
On-chip Trace
The MCDS module of the TriCore device generates information on the instruction execution and data accesses of up to six cores in parallel as well as on transfers on the on-chip
buses and interconnects. This information is stored into the onchip trace buffer.
- Up to 2 MB of on-chip trace memory
- Trigger and filter programming for an efficient use of the trace buffer
- Configurable on-chip memory allows a concurrent usage of the on-chip trace with third-party usage if necessary, e.g. calibration tools or the user application
- Fully supported by TRACE32 Debug Cable, approriate trace licenses required
|
|
Off-chip Trace
If the information generated about program execution and data access by cores/buses is conveyed to a serial trace port, it can be recorded by a TRACE32 Trace Tool. This
allows detailed qualification and analysis with its up to 8 GB trace buffer and includes the possibility of streaming to the host storage.
|
- Support of AGBT, SGBT
- Robust trace recording using Aurora
- Support of up to four differential lanes
- Maximum 6,25 Gbit/s lane speed
- Up to 8 G Byte trace memory
- Streaming to host storage
|
 | |
|
- 4 GByte trace memory
- Universal module that is prelicensed for a trace protocol on delivery, additional trace protocol licenses can be added
- Aurora-based trace protocols up to 8 RX lanes, up to 12.5 Gbit/s per lane
- Speed booster TRACE32 Preprocessor Aurora 2 delivers up to 22.5 Gbit/s per lane for up to 4 lanes
- PCIe 3.0-based trace protocol up to 8 RX/TX lanes, up to 8 Gbit/s per lane
- Maximum bandwidth of 100 GBit/s
- Reference-clock and bit-clock support
- Fast trace upload to the host computer
- Support for TRACE32 Streaming up to 400 MByte/s
- TRACE32 Standard Probe/Analog Probe option
|
 | |
DAP Streaming
DAP Streaming means that the contents of the onchip trace memory is streamed to the TRACE32 CombiProbe while the program execution is running.
As a medium bandwidth trace, the CombiProbe is the matching tool for AUTOSAR-aware profiling, Compact Function Trace and filtered data tracing.
|
- Debug cable and 512 MByte of trace memory
- Whisker cable with 26-pin automotive connector
- Same debug features as the TRACE32 Debug Cables for TriCore
- 512 MB of trace memory for DAP Streaming
- DAP Streaming with up to 30 MByte/s
- DAP streaming requires a DAP mode, for performance reasons, we recommend DAPWide at 160 MHz
- DAP Streaming for filtered tracing, GTM tracing and Compact Function Trace
- AUTOSAR-aware profiling
- TRACE32 Streaming for long-time recording
|
 | |
Trace Message Generation
The MCDS module of the TriCore device generates information on the instruction execution and data accesses of up to six cores in parallel as well as on transfers on the on-chip
buses and interconnects.
- Unfiltered and filtered trace messages
- Tracing of up to 6 TriCores, PPU (ARC), GTM and PCP (device dependent):
- Program flow or sync trace (generating trace messages on branches or on every MCDS clock cylce)
- Data trace
- Ownership information, e.g. PCP channel ID
- Support for tracing the processor buses such as LMB, SRI
- Support for tracing the peripheral buses such as SPB, RPB
- SRI trace allows tracing of data transfers to up to two SRI bus slaves in parallel
|
|
Trace Through Reset
Microcontroller applications are often secured by an internal and/or external watchdog or internal surveillance. In case of any unrecoverable error these systems reset the microcontroller in order to bring it back into a
known and safe state. When such a reset occurs during development, the engineers want to know why this error happens, and often how post-error handling after reboot behaves. If trace recording was enabled the information of
what happened prior to the reset is contained in the trace and can be displayed even after the reset. If the debugger is programmed to resume program execution the trace will continue recording (not possible for on-chip trace
in case of a hard reset, e.g. PORST).
Complex Trigger Language CTL
Complex test scenarios require precisely targeted triggers and filters. These can be configured intuitively with the CTL trigger language.

The screenshot above shows the following use case: According to the specification the interrupt stmlr1T0 is only allowed to interrupt the function
workSeries twice. The CTL trigger program now ensures that the program execution is stopped immediately as soon as this specification is violated.
|
- Allows re-debuggging of a traced program section
- Provides forward and backward debugging capabilities
- High-level language trace display including all local variables
- Timing and function nesting display
- Has the ability to fill most trace gaps caused by the limited bandwidth of trace port
|
 | |
|
- Detailed analysis of function run-times
- Detailed analysis of task run-times and state
- Graphical analysis of variable values over the time
- Analysis of the time interval of a single event (e.g. Interrupt)
- Analysis of the time interval between 2 defined events
|
 | |
|
- Provides all metrics for functional safety
- For standard trace protocols TRACE32 Code Coverage gets by with no or very little instrumentation, full instrumentation as fallback
- Suitable for long-term testing
- Automated report generation in multiple exchange formats
- TRACE32 Trace-Based Code Coverage is included in the scope of delivery of all TRACE32 Debug & Trace Tools at no additional cost
|
 | |
|