[torquedev] Build Problems
garrick at usc.edu
Fri Jul 23 18:39:19 MDT 2010
On Fri, Jul 23, 2010 at 05:27:11PM -0700, Michael Jennings alleged:
> I think David re-committed buildutils/ltmain.sh by accident. This is
> an autogenerated file, and the build will break if the libtool pulled
> in by autoreconf doesn't match the ltmain.sh that's there.
> Also, src/include/pbs_config.h.in is autogenerated and should be
> removed as well.
Fixed and fixed.
> And I discovered that I needed to unset my CFLAGS variable in order to
> run "make distcheck" due to -Werror being the default. Unless you
> specify exactly which warnings you turn on, this can result in build
> failures caused by someone (like me) employing -W flags that the
> developers don't (which I do). In the specific case of "make
> distcheck," the only way around this is to edit the Makefile.am to add
> --disable-gcc-warnings to the DISTCHECK_CONFIGURE_FLAGS or specifying
> it as part of the make distcheck command line. I assume this is being
> done to catch more errors, but it may be less confusing to users if
> the developers themselves select a set of warning flags they care
> about and then limit -Werror usage to their own testing and
> certification builds. Just a thought...
I think the use of -Wall -Werror has been very good for us. It catches little
errors slip by because what is only a warning on one platform is actually an
error on another platform.
I care about nearly all warnings. And I care about future types of warnings in
future compiler releases. I specifically exempt a few warnings (-Wno-long-long)
for ones that I don't care about it, but I want everything else watched.
So... what warning did you see? Have a patch yet?
Garrick Staples, GNU/Linux HPCC SysAdmin
University of Southern California
Life is Good!
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 189 bytes
Desc: not available
Url : http://www.supercluster.org/pipermail/torquedev/attachments/20100723/5cdabe84/attachment.bin
More information about the torquedev