<div class="gmail_quote">I have uploaded further fixes. It turns out that the changes in GMP were indeed just hiding the following problems:</div><div class="gmail_quote"><br></div><div class="gmail_quote">* msvc/ecl/config.h.msvc6 defined ECL_LONG_BITS to be 64 instead of 32, which is what should be with Windows64's model.</div>
<div class="gmail_quote"><br></div><div class="gmail_quote">* The routine that adds a fixnum and a bignum used GMP's *_ui/*_si function, which is incorrect (the fixnums do not fit in a long).</div><div class="gmail_quote">
<br></div><div class="gmail_quote">* A coercion error in ecl_to_int32_t(). For some reason the compiler was confused about an integer constant.</div><div class="gmail_quote"><br></div><div class="gmail_quote">I just build both 32 and 64 bit versions of ECL and ran the tests. In Win64 not everything runs fine, but we seem to be much closer.</div>
<div class="gmail_quote"><br></div><div class="gmail_quote">Juanjo</div><div><br></div>-- <br>Instituto de Física Fundamental, CSIC<br>c/ Serrano, 113b, Madrid 28006 (Spain) <br><a href="http://juanjose.garciaripoll.googlepages.com" target="_blank">http://juanjose.garciaripoll.googlepages.com</a><br>