From: Anton K. <an...@sw...> - 2011-03-13 18:06:13
|
Merinov Nikolay <kim...@gm...> writes: > Anton Kovalenko <an...@sw...> writes: > >> Major part of my effort during the last week or two had gone into 64-bit >> MS Windows port of SBCL for AMD64 architecture. I'm glad to announce >> that Windows AMD64 support is now fully functional. X86-64 binaries >> (MSI installer and an additional standalone executable) are available at >> [ http://www.siftsoft.com/inprogress/forknews.html ]. > > This port really can work? On my computer I get this > http://pastebin.com/raw.php?i=sQ7r5PuL result with > http://www.siftsoft.com/dl/sbcl-1.0.46.32.263.wth.kovalenko-x86-64.msi > binaries. Of course it can work, or how would I get the "warm" core? GCC supports SystemV ABI calling convention, even on Windows, and I use it for some functions (call_into_lisp, pthread_getspecific,...) -- that's why those places need not be rewritten. Please pay attention to attribute(sysv_abi) and AMD64_SYSV_ABI macro in the declarations. As of "long[]" arrays in bignum.h and vector.h -- theoretically, they should be replaced with uword_t[] (or u64), but those "arrays" are (almost?) never accessed from C code -- and slot accessors generated for Lisp code are correct. Don't know why it doesn't work on your system (trying lower --dynamic-space-size might help; ensuring there is no stale SBCL installation which sets SBCL_HOME may be useful too). -- Regards, Anton Kovalenko +7(916)345-34-02 | Elektrostal' MO, Russia |