Less strict memory ordering and bugfixes (#23)

Author: RainerZ

xcplib/README.md

@@ -1,5 +1,5 @@
 
-# XCPlite V1.x.x
+# XCPlite V1.0.0
 
 XCPlite is a lightweight pure C implementation of the ASAM XCP V1.4 standard protocol for measurement and calibration of electronic control units.  
 It supports XCP on TCP or UDP with jumboframes.  
@@ -19,17 +19,27 @@ Checkout the Vector XCP Book:
 Visit the Virtual VectorAcedemy for an E-Learning on XCP:  
 <https://elearning.vector.com/>  
 
+## Whats new in V1.0.0
+
+- Breaking changes to V6.  
+- Lockless transmit queue. Works on x86-64 strong and ARM-64 weak memory model.  
+- Measurement of and write access to variables on stack.  
+- Supports multiple calibration segments with working and reference page with independent page switching
+- Lock free and thread safe calibration parameter access, consistent calibration changes and page switches.  
+- Build as a library.  
+- Used (as FFI library) for the rust xcp-lite version.  
+
 ## Features
 
 - Supports XCP on TCP or UDP with jumbo frames.  
-- Thread safe, minimal thread lock and single copy event driven, timestamped high performance consistent data acquisition.  
+- Thread safe, minimal thread lock and single copy event driven, timestamped high performance and consistent data acquisition.  
 - Runtime A2L database file generation and upload.  
+- Prepared for PTP synchronized timestamps.  
+- Supports calibration and measurement of structures
 - User friendly code instrumentation to create calibration parameter segments, measurement variables and A2L metadata descriptions.  
 - Measurement of global or local stack variables.  
 - Thread safe, lock-free and wait-free ECU access to calibration data.  
 - Calibration page switching and consistent calibration.  
-- Can be build as a library.  
-- Used (as FFI library) in the rust xcp-lite version.  
 
 A list of restrictions compared to Vectors free XCPbasic or commercial XCPprof may be found in the source file xcpLite.c.  
 XCPbasic is an optimized implementation for smaller Microcontrollers and with CAN as Transport-Layer.

xcplib/src/xcpEthTl.c

@@ -569,10 +569,10 @@ void XcpEthTlGetInfo(bool *isTcp, uint8_t *mac, uint8_t *addr, uint16_t *port) {
 int32_t XcpTlHandleTransmitQueue(void) {
 
     // Simply polling transmit queue
-    // @@@@ TODO Optimize efficiency, use a convdar or something like that to wakeup/sleep the transmit thread
+    // @@@@ TODO Optimize efficiency, use a condvar or something like that to wakeup/sleep the transmit thread
     // @@@@ TODO Optimize the mutex
     // This is needed to assure XCP transport layer header counter consistency among response and DAQ packets
-    // In fact this is a XCP design flaw, CANape supports independand DAQ and response packet counters, but other tools don't
+    // In fact this is a XCP design flaw, CANape supports independent DAQ and response packet counters, but other tools don't
 
     // Timeout to give the caller a chance to do other heath checking or shutdown the server gracefully
     const uint32_t max_outer_loops = 100; // Number of outer loops before return
@@ -581,8 +581,8 @@ int32_t XcpTlHandleTransmitQueue(void) {
     // Sleep time in ms after burst or queue empty
     const uint32_t outer_loop_sleep_ms = 1; // Sleep time in ms for each outer loop
 
-    int32_t n = 0;      // Numer of bytes sent
-    bool flush = false; // Flush queue in regular intervalls
+    int32_t n = 0;      // Number of bytes sent
+    bool flush = false; // Flush queue in regular intervals
 
     for (uint32_t j = 0; j < max_outer_loops; j++) {
         for (uint32_t i = 0; i < max_inner_loops; i++) {