Allow for HALF/NOMINAL/DOUBLE frequency of MAX7456 SPI clock

src/main/cli/settings.c

@@ -331,7 +331,7 @@ static const char * const lookupTableBusType[] = {
 
 #ifdef USE_MAX7456
 static const char * const lookupTableMax7456Clock[] = {
-    "HALF", "DEFAULT", "FULL"
+    "HALF", "NOMINAL", "DOUBLE"
 };
 #endif
 

src/main/drivers/accgyro/accgyro_mpu.c

@@ -217,10 +217,7 @@ bool mpuAccReadSPI(accDev_t *acc)
     case GYRO_EXTI_INT:
     case GYRO_EXTI_NO_INT:
     {
-        // Ensure any prior DMA has completed before continuing
+        acc->gyro->dev.txBuf[0] = MPU_RA_ACCEL_XOUT_H | 0x80;
-        spiWaitClaim(&acc->gyro->dev);
-
-        acc->gyro->dev.txBuf[0] = acc->gyro->accDataReg | 0x80;
 
         busSegment_t segments[] = {
                 {NULL, NULL, 7, true, NULL},
@@ -274,8 +271,6 @@ bool mpuGyroReadSPI(gyroDev_t *gyro)
         gyro->gyroDmaMaxDuration = 5;
         if (gyro->detectedEXTI > GYRO_EXTI_DETECT_THRESHOLD) {
             if (spiUseDMA(&gyro->dev)) {
-                // Indicate that the bus on which this device resides may initiate DMA transfers from interrupt context
-                spiSetAtomicWait(&gyro->dev);
                 gyro->dev.callbackArg = (uint32_t)gyro;
                 gyro->dev.txBuf[0] = gyro->accDataReg | 0x80;
                 gyro->segments[0].len = gyro->gyroDataReg - gyro->accDataReg + 7;
@@ -299,10 +294,7 @@ bool mpuGyroReadSPI(gyroDev_t *gyro)
     case GYRO_EXTI_INT:
     case GYRO_EXTI_NO_INT:
     {
-        // Ensure any prior DMA has completed before continuing
+        gyro->dev.txBuf[0] = MPU_RA_GYRO_XOUT_H | 0x80;
-        spiWaitClaim(&gyro->dev);
-
-        gyro->dev.txBuf[0] = gyro->gyroDataReg | 0x80;
 
         busSegment_t segments[] = {
                 {NULL, NULL, 7, true, NULL},

src/main/drivers/accgyro/accgyro_spi_bmi270.c

@@ -316,9 +316,6 @@ static bool bmi270AccRead(accDev_t *acc)
     case GYRO_EXTI_INT:
     case GYRO_EXTI_NO_INT:
     {
-        // Ensure any prior DMA has completed before continuing
-        spiWaitClaim(&acc->gyro->dev);
-
         acc->gyro->dev.txBuf[0] = BMI270_REG_ACC_DATA_X_LSB | 0x80;
 
         busSegment_t segments[] = {
@@ -374,8 +371,6 @@ static bool bmi270GyroReadRegister(gyroDev_t *gyro)
         // Using DMA for gyro access upsets the scheduler on the F4
         if (gyro->detectedEXTI > GYRO_EXTI_DETECT_THRESHOLD) {
             if (spiUseDMA(&gyro->dev)) {
-                // Indicate that the bus on which this device resides may initiate DMA transfers from interrupt context
-                spiSetAtomicWait(&gyro->dev);
                 gyro->dev.callbackArg = (uint32_t)gyro;
                 gyro->dev.txBuf[0] = BMI270_REG_ACC_DATA_X_LSB | 0x80;
                 gyro->segments[0].len = 14;
@@ -399,9 +394,6 @@ static bool bmi270GyroReadRegister(gyroDev_t *gyro)
     case GYRO_EXTI_INT:
     case GYRO_EXTI_NO_INT:
     {
-        // Ensure any prior DMA has completed before continuing
-        spiWaitClaim(&gyro->dev);
-
         gyro->dev.txBuf[0] = BMI270_REG_GYR_DATA_X_LSB | 0x80;
 
         busSegment_t segments[] = {

src/main/drivers/bus.h

@@ -60,7 +60,6 @@ typedef struct busDevice_s {
         } mpuSlave;
     } busType_u;
     bool useDMA;
-    bool useAtomicWait;
     uint8_t deviceCount;
     dmaChannelDescriptor_t *dmaTx;
     dmaChannelDescriptor_t *dmaRx;

src/main/drivers/bus_spi.c

@@ -39,6 +39,8 @@
 #include "nvic.h"
 #include "pg/bus_spi.h"
 
+#define NUM_QUEUE_SEGS 5
+
 static uint8_t spiRegisteredDeviceCount = 0;
 
 spiDevice_t spiDevice[SPIDEV_COUNT];
@@ -157,34 +159,6 @@ bool spiIsBusy(const extDevice_t *dev)
     return (dev->bus->curSegment != (busSegment_t *)BUS_SPI_FREE);
 }
 
-// Indicate that the bus on which this device resides may initiate DMA transfers from interrupt context
-void spiSetAtomicWait(const extDevice_t *dev)
-{
-    dev->bus->useAtomicWait = true;
-}
-
-// Wait for DMA completion and claim the bus driver
-void spiWaitClaim(const extDevice_t *dev)
-{
-    // If there is a device on the bus whose driver might call spiSequence from an ISR then an
-    // atomic access is required to claim the bus, however if not, then interrupts need not be
-    // disabled as this can result in edge triggered interrupts being missed
-
-    if (dev->bus->useAtomicWait) {
-        // Prevent race condition where the bus appears free, but a gyro interrupt starts a transfer
-        do {
-            ATOMIC_BLOCK(NVIC_PRIO_MAX) {
-                if (dev->bus->curSegment == (busSegment_t *)BUS_SPI_FREE) {
-                    dev->bus->curSegment = (busSegment_t *)BUS_SPI_LOCKED;
-                }
-            }
-        } while (dev->bus->curSegment != (busSegment_t *)BUS_SPI_LOCKED);
-    } else {
-        // Wait for completion
-        while (dev->bus->curSegment != (busSegment_t *)BUS_SPI_FREE);
-    }
-}
-
 // Wait for DMA completion
 void spiWait(const extDevice_t *dev)
 {
@@ -201,9 +175,6 @@ void spiReadWriteBuf(const extDevice_t *dev, uint8_t *txData, uint8_t *rxData, i
             {NULL, NULL, 0, true, NULL},
     };
 
-    // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(dev);
-
     spiSequence(dev, &segments[0]);
 
     spiWait(dev);
@@ -233,9 +204,6 @@ uint8_t spiReadWrite(const extDevice_t *dev, uint8_t data)
             {NULL, NULL, 0, true, NULL},
     };
 
-    // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(dev);
-
     spiSequence(dev, &segments[0]);
 
     spiWait(dev);
@@ -255,9 +223,6 @@ uint8_t spiReadWriteReg(const extDevice_t *dev, uint8_t reg, uint8_t data)
             {NULL, NULL, 0, true, NULL},
     };
 
-    // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(dev);
-
     spiSequence(dev, &segments[0]);
 
     spiWait(dev);
@@ -274,9 +239,6 @@ void spiWrite(const extDevice_t *dev, uint8_t data)
             {NULL, NULL, 0, true, NULL},
     };
 
-    // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(dev);
-
     spiSequence(dev, &segments[0]);
 
     spiWait(dev);
@@ -292,9 +254,6 @@ void spiWriteReg(const extDevice_t *dev, uint8_t reg, uint8_t data)
             {NULL, NULL, 0, true, NULL},
     };
 
-    // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(dev);
-
     spiSequence(dev, &segments[0]);
 
     spiWait(dev);
@@ -323,9 +282,6 @@ void spiReadRegBuf(const extDevice_t *dev, uint8_t reg, uint8_t *data, uint8_t l
             {NULL, NULL, 0, true, NULL},
     };
 
-    // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(dev);
-
     spiSequence(dev, &segments[0]);
 
     spiWait(dev);
@@ -360,9 +316,6 @@ void spiWriteRegBuf(const extDevice_t *dev, uint8_t reg, uint8_t *data, uint32_t
             {NULL, NULL, 0, true, NULL},
     };
 
-    // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(dev);
-
     spiSequence(dev, &segments[0]);
 
     spiWait(dev);
@@ -379,9 +332,6 @@ uint8_t spiReadReg(const extDevice_t *dev, uint8_t reg)
             {NULL, NULL, 0, true, NULL},
     };
 
-    // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(dev);
-
     spiSequence(dev, &segments[0]);
 
     spiWait(dev);
@@ -414,6 +364,19 @@ uint16_t spiCalculateDivider(uint32_t freq)
     return divisor;
 }
 
+uint32_t spiCalculateClock(uint16_t spiClkDivisor)
+{
+#if defined(STM32F4) || defined(STM32G4) || defined(STM32F7)
+    uint32_t spiClk = SystemCoreClock / 2;
+#elif defined(STM32H7)
+    uint32_t spiClk = 100000000;
+#else
+#error "Base SPI clock not defined for this architecture"
+#endif
+
+    return spiClk / spiClkDivisor;
+}
+
 // Interrupt handler for SPI receive DMA completion
 static void spiIrqHandler(const extDevice_t *dev)
 {
@@ -448,9 +411,10 @@ static void spiIrqHandler(const extDevice_t *dev)
         if (nextSegment->txData) {
             const extDevice_t *nextDev = (const extDevice_t *)nextSegment->txData;
             busSegment_t *nextSegments = (busSegment_t *)nextSegment->rxData;
-            nextSegment->txData = NULL;
             // The end of the segment list has been reached
-            spiSequenceStart(nextDev, nextSegments);
+            bus->curSegment = nextSegments;
+            nextSegment->txData = NULL;
+            spiSequenceStart(nextDev);
         } else {
             // The end of the segment list has been reached, so mark transactions as complete
             bus->curSegment = (busSegment_t *)BUS_SPI_FREE;
@@ -559,7 +523,6 @@ bool spiSetBusInstance(extDevice_t *dev, uint32_t device)
 
     bus->busType = BUS_TYPE_SPI;
     bus->useDMA = false;
-    bus->useAtomicWait = false;
     bus->deviceCount = 1;
     bus->initTx = &dev->initTx;
     bus->initRx = &dev->initRx;
@@ -743,28 +706,62 @@ uint8_t spiGetExtDeviceCount(const extDevice_t *dev)
 void spiSequence(const extDevice_t *dev, busSegment_t *segments)
 {
     busDevice_t *bus = dev->bus;
+    busSegment_t *queuedEndSegments[NUM_QUEUE_SEGS];
+    uint8_t queuedEndSegmentsIndex = 0;
 
     ATOMIC_BLOCK(NVIC_PRIO_MAX) {
-        if ((bus->curSegment != (busSegment_t *)BUS_SPI_LOCKED) && spiIsBusy(dev)) {
+        if (spiIsBusy(dev)) {
-            /* Defer this transfer to be triggered upon completion of the current transfer. Blocking calls
+            busSegment_t *endSegment;
-             * and those from non-interrupt context will have already called spiWaitClaim() so this will
-             * only happen for non-blocking calls called from an ISR.
-             */
-            busSegment_t *endSegment = bus->curSegment;
 
-            if (endSegment) {
+            // Defer this transfer to be triggered upon completion of the current transfer
-                // Find the last segment of the current transfer
+            // Find the last segment of the current transfer
-                for (; endSegment->len; endSegment++);
+            for (endSegment = segments; endSegment->len; endSegment++);
+
+            busSegment_t *endCmpSegment = bus->curSegment;
+
+            if (endCmpSegment) {
+                while (true) {
+                    // Find the last segment of the current transfer
+                    for (; endCmpSegment->len; endCmpSegment++);
+
+                    if (endCmpSegment == endSegment) {
+                        // Attempt to use the new segment list twice in the same queue. Abort.
+                        return;
+                    }
+
+                    for (uint8_t n = 0; n < queuedEndSegmentsIndex; n++) {
+                        if (endCmpSegment == queuedEndSegments[n]) {
+                            // Attempt to use the same segment list twice in the same queue. Abort.
+                            return;
+                        }
+                    }
+
+                    queuedEndSegments[queuedEndSegmentsIndex++] = endCmpSegment;
+
+                    if (queuedEndSegmentsIndex == NUM_QUEUE_SEGS) {
+                        // Queue is too long. Abort.
+                        return;
+                    }
+
+                    if (endCmpSegment->txData == NULL) {
+                        break;
+                    } else {
+                        endCmpSegment = (busSegment_t *)endCmpSegment->rxData;
+                    }
+                }
 
                 // Record the dev and segments parameters in the terminating segment entry
-                endSegment->txData = (uint8_t *)dev;
+                endCmpSegment->txData = (uint8_t *)dev;
-                endSegment->rxData = (uint8_t *)segments;
+                endCmpSegment->rxData = (uint8_t *)segments;
 
                 return;
             }
+        } else {
+            // Claim the bus with this list of segments
+            bus->curSegment = segments;
         }
     }
 
-    spiSequenceStart(dev, segments);
+    spiSequenceStart(dev);
 }
 #endif

src/main/drivers/bus_spi.h

@@ -116,6 +116,8 @@ SPI_TypeDef *spiInstanceByDevice(SPIDevice device);
 bool spiSetBusInstance(extDevice_t *dev, uint32_t device);
 // Determine the divisor to use for a given bus frequency
 uint16_t spiCalculateDivider(uint32_t freq);
+// Return the SPI clock based on the given divisor
+uint32_t spiCalculateClock(uint16_t spiClkDivisor);
 // Set the clock divisor to be used for accesses by the given device
 void spiSetClkDivisor(const extDevice_t *dev, uint16_t divider);
 // Set the clock phase/polarity to be used for accesses by the given device
@@ -127,10 +129,6 @@ void spiDmaEnable(const extDevice_t *dev, bool enable);
 void spiSequence(const extDevice_t *dev, busSegment_t *segments);
 // Wait for DMA completion
 void spiWait(const extDevice_t *dev);
-// Indicate that the bus on which this device resides may initiate DMA transfers from interrupt context
-void spiSetAtomicWait(const extDevice_t *dev);
-// Wait for DMA completion and claim the bus driver - use this when waiting for a prior access to complete before starting a new one
-void spiWaitClaim(const extDevice_t *dev);
 // Return true if DMA engine is busy
 bool spiIsBusy(const extDevice_t *dev);
 

src/main/drivers/bus_spi_impl.h

@@ -33,7 +33,6 @@
 #endif
 
 #define BUS_SPI_FREE   0x0
-#define BUS_SPI_LOCKED 0x4
 
 typedef struct spiPinDef_s {
     ioTag_t pin;
@@ -90,5 +89,5 @@ void spiInternalStartDMA(const extDevice_t *dev);
 void spiInternalStopDMA (const extDevice_t *dev);
 void spiInternalResetStream(dmaChannelDescriptor_t *descriptor);
 void spiInternalResetDescriptors(busDevice_t *bus);
-void spiSequenceStart(const extDevice_t *dev, busSegment_t *segments);
+void spiSequenceStart(const extDevice_t *dev);
 

src/main/drivers/bus_spi_ll.c

@@ -539,7 +539,7 @@ void spiInternalStopDMA (const extDevice_t *dev)
 }
 
 // DMA transfer setup and start
-void spiSequenceStart(const extDevice_t *dev, busSegment_t *segments)
+void spiSequenceStart(const extDevice_t *dev)
 {
     busDevice_t *bus = dev->bus;
     SPI_TypeDef *instance = bus->busType_u.spi.instance;
@@ -549,7 +549,6 @@ void spiSequenceStart(const extDevice_t *dev, busSegment_t *segments)
     uint32_t segmentCount = 0;
 
     bus->initSegment = true;
-    bus->curSegment = segments;
 
     // Switch bus speed
 #if !defined(STM32H7)
@@ -689,8 +688,10 @@ void spiSequenceStart(const extDevice_t *dev, busSegment_t *segments)
         if (bus->curSegment->txData) {
             const extDevice_t *nextDev = (const extDevice_t *)bus->curSegment->txData;
             busSegment_t *nextSegments = (busSegment_t *)bus->curSegment->rxData;
-            bus->curSegment->txData = NULL;
+            busSegment_t *endSegment = bus->curSegment;
-            spiSequenceStart(nextDev, nextSegments);
+            bus->curSegment = nextSegments;
+            endSegment->txData = NULL;
+            spiSequenceStart(nextDev);
         } else {
             // The end of the segment list has been reached, so mark transactions as complete
             bus->curSegment = (busSegment_t *)BUS_SPI_FREE;

src/main/drivers/bus_spi_stdperiph.c

@@ -320,7 +320,7 @@ void spiInternalStopDMA (const extDevice_t *dev)
 }
 
 // DMA transfer setup and start
-void spiSequenceStart(const extDevice_t *dev, busSegment_t *segments)
+void spiSequenceStart(const extDevice_t *dev)
 {
     busDevice_t *bus = dev->bus;
     SPI_TypeDef *instance = bus->busType_u.spi.instance;
@@ -329,7 +329,6 @@ void spiSequenceStart(const extDevice_t *dev, busSegment_t *segments)
     uint32_t segmentCount = 0;
 
     dev->bus->initSegment = true;
-    dev->bus->curSegment = segments;
 
     SPI_Cmd(instance, DISABLE);
 
@@ -415,8 +414,10 @@ void spiSequenceStart(const extDevice_t *dev, busSegment_t *segments)
         if (bus->curSegment->txData) {
             const extDevice_t *nextDev = (const extDevice_t *)bus->curSegment->txData;
             busSegment_t *nextSegments = (busSegment_t *)bus->curSegment->rxData;
-            bus->curSegment->txData = NULL;
+            busSegment_t *endSegment = bus->curSegment;
-            spiSequenceStart(nextDev, nextSegments);
+            bus->curSegment = nextSegments;
+            endSegment->txData = NULL;
+            spiSequenceStart(nextDev);
         } else {
             // The end of the segment list has been reached, so mark transactions as complete
             bus->curSegment = (busSegment_t *)BUS_SPI_FREE;

src/main/drivers/flash_m25p16.c

@@ -288,7 +288,7 @@ static void m25p16_eraseSector(flashDevice_t *fdevice, uint32_t address)
     };
 
     // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(fdevice->io.handle.dev);
+    spiWait(fdevice->io.handle.dev);
 
     m25p16_setCommandAddress(&sectorErase[1], address, fdevice->isLargeFlash);
 
@@ -312,9 +312,6 @@ static void m25p16_eraseCompletely(flashDevice_t *fdevice)
             {NULL, NULL, 0, true, NULL},
     };
 
-    // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(fdevice->io.handle.dev);
-
     spiSequence(fdevice->io.handle.dev, segments);
 
     // Block pending completion of SPI access, but the erase will be ongoing
@@ -346,7 +343,7 @@ static uint32_t m25p16_pageProgramContinue(flashDevice_t *fdevice, uint8_t const
     };
 
     // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(fdevice->io.handle.dev);
+    spiWait(fdevice->io.handle.dev);
 
     // Patch the pageProgram segment
     segments[PAGE_PROGRAM].len = fdevice->isLargeFlash ? 5 : 4;
@@ -427,7 +424,7 @@ static int m25p16_readBytes(flashDevice_t *fdevice, uint32_t address, uint8_t *b
     STATIC_DMA_DATA_AUTO uint8_t readBytes[5] = { M25P16_INSTRUCTION_READ_BYTES };
 
     // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(fdevice->io.handle.dev);
+    spiWait(fdevice->io.handle.dev);
 
     busSegment_t segments[] = {
             {readStatus, readyStatus, sizeof(readStatus), true, m25p16_callbackReady},

src/main/drivers/flash_w25n01g.c

@@ -154,9 +154,6 @@ static void w25n01g_performOneByteCommand(flashDeviceIO_t *io, uint8_t command)
                 {NULL, NULL, 0, true, NULL},
         };
 
-        // Ensure any prior DMA has completed before continuing
-        spiWaitClaim(dev);
-
         spiSequence(dev, &segments[0]);
 
         // Block pending completion of SPI access
@@ -182,9 +179,6 @@ static void w25n01g_performCommandWithPageAddress(flashDeviceIO_t *io, uint8_t c
                 {NULL, NULL, 0, true, NULL},
         };
 
-        // Ensure any prior DMA has completed before continuing
-        spiWaitClaim(dev);
-
         spiSequence(dev, &segments[0]);
 
         // Block pending completion of SPI access
@@ -419,9 +413,6 @@ static void w25n01g_programDataLoad(flashDevice_t *fdevice, uint16_t columnAddre
                  {NULL, NULL, 0, true, NULL},
          };
 
-         // Ensure any prior DMA has completed before continuing
-         spiWaitClaim(dev);
-
          spiSequence(dev, &segments[0]);
 
          // Block pending completion of SPI access
@@ -453,9 +444,6 @@ static void w25n01g_randomProgramDataLoad(flashDevice_t *fdevice, uint16_t colum
                 {NULL, NULL, 0, true, NULL},
         };
 
-        // Ensure any prior DMA has completed before continuing
-        spiWaitClaim(dev);
-
         spiSequence(dev, &segments[0]);
 
         // Block pending completion of SPI access
@@ -699,9 +687,6 @@ int w25n01g_readBytes(flashDevice_t *fdevice, uint32_t address, uint8_t *buffer,
                 {NULL, NULL, 0, true, NULL},
         };
 
-        // Ensure any prior DMA has completed before continuing
-        spiWaitClaim(dev);
-
         spiSequence(dev, &segments[0]);
 
         // Block pending completion of SPI access
@@ -867,9 +852,6 @@ void w25n01g_readBBLUT(flashDevice_t *fdevice, bblut_t *bblut, int lutsize)
                 {NULL, NULL, 0, true, NULL},
         };
 
-        // Ensure any prior DMA has completed before continuing
-        spiWaitClaim(dev);
-
         spiSequence(dev, &segments[0]);
 
         // Block pending completion of SPI access

src/main/drivers/max7456.c

@@ -56,6 +56,7 @@
 #define DEBUG_MAX7456_SPICLOCK_OVERCLOCK   0
 #define DEBUG_MAX7456_SPICLOCK_DEVTYPE     1
 #define DEBUG_MAX7456_SPICLOCK_DIVISOR     2
+#define DEBUG_MAX7456_SPICLOCK_X100        3
 
 // VM0 bits
 #define VIDEO_BUFFER_DISABLE        0x01
@@ -192,7 +193,7 @@ extDevice_t max7456Device;
 extDevice_t *dev = &max7456Device;
 
 static bool max7456DeviceDetected = false;
-static uint16_t max7456SpiClock;
+static uint16_t max7456SpiClockDiv;
 
 uint16_t maxScreenSize = VIDEO_BUFFER_CHARS_PAL;
 
@@ -394,32 +395,35 @@ max7456InitStatus_e max7456Init(const max7456Config_t *max7456Config, const vcdP
     }
 
 #if defined(USE_OVERCLOCK)
+    max7456SpiClockDiv = spiCalculateDivider(MAX7456_MAX_SPI_CLK_HZ);
+
     // Determine SPI clock divisor based on config and the device type.
 
     switch (max7456Config->clockConfig) {
     case MAX7456_CLOCK_CONFIG_HALF:
-        max7456SpiClock = spiCalculateDivider(MAX7456_MAX_SPI_CLK_HZ / 2);
+        max7456SpiClockDiv <<= 1;
         break;
 
-    case MAX7456_CLOCK_CONFIG_OC:
+    case MAX7456_CLOCK_CONFIG_NOMINAL:
-        max7456SpiClock = (cpuOverclock && (max7456DeviceType == MAX7456_DEVICE_TYPE_MAX)) ? spiCalculateDivider(MAX7456_MAX_SPI_CLK_HZ / 2) : spiCalculateDivider(MAX7456_MAX_SPI_CLK_HZ);
+    default:
         break;
 
-    case MAX7456_CLOCK_CONFIG_FULL:
+    case MAX7456_CLOCK_CONFIG_DOUBLE:
-        max7456SpiClock = spiCalculateDivider(MAX7456_MAX_SPI_CLK_HZ);
+        max7456SpiClockDiv >>= 1;
         break;
     }
 
     DEBUG_SET(DEBUG_MAX7456_SPICLOCK, DEBUG_MAX7456_SPICLOCK_OVERCLOCK, cpuOverclock);
     DEBUG_SET(DEBUG_MAX7456_SPICLOCK, DEBUG_MAX7456_SPICLOCK_DEVTYPE, max7456DeviceType);
-    DEBUG_SET(DEBUG_MAX7456_SPICLOCK, DEBUG_MAX7456_SPICLOCK_DIVISOR, max7456SpiClock);
+    DEBUG_SET(DEBUG_MAX7456_SPICLOCK, DEBUG_MAX7456_SPICLOCK_DIVISOR, max7456SpiClockDiv);
+    DEBUG_SET(DEBUG_MAX7456_SPICLOCK, DEBUG_MAX7456_SPICLOCK_X100, spiCalculateClock(max7456SpiClockDiv) / 10000);
 #else
     UNUSED(max7456Config);
     UNUSED(cpuOverclock);
-    max7456SpiClock = spiCalculateDivider(MAX7456_MAX_SPI_CLK_HZ);
+    max7456SpiClockDiv = spiCalculateDivider(MAX7456_MAX_SPI_CLK_HZ);
 #endif
 
-    spiSetClkDivisor(dev, max7456SpiClock);
+    spiSetClkDivisor(dev, max7456SpiClockDiv);
 
     // force soft reset on Max7456
     spiWriteReg(dev, MAX7456ADD_VM0, MAX7456_RESET);
@@ -631,9 +635,6 @@ bool max7456DrawScreen(void)
             return true;
         }
 
-        // Ensure any prior DMA has completed before overwriting the buffer
-        spiWaitClaim(dev);
-
         // Allow for 8 bytes followed by an ESCAPE and reset of DMM at end of buffer
         maxSpiBufStartIndex -= 12;
 

src/main/drivers/rx/rx_spi.c

@@ -98,8 +98,6 @@ bool rxSpiDeviceInit(const rxSpiConfig_t *rxSpiConfig)
         return false;
     }
 
-    spiSetAtomicWait(dev);
-
     const IO_t rxCsPin = IOGetByTag(rxSpiConfig->csnTag);
     IOInit(rxCsPin, OWNER_RX_SPI_CS, 0);
     IOConfigGPIO(rxCsPin, SPI_IO_CS_CFG);

src/main/drivers/sdcard_spi.c

@@ -174,9 +174,6 @@ static bool sdcard_waitForIdle(int maxBytesToWait)
 
     sdcard.idleCount = maxBytesToWait;
 
-    // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(&sdcard.dev);
-
     spiSequence(&sdcard.dev, &segments[0]);
 
     // Block pending completion of SPI access
@@ -202,9 +199,6 @@ static uint8_t sdcard_waitForNonIdleByte(int maxDelay)
 
     sdcard.idleCount = maxDelay;
 
-    // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(&sdcard.dev);
-
     spiSequence(&sdcard.dev, &segments[0]);
 
     // Block pending completion of SPI access
@@ -248,9 +242,6 @@ static uint8_t sdcard_sendCommand(uint8_t commandCode, uint32_t commandArgument)
 
     sdcard.idleCount = SDCARD_MAXIMUM_BYTE_DELAY_FOR_CMD_REPLY;
 
-    // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(&sdcard.dev);
-
     spiSequence(&sdcard.dev, &segments[0]);
 
     // Block pending completion of SPI access
@@ -374,9 +365,6 @@ static sdcardReceiveBlockStatus_e sdcard_receiveDataBlock(uint8_t *buffer, int c
             {NULL, NULL, 0, false, NULL},
         };
 
-    // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(&sdcard.dev);
-
     spiSequence(&sdcard.dev, &segments[0]);
 
     // Block pending completion of SPI access
@@ -396,9 +384,6 @@ static bool sdcard_sendDataBlockFinish(void)
             {NULL, NULL, 0, false, NULL},
         };
 
-    // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(&sdcard.dev);
-
     spiSequence(&sdcard.dev, &segments[0]);
 
     // Block pending completion of SPI access
@@ -440,9 +425,6 @@ static void sdcard_sendDataBlockBegin(uint8_t *buffer, bool multiBlockWrite)
     segments[2].txData = buffer;
     segments[2].len = spiUseDMA(&sdcard.dev) ? SDCARD_BLOCK_SIZE : SDCARD_NON_DMA_CHUNK_SIZE;
 
-    // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(&sdcard.dev);
-
     spiSequence(&sdcard.dev, &segments[0]);
 
     // Don't block pending completion of SPI access
@@ -584,9 +566,6 @@ static void sdcardSpi_init(const sdcardConfig_t *config, const spiPinConfig_t *s
             {NULL, NULL, 0, false, NULL},
         };
 
-    // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(&sdcard.dev);
-
     spiSequence(&sdcard.dev, &segments[0]);
 
     // Block pending completion of SPI access
@@ -639,9 +618,6 @@ static sdcardOperationStatus_e sdcard_endWriteBlocks(void)
             {NULL, NULL, 0, false, NULL},
         };
 
-    // Ensure any prior DMA has completed before continuing
-    spiWaitClaim(&sdcard.dev);
-
     spiSequence(&sdcard.dev, &segments[0]);
 
     // Block pending completion of SPI access

src/main/pg/max7456.c

@@ -34,7 +34,7 @@ PG_REGISTER_WITH_RESET_FN(max7456Config_t, max7456Config, PG_MAX7456_CONFIG, 0);
 
 void pgResetFn_max7456Config(max7456Config_t *config)
 {
-    config->clockConfig = MAX7456_CLOCK_CONFIG_DEFAULT;
+    config->clockConfig = MAX7456_CLOCK_CONFIG_NOMINAL;
     config->csTag = IO_TAG(MAX7456_SPI_CS_PIN);
     config->spiDevice = SPI_DEV_TO_CFG(spiDeviceByInstance(MAX7456_SPI_INSTANCE));
     config->preInitOPU = false;

src/main/pg/max7456.h

@@ -24,15 +24,15 @@
 #include "pg/pg.h"
 
 typedef struct max7456Config_s {
-    uint8_t clockConfig; // SPI clock based on device type and overclock state (MAX7456_CLOCK_CONFIG_xxxx)
+    uint8_t clockConfig; // SPI clock modifier
     ioTag_t csTag;
     uint8_t spiDevice;
     bool preInitOPU;
 } max7456Config_t;
 
 // clockConfig values
-#define MAX7456_CLOCK_CONFIG_HALF 0  // Force half clock
+#define MAX7456_CLOCK_CONFIG_HALF       0  // Force half clock
-#define MAX7456_CLOCK_CONFIG_OC   1  // Half clock if OC
+#define MAX7456_CLOCK_CONFIG_NOMINAL    1  // Nominal clock (default)
-#define MAX7456_CLOCK_CONFIG_FULL 2  // Force full clock
+#define MAX7456_CLOCK_CONFIG_DOUBLE     2  // Double clock
 
 PG_DECLARE(max7456Config_t, max7456Config);