diff options
| author | Jonathan Wakely <jwakely@redhat.com> | 2024-10-24 11:06:42 +0100 |
|---|---|---|
| committer | Jonathan Wakely <redi@gcc.gnu.org> | 2024-10-30 12:17:10 +0000 |
| commit | 30563d8978a39c402a425e958b5b51936ee76cff (patch) | |
| tree | ce65f34ee082fa8e72742fd7ee317f46c89658c2 /gcc | |
| parent | bc0eeccf27a084461a2d5661e23468350acb43da (diff) | |
| download | gcc-30563d8978a39c402a425e958b5b51936ee76cff.zip gcc-30563d8978a39c402a425e958b5b51936ee76cff.tar.gz gcc-30563d8978a39c402a425e958b5b51936ee76cff.tar.bz2 | |
libstdc++: Define __memcpyable<float*, _Float32*> as true
This allows optimizing copying ranges of floating-point types when they
have the same size and representation, e.g. between _Float32 and float
when we know that float uses the same IEEE binary32 format as _Float32.
On some targets double and long double both use IEEE binary64 format so
we could enable memcpy between those types, but we don't have existing
macros to check for that case.
libstdc++-v3/ChangeLog:
* include/bits/cpp_type_traits.h (__memcpyable): Add
specializations for compatible floating-point types.
Reviewed-by: Patrick Palka <ppalka@redhat.com>
Diffstat (limited to 'gcc')
0 files changed, 0 insertions, 0 deletions