351b42bbc9b 100644
@@ -5,10 +5,19 @@ config MACH_INGENIC_GENERIC

  select MACH_INGENIC
  select MACH_JZ4740
  select MACH_JZ4725B

arch/mips/ingenic/Kconfig +3 -0
+	select MACH_JZ4750
+	select MACH_JZ4755
+	select MACH_JZ4760

  select MACH_JZ4770

arch/mips/ingenic/Kconfig +1 -0
+	select MACH_JZ4775

  select MACH_JZ4780
  select MACH_X1000

arch/mips/ingenic/Kconfig config MACH_INGENIC_GENERIC · line 15

Quoted 1 line
arch/mips/ingenic/Kconfig +1 -0
+	select MACH_X1500

You select MACH_X1500 but there's no CONFIG_MACH_X1500 symbol. No entry in enum ingenic_machine_type either, or compatible string for it.

Your patch didn't add a CONFIG_MACH_X2100 either.

Quoted 33 lines
arch/mips/ingenic/Kconfig +1 -0
+	select MACH_X1600

  select MACH_X1830

arch/mips/ingenic/Kconfig +3 -0
+	select MACH_X2000
+	select MACH_X2100
+	select MACH_X2600


 choice
  prompt "Machine type"

arch/mips/ingenic/Kconfig +0 -0
@@ -62,8 +71,10 @@ config MACH_JZ4770

 config MACH_JZ4780
  bool
  select MIPS_CPU_SCACHE

arch/mips/ingenic/Kconfig +1 -0
+	select GENERIC_CLOCKEVENTS_BROADCAST if SMP

  select SYS_HAS_CPU_MIPS32_R2
  select SYS_SUPPORTS_HIGHMEM

arch/mips/ingenic/Kconfig +1 -0
+	select SYS_SUPPORTS_SMP


 config MACH_X1000
  bool

arch/mips/ingenic/Kconfig +0 -0
@@ -71,8 +82,35 @@ config MACH_X1000

  select SYS_HAS_CPU_MIPS32_R2
  select SYS_SUPPORTS_HIGHMEM

arch/mips/ingenic/Kconfig +5 -0
+config MACH_X1600
+	bool
+	select SYS_HAS_CPU_MIPS32_R2
+	select SYS_SUPPORTS_HIGHMEM
+

 config MACH_X1830
  bool
  select MIPS_CPU_SCACHE
  select SYS_HAS_CPU_MIPS32_R2
  select SYS_SUPPORTS_HIGHMEM

arch/mips/ingenic/Kconfig config MACH_X1000 · line 103

Quoted 9 lines
arch/mips/ingenic/Kconfig +9 -0
+
+config MACH_X2000
+	bool
+	select MIPS_CPU_SCACHE
+	select GENERIC_CLOCKEVENTS_BROADCAST if SMP
+	select SYS_HAS_CPU_MIPS32_R5
+	select SYS_SUPPORTS_HIGHMEM
+	select SYS_SUPPORTS_SMP
+	select MIPS_O32_FP64_SUPPORT

I know MIPS_O32_FP64_SUPPORT was supported on the JZ4770, and most likely on the JZ4760 too (the first one with a FPU AFAIK). So this could probably be moved to CONFIG_MACH_INGENIC.

Cheers,
-Paul

Quoted 13 lines
arch/mips/ingenic/Kconfig +13 -0
+	select WEAK_ORDERING
+	select WEAK_REORDERING_BEYOND_LLSC
+
+config MACH_X2600
+	bool
+	select MIPS_CPU_SCACHE
+	select GENERIC_CLOCKEVENTS_BROADCAST if SMP
+	select SYS_HAS_CPU_MIPS32_R5
+	select SYS_SUPPORTS_HIGHMEM
+	select SYS_SUPPORTS_SMP
+	select MIPS_O32_FP64_SUPPORT
+	select WEAK_ORDERING
+	select WEAK_REORDERING_BEYOND_LLSC
#19 lore raw

Re: [PATCH v2 7/9] MIPS: Ingenic: Expand Kconfig for X1600, X2000, and X2600 SoCs

Quoted 9 lines

Am 12.11.2026 um 15:16 schrieb Paul Cercueil <[email protected]>:

Quoted 4 lines · H. Nikolaus Schaller

Expand MACH_INGENIC_GENERIC to select all existing and newly added machine variants (including JZ4750/55/60/75, X1500, X1600, X2000, X2100, and X2600).

Quoted 8 lines
Quoted 6 lines

Introduce individual Kconfig blocks for the following SoCs:
- MACH_X1600: Selects MIPS32_R2 and HIGHMEM.
- MACH_X2000 / MACH_X2600: Selects MIPS32_R5, SMP support, broadcast clockevents, O32 FP64 support, and weak ordering optimizations.

In parts developed by Reimu NotMoe <[email protected]>

Co-developed-by? (+ his Sob)

I have to check what he really has developed. I think we did take this from his github tree (https://github.com/Ingenic-community/linux/tree/v6.1.y) but were never in contact with him. So I am not sure if we can get a SoB:.

Quoted 26 lines
Quoted 12 lines
---
 arch/mips/ingenic/Kconfig | 38 ++++++++++++++++++++++++++++++++++++++
 1 file changed, 38 insertions(+)
arch/mips/ingenic/Kconfig +0 -0
index f595b339a4b84..5d351b42bbc9b 100644
@@ -5,10 +5,19 @@ config MACH_INGENIC_GENERIC
  select MACH_INGENIC

arch/mips/ingenic/Kconfig config MACH_INGENIC_GENERIC · line 15

Quoted 10 lines
arch/mips/ingenic/Kconfig +5 -0
  select MACH_JZ4740
  select MACH_JZ4725B
+ select MACH_JZ4750
+ select MACH_JZ4755
+ select MACH_JZ4760
  select MACH_JZ4770
+ select MACH_JZ4775
  select MACH_JZ4780
  select MACH_X1000
+ select MACH_X1500

You select MACH_X1500 but there's no CONFIG_MACH_X1500 symbol. No entry in enum ingenic_machine_type either, or compatible string for it.

Quoted 1 line

Your patch didn't add a CONFIG_MACH_X2100 either.

Ok, that may be from taking the Reimu NotMoe patch without cross-checking.

Will be removed in V3.

Quoted 39 lines
Quoted 39 lines
arch/mips/ingenic/Kconfig +16 -0
+ select MACH_X1600
  select MACH_X1830
+ select MACH_X2000
+ select MACH_X2100
+ select MACH_X2600
 
 choice
  prompt "Machine type"
@@ -62,8 +71,10 @@ config MACH_JZ4770
 config MACH_JZ4780
  bool
  select MIPS_CPU_SCACHE
+ select GENERIC_CLOCKEVENTS_BROADCAST if SMP
  select SYS_HAS_CPU_MIPS32_R2
  select SYS_SUPPORTS_HIGHMEM
+ select SYS_SUPPORTS_SMP
 
 config MACH_X1000
  bool
@@ -71,8 +82,35 @@ config MACH_X1000
  select SYS_HAS_CPU_MIPS32_R2
  select SYS_SUPPORTS_HIGHMEM
 
+config MACH_X1600
+ bool
+ select SYS_HAS_CPU_MIPS32_R2
+ select SYS_SUPPORTS_HIGHMEM
+
 config MACH_X1830
  bool
  select MIPS_CPU_SCACHE
  select SYS_HAS_CPU_MIPS32_R2
  select SYS_SUPPORTS_HIGHMEM
+
+config MACH_X2000
+ bool
+ select MIPS_CPU_SCACHE
+ select GENERIC_CLOCKEVENTS_BROADCAST if SMP
Quoted 8 lines

arch/mips/ingenic/Kconfig config MACH_X1000 · line 103

Quoted 4 lines
arch/mips/ingenic/Kconfig +4 -0
+ select SYS_HAS_CPU_MIPS32_R5
+ select SYS_SUPPORTS_HIGHMEM
+ select SYS_SUPPORTS_SMP
+ select MIPS_O32_FP64_SUPPORT

I know MIPS_O32_FP64_SUPPORT was supported on the JZ4770, and most likely on the JZ4760 too (the first one with a FPU AFAIK). So this could probably be moved to CONFIG_MACH_INGENIC.

That should be clearly checked. If I find a little time I can add it or it can be done later.

BR and thanks,
Nikolaus

Quoted 17 lines

Cheers,
-Paul

Quoted 13 lines
arch/mips/ingenic/Kconfig +13 -0
+ select WEAK_ORDERING
+ select WEAK_REORDERING_BEYOND_LLSC
+
+config MACH_X2600
+ bool
+ select MIPS_CPU_SCACHE
+ select GENERIC_CLOCKEVENTS_BROADCAST if SMP
+ select SYS_HAS_CPU_MIPS32_R5
+ select SYS_SUPPORTS_HIGHMEM
+ select SYS_SUPPORTS_SMP
+ select MIPS_O32_FP64_SUPPORT
+ select WEAK_ORDERING
+ select WEAK_REORDERING_BEYOND_LLSC
#20 lore raw

Re: [PATCH v2 7/9] MIPS: Ingenic: Expand Kconfig for X1600, X2000, and X2600 SoCs

In reply to H. Nikolaus Schaller

arch/mips/ingenic/Kconfig config MACH_X1000 · line 111

Quoted 5 lines · H. Nikolaus Schaller
arch/mips/ingenic/Kconfig +5 -0
+config MACH_X2600
+	bool
+	select MIPS_CPU_SCACHE
+	select GENERIC_CLOCKEVENTS_BROADCAST if SMP
+	select SYS_HAS_CPU_MIPS32_R5

With only R5 here, a kernel for just the X2600 is built for MIPS32r5, and it dies at the first interrupt return:

Reserved instruction in kernel code[#1]:
epc : arch_local_irq_restore+0xc/0x18
PrId : 00132000 (Ingenic XBurst II)

RESTORE_SP_AND_RET uses eretnc on R5, which XBurst II doesn't have (it predates R5.04). Zhou fixed this in 2020, using eretnc only on R6:
"MIPS: stackframe: Fix the problem caused by the eretnc instruction." With it the same kernel boots. Since this patch makes an X2600 kernel R5-only, I think Zhou's fix should come before it in the series.

Hunter

#21 lore raw

PATCH v2 8/9 MIPS: Ingenic: Expand Kconfig by MACH_JZ4730

In reply to H. Nikolaus Schaller

The jz4730 is a single core SoC by Ingenic with peripherals like LCD controller, NAND controller, Timers, PWM, GPIOs

---
 arch/mips/ingenic/Kconfig | 5 +++++
 1 file changed, 5 insertions(+)
arch/mips/ingenic/Kconfig +5 -0
index 5d351b42bbc9b..62b494ad77a61 100644
@@ -3,6 +3,7 @@
 config MACH_INGENIC_GENERIC
 	bool
 	select MACH_INGENIC
+	select MACH_JZ4730
 	select MACH_JZ4740
 	select MACH_JZ4725B
 	select MACH_JZ4750
@@ -58,6 +59,10 @@ config MACH_JZ4725B
 	bool
 	select SYS_HAS_CPU_MIPS32_R1
 
+config MACH_JZ4730
+	bool
+	select SYS_HAS_CPU_MIPS32_R1
+
 config MACH_JZ4740
 	bool
 	select SYS_HAS_CPU_MIPS32_R1
Signature · 1 line
2.56.0
#22 lore raw

PATCH v2 9/9 MIPS: Ingenic: Add Platform makefile and load addresses for X1600, X2000, X2600

In reply to H. Nikolaus Schaller

Introduce the arch/mips/ingenic/Platform file to define the default kernel load addresses for X1600, X2000 and X2600 SoCs.

All platforms are configured to use 0xffffffff80010000.

---
 arch/mips/ingenic/Platform | 3 +++
 1 file changed, 3 insertions(+)
 create mode 100644 arch/mips/ingenic/Platform
arch/mips/ingenic/Platform new file +3 -0
new file mode 100644
index 0000000000000..bffb4529d799d
@@ -0,0 +1,3 @@
+load-$(CONFIG_MACH_X1600)	:= 0xffffffff80010000
+load-$(CONFIG_MACH_X2000)	:= 0xffffffff80010000
+load-$(CONFIG_MACH_X2600)	:= 0xffffffff80010000
Signature · 1 line
2.56.0