Age | Commit message (Collapse) | Author |
|
Bug: 37670543
Bug: 36006390
Bug: 37679566
Bug: 37636434
In debug context, a script is forced to be recompiled every time it
is initialized.
To avoid the same .so file being written to by another thread,
while it is being loaded and used by one thread, do not save the .so
file. Delete it right after loading it.
Test: RefocusTest and RSTest (including ScriptGroup2 tests) with
debug.rs.debug set to 1 and CTS on Angler
Change-Id: If63e3d21e3d9abd007a66e0ec79c9e6f1c9f13a0
|
|
|
|
The debug version of RS runtime library libclcore_debug.bc did not have
the untyped variants of allocation accessors, causing some CTS tests to
fail in debug context. This CL fixed them by making the non-debug untyped
accessors available in debug context.
Affected tests are:
android.renderscript.cts.AllocationTest#testSetElementAt
android.renderscript.cts.AllocationTest#testSetElementAt2D
Test: CTS under normal context, and affected tests under debug context
Bug: 37636434
Change-Id: Icf3254a6391f15fdc6ad1bf5a12e2f185b011e22
|
|
|
|
to sizeof(*name) for clarity." into oc-dev
|
|
|
|
There was a line in VrView.java that would throw exception past API
23 per hoford@. On top of that, the line never did anything. This CL
removed that offending line and fixed the crashing problem.
Bug: b/37719569
Test: mm
Test: interact with the VrDemo app with and without debug context
Change-Id: Iba374383cf4563e15721bec62835291a36548b99
|
|
sizeof(*name) for clarity.
Fixes minor issues addressed in aosp/380576
Bug: 32511607
Test: mma, cts, vts
Change-Id: I8f2eca4bcc8c93671668a542dc6117df58979b56
(cherry picked from commit 94979da78cf5762c609a04dd42527edae03719a8)
|
|
Bug: 33350696
Test: compile, boot device, cts
Change-Id: I7168eae109abc2ce204e1b011574a4f2d3f25bc3
|
|
|
|
|
|
|
|
Bug: 30596040
In case such Allocations lose their other references and get deleted,
while being used by a rsForEach() call.
Test: CTS on x86_64 emulator
Change-Id: I01bd9ef9235ad963b461d235f4d63c481e7c034d
(cherry picked from commit 972be5160b8b52d07582fd0f0585524e2f7e143d)
|
|
|
|
Replace unsafe strcpy and strcat with bounded strlcpy and strlcat.
Bug: 32511607
Test: mma, cts, vts
Change-Id: I92bc8142f82eaf78f4465bf061871cad864c53b8
|
|
Bug: 24555166
Test: RSTest on x86_64 emulator
Change-Id: I039da54f2196b5981ae8c8bcb02d157cf05a75ac
(cherry picked from commit eb9a4d343f538c8a51a5eb82b4cfc728bd2e418f)
|
|
Bug: 24555166
This is for support lib.
Test: RSTest on x86_64 emulator
Change-Id: If50d28fc12d2d7a2df38b427b6d82cb89493dce6
(cherry picked from commit 97cf762f101b0a62672712016f4fd24e039e9a6e)
|
|
|
|
Bug: 27870945
Fail context creation, if the IO Fifo fails to create due to socket
creation error, which was never checked.
Test: CTS on Angler and x86_64 emulator
Change-Id: I61aa973ccd0ef4bd846f891c47c2df28899bfd07
(cherry picked from commit 436444f18d75827d1040ccfdaad1acae075ee630)
|
|
|
|
Breakpoints set on named ScriptGroups were never resolved by the
debugger due to aggressive optimizations by clang in release builds.
`debugHintScriptGroup2`, used by the debugger to inspect ScriptGroup
compilations at `optlevel == 0`, had its inner loop optimized away
completely as there are no visible side effects (where the ALOGV calls
are actually empty macros). This was fixed by using an `__asm__
__volatile__ ("")` construct as per the gcc docs:
noinline This function attribute prevents a function from being
considered for inlining. If the function does not have side-effects,
there are optimizations other than inlining that causes function
calls to be optimized away, although the function call is live. To
keep such calls from being optimized away, put `asm ("")` in the
called function, to serve as a special side-effect.
(see also
https://gcc.gnu.org/onlinedocs/gcc/Alternate-Keywords.html
https://gcc.gnu.org/onlinedocs/gcc/Common-Function-Attributes.html#Common-Function-Attributes)
Although preventing the inner loop from being optimized away is enough
to get the function called, it is apparent that clang only sets up the
call frame with arguments that it can see are used inside the function
body, meaning that we had garbage values for `groupName` and
`groupNameSize` unless the symbol is weakly visible so that the compiler
can no longer introspect that the arguments are ignored.
Test: aosp_x86_64-eng (emulator)
- RsTest (32-bit, 64-bit)
- cts -m RenderscriptTest
- rs/lldb (required various other pending CLs)
Bug: 23535482
Change-Id: I24e23895e67d1ede8bf8e671165a3a6843a11082
Signed-off-by: Luke Drummond <luke.drummond@codeplay.com>
(cherry picked from commit 6e64e51a6a252bd88acfa72dac8372c863f524e0)
|
|
* changes:
Update tests to reflect recent changes in lldb upstream and librs
Add .rs.info versionInfo parser to the cpu_ref driver
|
|
ScriptIntrinsicBlend's forEachDstAtop had broken NEON and ASIMD
implementations and were bypassed.
This CL fixed these two and turned them back on in
RsdCpuScriptIntrinsicBlend::kernel().
Bug: b/22047392
Test: CTS and RsTest on Angler
Change-Id: Ifa41d74222606f1c04b7b4e3fe1b43eb932ead89
(cherry picked from commit e0c6f1d59fa8e995cdd80a41cc8ed1d6af35652e)
|
|
Updates Android.mk files to use current SDK where applicable.
Bug: 24137209
Test: make -j32
Change-Id: I40dc7e5518cf73f1706fec4e7a221774819df1cc
|
|
- Recent work in lldb upstream has changed the output format for printing
resolved breakpoints to not print the function address in case the name
is known, and to now return `None` instead of "" in case a command has
no output
- The recent addition of licence headers to the RenderScript sources in
the lldb testsuite has caused upset to tests of breakpoints involving
specified file/lines. The copyright headers were introduced in
5bbcc3861c44435f89481f80946ef5c9c49968f2 which broke these tests
Test: aosp_x86_64-eng (emulator) rs/tests/lldb
- required various other pending CLs
Bug: 23535482
Change-Id: Ib65889865ea3d6442e7ee5c909afafee2fe947a0
Signed-off-by: Luke Drummond <luke.drummond@codeplay.com>
(cherry picked from commit 506520732420b196f4f17d67596fd81004faf009)
|
|
This is used by lldb to check for a versioning mismatch between the
frontent and backend, but is not used by libRS. However, libRS should
not fail in its presence, so this CL enables the runtime to skip this
section if it is present.
This field is not checked for by the compatibility lib.
Test: aosp_x86_64-eng (emulator)
- libbcc/tests
- RsTest (32-bit, 64-bit)
- cts -m RenderscriptTest
- rs/lldb (required various other pending CLs)
Bug: 23535482
Change-Id: I3056dd08e7e14ff1094ead60756fd568131bcf4d
Signed-off-by: Luke Drummond <luke.drummond@codeplay.com>
(cherry picked from commit 341f30630d310f52297cf2d9867618cb7a1e8969)
|
|
|
|
1) No reason to push native libraries onto target -- they should come from APKs.
2) Improve error reporting.
Test: aosp_x86_64-eng lldb/run_tests.py (on emulator)
Bug: 23535482
Change-Id: I633ffb650c28143996aa21bb5dd089f886efc0b5
(cherry picked from commit f2bde9ee29dac71db606c9bbb35fc5bd6cae4ca1)
|
|
|
|
Bug: 30316822
Test: CTS on x86_64 emulator
Change-Id: I5a4cfdb8a900ad684cff35815277bfbff2e15842
|
|
Bug: 34396220
Bug: 37067080
Test: mm
Change-Id: I2e623e3e64e4dead0d1407361274b556d52d4ab6
|
|
Bug: 34396220
Bug: 37067080
Test: mm
Change-Id: Ic35879e341f1a4357c6a8e9bdec8312d2507223f
|
|
|
|
- Use EGL sync fence instead of gui/DisplayEventReceiver.
- Remove obselete inclusion of ui/PixelFormat.h and gui/Surface.h.
- Remove dead code using SurfaceTexture in rsdAllocation.cpp, as
drv->surfaceTexture was never initialized.
Bug: 34396220
Bug: 37067080
Test: mm, boot Angler, all CTS tests pass, and example RS graphic apps
all still fine.
Change-Id: I5369c382a9815095301660735de61573e348a1dc
|
|
c/299496/ revised the protocol between bcc and bcinfo; bcc now injects
some metadata into the module which is in turn parsed by the
bcinfo::MetadataExtractor(Module *) constructor. Since RSoV uses bcinfo
on non-bcc-processed bitcodes, those missing data caused the constructor
to fail.
This CL uses a different MetadataExtractor constructor so that all the
required information are passed to the MetadataExtractor via full bitcode.
Bug: 30964317
Bug: 36657118
Test: RSoV LIT test, RSoVTest on Angler
Change-Id: I9b423ad8d0bcab5d261e595c10d4676486ec4b3b
(cherry picked from commit 56813de7a00f872fad490b1b3fe5ab1b620c1be5)
|
|
|
|
Don't rely on this being included transitively by some other header.
Test: m -j libRS
Change-Id: Ieeea9ca80f76f854bf9c557c59a0dde7ebda028e
|
|
- USAGE_IO_OUTPUT will use ANativeWindow APIs to deqeue and queue
buffers.
- USAGE_IO_INPUT will use AImage_getHardwareBuffer to get the handle to
hardware buffers.
Bug: 34396220
Bug: 36869607
Test: mm, boot sailfish, CTS pass and ImageProcessingJB works fine.
Change-Id: I71bdc21ed18102026cd842364145fd94666a0faf
|
|
|
|
Our current stack:
API->API_TO_HAL_translator-> HAL
->HAL_TO_Implementation_translator->Implementation
For most APIs:
- API passes objectCount.
- HAL expects objectCount.
- Implementation expects objectCount.
For APIs like ScriptGroupCreate:
- API passes byteCount. And unfortunately, these APIs are part of
NDK, we could not make them also passing objectCount like others.
- HAL expects objectCount.
- Implementation expects byteCount.
So that both API_TO_HAL_translator and
HAL_TO_Implementation_translator should correctly convert input
objectCount/byteCount to byteCount/objectCount.
This CL only fixes the API_TO_HAL_translator part.
API_TO_HAL_translator was mistakenly using byteCount as
objectCount, causing potential out-of-bound access.
Bug: 36404879
Test: mm
Change-Id: Ic3d4e6f4cfaa81239e9f732e739d5a79e08b1b64
(cherry picked from commit 699eec4b9a89078098717e98e20b8a0850daab84)
|
|
Bug: http://b/29154200
Bug: http://b/28070272
Test: (aosp_x86_64-eng emulator, full_fugu-eng, aosp_angler-eng) x
(RsTest 32-bit, RsTest 64-bit)
- Tried ( modified slang, unmodified bcc) and
( modified slang, modified bcc)
- "Modified bcc" is a forthcoming bcc change to turn off old struct
layout fixes, and to verify that front end (Module) and back end
(TargetMachine) agree on the layout of every exported struct type.
https://android-review.googlesource.com/#/c/358954
- "Modified slang" is a forthcoming slang change to add explicit padding
to struct types, thereby fixing the struct layout bugs.
https://android-review.googlesource.com/#/c/359135/
Change-Id: I393230bd9a8d5e87f0e3a94a2fa2bf6e89d00929
(cherry picked from commit 08012d1f91f78dcacdffa298d9bf3414dea75209)
|
|
am: 0fbf266df8
am: 0f64dbd00e
Change-Id: I4e5e048046f0d7c146ddf073e793ef8f3bc52ae9
|
|
am: 0fbf266df8
Change-Id: Ie33e60524637f96d2c2027cb23776867b4e40cd2
|
|
am: df1e54ae6c
Change-Id: Ie11bfa44f9b3234190c5ffab60cdc25506779bea
|
|
|
|
am: b6bc4300ff
Change-Id: Iee45b66f9f5472fe0123a5560d9af9f03a220280
|
|
am: 8607bbce71
Change-Id: I615cf80c5e6b80be5026abf8a9a60074e31be88c
|
|
am: 10657f54a0
Change-Id: I17e9237b8d7284b8271323b17cc3ecee8f56bf3c
|
|
|
|
|