MLECO-2344: Documentation improvement

Documenting how the target platform's SRAM size impacts configuration files,
sources and linker scripts.

Change-Id: I8647ab67b73bafd0c44e6c586a1b5f2602bf03f5

3 files changed, 114 insertions, 11 deletions

diff --git a/docs/sections/memory_considerations.md b/docs/sections/memory_considerations.md
index 599c902..fc81f8f 100644
--- a/docs/sections/memory_considerations.md
+++ b/docs/sections/memory_considerations.md
@@ -2,6 +2,8 @@
 
 - [Memory considerations](#memory-considerations)
   - [Introduction](#introduction)
+  - [Memory available on the target platform](#memory-available-on-the-target-platform)
+    - [Parameters linked to SRAM size definitions](#parameters-linked-to-sram-size-definitions)
   - [Understanding memory usage from Vela output](#understanding-memory-usage-from-vela-output)
     - [Total SRAM used](#total-sram-used)
     - [Total Off-chip Flash used](#total-off_chip-flash-used)
@@ -20,7 +22,7 @@ applicable for other platforms too. The Arm® *Corstone™-300* is composed of b
 
 The Arm® *Ethos™-U* NPU interacts with the system through two AXI interfaces. The first one, is envisaged to be the
 higher-bandwidth, lower-latency, interface. In a typical system, this is wired to an SRAM as it is required to service
-frequent Read and Write traffic.
+frequent read and write traffic.
 
 The second interface is expected to have a higher-latency, lower-bandwidth characteristic, and is typically wired to a
 flash device servicing read-only traffic. In this configuration, the Arm® *Cortex™-M55* CPU and Arm® *Ethos™-U* NPU
@@ -31,6 +33,63 @@ The input and output tensors, along with any intermediate computation buffers, a
 Arm® *Cortex™-M55* CPU and Arm® *Ethos™-U* NPU would be reading, or writing, to this region when running an inference.
 The Arm® *Ethos™-U* NPU requests these Read and Write transactions over the first AXI bus.
 
+## Memory available on the target platform
+
+Embedded target platforms supported have a description in the form of CMake files. These files
+have definitions that describe the memory regions and the peripheral base addresses.
+
+See the example for Arm® *Corstone™-300* description file [corstone-sse-300.cmake](../../scripts/cmake/subsystem-profiles/corstone-sse-300.cmake). For the discussion on this page, it is useful to note the following definitions:
+
+```
+set(ISRAM0_SIZE           "0x00200000" CACHE STRING "ISRAM0 size:       2 MiB")
+set(ISRAM1_SIZE           "0x00200000" CACHE STRING "ISRAM1 size:       2 MiB")
+...
+# SRAM size reserved for activation buffers
+math(EXPR ACTIVATION_BUF_SRAM_SZ "${ISRAM0_SIZE} + ${ISRAM1_SIZE}" OUTPUT_FORMAT HEXADECIMAL)
+```
+This will set `ACTIVATION_BUF_SRAM_SZ` to be **4 MiB** for Arm® *Corstone™-300* target platform.
+As mentioned in the comments within the file, this size is directly linked to the size mentioned
+in the linker scripts, and therefore, it should not be changed without corresponding changes
+in the linker script too. For example, a snippet from the scatter file for Corstone™-300 shows:
+
+```
+;-----------------------------------------------------
+; SSE-300's internal SRAM of 4MiB - reserved for
+; activation buffers.
+; This region should have 3 cycle read latency from
+; both Cortex-M55 and Ethos-U NPU
+;-----------------------------------------------------
+isram.bin       0x31000000  UNINIT ALIGN 16 0x00400000
+{
+  ...
+}
+```
+If the usable size of the internal SRAM was to be increased/decreased, the change should be
+made in both the linker script as well as the `corstone-300.cmake` definition.
+
+### Parameters linked to SRAM size definitions
+
+Other than the obvious link between the linker script and the target profile description in
+CMake files, there are other parameters linked to what the reserved space for activation
+buffers is. These are:
+
+- The file [default_vela.ini](../../scripts/vela/default_vela.ini) contains a parameter called
+  `arena_cache_size` under `Shared_Sram` memory mode. For example:
+  ```
+  [Memory_Mode.Shared_Sram]
+  const_mem_area=Axi1
+  arena_mem_area=Axi0
+  cache_mem_area=Axi0
+  arena_cache_size=4194304
+  ```
+  This size of **4 MiB** here is provided here to allow the default vela optimisation process to
+  use this size as a hint for the available SRAM size for use by the CPU and the NPU.
+
+- In every `usecase.cmake` file (present within each use case's source directory), there is
+  a parameter called `${use_case}_ACTIVATION_BUF_SZ` set to a fixed value by default. This
+  default value should be less than the `ACTIVATION_BUF_SRAM_SZ` if the activation buffer needs
+  to be reserved in the target platform's SRAM region.
+
 ## Understanding memory usage from Vela output
 
 ### Total SRAM used
diff --git a/scripts/cmake/subsystem-profiles/corstone-sse-300.cmake b/scripts/cmake/subsystem-profiles/corstone-sse-300.cmake
index 38930af..7e27f3c 100644
--- a/scripts/cmake/subsystem-profiles/corstone-sse-300.cmake
+++ b/scripts/cmake/subsystem-profiles/corstone-sse-300.cmake
@@ -17,15 +17,6 @@
 
 # CMake configuration file for peripheral memory map for MPS3 as per SSE-300 design
 ###################################################################################################
-#                              Application specific config                                        #
-###################################################################################################
-
-# This parameter is based on the linker/scatter script for SSE-300. Do not change this parameter
-# in isolation.
-set(ACTIVATION_BUF_SRAM_SZ "0x00400000" CACHE STRING "Maximum SRAM size for activation buffers")
-set(DESIGN_NAME            "Arm Corstone-300 (SSE-300)" CACHE STRING "Design name")
-
-###################################################################################################
 #                                         Mem sizes                                               #
 ###################################################################################################
 set(ITCM_SIZE             "0x00080000" CACHE STRING "ITCM size:       512 kiB")
@@ -67,6 +58,17 @@ set(DDR4_BLK2_BASE_S      "0xB0000000" CACHE STRING "DDR4 block 2 Secure base ad
 set(DDR4_BLK3_BASE_S      "0xD0000000" CACHE STRING "DDR4 block 3 Secure base address")
 
 ###################################################################################################
+#                              Application specific config                                        #
+###################################################################################################
+
+# This parameter is based on the linker/scatter script for SSE-300. Do not change this parameter
+# in isolation.
+set(DESIGN_NAME            "Arm Corstone-300 (SSE-300)" CACHE STRING "Design name")
+
+# SRAM size reserved for activation buffers
+math(EXPR ACTIVATION_BUF_SRAM_SZ "${ISRAM0_SIZE} + ${ISRAM1_SIZE}" OUTPUT_FORMAT HEXADECIMAL)
+
+###################################################################################################
 #           Base addresses for dynamic loads (to be used for FVP form only)                       #
 ###################################################################################################
 # This parameter is also mentioned in the linker/scatter script for SSE-300. Do not change these
diff --git a/scripts/cmake/subsystem-profiles/simple_platform.cmake b/scripts/cmake/subsystem-profiles/simple_platform.cmake
index 69a69b5..664697b 100644
--- a/scripts/cmake/subsystem-profiles/simple_platform.cmake
+++ b/scripts/cmake/subsystem-profiles/simple_platform.cmake
@@ -19,6 +19,45 @@
 # version of Arm Corstone-300 platform with minimal peripherals to be able to use Ethos-U55. However,
 # for ease of integration with Arm FastModel Tools, it uses PL011 as the UART component instead of
 # the CMSDK UART block used by the MPS3 FPGA and FVP implementations.
+###################################################################################################
+#                                         Mem sizes                                               #
+###################################################################################################
+set(ITCM_SIZE             "0x00080000" CACHE STRING "ITCM size:       512 kiB")
+set(DTCM_BLK_SIZE         "0x00020000" CACHE STRING "DTCM size:       128 kiB, 4 banks")
+set(BRAM_SIZE             "0x00200000" CACHE STRING "BRAM size:         2 MiB")
+set(ISRAM0_SIZE           "0x00200000" CACHE STRING "ISRAM0 size:       2 MiB")
+set(ISRAM1_SIZE           "0x00200000" CACHE STRING "ISRAM1 size:       2 MiB")
+set(DDR4_BLK_SIZE         "0x10000000" CACHE STRING "DDR4 block size: 256 MiB")
+
+###################################################################################################
+#                                Base addresses for memory regions                                #
+###################################################################################################
+set(ITCM_BASE_NS          "0x00000000" CACHE STRING "Instruction TCM Non-Secure base address")
+set(BRAM_BASE_NS          "0x01000000" CACHE STRING "CODE SRAM Non-Secure base address")
+set(DTCM0_BASE_NS         "0x20000000" CACHE STRING "Data TCM block 0 Non-Secure base address")
+set(DTCM1_BASE_NS         "0x20020000" CACHE STRING "Data TCM block 1 Non-Secure base address")
+set(DTCM2_BASE_NS         "0x20040000" CACHE STRING "Data TCM block 2 Non-Secure base address")
+set(DTCM3_BASE_NS         "0x20060000" CACHE STRING "Data TCM block 3 Non-Secure base address")
+set(ISRAM0_BASE_NS        "0x21000000" CACHE STRING "Internal SRAM Area Non-Secure base address")
+set(ISRAM1_BASE_NS        "0x21200000" CACHE STRING "Internal SRAM Area Non-Secure base address")
+set(QSPI_SRAM_BASE_NS     "0x28000000" CACHE STRING "QSPI SRAM Non-Secure base address")
+set(DDR4_BLK0_BASE_NS     "0x60000000" CACHE STRING "DDR4 block 0 Non-Secure base address")
+set(DDR4_BLK1_BASE_NS     "0x80000000" CACHE STRING "DDR4 block 1 Non-Secure base address")
+set(DDR4_BLK2_BASE_NS     "0xA0000000" CACHE STRING "DDR4 block 2 Non-Secure base address")
+set(DDR4_BLK3_BASE_NS     "0xC0000000" CACHE STRING "DDR4 block 3 Non-Secure base address")
+
+set(ITCM_BASE_S           "0x10000000" CACHE STRING "Instruction TCM Secure base address")
+set(BRAM_BASE_S           "0x11000000" CACHE STRING "CODE SRAM Secure base address")
+set(DTCM0_BASE_S          "0x30000000" CACHE STRING "Data TCM block 0 Secure base address")
+set(DTCM1_BASE_S          "0x30020000" CACHE STRING "Data TCM block 1 Secure base address")
+set(DTCM2_BASE_S          "0x30040000" CACHE STRING "Data TCM block 2 Secure base address")
+set(DTCM3_BASE_S          "0x30060000" CACHE STRING "Data TCM block 3 Secure base address")
+set(ISRAM0_BASE_S         "0x31000000" CACHE STRING "Internal SRAM Area Secure base address")
+set(ISRAM1_BASE_S         "0x31200000" CACHE STRING "Internal SRAM Area Secure base address")
+set(DDR4_BLK0_BASE_S      "0x70000000" CACHE STRING "DDR4 block 0 Secure base address")
+set(DDR4_BLK1_BASE_S      "0x90000000" CACHE STRING "DDR4 block 1 Secure base address")
+set(DDR4_BLK2_BASE_S      "0xB0000000" CACHE STRING "DDR4 block 2 Secure base address")
+set(DDR4_BLK3_BASE_S      "0xD0000000" CACHE STRING "DDR4 block 3 Secure base address")
 
 ###################################################################################################
 #                              Application specific config                                        #
@@ -26,9 +65,12 @@
 
 # This parameter is based on the linker/scatter script for simple platform. Do not change this
 # parameter in isolation.
-set(ACTIVATION_BUF_SRAM_SZ "0x00200000"      CACHE STRING "Maximum SRAM size for activation buffers")
 set(DESIGN_NAME            "Simple platform" CACHE STRING "Design name")
 
+# SRAM size reserved for activation buffers
+math(EXPR ACTIVATION_BUF_SRAM_SZ "${ISRAM0_SIZE} + ${ISRAM1_SIZE}" OUTPUT_FORMAT HEXADECIMAL)
+
+
 ###################################################################################################
 #                                         Base addresses                                          #
 ###################################################################################################