Wally4000
d40f4d24d7
Simplify Endian check Code
2022-06-22 23:35:36 +10:00
Carlos R
e213b81291
Fixes for DMA SP and memcpy swizzle
...
[*] Fixed memcpy usage for aligned path for the memcpy swizzle and also enabled it for Linux/OSX
[*] Refactored unaligned path for memcpy swizzle and also fixed memcpy_test
[*] Set correctly sp mem region for DMA_SP
2020-06-10 09:25:14 -07:00
salvy
b2dcc58c62
[!] Fixed bug in memcpy_byteswap when size is less than 4
...
[-] Removed memcpy_byteswap32 (no longer used)
[!] Fetch libraries' headers from third_party folder (W32 build)
2013-06-11 23:08:05 -07:00
Paul Holden
49a06ed319
[~] Make header guards consistently use FULL_PATH_TO_FILE_H_, with script
2013-06-04 22:02:30 +01:00
strmnnrmn
ffcbe649b1
[~] stdafx cleanup: Include stdio.h and stdlib.h where needed
...
git-svn-id: https://subversion.assembla.com/svn/Daedalusx64/trunk@1703 42e9bfbe-799a-4a2d-bad1-236e862a387a
2013-05-19 08:53:29 +00:00
salvy
9ab765a677
[!] Optimizations to accurate tmem
...
git-svn-id: https://subversion.assembla.com/svn/Daedalusx64/trunk@1399 42e9bfbe-799a-4a2d-bad1-236e862a387a
2013-03-24 21:10:55 +00:00
strmnnrmn
6d36f77ca0
[~] Define memcpy_swizzle first - it's a bit tidier
...
git-svn-id: https://subversion.assembla.com/svn/Daedalusx64/trunk@903 42e9bfbe-799a-4a2d-bad1-236e862a387a
2013-01-27 19:31:58 +00:00
strmnnrmn
422823e52e
[~] FastMemcpy.h was a bit broken for big-endian targets. fast_memcpy_swizzle correctly just called memcpy, but memcpy_swizzle ended up incorrectly byteswapping. I've renamed the underlying implementations to _byteswap (which always byteswap, regardless of enianness) and make the _swizzle functions correctly select between memcpy/memcpy_byteswap as appropriate. I think it would probably be clearer to rename memcpy_swizzle to n64_memcpy, or something like that?
...
git-svn-id: https://subversion.assembla.com/svn/Daedalusx64/trunk@902 42e9bfbe-799a-4a2d-bad1-236e862a387a
2013-01-27 19:31:53 +00:00
salvy
5fb82a2cc6
[!] Fixed async audio broken (VFPU and MediaEngine don't mix!)
...
[!] Made fast memcpy truly platform independent
[~] Some more fixes and improvements to memcpy_vfpu_swizzle from last commit
git-svn-id: https://subversion.assembla.com/svn/Daedalusx64/trunk@862 42e9bfbe-799a-4a2d-bad1-236e862a387a
2012-12-13 02:23:32 +00:00
Corn
3b94307914
[!] Fixed normal and swizzled versions of VFPU memcpy
...
[+] Enabled VFPU memcpy in PSPrenderer and DMA
git-svn-id: https://subversion.assembla.com/svn/Daedalusx64/trunk@856 42e9bfbe-799a-4a2d-bad1-236e862a387a
2012-12-08 18:37:03 +00:00
strmnnrmn
3a37f8ac57
Make FastMemcpy a bit more platform independenty and move to the main Utility dir. I removed the timing test and wrapped it up in a function (what it loses from the function call, it gains from being able to inline GetCurrent)
...
git-svn-id: https://subversion.assembla.com/svn/Daedalusx64/trunk@844 42e9bfbe-799a-4a2d-bad1-236e862a387a
2012-12-08 14:19:36 +00:00
