8ab9960ca0
lightpreview: don't crash when vis is run more than once
2023-12-24 17:22:01 -07:00
6351cc50e6
rename: vec_t -> double
2023-12-23 21:09:16 -07:00
b45871efdb
vis algorithm improvement (2h42 -> 57m on ad_dm5) ( #361 )
...
* vis algorithm improvement, new commandline option -targetchecks
* vis: only calculate num_expected_targetchecks if needed
Fixes a bit of performance regression when -targetchecks was added, when running
with -targetchecks 0, apparently due to the cost of all the std::popcount's.
gmsp3v2.bsp, 32 threads:
18s (brushbsp branch 9388d43b1c
2023-12-02 13:41:56 -07:00
5bcb2d4657
vis: remove MAX_PORTALS_ON_LEAF
...
reduces memory use a bit
gmsp3v2.bsp - 51.96s elapsed
2023-11-11 13:10:01 -07:00
37fc62ca0e
vis: remove dead code from header
2023-11-11 12:29:30 -07:00
f1bf0f546d
add missing #pragma once's
2023-11-08 21:30:44 -07:00
c9edf4ddaa
vis: huge speedup from fixing false sharing with stat counters
...
gmsp3v2.bsp, 32 threads 3950x, 157s -> 61s
2023-11-08 00:05:08 -07:00
e9532346c7
vis: partly fix perf regression of Oct 2021
...
- drops usage of winding_storage_hybrid
- makes pstack_t trivially default initializable again (like the C
original)
- we weren't making use of the unlimited size of winding_storage_hybrid;
the original vis already has an allowance for if a winding exceeds
the fixed stack allocation size (24) and we were still applying that
limit
before:
cmake-build-release\tests\tests.exe -r=xml -sf=*benchmark.cc
Testing started at 1:53 PM ...
| ns/op | op/s | err% | total | benchmark
|--------------------:|--------------------:|--------:|----------:|:----------
| 45.15 | 22,149,002.06 | 0.1% | 0.01 | `create pstack_t`
| 47.04 | 21,256,429.80 | 0.1% | 0.01 | `create pstack_t + 1x AllocStackWinding`
| 48.28 | 20,710,594.32 | 0.2% | 0.01 | `create pstack_t + 2x AllocStackWinding`
| 128.52 | 7,780,603.30 | 0.3% | 0.01 | `setup + ClipStackWinding`
Benchmark from: https://www.quaddicted.com/tools/vis_benchmark
cmake-build-release\vis\vis.exe -nostate -nopercent gmsp3v2.bsp
153.31s elapsed
after:
cmake-build-release\tests\tests.exe -r=xml -sf=*benchmark.cc
Testing started at 1:46 PM ...
| ns/op | op/s | err% | total | benchmark
|--------------------:|--------------------:|--------:|----------:|:----------
| 1.64 | 610,528,630.06 | 0.3% | 0.01 | `create pstack_t`
| 3.29 | 303,895,535.45 | 0.6% | 0.01 | `create pstack_t + 1x AllocStackWinding`
| 5.20 | 192,172,148.84 | 0.2% | 0.01 | `create pstack_t + 2x AllocStackWinding`
| 66.84 | 14,960,818.66 | 1.0% | 0.01 | `setup + ClipStackWinding`
Benchmark from: https://www.quaddicted.com/tools/vis_benchmark
cmake-build-release\vis\vis.exe -nostate -nopercent gmsp3v2.bsp
141.42s elapsed
2023-11-05 13:58:36 -07:00
e069d7b9b8
style: update settings api to follow current coding style
2023-04-16 11:55:38 -06:00
b17633f746
cleanup: set PackConstructorInitializers: Never
2023-02-20 10:23:59 -07:00
eda6155844
cleanup: run ./clang-format-all.sh
...
clang-format version 15.0.7
2023-02-18 22:59:43 -07:00
8f79f6a972
vis: rename portal_t -> visportal_t
...
fixes symbol collision causing crashes on macOS
2022-12-21 01:03:03 -07:00
331dbfc20d
vis: refactor output_group and advanced_group to avoid name clash with light
2022-12-08 21:32:39 -07:00
2ff9eda020
vis: refactor winding_t -> viswinding_t
2022-12-08 21:29:22 -07:00
642c4b589c
tests: add vis_main() taking std::vector
2022-12-08 21:27:37 -07:00
35ffc00462
all: add LIGHT_/QBSP_/VIS_ prefixes to epsilons where necessary
2022-10-30 15:08:10 -06:00
636b376577
squash brushbsp-compilation-speedup branch
...
move function implementations out of headers
2022-10-21 18:00:58 -06:00
06a116670b
winding_t: make non-copyable
2022-08-12 02:06:23 -06:00
53adb9fed5
split winding_base_t into two types, so that we can switch out the backing storage
...
default winding_t for the main polylib to use heap-based, exact-memory-usage-only, winding
2022-08-02 06:09:20 -04:00
1d8a0c1df3
clang-format pass
2022-07-28 06:14:54 -04:00
b7969eee49
vis: rename options -> vis_options
2022-07-10 17:26:55 -06:00
1718512b09
remove unnecessary "uncompressed_q2" - can just be a single buffer now
...
another ptr->vector
2022-06-23 17:08:50 -04:00
3ac1d02624
fix missing noexcept
...
explicit initialization in some cases for vis winding_t's parameters
more vector
2022-06-23 17:02:18 -04:00
ee596f3ff1
use vector for vis state
...
set autoclean default to true, make it invertable
2022-06-23 16:18:52 -04:00
6f6185e1a3
use unique_ptr for leafbits allocations
2022-06-22 16:04:55 -04:00
3fbdfa2a36
allow removing the vis file if completed
2022-06-21 20:46:46 -04:00
b54e72a184
re-implement the vis changes
2022-06-18 18:42:00 -04:00
b21920841d
revert some of the vis stuff, but re-implement it in a more working state
2022-06-18 18:36:09 -04:00
b0dad9d01d
save a few more cycles by just using raw pointers. it's a bit uglier, but this is hard to represent with shared_ptr and I think it was the wrong choice anyways since the ownership chain is difficult to represent in a shared_ptr here.
2022-06-17 14:56:17 -04:00
c3bb07acaa
~33% speed improvement on vis by using already-allocated memory for the vis stack
2022-06-17 13:53:35 -04:00
f6309a91cb
vis: move .prt reader to common
2022-06-16 21:55:03 -06:00
812797f7ed
enable lighting for warp/sky in Q2
...
simplify bitflag definitions
change nudging light warning to make more sense
another #define -> constexpr pass
2022-06-10 06:29:14 -04:00
0f2ea5333d
move CompressRow/DecompressRow from vis to common
2022-05-24 21:01:41 -06:00
f6ba5d77f9
allow recompilation of PHS from PVS data via -phsonly
2022-05-21 02:45:30 -04:00
8fe5d9d694
common: adjust formatting of usage line
...
-remove version hash as it's printed above
-rename `usage` variable to `programDescription`
2022-05-15 17:44:17 -06:00
49c78292dd
testvis: stub setup
2022-05-15 15:52:54 -06:00
0c27c92e41
vis: fix 32-bit shift warning
...
"result of 32-bit shift implicitly converted to 64 bits (was 64-bit shift intended?)"
2022-05-12 00:37:22 -06:00
abc799554d
build: fix all clang-cl warnings
2022-05-10 21:57:17 -06:00
d763b2e336
build: support clang-cl
2022-05-10 21:39:58 -06:00
c81e9ace28
remove old threading code entirely
...
add new parallel logging stuff in common/parallel which wraps parallel_for(_each) and provides automatic percentage indicator
fix issue with light not pulling in settings
move logging stuff into its own namespace and simplify
2022-03-01 10:28:59 -05:00
df92d9ce12
Command line args systems ( #2 )
...
* cmd
* cmake change for tbb DLLs
add in test for double-hyphens
* cmd
* game version
* settings!
* bsp texture usage
decompile fix for q1bsp
* finish up the options for qbsp
* light finished?
* finish up the few missing bits
* move scalar/int32 up
* Run clang-format pass
* natvis for our mathlib
use slightly faster storage for winding
fix light tests
* allow dictionary to be a base type
* re-run style
* Remove cfg_static
Fix options summary
* natvis for lockable_value<T>
* last-ish pass?
* format
* forgot last few lockables
2022-02-27 00:01:57 -05:00
74253fd189
Fixed Unix/Clang compilation
...
Adjusted makefile to work with both clang and msvc on win32
Removed "WIN32", use the standard _WIN32 instead
Fixed bad VectorCopy's
Use qvec4b instead of special color_rgba type
qvec now follows built-in type procedures for handling operators; this means that (qvec3i + qvec3f) will now properly produce a float, as one would expect, instead of adding the float to the integers without casting to the common type.
2021-10-23 07:35:41 -04:00
a21be0362c
replace VectorCompare, CrossProduct, DotProduct
...
qv::dot handles inputs a bit more gracefully now
plane::distance_to for distance checks, and dplane has the fast version
2021-10-21 23:25:19 -04:00
a7de4bdae2
Remove unused thingies
...
Use qplane3 where we can
2021-10-17 06:24:02 -04:00
fd18fc154c
Fix various warnings by testing a clang-cl compile
2021-10-12 14:39:15 -04:00
a2edc09018
Resolve GCC wanting `template` before emplacements
2021-10-05 22:54:25 -04:00
52839d4e7f
Run clang-format pass
2021-10-05 15:54:06 -04:00
9d828d2085
Structured texture data!
...
Move vis decompress/compress back to vis since phs is handled there now too
Fix missing textures in Q1BSP
Comment areaportaling for now
2021-09-30 11:33:04 -04:00
a6f6cc2d15
qvec now has an empty default initializer, for speed purposes; fixed up places that relied on this behavior
...
simplified constructor for qvec(T...) that uses fold expressions
simplified rgba_miptex_t
fixed a lot of broken winding_base_t behavior
merged all winding types into winding_base_t
removed now-empty files
2021-09-27 20:44:15 -04:00
87d31e6d4f
A lot of type-y stuff here:
...
- bsp store vectors now
- vis stored in a way that is compatible with both Q2 and Q1
- binary reading of stuff, so swapping is done at load time now
- vis cleanup, although there's still a lot of raw long modification that should get simplified later
2021-09-25 23:58:02 -04:00
Eric Wasylishen
Erika
Jonathan