We were using sw, the namespace used in the rest of SwiftShader, as the
namespace for Reactor. Putting Reactor code into its own namespace
makes it easier to untangle dependencies.

Bug swiftshader:16
Bug b/115344057

Change-Id: Iea4e049a4796323bf80b4f5e6e17778ccf17b995
Reviewed-on: https://swiftshader-review.googlesource.com/c/21308Reviewed-by: Alexis Hétu <sugoi@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>

Adds commit '65ce2e56889' from https://git.llvm.org/git/llvm as
'third_party/llvm-7.0/llvm'

Commands:
git remote add llvm https://git.llvm.org/git/llvm.git
git subtree add --prefix third_party/llvm-7.0/llvm llvm release_70 --squash

Bug b/115344057

Change-Id: I981f7e2fc47639ca8a8998b188e837bc3f268de5

git-subtree-dir: third_party/llvm-7.0/llvm
git-subtree-split: 65ce2e56889af84e8be8e311f484a4dfe4b62d7a

LLVM 7.0 will be added back as a subtree instead.

This reverts commit 1841c50c.

Bug b/115344057

Change-Id: I5fc51be1f107f3de0e97e8b53e282b0cc9557bed

Change-Id: Idc041213e40c0650b52209de199a098bd26a1b39
Reviewed-on: https://swiftshader-review.googlesource.com/c/21328Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Alexis Hétu <sugoi@google.com>

This reverts part of https://swiftshader-review.googlesource.com/21108

Bug b/73656151

Change-Id: I35c91cb5b28ede65eadebcb5e8141be4cf3f2e0e
Reviewed-on: https://swiftshader-review.googlesource.com/c/21368Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Shannon Woods <shannonwoods@google.com>

Bug b/116336664

Change-Id: I388b3c602d0b697ecedf19e390d8008ada0ea849
Reviewed-on: https://swiftshader-review.googlesource.com/c/21289Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Alexis Hétu <sugoi@google.com>

Bug b/117152542

Change-Id: I8bfa40d0e912f90946109c6d80889889ef7c5c55
Reviewed-on: https://swiftshader-review.googlesource.com/c/21249Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Alexis Hétu <sugoi@google.com>

The Vulkan implementation needs a directory for each architectural
layer, similar to the OpenGL ES stack. The entire rendering stack is
duplicated, leaving only Reactor common between them:

Renderer -> Device
Shader -> Pipeline
Common -> System
Main -> WSI

Bug b/117152542

Change-Id: I9c26b23654016d637f88ec2416f019ef65b9afbd
Reviewed-on: https://swiftshader-review.googlesource.com/c/21248Reviewed-by: Alexis Hétu <sugoi@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>

Although we only produce libraries, 'lib' is confusing as an output
directory. We already used 'out' for the CMake build.

Bug b/116336664
Bug b/29024574

Change-Id: I2bc1015a72100f81f734fc969a32e9f5a967e17c
Reviewed-on: https://swiftshader-review.googlesource.com/c/21228Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Alexis Hétu <sugoi@google.com>

The Vulkan implementation shouldn't depend on any of the legacy
directories. Also, changes to the environment-wide include paths
may have undesired consequences.

Bug b/116336664

Change-Id: Ied3c8f4e56994379db3518a4fbf0268451d84079
Reviewed-on: https://swiftshader-review.googlesource.com/c/21188Reviewed-by: Alexis Hétu <sugoi@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>

Bug swiftshader:116

Change-Id: I93208ea733cfe69474477ad2e00e52c8e86d1e25
Reviewed-on: https://swiftshader-review.googlesource.com/c/21268Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Alexis Hétu <sugoi@google.com>

##__VA_ARGS__ causes an error with clang if it's not preceded by a comma. The ## is
wholly unnecessary in this case so it can just be removed.

Bug b/73656151

Change-Id: Icc8a67a91b270d0e9b006cae7a438cdefa412f87
Reviewed-on: https://swiftshader-review.googlesource.com/21208Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Corentin Wallez <cwallez@google.com>

Starting from Android O, <cutils/log.h> is planned to be deprecated in the
future. The use of the original <cutils/log.h> is split into 2 header files.

Per Treble team's (b/78370064) description, <log/log.h> is for native shared
libs or executables, and <android/log.h> is for app JNI.

So a conditional preprocessor is put in place to look at the
ANDROID_PLATFORM_SDK_VERSION definition. If the value is less than 27, means
it's Android N or older, and it ought to include <cutils/log.h>.

Otherwise, it should be 27 (Android O) or any greater version, and it's
supposed to include the new <log/log.h> file.

If ANDROID_PLATFORM_SDK_VERSION is not definied, Andorid build system catches
it and stops the building process with an error message.

Bug: b/116855807
Test: pi-dev(28), oc-mr1-dev(27), and nyc-mr2-dev(26) branches
Test: aosp_arm-eng target
Change-Id: I732803e900144bf291feeb2cba1d632301c4fa21
Reviewed-on: https://swiftshader-review.googlesource.com/21130Tested-by: Merck Hung <merckhung@google.com>
Reviewed-by: Merck Hung <merckhung@google.com>
Reviewed-by: Alistair Strachan <astrachan@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>

SwiftShader shouldn't print anything to stdout in Release builds.

Also refactor UNIMPLEMENTED macro to be able to take a formatted string
with arguments.

Bug b/73656151

Change-Id: Ibadbfba3371324ba1bd608bd51bdac5e5923a20e
Reviewed-on: https://swiftshader-review.googlesource.com/21108Reviewed-by: Merck Hung <merckhung@google.com>
Reviewed-by: Alexis Hétu <sugoi@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>

Bug b/115344057

Change-Id: Iac34e89f56a1f67301beae3fcf492596522d08a2
Reviewed-on: https://swiftshader-review.googlesource.com/21168Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Alexis Hétu <sugoi@google.com>

Locking stencil buffers with a NULL format resulted in nullptr being
returned without acquiring the resource lock, while unlocking did
release the lock.

Instead of not releasing it on unlock, we must acquire it on lock
because otherwise the object could get destroyed before draw operations
end and attempt to unlock.

Bug b/116876483

Change-Id: Ie883115a1d6df3b40eb7c1feb5dcfde7316ec970
Reviewed-on: https://swiftshader-review.googlesource.com/21148Reviewed-by: Krzysztof Kosiński <krzysio@google.com>
Reviewed-by: Alexis Hétu <sugoi@google.com>
Tested-by: Krzysztof Kosiński <krzysio@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>

Bug b/115784742

Change-Id: Id3225adcaf77e1d26560151c506e885318bd0ff6
Reviewed-on: https://swiftshader-review.googlesource.com/21088Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Alexis Hétu <sugoi@google.com>

Replace all-makefiles-under makefile-macro with explicitly including
Android.mk(s) from subdirectories that are related to OpenGL and are
only needed by Android builds.

The change are mainly made to the top-level Android.mk and the
src/Android.mk. The other three sub-Android.mk(s) are deprecated and
deleted. Soong compile-time is expected to shorten with no use of
all-makefiles-under macros.

Since the project co-existed with both Android.bp and Android.mk.
The test was performed in absence of Android.bp on oc-mr1-dev branch
against aosp_arm-eng and aosp_arm64-eng targets.

NOTE: Android Pie uses Android.bp only.

Bug: b/29023322
Test: oc-mr1-dev branch, aosp_arm-eng (LLVM3) & aosp_arm64-eng (LLVM7) targets
Change-Id: I05ac46213f849747621c7084bca2549a22513881
Reviewed-on: https://swiftshader-review.googlesource.com/21068Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Merck Hung <merckhung@google.com>

The spec for glDeleteVertexArrays states that "Unused names in arrays
are silently ignored, as is the value zero."

b/116699321

Change-Id: I7cca0336dd9841b360f8fa20f6c0ac9677ec3ce1
Reviewed-on: https://swiftshader-review.googlesource.com/21048Reviewed-by: Chris Forbes <chrisforbes@google.com>
Reviewed-by: Alexis Hétu <sugoi@google.com>
Tested-by: Nicolas Capens <nicolascapens@google.com>

This commit fixes `Trunc(Float)` and `Trunc(Float4)` generic LLVM code
generation.  If `Trunc(x)` is implemented with `Float(Int(x))`, it will
result in quality warning in dEQP.

Bug: b/115344057
Test: dEQP-GLES3.functional.shaders.builtin_functions.precision.trunc
Test: dEQP-GLES3.functional.shaders.builtin_functions.precision.modf
Change-Id: I62d2dd1907e345fb00307b6c0d4d74613237f94b
Reviewed-on: https://swiftshader-review.googlesource.com/21029Tested-by: Chris Forbes <chrisforbes@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Chris Forbes <chrisforbes@google.com>

This commit fixes `Frac(Float)` and `Frac(Float4)` generic LLVM code
generation.

See also: https://bugs.chromium.org/p/swiftshader/issues/detail?id=74

Bug: b/115344057
Test: functional.shaders.builtin_functions.precision.fract.highp_vertex
Test: functional.shaders.builtin_functions.precision.fract.highp_fragment
Change-Id: I027b7ab44ba3060dc100a220ba19e5275f41d4c9
Reviewed-on: https://swiftshader-review.googlesource.com/20933Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Chris Forbes <chrisforbes@google.com>
Tested-by: Chris Forbes <chrisforbes@google.com>

This commit fixes `RoundInt(Float)` and `RoundInt(Float4)` generic LLVM
code generation.

Bug: b/115344057
Test: functional.shaders.builtin_functions.precision.sinh.highp_vertex
Test: functional.shaders.builtin_functions.precision.sinh.highp_fragment
Test: functional.shaders.builtin_functions.precision.cosh.highp_vertex
Test: functional.shaders.builtin_functions.precision.cosh.highp_fragment
Change-Id: I35de5cf6bec6607725523a10ebf098333957f39a
Reviewed-on: https://swiftshader-review.googlesource.com/20931Tested-by: Chris Forbes <chrisforbes@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Chris Forbes <chrisforbes@google.com>

This commit fixes `Round(Float)` and `Round(Float4)` generic LLVM code
generation.

Bug: b/115344057
Test: dEQP-GLES3.functional.shaders.builtin_functions.precision.sin
Test: dEQP-GLES3.functional.shaders.builtin_functions.precision.cos
Change-Id: I9758ac4ca69bb0f5a5b192b4be3310827625fac0
Reviewed-on: https://swiftshader-review.googlesource.com/20932Tested-by: Chris Forbes <chrisforbes@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Chris Forbes <chrisforbes@google.com>

Bug: b/115344057
Change-Id: I57d9cc936349c9f788a9bf791f55958caa6cbb73
Reviewed-on: https://swiftshader-review.googlesource.com/20930Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Chris Forbes <chrisforbes@google.com>
Tested-by: Chris Forbes <chrisforbes@google.com>

This commit fixes Floor() generic LLVM code generation.

Bug: b/115344057
Test: dEQP-GLES3.functional.shaders.builtin_functions.precision.floor
Change-Id: I70ec0babfe9778b7963296734f1901a73b1f696f
Reviewed-on: https://swiftshader-review.googlesource.com/20929Tested-by: Chris Forbes <chrisforbes@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Chris Forbes <chrisforbes@google.com>

This commit fixes saturated add/sub instructions in generic LLVM code
generation path.

Bug: b/115344057
Test: dEQP-GLES3.functional.texture.wrap.rgba8
Change-Id: Ie3e3b708565b3ad255804090e8a3ee5521f42982
Reviewed-on: https://swiftshader-review.googlesource.com/20928Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Chris Forbes <chrisforbes@google.com>
Tested-by: Chris Forbes <chrisforbes@google.com>

Use mixed indentations so that the code are aligned regardless the tab
width.

Bug: b/115344057
Change-Id: I9476c1b8ba5a518787dc76f50e1805b1b9d06e82
Reviewed-on: https://swiftshader-review.googlesource.com/21028Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Chris Forbes <chrisforbes@google.com>
Tested-by: Chris Forbes <chrisforbes@google.com>

Change-Id: I4e8a860446bde7f76ea289888bf592d78c59b9c4
Reviewed-on: https://swiftshader-review.googlesource.com/21008Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Merck Hung <merckhung@google.com>

Change-Id: I34b47a5327055f804951793d9a8a58f771d9b792
Bug: b/116336664
Reviewed-on: https://swiftshader-review.googlesource.com/20988Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Chris Forbes <chrisforbes@google.com>

Bug chromium:101600
Bug swiftshader:118

Change-Id: Id167a84c4d8781989d4d903384c4e6fe6f45fb85
Reviewed-on: https://swiftshader-review.googlesource.com/20868Tested-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Alexis Hétu <sugoi@google.com>

This patch contains an empty Vulkan shell with only GetProcAddr implemented

Initial version of the VulkanAPI, empty and unimplemented.
Successfully builds vk_swiftshader.dll for all platforms/configuration in Visual Studio.

To test using dEQP:
- Edit environment variables
	- Define VK_ICD_FILENAMES to <SwiftShader's source directory>\src\Vulkan\vk_swiftshader_icd.json
- If the location of vk_swiftshader.dll you're using is different than the one specified in
  src\Vulkan\vk_swiftshader_icd.json, modify it to point to the vk_swiftshader.dll file you want to use

Bug b/116336664

Change-Id: I560b53c3e4340b9c5ccd244a6693ff2f9f994a6f
Reviewed-on: https://swiftshader-review.googlesource.com/20788Reviewed-by: Alexis Hétu <sugoi@google.com>
Tested-by: Alexis Hétu <sugoi@google.com>

This commit switches several internal (implied) Makefile variables to
lower cases.  This follows the variable naming convention recommended by
GNU Makefile Manual.

Change-Id: I79819ef0cfd3828107dab665118f1511ef8a8337
Reviewed-on: https://swiftshader-review.googlesource.com/20848Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Logan Chien <loganchien@google.com>

Tweak ContextPtr to handle move construction/assignment correctly.
This was overlooked when I made the last minute change to use explicit lock/unlock.

Tweak LockGuard to allow construction from ptr to MutexLock, and behave
correctly if mutex == nullptr. This is required because the locking
machinery for EGL is /outside/ the error generation for bad EGLDisplay.

Add 'Big EGL Lock' around most of the EGL API

Bug: b/112184433
Change-Id: Ic4cd5fa9e6a16ae81cbb2b46e1cb881a850749c5
Reviewed-on: https://swiftshader-review.googlesource.com/20628Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Chris Forbes <chrisforbes@google.com>

Fixes: dEQP-GLES3.functional.shaders.constant_expressions.other.complex*
Bug: b/116598062
Change-Id: Ife2dd2edad17d6ebbd02653f484637672ce6af0a
Reviewed-on: https://swiftshader-review.googlesource.com/20828Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Chris Forbes <chrisforbes@google.com>

Bug: b/116263076
Change-Id: Id8657fdefb988827f0d78556884c4f55eb096ca3
Reviewed-on: https://swiftshader-review.googlesource.com/20808Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Chris Forbes <chrisforbes@google.com>

As mentioned in the patch itself, this is fairly benign, but still junk that a later pass has to be able to eliminate. Is easier to just not produce it in the first place.

Bug: b/116263076
Change-Id: I199672bb9b53eb4ee9d4a8e3a23fc2ce42f0dec6
Reviewed-on: https://swiftshader-review.googlesource.com/20770Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Chris Forbes <chrisforbes@google.com>

Matrix is laid out in column-major order, so the stride between elements
on the main diagonal is #rows + 1, not #columns + 1.

Bug: b/116263076
Change-Id: I00994b663bc4229e24f8ae1b7fa518caa0e7ea71
Reviewed-on: https://swiftshader-review.googlesource.com/20768Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Tested-by: Chris Forbes <chrisforbes@google.com>

Cleans up some noise that appears in dumps of pretty much every GLSL
shader.

Change-Id: I602a6356110803e30fcd8ba9818e56efa70bf5d6
Reviewed-on: https://swiftshader-review.googlesource.com/20769Tested-by: Chris Forbes <chrisforbes@google.com>
Reviewed-by: Nicolas Capens <nicolascapens@google.com>

Khronos test CL: https://gerrit.khronos.org/#/c/2944/
Bug: b/116259485

Change-Id: Ie10b3eb16da39f332966a5b957acb0232140a942
Reviewed-on: https://swiftshader-review.googlesource.com/20748Reviewed-by: Nicolas Capens <nicolascapens@google.com>
Reviewed-by: Alexis Hétu <sugoi@google.com>
Tested-by: Chris Forbes <chrisforbes@google.com>