This directory contains scripts that are used by the regression
|
This directory contains scripts that are used by the regression
|
tester,
|
tester,
|
|
|
The primary script is 'btest-gcc.sh'. This is the script that is run
|
The primary script is 'btest-gcc.sh'. This is the script that is run
|
to actually test the compiler.
|
to actually test the compiler.
|
|
|
'objs-gcc.sh' takes a combined tree and builds (but does not test) the
|
'objs-gcc.sh' takes a combined tree and builds (but does not test) the
|
tools required for 'btest-gcc.sh'. It is run periodically to update
|
tools required for 'btest-gcc.sh'. It is run periodically to update
|
the tools. This script is followed by running 'btest-gcc.sh' using
|
the tools. This script is followed by running 'btest-gcc.sh' using
|
the newly-build tools to check that they will not cause regressions.
|
the newly-build tools to check that they will not cause regressions.
|
|
|
'site.exp' is what $DEJAGNU points to when the regression tester runs
|
'site.exp' is what $DEJAGNU points to when the regression tester runs
|
these scripts.
|
these scripts.
|
|
|
'GCC_Regression_Tester.wdgt' is a Dashboard widget that displays the
|
'GCC_Regression_Tester.wdgt' is a Dashboard widget that displays the
|
current state of the tester using Javascript. You can use it without
|
current state of the tester using Javascript. You can use it without
|
needing Dashboard by pointing your web browser at
|
needing Dashboard by pointing your web browser at
|
'GCC_Regression_Tester.wdgt/widget.html', if your browser supports
|
'GCC_Regression_Tester.wdgt/widget.html', if your browser supports
|
and permits it.
|
and permits it.
|
|
|
Note that any changes made here need to be approved by the regression
|
Note that any changes made here need to be approved by the regression
|
tester's maintainer (see MAINTAINERS). The changes will be used on
|
tester's maintainer (see MAINTAINERS). The changes will be used on
|
the tester's next run, so `experimental' changes are very strongly
|
the tester's next run, so `experimental' changes are very strongly
|
discouraged :-).
|
discouraged :-).
|
|
|
