about summary refs log tree commit diff stats
path: root/src/dynarec/dynarec_native_pass.c (follow)
Commit message (Collapse)AuthorAgeFilesLines
...
* [DYNAREC] Small improvment to Dynarec infrastructureptitSeb2024-06-041-1/+4
|
* [ARM64_DYNAREC] Fixed handling of YMM0 on forward jumpptitSeb2024-06-021-2/+2
|
* [ARM64_DYNAREC] Added AVX.66.F3A 19 opcodes and some various avx helper fixesptitSeb2024-05-301-0/+1
|
* [ARM64_DYNAREC] That first avx opcode now is 256bits enabledptitSeb2024-05-301-2/+2
|
* [ARM64_DYNAREC] Added a fisrt 128bits only AVX opcodeptitSeb2024-05-301-0/+2
|
* [DYNAREC] Small fixes to when to stop a block (fixes regression on EALauncher)ptitSeb2024-05-151-9/+8
|
* [DYNAREC] Stop a block when a unimplemented opcode is foundptitSeb2024-05-021-2/+2
|
* [COSIM] Added thread-safe tests (#1477)Yang Liu2024-04-301-3/+3
| | | | | | | | | | | | | * [COSIM] Added thread-safe tests * Added more notest * Run cosim in CI * Disable tests for RISC-V for now * Simplify * Initialize notest
* [DYNAREC] Change method to detect dead code in block (fixes Spintires on ↵ptitSeb2024-04-281-2/+2
| | | | Wow64, maybe some other too)
* [DYNAREC] More fixes on BIGBLOCKptitSeb2024-04-261-1/+1
|
* [DYNAREC] Stop building a block if opcode is garbbage or not redeable ↵ptitSeb2024-04-261-1/+8
| | | | (should help BIGBLOCK=2)
* Changed x87 way of handling FFREE opcode ([DYNAREC] too, improving x87 ↵ptitSeb2024-04-241-3/+6
| | | | robustness overall)
* [ARM64_DYNAREC] Better fix for #1366, that should help #1383)ptitSeb2024-04-021-0/+4
|
* [ARM64_DYNAREC] Fixed some issue with x87 stack managmentptitSeb2024-03-291-2/+3
|
* [ARM64_DYNAREC] More check to abort a block if x87 stack cache is incoherentptitSeb2024-03-261-1/+1
|
* [DYNAREC] Limit temporary memory allocation on FillBlock64ptitSeb2024-03-101-2/+5
|
* [DYNAREC] Try to fix some issue with x87 stack and skipped code being incoherentptitSeb2024-02-171-2/+3
|
* [DYNAREC] Tests for emited coded size limit (usefull for DYNAREC_TEST for ↵ptitSeb2024-02-071-0/+7
| | | | example)
* [ARM64_DYNAREC] More fixes to F0 0F C7 opcodeptitSeb2024-02-071-1/+1
|
* [ARM64_DYNAREC][32BITS] Small optim on jump table for 32bits access (1 less ↵ptitSeb2024-02-041-1/+1
| | | | read, or 2 in SAVE_MEM configuration)
* [DYNAREC] Fixed a warningptitSeb2024-01-181-1/+1
|
* [DYNAREC] Test when changing page on building a dynablock if it's safeptitSeb2024-01-181-0/+18
|
* [DYNAREC] Made callret default, improved callret efficiency, and dynarec ↵ptitSeb2023-12-131-6/+4
| | | | speed on larger blocks
* [DYNAREC] New attempt to make callret optim stable and usableptitSeb2023-12-121-9/+18
|
* [DYNAREC] Rollback most callret changes for now, they are not good enoughptitSeb2023-12-111-18/+9
|
* [DYNAREC] Try to improve stability of BOX64_DYNAREC_CALLRET=1 (not really ↵ptitSeb2023-12-111-9/+18
| | | | support on RV64 yet)
* Better way to handle multiple 3E/26 and F2/F3 prefixes ([DYNAREC] too)ptitSeb2023-12-071-6/+7
|
* [DYNAREC] Removed unused code in strongmem emulationptitSeb2023-12-051-1/+0
|
* [COSIM] Fix GO_TRACE (#1059)Yang Liu2023-11-121-2/+2
| | | x1 is the second parameter when calling native functions in GO_TRACE, so we cannot use it as a scratch register for xthead optimization.
* [ARM64_DYNAREC] More changes to Strong memory model emulationptitSeb2023-10-181-0/+1
|
* [RV64_DYNAREC] Added more support for XTheadMemPair extension (#995)Yang Liu2023-09-271-6/+6
| | | | | | | | | * Reformat * [RV64_DYNAREC] Added more support for XTheadMemPair extension * Align 'em up * Fix spec typos
* [DYNAREC] Some spelling fixesAndré Zwing2023-08-021-4/+4
|
* [DYNAREC] Improvments to BOX64_DYNAREC_FORWARD handlingptitSeb2023-07-231-7/+14
|
* [32BITS] added D6 undocumented opcodeptitSeb2023-07-231-2/+12
|
* Merge pull request #896 from josch/exeptitSeb2023-07-211-0/+0
|\ | | | | remove executable bits
| * remove executable bitsJohannes Schauer Marin Rodrigues2023-07-211-0/+0
| | | | | | | | Closes: #895
* | Improved BOX64_MALLOC_HACK=2 and the glAleternate mecanism (help steam & ↵ptitSeb2023-07-081-2/+2
| | | | | | | | heroic and other libcef based software)
* | A few cosmetic fixes (#858)Alexandre Julliard2023-06-241-1/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * [DYNAREC] Don't include pthread.h in C files. It's already included from box64context.h. Since there's no pthread.h on Win32, including it only once avoids having to add ifdefs everywhere. * [DYNAREC] Remove some duplicate definitions of the GETG helper macros. * Declare void functions with an explicit void. To avoid 'function declaration is not a prototype' warnings when using -Wstrict-prototypes. * Avoid including bridge_private.h in files that don't need it. * Avoid defining ZYDIS_RUNTIME_ADDRESS_NONE. It's not used, and conflicts with the original Zydis headers.
* | [32BITS][DYNAREC] Preparing work for 32bits dynarecptitSeb2023-06-231-8/+7
| |
* | [32BITS][DYNAREC] Preparing Dynarec to handle 32bits codeptitSeb2023-06-221-1/+4
|/
* [DYNAREC] More improvement to TEST_DYNARECptitSeb2023-03-311-0/+3
|
* [DYNAREC] Better handling of forward blocks with unused opcode in betweenptitSeb2023-03-301-2/+12
|
* [DYNAREC] Fixed handling of x87 top and stack on TEST_INTERPRETERptitSeb2023-03-301-0/+2
|
* [DYNAREC] Improved TEST_INTERPRETER with inter-block testingptitSeb2023-03-281-9/+3
|
* [DYNAREC] Added BOX64_DYNAREC_TEST to run interpeter and dynarec at the same ↵ptitSeb2023-03-271-2/+16
| | | | time and compare states
* [DYNAREC] Fixed rolling back for failled forward blockptitSeb2023-03-261-1/+1
|
* [DYNAREC] Some fixes and small improvment to common codeblock generatorptitSeb2023-03-251-6/+20
|
* [RV64_DYNAREC] Small optim on last_ipptitSeb2023-03-181-1/+1
|
* [ARM64_DYNAREC] Fixed some special cases for the Double Pop optim (fixed ↵ptitSeb2023-03-181-0/+1
| | | | steamwebhelper and all chrome/libcef based apps)
* [DYNAREC] Fixed missing X_PEND need at end of block for unimplemented opcodeptitSeb2023-03-131-1/+13
|