Opened 11 years ago

Closed 5 years ago

#655 closed defect (wontfix)

Geos make check failure windows XP 32-bit gcc 4.8.0

Reported by: robe Owned by: geos-devel@…
Priority: major Milestone: GEOS Fund Me
Component: Default Version: 3.4.0
Severity: Unassigned Keywords:
Cc:

Description

I have to retry this on my windows 7 64-bit and 32-bit GCC 4.8.0 mingw64. For those I just use CMake and don't run into these issues.

This fails on make check compile

In file included from ../include/geos/geom/Coordinate.h:158:0,
                 from ../include/geos/geom/Envelope.h:25,
                 from ../include/geos/geom/Geometry.h:27,
                 from geos_ts_c.cpp:20:
../include/geos/geom/Coordinate.inl: In static member function 'static int geos:
:geom::Coordinate::hashCode(double)':
../include/geos/geom/Coordinate.inl:103:21: warning: right shift count >= width
of type [enabled by default]
  return (int)(f^(f>>32));
                     ^
In file included from ../../include/geos/geom/Coordinate.h:158:0,
                 from ../../include/geos/geom/Envelope.h:25,
                 from ../../include/geos/geom/Geometry.h:27,
                 from ../../include/geos/geom/GeometryFactory.h:23,
                 from bug234.cpp:3:
../../include/geos/geom/Coordinate.inl: In static member function 'static int ge
os::geom::Coordinate::hashCode(double)':
../../include/geos/geom/Coordinate.inl:103:21: warning: right shift count >= wid
th of type [enabled by default]
  return (int)(f^(f>>32));
                     ^
In file included from ../../include/geos/geom/Coordinate.h:158:0,
                 from ../../include/geos/geom/Envelope.h:25,
                 from ../../include/geos/geom/Geometry.h:27,
                 from ../../include/geos/geom/GeometryFactory.h:23,
                 from TestSweepLineSpeed.cpp:24:
../../include/geos/geom/Coordinate.inl: In static member function 'static int geos::geom::Coordinate::hashCode(double)':
../../include/geos/geom/Coordinate.inl:103:21: warning: right shift count >= width of type [enabled by default]
  return (int)(f^(f>>32));
                     ^
In file included from ../../include/geos/geom/Coordinate.h:158:0,
                 from ../../include/geos/geom/Envelope.h:25,
                 from ../../include/geos/geom/Geometry.h:27,
                 from ../../include/geos/geom/GeometryFactory.h:23,
                 from GeometryTestFactory.cpp:17:
../../include/geos/geom/Coordinate.inl: In static member function 'static int geos::geom::Coordinate::hashCode(double)':
../../include/geos/geom/Coordinate.inl:103:21: warning: right shift count >= width of type [enabled by default]
  return (int)(f^(f>>32));
                     ^
EXCEPTION on case 1 test 2: IllegalArgumentException: RobustDeterminant encounte
red non-finite numbers
EXCEPTION on case 1 test 1: IllegalArgumentException: RobustDeterminant encounte
red non-finite numbers
EXCEPTION on case 1 test 3: IllegalArgumentException: RobustDeterminant encountered non-finite numbers
make[4]: *** [check-TESTS] Error 1
make[3]: *** [check-am] Error 2
make[2]: *** [check-recursive] Error 1
make[1]: *** [check-recursive] Error 1
make: *** [check] Error 2

This may be a dupe all all those NaN errors people have been talking about

This was run on r3897

Change History (6)

comment:1 by strk, 11 years ago

Didn't ./configure find a 64-bit integer ?

comment:2 by robe, 11 years ago

Just tried on my window 7 64-bit (gcc 4.8.0) but using the same chain as I had on windows xp. Have that one turned off. It gives different error, but the 64-bit is not being picked up

Here is from config.log

configure:17207: result: no
configure:17225: checking whether int64_t is 64 bits
configure:17285: i686-w64-mingw32-gcc -o conftest.exe -g -O2   conftest.c  >&5
conftest.c:35:1: error: unknown type name 'int64_t'
 typedef int64_t ac_int64;
 ^
conftest.c: In function 'main':
conftest.c:60:5: warning: incompatible implicit declaration of built-in function 'exit' [enabled by default]
     exit(0);
     ^
configure:17285: $? = 1

I think it might have been that mingw64 patch at #568 that the other Sandro sent which seems to work fine under my gcc 4.5.4 install. I'll see if I can fix this myself in such a fashion to make both happy welcome for ideas though.

The reason I didn't notice this on my gcc 4.8.0 chain is that I use CMake for gcc 4.8.0 (postgresql 9.3) and have never bothered using libtool for it. On cmake it passes with flying colors.

comment:3 by robe, 11 years ago

I meant to say compiling my 32-bit mingw64-w2 (gcc 4.8.0) on my windows 7 64-bit box.

gcc 4.5.4 mingw64-w32 just gives those 3 errors I already complained about.

comment:4 by strk, 9 years ago

robe can this be closed ?

comment:5 by robe, 6 years ago

Milestone: GEOS FutureGEOS Fund Me

Milestone renamed

comment:6 by robe, 5 years ago

Resolution: wontfix
Status: newclosed
Note: See TracTickets for help on using tickets.