diff options
author | Alexandre Ratchov <ratchov@cvs.openbsd.org> | 2021-05-25 08:06:13 +0000 |
---|---|---|
committer | Alexandre Ratchov <ratchov@cvs.openbsd.org> | 2021-05-25 08:06:13 +0000 |
commit | cdc615e735f7d4f02ed5313c889b11295d1bc49a (patch) | |
tree | 699e19ed2de42ab8fab201509ef882100b446763 /sys/net | |
parent | 8804ad53f6cd2b8679c69d750448ecec534e7a43 (diff) |
Drop assembly for 24-bit fixed-point operations
Now i386 compiler is smart enough to generate a single imul
instruction per fixed-point multiplication. This change also allows
24-bit precision to be used on all archs (if -DADATA_BITS=24 is used).
Tested on arm64 by Doug Moss <dougmoss710 at yahoo.com>, thanks
Suggested and ok naddy@
Diffstat (limited to 'sys/net')
0 files changed, 0 insertions, 0 deletions