perl Configure.pl --without-gmp --cc=gcc --ccflags='-fno-common -pipe
-I/usr/local/include -pipe -fno-common'
Then chromatic suggested manually editing the Makefile to delete '-
bundle' from the following line.
LD_LOAD_FLAGS = -bundle -undefined suppress
make 2>&1 | tee ~/learn/perl/parrot.make.output.5.txt
make test
The t/library/pcre.t is fixed by the patch in RT #40818.
The problem persists. Tonight I attempt to rebuild Parrot for the first time since my attempts
at the Chicago Perl Hackathon on Nov 11.
I did 'svn update'. And I backed out my own patches to pmc2c.pl.
I called Configure as follows:
perl Configure.pl --without-gmp --cc=gcc --ccflags='-fno-common -pipe -I/usr/local/
include -pipe -fno-common'
See 'myconfig' attached.
I then called 'make'. The first time through, it failed in this section:
make[1]: Leaving directory `/Users/jimk/work/parrot/docs'
src/nci_test.c
c++ -bundle -undefined suppress -L/usr/local/lib -L/Users/jimk/work/parrot/blib/lib -
flat_namespace \
-o runtime/parrot/dynext/libnci_test.bundle src/nci_test.o -lm
c++: couldn't run 'undle-gcc-4.0.3': No such file or directory
make: *** [runtime/parrot/dynext/libnci_test.bundle] Error 1
Note the 'undle-gcc-4.0.3'; where did that initial 'b' on 'bundle' go? This is exactly the same
problem I encountered three weeks ago, and it was solved in exactly the same way that
chromatic suggested at the Hackathon: Edit Makefile to eliminate '-bundle' from the
following line:
LD_LOAD_FLAGS = -bundle -undefined suppress
Once I did that, make purred away to a successful conclusion (see attachment for successful
output from 'make'). I have not yet attempted 'make test' (enough tsuris for one night) --
but can anyone explain why I have to manually hack at the Makefile to get it to 'make'?
kid51
a
--
Andy Bach, Sys. Mangler
Internet: andy...@wiwb.uscourts.gov
VOICE: (608) 261-5738 FAX 264-5932
"Capital is only the fruit of labor. Labor is the superior of capital and
deserves much the higher consideration."
Abraham Lincoln, first annual address to Congress 1861
> James Keenan via RT wrote:
> > On Sat Nov 11 10:17:33 2006, jk...@verizon.net wrote:
> >> perl Configure.pl --without-gmp --cc=gcc --ccflags='-fno-common -pipe
> >> -I/usr/local/include -pipe -fno-common'
> >>
> >> Then chromatic suggested manually editing the Makefile to delete '-
> >> bundle' from the following line.
> >> LD_LOAD_FLAGS = -bundle -undefined suppress
>
> It appears (from google) that "-bundle" is a MacOS specific option to
> their version of gcc's c++ (also Intel's MacOS c++) but, I'm guessing,
> your version of c++, which is then treating the -bundle as -b undle or
> somesuch. So, perhaps the MacOS guessing code needs to poke a tad
> harder at the versions of gcc/c++ its getting its hands on.
What's weirder to me is that Configure.pl picks up g++ as the linker while
using cc as the compiler. I couldn't figure that out at the hackathon;
something's definitely weird there.
I think we opened a ticket asking for an override to specify "Link with gcc
instead of g++", but I don't know the status of that.
-- c
A 'diff' of the 'make test' summaries suggests that 37 more tests are now passing. Test
failures are now confined to t/pmc/nci.t and t/pmc/resizablepmcarray.t. The Nov 11 2006
run is first.
< t/codingstd/tabs.t 1 256 1 1 1
< t/codingstd/trailing_space.t 1 256 1 1 1
< t/compilers/imcc/syn/pcc.t 1 256 21 1 11
< t/dynpmc/subclass_with_pir_method.t 1 256 2 1 2
< t/examples/library.t 1 256 4 1 3
< t/examples/shootout.t 2 512 20 2 14-15
< t/library/pcre.t 1 256 1 1 1
< t/library/pg.t 0 6 43 0 ??
< t/op/sprintf.t 308 27 32-33 37-43 194-196
< 216-222 225-232
< t/pmc/capture.t 1 256 9 1 8
< t/pmc/nci.t 59 15104 60 59 1-56 58-60
< t/pmc/object-meths.t 1 256 34 1 32
< t/pmc/parrotinterpreter.t 1 256 3 1 3
< 11 tests and 602 subtests skipped.
< Failed 13/262 test scripts. 97/6672 subtests failed.
< Files=262, Tests=6672, 1635 wallclock secs (560.06 cusr + 307.48 csys = 867.54 CPU)
< Failed 13/262 test programs. 97/6672 subtests failed.
< make: *** [test] Error 1
---
> t/pmc/nci.t 59 15104 60 59 1-56 58-60
> t/pmc/resizablepmcarray.t 1 256 38 1 38
> 11 tests and 604 subtests skipped.
> Failed 2/264 test scripts. 60/6701 subtests failed.
> Files=264, Tests=6701, 1241 wallclock secs (521.75 cusr + 289.33 csys = 811.08 CPU)
> Failed 2/264 test programs. 60/6701 subtests failed.
> make: *** [test] Error 255
> What's weirder to me is that Configure.pl picks up g++ as the linker while
> using cc as the compiler. I couldn't figure that out at the hackathon;
> something's definitely weird there.
ICU.
(ICU was mandatory, and ICU is C++, so using g++ as the linker means that
all the C++ runtime libraries come in correctly)
Nicholas Clark