2 Aug 2005 21:35
[patch][commit] Floating Point Support for SID
Dave Brolley <brolley <at> redhat.com>
2005-08-02 19:35:14 GMT
2005-08-02 19:35:14 GMT
Hi, I've committed this patch which makes a small change to the generated code for accurate floating point ops. The change is to not dereference CGEN_CPU_FPU (current_cpu)->ops->XXXXX. This makes no difference for cgen simulators generated for sim/common, but does allow the implementation of 'ops' to be a C++ class with virtual member functions for each operation for SID. The SID patch which makes use of this will follow shortly. Dave
2005-08-02 Dave Brolley <brolley <at> redhat.com> * rtl-c.scm (s-unop): Don't dereference CGEN_CPU_FPU (current_cpu)->ops->xxxxx in the generated code. (s-binop, s-convop, s-cmpop): Likewise.
Index: cgen/rtl-c.scm =================================================================== RCS file: /cvs/src/src/cgen/rtl-c.scm,v retrieving revision 1.6 diff -c -p -r1.6 rtl-c.scm *** cgen/rtl-c.scm 15 Jun 2005 21:28:18 -0000 1.6 --- cgen/rtl-c.scm 2 Aug 2005 19:25:38 -0000 ***************(Continue reading)
RSS Feed