I have a pipeline job that parses warnings for several builds, using the same warnings parser. For example, building 32 and 64 bit cmake with gcc, so parsing output for 32 and 64 bit build independently. This results into 2 (in my case more) trend graphs to be visible in the Job view. This happens on the same build node in the same pipeline stage in parallel executions. The warnings trend graphs now mix results of different warnings(...) calls.

What is totally unclear to me, and I have not found any documentation (while searching for an hour) how warnings calls are assigned to a trend graph, i.e. which call appends the results to which graph? Can somebody clarify how this is done? By node? by stage? or by time order? or something else?

Another visual problem is, that the trend graph shows the name of the warnings parser only. For example "Doxygen Warnings Trend". When using the same parser multiple times, you see the same name multiple times and it is impossible to know what the graph is really related to. It would be nice to be able to give the trend graph a unique name during the warnings step. Then it would be possible to uniquely assign parsed warnings to the right graph and a user would know what the graph is about.