John,
as I said during the meeting a couple of minutes before, I can confirm
this issue and have made a patch for the OSCI implementation.
Note, that not only sc_fxnum and sc_fxnum_fast are affected, but also
(most of) the other fixed-point classes. The classes in sections
8.10.10 (sc_fxnum) to 8.10.21 (sc_ufixed_fast) of the latest Draft LRM,
that currently only provide constructors and operators for double should
be extended to support float explicitly as well.
This boils down to adding an corresponding entry for 'float' for every
occurrence of 'DECL_.*(op, double)' (operations, assignments,
constructors). So for these DECL_ lines, the attached patches exactly
contain the changes required to the LRM, since the same code is included
there verbatim.
Hope that helps.
Greetings from Oldenburg,
Philipp
-- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean.Received on Wed Oct 27 08:46:57 2010
This archive was generated by hypermail 2.1.8 : Wed Oct 27 2010 - 09:14:11 PDT