/* esp8266_waveform - General purpose waveform generation and control, supporting outputs on all pins in parallel. Copyright (c) 2018 Earle F. Philhower, III. All rights reserved. The core idea is to have a programmable waveform generator with a unique high and low period (defined in microseconds or CPU clock cycles). TIMER1 is set to 1-shot mode and is always loaded with the time until the next edge of any live waveforms. Up to one waveform generator per pin supported. Each waveform generator is synchronized to the ESP clock cycle counter, not the timer. This allows for removing interrupt jitter and delay as the counter always increments once per 80MHz clock. Changes to a waveform are contiguous and only take effect on the next waveform transition, allowing for smooth transitions. This replaces older tone(), analogWrite(), and the Servo classes. Everywhere in the code where "cycles" is used, it means ESP.getCycleCount() clock cycle count, or an interval measured in CPU clock cycles, but not TIMER1 cycles (which may be 2 CPU clock cycles @ 160MHz). This library is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation; either version 2.1 of the License, or (at your option) any later version. This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details. You should have received a copy of the GNU Lesser General Public License along with this library; if not, write to the Free Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */ #ifdef ESP8266 #include #include "ets_sys.h" #include "core_esp8266_waveform.h" #include "user_interface.h" extern "C" { // Internal-only calls, not for applications extern void _setPWMPeriodCC(uint32_t cc); extern bool _stopPWM(int pin); extern bool _setPWM(int pin, uint32_t cc); extern int startWaveformClockCycles(uint8_t pin, uint32_t timeHighCycles, uint32_t timeLowCycles, uint32_t runTimeCycles); // Maximum delay between IRQs #define MAXIRQUS (10000) // Set/clear GPIO 0-15 by bitmask #define SetGPIO(a) do { GPOS = a; } while (0) #define ClearGPIO(a) do { GPOC = a; } while (0) // Waveform generator can create tones, PWM, and servos typedef struct { uint32_t nextServiceCycle; // ESP cycle timer when a transition required uint32_t expiryCycle; // For time-limited waveform, the cycle when this waveform must stop uint32_t timeHighCycles; // Currently running waveform period uint32_t timeLowCycles; // uint32_t desiredHighCycles; // Currently running waveform period uint32_t desiredLowCycles; // uint32_t gotoTimeHighCycles; // Copied over on the next period to preserve phase uint32_t gotoTimeLowCycles; // uint32_t lastEdge; // } Waveform; class WVFState { public: Waveform waveform[17]; // State of all possible pins uint32_t waveformState = 0; // Is the pin high or low, updated in NMI so no access outside the NMI code uint32_t waveformEnabled = 0; // Is it actively running, updated in NMI so no access outside the NMI code // Enable lock-free by only allowing updates to waveformState and waveformEnabled from IRQ service routine uint32_t waveformToEnable = 0; // Message to the NMI handler to start a waveform on a inactive pin uint32_t waveformToDisable = 0; // Message to the NMI handler to disable a pin from waveform generation int32_t waveformToChange = -1; uint32_t waveformNewHigh = 0; uint32_t waveformNewLow = 0; uint32_t (*timer1CB)() = NULL; // Optimize the NMI inner loop by keeping track of the min and max GPIO that we // are generating. In the common case (1 PWM) these may be the same pin and // we can avoid looking at the other pins. int startPin = 0; int endPin = 0; }; static WVFState wvfState; // Ensure everything is read/written to RAM #define MEMBARRIER() { __asm__ volatile("" ::: "memory"); } // Non-speed critical bits #pragma GCC optimize ("Os") // Interrupt on/off control static ICACHE_RAM_ATTR void timer1Interrupt(); static bool timerRunning = false; static void initTimer() { if (!timerRunning) { timer1_disable(); ETS_FRC_TIMER1_INTR_ATTACH(NULL, NULL); ETS_FRC_TIMER1_NMI_INTR_ATTACH(timer1Interrupt); timer1_enable(TIM_DIV1, TIM_EDGE, TIM_SINGLE); timerRunning = true; timer1_write(microsecondsToClockCycles(10)); } } static ICACHE_RAM_ATTR void forceTimerInterrupt() { if (T1L > microsecondsToClockCycles(10)) { T1L = microsecondsToClockCycles(10); } } // PWM implementation using special purpose state machine // // Keep an ordered list of pins with the delta in cycles between each // element, with a terminal entry making up the remainder of the PWM // period. With this method sum(all deltas) == PWM period clock cycles. // // At t=0 set all pins high and set the timeout for the 1st edge. // On interrupt, if we're at the last element reset to t=0 state // Otherwise, clear that pin down and set delay for next element // and so forth. constexpr int maxPWMs = 8; // PWM machine state typedef struct { uint32_t mask; // Bitmask of active pins uint32_t cnt; // How many entries uint32_t idx; // Where the state machine is along the list uint8_t pin[maxPWMs + 1]; uint32_t delta[maxPWMs + 1]; uint32_t nextServiceCycle; // Clock cycle for next step } PWMState; static PWMState pwmState; static PWMState *pwmUpdate = nullptr; // Set by main code, cleared by ISR static uint32_t pwmPeriod = microsecondsToClockCycles(1000000UL) / 1000; static ICACHE_RAM_ATTR void disableIdleTimer() { if (timerRunning && !wvfState.waveformEnabled && !pwmState.cnt && !wvfState.timer1CB) { ETS_FRC_TIMER1_NMI_INTR_ATTACH(NULL); timer1_disable(); timer1_isr_init(); timerRunning = false; } } // Called when analogWriteFreq() changed to update the PWM total period void _setPWMPeriodCC(uint32_t cc) { if (cc == pwmPeriod) { return; } if (pwmState.cnt) { // Adjust any running ones to the best of our abilities by scaling them // Used FP math for speed and code size uint64_t oldCC64p0 = ((uint64_t)pwmPeriod); uint64_t newCC64p16 = ((uint64_t)cc) << 16; uint64_t ratio64p16 = (newCC64p16 / oldCC64p0); PWMState p; // The working copy since we can't edit the one in use p = pwmState; uint32_t ttl = 0; for (uint32_t i = 0; i < p.cnt; i++) { uint64_t val64p16 = ((uint64_t)p.delta[i]) << 16; uint64_t newVal64p32 = val64p16 * ratio64p16; p.delta[i] = newVal64p32 >> 32; ttl += p.delta[i]; } p.delta[p.cnt] = cc - ttl; // Final cleanup exactly cc total cycles // Update and wait for mailbox to be emptied pwmUpdate = &p; MEMBARRIER(); forceTimerInterrupt(); while (pwmUpdate) { delay(0); // No mem barrier. The external function call guarantees it's re-read } } pwmPeriod = cc; } // Helper routine to remove an entry from the state machine static ICACHE_RAM_ATTR void _removePWMEntry(int pin, PWMState *p) { uint32_t i; // Find the pin to pull out... for (i = 0; p->pin[i] != pin; i++) { /* no-op */ } auto delta = p->delta[i]; // Add the removed previous pin delta to preserve absolute position p->delta[i+1] += delta; // Move everything back one for (i++; i <= p->cnt; i++) { p->pin[i-1] = p->pin[i]; p->delta[i-1] = p->delta[i]; } // Remove the pin from the active list p->mask &= ~(1<cnt--; } // Called by analogWrite(0/100%) to disable PWM on a specific pin ICACHE_RAM_ATTR bool _stopPWM(int pin) { if (!((1<= maxPWMs) { return false; // No space left } else if (p.cnt == 0) { // Starting up from scratch, special case 1st element and PWM period p.pin[0] = pin; p.delta[0] = cc; p.pin[1] = 0xff; p.delta[1] = pwmPeriod - cc; p.cnt = 1; p.mask = 1<high transition. For immediate change, stopWaveform() // first, then it will immediately begin. int startWaveform(uint8_t pin, uint32_t timeHighUS, uint32_t timeLowUS, uint32_t runTimeUS) { return startWaveformClockCycles(pin, microsecondsToClockCycles(timeHighUS), microsecondsToClockCycles(timeLowUS), microsecondsToClockCycles(runTimeUS)); } int startWaveformClockCycles(uint8_t pin, uint32_t timeHighCycles, uint32_t timeLowCycles, uint32_t runTimeCycles) { if ((pin > 16) || isFlashInterfacePin(pin)) { return false; } Waveform *wave = &wvfState.waveform[pin]; wave->expiryCycle = runTimeCycles ? ESP.getCycleCount() + runTimeCycles : 0; if (runTimeCycles && !wave->expiryCycle) { wave->expiryCycle = 1; // expiryCycle==0 means no timeout, so avoid setting it } _stopPWM(pin); // Make sure there's no PWM live here uint32_t mask = 1<= 0) { delay(0); // Wait for waveform to update // No mem barrier here, the call to a global function implies global state updated } } else { // if (!(wvfState.waveformEnabled & mask)) { wave->timeHighCycles = timeHighCycles; wave->timeLowCycles = timeLowCycles; wave->desiredHighCycles = wave->timeHighCycles; wave->desiredLowCycles = wave->timeLowCycles; wave->lastEdge = 0; wave->gotoTimeHighCycles = wave->timeHighCycles; wave->gotoTimeLowCycles = wave->timeLowCycles; // Actually set the pin high or low in the IRQ service to guarantee times wave->nextServiceCycle = ESP.getCycleCount() + microsecondsToClockCycles(1); wvfState.waveformToEnable |= mask; MEMBARRIER(); initTimer(); forceTimerInterrupt(); while (wvfState.waveformToEnable) { delay(0); // Wait for waveform to update // No mem barrier here, the call to a global function implies global state updated } } return true; } // Set a callback. Pass in NULL to stop it void setTimer1Callback(uint32_t (*fn)()) { wvfState.timer1CB = fn; if (fn) { initTimer(); forceTimerInterrupt(); } disableIdleTimer(); } // Speed critical bits #pragma GCC optimize ("O2") // Normally would not want two copies like this, but due to different // optimization levels the inline attribute gets lost if we try the // other version. static inline ICACHE_RAM_ATTR uint32_t GetCycleCountIRQ() { uint32_t ccount; __asm__ __volatile__("rsr %0,ccount":"=a"(ccount)); return ccount; } static inline ICACHE_RAM_ATTR uint32_t min_u32(uint32_t a, uint32_t b) { if (a < b) { return a; } return b; } // Stops a waveform on a pin int ICACHE_RAM_ATTR stopWaveform(uint8_t pin) { // Can't possibly need to stop anything if there is no timer active if (!timerRunning) { return false; } // If user sends in a pin >16 but <32, this will always point to a 0 bit // If they send >=32, then the shift will result in 0 and it will also return false if (wvfState.waveformEnabled & (1UL << pin)) { wvfState.waveformToDisable = 1UL << pin; forceTimerInterrupt(); while (wvfState.waveformToDisable) { MEMBARRIER(); // If it wasn't written yet, it has to be by now /* no-op */ // Can't delay() since stopWaveform may be called from an IRQ } } disableIdleTimer(); return true; } // The SDK and hardware take some time to actually get to our NMI code, so // decrement the next IRQ's timer value by a bit so we can actually catch the // real CPU cycle counter we want for the waveforms. // The SDK also sometimes is running at a different speed the the Arduino core // so the ESP cycle counter is actually running at a variable speed. // adjust(x) takes care of adjusting a delta clock cycle amount accordingly. #if F_CPU == 80000000 #define DELTAIRQ (microsecondsToClockCycles(3)) #define adjust(x) ((x) << (turbo ? 1 : 0)) #else #define DELTAIRQ (microsecondsToClockCycles(2)) #define adjust(x) ((x) >> (turbo ? 0 : 1)) #endif #define ENABLE_ADJUST // Adjust takes 36 bytes #define ENABLE_FEEDBACK // Feedback costs 68 bytes #define ENABLE_PWM // PWM takes 160 bytes #ifndef ENABLE_ADJUST #undef adjust #define adjust(x) (x) #endif static ICACHE_RAM_ATTR void timer1Interrupt() { // Flag if the core is at 160 MHz, for use by adjust() bool turbo = (*(uint32_t*)0x3FF00014) & 1 ? true : false; uint32_t nextEventCycles = microsecondsToClockCycles(MAXIRQUS); uint32_t timeoutCycle = GetCycleCountIRQ() + microsecondsToClockCycles(14); if (wvfState.waveformToEnable || wvfState.waveformToDisable) { // Handle enable/disable requests from main app wvfState.waveformEnabled = (wvfState.waveformEnabled & ~wvfState.waveformToDisable) | wvfState.waveformToEnable; // Set the requested waveforms on/off wvfState.waveformState &= ~wvfState.waveformToEnable; // And clear the state of any just started wvfState.waveformToEnable = 0; wvfState.waveformToDisable = 0; // No mem barrier. Globals must be written to RAM on ISR exit. // Find the first GPIO being generated by checking GCC's find-first-set (returns 1 + the bit of the first 1 in an int32_t) wvfState.startPin = __builtin_ffs(wvfState.waveformEnabled) - 1; // Find the last bit by subtracting off GCC's count-leading-zeros (no offset in this one) wvfState.endPin = 32 - __builtin_clz(wvfState.waveformEnabled); #ifdef ENABLE_PWM } else if (!pwmState.cnt && pwmUpdate) { // Start up the PWM generator by copying from the mailbox pwmState.cnt = 1; pwmState.idx = 1; // Ensure copy this cycle, cause it to start at t=0 pwmState.nextServiceCycle = GetCycleCountIRQ(); // Do it this loop! // No need for mem barrier here. Global must be written by IRQ exit #endif } else if (wvfState.waveformToChange >= 0) { wvfState.waveform[wvfState.waveformToChange].gotoTimeHighCycles = wvfState.waveformNewHigh; wvfState.waveform[wvfState.waveformToChange].gotoTimeLowCycles = wvfState.waveformNewLow; wvfState.waveformToChange = -1; // No need for memory barrier here. The global has to be written before exit the ISR. } bool done = false; if (wvfState.waveformEnabled || pwmState.cnt) { do { nextEventCycles = microsecondsToClockCycles(MAXIRQUS); #ifdef ENABLE_PWM // PWM state machine implementation if (pwmState.cnt) { uint32_t now = GetCycleCountIRQ(); int32_t cyclesToGo = pwmState.nextServiceCycle - now; if (cyclesToGo < 0) { if (pwmState.idx == pwmState.cnt) { // Start of pulses, possibly copy new if (pwmUpdate) { // Do the memory copy from temp to global and clear mailbox pwmState = *(PWMState*)pwmUpdate; pwmUpdate = nullptr; } GPOS = pwmState.mask; // Set all active pins high // GPIO16 isn't the same as the others if (pwmState.mask & (1<<16)) { GP16O = 1; } pwmState.idx = 0; } else { do { // Drop the pin at this edge GPOC = 1<expiryCycle) { int32_t expiryToGo = wave->expiryCycle - now; if (expiryToGo < 0) { // Done, remove! wvfState.waveformEnabled &= ~mask; if (i == 16) { GP16O = 0; } else { ClearGPIO(mask); } continue; } } // Check for toggles int32_t cyclesToGo = wave->nextServiceCycle - now; if (cyclesToGo < 0) { uint32_t nextEdgeCycles; uint32_t desired = 0; uint32_t *timeToUpdate; wvfState.waveformState ^= mask; if (wvfState.waveformState & mask) { if (i == 16) { GP16O = 1; // GPIO16 write slow as it's RMW } else { SetGPIO(mask); } if (wave->gotoTimeHighCycles) { // Copy over next full-cycle timings wave->timeHighCycles = wave->gotoTimeHighCycles; wave->desiredHighCycles = wave->gotoTimeHighCycles; wave->timeLowCycles = wave->gotoTimeLowCycles; wave->desiredLowCycles = wave->gotoTimeLowCycles; wave->gotoTimeHighCycles = 0; } else { #ifdef ENABLE_FEEDBACK if (wave->lastEdge) { desired = wave->desiredLowCycles; timeToUpdate = &wave->timeLowCycles; } } #endif nextEdgeCycles = wave->timeHighCycles; } else { if (i == 16) { GP16O = 0; // GPIO16 write slow as it's RMW } else { ClearGPIO(mask); } #ifdef ENABLE_FEEDBACK desired = wave->desiredHighCycles; timeToUpdate = &wave->timeHighCycles; #endif nextEdgeCycles = wave->timeLowCycles; } #ifdef ENABLE_FEEDBACK if (desired) { desired = adjust(desired); int32_t err = desired - (now - wave->lastEdge); if (abs(err) < desired) { // If we've lost > the entire phase, ignore this error signal err /= 2; *timeToUpdate += err; } } #endif nextEdgeCycles = adjust(nextEdgeCycles); wave->nextServiceCycle = now + nextEdgeCycles; nextEventCycles = min_u32(nextEventCycles, nextEdgeCycles); wave->lastEdge = now; } else { uint32_t deltaCycles = wave->nextServiceCycle - now; nextEventCycles = min_u32(nextEventCycles, deltaCycles); } } // Exit the loop if we've hit the fixed runtime limit or the next event is known to be after that timeout would occur uint32_t now = GetCycleCountIRQ(); int32_t cycleDeltaNextEvent = timeoutCycle - (now + nextEventCycles); int32_t cyclesLeftTimeout = timeoutCycle - now; done = (cycleDeltaNextEvent < 0) || (cyclesLeftTimeout < 0); } while (!done); } // if (wvfState.waveformEnabled) if (wvfState.timer1CB) { nextEventCycles = min_u32(nextEventCycles, wvfState.timer1CB()); } if (nextEventCycles < microsecondsToClockCycles(5)) { nextEventCycles = microsecondsToClockCycles(5); } nextEventCycles -= DELTAIRQ; // Do it here instead of global function to save time and because we know it's edge-IRQ T1L = nextEventCycles >> (turbo ? 1 : 0); TEIE |= TEIE1; // Edge int enable } }; #endif // ESP8266