zdoom/gzdoom
0
0
Fork 0
mirror of https://github.com/ZDoom/gzdoom.git synced 2024-11-19 02:42:05 +00:00
Code Issues Projects Releases Packages Wiki Activity
9934 commits 131 branches 149 tags 200 MiB
Commit graph

128 commits

Author SHA1 Message Date
Magnus Norddahl
2676ca1284 Rename wall drawer args and functions to closer match the general naming convention 2016-12-19 21:39:16 +01:00
Magnus Norddahl
9ac7a07be6 - Set colormap light info using R_SetColorMapLight and R_SetDSColorMapLight rather than manually calculating it 
- Move texture and span management into R_DrawMaskedColumn
 2016-12-18 11:58:32 +01:00
Christoph Oelckers
356e55f0a2 - made the general entry point to the wall drawer a single function instead of requiring the caller to check for NPOT textures. 
- renamed all functions in r_walldraw.cpp to give them names more in line with Doom's naming conventions. Since this is not Build code anymore it also shouldn't use Build names to avoid giving a false impression.
 2016-12-09 14:25:51 +01:00
Christoph Oelckers
eab06ef086 - more Build cleanup: Moved clearbufshort out of Build derived headers, removed the inline version for GCC 32 bit which was actually Build derived and renamed the C-version of this function, which does not originate from Build to 'fillshort'. 
- cleaned out a lot the SafeDivScale stuff in m_fixed.h. The only SafeDivScale variant still in use was #16 for FixedDiv, so all the SafeDivScale stuff has been removed and the 16 variant renamed to FixedDiv because that's the only form in which it is still being used. (2x in R_DrawVoxel and 1x in ACS's FixedDiv PCode.)
- removed Build notice from m_fixed.h because aside from the inlines includes there is nothing here from Build anymore.
 2016-12-09 13:04:53 +01:00
Magnus Norddahl
3ff91807b8 Move swrenderer into a namespace, add multithreading framework, and move drawers to commands 2016-12-07 09:34:49 +01:00
Edoardo Prezioso
46a311b23c - Fixed a typo in 1 pixel tall support addition. 
Found by Clang -Wparentheses warning.
 2016-11-01 09:59:59 +01:00
Marisa Heit
9388597443 Add 1 pixel tall and wide texture support to the renderer 
- These require manual detection and overriding of the scaling factors to
  0, because a right shift of (32-0) bits wraps around to 0 and results in
  no shift at all rather than leaving the register zeroed out.
 2016-11-01 00:08:16 -05:00
Braden Obrzut
43b2584f79 - Fixed: Reference to freed stack object in R_FindPlane. This caused a massive slowdown (90% drop in total performance) in R_FindPlane when built with optimizations on in GCC6. Although I don't really understand why since the comparison should have been O(1) regardless of memory contents and even if the check failed every plane it would still be pretty fast, this is what they mean when they say that anything can happen when undefined behavior is triggered. 2016-10-30 23:27:29 -04:00
Christoph Oelckers
19d070c9bd - fixed: averageColor swapped red and blue. 
- renamed some stuff
 2016-10-20 09:08:07 +02:00
Magnus Norddahl
2fe545a4fd Merge r_stretchsky and r_capsky into r_skymode 2016-10-20 00:59:51 +02:00
Magnus Norddahl
5de8112578 Add support for capping sky with a solid color 2016-10-19 23:52:09 +02:00
Magnus Norddahl
657abb374b Fix SkyViewpoint skyboxes not being rendered 2016-10-14 23:01:10 +02:00
Christoph Oelckers
d4352dd1a7 - fixed slope calculation error. 2016-07-17 23:13:10 +02:00
Randy Heit
09730bff73 Fix incorrect texture mapping vectors for slopes 2016-04-29 20:35:55 -05:00
Randy Heit
8f3921190f Keep values in floats longer for visplane drawing 2016-04-28 21:21:49 -05:00
Christoph Oelckers
9f0c5d5909 -let's better be cautious about precision and use the CRT sin and cos functions for rotated plane textures. 2016-04-28 14:49:59 +02:00
Randy Heit
ae7d048057 Store FTransform in visplane_t, not just a pointer 
- The transform values passed to R_CheckPlane might live on the stack, so
  it's not safe to only store a pointer to them.
 2016-04-27 23:07:25 -05:00
Randy Heit
44adff459a Remove halfviewwidth 
- Why I thought I needed this variable in addition to centerx has
  long ago left my memory.
 2016-04-27 22:49:36 -05:00
Randy Heit
31d842a894 Store FTransform in visplanes instead of converting to fixed_t 2016-04-27 22:17:33 -05:00
Christoph Oelckers
8bda356412 - fixed crash in slope drawing: Several overlooked occurences of fixed point math in R_CalcTiltedLighting. Slopes are still completely broken, though. 2016-04-27 04:08:06 +02:00
Christoph Oelckers
0f99d7d73f - floatified the few remaining linedef and sector properties that still were fixed_t. The only fixed point things left in r_defs.h are the nodes and a fixed point setter function for vertex_t, which is still needed by the node builder. 2016-04-24 12:15:09 +02:00
Christoph Oelckers
04e614daec - removed some fixed point functions that are no longer needed. 
- fixed: Visplane checks should only compare the plane flags that are relevant for rendering and mask out the rest.
- floatified FTransform and made the visplane checks a bit less verbose by moving the comparison as an operator into FTransform.

Note that this operator needs forceinline on Visual Studio so that it won't get called as a function.
 2016-04-23 12:42:07 +02:00
Christoph Oelckers
35a6994d0a - remove the 'F' prefix from the floating point access function to the sidedef texture info, now that the fixed point versions are gone. 2016-04-23 10:55:55 +02:00
Christoph Oelckers
4d4f31fd83 - floatified sidedef texture info and fixed two bugs involving this data. 2016-04-23 09:41:59 +02:00
Randy Heit
f1b3f59bcc Renderer floatification: Use floating point for wall scale and yrepeat 2016-04-22 23:12:53 -05:00
Randy Heit
3ef5a678d5 Renderer floatification: Use floating point for texturemids and Y scales 2016-04-22 23:12:53 -05:00
Randy Heit
ec8d038c99 Renderer floatification: Use floats for lighting values 2016-04-22 23:10:34 -05:00
Randy Heit
4416d88eb2 Renderer floatification: Use floating point for visibility values 2016-04-22 23:10:02 -05:00
Randy Heit
ded3f92452 Commence renderer floatification 2016-04-22 23:09:13 -05:00
Christoph Oelckers
082042818b - refactored sector portal data so that it does not rely on actors. 
This is necessary because otherwise the level data cannot be serialized before the actors.
 2016-04-20 19:20:11 +02:00
Christoph Oelckers
d88a5ac353 - make distancecheck 3D and use the actual view position for calculation. 2016-04-13 10:48:39 +02:00
Christoph Oelckers
db86385cf6 - removed STACK_ARGS. 
The only reason this even existed was that ZDoom's original VC projects used __fastcall. The CMake generated project do not, they stick to __cdecl.
Since no performance gain can be seen by using __fastcall the best course of action is to just remove all traces of it from the source and forget that it ever existed.
 2016-04-11 10:46:30 +02:00
Christoph Oelckers
fc5f98a0be - floatified the sector planes. 
This should conclude the floating point conversions for now.
 2016-04-03 19:28:53 +02:00
Christoph Oelckers
8680391e0b - floatified the global view variables. 
The software renderer still uses internal copies of the fixed point version.
 2016-04-01 12:22:16 +02:00
Christoph Oelckers
a5e14425aa - fixed angular interpolation of skybox viewpoints. 2016-04-01 11:55:37 +02:00
Christoph Oelckers
8fd76f0c8a - floatified bmaporgx and bmaporgy, allowing to remove the gross overflow prevention hacks present in the blockmap code. 2016-03-31 09:23:14 +02:00
Christoph Oelckers
66929cbaff - floatified p_trace, p_slopes and p_udmf.cpp. 
- major cleanup of unused code.
 2016-03-30 16:30:22 +02:00
Christoph Oelckers
25f5e8449a - replaced all direct access to sector plane coefficients with wrapper functions. 2016-03-29 12:40:41 +02:00
Christoph Oelckers
0c39bdd04c - floatified texture scale values. 2016-03-26 13:37:44 +01:00
Christoph Oelckers
fb8e03d5eb - floatified FLineOpening. 
- some smaller fixes.
 2016-03-25 18:43:37 +01:00
Christoph Oelckers
1125101b37 - floatified AActor::Prev plus the stuff using it. 
- rewrote some coordinate functions in AActor to use real float math instead of converting back and forth between float and fixed.
 2016-03-25 16:25:25 +01:00
Christoph Oelckers
3a598d672e - removed the angle_t type from the VM, now that nothing in the interface requires it anymore. 2016-03-25 14:55:01 +01:00
Christoph Oelckers
4e60ea0252 - made AActor::alpha a floating point value 
- replaced some uses of FRACUNIT with OPAQUE when it was about translucency.
- simplified some overly complicated translucency multiplications in the SBARINFO code.
 2016-03-21 12:18:46 +01:00
Christoph Oelckers
51b05d331d - replaced AActor::vel and player_t::Vel with a floating point version. 
- Converted P_MovePlayer and all associated variables to floating point because this wasn't working well with a mixture between float and fixed.

Like the angle commit this has just been patched up to compile, the bulk of work is yet to be done.
 2016-03-20 00:54:18 +01:00
Christoph Oelckers
671291227e - first stage of converting actor angles to float complete 
Patched up everything so that it compiles without errors again. This only addresses code related to some compile error. A large portion of the angle code still uses angle_t and converts back and forth.
 2016-03-16 12:41:26 +01:00
Christoph Oelckers
e11a0986ce - encountered a strange crash that after changing a map, InSubsector was not NULL and pointing to invalid data. So let's better NULL this variable explicitly each time a render loop is started. 2016-03-08 12:27:01 +01:00
Randy Heit
55142078d8 Normalize line endings 2016-03-01 09:47:10 -06:00
Christoph Oelckers
02d7572343 - some header dependency cleanup so that it is no longer needed to include portal.h to get the inline functions. Portal.h has been reduced of most dependencies now so that including it is cheap and can be done in other headers. 
- some consolidation in p_map.cpp. PIT_CheckLine and PIT_FindFloorCeiling had quite a bit of redundancy which has been merged.
- čontinued work on FMultiBlockLinesIterator. It's still not completely finished.
 2016-02-19 14:08:41 +01:00
Christoph Oelckers
158caf78a0 - more float to double conversion. 
In particular this removes all assignments of FIXED2FLOAT to double variables because they not only lose precision but also generate unnecessary code.
 2016-02-11 21:33:30 +01:00
Christoph Oelckers
6d0ef7a9da - added conversion macros to convert floating point angles to angle_t, using xs_Float.h, and replaced all occurences in the code with them (let's hope I found everything.) 
Converting a floating point value that is out of range for a signed integer will result in 0x80000000 with SSE math, which is used exclusively for this purpose on modern Visual C++ compilers, so this cannot be used anywhere.
On ARM there's problems with float to unsigned int conversions.

xs_Float does not depend on these
 2016-02-08 12:10:53 +01:00