'make check' on Win 2k

[Sisyphus]

Hi,


On Windows 2000, GMP-4.1.2, ran:
./configure --disable-static --enable-shared

Runs fine, and 'make' also runs fine.

But 'make check' builds 't-bswap.exe', 't-constants.exe', 
t-count_zeros.exe', 't-gmpmax.exe' and 't-modlinv.exe' in the 
'tests/.libs' folder. Consequently these 5 tests fail, because the 
executables cannot be found:

FAIL: t-bswap.exe
/bin/sh: ./t-bswap.exe: No such file or directory
etc.

After the failure of these 5 tests 'make check'exits.
(The same thing happened with GMP-4.1, btw.)

Tried building with:
1) MSYS-1.0.8 and MinGW-2.0.0-3
2) MSYS-1.0.10-rc-3 and MinGW-3.1.0-1

Makes no difference whichever I use.

Ok ... so I copy the 5 files from the 'tests/.libs' folder up one level 
and re-run 'make check'. The 5 tests then pass and 'make check' proceeds 
to build the 5 mpn test executables, again in the wrong directory 
('tests/mpn/.libs'). Once again these tests fail because the executables 
cannot be found and 'make check' exits.

So ... I copy the mpn test executables up one level and re-run 'make 
check'. The 5 mpn tests pass (along, of course, with the other 5 tests) 
and make check proceeds to build the mpz test executables in the wrong 
place ('tests/mpz/.libs'). The tests once again fail because the 
executables cannot be found and 'make check' dies again.

I copy the 52 executables up one level and re-run 'make check'.
This time one of the tests fail and 'make check' refuses to proceed 
beyond these checks.

The test that fails is 't-fib_ui.exe':

abnormal program termination
t-fib_ui.c:63: GNU MP assertion failed: (__gmp_fib_table[(i)+1]) != 0
FAIL: t-fib_ui.exe

The other 51 tests pass.

Cheers,
Rob
--
Any emails containing attachments will be deleted from my ISP's mail 
server before I even get to see them. If you wish to email me an 
attachment, please provide advance warning so that I can make the 
necessary arrangements.