Commit graph

2272 commits

Author SHA1 Message Date
Nekotekina 8c28c4e8ec SPU: Make recompilers lock-free. 2019-10-26 23:33:18 +03:00
Nekotekina 9ac6ef6494 SPU: cleanup former OOM handling
Remove cpu_flag::jit_return.
It's obsolete now, and worked only in SPU ASMJIT anyway.
2019-10-26 21:24:12 +03:00
MSuih f3ed26e9db Small warnings cleanup (#6671)
* Ignore more warnings

These are intentional

* Signed/unsigned mismatch when comparing

* Explictly cast values

* Intentionally discard a nodiscard value

* Change ppu_tid to u32

* Do not use POSIX function name on Windows

* Qt: Use horizontalAdvance instead of width

* Change progress variables to u32
2019-10-25 13:32:21 +03:00
arabek 63a0b97dd1 Normalize audio when downmixing to avoid clipping (#6867)
* Normalize audio when downmixing to avoid clipping

Idea came from this topic:
https://hydrogenaud.io/index.php?topic=104214.msg855199#msg855199

Fixes very loud audio in Motorstorm (and probably other games
when playing over headphones/stereo speakers with
Downmix to Stereo option enabled)
2019-10-25 12:56:47 +03:00
Marin Baron 4599d58413 Fix debug build gcc/clang linker.
Create temporary variable to resolve undefined reference.

[ 98%] Linking CXX executable ../bin/rpcs3
/usr/bin/ld : Emu/librpcs3_emu.a(sys_vm.cpp.o) : dans la fonction « sys_vm_t::~sys_vm_t() » :
/mnt/raid/Git/rpcs3/rpcs3/Emu/Cell/lv2/sys_vm.cpp:21 : référence indéfinie vers « id_manager::id_traits<sys_vm_t, void>::invalid »
collect2: erreur: ld a retourné le statut de sortie 1
make[2]: *** [rpcs3/CMakeFiles/rpcs3.dir/build.make:426: bin/rpcs3] Error 1
2019-10-24 14:34:38 +03:00
JohnHolmesII 0842724f94 Update cellVoice.cpp
Caused warnings. Not sure what the actual intention was, this may need to be inverted.
This commit assumes that erase() returning 0 is a sign that deletion
failed, or that there was corruption. There should be a port there.
2019-10-24 14:19:52 +03:00
Nekotekina 6e19881b82 Update LLVM (10) 2019-10-23 16:01:14 +03:00
Emmanuel Gil Peyrot 9d1b7af2eb Avoid std::move() for copy elision
This could disable zero-copy optimisations, which is the opposite of
what we want.
2019-10-22 16:11:43 +03:00
Eladash 586fe11e22 Fix cellGcm HLE regression
Also correct flags.
2019-10-22 13:45:09 +03:00
Nekotekina b329bb604c SPU LLVM: implemented asynchronous compilation
Implemented interpreter-based pre-recompiler.
Interpreter functions are build with SPU LLVM.
2019-10-21 19:29:34 +03:00
Eladash cbecbc264e cellGcm: Fix driver_info->memory_size
based hw test
2019-10-21 15:31:45 +03:00
eladash 730e9cde84 sys_rsx: Improve allocations and error checks
* allow sys_rsx_device_map to be called twice: in this case the DEVICE address retrived from the previous call returned
* Add ENOMEM checks for sys_rsx_memory_allocate and sys_rsx_context_allocate
* add EINVAL check for sys_rsx_context_allocate if memory handle is not found
* Separate sys_rsx_device_map allocation from sys_rsx_context_allocate's
* Implement sys_rsx_memory_free; used by cellGcmInit upon failure
* Added context_id checks
* Throw if sys_rsx_context_allocate was called twice.
2019-10-21 15:31:45 +03:00
Eladash c16319f959 PPU Precise: Fix fused float ops misaccuracy by using fma
Also cleanup add64_flags, remove redundent copy of s_ppu_itype.
2019-10-20 20:58:31 +03:00
Eladash 79a3a7ce4c Add exception for sys_usbd_initialize 2019-10-20 15:28:26 +01:00
Eladash a1e7cdf1e9 Forcefully awake waiters in sys_usbd_finalize
This is a no-op from cellUsbd side but happens on lv2 side as well.
2019-10-20 15:28:26 +01:00
Eladash d90696be90 sys_usbd Tiny update 2019-10-20 15:28:26 +01:00
Eladash 6c2307fb96 cellVoice: Initial Implementation
Implement system service events, error checking and some functions.
2019-10-20 15:16:53 +01:00
Eladash fa73175754 sceNpBasicGetEvent error checks improved 2019-10-20 14:56:44 +01:00
Eladash d9a97d8674 sceNpUtil Improvements 2019-10-20 14:56:44 +01:00
Megamouse 0ef7ad129d cellNetCtl: improve param checks 2019-10-19 19:20:18 +02:00
Megamouse e6aeb7ecb9 stub cellKey2Char 2019-10-19 19:13:33 +02:00
Megamouse 0bad56da67 cellScreenshot: initial implementation 2019-10-17 21:24:25 +02:00
Megamouse 60f3ed7607 stub cellSsl 2019-10-17 20:32:33 +02:00
Nekotekina 00895862e1 Fix sys_vm_t destructor
It's not a destructor's job to free vm like this.
2019-10-17 19:57:29 +03:00
eladash 95752607ea sys_spu: Don't reset snr config at group_start()
Also first check for EINVAL in sys_spu_thread_set_spu_cfg
2019-10-16 21:11:29 +03:00
Eladash 6dd96d3db9 Fixes for cellUserInfoGetList
- Don't allow all args to be nullptr at once.
- Fill listBuf with zeroes for unwritten entries
- Fix userId set in listBuf

Similarly to what the firmware does
2019-10-16 21:11:29 +03:00
Eladash 47c3d945aa spu/ppu-diasm: Fixup for #6772, #6722 2019-10-16 21:11:29 +03:00
Nekotekina 0fe46934de Fix "SPU Interpreter (fast)"
Regression after #6770
2019-10-16 14:20:35 +03:00
Eladash 4ed2c7b594 error checking bugfix for sceNpTrophyCreateContext
options = 1 is allowed.
2019-10-16 00:50:58 +03:00
Eladash b03c2fc856 sys_cond_wait: Fix mutex acquire when signaling thread didnt pass ownership immediatly 2019-10-16 00:50:58 +03:00
Eladash ab02c56725 sceNpTrophyRegisterContext verbose logging in TROPHY_ERROR_ILLEGAL_UPDATE error
Also use safer checking when creating trophy ctxt dir
2019-10-16 00:50:58 +03:00
Eladash 80b4ac23e5 spu disasm: Show interrupts status modifiers on indirect branches
That's how the assembler represents it as well.
2019-10-16 00:50:58 +03:00
Nekotekina eafbc77c0d SPU LLVM: Always use linux-gnu target triple (affects Windows)
Unify internal code generation to make better use of GHC calling convention.
Ideally, it would just work on Windows as well, but some random bug appeared.
This bug was causing freezes on SPU LLVM compilation.
This commit desperately attempts to workaround it.
2019-10-15 18:04:50 +03:00
Nekotekina c69fe0f664 SPU Profiler preview
Add option "SPU Profiler" (disabled by default).
Works only with SPU recompilers.
Results are flushed on pausing.
2019-10-15 17:42:57 +03:00
Malcolm Jestadt 1a9e06d3c6 Linux: Change default Sleep Timers accuracy to host
- This doesn't change existing configs
- Also sets the host_min_quantum to the true value
- Restores lost TODO: comment
2019-10-15 00:32:31 +03:00
Eladash 7e08fff91d ppu disasm: Improve CR bits manipulation instructions formatting 2019-10-14 20:44:23 +03:00
Eladash 3910b2465e Fix unknown BCLR forms formatting 2019-10-14 20:44:23 +03:00
Eladash 75ee668922 Log fs::g_tls_error on ACCESS_ERROR errors 2019-10-14 20:44:23 +03:00
Nekotekina 076a80d71e Random space fixes 2019-10-14 19:42:35 +03:00
Eladash 1e303e9f97 ppu disasm: Implement BCLR's, BCCTR's and BC's extended mnemonics
Also:
BCCTR cr bit is now shown as cr(bit/4)[bit % 4]
BLRL is now shown properly
BDN?Z[TF](LR?)L?A? branches are now implemented.
2019-10-14 13:54:56 +01:00
Eladash 5978b1f28f ppu disasm: Fix AA BC formatting 2019-10-14 13:54:56 +01:00
Eladash 1d07c40aa5 Fix state reset in cellMic, cellAudio, cellCamera (#6761) 2019-10-13 19:26:11 +03:00
Nekotekina ac712bf4e3 lf_queue: remove explicit notify(), notify in push() instead
Should be cheap enough now if there are no waiters.
2019-10-12 22:45:17 +03:00
Nekotekina 5624b001ae lf_queue: add range-for support (endless loop with waiting) 2019-10-12 22:37:52 +03:00
Nekotekina 16dd72b3e3 Replace some uses of Emu.IsStopped()
Poll thread state instead.
2019-10-12 21:12:47 +03:00
Nekotekina 49e96b39dd [SPU, TSX] Fix reservation corruption in PUTLLC
Change reservation locking logic.
2019-10-12 15:41:24 +03:00
Megamouse 07022fd3b6 stub cellVoice 2019-10-12 10:16:01 +02:00
Eladash 0172e097f7 Fix cellCameraInit (camera_handler::null) 2019-10-11 21:55:10 +02:00
Eladash 5cc12ef22e cellCamera: Log error code 2019-10-11 21:55:10 +02:00
plappermaul 925f2ce02f Use Linux timers for sleeps up to 1ms (#6697)
* Use Linux timers for sleeps up to 1ms (v3)
The current sleep timer implementation basically offers two variants. Either
wait the specified time exactly with a condition variable (as host) or use a
combination of it with a thread yielding busy loop afterwards (usleep timer).

While the second one is very precise it consumes CPU loops for each wait call
below 50us. Games like Bomberman Ultra spam 30us waits and the emulator hogs
low power CPUs. Switching to host mode reduces CPU consumption but gives a
~50us penalty for each wait call. Thus extending all sleeps by a factor of
more than two.

The following bugfix tries to improve the system timer for Linux by using
Linux native timers for small wait calls below 1ms. This has two effects.

- Host wait setting has much less wait overhead
- usleep wait setting produces lower CPU overhead
2019-10-09 20:03:34 +03:00