about summary refs log tree commit diff stats
Commit message (Collapse)AuthorAgeFilesLines
...
* [DYNAREC] Fixed a potential issue where defered flags are not computed/reset ↵ptitSeb2025-07-024-4/+5
| | | | on internal jump
* [TRACE] Fixed the jump from instruction idx for Extend blcok in traceptitSeb2025-07-021-1/+1
|
* Simplified a bit ROLLING_LOG buffer handlingptitSeb2025-07-015-23/+15
|
* [ARM64_DYNAREC] Optimized propate XMM/YMM unused helpersptitSeb2025-07-011-26/+11
|
* [ARM64_DYNAREC] Fixed a regression introduced with ↵ptitSeb2025-06-301-1/+1
| | | | 4903177bab1f3324a0faeedd968fed5bf4ea8772 (for ForzaHorizon4, maybe other too)
* [DYNAREC] Refactor hotpage detection and dynarec_dirty 1 & 2. Also adjust ↵ptitSeb2025-06-308-52/+256
| | | | some launcher/games flags
* [ARM64_DYNAREC] Added 66 8D opcodeptitSeb2025-06-291-1/+16
|
* [ARM64_DYNAREC] Added 67 9C opcodeptitSeb2025-06-291-0/+6
|
* [ARM64_DYNAREC] Made (V)M(IN/AX)P(D/S) opcodes always exact, it's ↵ptitSeb2025-06-293-38/+16
| | | | inexpensive and make FASTNAN=0 less usefull
* [INTERP] Added 67 9C opcode (for LEGO2Kdrive)ptitSeb2025-06-281-0/+6
|
* [INTERP] Small imrpvoment for 0F 03 opcodeptitSeb2025-06-281-1/+1
|
* [INTERP] Added 0F 03 opcode (for LEGO2Kdrive)ptitSeb2025-06-281-0/+14
|
* [ARM64_DYNAREC] Use Unaligned sepcial handling to handle rare case of 32bits ↵ptitSeb2025-06-281-4/+21
| | | | access out of 32bits bounds (for tasty platnet 2 on steam)
* [LA64_DYNAREC] Removed some TABLE64 usage (#2782)Yang Liu2025-06-271-21/+12
|
* [LA64_DYNAREC] Add la64 avx load/store ops part 4. (#2775)phorcys2025-06-271-0/+252
| | | * VEX.66.0f.38 VMOVSX{BW,BD,BQ,WD,WQ,DQ}/VMOVZX{BW,BD,BQ,WD,WQ,DQ}
* [LA64_DYNAREC] Optimized GETIP macro (#2781)Yang Liu2025-06-276-51/+49
|
* [LA64_DYNAREC] Removed some redundant macro definitions (#2778)Yang Liu2025-06-261-2/+0
|
* [DYNACACHE][LA64] More work on internal reloc (#2779)Yang Liu2025-06-265-13/+23
|
* [LA64_DYNAREC] Add la64 avx load/store ops part 3. (#2774)phorcys2025-06-266-10/+416
| | | | | * VEX.0f VMOVMSKPS * VEX.66.0f VMOVMSKPD/VPMOVMSKB/VMASKMOVDQU * VEX.66.0f.38 VMASKMOVPS/VMASKMOVPD/VPMASKMOVD/VPMASKMOVQ
* [LA64_DYNAREC] Add la64 avx load/store ops part 2. (#2773)phorcys2025-06-262-0/+113
| | | | * VEX.0f VMOVLPS/VMOVHPS/VMOVLHPS/VMOVHLPS * VEX.66.0f VMOVLPD/VMOVHPD
* [ARM64_DYNAREC] Fixed opcode name for VDIVPDptitSeb2025-06-251-1/+1
|
* [ARM64_DYNAREC] Fixed some extended instance of VCMPSD opcodesptitSeb2025-06-251-3/+3
|
* [INTERP] Improved 32bits to 16bits float conversionptitSeb2025-06-251-6/+21
|
* [INTERP] More work on UD flagsptitSeb2025-06-251-5/+17
|
* [INTERP] More work on UD flagsptitSeb2025-06-251-6/+11
|
* [ARM64_DYNAREC] Adjusted some UD flags in BLSMSK opcodeptitSeb2025-06-251-4/+14
|
* [INTERP] Adjusted some UD flags in BLSMSK opcodeptitSeb2025-06-251-4/+18
|
* [ARM64_DYNAREC] Fixed CF flag of BLSI opcodeptitSeb2025-06-251-1/+1
|
* [INTERP] Fixed CF flag of BLSI opcodeptitSeb2025-06-251-1/+1
|
* [ARM64_DYNAREC] Fixed BEXTR opcodeptitSeb2025-06-251-11/+13
|
* [INTERP] Added 66 F0 F7 /2 opcode (aligned only)ptitSeb2025-06-242-1/+36
|
* [ARM64_DYNAREC] Improved BTx opcodes (and fixed one BTC opcode)ptitSeb2025-06-241-41/+141
|
* [INTERP] Added F0 BB and improved F0 BA /7 opcodesptitSeb2025-06-241-9/+94
|
* [DYNAREC] Better check of limit for a dynablockptitSeb2025-06-242-2/+2
|
* [WRAPPER] additional wrapped symbols (#2765)airidosas2522025-06-242-0/+16
| | | | | | | | | | | * [WRAPPER] Wrap libiconv library Seems to be required with newer versions of Wine for whatever reason. * [WRAPPER] Wrap _ctype_ symbol (ANDROID only) For whatever reason, this symbol is used on bionic at least. Not quite sure if it's the same on Linux. * [WRAPPER} Fix a signature typo for libiconv function
* [LA64_DYNAREC] Add la64 avx load/store ops part 1. (#2766)phorcys2025-06-245-12/+194
| | | | | | | * VEX.0f VMOVNTPS/VLDMXCSR/VSTMXCSR * VEX.66.0f VMOVNTPD/VMOVD/VMOVNTDQ * VEX.66.0f.3a VMOVNTDQA * VEX.f2.0f VMOVDDUP/VLDDQU * VEX.f3.0f VMOVSLDUP/VMOVSHDUP/VMOVD
* [ARM64_DYNAREC] Improved handling of last_ipptitSeb2025-06-245-2/+5
|
* [DYNACACHE][LA64] Added const table for later use in internal relocation (#2770)Yang Liu2025-06-2413-69/+318
|
* [LA64_DYNAREC] This should help certain builds (for #2769)ptitSeb2025-06-242-5/+10
|
* [CI] Upgraded QEMU and loongarch64 toolchains (#2768)Yang Liu2025-06-241-4/+6
|
* [DYNAREC] Removed some unused code (#2767)Yang Liu2025-06-246-26/+0
|
* [LA64_DYNAREC]Add basic avx support for la64. (#2745)phorcys2025-06-2316-70/+5585
| | | | | | | | | * basic infra for avx * some basic ops for avx VMOVDQU/VMOVDQA/VMOVUPS/VMOVAPS/VMOVUPD/VMOVAPD VZEROUPPER/VZEROALL VMOVD/VMOVSD/VMOVSS VINSERTF128/VINSERTI128/VEXTRACTF128/VEXTRACTI128 VBROADCASTSS/VBROADCASTSD/VBROADCASTF128
* Fix wowbox64 buildptitSeb2025-06-231-0/+12
|
* [DYNACACHE] Added support for unaligned addressesptitSeb2025-06-232-4/+49
|
* [ARM64_DYNAREC] Removed commented codeptitSeb2025-06-231-1/+0
|
* Add `map64_customMalloc` (#2764)Chi-Kuan Chiu2025-06-231-6/+140
| | | | | | | | | | | | | Most allocations target ~56 bytes, causing internal fragmentation in `map128_customMalloc`. Introduce `map64_customMalloc`, a 64-byte–chunk variant of `map128_customMalloc`. The new allocator uses a `uint16_t` “map” array: each bit in a 16-bit entry tracks the in-use status of one 64-byte chunk (16 chunks per map entry). The bitmap region is fixed at 256 bytes. Note that “map” here is a fast-lookup structure, distinct from the bitmap itself. All chunk- size, shift, and map-size calculations have been updated accordingly. Relates to #2740
* [DYNAREC] Refactored a bit BARRIER_FLOAT ([ARM64] olny for now, todo for ↵ptitSeb2025-06-2316-42/+118
| | | | RV64 and LA64)
* [ARM64_DYNAREC] Small fix in arch_buildptitSeb2025-06-221-1/+1
|
* [DYNACACHE] Introduced B64X_DYNACACHE=2 to use cache but not create new ↵ptitSeb2025-06-205-15/+19
| | | | ones, and made it default
* [DYNACACHE] Fixed a WarningptitSeb2025-06-201-1/+1
|