R5Reloaded/r5sdk
1
0
Fork 0
mirror of https://github.com/Mauler125/r5sdk.git synced 2025-02-09 19:15:03 +01:00
Code Issues Packages Projects Releases Wiki Activity
r5sdk/r5dev/bsplib/bsplib.cpp

446 lines
18 KiB
C++
Raw Normal View History

Code base refactor + major performance and readability improvement. Read description for details. * Codebase restructured to SourceSDK codebase style and .cpp/.h assertion paths in the game executable. * Document most functions with valve style 'Purpose' blocks. * Rename variables to match the rest of the codebase and Valve's naming convention. * Dedicated DLL and the SDKLauncher now share the same codebase as the DevSDK. * Obtain globals or pointers directly instead of waiting for runtime initialized data. * Dynamically search for all functions and globals (this doesn't count for dedicated yet!). * Initialize most in-SDK variables. * Move certain prints and other utilities under ConVars to reduce verbosity and increase performance. * Print all pattern scan results through a virtual function to make it easier to add and debug new patterns in the future. * Type global var pointers appropriately if class or type is known and implemented. * Forward declare 'CClient' class to avoid having 2 'g_pClient' copies. * Add IDA's pseudo definitions for easier prototyping with decompiled assembly code. * RPAK decompress Command callback implementation. * Load decompressed RPaks from 'paks\Win32\' overriding the ones in 'paks\Win64\' (the decompress callback will automatically fix the header and write it to 'paks\Win32\'). * VPK decompress Command callback implementation. * Move CRC32 ands Adler32 to implementation files. * Server will print out more details about the connecting client. * Upgrade ImGui lib to v1.86. * Don't compile id3dx.h for dedicated. * Don't compile id3dx.cpp for dedicated * Implement DevMsg print function allowing to print information to the in-game VGUI/RUI console overlay, ImGui console overlay and the external windows console * Fixed bug where the Error function would not properly terminate the process when an error is called. This caused access violations for critical/non-recoverable errors. * Fixed bug where the game would crash if the console or server browser was enabled while the game was still starting up. * Several bug fixes for the dedicated server (warning: dedicated is still considered work-in-progress!).
2021-12-25 22:36:38 +01:00
#include "core/stdafx.h"
Move ConVar/ConCommand stuff to tier1 instead and cleanup KeyValues/KeyValuesSystem
2022-04-09 16:16:40 +02:00
#include "tier1/cvar.h"
See description Fixed rare crash in CMDLCache when buffer itteration returns invalid pointer (not null!) for handle. Made all variables and functions used by BuildPropStaticFrustumCullMap searched dynamically (these should also work on s0, s1 and s2, they are compared but at the moment untested). TODO: Fix deref crash properly (see bsplib.cpp l291).
2022-05-01 05:38:51 +02:00
#include "datacache/mdlcache.h"
Git client decided not to commit this.
2021-12-25 22:58:51 +01:00
#include "common/pseudodefs.h"
See description Fixed rare crash in CMDLCache when buffer itteration returns invalid pointer (not null!) for handle. Made all variables and functions used by BuildPropStaticFrustumCullMap searched dynamically (these should also work on s0, s1 and s2, they are compared but at the moment untested). TODO: Fix deref crash properly (see bsplib.cpp l291).
2022-05-01 05:38:51 +02:00
#include "materialsystem/cmaterialglue.h"
fixing CalcPropStaticFrustumCulling up.
2022-03-01 22:49:57 +01:00
#include "engine/host_state.h"
Initial working implementation of frustum building in SDK
2022-05-01 01:32:23 +02:00
#include "engine/modelloader.h"
See description Fixed rare crash in CMDLCache when buffer itteration returns invalid pointer (not null!) for handle. Made all variables and functions used by BuildPropStaticFrustumCullMap searched dynamically (these should also work on s0, s1 and s2, they are compared but at the moment untested). TODO: Fix deref crash properly (see bsplib.cpp l291).
2022-05-01 05:38:51 +02:00
#include "bsplib/bsplib.h"
Fill out cull map
2022-05-03 13:17:42 +02:00
//void* __fastcall BuildPropStaticFrustumCullMap(int64_t a1, int64_t a2, unsigned int a3, unsigned int a4, int64_t a5, int64_t a6, int64_t a7)
Refactored BuildPropStaticFrustumCullMap (see description) Redacted most code to only perform operation to obtain studiohdr and staticprop material (read only operation, function does not modify memory used by the original 'v_BuildPropStaticFrustumCullMap()' implementation!) The new function checks if pointer is within bounds (must stay within .data memory), if this isn't the case, the 'batch' is considered corrupt and function call to 'v_BuildPropStaticFrustumCullMap' won't proceed (note that the loop checks all bodygroups/LOD's of a model, it is possible that one doesn't have corrupt data, but a model where only one of its submeshes culls properly in the level is too much effort to fix as this function is based of disassembled output of the engine implementation, which in it self, comes with its own limitations). For a better implementation we should proceed fully rebuilding the function (see commented function body). The current and only problem with this is that it doesn't compute cull data properly (see bsplib.cpp line '222').
2022-05-03 02:37:37 +02:00
//{
// if (staticProp_defaultBuildFrustum->GetBool())
// return v_BuildPropStaticFrustumCullMap(a1, a2, a3, a4, a5, a6, a7);
//
// float v9; // xmm6_4
// char v10; // r13
// double v11; // xmm7_8
// __int64 v12; // rsi
// MDLHandle_t mdlhandle; // dx
// studiohdr_t* studio; // rbx
// unsigned __int8 v15; // r8
// const char* v16; // rax
// int v17; // eax
// char v18; // dl
// bool v19; // cc
// char v20; // al
// __int16 v21; // ax
// __int64 v22; // rax
// __m128 v23; // xmm9
// __m128 v24; // xmm10
// __m128 v25; // xmm11
// unsigned __int64 v26; // rbx
// __m128 v27; // xmm8
// __int64 v28; // rax
// __int64 v29; // rcx
// __int64 v30; // xmm1_8
// int v31; // er8
// __m128i v32; // xmm0
// __int64 v33; // rcx
// unsigned __int64 v34; // rdx
// __int64 v35; // rax
// __int64 v36; // rax
// unsigned __int64 v37; // rcx
// __int64 v38; // rax
// unsigned __int64 v39; // rcx
// __int64 v40; // rax
// float v41; // xmm2_4
// studiohdr_t* v42; // rbx
// float v43; // xmm2_4
// float v44; // xmm4_4
// __m128i v45; // xmm1
// __m128i v46; // xmm2
// __m128i v47; // xmm3
// unsigned int v48; // eax
// float v49; // xmm0_4
// float v50; // xmm2_4
// float v51; // xmm1_4
// float v52; // xmm2_4
// unsigned __int8 v53; // si
// __int64 v54; // rax
// __int64 v55; // rcx
// int v56; // eax
// __int64 v57; // rcx
// int v58; // edx
// __int64 v59; // rax
// __int64 v60; // r13
// int v61; // eax
// char* v62; // r13
// __int64 v63; // rdx
// int v64; // er14
// char* v65; // rdi
// __int64 v66; // r13
// __int64 v67; // r15
// void* v68; // rbx
// int32_t v69; // rcx
Fill out cull map
2022-05-03 13:17:42 +02:00
// void* result; // rax
Refactored BuildPropStaticFrustumCullMap (see description) Redacted most code to only perform operation to obtain studiohdr and staticprop material (read only operation, function does not modify memory used by the original 'v_BuildPropStaticFrustumCullMap()' implementation!) The new function checks if pointer is within bounds (must stay within .data memory), if this isn't the case, the 'batch' is considered corrupt and function call to 'v_BuildPropStaticFrustumCullMap' won't proceed (note that the loop checks all bodygroups/LOD's of a model, it is possible that one doesn't have corrupt data, but a model where only one of its submeshes culls properly in the level is too much effort to fix as this function is based of disassembled output of the engine implementation, which in it self, comes with its own limitations). For a better implementation we should proceed fully rebuilding the function (see commented function body). The current and only problem with this is that it doesn't compute cull data properly (see bsplib.cpp line '222').
2022-05-03 02:37:37 +02:00
// __m128i v71; // [rsp+38h] [rbp-D0h] BYREF
// __int64 v72{}; // [rsp+48h] [rbp-C0h]
// __int64 v73; // [rsp+50h] [rbp-B8h]
// __m128 v74; // [rsp+58h] [rbp-B0h] BYREF
// __int64 v75{}; // [rsp+68h] [rbp-A0h]
// __int64 v76; // [rsp+78h] [rbp-90h]
// __m128 v77[3]; // [rsp+88h] [rbp-80h] BYREF
// char Destination[376]; // [rsp+B8h] [rbp-50h] BYREF
// __int64 v79; // [rsp+278h] [rbp+170h]
// int v82; // [rsp+288h] [rbp+180h]
// int v83; // [rsp+290h] [rbp+188h]
// studiohdr_t* v84; // [rsp+298h] [rbp+190h]
// __int64 v85; // [rsp+2A8h] [rbp+1A0h]
//
// v9 = 1.0f;
// v10 = a4;
// *(_QWORD*)(a1 + 20) = *(_QWORD*)a5;
// *(_DWORD*)(a1 + 28) = *(_DWORD*)(a5 + 8);
// *(_DWORD*)(a1 + 8) = a4;
// *(_QWORD*)&v11 = *(unsigned int*)(a5 + 24);
// v12 = a4 >> 1;
// *(float*)(a1 + 12) = 1.0f / (float)(*(float*)&v11 * *(float*)&v11);
// mdlhandle = *(unsigned __int16*)(a7 + 0x140);
// *(MDLHandle_t*)a1 = mdlhandle;
// studio = CMDLCache::FindMDL(g_MDLCache, mdlhandle, nullptr);
// v84 = studio;
// if ((studio->flags & STUDIOHDR_FLAGS_STATIC_PROP) == 0 && (*dword_1696A9D20) < 0x64)
// ++(*dword_1696A9D20);
// v15 = *(_BYTE*)(a5 + 30);
// if (v15 > 2u && (unsigned __int8)(v15 - 6) > 2u)
// {
// v16 = (const char*)(*((__int64(__fastcall**)(CModelLoader**, __int64))g_pModelLoader + 4))(&g_pModelLoader, a7);
// strncpy(Destination, v16, 0x104ui64);
// v15 = 0;
// }
// v17 = *(unsigned __int8*)(a5 + 32);
// *(_BYTE*)(a1 + 4) = v17;
// v18 = v17;
// v19 = v17 < studio->numskinfamilies;
// *(_BYTE*)(a1 + 5) = v15;
// if (!v19)
// v18 = 0;
// *(_BYTE*)(a1 + 4) = v18;
// v20 = *(_BYTE*)(a5 + 31);
// if ((v20 & 4) != 0)
// {
// v15 |= 0x40u;
// *(_BYTE*)(a1 + 5) = v15;
// v20 = *(_BYTE*)(a5 + 31);
// }
// if ((v20 & 8) != 0)
// {
// v15 |= 0x20u;
// *(_BYTE*)(a1 + 5) = v15;
// v20 = *(_BYTE*)(a5 + 31);
// }
// if ((v20 & 0x30) != 16)
// *(_BYTE*)(a1 + 5) = v15 | 0x10;
// v21 = 0;
// if (*(_WORD*)(a5 + 34) != 0xFFFF)
// v21 = *(_WORD*)(a5 + 34);
// *(_WORD*)(a1 + 2) = v21;
// sub_1404365A0(v77, (const __m128i*)a5, (__m128i*)(a5 + 12), v11);
// v22 = *qword_141744EA8;
// v23 = v77[0];
// v24 = v77[1];
// v25 = v77[2];
// v26 = (unsigned __int64)(unsigned int)v12 << 6;
// *(__m128*)(v26 + (*qword_141744EA8)) = v77[0];
// *(__m128*)(v26 + v22 + 16) = v24;
// *(__m128*)(v26 + v22 + 32) = v25;
// __m128i m2 = { 0 };
// __m128i m3 = { 0 };
// v74 = _mm_mul_ps(_mm_cvtepi32_ps(_mm_unpacklo_epi16(_mm_unpacklo_epi8(_mm_cvtsi32_si128(*(_DWORD*)(a5 + 52)), m2), m3)), (*xmmword_1415BD270));
// v27 = sub_140270130(&v74);
// *(__m128*)(v26 + (*qword_141744EA8) + 48) = v27;// DIFF 1
// (*((void(__fastcall**)(void**, __int64, __m128i*, __int8*)) * (void**)off_141731448 + 10))(&*(void**)off_141731448, a7, &v71, &v71.m128i_i8[12]);
// sub_14028F170((__int64)&v74, (__int64)&v74.m128_i64[1] + 4, v77, &v71, (const __m128i*) & v71.m128i_i8[12]);
// v28 = *qword_141744EA0;
// v29 = 3 * v12;
// v30 = v75;
// *(__m128*)((*qword_141744EA0) + 8 * v29) = v74;
// *(_QWORD*)(v28 + 8 * v29 + 16) = v30;
// if ((v10 & 1) != 0)
// {
// v31 = *dword_141744EBC;
// v32 = v71;
// *(_DWORD*)a2 = *(_DWORD*)(a6 + 48);
// *(_DWORD*)(a2 + 4) = *(_DWORD*)(a6 + 52);
// *(_QWORD*)(a2 + 8) = 0i64;
// v33 = 3i64 * (unsigned int)(v31 + v12);
// v34 = (unsigned __int64)(unsigned int)(v31 + v12) << 6;
// v35 = *qword_141744EA0;
// *(__m128i*)((*qword_141744EA0) + 8 * v33) = v32;
// *(_QWORD*)(v35 + 8 * v33 + 16) = v72;
// v36 = *qword_141744EA8;
// v37 = (unsigned __int64)(unsigned int)(v12 + 2 * v31) << 6;
// *(__m128*)(v34 + (*qword_141744EA8)) = v23;
// *(__m128*)(v34 + v36 + 16) = v24;
// *(__m128*)(v34 + v36 + 32) = v25;
// *(__m128*)(v34 + (*qword_141744EA8) + 48) = v27;
// v38 = *qword_141744EA8;
// *(__m128*)(v37 + (*qword_141744EA8)) = v23;
// *(__m128*)(v37 + v38 + 16) = v24;
// *(__m128*)(v37 + v38 + 32) = v25;
// *(__m128*)(v37 + (*qword_141744EA8) + 48) = v27;
// v39 = (unsigned __int64)(unsigned int)(v31 + v12 + 2 * v31) << 6;
// *(__m128*)(v39 + (*qword_141744EA8) + 48) = v27;
// v40 = *qword_141744EA8;
// *(__m128*)(v39 + (*qword_141744EA8)) = *(__m128*)a6; //*(_OWORD*)(v39 + qword_141744EA8) = *(_OWORD*)a6;
// *(__m128*)(v39 + v40 + 16) = *(__m128*)(a6 + 16); //*(_OWORD*)(v39 + v40 + 16) = *(_OWORD*)(a6 + 16);
// *(__m128*)(v39 + v40 + 32) = *(__m128*)(a6 + 32); //*(__m128*)(v40 + v41 + 32) = *(__m128*)(a6 + 32);
// }
// v41 = *(float*)(a5 + 36);
// v42 = v84;
// if (v41 <= 0.0f)
// {
// if ((v84->flags & STUDIOHDR_FLAGS_NO_FORCED_FADE) != 0)
// {
// v41 = 227023.36f;
// }
// else
// {
// v43 = v84->fadeDist;
// if (v43 <= 0.0f)
// v41 = fmaxf(
// (float)((float)(sqrtf(
// (float)((float)((float)(*(float*)v71.m128i_i32 - *(float*)&v71.m128i_i32[3]) * (float)(*(float*)v71.m128i_i32 - *(float*)&v71.m128i_i32[3]))
// + (float)((float)(*(float*)&v71.m128i_i32[1] - *(float*)&v72) * (float)(*(float*)&v71.m128i_i32[1] - *(float*)&v72)))
// + (float)((float)(*(float*)&v71.m128i_i32[2] - *((float*)&v72 + 1)) * (float)(*(float*)&v71.m128i_i32[2] - *((float*)&v72 + 1))))
// * 0.5)
// * *(float*)&v11)
// * model_defaultFadeDistScale->GetFloat(),
// model_defaultFadeDistMin->GetFloat());
// else
// v41 = v43 * *(float*)&v11;
// }
// }
// v44 = fmaxf(v41, 100.0f);
// *(float*)(a1 + 16) = v44 * v44; // <-- Data written here is incorrect [v41 used in the 'fmaxf' operation is most likely computed wrong]!
// v45 = _mm_castps_si128(_mm_load_ss(&v84->gatherSize));
// if (*(float*)v45.m128i_i32 <= 0.0f)
// {
// LOWORD(v48) = 0;
// }
// else
// {
// *(float*)v45.m128i_i32 = *(float*)v45.m128i_i32 * *(float*)v45.m128i_i32;
// v46 = v45;
// *(float*)v46.m128i_i32 = fmaxf(*(float*)v45.m128i_i32, 1.0004883f);
// v47 = v46;
// *(float*)v47.m128i_i32 = fminf(*(float*)v46.m128i_i32, 4293918700.0f);
// v48 = (unsigned int)(_mm_cvtsi128_si32(v47) - 1065351168) >> 12;
// }
// *(_WORD*)(a1 + 6) = v48; // <-- DEBUG!
// v49 = *(float*)((*qword_141744EA0) + 24 * ((unsigned __int64)*(unsigned int*)(a1 + 8) >> 1) + 8) - *(float*)((*qword_141744EA0) + 24 * ((unsigned __int64)*(unsigned int*)(a1 + 8) >> 1) + 20);
// v50 = *(float*)((*qword_141744EA0) + 24 * ((unsigned __int64)*(unsigned int*)(a1 + 8) >> 1) + 4) - *(float*)((*qword_141744EA0) + 24 * ((unsigned __int64)*(unsigned int*)(a1 + 8) >> 1) + 16);
// v51 = *(float*)((*qword_141744EA0) + 24 * ((unsigned __int64)*(unsigned int*)(a1 + 8) >> 1)) - *(float*)((*qword_141744EA0) + 24 * ((unsigned __int64)*(unsigned int*)(a1 + 8) >> 1) + 12);
// v52 = (float)((float)(v50 * v50) + (float)(v51 * v51)) + (float)(v49 * v49);
// if (v44 >= 227023.363449684f)
// v9 = staticProp_no_fade_scalar->GetFloat();
// v53 = 0;
// *(float*)((*qword_141744E88) + 8i64 * a3) = v9 * (float)(1.0f / (float)(v52 * staticProp_gather_size_weight->GetFloat()));
// v54 = *qword_141744E88;
// *(_BYTE*)((*qword_141744E88) + 8i64 * a3 + 4) &= 0xFEu;
// *(_BYTE*)(v54 + 8i64 * a3 + 4) |= v44 >= 227023.363449684f;
// v55 = (__int64)CMDLCache::GetStudioMaterialGlue(g_MDLCache, *(unsigned __int16*)(a7 + 320)); // Gets some object containing pointer to 2 CMaterialGlue vtables.
// v56 = *(unsigned __int16*)(a5 + 0x20);
// v76 = *(__int64*)v55;
// v57 = (__int64)v84 + 2 * v56 * v84->numskinref + v84->skinindex;
// v58 = 0;
// v85 = v57;
// v83 = 0;
// if (v84->numbodyparts <= 0)
// return 0i64;
// v59 = 0i64;
// v73 = 0i64;
// do
// {
// v60 = v59 + v42->bodypartindex;
// v61 = 0;
// v62 = (char*)v42 + v60;
// v82 = 0;
// v74.m128_u64[0] = (unsigned __int64)v62;
// if (*((int*)v62 + 1) > 0)
// {
// v63 = 0i64;
// v79 = 0i64;
// do
// {
// v64 = 0;
// v65 = &v62[v63 + *((int*)v62 + 3)];
// if (*((int*)v65 + 19) > 0)
// {
// v66 = v76;
// v67 = 0i64;
// do
// {
// v68 = *(void**)(v66 + 8i64 * *(__int16*)(v57 + 2i64 * *(int*)(v67 + *(int*)(v65 + 80) + v65)));
Fill out cull map
2022-05-03 13:17:42 +02:00
//
// static CModule::ModuleSections_t mData = g_mGameDll.GetSectionByName(".data");
// static CModule::ModuleSections_t mPData = g_mGameDll.GetSectionByName(".pdata");
// if (reinterpret_cast<uintptr_t>(v68) < mData.m_pSectionBase ||
// reinterpret_cast<uintptr_t>(v68) > mPData.m_pSectionBase) // Check bounds (data could only be within the '.data' segment.
// continue;
//
Refactored BuildPropStaticFrustumCullMap (see description) Redacted most code to only perform operation to obtain studiohdr and staticprop material (read only operation, function does not modify memory used by the original 'v_BuildPropStaticFrustumCullMap()' implementation!) The new function checks if pointer is within bounds (must stay within .data memory), if this isn't the case, the 'batch' is considered corrupt and function call to 'v_BuildPropStaticFrustumCullMap' won't proceed (note that the loop checks all bodygroups/LOD's of a model, it is possible that one doesn't have corrupt data, but a model where only one of its submeshes culls properly in the level is too much effort to fix as this function is based of disassembled output of the engine implementation, which in it self, comes with its own limitations). For a better implementation we should proceed fully rebuilding the function (see commented function body). The current and only problem with this is that it doesn't compute cull data properly (see bsplib.cpp line '222').
2022-05-03 02:37:37 +02:00
// if (!(*(unsigned __int8(__fastcall**)(void*))(*(_QWORD*)v68 + 688i64))(v68))
// {
// if (!(*(unsigned __int8(__fastcall**)(void*))(*(_QWORD*)v68 + 256i64))(v68) && (*(unsigned __int8(__fastcall**)(void*))(*(_QWORD*)v68 + 248i64))(v68))
// {
// v69 = 0;
// if (*dword_141744EE8)
// {
// // Compares equality of pointers to the CMaterial object
// // between the current offset [v69] in the CStaticPropMgr buffer
// // and that of v68 (obtained via operation in parent loop).
//
// while (*(&*(void**)off_141744E70 + v69 + 16) != v68)
// {
// v69++;// = (unsigned int)(v69 + 1);
//
// if (static_cast<int32_t>(v69) >= *dword_141744EE8)
// goto LABEL_42;
// }
// }
// else
// {
// LABEL_42:
// *(&*(void**)off_141744E70 + (*dword_141744EE8)++ + 16) = v68;
// }
// }
// if ((*(unsigned __int8(__fastcall**)(void*))(*(_QWORD*)v68 + 168i64))(v68) && (*(unsigned int(__fastcall**)(void*, __int64))(*(_QWORD*)v68 + 144i64))(v68, 1i64))
// *(_BYTE*)(a1 + 5) |= 0x80u;
// if (!(*(unsigned __int8(__fastcall**)(void*))(*(_QWORD*)v68 + 384i64))(v68) && (*(unsigned int(__fastcall**)(void*, __int64))(*(_QWORD*)v68 + 144i64))(v68, 21844i64))
// v53 |= 2u;
// v53 |= (*(unsigned __int8(__fastcall**)(void*))(*(_QWORD*)v68 + 384i64))(v68) != 0;
// }
// v57 = v85;
// ++v64;
// v67 += 92i64;
// } while (v64 < *((_DWORD*)v65 + 19));
// v62 = (char*)v74.m128_u64[0];
// v61 = v82;
// v63 = v79;
// }
// ++v61;
// v63 += 136i64;
// v82 = v61;
// v79 = v63;
// } while (v61 < *((_DWORD*)v62 + 1));
// v42 = v84;
// v58 = v83;
// }
// ++v58;
// v59 = v73 + 16;
// v83 = v58;
// v73 += 16i64;
// } while (v58 < v42->numbodyparts);
Fill out cull map
2022-05-03 13:17:42 +02:00
// result = (void*)v53;
Refactored BuildPropStaticFrustumCullMap (see description) Redacted most code to only perform operation to obtain studiohdr and staticprop material (read only operation, function does not modify memory used by the original 'v_BuildPropStaticFrustumCullMap()' implementation!) The new function checks if pointer is within bounds (must stay within .data memory), if this isn't the case, the 'batch' is considered corrupt and function call to 'v_BuildPropStaticFrustumCullMap' won't proceed (note that the loop checks all bodygroups/LOD's of a model, it is possible that one doesn't have corrupt data, but a model where only one of its submeshes culls properly in the level is too much effort to fix as this function is based of disassembled output of the engine implementation, which in it self, comes with its own limitations). For a better implementation we should proceed fully rebuilding the function (see commented function body). The current and only problem with this is that it doesn't compute cull data properly (see bsplib.cpp line '222').
2022-05-03 02:37:37 +02:00
// if (v53)
// *(_BYTE*)(a1 + 5) &= 0x7Fu;
// return result;
//}
Code base refactor + major performance and readability improvement. Read description for details. * Codebase restructured to SourceSDK codebase style and .cpp/.h assertion paths in the game executable. * Document most functions with valve style 'Purpose' blocks. * Rename variables to match the rest of the codebase and Valve's naming convention. * Dedicated DLL and the SDKLauncher now share the same codebase as the DevSDK. * Obtain globals or pointers directly instead of waiting for runtime initialized data. * Dynamically search for all functions and globals (this doesn't count for dedicated yet!). * Initialize most in-SDK variables. * Move certain prints and other utilities under ConVars to reduce verbosity and increase performance. * Print all pattern scan results through a virtual function to make it easier to add and debug new patterns in the future. * Type global var pointers appropriately if class or type is known and implemented. * Forward declare 'CClient' class to avoid having 2 'g_pClient' copies. * Add IDA's pseudo definitions for easier prototyping with decompiled assembly code. * RPAK decompress Command callback implementation. * Load decompressed RPaks from 'paks\Win32\' overriding the ones in 'paks\Win64\' (the decompress callback will automatically fix the header and write it to 'paks\Win32\'). * VPK decompress Command callback implementation. * Move CRC32 ands Adler32 to implementation files. * Server will print out more details about the connecting client. * Upgrade ImGui lib to v1.86. * Don't compile id3dx.h for dedicated. * Don't compile id3dx.cpp for dedicated * Implement DevMsg print function allowing to print information to the in-game VGUI/RUI console overlay, ImGui console overlay and the external windows console * Fixed bug where the Error function would not properly terminate the process when an error is called. This caused access violations for critical/non-recoverable errors. * Fixed bug where the game would crash if the console or server browser was enabled while the game was still starting up. * Several bug fixes for the dedicated server (warning: dedicated is still considered work-in-progress!).
2021-12-25 22:36:38 +01:00
//-----------------------------------------------------------------------------
// Purpose: calculates the view frustum culling data per static prop
//-----------------------------------------------------------------------------
More cleanup and slight performance boost
2022-05-03 03:40:47 +02:00
void* __fastcall BuildPropStaticFrustumCullMap(int64_t a1, int64_t a2, unsigned int a3, unsigned int a4, int64_t a5, int64_t a6, int64_t a7)
Code base refactor + major performance and readability improvement. Read description for details. * Codebase restructured to SourceSDK codebase style and .cpp/.h assertion paths in the game executable. * Document most functions with valve style 'Purpose' blocks. * Rename variables to match the rest of the codebase and Valve's naming convention. * Dedicated DLL and the SDKLauncher now share the same codebase as the DevSDK. * Obtain globals or pointers directly instead of waiting for runtime initialized data. * Dynamically search for all functions and globals (this doesn't count for dedicated yet!). * Initialize most in-SDK variables. * Move certain prints and other utilities under ConVars to reduce verbosity and increase performance. * Print all pattern scan results through a virtual function to make it easier to add and debug new patterns in the future. * Type global var pointers appropriately if class or type is known and implemented. * Forward declare 'CClient' class to avoid having 2 'g_pClient' copies. * Add IDA's pseudo definitions for easier prototyping with decompiled assembly code. * RPAK decompress Command callback implementation. * Load decompressed RPaks from 'paks\Win32\' overriding the ones in 'paks\Win64\' (the decompress callback will automatically fix the header and write it to 'paks\Win32\'). * VPK decompress Command callback implementation. * Move CRC32 ands Adler32 to implementation files. * Server will print out more details about the connecting client. * Upgrade ImGui lib to v1.86. * Don't compile id3dx.h for dedicated. * Don't compile id3dx.cpp for dedicated * Implement DevMsg print function allowing to print information to the in-game VGUI/RUI console overlay, ImGui console overlay and the external windows console * Fixed bug where the Error function would not properly terminate the process when an error is called. This caused access violations for critical/non-recoverable errors. * Fixed bug where the game would crash if the console or server browser was enabled while the game was still starting up. * Several bug fixes for the dedicated server (warning: dedicated is still considered work-in-progress!).
2021-12-25 22:36:38 +01:00
{
Initial working implementation of frustum building in SDK
2022-05-01 01:32:23 +02:00
if (staticProp_defaultBuildFrustum->GetBool())
See description Fixed rare crash in CMDLCache when buffer itteration returns invalid pointer (not null!) for handle. Made all variables and functions used by BuildPropStaticFrustumCullMap searched dynamically (these should also work on s0, s1 and s2, they are compared but at the moment untested). TODO: Fix deref crash properly (see bsplib.cpp l291).
2022-05-01 05:38:51 +02:00
return v_BuildPropStaticFrustumCullMap(a1, a2, a3, a4, a5, a6, a7);
Initial working implementation of frustum building in SDK
2022-05-01 01:32:23 +02:00
More cleanup and slight performance boost
2022-05-03 03:40:47 +02:00
MDLHandle_t handle; // dx
studiohdr_t *studio; // rbx
Fill out cull map
2022-05-03 13:17:42 +02:00
double v54; // xmm7_8
More cleanup and slight performance boost
2022-05-03 03:40:47 +02:00
int64_t v55; // rcx
int v56; // eax
int64_t v57; // rcx
int v58; // edx
int64_t v59; // rax
int64_t v60; // r13
int v61; // eax
char *v62; // r13
int64_t v63; // rdx
int v64; // er14
char *v65; // rdi
int64_t v67; // r15
void *v68; // rbx
int64_t v73; // [rsp+50h] [rbp-B8h]
Fill out cull map
2022-05-03 13:17:42 +02:00
bool error = false;
Code base refactor + major performance and readability improvement. Read description for details. * Codebase restructured to SourceSDK codebase style and .cpp/.h assertion paths in the game executable. * Document most functions with valve style 'Purpose' blocks. * Rename variables to match the rest of the codebase and Valve's naming convention. * Dedicated DLL and the SDKLauncher now share the same codebase as the DevSDK. * Obtain globals or pointers directly instead of waiting for runtime initialized data. * Dynamically search for all functions and globals (this doesn't count for dedicated yet!). * Initialize most in-SDK variables. * Move certain prints and other utilities under ConVars to reduce verbosity and increase performance. * Print all pattern scan results through a virtual function to make it easier to add and debug new patterns in the future. * Type global var pointers appropriately if class or type is known and implemented. * Forward declare 'CClient' class to avoid having 2 'g_pClient' copies. * Add IDA's pseudo definitions for easier prototyping with decompiled assembly code. * RPAK decompress Command callback implementation. * Load decompressed RPaks from 'paks\Win32\' overriding the ones in 'paks\Win64\' (the decompress callback will automatically fix the header and write it to 'paks\Win32\'). * VPK decompress Command callback implementation. * Move CRC32 ands Adler32 to implementation files. * Server will print out more details about the connecting client. * Upgrade ImGui lib to v1.86. * Don't compile id3dx.h for dedicated. * Don't compile id3dx.cpp for dedicated * Implement DevMsg print function allowing to print information to the in-game VGUI/RUI console overlay, ImGui console overlay and the external windows console * Fixed bug where the Error function would not properly terminate the process when an error is called. This caused access violations for critical/non-recoverable errors. * Fixed bug where the game would crash if the console or server browser was enabled while the game was still starting up. * Several bug fixes for the dedicated server (warning: dedicated is still considered work-in-progress!).
2021-12-25 22:36:38 +01:00
More cleanup and slight performance boost
2022-05-03 03:40:47 +02:00
handle = *reinterpret_cast<uint16_t*>(a7 + 0x140);
Fill out cull map
2022-05-03 13:17:42 +02:00
studio = g_MDLCache->FindMDL(g_MDLCache, handle, nullptr);
v55 = *reinterpret_cast<int64_t*>(g_MDLCache->GetStudioMaterialGlue(g_MDLCache, *reinterpret_cast<uint16_t*>((a7 + 320)))); // Gets some object containing pointer to 2 CMaterialGlue vtables.
More cleanup and slight performance boost
2022-05-03 03:40:47 +02:00
v56 = *reinterpret_cast<uint16_t*>(a5 + 0x20);
Fill out cull map
2022-05-03 13:17:42 +02:00
v57 = reinterpret_cast<int64_t>(studio) + 2i64 * v56 * studio->numskinref + studio->skinindex;
fixing CalcPropStaticFrustumCulling up.
2022-03-01 22:49:57 +01:00
v58 = 0;
Refactored BuildPropStaticFrustumCullMap (see description) Redacted most code to only perform operation to obtain studiohdr and staticprop material (read only operation, function does not modify memory used by the original 'v_BuildPropStaticFrustumCullMap()' implementation!) The new function checks if pointer is within bounds (must stay within .data memory), if this isn't the case, the 'batch' is considered corrupt and function call to 'v_BuildPropStaticFrustumCullMap' won't proceed (note that the loop checks all bodygroups/LOD's of a model, it is possible that one doesn't have corrupt data, but a model where only one of its submeshes culls properly in the level is too much effort to fix as this function is based of disassembled output of the engine implementation, which in it self, comes with its own limitations). For a better implementation we should proceed fully rebuilding the function (see commented function body). The current and only problem with this is that it doesn't compute cull data properly (see bsplib.cpp line '222').
2022-05-03 02:37:37 +02:00
if (studio->numbodyparts <= 0)
Little cleanup
2022-05-03 03:17:45 +02:00
return nullptr;
fixing CalcPropStaticFrustumCulling up.
2022-03-01 22:49:57 +01:00
v59 = 0i64;
v73 = 0i64;
do
{
Refactored BuildPropStaticFrustumCullMap (see description) Redacted most code to only perform operation to obtain studiohdr and staticprop material (read only operation, function does not modify memory used by the original 'v_BuildPropStaticFrustumCullMap()' implementation!) The new function checks if pointer is within bounds (must stay within .data memory), if this isn't the case, the 'batch' is considered corrupt and function call to 'v_BuildPropStaticFrustumCullMap' won't proceed (note that the loop checks all bodygroups/LOD's of a model, it is possible that one doesn't have corrupt data, but a model where only one of its submeshes culls properly in the level is too much effort to fix as this function is based of disassembled output of the engine implementation, which in it self, comes with its own limitations). For a better implementation we should proceed fully rebuilding the function (see commented function body). The current and only problem with this is that it doesn't compute cull data properly (see bsplib.cpp line '222').
2022-05-03 02:37:37 +02:00
v60 = v59 + studio->bodypartindex;
fixing CalcPropStaticFrustumCulling up.
2022-03-01 22:49:57 +01:00
v61 = 0;
More cleanup and slight performance boost
2022-05-03 03:40:47 +02:00
v62 = reinterpret_cast<char*>(studio) + v60;
Some 'BuildPropStaticFrustumCullMap' improvements Artifact problem for large maps is still not solved.
2022-05-02 03:13:39 +02:00
if (*((int*)v62 + 1) > 0)
fixing CalcPropStaticFrustumCulling up.
2022-03-01 22:49:57 +01:00
{
v63 = 0i64;
do
{
v64 = 0;
Some 'BuildPropStaticFrustumCullMap' improvements Artifact problem for large maps is still not solved.
2022-05-02 03:13:39 +02:00
v65 = &v62[v63 + *((int*)v62 + 3)];
if (*((int*)v65 + 19) > 0)
fixing CalcPropStaticFrustumCulling up.
2022-03-01 22:49:57 +01:00
{
v67 = 0i64;
do
{
Refactored BuildPropStaticFrustumCullMap (see description) Redacted most code to only perform operation to obtain studiohdr and staticprop material (read only operation, function does not modify memory used by the original 'v_BuildPropStaticFrustumCullMap()' implementation!) The new function checks if pointer is within bounds (must stay within .data memory), if this isn't the case, the 'batch' is considered corrupt and function call to 'v_BuildPropStaticFrustumCullMap' won't proceed (note that the loop checks all bodygroups/LOD's of a model, it is possible that one doesn't have corrupt data, but a model where only one of its submeshes culls properly in the level is too much effort to fix as this function is based of disassembled output of the engine implementation, which in it self, comes with its own limitations). For a better implementation we should proceed fully rebuilding the function (see commented function body). The current and only problem with this is that it doesn't compute cull data properly (see bsplib.cpp line '222').
2022-05-03 02:37:37 +02:00
v68 = *(void**)(v55 + 8i64 * *(__int16*)(v57 + 2i64 * *(int*)(v67 + *(int*)(v65 + 80) + v65)));
Fill out cull map
2022-05-03 13:17:42 +02:00
++v64;
v67 += 92i64;
Improved BuildPropStaticFrustumCullMap * Fixed bug where the address of g_pStaticPropMgr wasn't properly dereferenced and passed to function via its reference (engine addr). * Fixed dword dereference. Problems: engine still culls models improperly.
2022-05-01 21:09:05 +02:00
Refactored BuildPropStaticFrustumCullMap (see description) Redacted most code to only perform operation to obtain studiohdr and staticprop material (read only operation, function does not modify memory used by the original 'v_BuildPropStaticFrustumCullMap()' implementation!) The new function checks if pointer is within bounds (must stay within .data memory), if this isn't the case, the 'batch' is considered corrupt and function call to 'v_BuildPropStaticFrustumCullMap' won't proceed (note that the loop checks all bodygroups/LOD's of a model, it is possible that one doesn't have corrupt data, but a model where only one of its submeshes culls properly in the level is too much effort to fix as this function is based of disassembled output of the engine implementation, which in it self, comes with its own limitations). For a better implementation we should proceed fully rebuilding the function (see commented function body). The current and only problem with this is that it doesn't compute cull data properly (see bsplib.cpp line '222').
2022-05-03 02:37:37 +02:00
static CModule::ModuleSections_t mData = g_mGameDll.GetSectionByName(".data");
static CModule::ModuleSections_t mPData = g_mGameDll.GetSectionByName(".pdata");
Fill out cull map
2022-05-03 13:17:42 +02:00
if (reinterpret_cast<uintptr_t>(v68) < mData.m_pSectionBase ||
reinterpret_cast<uintptr_t>(v68) > mPData.m_pSectionBase || error) // Check bounds (data could only be within the '.data' segment.
{
error = true;
continue;
}
Little cleanup
2022-05-03 03:17:45 +02:00
} while (v64 < *((int*)v65 + 19));
fixing CalcPropStaticFrustumCulling up.
2022-03-01 22:49:57 +01:00
}
++v61;
v63 += 136i64;
Little cleanup
2022-05-03 03:17:45 +02:00
} while (v61 < *((int*)v62 + 1));
fixing CalcPropStaticFrustumCulling up.
2022-03-01 22:49:57 +01:00
}
++v58;
v59 = v73 + 16;
v73 += 16i64;
Refactored BuildPropStaticFrustumCullMap (see description) Redacted most code to only perform operation to obtain studiohdr and staticprop material (read only operation, function does not modify memory used by the original 'v_BuildPropStaticFrustumCullMap()' implementation!) The new function checks if pointer is within bounds (must stay within .data memory), if this isn't the case, the 'batch' is considered corrupt and function call to 'v_BuildPropStaticFrustumCullMap' won't proceed (note that the loop checks all bodygroups/LOD's of a model, it is possible that one doesn't have corrupt data, but a model where only one of its submeshes culls properly in the level is too much effort to fix as this function is based of disassembled output of the engine implementation, which in it self, comes with its own limitations). For a better implementation we should proceed fully rebuilding the function (see commented function body). The current and only problem with this is that it doesn't compute cull data properly (see bsplib.cpp line '222').
2022-05-03 02:37:37 +02:00
} while (v58 < studio->numbodyparts);
Fill out cull map
2022-05-03 13:17:42 +02:00
if (error) // Don't use engine's implementation if batch contains errors!
{
*(MDLHandle_t*)a1 = handle;
*(_QWORD*)&v54 = *(unsigned int*)(a5 + 24);
*(_DWORD*)(a1 + 8) = a4;
*(float*)(a1 + 12) = 1.0f / (float)(*(float*)&v54 * *(float*)&v54);
*(float*)(a1 + 16) = 227023.36f * 227023.36f; // STUDIOHDR_FLAGS_NO_FORCED_FADE
*(_QWORD*)(a1 + 20) = *(_QWORD*)a5;
*(_DWORD*)(a1 + 28) = *(_DWORD*)(a5 + 8);
return nullptr;
}
Refactored BuildPropStaticFrustumCullMap (see description) Redacted most code to only perform operation to obtain studiohdr and staticprop material (read only operation, function does not modify memory used by the original 'v_BuildPropStaticFrustumCullMap()' implementation!) The new function checks if pointer is within bounds (must stay within .data memory), if this isn't the case, the 'batch' is considered corrupt and function call to 'v_BuildPropStaticFrustumCullMap' won't proceed (note that the loop checks all bodygroups/LOD's of a model, it is possible that one doesn't have corrupt data, but a model where only one of its submeshes culls properly in the level is too much effort to fix as this function is based of disassembled output of the engine implementation, which in it self, comes with its own limitations). For a better implementation we should proceed fully rebuilding the function (see commented function body). The current and only problem with this is that it doesn't compute cull data properly (see bsplib.cpp line '222').
2022-05-03 02:37:37 +02:00
return v_BuildPropStaticFrustumCullMap(a1, a2, a3, a4, a5, a6, a7);
Code base refactor + major performance and readability improvement. Read description for details. * Codebase restructured to SourceSDK codebase style and .cpp/.h assertion paths in the game executable. * Document most functions with valve style 'Purpose' blocks. * Rename variables to match the rest of the codebase and Valve's naming convention. * Dedicated DLL and the SDKLauncher now share the same codebase as the DevSDK. * Obtain globals or pointers directly instead of waiting for runtime initialized data. * Dynamically search for all functions and globals (this doesn't count for dedicated yet!). * Initialize most in-SDK variables. * Move certain prints and other utilities under ConVars to reduce verbosity and increase performance. * Print all pattern scan results through a virtual function to make it easier to add and debug new patterns in the future. * Type global var pointers appropriately if class or type is known and implemented. * Forward declare 'CClient' class to avoid having 2 'g_pClient' copies. * Add IDA's pseudo definitions for easier prototyping with decompiled assembly code. * RPAK decompress Command callback implementation. * Load decompressed RPaks from 'paks\Win32\' overriding the ones in 'paks\Win64\' (the decompress callback will automatically fix the header and write it to 'paks\Win32\'). * VPK decompress Command callback implementation. * Move CRC32 ands Adler32 to implementation files. * Server will print out more details about the connecting client. * Upgrade ImGui lib to v1.86. * Don't compile id3dx.h for dedicated. * Don't compile id3dx.cpp for dedicated * Implement DevMsg print function allowing to print information to the in-game VGUI/RUI console overlay, ImGui console overlay and the external windows console * Fixed bug where the Error function would not properly terminate the process when an error is called. This caused access violations for critical/non-recoverable errors. * Fixed bug where the game would crash if the console or server browser was enabled while the game was still starting up. * Several bug fixes for the dedicated server (warning: dedicated is still considered work-in-progress!).
2021-12-25 22:36:38 +01:00
}
void BspLib_Attach()
{
Optimize some CHLClient functions away for dedicated
2022-05-04 21:10:46 +02:00
#ifndef DEDICATED
See description Fixed rare crash in CMDLCache when buffer itteration returns invalid pointer (not null!) for handle. Made all variables and functions used by BuildPropStaticFrustumCullMap searched dynamically (these should also work on s0, s1 and s2, they are compared but at the moment untested). TODO: Fix deref crash properly (see bsplib.cpp l291).
2022-05-01 05:38:51 +02:00
DetourAttach((LPVOID*)&v_BuildPropStaticFrustumCullMap, &BuildPropStaticFrustumCullMap);
Optimize some CHLClient functions away for dedicated
2022-05-04 21:10:46 +02:00
#endif // !DEDICATED
Code base refactor + major performance and readability improvement. Read description for details. * Codebase restructured to SourceSDK codebase style and .cpp/.h assertion paths in the game executable. * Document most functions with valve style 'Purpose' blocks. * Rename variables to match the rest of the codebase and Valve's naming convention. * Dedicated DLL and the SDKLauncher now share the same codebase as the DevSDK. * Obtain globals or pointers directly instead of waiting for runtime initialized data. * Dynamically search for all functions and globals (this doesn't count for dedicated yet!). * Initialize most in-SDK variables. * Move certain prints and other utilities under ConVars to reduce verbosity and increase performance. * Print all pattern scan results through a virtual function to make it easier to add and debug new patterns in the future. * Type global var pointers appropriately if class or type is known and implemented. * Forward declare 'CClient' class to avoid having 2 'g_pClient' copies. * Add IDA's pseudo definitions for easier prototyping with decompiled assembly code. * RPAK decompress Command callback implementation. * Load decompressed RPaks from 'paks\Win32\' overriding the ones in 'paks\Win64\' (the decompress callback will automatically fix the header and write it to 'paks\Win32\'). * VPK decompress Command callback implementation. * Move CRC32 ands Adler32 to implementation files. * Server will print out more details about the connecting client. * Upgrade ImGui lib to v1.86. * Don't compile id3dx.h for dedicated. * Don't compile id3dx.cpp for dedicated * Implement DevMsg print function allowing to print information to the in-game VGUI/RUI console overlay, ImGui console overlay and the external windows console * Fixed bug where the Error function would not properly terminate the process when an error is called. This caused access violations for critical/non-recoverable errors. * Fixed bug where the game would crash if the console or server browser was enabled while the game was still starting up. * Several bug fixes for the dedicated server (warning: dedicated is still considered work-in-progress!).
2021-12-25 22:36:38 +01:00
}
void BspLib_Detach()
{
Optimize some CHLClient functions away for dedicated
2022-05-04 21:10:46 +02:00
#ifndef DEDICATED
See description Fixed rare crash in CMDLCache when buffer itteration returns invalid pointer (not null!) for handle. Made all variables and functions used by BuildPropStaticFrustumCullMap searched dynamically (these should also work on s0, s1 and s2, they are compared but at the moment untested). TODO: Fix deref crash properly (see bsplib.cpp l291).
2022-05-01 05:38:51 +02:00
DetourDetach((LPVOID*)&v_BuildPropStaticFrustumCullMap, &BuildPropStaticFrustumCullMap);
Optimize some CHLClient functions away for dedicated
2022-05-04 21:10:46 +02:00
#endif // !DEDICATED
Code base refactor + major performance and readability improvement. Read description for details. * Codebase restructured to SourceSDK codebase style and .cpp/.h assertion paths in the game executable. * Document most functions with valve style 'Purpose' blocks. * Rename variables to match the rest of the codebase and Valve's naming convention. * Dedicated DLL and the SDKLauncher now share the same codebase as the DevSDK. * Obtain globals or pointers directly instead of waiting for runtime initialized data. * Dynamically search for all functions and globals (this doesn't count for dedicated yet!). * Initialize most in-SDK variables. * Move certain prints and other utilities under ConVars to reduce verbosity and increase performance. * Print all pattern scan results through a virtual function to make it easier to add and debug new patterns in the future. * Type global var pointers appropriately if class or type is known and implemented. * Forward declare 'CClient' class to avoid having 2 'g_pClient' copies. * Add IDA's pseudo definitions for easier prototyping with decompiled assembly code. * RPAK decompress Command callback implementation. * Load decompressed RPaks from 'paks\Win32\' overriding the ones in 'paks\Win64\' (the decompress callback will automatically fix the header and write it to 'paks\Win32\'). * VPK decompress Command callback implementation. * Move CRC32 ands Adler32 to implementation files. * Server will print out more details about the connecting client. * Upgrade ImGui lib to v1.86. * Don't compile id3dx.h for dedicated. * Don't compile id3dx.cpp for dedicated * Implement DevMsg print function allowing to print information to the in-game VGUI/RUI console overlay, ImGui console overlay and the external windows console * Fixed bug where the Error function would not properly terminate the process when an error is called. This caused access violations for critical/non-recoverable errors. * Fixed bug where the game would crash if the console or server browser was enabled while the game was still starting up. * Several bug fixes for the dedicated server (warning: dedicated is still considered work-in-progress!).
2021-12-25 22:36:38 +01:00
}
Reference in New Issue Copy Permalink