Commit Graph

577 Commits

Author SHA1 Message Date
UnknownShadow200
5b5e6d3ee6 Fix alloc_count_continuous calculation 2025-02-03 06:53:16 +11:00
Luke Benstead
acee4b5c37 Merge branch 'quicker_transform' into 'master'
Quicker vertex transform in non-fast path

See merge request simulant/GLdc!144
2025-02-02 08:34:36 +00:00
Luke Benstead
9100660c78 Merge branch 'attribute_speedup' into 'master'
Move attribute parsing into separate file and avoid recomputing parse/read functions unless necessary

See merge request simulant/GLdc!143
2025-02-02 08:32:06 +00:00
UnknownShadow200
a566bba082 Call TransformVertex directly instead of using TransformVertices 2025-02-02 15:58:19 +11:00
UnknownShadow200
3b2e549934 Optimise vertex transform in non-fast path to avoid storing xyz to memory and then loading it again 2025-02-02 14:36:40 +11:00
UnknownShadow200
9d717800bd Reorganise attribute parsing functions 2025-02-02 13:51:30 +11:00
Luke Benstead
ea4073fbdc Merge branch 'error_coldpath' into 'master'
Move error handling code into the cold path

See merge request simulant/GLdc!142
2025-02-01 11:10:52 +00:00
UnknownShadow200
e2c3ff0ed6 Fix not building for desktop 2025-02-01 20:12:23 +11:00
UnknownShadow200
e2fc2da055 Move error handling code into the cold path 2025-02-01 19:58:22 +11:00
UnknownShadow200
5318f11f11 Defer recomputing attributes state when possible 2025-02-01 18:17:28 +11:00
UnknownShadow200
3550c89501 Make non-textured non-fast drawing a little bit faster 2025-02-01 17:04:42 +11:00
UnknownShadow200
a35659ce3d Fix enable/disable attribute calls not updating read functions 2025-02-01 16:05:07 +11:00
UnknownShadow200
bdecf90d64 Calculate read attribute functions when attribute states changes, instead of at every draw call 2025-02-01 15:38:59 +11:00
UnknownShadow200
3b4b1b6975 Fix compile error 2025-02-01 14:00:49 +11:00
UnknownShadow200
71d0094ac4 Prefer relying on built in codegen rather than relying on sh4_math.h 2025-02-01 13:06:37 +11:00
Falco Girgis
dc0c8148cb API FP ABI conformity fixes. 2025-01-27 00:53:23 -06:00
Falco Girgis
9653bf8c12 Got rid of a memset4() call from KOS. 2025-01-23 21:27:34 -06:00
Luke Benstead
b4c2dd0a5e Merge branch 'micro_opts' into 'master'
Few micro optimisations

See merge request simulant/GLdc!133
2024-12-22 22:31:47 +00:00
Luke Benstead
aa383913cd Merge branch 'nonPOT2glTexSubImage2D' into 'master'
update for glTexSubImage2D, now checks the active texture size and not the...

See merge request simulant/GLdc!136
2024-12-07 23:27:52 +00:00
GPF
f7830df480 removed the commented out debug printing 2024-12-07 16:24:05 -07:00
Luke Benstead
55a912b621 Fix continuous free count 2024-12-07 19:54:16 +00:00
Luke Benstead
c4786a62e6 Fix bug in allocations (thanks @skmp) 2024-12-07 19:35:40 +00:00
GPF
553f6016f3 Only initialize the texture buffer with zeros if it's a partial update 2024-10-15 12:11:17 -07:00
GPF
b44dddbaff removed some unneeded code 2024-10-15 12:04:17 -07:00
GPF
4adf11543b fixed the case where xoffset and yoffset !=0 in glTexSubImage2D 2024-10-15 11:25:56 -07:00
GPF
a11847e513 update for glTexSubImage2D, now checks the active texture size and not the subimage size to draw the glTexSubImage2D pixelas into drawing with the xoffset and yoffset into the existing POT2 texture, so subimage size doesn't need to be a POT2. 2024-10-14 13:26:17 -07:00
GPF
72136bb7f5 quick implementation of glTexSubImage2D/_glTexSubImage2DValidate, update to nehe06 example to show usage of glTexSubImage2D, and a compile issue to primitive_modes fixed. 2024-10-13 12:58:09 -07:00
Falco Girgis
6407353121 Removed opb_count hack for older KOS versions 2024-10-09 22:20:32 -05:00
UnknownShadow200
490bf3de6d Unify and slightly simplify specialised matrix multiply functions 2024-10-07 18:06:42 +11:00
UnknownShadow200
91b59ceca8 glXYZPointer methods now avoid recomputing state when only the data pointer changes
It is relatively common to call glXYZPointer multiple times with a different pointer but the same type/size
2024-10-07 18:06:03 +11:00
Luke Benstead
31676d9dc8 Merge branch 'sq_fix' into 'master'
Use direct render APIs instead of manually locking/unlocking store queues

See merge request simulant/GLdc!129
2024-10-04 14:29:12 +00:00
Luke Benstead
98577ebb31 Merge branch 'twid_speedup' into 'master'
Only check whether twiddle table needs updating once instead of every pixel

See merge request simulant/GLdc!130
2024-10-04 14:28:16 +00:00
UnknownShadow200
6f0acd5218 Only check whether twiddle table needs updating once instead of every pixel
Based on checking disassembly output in Release mode
2024-10-04 22:11:42 +10:00
UnknownShadow200
ea56577ca9 Use direct render APIs instead of manually locking/unlocking store queues
This ensures that GLdc works in both 2.1.0 and latest matter
2024-10-04 19:51:33 +10:00
UnknownShadow200
c30b16b97b Fix points/lines being improperly culled 2024-10-03 08:08:33 +10:00
UnknownShadow200
2b06418fed Fix not compiling on desktop due to relying on KOS specific frsqrt 2024-10-02 20:27:28 +10:00
UnknownShadow200
4513b040e2 Don't inline less optimal primitive modes 2024-10-02 20:07:36 +10:00
UnknownShadow200
165eb9d5c8 Add primitive modes sample, quad strips partially work
Although the second triangle is wrongly culled
2024-10-02 19:28:02 +10:00
UnknownShadow200
19a76cbefd Fix lines 2024-10-02 19:00:05 +10:00
UnknownShadow200
a47f331cee Initial work on supporting missing primitive modes 2024-10-02 07:48:44 +10:00
UnknownShadow200
d978e3187c Fix array drawing mode not rendering properly in non-fast mode.
Also move vertex transform to be done inline in non-fast path, instead of all entirely at the end
2024-10-01 07:13:02 +10:00
Luke Benstead
75b992f52e Fix multiple bugs in the allocator 2024-09-14 07:51:13 +01:00
Luke Benstead
42786f775e Fix a bug in the allocator# 2024-09-14 07:46:11 +01:00
Luke Benstead
dc07cbc4d2 Merge branch 'blend2' into 'master'
Try to fix not being able to use GL_SRC_COLOR and GL_ONE_MINUS_SRC_COLOR as destination blend factor

See merge request simulant/GLdc!125
2024-09-06 13:26:55 +00:00
UnknownShadow200
d570030070 Try to fix not being able to use GL_SRC_COLOR and GL_ONE_MINUS_SRC_COLOR as destination blend factor 2024-09-06 22:13:31 +10:00
UnknownShadow200
5f28cae804 Tidy up textures table and add some missing formats 2024-09-04 21:26:46 +10:00
Luke Benstead
ab24e2db35 Make GL_MULTISAMPLE not throw an error 2024-08-28 22:39:05 +01:00
Falco Girgis
b6995edc66 Merge branch 'master' of https://gitlab.com/simulant/GLdc 2024-08-11 09:19:59 -05:00
Unknown Shadow
d737961e88 Merge branch 'master' into 'FasterImmediate'
# Conflicts:
#   GL/immediate.c
2024-08-11 11:11:25 +00:00
UnknownShadow200
92ec1db643 Only update attribute pointers at end instead of per vertex 2024-08-11 19:39:27 +10:00