Commit Graph

2470 Commits

Author SHA1 Message Date
Gregory Hainaut 9c84712640 gsdx hw: enable palette read back on Jak2 2016-10-01 18:47:21 +02:00
Gregory Hainaut 5d93a392e4 gsdx hw: xman half screen correction 2016-09-30 22:25:24 +02:00
Gregory Hainaut 29c25bc375 gsdx clut: disable some assertions often fired 2016-09-30 22:24:38 +02:00
Gregory Hainaut 34686394a6 gsdx hw: add an hack for Jak3
I'm disappointed...

Read back palette written by the GPU. Avoid strange color on the skin.
2016-09-30 19:22:16 +02:00
Gregory Hainaut df2f8684d6 gsdx: improve TC cache log 2016-09-30 19:18:58 +02:00
Gregory Hainaut 8d24aa2b39 gsdx: GoW2 half screen bug
... Seriously ...
2016-09-30 19:15:25 +02:00
Gregory Hainaut 8d4fadb3c8 Revert "gsdx state: TEX0 write behavior might depends on the context"
This reverts commit b34243fa7b.

Unfortunately we have a regression on GT4. I'm doomed
2016-09-28 20:00:56 +02:00
Gregory Hainaut 90ad9cf668 gsdx hw mipmap: tune the rounding + remove previous hack
GS really uses an invalid texture located at 0.

Improve the rounding for R&C. The idea is to avoid the corner case were only
the corner of the triangle touch the 7/16 edge.
2016-09-28 19:41:12 +02:00
Gregory Hainaut adb8be3287 gsdx: add/update comment
Need to investigate why the number of vertex is different in upscaled resolution
2016-09-28 19:41:12 +02:00
Gregory Hainaut 749037346c gsdx: report an error if PSM is invalid 2016-09-28 19:41:12 +02:00
Gregory Hainaut fea31f7422 gsdx: don't use virtual when it is useless 2016-09-28 19:41:11 +02:00
Gregory Hainaut 97172f5e51 gsdx debug: count draw call in a different way
* Always do +1 before the draw call
* Prefix texture name with i (as input) to keep them before the FB

Goal is to ensure that all renderers share the same draw call value.
2016-09-28 19:41:11 +02:00
Gregory Hainaut b34243fa7b gsdx state: TEX0 write behavior might depends on the context
Game: harley davidson
* write tex0 ctx0
* write tex0 ctx1
* draw  ctx 0

Previous GSdx behavior will load the clut every write of TEX0. In the
above case the draw will take the wrong clut.

To be honest, it could be a wrong emulation on the EE core emulation.
The hardware likely got a single clut (1KB cache is quite expensive)
So clut loading must be skipped if the context is wrong.

Next draw will use the ctx1 clut so I apply TEX0 when the context is switched

Please test harley davidson :)

v2: detect context switch from UpdateContext function
V3: always set m_env.CTXT[i].offset.tex, avoid crash (Thanks to FlatOutPS2 that spot the issue)
V4: move bad psm correction code (rebase put it in the wrong place)
2016-09-27 10:04:09 +02:00
Gregory Hainaut e0961d81a0 gsdx HW: use after free crash
Free bt
3 0xe676d194 in ~Source ../plugins/GSdx/GSTextureCache.cpp:1526
4 0xe676d194 in GSTextureCache::SourceMap::RemoveAt(GSTextureCache::Source*) ../plugins/GSdx/GSTextureCache.cpp:1990
5 0xe676f0fe in GSTextureCache::IncAge() ../plugins/GSdx/GSTextureCache.cpp:1022

Use bt
0 0xe6772a83 in GSTextureCache::LookupSource(GIFRegTEX0 const&, GIFRegTEXA const&, GSVector4i const&) ../plugins/GSdx/GSTextureCache.cpp:204
1 0xe66b0c9f in GSRendererHW::Draw() ../plugins/GSdx/GSRendererHW.cpp:579
2 0xe66fb43e in GSState::FlushPrim() ../plugins/GSdx/GSState.cpp:1509

Hypothesis the m_map array of list contains an invalid pointer
It is populated GSTextureCache::SourceMap::Add based on the coverage. The coverage is based on the offset.
So offset is potentially wrong. As mipmap code hack the offset value. It would be a nice culprit.

This commit avoids a potential bad transition between MIPMAP (which
overwrite the "offset") and the base layer (which wrongly keep an old "offset")

Conclusion, pray for my soul as it is very hard to reproduce
2016-09-26 20:34:38 +02:00
Gregory Hainaut 04d82d8127 gsdx: size doesn't count but 0.625 might change your life :p 2016-09-26 19:42:07 +02:00
Gregory Hainaut be9995bd1e gsdx mipmap: improve robustness to avoid potential crash 2016-09-26 09:27:03 +02:00
Gregory Hainaut ebb2a8ad5a gsdx debug: Add DISABLE_WIP_ASSERTION
To easily disable assertion on codes that only require investigation
2016-09-25 20:00:51 +02:00
Gregory Hainaut ca4f265ce6 gsdx debug: dump mipmap register could be useful now :) 2016-09-25 20:00:51 +02:00
Gregory Hainaut ef9a144437 gsdx hw: extra trick for the mipmap
Ratchet & Clank (the third) uses an address of 0 for invalid mipmap.

It would be very awkward to put the middle layer of texture in start of
memory. So let's use this information to correct the lod.

It make the game more robust on the lod rounding
2016-09-25 11:54:00 +02:00
Gregory Hainaut 78b2848622 gsdx: factorize the TEX0 generation for mipmap 2016-09-25 11:51:48 +02:00
Gregory Hainaut ccd9ce1728 gsdx hack: move tales of legendia hack to full move
Also remove the OI hack which is handled by the generic double half clear
2016-09-25 10:17:08 +02:00
Gregory Hainaut e229dc1549 gsdx hack: move Resident Evil 4 to full level
The CRC hack removes the fog, without the hack the screen turns green with D3D,, but the fog emulates correctly on OGL with hardware depth.
2016-09-25 10:13:21 +02:00
Gregory Hainaut ad8a84a532 gsdx hw: reduce hack of Spyro series
Based on the comment, it ought to be fixed with previous commit.
2016-09-25 10:08:55 +02:00
Gregory Hainaut 7f871b0cc6 gsdx hw: remove various CRC hack that ought to be fixed with previous commit
* SpidermanWoS
* TyTasmanianTiger
* TyTasmanianTiger2
* DigimonRumbleArena2
* XmenOriginsWolverine
* CallofDutyFinalFronts
* GoW2 (only in openGL)
2016-09-25 10:08:54 +02:00
Gregory Hainaut 1e343084e2 gsdx hw: group the double half clear with gs mem clear hack
It avoid redundant check of the GS context

You can disable the extra clear with UserHacks_DisableGsMemClear = 1
2016-09-25 10:08:54 +02:00
Gregory Hainaut 6c6ed5f443 gsdx ogl: trigger the special RW framebuffer shader when only Alpha is written (#1588)
* gsdx ogl: trigger the special RW framebuffer shader when only Alpha is written

It will work for both Jak and Ratchet&Clank and without CRC
2016-09-25 10:05:44 +02:00
Gregory Hainaut 6e2bf98d5f gsdx hw: round mipmap lod value
Will avoid some corruption based on distance/angle
2016-09-24 20:05:39 +02:00
Gregory Hainaut 719982e4a3 gsdx hw: forget to update a parameter
OMG, fixes various mipmap issues
2016-09-24 18:57:47 +02:00
Gregory Hainaut 76494e5f9c gsdx hw: add an userhack for hw mipmap 2016-09-24 12:39:31 +02:00
Gregory Hainaut e38aba0dff gsdx: rename SW texture dump filename
* Use texraw for the unconverted texture (keep index fmt)
  => avoid bad filename order with the multiple texture layers
* add the real mipmap address
* Use a nice string format
2016-09-24 09:15:40 +02:00
Gregory Hainaut d89b09427f gsdx: gl logging is only available in MTGS thread
Fix crash in debug build
2016-09-22 17:54:44 +02:00
Gregory Hainaut b00ae974d2 gsdx ogl: fix bad interaction between accurate blending and geometry shader
Recent optimization allow to use triangle to render sprite even when gs is supported

close #1574
2016-09-20 09:30:24 +02:00
Gregory Hainaut b91bd3ea01 gsdx dx/gl: fix potential bug in alpha test handling
Depth test must remain disabled if write were masked.

Bug was detected on ICO (sea/sky) but the previous commit removes
the wrong ATE draw call.
2016-09-19 08:56:49 +02:00
Gregory Hainaut 6966e08306 gsdx: move alpha test optimization in base class
One code for all renderers :)
2016-09-19 08:48:32 +02:00
Gregory Hainaut 38b77397e0 gsdx hw: remove custom Ty Tasmanian Tiger / Simposons OI hack
Replaced by the generic OI_DoubleHalfClear. If OI_DoubleHalfClear is fast/safe enough, it
will enabled by default in the future.
2016-09-18 18:50:16 +02:00
Gregory Hainaut 6558317a7e gsdx: s/m_eq.xzyf & 4/m_eq.z/
Much better this way
2016-09-18 18:50:15 +02:00
Gregory Hainaut 7a40b27931 gsdx CRC: move GTA hack to full level
You can now emulate the radiosity effect with the UserHacks_AutoFlush hack.
2016-09-18 18:50:15 +02:00
Gregory Hainaut 0f387a8e3c Merge pull request #1535 from PCSX2/clang-tidy-macro-parenthesis
clang-tidy: use parenthesis around macro parameters
2016-09-18 18:48:04 +02:00
Gregory Hainaut 310f13a2f7 gsdx ogl: only use geometry shader to convert big enough draw call (#1554)
gsdx ogl: only use geometry shader to convert big enough draw call

The purpose of geometry shader is to reduce bandwidth (72 bytes by sprite)
and CPU load.

Unfortunately it increases CPU load due to extra shader validations.

So geometry shader will only be enabled for draw call with more than
16 sprites (arbitrarily, smallest number before shadow hearts plummet)

v2: don't disable geometry shader in replayer.
It is easier to spot sprite rendering and to manually read vertex info.
2016-09-18 16:58:02 +02:00
Gregory Hainaut 2a60264e40 gsdx replayer: improve error message when file is corrupted 2016-09-18 16:52:35 +02:00
Gregory Hainaut 6e6eae7844 pcsx2:gsdx:spu2x: use parenthesis around macro parameters
For safety reasons.
2016-09-18 16:13:55 +02:00
Gregory Hainaut b1d446895b GSdx: replace IsTopLeftAligned macro with a function 2016-09-18 16:13:28 +02:00
Gregory Hainaut 5ef2d538a0 gsdx hw: only keep a single half clear implementation 2016-09-18 13:16:59 +02:00
Gregory Hainaut a0c1719695 gsdx hw: DoC takes 3
Add support for 16 bits format too. Fix shadows.
2016-09-17 12:06:43 +02:00
Gregory Hainaut ebdf54a60b gsdx sw: don't enable auto flush hack in replayer
Otherwise we don't have same number of draw call in HW/SW which is
awful to compare
2016-09-17 12:05:05 +02:00
Gregory Hainaut 3546d54e3b gsdx hw: improve the vertical half clear for DoC
Previous assumption that crazyness was used to clear the buffer is wrong.

It was also used to write a constant color to the buffer.
2016-09-17 10:34:35 +02:00
Jonathan Li 4d5e051387 gsdx: Fix missing "\n" in rc file
It stops Visual Studio from asking to normalise the line endings
everytime the resource editor is used.
2016-09-16 23:42:54 +01:00
Jonathan Li 91b2e5bdbb Merge pull request #1564 from FlatOutPS2/master
GSdx: Add Windows UI options for HW hacks
2016-09-16 23:35:45 +01:00
Gregory Hainaut 7b9654f164 gsdx hw: CRC based solution to solve half screen issue on FFVII Dirge Of Cerberus
The trick is that the depth buffer is actually the right of the frame buffer
2016-09-17 00:02:32 +02:00
FlatOutPS2 55b0be36cd GSdx: Add cancel button to Hacks dialog
Add a cancel button to the Hacks dialog to make it easier for the user
to exit the window without making changes.
2016-09-16 23:22:13 +02:00
FlatOutPS2 5420fcaf3a GSdx: Add Windows UI options for HW hacks
Add UI options for Auto Flush and Unscale Point Line HW hacks.
2016-09-16 23:19:05 +02:00
Gregory Hainaut 3994141d40 gsdx sw: always enable the auto flush hack on the SW renderer
There are enough option. Will remove it if too slow.

Positively impact Jak, GTA, Battlefield 2
2016-09-15 17:24:40 +02:00
Gregory Hainaut 1a1d6ab478 gsdx: handle the invalid format 0x3
International Super Star Soccer seems to use it.

It could be a core bug, in doubt uses the likely expected standard 0x13
2016-09-14 19:22:45 +02:00
Gregory Hainaut 4796803c33 pcsx2: Remove == true/false for boolean logic (#1556)
As discussed in #1553

Clang Tidy reports goes from 156 to 9.

Remain some macro in spu2x + a deadcode line
2016-09-10 20:08:14 +02:00
Jonathan Li 72d3e4e03f windows: Add freetype to solution and link to GSdx 2016-09-10 12:36:56 +01:00
Gregory Hainaut ccf198a80b gsdx ogl: fix a regression on Okami
Strangely the game uses large texture to handle texture buffer.

I think it plays with WMS/WMT. I'm not sure texture shuffling is 100%
correct here. But without it, it's completely broken.
2016-09-09 21:51:23 +02:00
Gregory Hainaut 8c1722faa2 gsdx: add a note that Jak shadows can be handled with HW GL without auto flush hack 2016-09-09 21:08:15 +02:00
Gregory Hainaut 64a97483fb gsdx: add Jak3 US too 2016-09-09 21:05:00 +02:00
Gregory Hainaut de480ec0f5 gsdx: add crc hack for Jak
* 0x9C712FF0, Jak1, EU
* 0x472E7699, Jak1, US
* 0x2479F4A9, Jak2, EU
* 0x12804727, Jak3, EU
* 0xDF659E77, JakX, EU

Please report me the CRC of the US version too so I can add them.

Please test the shadows rendering (openGL HW + accurate blending at least basic)
2016-09-09 19:44:04 +02:00
Gregory Hainaut d27d2dd520 fix compilation issue 2016-09-09 00:11:02 +02:00
Gregory Hainaut c2b67ccb78 gsdx ogl: implement an alternate shader for Jak Shadows
The game sets the framebuffer as an input texture. So I did the same for
openGL. Code is protected with a CRC. It is working because the game want to sample
pixels.

For the record, I tested it GTA too, it doesn't work as expected because
the game will resize the framebuffer to a smaller one. So you don't have
the guarantee that pixel will be read before a data write.

Note: it requires at least accurate blending set on basic

Note: I need CRC of all Jak games that suffers of this issue. Thanks you :)
2016-09-08 23:58:13 +02:00
Gregory Hainaut f9d8cb97c0 Merge pull request #1508 from PCSX2/gsdx-allow-sso-amd
Finally AMD delivered the fix to support separate shader object !!!
2016-09-08 21:47:32 +02:00
Gregory Hainaut 6ee841319f gsdx: help stupid compiler to allow 32 byte move
Default copy-constructor is eight 32 bits move

GSRendererOGL::Lines2Sprites code shrinks from 510B to 398B
(loop of the function 296B => 181B). Hopefully it will reduce the cost
to convert line to sprite on the CPU (i.e. when geometry shader is disabled)
2016-09-08 20:01:05 +02:00
Gregory Hainaut aba0c733cd gsdx linux: add a gui check box to control previous hack 2016-09-08 17:34:28 +02:00
Gregory Hainaut 029d74f855 gsdx: add an auto flush hack to implement a Read Write GS effect.
It impacts all renderers. It ought to fix issue in GTA radiosity,
Shadows in Jak series. (note shadows will suck in upscaling)

Implementation is really brutal. Expect a massive slow down, but at least we can test the effect
easily.

Normally perf impact will remain reasonable if the game doesn't use a Read-Write effect

Performances number are welcomes (my guess is really awful in HW mode, slow in SW mode).
You can enable it with "UserHacks_AutoFlush = 1"
2016-09-08 17:25:59 +02:00
Gregory Hainaut bcb951077d gsdx: c_str() is useless when parameter is a string
Reported by clang tidy
2016-09-07 22:10:50 +02:00
Gregory Hainaut 145d6e29c2 gsdx: increase the allocation of the GS memory
GS memory is only 4MB but rendering is allowed to be 2048x2048
with 32 bits format (so 16MB). Technically the frame/depth buffer can start
at the end of the GS memory. Let's not waste too much memory.

Fix crash with BASARAX

(game draws a 2048x1664 32 bits area)
2016-09-07 08:34:24 +02:00
Gregory Hainaut c638e5ec87 gsdx debug: use a pretty format name when dumping texture
Much faster to read the format on the file name :)

+ remove a couple of useless ifdef
2016-09-07 08:34:24 +02:00
Gregory Hainaut 6abd806539 gsdx perf: print frame time distribution 2016-09-07 08:34:24 +02:00
Gregory Hainaut cedc4f241d gsdx hw: add a crc hack for Dragon Quest & Final Fantasy in Itadaki Street
The hack only fix the HW renderer but not the SW renderer. However I'm not sure
the issue is from GSdx.

The hack will disable alpha test that used to generate empty draw call.
2016-09-04 19:41:31 +02:00
Gregory Hainaut bb50b3419e Merge pull request #1542 from PCSX2/gsdx-texture-transfer-motocross
gsdx: save the blit buffer register when a transfer is started
2016-09-04 16:55:08 +02:00
Gregory Hainaut 52a2bb5524 gsdx ogl: disable texture shuffle when a render a basic sprite
Fix shadow issue on "Shadow of Memories"
2016-09-03 12:18:52 +02:00
Gregory Hainaut 79209c31bf gsdx ogl: avoid bad logging of framebuffer/texture overlap 2016-09-03 12:18:51 +02:00
Gregory Hainaut a123e65185 gsdx sw: handle the invalidation of the palette when written by the GS
Manual gives all setup to upload a palette from the host. But nothing forbid to render
directly in the palette buffer. (GS rule nb 1, there is no rule ^^)

Fix Virtua Fighter 2 dark colors

However I'm not sure we can fix HW renderer. Rendering is done on the GPU but palette
handling is done on the CPU... So we need to read back data (ouch, and slow). A quick
test didn't get the expected results. Potentially there are others bugs (aka not gonna
happen on the HW renderer)
2016-09-03 00:21:56 +02:00
Gregory Hainaut d01194da0b gsdx sw: help debug
* quick way to disable verbose fixed tex0
* use palette address/format when the palette is dumped :)
2016-09-03 00:10:23 +02:00
Gregory Hainaut 6323268f40 gsdx tc: reset dirty alpha flag when we lookup a frame
Fix black skin regression for haunting ground (at least on my gs dump)
2016-09-02 21:31:13 +02:00
Gregory Hainaut c9a1558c60 gsdx: save the blit buffer register when a transfer is started
Fix motocross mania missing texture. Close #1319

As far as I understand, transfer is initialized in DIR. But the real
write only occured later so the blit buffer could have been overwritten
by a new value.

BLIT 0 13700
TREG 40 40
DIR 0 0
BLIT 0 13f00 <=== the bad guy
Write! ...  => 0x3f00 W:1 F:C_32 (DIR 00), dPos(0 0) size(64 64)

v2: set a value in m_tr.m_blit for load state
2016-09-01 09:59:38 +02:00
Gregory Hainaut b546d096bd Finally AMD delivered the fix to support separate shader object !!!
v2: blacklist AMD driver from the start of 2016
Please note that it is highly recommended to upgrade to a stable&working driver
2016-09-01 09:47:57 +02:00
Gregory Hainaut 448fee76c1 gsdx ogl: move GS setup in SetupIA
Fix regression when geometry shader is disabled
2016-09-01 09:47:46 +02:00
Gregory Hainaut 165bcf5662 Merge pull request #1539 from FlatOutPS2/DX
GSdx DX: extend ATE to better support additional corner case DX port
2016-08-29 11:39:22 +02:00
Gregory Hainaut 19a9d639b1 Merge pull request #1538 from PCSX2/gsdx-improved-ate
gsdx ogl: extend ATE to better support additional corner case (issue …
2016-08-29 11:38:57 +02:00
Gregory Hainaut da688114f4 Merge pull request #1522 from PCSX2/gsdx-issue-1438
gsdx tc: don't partially invalidate small target
2016-08-28 16:51:25 +02:00
Gregory Hainaut e4337b6631 glsl2h: add input files in the header
Close #1404
2016-08-28 16:31:35 +02:00
Gregory Hainaut a1cdf7189a gsdx tc: don't partially invalidate small target
It creates a regression on game that uses a small temporary target to
upload textures of various sizes. Inital code was done to handle direct
frame write (background, FMV) so big target
2016-08-28 16:24:02 +02:00
FlatOutPS2 e5c6b2e952 GSdx DX: extend ATE to better support additional corner case DX port
Port for Direct3D renderers of PR #1538.
2016-08-28 15:13:15 +02:00
Gregory Hainaut 7103707aeb gsdx: new linux option linear_present
Purpose is to control the filtering when final image is displayed on the screen

Could improve the sharpness of the output in some games (ofc, it will be pixelated)
2016-08-27 19:23:12 +02:00
Gregory Hainaut 2b925e0103 gsdx ogl: extend ATE to better support additional corner case (issue #1499)
When depth primitive is constant and depth test is greater or equal, we can
execute the depth write after color (depth status will only depends on the initial
value)

New case for RGB_ONLY ate:
If the blending equation uses a fixed alpha or a source alpha. We can postpone the alpha write
in a 2nd pass.
If depth can also be postponed, we can guarantee the order of correctness of the value.
1st pass => do RGB
2nd pass => do Alpha & Depth

It fixed Stuntman letter rendering :) Remaining of the game is still broken :(
2016-08-27 14:46:10 +02:00
Gregory Hainaut b6693c4345 gsdx dx/gl: fix depth optimization
4th bit is the fog value, not the depth
2016-08-27 13:57:39 +02:00
Gregory Hainaut 411a25b610 Merge pull request #1531 from PCSX2/clang-tidy-empty-instead-of-size
pcsx2|common|gsdx: use empty() instead of .size() ==/!= 0 check
2016-08-26 10:05:01 +02:00
Gregory Hainaut 5a1cb00a07 Merge pull request #1517 from FlatOutPS2/DX
Gsdx alpha test improvement DX
2016-08-25 09:25:19 +02:00
Gregory Hainaut 1fb2c66a21 gsdx ogl: Unscale line
Line thickness will be increased to N pixels (N is the upscaling factor).

Code will also be enabled by UserHacks_unscale_point_line = 1
2016-08-22 18:18:12 +02:00
Gregory Hainaut 018895067b gsdx ogl: restore code to unscale point
enabled by UserHacks_unscale_point_line = 1

Point will be transformed into a NxN square sprite. (N is the upscaling factor)
2016-08-22 18:18:12 +02:00
Gregory Hainaut 1a8825b374 pcsx2|common|gsdx: use empty() instead of .size() ==/!= 0 check
Enhance readability reported by clang tidy
2016-08-21 17:20:13 +02:00
Gregory Hainaut 8ee2d3d367 gsdx: use static assert when possible
reported by clang tidy
2016-08-21 15:22:09 +02:00
Jonathan Li 2b2042e1c2 gsdx: Allow screenshot compression level to be changed
At higher resolutions it takes too much time to save a screenshot at the
maximum compression level. So let's allow the user to set the
compression level.

This re-uses the png_compression_level setting. The default compression
level is 1 for speed, but if the user wishes to increase the compression
level (without using an external tool) and doesn't mind if the
screenshot takes more time to save then they can increase the
compression level up to a maximum of 9 (which can take quite a while).

Fixes #1527.
2016-08-20 14:01:24 +01:00
Gregory Hainaut d6b834e8af gsdx hw: don't execute blit fmv with target
Avoid invalid operation due to depth buffer
2016-08-20 12:56:30 +02:00
Gregory Hainaut fa826b3167 gsdx tc: check compatible bit when wrote in middle of target
All maths are wrong otherwise. Fix half screen issue in WRC
2016-08-20 11:57:15 +02:00
Gregory Hainaut a49b3c9bf6 gsdx tc: log more stuff on texture cache dirty & frame format 2016-08-20 11:52:22 +02:00
FlatOutPS2 87eed3965e GSDx GSRendererDX code improvement
Replace local copies context and env with m_ prefix originals.
2016-08-19 16:02:49 +02:00
FlatOutPS2 c5cd716c20 Gsdx alpha test improvement DX
Port for the DX renderers of the alpha test improvement for OGL created by gregory38.
2016-08-19 15:59:54 +02:00
Gregory Hainaut 765b68458a gsdx: improve logging
Don't enable UpdateValidity print by default (+20~25% on log size)
Only useful in rare cases
2016-08-17 21:23:06 +02:00
Gregory Hainaut 19ceea4f1e Merge branch 'strict-aliasing' 2016-08-17 18:53:08 +02:00
Jonathan Li d36002a02a gsdx:cmake: Use PNG_LIBRARIES instead of PNG_LIBRARY
PNG_LIBRARIES adds both libpng and zlib to the command line.
PNG_LIBRARY only adds libpng to the linker command line, and the cmake
documentation also suggests not to use it.
2016-08-16 20:35:21 +01:00
Gregory Hainaut fa249a3f78 gsdx ogl: don't rely on the Z value in m_vt
Value seems wrongly rounded and you can't distinguish 0xFFFF from 0xFFFE

Instead check that depth is constant for the draw call and the value from the vertex buffer

Fix recent regression on GTA (and likely various games)
2016-08-16 07:30:52 +02:00
Gregory Hainaut 5c7e2432bd gsdx-ogl: improve alpha test for GTA vice city (letters)
I don't understand why but it seems depth is rounded down even in 16 bits.
So uses 0xFFFE to enable ate_all_color_then_depth in 16 bits format too :)
2016-08-15 11:38:07 +02:00
Gregory Hainaut 5fbf702500 gsdx ogl: new optimization to bypass the alpha test
In FB_ONLY mode the alpha test impacts (discard) only the depth value.
If there is no depth buffer, we don't care about depth write. So alpha
test is useless and we can do the draw with a single draw call and no program
switch
2016-08-15 11:35:24 +02:00
Gregory Hainaut b62859ffa2 gsdx ogl: only enable the alternate alpha test in FB_ONLY
RGB_ONLY requires to handle the alpha channel in the alpha test
2016-08-15 11:25:19 +02:00
Gregory Hainaut 0417a10483 gsdx: allow strict aliasing optimization 2016-08-14 21:01:40 +02:00
Gregory Hainaut 2e7fab7813 gsdx sw: avoid aliasing issue, clear function was wrong anyway
Value must be un-normalized. Function is only used with 0.0 so
API was updated to only use the 0 case.
2016-08-14 21:01:09 +02:00
Gregory Hainaut 90707f8d4e gsdx: avoid aliasing issue
Extend GSVector to support float move

Initial code likely used integer move for performance reason. However due to
the nan correction, register is now in float domain.
2016-08-14 20:48:34 +02:00
Gregory Hainaut 447e0d0336 Merge pull request #1514 from PCSX2/gsdx-nouveau-driver
Gsdx alpha test improvement
2016-08-14 20:29:09 +02:00
Gregory Hainaut 15db7eeb81 gsdx: use correct format in printf
u => unsigned
d => signed

zu => size_t
2016-08-13 12:48:35 +02:00
Gregory Hainaut 0a1454ff6c gsdx: use prefix increment operator for non-primitive type
I think compiler is clever enough to optimize the code but it won't hurt
and it reduces the number of cpp check warning.
2016-08-12 23:11:26 +02:00
Gregory Hainaut dc2864ad7a gsdx hw: avoid a null deferenced false-positive in coverity
It is a false positive because it supposes that PRIM->TME is both true and false.

It is the first Schrodinger's variable ;)
2016-08-12 23:02:06 +02:00
Gregory Hainaut 1a09712f89 gsdx: init m_custom_width/height to a default value
Avoid useless warning on coverity
2016-08-12 22:48:20 +02:00
Gregory Hainaut c41cf6c444 gsdx: init variable in constructor
CID 168626 (#1 of 1): Uninitialized scalar field
uninit_member: Non-static class member m_end_block is not initialized in this constructor nor in any functions that it calls.
2016-08-12 22:45:17 +02:00
Gregory Hainaut 41afd85a5d gsdx sw: fix GCC warning enumeral and non-enumeral type in conditional expression 2016-08-12 19:30:14 +02:00
Gregory Hainaut 88b7470c07 gsdx: use const qualifier to avoid gcc (false positive) warning
variable  might be clobbered by ‘longjmp’ or ‘vfork’

Only remains warning for 2 variables: success & image
2016-08-12 19:30:14 +02:00
Gregory Hainaut 19e992869a Merge pull request #1512 from PCSX2/gsdx-dynamic-thread-height
gsdx sw: add extrathreads_height to control the quantity of pixels pr…
2016-08-12 15:52:13 +02:00
Gregory Hainaut 332ef5892b gsdx: replace hardcoded constant with nice enum 2016-08-11 22:30:34 +02:00
Gregory Hainaut 8b94320524 gsdx ogl: implement another algo to handle alpha test
Fix rendering issue on letters on Kengo/burnout 3/...

Default algo will execute the alpha test in 2 passes. However due to blending
you can't handle accurately the color.

Fortunately for us, the rendering uses an always pass depth test so you
can execute first all the color rendering (which doesn't depends on the alpha test)
And then the depth part which depends on the alpha test.
2016-08-11 21:15:17 +02:00
Gregory Hainaut 866173a481 gsdx ogl: move zbuffer emulation code into a dedicated function
* Code was factorized a bit with the help of max_z
* Add an extra optimization if test is ZTST_GEQUAL and min z value is
the biggest value. Z test will always be pass.
Note: due to float rounding (23 bits mantissa vs 24 bits depth) the test
is done against 0xFF_FFFE and not 0xFF_FFFF. It is wrong but GPU will
also use float so impact will be null.
2016-08-11 21:14:58 +02:00
Gregory Hainaut 1fa9e3fc93 gsdx ogl: align shader self test to atst previous change 2016-08-10 11:30:13 +02:00
Gregory Hainaut 74822d6ba3 gsdx ogl: rework atst handling
* Move the rounding operation in the constant buffer
* Merge less with less equal. And merge greater with greater equal

Need test
2016-08-10 11:30:11 +02:00
Gregory Hainaut 9a188a87c2 gsdx ogl: move ATST emulation in a dedicated function
Future commit will try to reduce the number of Alpha Test possibilities
2016-08-09 20:53:44 +02:00
Gregory Hainaut fb2182dd9b gsdx small nitpick 2016-08-09 15:15:30 +02:00
Gregory Hainaut 0dcce91a6e Merge pull request #1504 from ssakash/RemoveFunction
GSDX: Remove GetDeviceSize()
2016-08-08 16:15:07 +02:00
Akash bc24e90070 GSDX: Remove GetDeviceSize()
It's not used anywhere and it doesn't seem needed.
2016-08-07 12:43:03 +05:30
Jonathan Li af60501fb6 windows: Remove unused solution files and property sheets
The solution files are unused and for ancient Visual Studio versions -
GSDumpGUI has its own solution file, and bin2cpp is included in the main
solution file.

The property sheets have either fallen out of use or were never used in
the first place.
2016-08-04 23:09:08 +01:00
Jonathan Li 869f4d1898 gsdx: Only set ini path in GSdxApp::Init() if it's empty
Fixes a regression introduced by 46ba9aa117,
where the Linux GS replayer would always use the options in inis/GSdx.ini
(or use the default options if that doesn't exist) to replay the dump,
instead of using the GSdx.ini from the specified ini folder.
2016-08-04 00:53:50 +01:00
Gregory Hainaut e9e1b33884 gsdx: log a nice string instead of the hex format of PSM 2016-08-03 18:07:54 +02:00
Gregory Hainaut 41bede60eb gsdx replayer: fix broken init in replayer
"regression" from previous defered init
2016-08-03 11:57:56 +02:00
Gregory Hainaut 27f313ee16 gsdx hw: limit OI_GsMemClear to large clear
Help to reduce the speed impact
2016-08-03 11:23:23 +02:00
Jonathan Li f978f9a07d Merge pull request #1477 from turtleli/gsdx-defer-init
gsdx: Avoid illegal instruction crash on older CPUs
2016-08-02 23:00:19 +01:00
Gregory Hainaut 16affc9ef4 gsdx tc: potentially fix a regression
Update done on f712c5c6d0

Previous code use the size of the draw to compute latest block. I
don't know why I use .x/.y which are the origin offset so the start of the block.
2016-08-02 15:35:28 +02:00
Jonathan Li 44f90efb93 gsdx:psx: Fix illegal instruction crash on old CPUs
Check the instruction set first in GPUinit, GPUconfigure and GPUtext
to prevent unsupported vector instructions from being executed.

Move the vector initialisation in GPUinit to a separate function - it
avoids a vzeroupper instruction.
2016-08-01 19:38:23 +01:00
Gregory Hainaut 92ae8a5be7 gsdx sw: print current draw call of GetSizeFixedTEX0
To easily detect compare with hardware renderer
2016-07-31 13:19:38 +02:00
Gregory Hainaut 1ea83ba3a7 gsdx debug: log current area of the draw
This way it is easier to see which part of the texture is read, and which part
of the framebuffer is updated
2016-07-31 13:19:38 +02:00
Adam Diffin 6376e8d6e7 Just a simple grammar fix. (#1489) 2016-07-31 01:23:35 +01:00
Gregory Hainaut 140fe74ca2 gsdx ogl: fix bad blending regression
(when accurate blending is disabled)

Regression was introduced in 29c97a9bf2 (11th June)
2016-07-30 09:59:53 +02:00
Gregory Hainaut 242ac26299 Merge pull request #1485 from ssakash/HPO_Custom
GSDX-TextureCache: Port Half pixel offset hack for custom resolutions
2016-07-29 15:07:02 +02:00
FlatOutPS2 02b0451d3c GSdx: Star Ocean 3 depth issue fix
Fixes issue #1475
2016-07-29 12:13:26 +02:00
FlatOutPS2 91e07727e3 GSdx Add missing CRC hacks 2016-07-29 12:12:52 +02:00
Akash dbad57b8e8 GSDX: Silence an implicit conversion warning
OMSetBlendState() doesn't accept passing a float value as it's argument
(AKA actual parameter).
2016-07-28 18:58:06 +05:30
Akash 47f69f694a GSDX-TextureCache: Port Half pixel offset hack for custom resolutions 2016-07-28 18:45:23 +05:30
Gregory Hainaut 620876e0e7 gsdx: fix compilation error with ICC
just require the -restrict compilation flag
2016-07-28 11:01:28 +02:00
Jonathan Li a9f9c1406c Merge pull request #1474 from turtleli/windows-64-bit-fixes
Windows 64-bit compile fixes
2016-07-28 00:19:04 +01:00
Jonathan Li 7e30d16797 gsdx: Use char array instead of string for name
"static string str;" causes a SIGILL signal on a "Nehalem" (SSE4.2)
QEMU VM when compiled with GCC 6.1.1.
2016-07-27 23:34:16 +01:00
Jonathan Li 46ba9aa117 gsdx: Defer GSdxApp initialisation on Linux only
vector push_back causes a SIGILL signal on a Nehalem (SSE4.2) QEMU VM
when compiled with GCC 6.1.1.

However, an empty constructor causes illegal instruction exceptions to be
generated on a Windows VM.

So here's an inbetween that looks stupid but works on what I've tested.
2016-07-27 23:34:15 +01:00
Jonathan Li 9f53987e3b gsdx: Use map for GSCRC::m_map
Visual Studio 2015 initialises unordered_map using vector instructions,
which can cause PCSX2 to crash if the instructions aren't supported.
2016-07-27 23:34:14 +01:00