Files
allwpilib/wpilibj/src/main/java/edu/wpi/first/wpilibj/DriverStation.java

1294 lines
39 KiB
Java
Raw Normal View History

// Copyright (c) FIRST and other WPILib contributors.
// Open Source Software; you can modify and/or share it under the terms of
// the WPILib BSD license file in the root directory of this project.
package edu.wpi.first.wpilibj;
import edu.wpi.first.hal.AllianceStationID;
import edu.wpi.first.hal.ControlWord;
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
import edu.wpi.first.hal.DriverStationJNI;
import edu.wpi.first.hal.HAL;
import edu.wpi.first.hal.MatchInfoData;
2022-10-08 10:01:31 -07:00
import edu.wpi.first.networktables.BooleanPublisher;
import edu.wpi.first.networktables.IntegerPublisher;
import edu.wpi.first.networktables.NetworkTable;
import edu.wpi.first.networktables.NetworkTableInstance;
2022-10-08 10:01:31 -07:00
import edu.wpi.first.networktables.StringPublisher;
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
import edu.wpi.first.util.EventVector;
import edu.wpi.first.util.WPIUtilJNI;
import edu.wpi.first.util.datalog.BooleanArrayLogEntry;
import edu.wpi.first.util.datalog.BooleanLogEntry;
import edu.wpi.first.util.datalog.DataLog;
import edu.wpi.first.util.datalog.FloatArrayLogEntry;
import edu.wpi.first.util.datalog.IntegerArrayLogEntry;
import java.nio.ByteBuffer;
import java.util.concurrent.locks.ReentrantLock;
/** Provide access to the network communication data to / from the Driver Station. */
public final class DriverStation {
/** Number of Joystick Ports. */
public static final int kJoystickPorts = 6;
private static class HALJoystickButtons {
public int m_buttons;
public byte m_count;
}
private static class HALJoystickAxes {
public float[] m_axes;
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
public int m_count;
2017-10-17 21:47:55 -07:00
HALJoystickAxes(int count) {
m_axes = new float[count];
}
}
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
private static class HALJoystickAxesRaw {
public int[] m_axes;
public int m_count;
HALJoystickAxesRaw(int count) {
m_axes = new int[count];
}
}
private static class HALJoystickPOVs {
public short[] m_povs;
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
public int m_count;
2017-10-17 21:47:55 -07:00
HALJoystickPOVs(int count) {
m_povs = new short[count];
for (int i = 0; i < count; i++) {
m_povs[i] = -1;
}
}
}
/** The robot alliance that the robot is a part of. */
public enum Alliance {
Red,
Blue,
Invalid
}
public enum MatchType {
None,
Practice,
Qualification,
Elimination
}
private static final double JOYSTICK_UNPLUGGED_MESSAGE_INTERVAL = 1.0;
private static double m_nextMessageTime;
@SuppressWarnings("MemberName")
private static class MatchDataSender {
NetworkTable table;
2022-10-08 10:01:31 -07:00
StringPublisher typeMetadata;
StringPublisher gameSpecificMessage;
StringPublisher eventName;
IntegerPublisher matchNumber;
IntegerPublisher replayNumber;
IntegerPublisher matchType;
BooleanPublisher alliance;
IntegerPublisher station;
IntegerPublisher controlWord;
boolean oldIsRedAlliance = true;
int oldStationNumber = 1;
String oldEventName = "";
String oldGameSpecificMessage = "";
int oldMatchNumber;
int oldReplayNumber;
int oldMatchType;
int oldControlWord;
MatchDataSender() {
table = NetworkTableInstance.getDefault().getTable("FMSInfo");
2022-10-08 10:01:31 -07:00
typeMetadata = table.getStringTopic(".type").publish();
typeMetadata.set("FMSInfo");
gameSpecificMessage = table.getStringTopic("GameSpecificMessage").publish();
gameSpecificMessage.set("");
eventName = table.getStringTopic("EventName").publish();
eventName.set("");
matchNumber = table.getIntegerTopic("MatchNumber").publish();
matchNumber.set(0);
replayNumber = table.getIntegerTopic("ReplayNumber").publish();
replayNumber.set(0);
matchType = table.getIntegerTopic("MatchType").publish();
matchType.set(0);
alliance = table.getBooleanTopic("IsRedAlliance").publish();
alliance.set(true);
station = table.getIntegerTopic("StationNumber").publish();
station.set(1);
controlWord = table.getIntegerTopic("FMSControlData").publish();
controlWord.set(0);
}
private void sendMatchData() {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
AllianceStationID allianceID = DriverStationJNI.getAllianceStation();
boolean isRedAlliance = false;
int stationNumber = 1;
switch (allianceID) {
case Blue1:
isRedAlliance = false;
stationNumber = 1;
break;
case Blue2:
isRedAlliance = false;
stationNumber = 2;
break;
case Blue3:
isRedAlliance = false;
stationNumber = 3;
break;
case Red1:
isRedAlliance = true;
stationNumber = 1;
break;
case Red2:
isRedAlliance = true;
stationNumber = 2;
break;
default:
isRedAlliance = true;
stationNumber = 3;
break;
}
String currentEventName;
String currentGameSpecificMessage;
int currentMatchNumber;
int currentReplayNumber;
int currentMatchType;
int currentControlWord;
m_cacheDataMutex.lock();
try {
currentEventName = DriverStation.m_matchInfo.eventName;
currentGameSpecificMessage = DriverStation.m_matchInfo.gameSpecificMessage;
currentMatchNumber = DriverStation.m_matchInfo.matchNumber;
currentReplayNumber = DriverStation.m_matchInfo.replayNumber;
currentMatchType = DriverStation.m_matchInfo.matchType;
} finally {
m_cacheDataMutex.unlock();
}
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
currentControlWord = DriverStationJNI.nativeGetControlWord();
if (oldIsRedAlliance != isRedAlliance) {
2022-10-08 10:01:31 -07:00
alliance.set(isRedAlliance);
oldIsRedAlliance = isRedAlliance;
}
if (oldStationNumber != stationNumber) {
2022-10-08 10:01:31 -07:00
station.set(stationNumber);
oldStationNumber = stationNumber;
}
if (!oldEventName.equals(currentEventName)) {
2022-10-08 10:01:31 -07:00
eventName.set(currentEventName);
oldEventName = currentEventName;
}
if (!oldGameSpecificMessage.equals(currentGameSpecificMessage)) {
2022-10-08 10:01:31 -07:00
gameSpecificMessage.set(currentGameSpecificMessage);
oldGameSpecificMessage = currentGameSpecificMessage;
}
if (currentMatchNumber != oldMatchNumber) {
2022-10-08 10:01:31 -07:00
matchNumber.set(currentMatchNumber);
oldMatchNumber = currentMatchNumber;
}
if (currentReplayNumber != oldReplayNumber) {
2022-10-08 10:01:31 -07:00
replayNumber.set(currentReplayNumber);
oldReplayNumber = currentReplayNumber;
}
if (currentMatchType != oldMatchType) {
2022-10-08 10:01:31 -07:00
matchType.set(currentMatchType);
oldMatchType = currentMatchType;
}
if (currentControlWord != oldControlWord) {
2022-10-08 10:01:31 -07:00
controlWord.set(currentControlWord);
oldControlWord = currentControlWord;
}
}
}
private static class JoystickLogSender {
JoystickLogSender(DataLog log, int stick, long timestamp) {
m_stick = stick;
m_logButtons = new BooleanArrayLogEntry(log, "DS:joystick" + stick + "/buttons", timestamp);
m_logAxes = new FloatArrayLogEntry(log, "DS:joystick" + stick + "/axes", timestamp);
m_logPOVs = new IntegerArrayLogEntry(log, "DS:joystick" + stick + "/povs", timestamp);
appendButtons(timestamp);
appendAxes(timestamp);
appendPOVs(timestamp);
}
public void send(long timestamp) {
if (m_joystickButtonsCache[m_stick].m_count != m_joystickButtons[m_stick].m_count
|| m_joystickButtonsCache[m_stick].m_buttons != m_joystickButtons[m_stick].m_buttons) {
appendButtons(timestamp);
}
if (m_joystickAxesCache[m_stick].m_count != m_joystickAxes[m_stick].m_count) {
appendAxes(timestamp);
} else {
int count = m_joystickAxesCache[m_stick].m_count;
for (int i = 0; i < count; i++) {
if (m_joystickAxesCache[m_stick].m_axes[i] != m_joystickAxes[m_stick].m_axes[i]) {
appendAxes(timestamp);
break;
}
}
}
if (m_joystickPOVsCache[m_stick].m_count != m_joystickPOVs[m_stick].m_count) {
appendPOVs(timestamp);
} else {
int count = m_joystickPOVsCache[m_stick].m_count;
for (int i = 0; i < count; i++) {
if (m_joystickPOVsCache[m_stick].m_povs[i] != m_joystickPOVs[m_stick].m_povs[i]) {
appendPOVs(timestamp);
break;
}
}
}
}
void appendButtons(long timestamp) {
int count = m_joystickButtonsCache[m_stick].m_count;
if (m_sizedButtons == null || m_sizedButtons.length != count) {
m_sizedButtons = new boolean[count];
}
int buttons = m_joystickButtonsCache[m_stick].m_buttons;
for (int i = 0; i < count; i++) {
m_sizedButtons[i] = (buttons & (1 << i)) != 0;
}
m_logButtons.append(m_sizedButtons, timestamp);
}
void appendAxes(long timestamp) {
int count = m_joystickAxesCache[m_stick].m_count;
if (m_sizedAxes == null || m_sizedAxes.length != count) {
m_sizedAxes = new float[count];
}
System.arraycopy(m_joystickAxesCache[m_stick].m_axes, 0, m_sizedAxes, 0, count);
m_logAxes.append(m_sizedAxes, timestamp);
}
void appendPOVs(long timestamp) {
int count = m_joystickPOVsCache[m_stick].m_count;
if (m_sizedPOVs == null || m_sizedPOVs.length != count) {
m_sizedPOVs = new long[count];
}
for (int i = 0; i < count; i++) {
m_sizedPOVs[i] = m_joystickPOVsCache[m_stick].m_povs[i];
}
m_logPOVs.append(m_sizedPOVs, timestamp);
}
final int m_stick;
boolean[] m_sizedButtons;
float[] m_sizedAxes;
long[] m_sizedPOVs;
final BooleanArrayLogEntry m_logButtons;
final FloatArrayLogEntry m_logAxes;
final IntegerArrayLogEntry m_logPOVs;
}
private static class DataLogSender {
DataLogSender(DataLog log, boolean logJoysticks, long timestamp) {
m_logEnabled = new BooleanLogEntry(log, "DS:enabled", timestamp);
m_logAutonomous = new BooleanLogEntry(log, "DS:autonomous", timestamp);
m_logTest = new BooleanLogEntry(log, "DS:test", timestamp);
m_logEstop = new BooleanLogEntry(log, "DS:estop", timestamp);
// append initial control word values
m_wasEnabled = m_controlWordCache.getEnabled();
m_wasAutonomous = m_controlWordCache.getAutonomous();
m_wasTest = m_controlWordCache.getTest();
m_wasEstop = m_controlWordCache.getEStop();
m_logEnabled.append(m_wasEnabled, timestamp);
m_logAutonomous.append(m_wasAutonomous, timestamp);
m_logTest.append(m_wasTest, timestamp);
m_logEstop.append(m_wasEstop, timestamp);
if (logJoysticks) {
m_joysticks = new JoystickLogSender[kJoystickPorts];
for (int i = 0; i < kJoystickPorts; i++) {
m_joysticks[i] = new JoystickLogSender(log, i, timestamp);
}
} else {
m_joysticks = new JoystickLogSender[0];
}
}
public void send(long timestamp) {
// append control word value changes
boolean enabled = m_controlWordCache.getEnabled();
if (enabled != m_wasEnabled) {
m_logEnabled.append(enabled, timestamp);
}
m_wasEnabled = enabled;
boolean autonomous = m_controlWordCache.getAutonomous();
if (autonomous != m_wasAutonomous) {
m_logAutonomous.append(autonomous, timestamp);
}
m_wasAutonomous = autonomous;
boolean test = m_controlWordCache.getTest();
if (test != m_wasTest) {
m_logTest.append(test, timestamp);
}
m_wasTest = test;
boolean estop = m_controlWordCache.getEStop();
if (estop != m_wasEstop) {
m_logEstop.append(estop, timestamp);
}
m_wasEstop = estop;
// append joystick value changes
for (JoystickLogSender joystick : m_joysticks) {
joystick.send(timestamp);
}
}
boolean m_wasEnabled;
boolean m_wasAutonomous;
boolean m_wasTest;
boolean m_wasEstop;
final BooleanLogEntry m_logEnabled;
final BooleanLogEntry m_logAutonomous;
final BooleanLogEntry m_logTest;
final BooleanLogEntry m_logEstop;
final JoystickLogSender[] m_joysticks;
}
// Joystick User Data
private static HALJoystickAxes[] m_joystickAxes = new HALJoystickAxes[kJoystickPorts];
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
private static HALJoystickAxesRaw[] m_joystickAxesRaw = new HALJoystickAxesRaw[kJoystickPorts];
private static HALJoystickPOVs[] m_joystickPOVs = new HALJoystickPOVs[kJoystickPorts];
private static HALJoystickButtons[] m_joystickButtons = new HALJoystickButtons[kJoystickPorts];
private static MatchInfoData m_matchInfo = new MatchInfoData();
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
private static ControlWord m_controlWord = new ControlWord();
private static EventVector m_refreshEvents = new EventVector();
// Joystick Cached Data
private static HALJoystickAxes[] m_joystickAxesCache = new HALJoystickAxes[kJoystickPorts];
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
private static HALJoystickAxesRaw[] m_joystickAxesRawCache =
new HALJoystickAxesRaw[kJoystickPorts];
private static HALJoystickPOVs[] m_joystickPOVsCache = new HALJoystickPOVs[kJoystickPorts];
private static HALJoystickButtons[] m_joystickButtonsCache =
new HALJoystickButtons[kJoystickPorts];
private static MatchInfoData m_matchInfoCache = new MatchInfoData();
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
private static ControlWord m_controlWordCache = new ControlWord();
// Joystick button rising/falling edge flags
private static int[] m_joystickButtonsPressed = new int[kJoystickPorts];
private static int[] m_joystickButtonsReleased = new int[kJoystickPorts];
// preallocated byte buffer for button count
private static final ByteBuffer m_buttonCountBuffer = ByteBuffer.allocateDirect(1);
private static final MatchDataSender m_matchDataSender;
private static DataLogSender m_dataLogSender;
private static final ReentrantLock m_cacheDataMutex = new ReentrantLock();
private static boolean m_silenceJoystickWarning;
/**
* DriverStation constructor.
*
* <p>The single DriverStation instance is created statically with the instance static member
* variable.
*/
private DriverStation() {}
static {
HAL.initialize(500, 0);
for (int i = 0; i < kJoystickPorts; i++) {
m_joystickButtons[i] = new HALJoystickButtons();
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_joystickAxes[i] = new HALJoystickAxes(DriverStationJNI.kMaxJoystickAxes);
m_joystickAxesRaw[i] = new HALJoystickAxesRaw(DriverStationJNI.kMaxJoystickAxes);
m_joystickPOVs[i] = new HALJoystickPOVs(DriverStationJNI.kMaxJoystickPOVs);
m_joystickButtonsCache[i] = new HALJoystickButtons();
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_joystickAxesCache[i] = new HALJoystickAxes(DriverStationJNI.kMaxJoystickAxes);
m_joystickAxesRawCache[i] = new HALJoystickAxesRaw(DriverStationJNI.kMaxJoystickAxes);
m_joystickPOVsCache[i] = new HALJoystickPOVs(DriverStationJNI.kMaxJoystickPOVs);
}
m_matchDataSender = new MatchDataSender();
}
/**
* Report error to Driver Station. Optionally appends Stack trace to error message.
*
* @param error The error to report.
* @param printTrace If true, append stack trace to error string
*/
public static void reportError(String error, boolean printTrace) {
reportErrorImpl(true, 1, error, printTrace);
}
/**
* Report error to Driver Station. Appends provided stack trace to error message.
*
* @param error The error to report.
* @param stackTrace The stack trace to append
*/
public static void reportError(String error, StackTraceElement[] stackTrace) {
reportErrorImpl(true, 1, error, stackTrace);
}
/**
* Report warning to Driver Station. Optionally appends Stack trace to warning message.
*
* @param warning The warning to report.
* @param printTrace If true, append stack trace to warning string
*/
public static void reportWarning(String warning, boolean printTrace) {
reportErrorImpl(false, 1, warning, printTrace);
}
/**
* Report warning to Driver Station. Appends provided stack trace to warning message.
*
* @param warning The warning to report.
* @param stackTrace The stack trace to append
*/
public static void reportWarning(String warning, StackTraceElement[] stackTrace) {
reportErrorImpl(false, 1, warning, stackTrace);
}
private static void reportErrorImpl(boolean isError, int code, String error, boolean printTrace) {
reportErrorImpl(isError, code, error, printTrace, Thread.currentThread().getStackTrace(), 3);
}
private static void reportErrorImpl(
boolean isError, int code, String error, StackTraceElement[] stackTrace) {
reportErrorImpl(isError, code, error, true, stackTrace, 0);
}
private static void reportErrorImpl(
boolean isError,
int code,
String error,
boolean printTrace,
StackTraceElement[] stackTrace,
int stackTraceFirst) {
String locString;
if (stackTrace.length >= stackTraceFirst + 1) {
locString = stackTrace[stackTraceFirst].toString();
} else {
2016-12-08 00:24:44 -05:00
locString = "";
}
StringBuilder traceString = new StringBuilder();
if (printTrace) {
boolean haveLoc = false;
for (int i = stackTraceFirst; i < stackTrace.length; i++) {
String loc = stackTrace[i].toString();
2018-06-03 10:00:53 -07:00
traceString.append("\tat ").append(loc).append('\n');
// get first user function
if (!haveLoc && !loc.startsWith("edu.wpi.first")) {
locString = loc;
haveLoc = true;
}
}
}
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
DriverStationJNI.sendError(
isError, code, false, error, locString, traceString.toString(), true);
}
/**
* The state of one joystick button. Button indexes begin at 1.
*
* @param stick The joystick to read.
* @param button The button index, beginning at 1.
* @return The state of the joystick button.
*/
public static boolean getStickButton(final int stick, final int button) {
if (stick < 0 || stick >= kJoystickPorts) {
throw new IllegalArgumentException("Joystick index is out of range, should be 0-5");
}
if (button <= 0) {
reportJoystickUnpluggedError("Button indexes begin at 1 in WPILib for C++ and Java\n");
return false;
}
m_cacheDataMutex.lock();
try {
if (button <= m_joystickButtons[stick].m_count) {
return (m_joystickButtons[stick].m_buttons & 1 << (button - 1)) != 0;
}
} finally {
m_cacheDataMutex.unlock();
}
reportJoystickUnpluggedWarning(
"Joystick Button "
+ button
+ " on port "
+ stick
+ " not available, check if controller is plugged in");
return false;
}
/**
* Whether one joystick button was pressed since the last check. Button indexes begin at 1.
*
* @param stick The joystick to read.
* @param button The button index, beginning at 1.
* @return Whether the joystick button was pressed since the last check.
*/
public static boolean getStickButtonPressed(final int stick, final int button) {
if (button <= 0) {
reportJoystickUnpluggedError("Button indexes begin at 1 in WPILib for C++ and Java\n");
return false;
}
if (stick < 0 || stick >= kJoystickPorts) {
throw new IllegalArgumentException("Joystick index is out of range, should be 0-5");
}
m_cacheDataMutex.lock();
try {
if (button <= m_joystickButtons[stick].m_count) {
// If button was pressed, clear flag and return true
if ((m_joystickButtonsPressed[stick] & 1 << (button - 1)) != 0) {
m_joystickButtonsPressed[stick] &= ~(1 << (button - 1));
return true;
} else {
return false;
}
}
} finally {
m_cacheDataMutex.unlock();
}
reportJoystickUnpluggedWarning(
"Joystick Button "
+ button
+ " on port "
+ stick
+ " not available, check if controller is plugged in");
return false;
}
/**
* Whether one joystick button was released since the last check. Button indexes begin at 1.
*
* @param stick The joystick to read.
* @param button The button index, beginning at 1.
* @return Whether the joystick button was released since the last check.
*/
public static boolean getStickButtonReleased(final int stick, final int button) {
if (button <= 0) {
reportJoystickUnpluggedError("Button indexes begin at 1 in WPILib for C++ and Java\n");
return false;
}
if (stick < 0 || stick >= kJoystickPorts) {
throw new IllegalArgumentException("Joystick index is out of range, should be 0-5");
}
m_cacheDataMutex.lock();
try {
if (button <= m_joystickButtons[stick].m_count) {
// If button was released, clear flag and return true
if ((m_joystickButtonsReleased[stick] & 1 << (button - 1)) != 0) {
m_joystickButtonsReleased[stick] &= ~(1 << (button - 1));
return true;
} else {
return false;
}
}
} finally {
m_cacheDataMutex.unlock();
}
reportJoystickUnpluggedWarning(
"Joystick Button "
+ button
+ " on port "
+ stick
+ " not available, check if controller is plugged in");
return false;
}
/**
* Get the value of the axis on a joystick. This depends on the mapping of the joystick connected
* to the specified port.
*
* @param stick The joystick to read.
* @param axis The analog axis value to read from the joystick.
* @return The value of the axis on the joystick.
*/
public static double getStickAxis(int stick, int axis) {
if (stick < 0 || stick >= kJoystickPorts) {
2018-06-03 10:00:53 -07:00
throw new IllegalArgumentException("Joystick index is out of range, should be 0-5");
}
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
if (axis < 0 || axis >= DriverStationJNI.kMaxJoystickAxes) {
2018-06-03 10:00:53 -07:00
throw new IllegalArgumentException("Joystick axis is out of range");
}
m_cacheDataMutex.lock();
try {
if (axis < m_joystickAxes[stick].m_count) {
return m_joystickAxes[stick].m_axes[axis];
}
} finally {
m_cacheDataMutex.unlock();
}
reportJoystickUnpluggedWarning(
"Joystick axis "
+ axis
+ " on port "
+ stick
+ " not available, check if controller is plugged in");
return 0.0;
}
/**
* Get the state of a POV on the joystick.
*
* @param stick The joystick to read.
* @param pov The POV to read.
* @return the angle of the POV in degrees, or -1 if the POV is not pressed.
*/
public static int getStickPOV(int stick, int pov) {
if (stick < 0 || stick >= kJoystickPorts) {
2018-06-03 10:00:53 -07:00
throw new IllegalArgumentException("Joystick index is out of range, should be 0-5");
}
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
if (pov < 0 || pov >= DriverStationJNI.kMaxJoystickPOVs) {
2018-06-03 10:00:53 -07:00
throw new IllegalArgumentException("Joystick POV is out of range");
}
m_cacheDataMutex.lock();
try {
if (pov < m_joystickPOVs[stick].m_count) {
return m_joystickPOVs[stick].m_povs[pov];
}
} finally {
m_cacheDataMutex.unlock();
}
reportJoystickUnpluggedWarning(
"Joystick POV "
+ pov
+ " on port "
+ stick
+ " not available, check if controller is plugged in");
return -1;
}
/**
* The state of the buttons on the joystick.
*
* @param stick The joystick to read.
* @return The state of the buttons on the joystick.
*/
public static int getStickButtons(final int stick) {
if (stick < 0 || stick >= kJoystickPorts) {
throw new IllegalArgumentException("Joystick index is out of range, should be 0-5");
}
m_cacheDataMutex.lock();
try {
return m_joystickButtons[stick].m_buttons;
} finally {
m_cacheDataMutex.unlock();
}
}
/**
* Returns the number of axes on a given joystick port.
*
* @param stick The joystick port number
* @return The number of axes on the indicated joystick
*/
public static int getStickAxisCount(int stick) {
if (stick < 0 || stick >= kJoystickPorts) {
2018-06-03 10:00:53 -07:00
throw new IllegalArgumentException("Joystick index is out of range, should be 0-5");
}
m_cacheDataMutex.lock();
try {
return m_joystickAxes[stick].m_count;
} finally {
m_cacheDataMutex.unlock();
}
}
/**
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
* Returns the number of povs on a given joystick port.
*
* @param stick The joystick port number
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
* @return The number of povs on the indicated joystick
*/
public static int getStickPOVCount(int stick) {
if (stick < 0 || stick >= kJoystickPorts) {
2018-06-03 10:00:53 -07:00
throw new IllegalArgumentException("Joystick index is out of range, should be 0-5");
}
m_cacheDataMutex.lock();
try {
return m_joystickPOVs[stick].m_count;
} finally {
m_cacheDataMutex.unlock();
}
}
/**
* Gets the number of buttons on a joystick.
*
* @param stick The joystick port number
* @return The number of buttons on the indicated joystick
*/
public static int getStickButtonCount(int stick) {
if (stick < 0 || stick >= kJoystickPorts) {
2018-06-03 10:00:53 -07:00
throw new IllegalArgumentException("Joystick index is out of range, should be 0-5");
}
m_cacheDataMutex.lock();
try {
return m_joystickButtons[stick].m_count;
} finally {
m_cacheDataMutex.unlock();
}
}
/**
* Gets the value of isXbox on a joystick.
*
* @param stick The joystick port number
* @return A boolean that returns the value of isXbox
*/
public static boolean getJoystickIsXbox(int stick) {
if (stick < 0 || stick >= kJoystickPorts) {
2018-06-03 10:00:53 -07:00
throw new IllegalArgumentException("Joystick index is out of range, should be 0-5");
}
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
return DriverStationJNI.getJoystickIsXbox((byte) stick) == 1;
}
/**
* Gets the value of type on a joystick.
*
* @param stick The joystick port number
* @return The value of type
*/
public static int getJoystickType(int stick) {
if (stick < 0 || stick >= kJoystickPorts) {
2018-06-03 10:00:53 -07:00
throw new IllegalArgumentException("Joystick index is out of range, should be 0-5");
}
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
return DriverStationJNI.getJoystickType((byte) stick);
}
/**
* Gets the name of the joystick at a port.
*
* @param stick The joystick port number
* @return The value of name
*/
public static String getJoystickName(int stick) {
if (stick < 0 || stick >= kJoystickPorts) {
2018-06-03 10:00:53 -07:00
throw new IllegalArgumentException("Joystick index is out of range, should be 0-5");
}
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
return DriverStationJNI.getJoystickName((byte) stick);
}
2016-08-07 08:19:19 -07:00
/**
* Returns the types of Axes on a given joystick port.
*
* @param stick The joystick port number
* @param axis The target axis
* @return What type of axis the axis is reporting to be
*/
public static int getJoystickAxisType(int stick, int axis) {
2016-08-07 08:19:19 -07:00
if (stick < 0 || stick >= kJoystickPorts) {
2018-06-03 10:00:53 -07:00
throw new IllegalArgumentException("Joystick index is out of range, should be 0-5");
2016-08-07 08:19:19 -07:00
}
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
return DriverStationJNI.getJoystickAxisType((byte) stick, (byte) axis);
2016-08-07 08:19:19 -07:00
}
/**
* Returns if a joystick is connected to the Driver Station.
*
* <p>This makes a best effort guess by looking at the reported number of axis, buttons, and POVs
* attached.
*
* @param stick The joystick port number
* @return true if a joystick is connected
*/
public static boolean isJoystickConnected(int stick) {
return getStickAxisCount(stick) > 0
|| getStickButtonCount(stick) > 0
|| getStickPOVCount(stick) > 0;
}
/**
* Gets a value indicating whether the Driver Station requires the robot to be enabled.
*
* @return True if the robot is enabled, false otherwise.
*/
public static boolean isEnabled() {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_cacheDataMutex.lock();
try {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
return m_controlWord.getEnabled() && m_controlWord.getDSAttached();
} finally {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_cacheDataMutex.unlock();
}
}
/**
* Gets a value indicating whether the Driver Station requires the robot to be disabled.
*
* @return True if the robot should be disabled, false otherwise.
*/
public static boolean isDisabled() {
return !isEnabled();
}
/**
* Gets a value indicating whether the Robot is e-stopped.
*
* @return True if the robot is e-stopped, false otherwise.
*/
public static boolean isEStopped() {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_cacheDataMutex.lock();
try {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
return m_controlWord.getEStop();
} finally {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_cacheDataMutex.unlock();
}
}
/**
* Gets a value indicating whether the Driver Station requires the robot to be running in
* autonomous mode.
*
* @return True if autonomous mode should be enabled, false otherwise.
*/
public static boolean isAutonomous() {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_cacheDataMutex.lock();
try {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
return m_controlWord.getAutonomous();
} finally {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_cacheDataMutex.unlock();
}
}
/**
* Gets a value indicating whether the Driver Station requires the robot to be running in
* autonomous mode and enabled.
*
* @return True if autonomous should be set and the robot should be enabled.
*/
public static boolean isAutonomousEnabled() {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_cacheDataMutex.lock();
try {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
return m_controlWord.getAutonomous() && m_controlWord.getEnabled();
} finally {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_cacheDataMutex.unlock();
}
}
/**
* Gets a value indicating whether the Driver Station requires the robot to be running in
* operator-controlled mode.
*
* @return True if operator-controlled mode should be enabled, false otherwise.
*/
public static boolean isTeleop() {
return !(isAutonomous() || isTest());
}
/**
* Gets a value indicating whether the Driver Station requires the robot to be running in
* operator-controller mode and enabled.
*
* @return True if operator-controlled mode should be set and the robot should be enabled.
*/
public static boolean isTeleopEnabled() {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_cacheDataMutex.lock();
try {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
return !m_controlWord.getAutonomous()
&& !m_controlWord.getTest()
&& m_controlWord.getEnabled();
} finally {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_cacheDataMutex.unlock();
}
}
/**
* Gets a value indicating whether the Driver Station requires the robot to be running in test
* mode.
*
* @return True if test mode should be enabled, false otherwise.
*/
public static boolean isTest() {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_cacheDataMutex.lock();
try {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
return m_controlWord.getTest();
} finally {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_cacheDataMutex.unlock();
}
}
/**
* Gets a value indicating whether the Driver Station is attached.
*
* @return True if Driver Station is attached, false otherwise.
*/
public static boolean isDSAttached() {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_cacheDataMutex.lock();
try {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
return m_controlWord.getDSAttached();
} finally {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_cacheDataMutex.unlock();
}
}
/**
* Gets if the driver station attached to a Field Management System.
*
* @return true if the robot is competing on a field being controlled by a Field Management System
*/
public static boolean isFMSAttached() {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_cacheDataMutex.lock();
try {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
return m_controlWord.getFMSAttached();
} finally {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_cacheDataMutex.unlock();
}
}
/**
* Get the game specific message.
*
* @return the game specific message
*/
public static String getGameSpecificMessage() {
m_cacheDataMutex.lock();
try {
return m_matchInfo.gameSpecificMessage;
} finally {
m_cacheDataMutex.unlock();
}
}
/**
* Get the event name.
*
* @return the event name
*/
public static String getEventName() {
m_cacheDataMutex.lock();
try {
return m_matchInfo.eventName;
} finally {
m_cacheDataMutex.unlock();
}
}
/**
* Get the match type.
*
* @return the match type
*/
public static MatchType getMatchType() {
int matchType;
m_cacheDataMutex.lock();
try {
matchType = m_matchInfo.matchType;
} finally {
m_cacheDataMutex.unlock();
}
switch (matchType) {
case 1:
return MatchType.Practice;
case 2:
return MatchType.Qualification;
case 3:
return MatchType.Elimination;
default:
return MatchType.None;
}
}
/**
* Get the match number.
*
* @return the match number
*/
public static int getMatchNumber() {
m_cacheDataMutex.lock();
try {
return m_matchInfo.matchNumber;
} finally {
m_cacheDataMutex.unlock();
}
}
/**
* Get the replay number.
*
* @return the replay number
*/
public static int getReplayNumber() {
m_cacheDataMutex.lock();
try {
return m_matchInfo.replayNumber;
} finally {
m_cacheDataMutex.unlock();
}
}
/**
* Get the current alliance from the FMS.
*
* @return the current alliance
*/
public static Alliance getAlliance() {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
AllianceStationID allianceStationID = DriverStationJNI.getAllianceStation();
if (allianceStationID == null) {
return Alliance.Invalid;
}
switch (allianceStationID) {
case Red1:
case Red2:
case Red3:
return Alliance.Red;
case Blue1:
case Blue2:
case Blue3:
return Alliance.Blue;
default:
return Alliance.Invalid;
}
}
/**
* Gets the location of the team's driver station controls.
*
* @return the location of the team's driver station controls: 1, 2, or 3
*/
public static int getLocation() {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
AllianceStationID allianceStationID = DriverStationJNI.getAllianceStation();
if (allianceStationID == null) {
return 0;
}
switch (allianceStationID) {
case Red1:
case Blue1:
return 1;
case Red2:
case Blue2:
return 2;
case Blue3:
case Red3:
return 3;
default:
return 0;
}
}
/**
* Return the approximate match time. The FMS does not send an official match time to the robots,
* but does send an approximate match time. The value will count down the time remaining in the
* current period (auto or teleop). Warning: This is not an official time (so it cannot be used to
* dispute ref calls or guarantee that a function will trigger before the match ends) The Practice
* Match function of the DS approximates the behavior seen on the field.
*
* @return Time remaining in current match period (auto or teleop) in seconds
*/
public static double getMatchTime() {
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
return DriverStationJNI.getMatchTime();
}
/**
* Allows the user to specify whether they want joystick connection warnings to be printed to the
* console. This setting is ignored when the FMS is connected -- warnings will always be on in
* that scenario.
*
* @param silence Whether warning messages should be silenced.
*/
public static void silenceJoystickConnectionWarning(boolean silence) {
m_silenceJoystickWarning = silence;
}
/**
* Returns whether joystick connection warnings are silenced. This will always return false when
* connected to the FMS.
*
* @return Whether joystick connection warnings are silenced.
*/
public static boolean isJoystickConnectionWarningSilenced() {
return !isFMSAttached() && m_silenceJoystickWarning;
}
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
/**
* Refresh the passed in control word to contain the current control word cache.
*
* @param word Word to refresh.
*/
public static void refreshControlWordFromCache(ControlWord word) {
m_cacheDataMutex.lock();
try {
word.update(m_controlWord);
} finally {
m_cacheDataMutex.unlock();
}
}
/**
* Copy data from the DS task for the user. If no new data exists, it will just be returned,
* otherwise the data will be copied from the DS polling loop.
*/
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
public static void refreshData() {
DriverStationJNI.refreshDSData();
2022-12-26 14:32:13 -05:00
// Get the status of all the joysticks
for (byte stick = 0; stick < kJoystickPorts; stick++) {
m_joystickAxesCache[stick].m_count =
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
DriverStationJNI.getJoystickAxes(stick, m_joystickAxesCache[stick].m_axes);
m_joystickAxesRawCache[stick].m_count =
DriverStationJNI.getJoystickAxesRaw(stick, m_joystickAxesRawCache[stick].m_axes);
m_joystickPOVsCache[stick].m_count =
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
DriverStationJNI.getJoystickPOVs(stick, m_joystickPOVsCache[stick].m_povs);
m_joystickButtonsCache[stick].m_buttons =
DriverStationJNI.getJoystickButtons(stick, m_buttonCountBuffer);
m_joystickButtonsCache[stick].m_count = m_buttonCountBuffer.get(0);
}
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
DriverStationJNI.getMatchInfo(m_matchInfoCache);
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
DriverStationJNI.getControlWord(m_controlWordCache);
DataLogSender dataLogSender;
// lock joystick mutex to swap cache data
m_cacheDataMutex.lock();
try {
for (int i = 0; i < kJoystickPorts; i++) {
// If buttons weren't pressed and are now, set flags in m_buttonsPressed
m_joystickButtonsPressed[i] |=
~m_joystickButtons[i].m_buttons & m_joystickButtonsCache[i].m_buttons;
// If buttons were pressed and aren't now, set flags in m_buttonsReleased
m_joystickButtonsReleased[i] |=
m_joystickButtons[i].m_buttons & ~m_joystickButtonsCache[i].m_buttons;
}
// move cache to actual data
HALJoystickAxes[] currentAxes = m_joystickAxes;
m_joystickAxes = m_joystickAxesCache;
m_joystickAxesCache = currentAxes;
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
HALJoystickAxesRaw[] currentAxesRaw = m_joystickAxesRaw;
m_joystickAxesRaw = m_joystickAxesRawCache;
m_joystickAxesRawCache = currentAxesRaw;
HALJoystickButtons[] currentButtons = m_joystickButtons;
m_joystickButtons = m_joystickButtonsCache;
m_joystickButtonsCache = currentButtons;
HALJoystickPOVs[] currentPOVs = m_joystickPOVs;
m_joystickPOVs = m_joystickPOVsCache;
m_joystickPOVsCache = currentPOVs;
MatchInfoData currentInfo = m_matchInfo;
m_matchInfo = m_matchInfoCache;
m_matchInfoCache = currentInfo;
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
ControlWord currentWord = m_controlWord;
m_controlWord = m_controlWordCache;
m_controlWordCache = currentWord;
dataLogSender = m_dataLogSender;
} finally {
m_cacheDataMutex.unlock();
}
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
m_refreshEvents.wakeup();
m_matchDataSender.sendMatchData();
if (dataLogSender != null) {
dataLogSender.send(WPIUtilJNI.now());
}
}
[hal, wpilib] New DS thread model and implementation (#3787) The current DS thread model has some pretty major issues. It makes it difficult to know if all data is from the same remote packet, and if the data changes while the robot loop is running. Additionally, the DS thread is used for a few other things (MotorSafety and State Tracking for EducationalRobot). This also makes sim difficult, as user code has to wait for the thread to know it has new data. This change completely rethinks how threading works in the driver station model. First, the DS HAL system receives a new data callback, either from Netcomm or DriverStationSim. Inside the context of this callback, all the low latency data is read and put into a cache. Doing some investigation on the robot side, this is perfectly safe to do, and also ensures a ds packet will not be parsed before we finish reading the current packet data. After all data is read, the cache is swapped with a 2nd buffer. This buffer just stores the data, none of the HAL DS calls read from this buffer. An event is then fired, stating there is new data ready to go. Robot code calls HAL_UpdateDSData(). This swaps the 2nd buffer with a 3rd buffer, which always contains the current data. This data will not be updated until HAL_UpdateDSData is called again. Which solves the state problem. The high level driver station classes have. an updateData() call, which calls HAL_UpdateDSData, and then update button state variables, then data log and update the NT FMS data table (Java also caches across the JNI boundary here, but that could trivially be removed). An extra event provider is provided, allowing other threads to know when this call has been completed. IterativeRobotBase calls DS.updateData() at the beginning of each loop, and only once per loop. This means all commands will always have the same state. All of this means there is no longer a DS thread. Everything happens synchronously. This means Sim and testing is easier, as you can just call DriverStationSim.NotifyNewData(), and then DriverStation.UpdateData(), and you can guarantee that all the DriverStation.*** data is up to date. As for Motor Safety and Educational Robot State Handling, those can all be handled by their own threads. The Educational Thread only needs to run under EducationalRobot, and MotorSafety will only be started if there is a motor safety object enabled.
2022-10-21 22:01:55 -07:00
public static void provideRefreshedDataEventHandle(int handle) {
m_refreshEvents.add(handle);
}
public static void removeRefreshedDataEventHandle(int handle) {
m_refreshEvents.remove(handle);
}
/**
* Reports errors related to unplugged joysticks Throttles the errors so that they don't overwhelm
* the DS.
*/
private static void reportJoystickUnpluggedError(String message) {
double currentTime = Timer.getFPGATimestamp();
if (currentTime > m_nextMessageTime) {
reportError(message, false);
m_nextMessageTime = currentTime + JOYSTICK_UNPLUGGED_MESSAGE_INTERVAL;
}
}
/**
* Reports errors related to unplugged joysticks Throttles the errors so that they don't overwhelm
* the DS.
*/
private static void reportJoystickUnpluggedWarning(String message) {
if (isFMSAttached() || !m_silenceJoystickWarning) {
double currentTime = Timer.getFPGATimestamp();
if (currentTime > m_nextMessageTime) {
reportWarning(message, false);
m_nextMessageTime = currentTime + JOYSTICK_UNPLUGGED_MESSAGE_INTERVAL;
}
}
}
/**
* Starts logging DriverStation data to data log. Repeated calls are ignored.
*
* @param log data log
* @param logJoysticks if true, log joystick data
*/
@SuppressWarnings("PMD.NonThreadSafeSingleton")
public static void startDataLog(DataLog log, boolean logJoysticks) {
m_cacheDataMutex.lock();
try {
if (m_dataLogSender == null) {
m_dataLogSender = new DataLogSender(log, logJoysticks, WPIUtilJNI.now());
}
} finally {
m_cacheDataMutex.unlock();
}
}
/**
* Starts logging DriverStation data to data log, including joystick data. Repeated calls are
* ignored.
*
* @param log data log
*/
public static void startDataLog(DataLog log) {
startDataLog(log, true);
}
}