Opened 14 years ago
Last modified 7 years ago
#29336 assigned defect
Maude: universal variant fails
Reported by: | ryandesign (Ryan Carsten Schmidt) | Owned by: | |
---|---|---|---|
Priority: | Normal | Milestone: | |
Component: | ports | Version: | 1.9.2 |
Keywords: | universal | Cc: | |
Port: | Maude |
Description
Maude does not build when using the universal variant:
Making all in Utility ccache /usr/bin/g++-4.2 -DHAVE_CONFIG_H -I. -I. -I../.. -I/opt/local/include -pipe -O2 -arch x86_64 -arch i386 -c -o macros.o macros.cc macros.cc:111: error: integer constant is too large for 'long' type macros.cc:115: error: integer constant is too large for 'long' type macros.cc:117: error: integer constant is too large for 'long' type macros.cc: In function 'Int64 stringToInt64(const char*, bool&, int)': macros.cc:117: warning: overflow in implicit constant conversion macros.cc: In function 'void correctEcvt(double, int, char*, int&, int&)': macros.cc:258: warning: large integer implicitly truncated to unsigned type macros.cc:259: warning: large integer implicitly truncated to unsigned type macros.cc:266: warning: right shift count >= width of type macros.cc:268: warning: right shift count >= width of type lipo: can't open input file: /var/tmp//cccFBzK7.out (No such file or directory) make[3]: *** [macros.o] Error 1 make[2]: *** [all-recursive] Error 1 make[1]: *** [all-recursive] Error 1 make: *** [all] Error 2
Attachments (2)
Change History (5)
Changed 14 years ago by ryandesign (Ryan Carsten Schmidt)
Changed 14 years ago by dcolish@…
Attachment: | fix-universal.diff added |
---|
comment:1 Changed 14 years ago by dcolish@…
comment:2 Changed 14 years ago by ryandesign (Ryan Carsten Schmidt)
I'm skeptical about this proposed fix. It seems to me that the developers of maude want to make decisions in the build process based on whether we're building 32-bit or 64-bit, and that they make that determination at configure time and as a result define various constants in config.h that are then referenced in the source at build time. Instead, the determination should be made at build time, using existing preprocessor defines like __LP64__
.
comment:3 Changed 7 years ago by mf2k (Frank Schima)
Owner: | dcolish@… deleted |
---|---|
Status: | new → assigned |
See #54089.
Note: See
TracTickets for help on using
tickets.
I'm not 100% sure that diff is the best way to fix the build issue, but it does appear to do the trick.