This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug other/10109] documentation of target macro LIBCALL_VALUE
- From: "a dot cilio at et dot tudelft dot nl" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: 1 Aug 2003 15:44:54 -0000
- Subject: [Bug other/10109] documentation of target macro LIBCALL_VALUE
- References: <20030316145601.10109.a.cilio@et.tudelft.nl>
- Reply-to: gcc-bugzilla at gcc dot gnu dot org
PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=10109
------- Additional Comments From a dot cilio at et dot tudelft dot nl 2003-08-01 15:44 -------
Unlike stated in previous comment, I think that aggregate function return values
are a concern for library support functions. Aggregate return values occur in
library support functions that return a double-precision floating-point value
when the target architecture does not have a register class for double precision
floating-point values. In such situations, the macros that describe the target
architecture can specify that the doubles are passed/retuned as aggregates.
This appears to work correctly for normal compiled C functions.