On 02/22/2012 03:57 PM, Bob Friesenhahn wrote: > On Wed, 22 Feb 2012, Stefano Lattarini wrote: > >> Hi Bob, sorry for the delay. >> >> On 02/19/2012 07:55 PM, Bob Friesenhahn wrote: >>> I am again bit by automake not being able run the test suite on systems >>> with bounded command line length. Up to automake 1.11.2 I was able to >>> apply a patch by Ralf Wildenhues. >>> >> Which patch are you referring to exactly? > > This one: > http://lists.gnu.org/archive/html/automake-patches/2010-09/msg00068.html > I don't understand how that patch could actually work ... If there are too many tests in $(TESTS), there will be too many logs in $(TEST_LOGS), and since the recipe for $(TEST_SUITE_LOG) contents the expansion of $(TEST_LOGS), the command line length limit for "/bin/sh -c" will be exceeded anyway ...
In fact, it was Ralf who explained that to me, when I tried to "resurrect" its proposed patch (which he had already reverted for an unrelated regression); see my proposal: <http://debbugs.gnu.org/cgi/bugreport.cgi?bug=7868#14> and Ralf's explanation for why this couldn't work: <http://debbugs.gnu.org/cgi/bugreport.cgi?bug=7868#17> >>> With 1.11.3 the implementation is totally different and the patch >>> has no hope of applying. >>> I can rework it for the code in Automake 1.12 (soonish to appear) if you are still interested in it after having read the rationale above. >> How does this later patch of mine fares? >> >> <http://debbugs.gnu.org/cgi/bugreport.cgi?bug=7868#31> > > I will check when I have time. > >> Caveat: it works for GNU make only, and it's *really ugly*. >> So ugly that I don't feel like applying it to automake proper. > > A GNU make dependency would be a real problem. A dependency on a > properly-working standard make would not be such a problem. > Is the test suite actually failing due to "exceeded command line length" on any system different from Cygwin or MinGW? If not, I don't see how a GNU make dependency would be a problem (if not for the fact that it uglifies the already hideous patch even more). >> "Again" is not the correct word here -- automake, either patched or not, >> has *never* been able to run arbitrarily large test suites. Still, recent >> changes might have unintentionally reduced the maximal "testsuite size" >> automake is able to handle, and we'd want to fix that. How "big" is the >> testsuite you are referring to concretely? > > I have not run into any problems when using Ralf's patch, but it apparently > does not address all issues so Ralf backed it out. > In fact, it fundamentally fails to address the issues -- that's why Ralf backed it out. > My test suite is pretty slim at the moment. About 610 tests (used to be > about 1200) which expand to a list of tests of at least 17903 bytes when > builddir == srcdir, or 36813 bytes with one of my typical build > configurations with builddir != srcdir. > Are you referring to the GraphicsMagick testsuite? If yes, it seems to me that it could benefit greatly from the use of TAP once automake 1.12 is out; for example, all the 314 'tests/rwfile*.sh' tests could be rewritten as one single TAP test (and similarly for many other auto-generated tests), thus removing the need for a huge number of files without reducing granularity of the results. HTH, Stefano