| Rev | Author | Branch | Log Message | Date | SHA |
|---|---|---|---|---|---|
| r25986 | master | Fix textures crash when restarting resource |
5 December, 2025 | ||
| master | Fix serveer crash |
5 December, 2025 | |||
| master | Visual Studio Update Build Tools 2022: 17.14.36717.8 Build Tools 2026: 18.0.11222.15 ... |
5 December, 2025 | |||
| master | New Crowdin updates (PR #4581) |
5 December, 2025 | |||
| r25982 | master | Force rebuild |
4 December, 2025 | ||
| master | Addendum #2 to 61c5d9e |
4 December, 2025 | |||
| master | Addendum to 61c5d9e |
4 December, 2025 | |||
| master | Update client en_US pot [ci skip] |
4 December, 2025 | |||
| master | Fix build error in certain configurations (custom builds) #2 |
4 December, 2025 | |||
| master | Update client en_US pot [ci skip] |
4 December, 2025 | |||
| master | Fix build error in certain configurations (custom builds) |
4 December, 2025 | |||
| r25975 | master | Update CEF to 142.0.17+g60aac24+chromium-142.0.7444.176 This brings some stability fixes in CEF itself. |
3 December, 2025 | ||
| master | Visual Studio Update Build Tools 2022: 17.14.36717.8 Build Tools 2026: 18.0.11222.15 ... |
3 December, 2025 | |||
| r25973 | master | Crash handler: Always log registers (creating a full .dmp can fail) |
3 December, 2025 | ||
| master | Fix SA crash at 0x00897C28 (dump name wrongly identifies it as 0x003F0090), likely 'caused' by VS2026 but a latent stack offsets bug since its introduction (736660bf2). Theory: VS2026 changed changed how inline assembly works. Probably something with different stack alignment, optimization choices, register allocation, calling convention optimizations, or code generation. (This can all be relevant to spot similar bugs manifesting now across hooks). The broken, now patched, assembly code was reading stack values from wrong offsets, and while the old compiler's code generation might have accidentally put "harmless" values at those wrong locations, the new compiler's different stack layout or register usage puts critical values (like 0xC97C28 - the address of _RwD3DDevice) at those locations. ... |
3 December, 2025 | |||
| master | Update client en_US pot [ci skip] |
3 December, 2025 | |||
| master | Add ability to remove all domains (#4465) • Add ability to remove all domains • Fix conflict mistakes & improve button styling to match the others ... |
3 December, 2025 | |||
| r25969 | master | Fix memory waste in CProxyDirect3DVertexBuffer fallback • Remove unnecessary pre-allocation of fallback storage in constructor • Release fallback storage after successful unlock instead of holding indefinitely ... |
2 December, 2025 | ||
| master | Update client en_US pot [ci skip] |
2 December, 2025 | |||
| master | Addendum to ae0f075 & ba6efb5 |
2 December, 2025 | |||
| master | Fix old, top 5 SA crash @ 0x001D85AA (0x5D85AA) |
2 December, 2025 | |||
| master | Add PostFX Lua functions (#4587) |
2 December, 2025 | |||
| master | Fix server list related crash |
2 December, 2025 | |||
| master | Addendum to c19c949 to cover decodeString in the fix reliability revisions as well (similarly to ae0f075 that covered teaDecode better) |
2 December, 2025 | |||
| master | Addendum to 5d37348 ("Also cover teaEncode case (besides c19c949 that covers encodeString), as both represent the same crash trend and observations") to make it work as expected/at all |
2 December, 2025 | |||
| r25961 | master | 1.7: Better texture-related error handling and debugscript return values |
2 December, 2025 | ||
| master | Update client en_US pot [ci skip] |
2 December, 2025 | |||
| master | Prevent a class of cross-reconnect texture leaks |
2 December, 2025 | |||
| r25958 | master | Fix issues with texture shader loading (.fx) through addressing a couple of underlying bugs. Has more benefits. |
2 December, 2025 | ||
| master | Fix `getVehiclePaintjob` returning -1 when the game engine hasn't applied a paintjob yet during stream-in (#4585) Fixes #2227 |
2 December, 2025 | |||
| r25956 | master | Fix 1.7 streaming sectors crash when moving around a total replacement map (like VC) that presents itself as if in CShadows::CastShadowSectorList but it wasn't that straightforward.. 1. 0x6C65746E - "ntel" (corrupted pointer: executing string data) 2. 0x70A6DC` CShadows::CastShadowSectorList` ... |
2 December, 2025 | ||
| r25955 | master | Addendum to last |
2 December, 2025 | ||
| master | Addendum #2 to a235955 (and 1b40002) |
2 December, 2025 | |||
| r25953 | master | Addendum to a235955 (Actual crash fixes) |
2 December, 2025 | ||
| master | Also cover teaEncode case (besides c19c949 that covers encodeString), as both represent the same crash trend and observations. |
1 December, 2025 | |||
| master | Addendum to c19c949 |
1 December, 2025 | |||
| master | Try to avert #1 crash from model encryption system-memory abuse by (mostly pt-BR) servers, assuming it's only allocation peaks and that it can be retried later when allocation/mem pressure is lower. The use of decodeString with huge data payloads (or spammed requests) leads to the crash, that usually presents as KERNELBASE.dll 0x0013b702 in MTA as of writing. Servers should actually batch decode requests instead of abusing memory like this, but they turned to blame us instead. |
1 December, 2025 | |||
| master | Convert SString to std:: for JPEG handling code in 1.7 |
1 December, 2025 | |||
| master | Further improve JPEG implementation and error handling. Also addendum to 0f62917 & ce73b73 |
1 December, 2025 | |||
| master | Fix `JpegDecode` crash (#4589) |
1 December, 2025 | |||
| master | Visual Studio Update Build Tools 2022: 17.14.36717.8 Build Tools 2026: 18.0.11222.15 ... |
1 December, 2025 |