Commit graph

18394 commits

Author SHA1 Message Date
Scott Mansell
368867dba0 Merge pull request #2922 from aserna3/SDBlock
Implemented ability to block writes to the SD card
2015-08-31 04:51:50 +12:00
Ryan Houdek
fe1c6cba36 Merge pull request #2809 from Sonicadvance1/aarch64_cycle_counter_profiling
[AArch64] Support profiling by cycle counters if they are available to EL0
2015-08-30 10:31:34 -05:00
Ryan Houdek
b907576510 [AArch64] Support profiling by cycle counters if they are available to EL0 2015-08-30 10:25:16 -05:00
Ryan Houdek
5110574c1f Merge pull request #2921 from Sonicadvance1/aarch64_optimize_lmw
[AArch64] Optimize lmw.
2015-08-30 10:23:57 -05:00
flacs
cc5eb6d00b Merge pull request #2927 from aserna3/workshop
Fixed introduced compiler warning in Linux
2015-08-30 05:47:33 +02:00
Anthony Serna
0390bd61df Fixed introduced compiler warning in Linux 2015-08-29 20:41:59 -07:00
flacs
83a179b56a Merge pull request #2925 from lioncash/wx-mem
MemcardManager: Header cleanup
2015-08-29 16:44:59 +02:00
Lioncash
d58550e874 MemcardManager: Minor cleanup of header code 2015-08-29 05:19:51 -04:00
Lioncash
0f3e4c50e1 MemcardManager: Correct class indentation 2015-08-29 05:13:20 -04:00
Lioncash
072150589e Merge pull request #2924 from lioncash/scope
Hash: Narrow define scope
2015-08-29 03:12:18 -04:00
Lioncash
e7c7dcaa1f Merge pull request #2923 from lioncash/override
Jit_Util: Add missing override specifiers
2015-08-29 03:12:11 -04:00
Lioncash
310bb46967 Hash: Narrow define scope 2015-08-29 02:57:35 -04:00
Markus Wick
a16669231a Merge pull request #2917 from Sonicadvance1/android_fix_sgs6
[Android] Workaround Mali driver issue on the Samsung Galaxy S6.
2015-08-29 08:56:32 +02:00
Lioncash
df19f11cb9 Jit_Util: Add missing override specifiers 2015-08-29 00:30:18 -04:00
Anthony Serna
db7fe9507e Implemented ability to block writes to the SD card
Renamed variable to be more accurate
2015-08-28 17:32:29 -07:00
Markus Wick
6004ecc521 Merge pull request #2920 from rohit-n/build-pch
Fix building with PCH disabled.
2015-08-28 23:08:24 +02:00
Ryan Houdek
8d61706440 [AArch64] Optimize lmw.
This instruction is fairly heavily used by Ikaruga to load a bunch of registers from the stack.
In particular at the start of the second stage is a block that takes up ~20% CPU time that includes a usage of lmw to load half of the guest
registers.

Basic thing optimized here is changing from a single 32bit LDR to potentially a single 128bit LDR.
a single 32bit LDR is fairly slow, so we can optimize a few ways.
If we have four or more registers to load, do a 64bit LDP in to two host registers, byteswap, and then move the high 32bits of the host registers in
to the correct mapped guest register locations.
If we have two registers to load then do a 32bit LDP which will load two guest registers in a single instruction.
and then if we have only one register left to load, load it as before.

This saves quite a bit of cycles since the Cortex-A57 and A72's LDR instruction takes a few cycles.

Each 32bit LDR takes 4 cycles latency, plus 1 cycle for post-index(which typically happens in parallel.
Both the 32bit and 64bit LDP take the same amount of latency.

So we are improving latencies and reducing code bloat here.
2015-08-28 14:40:30 -05:00
Ryan Houdek
2c3fa8da28 [AArch64] Fix a bug in the register caches.
This is a bug that crops if BindToRegister() is called multiple times in a row without a R() function call between them.
How to reproduce the bug:
1) Have a completely filled cache with no host register remaining
2) Call BindToRegister() with different guest registers
3) Don't call R() between the BindToRegister() calls.

This issue typically wouldn't be seen for a couple of reasons. Typically we have /plenty/ of registers in the cache, and in most cases we only call
BindToRegister() once per instruction. In the off chance that it is called multiple times, it wouldn't update the last used counts and would flush the
same register as the previous call to it.
2015-08-28 14:36:14 -05:00
Rohit Nirmal
6252d2d71a Fix building with PCH disabled. 2015-08-28 14:13:28 -05:00
Lioncash
a6bd2fea28 Merge pull request #2919 from lioncash/vec
Vec3: Remove a memset call on the this pointer.
2015-08-28 15:05:02 -04:00
Lioncash
e787501528 Vec3: Simplify operator== code 2015-08-28 14:46:40 -04:00
Markus Wick
b11de5bddb Merge pull request #2918 from lioncash/memcpy
DataReader: Get rid of pointer casts
2015-08-28 20:45:15 +02:00
Lioncash
d86d5fae9f Merge pull request #2909 from aserna3/DollsAndElves
Implemented .elf and .dol support in gamelist
2015-08-28 14:28:09 -04:00
Lioncash
bb27f80a65 Vec3: Remove a memset call on the this pointer 2015-08-28 14:10:07 -04:00
Anthony Serna
faedf1bc5c Implemented .elf and .dol support in gamelist
Fixed a TON of structuring, formatting.

removed README.txt files from themes at MaJoR's request

Added platform icon for ELFs/DOLs
2015-08-28 11:10:03 -07:00
Ryan Houdek
01db003779 [Android] Workaround Mali driver issue on the Samsung Galaxy S6.
Samsung updated the video drivers on the SGS6 which introduced a bug when disabling vsync.
Both the driver versions are r5p0, but the md5sums of the blob differ.
To work around the issue, make sure to never disable vsync by calling eglSwapInterval.

We can't actually determine the driver version on Android yet.
So until the driver version lands that displays the driver version string in the GL_VERSION string
we will need to keep this workaround enabled at all times, which is a bit annoying.

Current mali drivers return the video driver version in one of the EGL strings you can query.
The issue with that is that Android eats all of those strings, so we can't query it.
2015-08-28 09:02:46 -05:00
flacs
d373dd372d Merge pull request #2913 from Tilka/fix_warning_fix
AVIDump: fix -Wsign-compare warning
2015-08-27 23:50:34 +02:00
Lioncash
4fb3a8b78d DataReader: Get rid of pointer casts 2015-08-27 13:43:04 -04:00
Ryan Houdek
d9b18862f3 Merge pull request #2908 from Sonicadvance1/gles_3_2
Support OpenGL ES 3.2.
2015-08-26 18:19:17 -05:00
Ryan Houdek
447b1b09e3 Support OpenGL ES 3.2.
OpenGL ES 3.2 adds a few things we care about supporting in core. In particular:
- GL_{ARB,EXT,OES}_draw_elements_base_vertex
- KHR_Debug
- Sample Shading
- GL_{ARB,EXT,OES,NV}_copy_image
- Geometry shaders
- Geometry shader instancing (If they support GL_{EXT,OES}_geometry_point_size)

Nvidia was the first to release an OpenGL ES 3.2 driver which I uesd to test this on.
This also enables GS Instancing on GLES 3.1 hardware if it supports all of the required extensions.
2015-08-26 17:57:51 -05:00
Ryan Houdek
6d25c469cf Merge pull request #2915 from degasus/arm
JitArm64: Implement rlwnmx
2015-08-26 15:52:37 -05:00
Markus Wick
54f882704a Merge pull request #2914 from JosJuice/fix-volumedirectory
Fix VolumeDirectory
2015-08-26 22:12:23 +02:00
degasus
e516d4ef59 JitArm64: Implement rlwnmx 2015-08-26 21:59:10 +02:00
JosJuice
d276d1abbb Fix VolumeDirectory
Fixes the regression from a225426 and clarifies a related comment.
2015-08-26 19:21:09 +02:00
Markus Wick
3e9dac3910 Merge pull request #2810 from Sonicadvance1/disassembler_improv
Have the disassembler show the PC next to host instructions.
2015-08-26 17:01:39 +02:00
flacs
99e88a7af7 Merge pull request #2887 from Tilka/swap
Jit64: some byte-swapping changes
2015-08-26 16:43:45 +02:00
flacs
eb6ac641be Merge pull request #2906 from Tilka/fpscr
Jit64: fix bugs in the FPSCR instructions
2015-08-26 16:43:28 +02:00
Markus Wick
2be69ae747 Merge pull request #2912 from Tilka/coretiming
CoreTiming: small cleanup
2015-08-26 16:38:42 +02:00
Tillmann Karras
6ec4bdf862 CoreTiming: remove unused functions 2015-08-26 15:40:15 +02:00
Tillmann Karras
0f4861cac2 CoreTiming: make loops easier to read 2015-08-26 14:53:58 +02:00
Markus Wick
a36b135473 Merge pull request #2911 from Sonicadvance1/aarch64_regcache_optimization
[AArch64] Optimize paired registers being used in double operations.
2015-08-26 13:10:04 +02:00
Ryan Houdek
ca51f1a4f6 [AArch64] Optimize paired registers being used in double operations.
In particular this optimizes the case where a 32bit float is loaded via lfs, and then used in double operations.
This happens very often in Gekko based code because the best way to load a 32bit value as a double is lfs since it automatically turns in to a double value.

There are a few other implications of this in practice as well. Like if both of the paired registers are loaded via psq_l and then used in double
operations it would be improved.
Also if we implement a double register we've got to be careful to make sure we understand if it is in "lower" register or the full 128bit register.
2015-08-26 05:50:04 -05:00
Markus Wick
5716d18d10 Merge pull request #2910 from Sonicadvance1/aarch64_regcache_fix
[AArch64] Fix a bug in the register cache.
2015-08-26 08:31:24 +02:00
Ryan Houdek
4f5f29a0fb [AArch64] Fix a bug in the register cache.
If the register was only a lower pair and it needed the full register, then we need to load the high 64bits.
Which we weren't doing before.
2015-08-26 01:21:43 -05:00
Markus Wick
43d17cb360 Merge pull request #2904 from Sonicadvance1/aarch64_more_inst
[AArch64] Implement fdivx/fdivsx/mfcr/mtcrf.
2015-08-26 07:48:24 +02:00
Tillmann Karras
ee4a12ffe2 Jit64: some byte-swapping changes 2015-08-26 05:41:18 +02:00
flacs
6015e2d812 Merge pull request #2900 from aroulin/x64emitter-rcp
x64Emitter: add RCPPS and RCPSS SSE instructions
2015-08-26 05:05:53 +02:00
Ryan Houdek
6729a36d8d [AArch64] Set BindToRegister's to_load correctly for double FP ops. 2015-08-25 21:29:27 -05:00
flacs
9ba7497a3f Merge pull request #2907 from lioncash/log
GCMemcard: Clean up memcard logging messages.
2015-08-26 04:03:07 +02:00
Lioncash
db4f692482 GCMemcard: Clean up memcard logging messages. 2015-08-25 21:55:52 -04:00