Improvements to heap-memory and PSRAM handling (#4791)

* Improved heap and PSRAM handling

- Segment `allocateData()` uses more elaborate DRAM checking to reduce fragmentation and allow for larger setups to run on low heap
- Segment data allocation fails if minimum contiguous block size runs low to keep the UI working
- Increased `MAX_SEGMENT_DATA` to account for better segment data handling
- Memory allocation functions try to keep enough DRAM for segment data
- Added constant `PSRAM_THRESHOLD` to improve PSARM usage
- Increase MIN_HEAP_SIZE to reduce risk of breaking UI due to low memory for JSON response
- ESP32 makes use of IRAM (no 8bit access) for pixeluffers, freeing up to 50kB of RAM
- Fix to properly get available heap on all platforms: added function `getFreeHeapSize()`
- Bugfix for effects that divide by SEGLEN: don't run FX in service() if segment is not active
-Syntax fix in AR: calloc() uses (numelements, size) as arguments

* Added new functions for allocation and heap checking

- added `allocate_buffer()` function that can be used to allocate large buffers: takes parameters to set preferred ram location, including 32bit accessible RAM on ESP32. Returns null if heap runs low or switches to PSRAM
- getFreeHeapSize() and getContiguousFreeHeap() helper functions for all platforms to correctly report free useable heap
- updated some constants
- updated segment data allocation to free the data if it is large

- replaced "psramsafe" variable with it's #ifdef: BOARD_HAS_PSRAM and made accomodating changes
- added some compile-time checks to handle invalid env. definitions
- updated all allocation functions and some of the logic behind them
- added use of fast RTC-Memory where available
- increased MIN_HEAP_SIZE for all systems (improved stability in tests)
- updated memory calculation in web-UI to account for required segment buffer
- added UI alerts if buffer allocation fails
- made getUsedSegmentData() non-private (used in buffer alloc function)
- changed MAX_SEGMENT_DATA
- added more detailed memory log to DEBUG output
- added debug output to buffer alloc function

wled00/FX.h

@@ -88,23 +88,26 @@ extern byte realtimeMode;           // used in getMappedPixelIndex()
 #endif
 #define FPS_CALC_SHIFT 7 // bit shift for fixed point math
 
-/* each segment uses 82 bytes of SRAM memory, so if you're application fails because of
-  insufficient memory, decreasing MAX_NUM_SEGMENTS may help */
+// heap memory limit for effects data, pixel buffers try to reserve it if PSRAM is available
 #ifdef ESP8266
   #define MAX_NUM_SEGMENTS  16
   /* How much data bytes all segments combined may allocate */
-  #define MAX_SEGMENT_DATA  5120
+  #define MAX_SEGMENT_DATA  (6*1024) // 6k by default
 #elif defined(CONFIG_IDF_TARGET_ESP32S2)
-  #define MAX_NUM_SEGMENTS  20
-  #define MAX_SEGMENT_DATA  (MAX_NUM_SEGMENTS*512)  // 10k by default (S2 is short on free RAM)
+  #define MAX_NUM_SEGMENTS  32
+  #define MAX_SEGMENT_DATA  (20*1024) // 20k by default (S2 is short on free RAM), limit does not apply if PSRAM is available
 #else
-  #define MAX_NUM_SEGMENTS  32  // warning: going beyond 32 may consume too much RAM for stable operation
-  #define MAX_SEGMENT_DATA  (MAX_NUM_SEGMENTS*1280) // 40k by default
+  #ifdef BOARD_HAS_PSRAM
+    #define MAX_NUM_SEGMENTS  64
+  #else
+    #define MAX_NUM_SEGMENTS  32
+  #endif
+  #define MAX_SEGMENT_DATA  (64*1024) // 64k by default, limit does not apply if PSRAM is available
 #endif
 
 /* How much data bytes each segment should max allocate to leave enough space for other segments,
   assuming each segment uses the same amount of data. 256 for ESP8266, 640 for ESP32. */
-#define FAIR_DATA_PER_SEG (MAX_SEGMENT_DATA / WS2812FX::getMaxSegments())
+#define FAIR_DATA_PER_SEG (MAX_SEGMENT_DATA / MAX_NUM_SEGMENTS)
 
 #define MIN_SHOW_DELAY   (_frametime < 16 ? 8 : 15)
 
@@ -533,7 +536,6 @@ class Segment {
 
   protected:
 
-    inline static unsigned getUsedSegmentData()            { return Segment::_usedSegmentData; }
     inline static void     addUsedSegmentData(int len)     { Segment::_usedSegmentData += len; }
 
     inline uint32_t *getPixels() const                              { return pixels; }
@@ -600,8 +602,8 @@ class Segment {
     , _t(nullptr)
     {
       DEBUGFX_PRINTF_P(PSTR("-- Creating segment: %p [%d,%d:%d,%d]\n"), this, (int)start, (int)stop, (int)startY, (int)stopY);
-      // allocate render buffer (always entire segment)
-      pixels = static_cast<uint32_t*>(d_calloc(sizeof(uint32_t), length())); // error handling is also done in isActive()
+      // allocate render buffer (always entire segment), prefer PSRAM if DRAM is running low. Note: impact on FPS with PSRAM buffer is low (<2% with QSPI PSRAM)
+      pixels = static_cast<uint32_t*>(allocate_buffer(length() * sizeof(uint32_t), BFRALLOC_PREFER_PSRAM | BFRALLOC_NOBYTEACCESS | BFRALLOC_CLEAR));
       if (!pixels) {
         DEBUGFX_PRINTLN(F("!!! Not enough RAM for pixel buffer !!!"));
         extern byte errorFlag;
@@ -623,7 +625,7 @@ class Segment {
       #endif
       clearName();
       deallocateData();
-      d_free(pixels);
+      p_free(pixels);
     }
 
     Segment& operator= (const Segment &orig); // copy assignment
@@ -646,7 +648,7 @@ class Segment {
     inline uint16_t groupLength()          const { return grouping + spacing; }
     inline uint8_t  getLightCapabilities() const { return _capabilities; }
     inline void     deactivate()                 { setGeometry(0,0); }
-    inline Segment &clearName()                  { d_free(name); name = nullptr; return *this; }
+    inline Segment &clearName()                  { p_free(name); name = nullptr; return *this; }
     inline Segment &setName(const String &name)  { return setName(name.c_str()); }
 
     inline static unsigned vLength()                       { return Segment::_vLength; }
@@ -672,6 +674,7 @@ class Segment {
     inline uint16_t dataSize() const { return _dataLen; }
     bool allocateData(size_t len);  // allocates effect data buffer in heap and clears it
     void deallocateData();          // deallocates (frees) effect data buffer from heap
+    inline static unsigned getUsedSegmentData()            { return Segment::_usedSegmentData; }
     /**
       * Flags that before the next effect is calculated,
       * the internal segment state should be reset.
@@ -868,8 +871,8 @@ class WS2812FX {
     }
 
     ~WS2812FX() {
-      d_free(_pixels);
-      d_free(_pixelCCT); // just in case
+      p_free(_pixels);
+      p_free(_pixelCCT); // just in case
       d_free(customMappingTable);
       _mode.clear();
       _modeData.clear();