[PATCH 0/4] memory: tegra: Add Tegra264 support

[PATCH 0/4] memory: tegra: Add Tegra264 support

[Thierry Reding]

From: Thierry Reding <treding@nvidia.com>

This set of patches extends the DT bindings for the memory controller
and external memory controller for Tegra264 and add the necessary DT
headers with memory client and stream ID definitions.

The driver changes in patch 4 are mostly an extension of existing code
and the bulk consists of the memory client table for the new chip as
well as the bandwidth manager calculations.

Thierry

Sumit Gupta (3):
  dt-bindings: memory: tegra: Add Tegra264 support
  dt-bindings: memory: tegra: Add Tegra264 definitions
  memory: tegra: Add Tegra264 MC and EMC support

Thierry Reding (1):
  dt-bindings: memory: tegra: Add Tegra264 stream IDs

 .../nvidia,tegra186-mc.yaml                   |  65 +++-
 drivers/memory/tegra/Makefile                 |   2 +
 drivers/memory/tegra/mc.c                     |   5 +-
 drivers/memory/tegra/mc.h                     |   9 +-
 drivers/memory/tegra/tegra186-emc.c           |   5 +-
 drivers/memory/tegra/tegra186.c               |  17 +-
 drivers/memory/tegra/tegra264-bwmgr.h         |  50 +++
 drivers/memory/tegra/tegra264.c               | 313 ++++++++++++++++++
 include/dt-bindings/memory/nvidia,tegra264.h  | 136 ++++++++
 9 files changed, 594 insertions(+), 8 deletions(-)
 create mode 100644 drivers/memory/tegra/tegra264-bwmgr.h
 create mode 100644 drivers/memory/tegra/tegra264.c
 create mode 100644 include/dt-bindings/memory/nvidia,tegra264.h

-- 
2.50.0

[PATCH 1/4] dt-bindings: memory: tegra: Add Tegra264 support

[Thierry Reding]

From: Sumit Gupta <sumitg@nvidia.com>

Add bindings for the Memory Controller (MC) and External Memory
Controller (EMC) found on the Tegra264 SoC. Tegra264 SoC has a different
number of interrupt lines for MC sub-units: UCF_SOC, hub, hub common,
syncpoint and MC channel. The total number of interrupt lines is eight.
Update maxItems for MC interrupts accordingly.

Signed-off-by: Sumit Gupta <sumitg@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 .../nvidia,tegra186-mc.yaml                   | 65 ++++++++++++++++++-
 1 file changed, 63 insertions(+), 2 deletions(-)

diff --git a/Documentation/devicetree/bindings/memory-controllers/nvidia,tegra186-mc.yaml b/Documentation/devicetree/bindings/memory-controllers/nvidia,tegra186-mc.yaml
index 935d63d181d9..52575a16349b 100644
--- a/Documentation/devicetree/bindings/memory-controllers/nvidia,tegra186-mc.yaml
+++ b/Documentation/devicetree/bindings/memory-controllers/nvidia,tegra186-mc.yaml
@@ -32,6 +32,7 @@ properties:
           - nvidia,tegra186-mc
           - nvidia,tegra194-mc
           - nvidia,tegra234-mc
+          - nvidia,tegra264-mc
 
   reg:
     minItems: 6
@@ -42,8 +43,12 @@ properties:
     maxItems: 18
 
   interrupts:
-    items:
-      - description: MC general interrupt
+    minItems: 1
+    maxItems: 8
+
+  interrupt-names:
+    minItems: 1
+    maxItems: 8
 
   "#address-cells":
     const: 2
@@ -74,6 +79,7 @@ patternProperties:
               - nvidia,tegra186-emc
               - nvidia,tegra194-emc
               - nvidia,tegra234-emc
+              - nvidia,tegra264-emc
 
       reg:
         minItems: 1
@@ -127,6 +133,15 @@ patternProperties:
             reg:
               minItems: 2
 
+      - if:
+          properties:
+            compatible:
+              const: nvidia,tegra264-emc
+        then:
+          properties:
+            reg:
+              minItems: 2
+
     additionalProperties: false
 
     required:
@@ -220,6 +235,52 @@ allOf:
             - const: ch14
             - const: ch15
 
+  - if:
+      properties:
+        compatible:
+          const: nvidia,tegra264-mc
+    then:
+      properties:
+        reg:
+          minItems: 17
+          description: 17 memory controller channels
+
+        reg-names:
+          items:
+            - const: broadcast
+            - const: ch0
+            - const: ch1
+            - const: ch2
+            - const: ch3
+            - const: ch4
+            - const: ch5
+            - const: ch6
+            - const: ch7
+            - const: ch8
+            - const: ch9
+            - const: ch10
+            - const: ch11
+            - const: ch12
+            - const: ch13
+            - const: ch14
+            - const: ch15
+
+        interrupts:
+          minItems: 8
+          maxItems: 8
+          description: One interrupt line for each MC component
+
+        interrupt-names:
+          items:
+            - const: mcf
+            - const: hub1
+            - const: hub2
+            - const: hub3
+            - const: hub4
+            - const: hub5
+            - const: sbs
+            - const: channel
+
 additionalProperties: false
 
 required:
-- 
2.50.0

Re: [PATCH 1/4] dt-bindings: memory: tegra: Add Tegra264 support

[Krzysztof Kozlowski]

On 08/07/2025 12:52, Thierry Reding wrote:
>    interrupts:
> -    items:
> -      - description: MC general interrupt
> +    minItems: 1
> +    maxItems: 8
> +
> +  interrupt-names:
> +    minItems: 1
> +    maxItems: 8
>  
>    "#address-cells":
>      const: 2
> @@ -74,6 +79,7 @@ patternProperties:
>                - nvidia,tegra186-emc
>                - nvidia,tegra194-emc
>                - nvidia,tegra234-emc
> +              - nvidia,tegra264-emc
>  
>        reg:
>          minItems: 1
> @@ -127,6 +133,15 @@ patternProperties:
>              reg:
>                minItems: 2
>  
> +      - if:
> +          properties:
> +            compatible:
> +              const: nvidia,tegra264-emc
> +        then:
> +          properties:
> +            reg:
> +              minItems: 2
> +
>      additionalProperties: false
>  
>      required:
> @@ -220,6 +235,52 @@ allOf:
>              - const: ch14
>              - const: ch15
>  
> +  - if:
> +      properties:
> +        compatible:
> +          const: nvidia,tegra264-mc
> +    then:
> +      properties:
> +        reg:
> +          minItems: 17

Missing maxItems

> +          description: 17 memory controller channels
> +
> +        reg-names:
> +          items:
> +            - const: broadcast
> +            - const: ch0
> +            - const: ch1
> +            - const: ch2
> +            - const: ch3
> +            - const: ch4
> +            - const: ch5
> +            - const: ch6
> +            - const: ch7
> +            - const: ch8
> +            - const: ch9
> +            - const: ch10
> +            - const: ch11
> +            - const: ch12
> +            - const: ch13
> +            - const: ch14
> +            - const: ch15
> +
> +        interrupts:
> +          minItems: 8
> +          maxItems: 8
> +          description: One interrupt line for each MC component
> +
> +        interrupt-names:
> +          items:
> +            - const: mcf
> +            - const: hub1
> +            - const: hub2
> +            - const: hub3
> +            - const: hub4
> +            - const: hub5
> +            - const: sbs
> +            - const: channel


Missing constraints for interrupts and interrupt-names for all other
variants. Now this patch claims they all have 8 interrupts with any name.



Best regards,
Krzysztof

Re: [PATCH 1/4] dt-bindings: memory: tegra: Add Tegra264 support

[Thierry Reding]

[-- Attachment #1: Type: text/plain, Size: 3209 bytes --]

On Wed, Jul 09, 2025 at 08:19:51PM +0200, Krzysztof Kozlowski wrote:
> On 08/07/2025 12:52, Thierry Reding wrote:
> >    interrupts:
> > -    items:
> > -      - description: MC general interrupt
> > +    minItems: 1
> > +    maxItems: 8
> > +
> > +  interrupt-names:
> > +    minItems: 1
> > +    maxItems: 8
> >  
> >    "#address-cells":
> >      const: 2
> > @@ -74,6 +79,7 @@ patternProperties:
> >                - nvidia,tegra186-emc
> >                - nvidia,tegra194-emc
> >                - nvidia,tegra234-emc
> > +              - nvidia,tegra264-emc
> >  
> >        reg:
> >          minItems: 1
> > @@ -127,6 +133,15 @@ patternProperties:
> >              reg:
> >                minItems: 2
> >  
> > +      - if:
> > +          properties:
> > +            compatible:
> > +              const: nvidia,tegra264-emc
> > +        then:
> > +          properties:
> > +            reg:
> > +              minItems: 2
> > +
> >      additionalProperties: false
> >  
> >      required:
> > @@ -220,6 +235,52 @@ allOf:
> >              - const: ch14
> >              - const: ch15
> >  
> > +  - if:
> > +      properties:
> > +        compatible:
> > +          const: nvidia,tegra264-mc
> > +    then:
> > +      properties:
> > +        reg:
> > +          minItems: 17
> 
> Missing maxItems

My recollection was that maxItems didn't have to be specified if we
already have minItems and they are both equal. That said, I see now
there are a few cases in existing bindings where both are used in
conjunction, so I must be misremembering. I've added "maxItems: 17".

> 
> > +          description: 17 memory controller channels
> > +
> > +        reg-names:
> > +          items:
> > +            - const: broadcast
> > +            - const: ch0
> > +            - const: ch1
> > +            - const: ch2
> > +            - const: ch3
> > +            - const: ch4
> > +            - const: ch5
> > +            - const: ch6
> > +            - const: ch7
> > +            - const: ch8
> > +            - const: ch9
> > +            - const: ch10
> > +            - const: ch11
> > +            - const: ch12
> > +            - const: ch13
> > +            - const: ch14
> > +            - const: ch15
> > +
> > +        interrupts:
> > +          minItems: 8
> > +          maxItems: 8
> > +          description: One interrupt line for each MC component
> > +
> > +        interrupt-names:
> > +          items:
> > +            - const: mcf
> > +            - const: hub1
> > +            - const: hub2
> > +            - const: hub3
> > +            - const: hub4
> > +            - const: hub5
> > +            - const: sbs
> > +            - const: channel
> 
> 
> Missing constraints for interrupts and interrupt-names for all other
> variants. Now this patch claims they all have 8 interrupts with any name.

I'll add back the interrupts property for the other variants. For the
interrupt-names it's slightly more tricky because on older variants we
don't need it since there's only one interrupt. It looks like I can do
"interrupt-names: false" in those cases.

Thanks,
Thierry

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [PATCH 1/4] dt-bindings: memory: tegra: Add Tegra264 support

[Krzysztof Kozlowski]

On 09/07/2025 22:36, Thierry Reding wrote:
> On Wed, Jul 09, 2025 at 08:19:51PM +0200, Krzysztof Kozlowski wrote:
>> On 08/07/2025 12:52, Thierry Reding wrote:
>>>    interrupts:
>>> -    items:
>>> -      - description: MC general interrupt
>>> +    minItems: 1
>>> +    maxItems: 8
>>> +
>>> +  interrupt-names:
>>> +    minItems: 1
>>> +    maxItems: 8
>>>  
>>>    "#address-cells":
>>>      const: 2
>>> @@ -74,6 +79,7 @@ patternProperties:
>>>                - nvidia,tegra186-emc
>>>                - nvidia,tegra194-emc
>>>                - nvidia,tegra234-emc
>>> +              - nvidia,tegra264-emc
>>>  
>>>        reg:
>>>          minItems: 1
>>> @@ -127,6 +133,15 @@ patternProperties:
>>>              reg:
>>>                minItems: 2
>>>  
>>> +      - if:
>>> +          properties:
>>> +            compatible:
>>> +              const: nvidia,tegra264-emc
>>> +        then:
>>> +          properties:
>>> +            reg:
>>> +              minItems: 2
>>> +
>>>      additionalProperties: false
>>>  
>>>      required:
>>> @@ -220,6 +235,52 @@ allOf:
>>>              - const: ch14
>>>              - const: ch15
>>>  
>>> +  - if:
>>> +      properties:
>>> +        compatible:
>>> +          const: nvidia,tegra264-mc
>>> +    then:
>>> +      properties:
>>> +        reg:
>>> +          minItems: 17
>>
>> Missing maxItems
> 
> My recollection was that maxItems didn't have to be specified if we
> already have minItems and they are both equal. That said, I see now

There was never such rule.

> there are a few cases in existing bindings where both are used in
> conjunction, so I must be misremembering. I've added "maxItems: 17".
> 
>>
>>> +          description: 17 memory controller channels
>>> +
>>> +        reg-names:
>>> +          items:
>>> +            - const: broadcast
>>> +            - const: ch0
>>> +            - const: ch1
>>> +            - const: ch2
>>> +            - const: ch3
>>> +            - const: ch4
>>> +            - const: ch5
>>> +            - const: ch6
>>> +            - const: ch7
>>> +            - const: ch8
>>> +            - const: ch9
>>> +            - const: ch10
>>> +            - const: ch11
>>> +            - const: ch12
>>> +            - const: ch13
>>> +            - const: ch14
>>> +            - const: ch15
>>> +
>>> +        interrupts:
>>> +          minItems: 8
>>> +          maxItems: 8
>>> +          description: One interrupt line for each MC component
>>> +
>>> +        interrupt-names:
>>> +          items:
>>> +            - const: mcf
>>> +            - const: hub1
>>> +            - const: hub2
>>> +            - const: hub3
>>> +            - const: hub4
>>> +            - const: hub5
>>> +            - const: sbs
>>> +            - const: channel
>>
>>
>> Missing constraints for interrupts and interrupt-names for all other
>> variants. Now this patch claims they all have 8 interrupts with any name.
> 
> I'll add back the interrupts property for the other variants. For the
> interrupt-names it's slightly more tricky because on older variants we
> don't need it since there's only one interrupt. It looks like I can do
> "interrupt-names: false" in those cases.

Yes.

> 
> Thanks,
> Thierry


Best regards,
Krzysztof

[PATCH 2/4] dt-bindings: memory: tegra: Add Tegra264 definitions

[Thierry Reding]

From: Sumit Gupta <sumitg@nvidia.com>

Add memory client ID defines for use by the interconnects property in
the device tree and tegra_mc_client table in the MC driver. Note that
these IDs are defined by the hardware, so the numbering doesn't start
at 0 and contains holes.

Signed-off-by: Sumit Gupta <sumitg@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 include/dt-bindings/memory/nvidia,tegra264.h | 86 ++++++++++++++++++++
 1 file changed, 86 insertions(+)
 create mode 100644 include/dt-bindings/memory/nvidia,tegra264.h

diff --git a/include/dt-bindings/memory/nvidia,tegra264.h b/include/dt-bindings/memory/nvidia,tegra264.h
new file mode 100644
index 000000000000..d6cb0c9698f2
--- /dev/null
+++ b/include/dt-bindings/memory/nvidia,tegra264.h
@@ -0,0 +1,86 @@
+/* SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) */
+/* Copyright (c) 2025, NVIDIA CORPORATION. All rights reserved. */
+
+#ifndef DT_BINDINGS_MEMORY_NVIDIA_TEGRA264_H
+#define DT_BINDINGS_MEMORY_NVIDIA_TEGRA264_H
+
+/*
+ * memory client IDs
+ */
+
+/* HOST1X read client */
+#define TEGRA264_MEMORY_CLIENT_HOST1XR		0x16
+/* VIC read client */
+#define TEGRA264_MEMORY_CLIENT_VICR		0x6c
+/* VIC Write client */
+#define TEGRA264_MEMORY_CLIENT_VICW		0x6d
+/* VI R5 Write client */
+#define TEGRA264_MEMORY_CLIENT_VIW		0x72
+#define TEGRA264_MEMORY_CLIENT_NVDECSRD2MC	0x78
+#define TEGRA264_MEMORY_CLIENT_NVDECSWR2MC	0x79
+/* Audio processor(APE) Read client */
+#define TEGRA264_MEMORY_CLIENT_APER		0x7a
+/* Audio processor(APE) Write client */
+#define TEGRA264_MEMORY_CLIENT_APEW		0x7b
+/* Audio DMA Read client */
+#define TEGRA264_MEMORY_CLIENT_APEDMAR		0x9f
+/* Audio DMA Write client */
+#define TEGRA264_MEMORY_CLIENT_APEDMAW		0xa0
+#define TEGRA264_MEMORY_CLIENT_GPUR02MC		0xb6
+#define TEGRA264_MEMORY_CLIENT_GPUW02MC		0xb7
+/* VI Falcon Read client */
+#define TEGRA264_MEMORY_CLIENT_VIFALCONR	0xbc
+/* VI Falcon Write client */
+#define TEGRA264_MEMORY_CLIENT_VIFALCONW	0xbd
+/* Read Client of RCE */
+#define TEGRA264_MEMORY_CLIENT_RCER		0xd2
+/* Write client of RCE */
+#define TEGRA264_MEMORY_CLIENT_RCEW		0xd3
+/* PCIE0/MSI Write clients */
+#define TEGRA264_MEMORY_CLIENT_PCIE0W		0xd9
+/* PCIE1/RPX4 Read clients */
+#define TEGRA264_MEMORY_CLIENT_PCIE1R		0xda
+/* PCIE1/RPX4 Write clients */
+#define TEGRA264_MEMORY_CLIENT_PCIE1W		0xdb
+/* PCIE2/DMX4 Read clients */
+#define TEGRA264_MEMORY_CLIENT_PCIE2AR		0xdc
+/* PCIE2/DMX4 Write clients */
+#define TEGRA264_MEMORY_CLIENT_PCIE2AW		0xdd
+/* PCIE3/RPX4 Read clients */
+#define TEGRA264_MEMORY_CLIENT_PCIE3R		0xde
+/* PCIE3/RPX4 Write clients */
+#define TEGRA264_MEMORY_CLIENT_PCIE3W		0xdf
+/* PCIE4/DMX8 Read clients */
+#define TEGRA264_MEMORY_CLIENT_PCIE4R		0xe0
+/* PCIE4/DMX8 Write clients */
+#define TEGRA264_MEMORY_CLIENT_PCIE4W		0xe1
+/* PCIE5/DMX4 Read clients */
+#define TEGRA264_MEMORY_CLIENT_PCIE5R		0xe2
+/* PCIE5/DMX4 Write clients */
+#define TEGRA264_MEMORY_CLIENT_PCIE5W		0xe3
+/* UFS Read client */
+#define TEGRA264_MEMORY_CLIENT_UFSR		0x15c
+/* UFS write client */
+#define TEGRA264_MEMORY_CLIENT_UFSW		0x15d
+/* HDA Read client */
+#define TEGRA264_MEMORY_CLIENT_HDAR		0x17c
+/* HDA Write client */
+#define TEGRA264_MEMORY_CLIENT_HDAW		0x17d
+/* Disp ISO Read Client */
+#define TEGRA264_MEMORY_CLIENT_DISPR		0x182
+/* MGBE0 Read mccif */
+#define TEGRA264_MEMORY_CLIENT_MGBE0R		0x1a2
+/* MGBE0 Write mccif */
+#define TEGRA264_MEMORY_CLIENT_MGBE0W		0x1a3
+/* MGBE1 Read mccif */
+#define TEGRA264_MEMORY_CLIENT_MGBE1R		0x1a4
+/* MGBE1 Write mccif */
+#define TEGRA264_MEMORY_CLIENT_MGBE1W		0x1a5
+/* VI1 R5 Write client */
+#define TEGRA264_MEMORY_CLIENT_VI1W		0x1a6
+/* SDMMC0 Read mccif */
+#define TEGRA264_MEMORY_CLIENT_SDMMC0R		0x1c2
+/* SDMMC0 Write mccif */
+#define TEGRA264_MEMORY_CLIENT_SDMMC0W		0x1c3
+
+#endif /* DT_BINDINGS_MEMORY_NVIDIA_TEGRA264_H */
-- 
2.50.0

Re: [PATCH 2/4] dt-bindings: memory: tegra: Add Tegra264 definitions

[Krzysztof Kozlowski]

On 08/07/2025 12:52, Thierry Reding wrote:
> From: Sumit Gupta <sumitg@nvidia.com>
> 
> Add memory client ID defines for use by the interconnects property in
> the device tree and tegra_mc_client table in the MC driver. Note that
> these IDs are defined by the hardware, so the numbering doesn't start
> at 0 and contains holes.
> 
> Signed-off-by: Sumit Gupta <sumitg@nvidia.com>
> Signed-off-by: Thierry Reding <treding@nvidia.com>
> ---
>  include/dt-bindings/memory/nvidia,tegra264.h | 86 ++++++++++++++++++++
>  1 file changed, 86 insertions(+)
>  create mode 100644 include/dt-bindings/memory/nvidia,tegra264.h
> 

This looks like values for new MC added in patch #1, so this should be
squashed there. If this not the case, commit msg should be more obvious
on that.

Best regards,
Krzysztof

Re: [PATCH 2/4] dt-bindings: memory: tegra: Add Tegra264 definitions

[Thierry Reding]

[-- Attachment #1: Type: text/plain, Size: 1090 bytes --]

On Wed, Jul 09, 2025 at 08:20:44PM +0200, Krzysztof Kozlowski wrote:
> On 08/07/2025 12:52, Thierry Reding wrote:
> > From: Sumit Gupta <sumitg@nvidia.com>
> > 
> > Add memory client ID defines for use by the interconnects property in
> > the device tree and tegra_mc_client table in the MC driver. Note that
> > these IDs are defined by the hardware, so the numbering doesn't start
> > at 0 and contains holes.
> > 
> > Signed-off-by: Sumit Gupta <sumitg@nvidia.com>
> > Signed-off-by: Thierry Reding <treding@nvidia.com>
> > ---
> >  include/dt-bindings/memory/nvidia,tegra264.h | 86 ++++++++++++++++++++
> >  1 file changed, 86 insertions(+)
> >  create mode 100644 include/dt-bindings/memory/nvidia,tegra264.h
> > 
> 
> This looks like values for new MC added in patch #1, so this should be
> squashed there. If this not the case, commit msg should be more obvious
> on that.

Will do. I thought splitting this up would make it a bit easier to
handle dependencies, but ultimately it doesn't matter very much. These
are all fairly small files overall.

Thierry

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

[PATCH 3/4] dt-bindings: memory: tegra: Add Tegra264 stream IDs

[Thierry Reding]

From: Thierry Reding <treding@nvidia.com>

Add the stream IDs for various hardware blocks found on Tegra264. These
are allocated as blocks of 256 IDs and each block can be subdivided for
additional fine-grained isolation if needed.

Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 include/dt-bindings/memory/nvidia,tegra264.h | 50 ++++++++++++++++++++
 1 file changed, 50 insertions(+)

diff --git a/include/dt-bindings/memory/nvidia,tegra264.h b/include/dt-bindings/memory/nvidia,tegra264.h
index d6cb0c9698f2..521405c01f84 100644
--- a/include/dt-bindings/memory/nvidia,tegra264.h
+++ b/include/dt-bindings/memory/nvidia,tegra264.h
@@ -4,6 +4,56 @@
 #ifndef DT_BINDINGS_MEMORY_NVIDIA_TEGRA264_H
 #define DT_BINDINGS_MEMORY_NVIDIA_TEGRA264_H
 
+#define TEGRA264_SID(x) ((x) << 8)
+
+/*
+ * SMMU stream IDs
+ */
+
+#define TEGRA264_SID_AON       TEGRA264_SID(0x01)
+#define TEGRA264_SID_APE       TEGRA264_SID(0x02)
+#define TEGRA264_SID_ETR       TEGRA264_SID(0x03)
+#define TEGRA264_SID_BPMP      TEGRA264_SID(0x04)
+#define TEGRA264_SID_DCE       TEGRA264_SID(0x05)
+#define TEGRA264_SID_EQOS      TEGRA264_SID(0x06)
+#define TEGRA264_SID_GPCDMA    TEGRA264_SID(0x08)
+#define TEGRA264_SID_DISP      TEGRA264_SID(0x09)
+#define TEGRA264_SID_HDA       TEGRA264_SID(0x0a)
+#define TEGRA264_SID_HOST1X    TEGRA264_SID(0x0b)
+#define TEGRA264_SID_ISP0      TEGRA264_SID(0x0c)
+#define TEGRA264_SID_ISP1      TEGRA264_SID(0x0d)
+#define TEGRA264_SID_PMA0      TEGRA264_SID(0x0e)
+#define TEGRA264_SID_FSI0      TEGRA264_SID(0x0f)
+#define TEGRA264_SID_FSI1      TEGRA264_SID(0x10)
+#define TEGRA264_SID_PVA       TEGRA264_SID(0x11)
+#define TEGRA264_SID_SDMMC0    TEGRA264_SID(0x12)
+#define TEGRA264_SID_MGBE0     TEGRA264_SID(0x13)
+#define TEGRA264_SID_MGBE1     TEGRA264_SID(0x14)
+#define TEGRA264_SID_MGBE2     TEGRA264_SID(0x15)
+#define TEGRA264_SID_MGBE3     TEGRA264_SID(0x16)
+#define TEGRA264_SID_MSSSEQ    TEGRA264_SID(0x17)
+#define TEGRA264_SID_SE        TEGRA264_SID(0x18)
+#define TEGRA264_SID_SEU1      TEGRA264_SID(0x19)
+#define TEGRA264_SID_SEU2      TEGRA264_SID(0x1a)
+#define TEGRA264_SID_SEU3      TEGRA264_SID(0x1b)
+#define TEGRA264_SID_PSC       TEGRA264_SID(0x1c)
+#define TEGRA264_SID_OESP      TEGRA264_SID(0x23)
+#define TEGRA264_SID_SB        TEGRA264_SID(0x24)
+#define TEGRA264_SID_XSPI0     TEGRA264_SID(0x25)
+#define TEGRA264_SID_TSEC      TEGRA264_SID(0x29)
+#define TEGRA264_SID_UFS       TEGRA264_SID(0x2a)
+#define TEGRA264_SID_RCE       TEGRA264_SID(0x2b)
+#define TEGRA264_SID_RCE1      TEGRA264_SID(0x2c)
+#define TEGRA264_SID_VI        TEGRA264_SID(0x2e)
+#define TEGRA264_SID_VI1       TEGRA264_SID(0x2f)
+#define TEGRA264_SID_VIC       TEGRA264_SID(0x30)
+#define TEGRA264_SID_XUSB_DEV  TEGRA264_SID(0x32)
+#define TEGRA264_SID_XUSB_DEV1 TEGRA264_SID(0x33)
+#define TEGRA264_SID_XUSB_DEV2 TEGRA264_SID(0x34)
+#define TEGRA264_SID_XUSB_DEV3 TEGRA264_SID(0x35)
+#define TEGRA264_SID_XUSB_DEV4 TEGRA264_SID(0x36)
+#define TEGRA264_SID_XUSB_DEV5 TEGRA264_SID(0x37)
+
 /*
  * memory client IDs
  */
-- 
2.50.0

Re: [PATCH 3/4] dt-bindings: memory: tegra: Add Tegra264 stream IDs

[Krzysztof Kozlowski]

On 08/07/2025 12:52, Thierry Reding wrote:
> From: Thierry Reding <treding@nvidia.com>
> 
> Add the stream IDs for various hardware blocks found on Tegra264. These
> are allocated as blocks of 256 IDs and each block can be subdivided for
> additional fine-grained isolation if needed.
> 
> Signed-off-by: Thierry Reding <treding@nvidia.com>
> ---
>  include/dt-bindings/memory/nvidia,tegra264.h | 50 ++++++++++++++++++++

Please squash it (you can take authorship and add co-developed), but
adding given bindings and bindings header for new device is one patch.
Not three.


Best regards,
Krzysztof

Re: [PATCH 3/4] dt-bindings: memory: tegra: Add Tegra264 stream IDs

[Thierry Reding]

[-- Attachment #1: Type: text/plain, Size: 717 bytes --]

On Wed, Jul 09, 2025 at 08:21:53PM +0200, Krzysztof Kozlowski wrote:
> On 08/07/2025 12:52, Thierry Reding wrote:
> > From: Thierry Reding <treding@nvidia.com>
> > 
> > Add the stream IDs for various hardware blocks found on Tegra264. These
> > are allocated as blocks of 256 IDs and each block can be subdivided for
> > additional fine-grained isolation if needed.
> > 
> > Signed-off-by: Thierry Reding <treding@nvidia.com>
> > ---
> >  include/dt-bindings/memory/nvidia,tegra264.h | 50 ++++++++++++++++++++
> 
> Please squash it (you can take authorship and add co-developed), but
> adding given bindings and bindings header for new device is one patch.
> Not three.

Will do.

Thanks,
Thierry

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

[PATCH 4/4] memory: tegra: Add Tegra264 MC and EMC support

[Thierry Reding]

From: Sumit Gupta <sumitg@nvidia.com>

Add support to enable Memory Controller (MC) and External Memory
Controller (EMC) drivers for Tegra264. The nodes for MC and EMC are
mostly the same as Tegra234 but differ in number of channels and
interrupt numbers.

The patch also adds the bandwidth manager definitions required for
Tegra264 and uses them to populate the memory client table. All of
these are needed to properly enable memory interconnect (ICC) support.

Signed-off-by: Sumit Gupta <sumitg@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/memory/tegra/Makefile         |   2 +
 drivers/memory/tegra/mc.c             |   5 +-
 drivers/memory/tegra/mc.h             |   9 +-
 drivers/memory/tegra/tegra186-emc.c   |   5 +-
 drivers/memory/tegra/tegra186.c       |  17 +-
 drivers/memory/tegra/tegra264-bwmgr.h |  50 ++++
 drivers/memory/tegra/tegra264.c       | 313 ++++++++++++++++++++++++++
 7 files changed, 395 insertions(+), 6 deletions(-)
 create mode 100644 drivers/memory/tegra/tegra264-bwmgr.h
 create mode 100644 drivers/memory/tegra/tegra264.c

diff --git a/drivers/memory/tegra/Makefile b/drivers/memory/tegra/Makefile
index 0750847dac3c..6334601e6120 100644
--- a/drivers/memory/tegra/Makefile
+++ b/drivers/memory/tegra/Makefile
@@ -10,6 +10,7 @@ tegra-mc-$(CONFIG_ARCH_TEGRA_210_SOC) += tegra210.o
 tegra-mc-$(CONFIG_ARCH_TEGRA_186_SOC) += tegra186.o
 tegra-mc-$(CONFIG_ARCH_TEGRA_194_SOC) += tegra186.o tegra194.o
 tegra-mc-$(CONFIG_ARCH_TEGRA_234_SOC) += tegra186.o tegra234.o
+tegra-mc-$(CONFIG_ARCH_TEGRA_264_SOC) += tegra186.o tegra264.o
 
 obj-$(CONFIG_TEGRA_MC) += tegra-mc.o
 
@@ -21,5 +22,6 @@ obj-$(CONFIG_TEGRA210_EMC) += tegra210-emc.o
 obj-$(CONFIG_ARCH_TEGRA_186_SOC) += tegra186-emc.o
 obj-$(CONFIG_ARCH_TEGRA_194_SOC) += tegra186-emc.o
 obj-$(CONFIG_ARCH_TEGRA_234_SOC) += tegra186-emc.o
+obj-$(CONFIG_ARCH_TEGRA_264_SOC) += tegra186-emc.o
 
 tegra210-emc-y := tegra210-emc-core.o tegra210-emc-cc-r21021.o
diff --git a/drivers/memory/tegra/mc.c b/drivers/memory/tegra/mc.c
index bd5b58f1fd42..6edb210287dc 100644
--- a/drivers/memory/tegra/mc.c
+++ b/drivers/memory/tegra/mc.c
@@ -1,6 +1,6 @@
 // SPDX-License-Identifier: GPL-2.0-only
 /*
- * Copyright (C) 2014 NVIDIA CORPORATION.  All rights reserved.
+ * Copyright (C) 2014-2025 NVIDIA CORPORATION.  All rights reserved.
  */
 
 #include <linux/clk.h>
@@ -48,6 +48,9 @@ static const struct of_device_id tegra_mc_of_match[] = {
 #endif
 #ifdef CONFIG_ARCH_TEGRA_234_SOC
 	{ .compatible = "nvidia,tegra234-mc", .data = &tegra234_mc_soc },
+#endif
+#ifdef CONFIG_ARCH_TEGRA_264_SOC
+	{ .compatible = "nvidia,tegra264-mc", .data = &tegra264_mc_soc },
 #endif
 	{ /* sentinel */ }
 };
diff --git a/drivers/memory/tegra/mc.h b/drivers/memory/tegra/mc.h
index c3f6655bec60..1d97cf4d3a94 100644
--- a/drivers/memory/tegra/mc.h
+++ b/drivers/memory/tegra/mc.h
@@ -1,6 +1,6 @@
 /* SPDX-License-Identifier: GPL-2.0-only */
 /*
- * Copyright (C) 2014 NVIDIA CORPORATION.  All rights reserved.
+ * Copyright (C) 2014-2025 NVIDIA CORPORATION.  All rights reserved.
  */
 
 #ifndef MEMORY_TEGRA_MC_H
@@ -182,6 +182,10 @@ extern const struct tegra_mc_soc tegra194_mc_soc;
 extern const struct tegra_mc_soc tegra234_mc_soc;
 #endif
 
+#ifdef CONFIG_ARCH_TEGRA_264_SOC
+extern const struct tegra_mc_soc tegra264_mc_soc;
+#endif
+
 #if defined(CONFIG_ARCH_TEGRA_3x_SOC) || \
     defined(CONFIG_ARCH_TEGRA_114_SOC) || \
     defined(CONFIG_ARCH_TEGRA_124_SOC) || \
@@ -193,7 +197,8 @@ extern const struct tegra_mc_ops tegra30_mc_ops;
 
 #if defined(CONFIG_ARCH_TEGRA_186_SOC) || \
     defined(CONFIG_ARCH_TEGRA_194_SOC) || \
-    defined(CONFIG_ARCH_TEGRA_234_SOC)
+    defined(CONFIG_ARCH_TEGRA_234_SOC) || \
+    defined(CONFIG_ARCH_TEGRA_264_SOC)
 extern const struct tegra_mc_ops tegra186_mc_ops;
 #endif
 
diff --git a/drivers/memory/tegra/tegra186-emc.c b/drivers/memory/tegra/tegra186-emc.c
index bc807d7fcd4e..d6cd90c7ad53 100644
--- a/drivers/memory/tegra/tegra186-emc.c
+++ b/drivers/memory/tegra/tegra186-emc.c
@@ -1,6 +1,6 @@
 // SPDX-License-Identifier: GPL-2.0-only
 /*
- * Copyright (C) 2019 NVIDIA CORPORATION.  All rights reserved.
+ * Copyright (C) 2019-2025 NVIDIA CORPORATION.  All rights reserved.
  */
 
 #include <linux/clk.h>
@@ -393,6 +393,9 @@ static const struct of_device_id tegra186_emc_of_match[] = {
 #endif
 #if defined(CONFIG_ARCH_TEGRA_234_SOC)
 	{ .compatible = "nvidia,tegra234-emc" },
+#endif
+#if defined(CONFIG_ARCH_TEGRA_264_SOC)
+	{ .compatible = "nvidia,tegra264-emc" },
 #endif
 	{ /* sentinel */ }
 };
diff --git a/drivers/memory/tegra/tegra186.c b/drivers/memory/tegra/tegra186.c
index 1b3183951bfe..aee11457bf8e 100644
--- a/drivers/memory/tegra/tegra186.c
+++ b/drivers/memory/tegra/tegra186.c
@@ -1,6 +1,6 @@
 // SPDX-License-Identifier: GPL-2.0-only
 /*
- * Copyright (C) 2017-2021 NVIDIA CORPORATION.  All rights reserved.
+ * Copyright (C) 2017-2025 NVIDIA CORPORATION.  All rights reserved.
  */
 
 #include <linux/io.h>
@@ -26,11 +26,24 @@
 static int tegra186_mc_probe(struct tegra_mc *mc)
 {
 	struct platform_device *pdev = to_platform_device(mc->dev);
+	struct resource *res;
 	unsigned int i;
 	char name[8];
 	int err;
 
-	mc->bcast_ch_regs = devm_platform_ioremap_resource_byname(pdev, "broadcast");
+	/*
+	 * From Tegra264, the SID region is not present in MC node and BROADCAST is first.
+	 * The common function 'tegra_mc_probe()' already maps first region entry from DT.
+	 * Check if the SID region is present in DT then map BROADCAST. Otherwise, consider
+	 * the first entry mapped in mc probe as the BROADCAST region. This is done to avoid
+	 * mapping the region twice when SID is not present and keep backward compatibility.
+	 */
+	res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "sid");
+	if (res)
+		mc->bcast_ch_regs = devm_platform_ioremap_resource_byname(pdev, "broadcast");
+	else
+		mc->bcast_ch_regs = mc->regs;
+
 	if (IS_ERR(mc->bcast_ch_regs)) {
 		if (PTR_ERR(mc->bcast_ch_regs) == -EINVAL) {
 			dev_warn(&pdev->dev,
diff --git a/drivers/memory/tegra/tegra264-bwmgr.h b/drivers/memory/tegra/tegra264-bwmgr.h
new file mode 100644
index 000000000000..93bfceaac9c8
--- /dev/null
+++ b/drivers/memory/tegra/tegra264-bwmgr.h
@@ -0,0 +1,50 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/* Copyright (C) 2025 NVIDIA CORPORATION.  All rights reserved. */
+
+#ifndef MEMORY_TEGRA_TEGRA264_BWMGR_H
+#define MEMORY_TEGRA_TEGRA264_BWMGR_H
+
+#define TEGRA264_BWMGR_ICC_PRIMARY	1
+#define TEGRA264_BWMGR_DEBUG		2
+#define TEGRA264_BWMGR_CPU_CLUSTER0	3
+#define TEGRA264_BWMGR_CPU_CLUSTER1	4
+#define TEGRA264_BWMGR_CPU_CLUSTER2	5
+#define TEGRA264_BWMGR_CPU_CLUSTER3	6
+#define TEGRA264_BWMGR_CPU_CLUSTER4	7
+#define TEGRA264_BWMGR_CPU_CLUSTER5	8
+#define TEGRA264_BWMGR_CPU_CLUSTER6	9
+#define TEGRA264_BWMGR_CACTMON		10
+#define TEGRA264_BWMGR_DISPLAY		11
+#define TEGRA264_BWMGR_VI		12
+#define TEGRA264_BWMGR_APE		13
+#define TEGRA264_BWMGR_VIFAL		14
+#define TEGRA264_BWMGR_GPU		15
+#define TEGRA264_BWMGR_EQOS		16
+#define TEGRA264_BWMGR_PCIE_0		17
+#define TEGRA264_BWMGR_PCIE_1		18
+#define TEGRA264_BWMGR_PCIE_2		19
+#define TEGRA264_BWMGR_PCIE_3		20
+#define TEGRA264_BWMGR_PCIE_4		21
+#define TEGRA264_BWMGR_PCIE_5		22
+#define TEGRA264_BWMGR_SDMMC_1		23
+#define TEGRA264_BWMGR_SDMMC_2		24
+#define TEGRA264_BWMGR_NVDEC		25
+#define TEGRA264_BWMGR_NVENC		26
+#define TEGRA264_BWMGR_NVJPG_0		27
+#define TEGRA264_BWMGR_NVJPG_1		28
+#define TEGRA264_BWMGR_OFAA		29
+#define TEGRA264_BWMGR_XUSB_HOST	30
+#define TEGRA264_BWMGR_XUSB_DEV		31
+#define TEGRA264_BWMGR_TSEC		32
+#define TEGRA264_BWMGR_VIC		33
+#define TEGRA264_BWMGR_APEDMA		34
+#define TEGRA264_BWMGR_SE		35
+#define TEGRA264_BWMGR_ISP		36
+#define TEGRA264_BWMGR_HDA		37
+#define TEGRA264_BWMGR_VI2FAL		38
+#define TEGRA264_BWMGR_VI2		39
+#define TEGRA264_BWMGR_RCE		40
+#define TEGRA264_BWMGR_PVA		41
+#define TEGRA264_BWMGR_NVPMODEL		42
+
+#endif
diff --git a/drivers/memory/tegra/tegra264.c b/drivers/memory/tegra/tegra264.c
new file mode 100644
index 000000000000..5203e6c11372
--- /dev/null
+++ b/drivers/memory/tegra/tegra264.c
@@ -0,0 +1,313 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2025, NVIDIA CORPORATION.  All rights reserved.
+ */
+
+#include <dt-bindings/memory/nvidia,tegra264.h>
+
+#include <linux/interconnect.h>
+#include <linux/of_device.h>
+#include <linux/tegra-icc.h>
+
+#include <soc/tegra/bpmp.h>
+#include <soc/tegra/mc.h>
+
+#include "mc.h"
+#include "tegra264-bwmgr.h"
+
+/*
+ * MC Client entries are sorted in the increasing order of the
+ * override and security register offsets.
+ */
+static const struct tegra_mc_client tegra264_mc_clients[] = {
+	{
+		.id = TEGRA264_MEMORY_CLIENT_HDAR,
+		.name = "hdar",
+		.bpmp_id = TEGRA264_BWMGR_HDA,
+		.type = TEGRA_ICC_ISO_AUDIO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_HDAW,
+		.name = "hdaw",
+		.bpmp_id = TEGRA264_BWMGR_HDA,
+		.type = TEGRA_ICC_ISO_AUDIO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_MGBE0R,
+		.name = "mgbe0r",
+		.bpmp_id = TEGRA264_BWMGR_EQOS,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_MGBE0W,
+		.name = "mgbe0w",
+		.bpmp_id = TEGRA264_BWMGR_EQOS,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_MGBE1R,
+		.name = "mgbe1r",
+		.bpmp_id = TEGRA264_BWMGR_EQOS,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_MGBE1W,
+		.name = "mgbe1w",
+		.bpmp_id = TEGRA264_BWMGR_EQOS,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_SDMMC0R,
+		.name = "sdmmc0r",
+		.bpmp_id = TEGRA264_BWMGR_SDMMC_1,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_SDMMC0W,
+		.name = "sdmmc0w",
+		.bpmp_id = TEGRA264_BWMGR_SDMMC_1,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_VICR,
+		.name = "vicr",
+		.bpmp_id = TEGRA264_BWMGR_VIC,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_VICW,
+		.name = "vicw",
+		.bpmp_id = TEGRA264_BWMGR_VIC,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_APER,
+		.name = "aper",
+		.bpmp_id = TEGRA264_BWMGR_APE,
+		.type = TEGRA_ICC_ISO_AUDIO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_APEW,
+		.name = "apew",
+		.bpmp_id = TEGRA264_BWMGR_APE,
+		.type = TEGRA_ICC_ISO_AUDIO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_APEDMAR,
+		.name = "apedmar",
+		.bpmp_id = TEGRA264_BWMGR_APEDMA,
+		.type = TEGRA_ICC_ISO_AUDIO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_APEDMAW,
+		.name = "apedmaw",
+		.bpmp_id = TEGRA264_BWMGR_APEDMA,
+		.type = TEGRA_ICC_ISO_AUDIO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_VIFALCONR,
+		.name = "vifalconr",
+		.bpmp_id = TEGRA264_BWMGR_VIFAL,
+		.type = TEGRA_ICC_ISO_VIFAL,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_VIFALCONW,
+		.name = "vifalconw",
+		.bpmp_id = TEGRA264_BWMGR_VIFAL,
+		.type = TEGRA_ICC_ISO_VIFAL,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_RCER,
+		.name = "rcer",
+		.bpmp_id = TEGRA264_BWMGR_RCE,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_RCEW,
+		.name = "rcew",
+		.bpmp_id = TEGRA264_BWMGR_RCE,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_PCIE0W,
+		.name = "pcie0w",
+		.bpmp_id = TEGRA264_BWMGR_PCIE_0,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_PCIE1R,
+		.name = "pcie1r",
+		.bpmp_id = TEGRA264_BWMGR_PCIE_1,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_PCIE1W,
+		.name = "pcie1w",
+		.bpmp_id = TEGRA264_BWMGR_PCIE_1,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_PCIE2AR,
+		.name = "pcie2ar",
+		.bpmp_id = TEGRA264_BWMGR_PCIE_2,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_PCIE2AW,
+		.name = "pcie2aw",
+		.bpmp_id = TEGRA264_BWMGR_PCIE_2,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_PCIE3R,
+		.name = "pcie3r",
+		.bpmp_id = TEGRA264_BWMGR_PCIE_3,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_PCIE3W,
+		.name = "pcie3w",
+		.bpmp_id = TEGRA264_BWMGR_PCIE_3,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_PCIE4R,
+		.name = "pcie4r",
+		.bpmp_id = TEGRA264_BWMGR_PCIE_4,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_PCIE4W,
+		.name = "pcie4w",
+		.bpmp_id = TEGRA264_BWMGR_PCIE_4,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_PCIE5R,
+		.name = "pcie5r",
+		.bpmp_id = TEGRA264_BWMGR_PCIE_5,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_PCIE5W,
+		.name = "pcie5w",
+		.bpmp_id = TEGRA264_BWMGR_PCIE_5,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_GPUR02MC,
+		.name = "gpur02mc",
+		.bpmp_id = TEGRA264_BWMGR_GPU,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_GPUW02MC,
+		.name = "gpuw02mc",
+		.bpmp_id = TEGRA264_BWMGR_GPU,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_NVDECSRD2MC,
+		.name = "nvdecsrd2mc",
+		.bpmp_id = TEGRA264_BWMGR_NVDEC,
+		.type = TEGRA_ICC_NISO,
+	}, {
+		.id = TEGRA264_MEMORY_CLIENT_NVDECSWR2MC,
+		.name = "nvdecswr2mc",
+		.bpmp_id = TEGRA264_BWMGR_NVDEC,
+		.type = TEGRA_ICC_NISO,
+	},
+};
+
+/*
+ * tegra264_mc_icc_set() - Pass MC client info to the BPMP-FW
+ * @src: ICC node for Memory Controller's (MC) Client
+ * @dst: ICC node for Memory Controller (MC)
+ *
+ * Passing the current request info from the MC to the BPMP-FW where
+ * LA and PTSA registers are accessed and the final EMC freq is set
+ * based on client_id, type, latency and bandwidth.
+ * icc_set_bw() makes set_bw calls for both MC and EMC providers in
+ * sequence. Both the calls are protected by 'mutex_lock(&icc_lock)'.
+ * So, the data passed won't be updated by concurrent set calls from
+ * other clients.
+ */
+static int tegra264_mc_icc_set(struct icc_node *src, struct icc_node *dst)
+{
+	struct tegra_mc *mc = icc_provider_to_tegra_mc(dst->provider);
+	struct mrq_bwmgr_int_request bwmgr_req = { 0 };
+	struct mrq_bwmgr_int_response bwmgr_resp = { 0 };
+	const struct tegra_mc_client *pclient = src->data;
+	struct tegra_bpmp_message msg;
+	int ret;
+
+	/*
+	 * Same Src and Dst node will happen during boot from icc_node_add().
+	 * This can be used to pre-initialize and set bandwidth for all clients
+	 * before their drivers are loaded. We are skipping this case as for us,
+	 * the pre-initialization already happened in Bootloader(MB2) and BPMP-FW.
+	 */
+	if (src->id == dst->id)
+		return 0;
+
+	if (!mc->bwmgr_mrq_supported)
+		return 0;
+
+	if (!mc->bpmp) {
+		dev_err(mc->dev, "BPMP reference NULL\n");
+		return -ENOENT;
+	}
+
+	if (pclient->type == TEGRA_ICC_NISO)
+		bwmgr_req.bwmgr_calc_set_req.niso_bw = src->avg_bw;
+	else
+		bwmgr_req.bwmgr_calc_set_req.iso_bw = src->avg_bw;
+
+	bwmgr_req.bwmgr_calc_set_req.client_id = pclient->bpmp_id;
+
+	bwmgr_req.cmd = CMD_BWMGR_INT_CALC_AND_SET;
+	bwmgr_req.bwmgr_calc_set_req.mc_floor = src->peak_bw;
+	bwmgr_req.bwmgr_calc_set_req.floor_unit = BWMGR_INT_UNIT_KBPS;
+
+	memset(&msg, 0, sizeof(msg));
+	msg.mrq = MRQ_BWMGR_INT;
+	msg.tx.data = &bwmgr_req;
+	msg.tx.size = sizeof(bwmgr_req);
+	msg.rx.data = &bwmgr_resp;
+	msg.rx.size = sizeof(bwmgr_resp);
+
+	ret = tegra_bpmp_transfer(mc->bpmp, &msg);
+	if (ret < 0) {
+		dev_err(mc->dev, "BPMP transfer failed: %d\n", ret);
+		goto error;
+	}
+	if (msg.rx.ret < 0) {
+		pr_err("failed to set bandwidth for %u: %d\n",
+		       bwmgr_req.bwmgr_calc_set_req.client_id, msg.rx.ret);
+		ret = -EINVAL;
+	}
+
+error:
+	return ret;
+}
+
+static int tegra264_mc_icc_aggregate(struct icc_node *node, u32 tag, u32 avg_bw,
+				     u32 peak_bw, u32 *agg_avg, u32 *agg_peak)
+{
+	struct icc_provider *p = node->provider;
+	struct tegra_mc *mc = icc_provider_to_tegra_mc(p);
+
+	if (!mc->bwmgr_mrq_supported)
+		return 0;
+
+	*agg_avg += avg_bw;
+	*agg_peak = max(*agg_peak, peak_bw);
+
+	return 0;
+}
+
+static int tegra264_mc_icc_get_init_bw(struct icc_node *node, u32 *avg, u32 *peak)
+{
+	*avg = 0;
+	*peak = 0;
+
+	return 0;
+}
+
+static const struct tegra_mc_icc_ops tegra264_mc_icc_ops = {
+	.xlate = tegra_mc_icc_xlate,
+	.aggregate = tegra264_mc_icc_aggregate,
+	.get_bw = tegra264_mc_icc_get_init_bw,
+	.set = tegra264_mc_icc_set,
+};
+
+const struct tegra_mc_soc tegra264_mc_soc = {
+	.num_clients = ARRAY_SIZE(tegra264_mc_clients),
+	.clients = tegra264_mc_clients,
+	.num_address_bits = 40,
+	.num_channels = 16,
+	.client_id_mask = 0x1ff,
+	.intmask = MC_INT_DECERR_ROUTE_SANITY |
+		   MC_INT_DECERR_GENERALIZED_CARVEOUT | MC_INT_DECERR_MTS |
+		   MC_INT_SECERR_SEC | MC_INT_DECERR_VPR |
+		   MC_INT_SECURITY_VIOLATION | MC_INT_DECERR_EMEM,
+	.has_addr_hi_reg = true,
+	.ops = &tegra186_mc_ops,
+	.icc_ops = &tegra264_mc_icc_ops,
+	.ch_intmask = 0x0000ff00,
+	.global_intstatus_channel_shift = 8,
+	/*
+	 * Additionally, there are lite carveouts but those are not currently
+	 * supported.
+	 */
+	.num_carveouts = 32,
+};
-- 
2.50.0

Re: [PATCH 4/4] memory: tegra: Add Tegra264 MC and EMC support

[Krzysztof Kozlowski]

On 08/07/2025 12:52, Thierry Reding wrote:
> From: Sumit Gupta <sumitg@nvidia.com>
> 
> Add support to enable Memory Controller (MC) and External Memory
> Controller (EMC) drivers for Tegra264. The nodes for MC and EMC are
> mostly the same as Tegra234 but differ in number of channels and
> interrupt numbers.
> 
> The patch also adds the bandwidth manager definitions required for
> Tegra264 and uses them to populate the memory client table. All of
> these are needed to properly enable memory interconnect (ICC) support.
> 
> Signed-off-by: Sumit Gupta <sumitg@nvidia.com>
> Signed-off-by: Thierry Reding <treding@nvidia.com>


Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>

Best regards,
Krzysztof

Re: [PATCH 0/4] memory: tegra: Add Tegra264 support

[Thierry Reding]

[-- Attachment #1: Type: text/plain, Size: 1977 bytes --]

On Tue, Jul 08, 2025 at 12:52:41PM +0200, Thierry Reding wrote:
> From: Thierry Reding <treding@nvidia.com>
> 
> This set of patches extends the DT bindings for the memory controller
> and external memory controller for Tegra264 and add the necessary DT
> headers with memory client and stream ID definitions.
> 
> The driver changes in patch 4 are mostly an extension of existing code
> and the bulk consists of the memory client table for the new chip as
> well as the bandwidth manager calculations.
> 
> Thierry
> 
> Sumit Gupta (3):
>   dt-bindings: memory: tegra: Add Tegra264 support
>   dt-bindings: memory: tegra: Add Tegra264 definitions
>   memory: tegra: Add Tegra264 MC and EMC support
> 
> Thierry Reding (1):
>   dt-bindings: memory: tegra: Add Tegra264 stream IDs
> 
>  .../nvidia,tegra186-mc.yaml                   |  65 +++-
>  drivers/memory/tegra/Makefile                 |   2 +
>  drivers/memory/tegra/mc.c                     |   5 +-
>  drivers/memory/tegra/mc.h                     |   9 +-
>  drivers/memory/tegra/tegra186-emc.c           |   5 +-
>  drivers/memory/tegra/tegra186.c               |  17 +-
>  drivers/memory/tegra/tegra264-bwmgr.h         |  50 +++
>  drivers/memory/tegra/tegra264.c               | 313 ++++++++++++++++++
>  include/dt-bindings/memory/nvidia,tegra264.h  | 136 ++++++++
>  9 files changed, 594 insertions(+), 8 deletions(-)
>  create mode 100644 drivers/memory/tegra/tegra264-bwmgr.h
>  create mode 100644 drivers/memory/tegra/tegra264.c
>  create mode 100644 include/dt-bindings/memory/nvidia,tegra264.h

Krzysztof,

There's a dependency between this series and a series adding device tree
files for Tegra264 (both the driver and DTS files include the memory
client IDs in dt-bindings/memory/nvidia,tegra264.h), so I think it'd be
easiest for me to take the driver patches here through the Tegra tree as
well.

Let me know if that works for you or not.

Thanks,
Thierry

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [PATCH 0/4] memory: tegra: Add Tegra264 support

[Krzysztof Kozlowski]

On 09/07/2025 14:18, Thierry Reding wrote:
> On Tue, Jul 08, 2025 at 12:52:41PM +0200, Thierry Reding wrote:
>> From: Thierry Reding <treding@nvidia.com>
>>
>> This set of patches extends the DT bindings for the memory controller
>> and external memory controller for Tegra264 and add the necessary DT
>> headers with memory client and stream ID definitions.
>>
>> The driver changes in patch 4 are mostly an extension of existing code
>> and the bulk consists of the memory client table for the new chip as
>> well as the bandwidth manager calculations.
>>
>> Thierry
>>
>> Sumit Gupta (3):
>>   dt-bindings: memory: tegra: Add Tegra264 support
>>   dt-bindings: memory: tegra: Add Tegra264 definitions
>>   memory: tegra: Add Tegra264 MC and EMC support
>>
>> Thierry Reding (1):
>>   dt-bindings: memory: tegra: Add Tegra264 stream IDs
>>
>>  .../nvidia,tegra186-mc.yaml                   |  65 +++-
>>  drivers/memory/tegra/Makefile                 |   2 +
>>  drivers/memory/tegra/mc.c                     |   5 +-
>>  drivers/memory/tegra/mc.h                     |   9 +-
>>  drivers/memory/tegra/tegra186-emc.c           |   5 +-
>>  drivers/memory/tegra/tegra186.c               |  17 +-
>>  drivers/memory/tegra/tegra264-bwmgr.h         |  50 +++
>>  drivers/memory/tegra/tegra264.c               | 313 ++++++++++++++++++
>>  include/dt-bindings/memory/nvidia,tegra264.h  | 136 ++++++++
>>  9 files changed, 594 insertions(+), 8 deletions(-)
>>  create mode 100644 drivers/memory/tegra/tegra264-bwmgr.h
>>  create mode 100644 drivers/memory/tegra/tegra264.c
>>  create mode 100644 include/dt-bindings/memory/nvidia,tegra264.h
> 
> Krzysztof,
> 
> There's a dependency between this series and a series adding device tree
> files for Tegra264 (both the driver and DTS files include the memory
> client IDs in dt-bindings/memory/nvidia,tegra264.h), so I think it'd be
> easiest for me to take the driver patches here through the Tegra tree as
> well.
> 
> Let me know if that works for you or not.

Works for me. I don't have anything for Tegra memory controllers and
it's close to end of cycle. Let me look at the patches and ack/comment.

Best regards,
Krzysztof