Age | Commit message (Collapse) | Author |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Makes easy to test without actually doing the regress in make build.
|
|
|
|
|
|
|
|
|
|
|
|
Thanks to pval@ for resolving the license stuff.
|
|
|
|
|
|
|
|
with respect to the input.
Additional testcase, based on libcurses, that failed in the past.
Activate t5, t6. Check cycles output everywhere.
|
|
|
|
|
|
Based on a test case found by Dragos Ruiu (dr@kyx.net)
|
|
|
|
|
|
|
|
|
|
|
|
Add VAR:sh=cmd check
|
|
|
|
parser gets fixed.
|
|
|
|
and the canonical output looks much better with a blank line removed.
Note that the last test needs large amounts of memory to finish (especially
the perl part).
|
|
|
|
|
|
|
|
|
|
39: sort /usr/share/dict/words (presorted)
40: sort large files
|
|
|
|
|
|
|
|
|
|
|
|
the test files to be executed actually have the execute bit set on them
since CTM doesn't preserve file permissions.
|
|
|
|
|
|
|
|
|
|
|
|
depending on .PHONY and such
|
|
|
|
|