Segger J-Link - Tracing with On-Chip Trace Buffer

To Next Page

To Previous Page

379

14.3 Tracing with on-chip trace buffer

Some target CPUs provide trace functionality also provide an on-chip trace buffer

that is used to store the trace data output by the trace hardware unit on the device.

This allows to also do trace on such targets with a regular J-Link, as the on-chip trace

buffer can be read out via the regular debug interface J-Link uses to communicate

with the target CPU. Downside of this implementation is that it needs RAM on the tar-

get CPU that can be used as a trace buffer. This trace buffer is very limited (usually

between 1 and 4 KB) and reduces the RAM that can be used by the target application,

while tracing is done.

Note: Streaming trace is not possible with this trace implementation

14.3.1 CPUs that provide tracing via pins and on-chip buffer

Some CPUs provide a choice to either use the on-chip trace buffer for tracing (e.g.

when the trace pins are needed as GPIOs etc. or are not available on all packages of

the device).

•

For J-Link: The on-chip trace buffer is automatically used, as this is the only

method J-Link supports.

•

For J-Trace: By default, tracing via trace pins is used. If, for some reason, the

on-chip trace buffer shall be used instead, the J-Link software needs to be made

aware of this. The trace source can be selected via the

SelectTraceSource com-

mand string. For more information about the syntax this command string, please

refer to

Command strings on page 223. For more information about how to use

command strings in different environments, please refer to

Using command

strings

on page 241.

Other manuals for Segger J-Link