This is the mail archive of the crossgcc@sources.redhat.com mailing list for the crossgcc project.

See the CrossGCC FAQ for lots more information.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: crosstool status


Robert P. J. Day wrote:
one more suggestion (as if you needed any more).  from the recent
traffic on this list, it seems that there are still some known bugs in
crosstool, which is to be expected, no problem with that. if that's
the case, though, i would make the argument that what is warranted is
yet another *release candidate*.  0.28-rc40, anyone? :-)

philosophically (at least in my experience), putting a regular version
number on something (0.29? 1.0?) means that it has passed at least
some level of Q/A testing, it appears to perform nominally and there
are no obvious flaws or show-stoppers in the code that is *already
there*.  these kinds of releases don't have to be fully-featured --
there could still be "coming soon" features, but the salient point is
that what's there works fairly reliably.

There are, and will always be, known problems in crosstool, just as there are in each new version of gcc and glibc. It's the nature of the beast.

That said, I *do* put new releases
through a QA procedure: they have to be able to build toolchains
and Linux kernels for "most" combinations of tools.  I include
the matrix of build results along with the release.  That way,
you can tell which combinations are likely to build for you.

I do *not* put crosstool through any sort of runtime QA procedure
on the target platforms.  I haven't got all the hardware needed
to do that, nor do I have the time.
Instead, I include detailed instructions on how you can run
your own QA using the gcc and glibc regression test suites,
and I encourage people who run the test suites to post the
results to the crossgcc mailing list.  Not many people do,
since it's a bit hard, but I imagine more might do it in the
future, especially now that targets are starting to get
enough RAM to actually run the testsuite scripts on the
target.  One of these days the doc needs to be enlarged
to cover how to run the tests that way; the current method
(using remote execution) is just hard enough to set up that
it scares people off.
- Dan

--
Trying to get a job as a c++ developer?  See http://kegel.com/academy/getting-hired.html

------
Want more information?  See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/
Want to unsubscribe? Send a note to crossgcc-unsubscribe@sources.redhat.com


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]