| ae2ec3febedfc29376b9104413fb4042028f1265 |
01-Jun-2016 |
David Gross <dgross@google.com> |
Delete simple reduction implementation.
Bug: 27298560
Change-Id: I8c3d568e98aaf0b7d86881c985d13ed5b8e95338
sCpuCore.cpp
sCpuCore.h
sCpuExecutable.cpp
sCpuExecutable.h
sCpuScript.cpp
sCpuScript.h
sd_cpu.h
|
| 56ecb64783b2637a0ff907c9459a7271881e5963 |
06-May-2016 |
Yang Ni <yangni@google.com> |
Merge "Avoid breaking on dead launch options" into nyc-dev
|
| 87643a32d4eaefe349d3a81513fffe0c4e3e8a80 |
04-May-2016 |
Yang Ni <yangni@google.com> |
Avoid breaking on dead launch options
Bug: 28463187
Launch options array[2-4] are not initialized by the support library
in the current SDK.
They may have arbitrary values that invalidate the checks that this CL
is removing, and cause a kernel launch to bail out unnecessarily.
This CL removes those checks, since the array.*Start and array.*End
fields in the launch options are not used anywhere in the current
runtime.
Change-Id: Ib8a2e551619f770b402f569acbde635ee047675a
sCpuScript.cpp
|
| d48481048a68725e79bbe8ba8ea75e710c646fe1 |
28-Apr-2016 |
Miao Wang <miaowang@google.com> |
Fix potential security vulnerability in rsCpuExecutable.cpp
Bug: 28318721
Change-Id: I69a4521bc8fa9c8036f71cac4b846c7118114caa
(cherry picked from commit 523ecf7672c0d7b4c85e9e0674b2d588c42af0dd)
sCpuExecutable.cpp
|
| d708d99149b92fad40fa8e6053ee231afaffe375 |
21-Apr-2016 |
Simon Hosie <simon.hosie@arm.com> |
Refactor ARM Blur prefill logic.
Bug: 28116032
Bug: 28076184
Refactor the prefill logic for ARM (and improve documentation along the way) so
as to fix some cases where data is read outside of the source image, and to
minimise the remaining cases which must fall back to the C implementation.
Change-Id: I3d06416b40c48dea06258e9f7bb5ddc246d7c710
(cherry picked from commit 5a1f196d68d54513c081958adf4ce3dcafed9ea2)
sCpuIntrinsicBlur.cpp
sCpuIntrinsics_advsimd_Blur.S
sCpuIntrinsics_neon_Blur.S
|
| 633beab5de5ffe4c5068cf1c7a1c0bd09ee9f195 |
20-Apr-2016 |
Yang Ni <yangni@google.com> |
Remove unnecessary logcat message
Bug: 28281435
Change-Id: If0865b20576322845cd23743916c4589ea199287
sCpuCore.cpp
sCpuExecutable.cpp
|
| 388d02df11f57b12ff7a62dff297e3131766e0c7 |
08-Apr-2016 |
David Gross <dgross@google.com> |
Merge "Multithreaded execution of general reduction kernels over 2D and 3D iteration spaces." into nyc-dev
|
| 415a249300cb9ca3f560814bfde52c6796fed1f7 |
07-Apr-2016 |
David Gross <dgross@google.com> |
Multithreaded execution of general reduction kernels over 2D and 3D iteration spaces.
Includes fix for bug in reduce_backward fz3 kernel (combiner function).
Bug: 27299475
Change-Id: I48ed2f99f53dfc786a85e04dc0206cb3ebe98034
sCpuCore.cpp
|
| 5e48002cc4a11e9ce63852a77488cfb9f6765c2b |
06-Apr-2016 |
Yang Ni <yangni@google.com> |
Avoid accessing TLS in rsClearObject()
Bug: 27874893
Bug: 27925095
rsClearObject() may be called on rsContextDestroy() in the JNI thread,
where TLS is not initialized.
This also reverted Icca50fade1199b6f530f963e9e694366b7d8cf67 that
made rsContextDestroy() go through the FIFO. rsContextDestroy() must
happen off the FIFO, since it joins the FIFO consuming thread.
Both the rsClearObject() change and the rsContextDestroy() revert need
to go together to make RenderScript and Camera CTS tests pass. Hence
this single CL for both changes.
Change-Id: I5f1d91a4cc1ce9bf2b3386fbef6f1171638b5266
(cherry picked from commit ade3137c22dd91694742145c2d5fe9d6bae17516)
sCpuExecutable.cpp
sCpuExecutable.h
sCpuScript.cpp
sCpuScriptGroup2.cpp
|
| 8ad1691f2e620ab6d5e9b3e61d03a205d295d056 |
01-Apr-2016 |
David Gross <dgross@google.com> |
Merge "Guard general reduction logging output under property "debug.rs.reduce"." into nyc-dev
|
| 6760f7ba7934ddd51938a8d0206fc41c2a7cb419 |
01-Apr-2016 |
David Gross <dgross@google.com> |
Guard general reduction logging output under property "debug.rs.reduce".
Bug: 27299475
Change-Id: I5be634fe38d20b9fe6867ad3c0c0b982442b52fd
sCpuCore.cpp
sCpuCore.h
|
| 644d5943a78b3d84a21a617e7fd7fcd8cff4500a |
31-Mar-2016 |
Miao Wang <miaowang@google.com> |
Fix and enable advsimd_Blend.S for arm64 IntrinsicBlend.
Bug: 27153652
- Fixed address arithmetic for the assembly.
- CTS tests pass.
Change-Id: I81ba4cca2ef830f40d17bca6abd60235b71f559e
ndroid.mk
sCpuIntrinsicBlend.cpp
sCpuIntrinsics_advsimd_Blend.S
|
| abe31c581489727afa84f9b756f0d2a0f4be057e |
30-Mar-2016 |
David Gross <dgross@google.com> |
Merge "Multithreaded execution of certain general reduction kernels; reduction test overhaul." into nyc-dev
|
| 10adb0c2029f112b5738228617d5645f6ecea0c5 |
29-Mar-2016 |
David Gross <dgross@google.com> |
Multithreaded execution of certain general reduction kernels; reduction test overhaul.
A reduction kernel is eligible for multithreaded execution if it has a
combiner function and it is launched over a 1D iteration space.
Note: Properties debug.rs.reduce-accum and debug.rs.reduce-split-accum
are added for debugging multithreaded reduction.
The following changes are made to reduction tests in RsTest:
- Overhaul the test framework -- now data-driven, and can execute the same
test multiple times with different seeds and input sizes, features
separate sets of quick correctness tests, full correctness tests,
and performance tests. (Performance tests are not run by default.)
- Report timing information for test execution.
- Report more information for fz* kernel testing.
- Remove dp kernel testing -- this involved floating-point arithmetic
which is not guaranteed to produce identical results between java
and rs or for different rs multithreaded executions.
- Add sumgcd kernel testing. This is intended to be representative of
a compute-heavy kernel.
- findMinAndMax kernel testing must compare cell value not cell index
-- two or more cells might have the same min or max value, and java
and various rs multithreaded executions are not guaranteed to find
the same cell.
- Fix bug in findMinAndMax kernel's combiner function. (It behaved
incorrectly when operating on an accumulator datum that has been
initialized but never passed to the accumulator function.)
- RsTest now requests largeHeap.
Bug: 27299475
Change-Id: I58f99c21389dbae5c8e3ad85d98700dc165664bb
sCpuCore.cpp
sCpuCore.h
sCpuExecutable.cpp
sCpuScript.cpp
sd_cpu.h
|
| 2e252d9a39464d67c9e0d9d72e7aaff80c7a3939 |
24-Mar-2016 |
Yang Ni <yangni@google.com> |
Merge "Separate libRS.so into public and internal libs" into nyc-dev
|
| 7c045dff08287a0273e6a0340d5ca88a90030363 |
06-Aug-2015 |
Yong Chen <yong.a.chen@intel.com> |
X86: Fix IntrinsicBlend bugs and re-enable BLEND_DST_ATOP optimization
Bug: 22047392
Signed-off-by: Yong Chen <yong.a.chen@intel.com>
(cherry picked from commit 7c892eea2bf1f924b56e68af0b4b03c03cec12ce)
Change-Id: I27f78e671e6d5a57ccc425ac5afff0321d514e6e
sCpuIntrinsicBlend.cpp
sCpuIntrinsics_x86.cpp
|
| a7481b21c4ad4127f54c02c7402039d068948a34 |
11-Mar-2016 |
Yang Ni <yangni@google.com> |
Fixed intrinsics inside a script group
Bug: 27439842
An RS intrinsic relies on RSExpandKernelDriverInfo.usr for the "this"
pointer to the ScriptIntrinsic object.
Need to to correctly set that field in a script execution for intrinsics.
Also fixed a bcc command-line option.
Added a test of script group with intrinsics to ImageProcessing GroupTest(native).
Change-Id: Ia8b45bca3396ac987ba167fd1709ee89123a2557
(cherry picked from commit 7a106adbb5e370b544f2cbf5b83c063560c81b57)
sCpuScript.cpp
sCpuScriptGroup2.cpp
|
| 85d0cb112010781ad07b5979f0824b7f0ae3987e |
18-Mar-2016 |
Yang Ni <yangni@google.com> |
Separate libRS.so into public and internal libs
Bug: 26546579
libRS_internal.so is not going to be whitelisted by the NDK linker,
therefore loadable only by the internal implementation, but not any
application code.
libRS is going to contain only publicly exported functions, i.e. the
functions declared in frameworks/rs/rs.h, including those in
rsApiFuncDecl.h (generated from rs.spec).
Change-Id: Iac58695316edc9aff87ade2bac3b4a44d1a68b23
ndroid.mk
|
| 838c2d0dc0b8b07d4aa8de4360765738c6d2770e |
25-Feb-2016 |
Miao Wang <miaowang@google.com> |
Better handle error log and nullptr reference.
Bug: 27367378
Change-Id: Ide29c702d7a70dfe087ddc600d3fd17279223196
(cherry picked from commit 68e00893f7400d1ce71a9ed92365ffdfd6730c03)
sCpuScript.cpp
|
| f7f9316f4cf373733285f947bea876ead7216715 |
18-Nov-2015 |
Verena Beckham <verena@codeplay.com> |
Add a new libclcore_g library which contains debug info.
Add a new libclcore_g.bc which is functionally equivalent to
libclcore.bc but is compiled with debug information and -O0.
To achieve this some allocation functions had to be written in C,
because they were previously available in LLVM IR without debug
information.
The RenderScript cts has been run to verify these changes, both as it
is and forcing it to link with the new libclcore_g library. Both runs
show no regressions.
Signed-off-by: Verena Beckham <verena@codeplay.com>
(cherry picked from commit fb99e0f905b0f73a505c7900b434531ce7c3e2e5)
Change-Id: I64c82ce6f571e766c3c607e926326040dff8257c
sCpuScript.cpp
|
| e4f999b761180a227864d97b172a42ca1d8c0df3 |
03-Feb-2016 |
Miao Wang <miaowang@google.com> |
Switch "transpose" for Matrix A & B, after gemmlowp change.
Change-Id: I26fcaebcca828388ef6fe53c6e9e4db8e60dd4d9
sCpuIntrinsicBLAS.cpp
|
| e0f52d61fe6790093ef19c734dda2e2c9c0c0fbf |
18-Nov-2015 |
Verena Beckham <verena@codeplay.com> |
Move scalar function implementations from into driver/runtime.
Move the implementation of all scalar SC_* functions from
rsCpuRuntimeMathFuncs.cpp to driver/runtime/rs_cl.c.
Removed the file rsCpuRuntimeMathFuncs.cpp since it is now empty.
Signed-off-by: Verena Beckham <verena@codeplay.com>
Change-Id: I2c5d063c67ad1d92bcf6b62841ca1fbb5a72bd9c
ndroid.mk
sCpuRuntimeMathFuncs.cpp
|
| a48442ad9cc9922d012407b3aebe648b5f0acf14 |
31-Jan-2016 |
Stephen Hines <srhines@google.com> |
Merge "Move implementation of rsFrac from cpu_ref into driver/runtime."
|
| 6c1876bbef1b2c89975dce91230a168bd2d2ce4c |
15-Jan-2016 |
David Gross <dgross@google.com> |
Support for general reduction kernels.
Requires coordinated change in frameworks/base.
Requires coordinated change in frameworks/compile/libbcc in order
for RsTest to run.
At present, general reduction kernels are run single-threaded.
Also: Remove dead struct field MTLaunchStructForEach::sig.
Bug: 23535724
Change-Id: Ice17ccf20a902f8a106eaa62ec071d46e3c0ad8c
sCpuCore.cpp
sCpuCore.h
sCpuExecutable.cpp
sCpuExecutable.h
sCpuScript.cpp
sCpuScript.h
sd_cpu.h
|
| 46c93e405b0ad7e8fca12d0b1f9eac3997798e45 |
10-Jan-2016 |
David Gross <dgross@google.com> |
Expect to see "exportReduceNewCount: 0" in metadata.
Needs corresponding change in frameworks/compile/libbcc: The driver
and bcinfo need to remain precisely in sync with regard to metadata
format in order for RenderScript to work at all.
Bug: 23535724
Change-Id: Iad82e495d03215e970d2e99ff45ccfa629688f03
sCpuExecutable.cpp
|
| d1be4d815ed18d5cf538ff5f36f5d36304eb337f |
18-Nov-2015 |
Verena Beckham <verena@codeplay.com> |
Move implementation of rsFrac from cpu_ref into driver/runtime.
Move the implementation of all the rsFrac functions from
rsCpuRuntimeMath.cpp to driver/runtime/rs_core.c.
Change-Id: I42bb4ba694fff76854c8deaefdd55671fec40acc
Signed-off-by: Verena Beckham <verena@codeplay.com>
sCpuRuntimeMath.cpp
|
| fef0cd45027f235126d4fb62bda5ea9037450d9c |
12-Nov-2015 |
Yang Ni <yangni@google.com> |
Various fixes in setting globals in a script group
Bug: 25602504
1) Fixed setting 64-bit globals on a 32-bit platform.
The long values were casted to pointer-size integers (uintptr_t) or void*
in the runtime implementation. This caused the upper 32 bits to be lost
on 32-bit platforms. Fixed that by casting to int64_t instead.
2) Fixed setting Allocation values (either as arg or a global) in a script
group.
The size for a value in a script is signed, with -1 indicating that the
value is an Allocation. The sign has to be kept across different layers
of our stack.
3) Added a test for both native and support libraries for passing
floating point values to a script group, and for passing 64-bit values
(long or double) to a script group.
Change-Id: Ifff099a76be7707df7b67c388395f5a00f9cae66
sCpuScriptGroup2.cpp
|
| e0bb9e833075eb665ac10b70c8b5bc8edf0e93a9 |
27-Aug-2015 |
Simon Hosie <simon.hosie@arm.com> |
Clarify vertical fetch loop in Blur assembly.
Change-Id: I50a12af2535c66e90ca847abde07f51fa910dc6d
ndroid.mk
sCpuIntrinsics_advsimd_Blur.S
sCpuIntrinsics_neon_Blur.S
|
| f5eb324d41433f81b64b5f49f7c4c5b656c7cdd8 |
03-Oct-2015 |
Stephen Hines <srhines@google.com> |
Add -Wextra to our RS build.
Change-Id: Ic1c430eec6840af4a8c19a1c13e8559ddf971eb3
ndroid.mk
|
| 9195e5188cb0e72d874512de01e7e58f1f47e0b7 |
15-Sep-2015 |
Miao Wang <miaowang@google.com> |
Update IntrinsicBLAS call to gemmlowp after rebase.
Change-Id: Id084ac7b53ea0b3c61311b4f4c78312f397b7c5f
sCpuIntrinsicBLAS.cpp
|
| e2af295f94c8fb17ba51d0e6a199d5ca265f92da |
24-Aug-2015 |
Chih-Hung Hsieh <chh@google.com> |
Rename some instructions to compile with llvm and gas.
* Rename .irep to .irp
* Rename "cmp x2, #-8" to "cmn x2, #8"
* Replace "vmov.s32" with "vmov"
* Replace "LSL #COMPONENT_SHIFT" with "LSL #(COMPONENT_SHIFT)"
* Nested .irp in *_Blur.S still cannot be compiled with llvm,
so -no-integrated-as is required.
* Verified before and after objdump binary codes are identical.
BUG: 23217766
Change-Id: I3c0d2eed44b79a39e3efcba3afadc3a14ca07874
ndroid.mk
sCpuIntrinsics_advsimd_3DLUT.S
sCpuIntrinsics_advsimd_Blur.S
sCpuIntrinsics_advsimd_Resize.S
sCpuIntrinsics_neon_3DLUT.S
sCpuIntrinsics_neon_Blur.S
sCpuIntrinsics_neon_Resize.S
|
| 14ce007a633b10e3b9a3fae29d8f53a7e8c9b59f |
31-Jul-2015 |
Matt Wala <wala@google.com> |
Add a basic implementation of the reduce kernel API to the CPU
reference implementation.
Bug: 22631253
For now, this just runs a serial reduction on one thread.
Change-Id: I34c96d24bb6f44274de72bb53160abcf79d143b0
sCpuCore.cpp
sCpuCore.h
sCpuExecutable.cpp
sCpuExecutable.h
sCpuIntrinsic.cpp
sCpuIntrinsic.h
sCpuScript.cpp
sCpuScript.h
sCpuScriptGroup.cpp
sCpuScriptGroup2.cpp
sd_cpu.h
|
| 83268ff957a17fc21132e923709ca43cea9b251c |
18-Jul-2015 |
Stephen Hines <srhines@google.com> |
Merge "Add a flag to the API to specify -O0 on the bcc command line."
|
| f5029803ae6ce9d92d70b76e7a7cdd8d484f31ca |
22-May-2015 |
verena beckham <verena@codeplay.com> |
Add a flag to the API to specify -O0 on the bcc command line.
Add a flag RS_CONTEXT_OPT_LEVEL_0 which sets a field mOptLevel in the Context
to 0 when calling bcc.
Using this flag will result in an object file that is easier to debug.
Change-Id: I9673506710cc9f34c415b694fe5d6bc7e967f1db
Signed-off-by: verena beckham <verena@codeplay.com>
sCpuScript.cpp
sCpuScriptGroup2.cpp
|
| 5d70cb591d78d62d10839a52302ec9087c6f3350 |
17-Jul-2015 |
Miao Wang <miaowang@google.com> |
Fix GetCpuInfo() routine to correctly check the cpuinfo file to make
sure we don't miss SIMD path if there is one.
Change-Id: I8c8841ba9924ee28ae56be8b3c66c50b5badf796
sCpuCore.cpp
|
| 223231fe99c9c958de4a1c8723aff88cb667de52 |
17-Jul-2015 |
Miao Wang <miaowang@google.com> |
Update eight_bit_int_gemm call after gemmlowp rebase and provide
non-optimal path for armv7 without NEON.
- gemmlowp will handle the optimal path for x86, NEON and aarch64
Change-Id: I67ce4c1e5b3195017a3d46895a8ce096682bc172
sCpuIntrinsicBLAS.cpp
|
| e941f18202b9c9883ff81c63710f7faec5c988e4 |
15-Jul-2015 |
Miao Wang <miaowang@google.com> |
Making libRSSupport able to optionally bundle libblas(V8) through dlopen
and dlsym.
Change-Id: I3ade3ad2802f3b8e5fc5661319b98a6212e6d8a2
sCpuBLAS.inc
sCpuBLASDispatch.h
sCpuIntrinsicBLAS.cpp
|
| 99d0e8130f5b4bb83d1a68d96496fa558e35193a |
07-Jul-2015 |
Miao Wang <miaowang@google.com> |
Update the BNNM cpu reference implementation with NEON friendly
gemmlowp.
Change-Id: I5bcfd0fa988d8075e70272f277d7d7fab93d5fea
ndroid.mk
sCpuIntrinsicBLAS.cpp
|
| 11fd9ec1ab8dfa7ae45c6edeea48dddc4633efea |
11-Jul-2015 |
Matt Wala <wala@google.com> |
CPU ref: Fix potential buffer over-read / uninitialized memory access.
GetCpuInfo() was reading /proc/cpuinfo into a string without properly
null terminating the result. The resulting unterminated string was
being passed to strstr().
Change the code to read the file with fgets(), which ensures the
result is null terminated. Also, document the GetCpuInfo() function
and the global variable that it sets.
Change-Id: I041331fdc25d79217ff7c1bf36a4aff2be8e0192
sCpuCore.cpp
sCpuCore.h
|
| cac180053adf5ab5900bf885f74a4cac174ece47 |
01-Jul-2015 |
Miao Wang <miaowang@google.com> |
Merge "Make support lib CPU driver able to access IntrinsicBLAS"
|
| 4b5b295f6437c580083997c8b7cbaad3e5975df8 |
18-Jun-2015 |
Miao Wang <miaowang@google.com> |
Make support lib CPU driver able to access IntrinsicBLAS
bug: 21902810
(cherry picked from commit 0d6b6f51b9d1f98478a32a270fa2304f0839ca8c)
Change-Id: I1f15306f1cf333d6932127ae62e9af8386ecb3fe
sCpuCore.cpp
|
| 06deda3751a4a7358a7c7e03fbf1e4325fafb807 |
30-Jun-2015 |
Miao Wang <miaowang@google.com> |
update the offset type for BLAS.BNNM
bug: 22184114
Change-Id: I6ec212f8d5feb46fc9d0f97862b206978af1675b
(cherry picked from commit 22cb808b0dfc9bd514d2e19b302a97f8455b5731)
sCpuIntrinsicBLAS.cpp
|
| 8b7ddc3c62f01923f2e416d0a4654d051d220e8a |
01-Jul-2015 |
Stephen Hines <srhines@google.com> |
Merge "Fix BLEND_DST_ATOP and disable vectorized optimizations temporarily."
|
| 1efae29f4bbe6c165caf6dfc4b89cf8a5f8c469b |
28-Jun-2015 |
Yang Ni <yangni@google.com> |
Avoid overwriting shared library file that is open
b/21949580
Writing different contents to an existing .so file, which is currently
open, may corrupt its globals and code. Choose a different file name
instead.
Change-Id: I3e649b09a55b43339283aa1c46f2844c66434c17
(cherry picked from commit 9844cffc8af28c4829012bc0c48f85d04b671f88)
sCpuExecutable.cpp
sCpuExecutable.h
sCpuScript.cpp
sCpuScriptGroup2.cpp
|
| 820e22b13fb48bc4719fa606ee4bc47e184970e0 |
29-Jun-2015 |
Stephen Hines <srhines@google.com> |
Fix BLEND_DST_ATOP and disable vectorized optimizations temporarily.
Bug: 22047392
This function needed to assign the 'w' component (alpha channel) of the
source input to the output. This patch does that for the C++ code, but
needs to be updated for the inline assembly that we have for
NEON/ASIMD/SSE. For the time being, we can just disable those paths.
Change-Id: Iaac02ae2d2cb670d299446d5a017d518c0e303da
sCpuIntrinsicBlend.cpp
|
| 565f98163670bf65d14803726b414c1e77791856 |
24-Jun-2015 |
Stephen Hines <srhines@google.com> |
Merge "Skip the checksum if we have precompiled code on the /system/ partition."
|
| 1c20667f7a174a7c0a1599d34a40c524fe24c615 |
19-Jun-2015 |
Yang Ni <yangni@google.com> |
Check for kernel chaining in ScriptGroup
b/21958851
bcc expects kernels chained up via input and output.
Check this in the runtime before passing ScriptGroup to bcc for
compiler fusion.
This serves the same purpose as
https://googleplex-android-review.git.corp.google.com/#/c/713267, but
provides additional checks in the runtime before bcc is even called.
Change-Id: I6004d577410ea1684f5043babadc1e1b885c4f6a
(cherry picked from commit 08e7693c13f895677b93537af3b6bc353318a81c)
sCpuScriptGroup2.cpp
|
| 6a236ad3a3760e8124b68a1b6220ed6e4fbfb152 |
22-Jun-2015 |
Stephen Hines <srhines@google.com> |
Skip the checksum if we have precompiled code on the /system/ partition.
Bug: 20894664
The system partition is read-only, and can/should be compiled correctly
during the offline creation of the system image. Since we cannot replace
these precompiled blobs (short of app update/OTA), there is no reason or
correct way to validate/replace the checksum.
Change-Id: Ia66bfdbe178bf215e146c3699f5bc7804222e978
sCpuScript.cpp
sCpuScript.h
|
| c5b28f7626368e0ad5dc9ba5223c4a25b6e8b8e5 |
29-May-2015 |
David Gross <dgross@google.com> |
Merge "Remove dead uses of RSCompilerDriver and of compiler callbacks."
|
| b043df0676fef226336deb3a00ead2f31e02343f |
29-May-2015 |
David Gross <dgross@google.com> |
Remove dead uses of RSCompilerDriver and of compiler callbacks.
Change-Id: Ibe8725074724b75e35c25a404daaba07ffbca2ab
ndroid.mk
sCpuCore.cpp
sCpuCore.h
sCpuScript.cpp
sCpuScript.h
sd_cpu.h
|
| b59f313ddb0e05d13c270862c62ac1fa2a5bb855 |
29-May-2015 |
Yang Ni <yangni@google.com> |
Merge "Avoid casting for CpuScript object"
|
| c246d91e490df18899490931469d277064070a63 |
22-May-2015 |
Yong Chen <yong.a.chen@intel.com> |
Fix memory leak when rs invoke function
Change-Id: I17c701dfda5b3031c1fc50bfc186e4b7f8b7ac1e
Signed-off-by: Yong Chen <yong.a.chen@intel.com>
sCpuScript.cpp
|
| cbff7bcc4aacdc39d56628fa5c7c50518d52748c |
27-May-2015 |
Yang Ni <yangni@google.com> |
Avoid casting for CpuScript object
b/20894664
Vendor driver may have different layout for the Script object.
Avoid casting mHal.drv but use the vendor provided lookupScript callback
instead.
Change-Id: Ie492dd98a435b35843ca222a8430c51ccfc8aea9
sCpuScriptGroup2.cpp
|
| cee9e9898ed549d00c1fd1c911feff66dbb4225e |
23-May-2015 |
Stephen Hines <srhines@google.com> |
Merge changes from topic 'global_info'
* changes:
Update RS driver to support extraction of global variable properties.
Use "override" instead of "virtual" when replacing methods.
|
| 5aa018cc36e589b07674957714d27ae3d1fa1c4e |
21-May-2015 |
Stephen Hines <srhines@google.com> |
Update RS driver to support extraction of global variable properties.
Bug: 20306487
This patch adds some new enums to classify properties (such as "static",
"constant", and "pointer") for global variables. The reference driver
is also extended to provide methods to examine these properties (when
the bitcode is compiled with bcc).
Change-Id: I331756f8a8990caf5ebdf85599060434a7cfdcb7
sCpuExecutable.cpp
sCpuExecutable.h
sCpuScript.cpp
sCpuScript.h
sd_cpu.h
|
| c060f1435e7b9405f3be8974417fa6f410f03753 |
14-May-2015 |
Stephen Hines <srhines@google.com> |
Use "override" instead of "virtual" when replacing methods.
Bug: 20306487
Change-Id: Ic83cb04cac153a7556f5d516e8f5ec88b5527b6f
sCpuCore.h
sCpuIntrinsic.h
sCpuIntrinsic3DLUT.cpp
sCpuIntrinsicBLAS.cpp
sCpuIntrinsicBlend.cpp
sCpuIntrinsicBlur.cpp
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsicConvolve3x3.cpp
sCpuIntrinsicConvolve5x5.cpp
sCpuIntrinsicHistogram.cpp
sCpuIntrinsicLUT.cpp
sCpuIntrinsicResize.cpp
sCpuIntrinsicYuvToRGB.cpp
sCpuScript.h
sCpuScriptGroup.h
sCpuScriptGroup2.h
sd_cpu.h
|
| 902077a7b2db5b34560289e28fed0fab36ef4e6c |
16-May-2015 |
Elliott Hughes <enh@google.com> |
Merge "Don't use TEMP_FAILURE_RETRY on close in frameworks/rs."
|
| 2df576745c52a04a42d615f83b03f19d803b7757 |
16-May-2015 |
Elliott Hughes <enh@google.com> |
Don't use TEMP_FAILURE_RETRY on close in frameworks/rs.
Bug: http://b/20501816
Change-Id: I63a7c690bc5bde3dc1354f6e04ad8d32b3b43015
sCpuScript.cpp
|
| 8237638f87ca0e265d050fbb13725b41a795fe5f |
13-May-2015 |
Yang Ni <yangni@google.com> |
Generate global info for fused kernel
b/21114118
Change-Id: I6332572f2f2e6f7857a50bf503e73984371cfe0b
(cherry picked from commit b616bcf3df4ce5d7ccc369acf8e90afe59c359fa)
sCpuScriptGroup2.cpp
|
| 5cb19fd40ad3225005f65b6960c875746c2136b9 |
12-May-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
Let vendor drivers access global-variable flags
bug 20894664
Expose flags for embedding global variables and skipping constant
globals in rsd_cpu.h. This allows vendor drivers to query or set these
flags.
Change-Id: I5c45355fb9254103ca8b0fa539edb691e6029ec7
sd_cpu.h
|
| 6063b3ef8f88f5b4f29994b58ef0c883c4bea066 |
11-May-2015 |
Miao Wang <miaowang@google.com> |
Merge "remove dead code (ALOGE) in rsCpuIntrinsicBLAS.cpp"
|
| e64c3d560186f976f9b8923e0f2c6ac3080913a2 |
07-May-2015 |
Miao Wang <miaowang@google.com> |
remove dead code (ALOGE) in rsCpuIntrinsicBLAS.cpp
bug: 21028875
Change-Id: Ia2d85a265f6e4a2617373f99b5c7bdc3810a7f24
sCpuIntrinsicBLAS.cpp
|
| 0be70c6dd8897603530dbc7411f27d9dbd4fee9a |
11-May-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
Merge "Add /system/vendor/{lib|lib64} to linker args"
|
| 1de97c307624f57d95281ebaa77c2129e66e21bc |
11-May-2015 |
Jason Sams <jsams@google.com> |
Merge "Add RSGlobalInfoPass information to RS driver."
|
| 682672e36b05349bc4d9dee74e9fab73ce804183 |
08-May-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
Add /system/vendor/{lib|lib64} to linker args
Add vendor-specific lib directory to ld.mc's library search path. Since
ld.mc takes multiple search directories, there is no need to derive a
single path at runtime.
bug 20894664
Change-Id: Id62d2acbd94914c70bf4668f553a94a33e2a0587
sCpuExecutable.cpp
sCpuScript.h
|
| 5e5ca4291845886257df24dbc4491b0698bd7f22 |
08-May-2015 |
Jason Sams <jsams@google.com> |
Merge "Add a way to link against different driver names."
|
| 08ef7b7f7977e9c991d8ba94a63860edcb88a3d9 |
30-Apr-2015 |
Miao Wang <miaowang@google.com> |
fix the CHER, CHPR, ZHER, ZHPR crash due to incorrect param order.
Change-Id: If91cbf969c75e01afc6d93b204bc8167180c9ef9
sCpuIntrinsicBLAS.cpp
|
| b75ba0fc7469d0bb4c1a6679664a846b3741792e |
27-Apr-2015 |
Miao Wang <miaowang@google.com> |
fix RsBlas_xgemv and RsBlas_xgbmv crash. (typo)
Change-Id: Ia948afa2bc4af22f99323618738d5eb7d415ca97
sCpuIntrinsicBLAS.cpp
|
| 4c368af7e705f0bcb77fa99495b2e33ef20d2699 |
06-May-2015 |
Stephen Hines <srhines@google.com> |
Add a way to link against different driver names.
The existing linker path hard-coded "-lRSDriver" into the command line, but
this won't work for partner drivers that have a different implementation
loaded. In order to still properly handle use of the CPU driver, this needs
to change depending on whether we actually loaded an OVERRIDE_RS_DRIVER or
not.
Change-Id: I0c4a4f12f5db819b234952bc8f364ac6300f147b
sCpuExecutable.cpp
sCpuExecutable.h
sCpuScript.cpp
sCpuScriptGroup2.cpp
|
| 8409d6414dd4a42aa59779fcfe9fce18648cb135 |
29-Apr-2015 |
Stephen Hines <srhines@google.com> |
Add RSGlobalInfoPass information to RS driver.
Bug: 20306487
This change enables vendor drivers to configure support for including
additional information about global variables in the emitted CPU code.
This information includes the number of total global variables, the
names of these variables, the addresses of these variables and the
sizes of these variables. The driver can also select whether the
information includes constant (immutable) globals or not.
The reference driver defaults to embedding information about each of
the existing, non-constant global variables.
Change-Id: I1e55fc3f08e518f04eeee3e4f9dc7b6ea3b80d7c
sCpuCore.cpp
sCpuCore.h
sCpuExecutable.cpp
sCpuExecutable.h
sCpuScript.cpp
sCpuScript.h
sd_cpu.h
|
| 80afd424b02291e6f088cab2961537c7ba1ee2a5 |
29-Apr-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
SetError if forEachMtlsSetup fails sanity check
bug 19866850
Explicitly set error when a kernel launch fails.
Change-Id: If42419f3cf7557c524afc5bb27a0753460ca7e4e
sCpuScript.cpp
|
| 38014b1ee631ee61e457b850138d3ea6eb73cb1b |
23-Apr-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
Merge "Tweak arguments to ld.mc"
|
| c2be413317efcc44e3b282fd38fca3334c11777e |
22-Apr-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
Tweak arguments to ld.mc
Change the arguments to match mclinker's revamped option parser.
Change-Id: I71135ebf81e3105192caf5c946b16297b2702d37
sCpuExecutable.cpp
|
| 247f8ee57196d6cf3264e6f7505f53e8f8a7860d |
19-Apr-2015 |
Logan Chien <tzuhsiang.chien@gmail.com> |
Code cleanup: Remove unused typedefs and declarations.
Change-Id: I48dafb2bc1dc335a52b289db2981397251f673c8
sCpuCore.cpp
sCpuScript.cpp
|
| f6dc05e18930bdb5ca74e5be773ee269263fb676 |
17-Apr-2015 |
Yang Ni <yangni@google.com> |
Merge "Dedup checksum calculation routines"
|
| cb17015fed6b11a5028f31cc804a3847e379945d |
16-Apr-2015 |
Yang Ni <yangni@google.com> |
Dedup checksum calculation routines
I introduced a separate routine to cacluate checksum for ScriptGroup
in my previous CL, in addition to the one we use for regular scripts.
This CL removes the new one and uses the old one.
While I am on it, I made some other minor changes, e.g., changing
mBuildChecksum in RsdCpuScriptIml from char* to uint32_t, and a few
other minor cleanups in ScriptGroup2 implementation.
Change-Id: I168fdbb4e7bd14f1549a687e7b0d0ca6dd4da866
sCpuExecutable.cpp
sCpuExecutable.h
sCpuScript.cpp
sCpuScript.h
sCpuScriptGroup2.cpp
sCpuScriptGroup2.h
|
| a9139c724f8312b3634d213599f2d6b3b2505db2 |
17-Apr-2015 |
Jason Sams <jsams@google.com> |
Fix allocation-less launches.
Change-Id: I6d6b46c55f3e88a810ebe51def3ebaccb1fd3fa2
sCpuCore.cpp
sCpuScript.cpp
|
| 0ced82a4a279c11411937afa0db1dc9d9782e35f |
16-Apr-2015 |
Tim Murray <timmurray@google.com> |
Merge "Rename BGEMM to BNNM. Modify layout of eight-bit GEMM-like intrinsic storage."
|
| 2b999883f2f390ee43ed18317d77c810a0c6657b |
13-Apr-2015 |
Tim Murray <timmurray@google.com> |
Rename BGEMM to BNNM. Modify layout of eight-bit GEMM-like intrinsic storage.
Change-Id: If4b1267dfd42d6dd65bedf20c0b674479eefab35
sCpuIntrinsicBLAS.cpp
|
| f02a2b0a2749d4a4f07edbc23eddff2e51d11b72 |
08-Apr-2015 |
Yang Ni <yangni@google.com> |
Compute checksum for script group generated code
Change-Id: I5f469266fd11b5031d1357f356ac40e00f5c0348
sCpuScript.cpp
sCpuScript.h
sCpuScriptGroup2.cpp
sCpuScriptGroup2.h
|
| 8b94222cc8ea83c780c98b22dd1921f392a2bcf6 |
03-Apr-2015 |
Yang Ni <yangni@google.com> |
Added a name to ScriptGroup2
Implemented in the runtime and CPU driver the new script group name.
Code caching is not implemented yet, and to be added in another CL.
Change-Id: I46e351c8412740512f56b7915b14f36183c6eeaf
sCpuScriptGroup2.cpp
|
| 6363723cacc46cfdf473f9c0579e661920f0901c |
10-Apr-2015 |
Jason Sams <jsams@google.com> |
remove dead code
Change-Id: I9ebda4e61be9cf270233ea8c0c51892517395487
ndroid.mk
sCpuRuntimeMath.cpp
sCpuRuntimeStubs.cpp
|
| ca3c91ff8b9a4a4ef092809ab2361ab48940a434 |
07-Apr-2015 |
Jason Sams <jsams@google.com> |
Merge "Assembly optimisations for Resize."
|
| aff744561bea3c8a7a7d59c0cb8cd9438f6dcd1c |
31-Mar-2015 |
Tim Murray <timmurray@google.com> |
Add eight-bit GEMM-like intrinsic.
Change-Id: I9b920900b4cb8b27e2ab27386d05f4175142d6b2
ndroid.mk
sCpuIntrinsicBLAS.cpp
|
| 3a98366c7f53b20f8550ffac82732d60ece794b4 |
27-Nov-2014 |
Simon Hosie <simon.hosie@arm.com> |
Assembly optimisations for Resize.
Assembly implementations of bicubic resize for AArch32 and AArch64.
Change-Id: I7336507482773a68c0dad40abb2114f1218cab9f
ndroid.mk
sCpuIntrinsicResize.cpp
sCpuIntrinsics_advsimd_Resize.S
sCpuIntrinsics_neon_Resize.S
|
| bd0af2d161e36e52e6782ccb2d15dd5a36467704 |
24-Mar-2015 |
Yang Ni <yangni@google.com> |
Clean up ScriptGroup2 implementation
b/19944127
- Adjust ref counting
- Fix value size type in closure creation
Change-Id: I1179d34aa67f845578740e71cc2da4f82419f251
sCpuScriptGroup2.cpp
|
| 7ce3943ca8b2773738710f9dfa145727ba16416c |
27-Mar-2015 |
Dan Albert <danalbert@google.com> |
Remove uses of libcxx.mk.
This is a no-op.
Change-Id: Id799af86c5de18d7f1b816c4964e0d32acb0840b
ndroid.mk
|
| 2fa8a238dd69afebdeb757adcb1d674043d78e32 |
26-Mar-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
Wrap TEMP_FAILURE_RETRY around system calls
BUG 19934827
Wrap TEMP_FAILURE_RETRY around system calls that can return EINTR
(waitpid, close).
Refactor fork/exec flows in various places into a utility function
and log errors so we can better understand failures in the test server.
Fix a small use-after-free issue in ScriptGroups.
Change-Id: I60b192f83c395a13c27cd6bd2289c44132b84791
sCpuExecutable.cpp
sCpuScript.cpp
sCpuScriptGroup2.cpp
|
| d6f44bc5aecd024bf2311b9ae844300fbdc3f584 |
18-Mar-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
Remove unnecessary include of RSInfo.h
RSInfo.h is obsolete and not needed here.
Change-Id: Ifc0e21ac0a393224b1e2dff63946aed16e5aa36d
sCpuExecutable.cpp
|
| 59b35f29cc576243a322bf88bc16063a9810da55 |
17-Mar-2015 |
Jason Sams <jsams@google.com> |
Fix issues with >2D launches
mtls->fep was being passed to setup in place of per-thread fep.
Change-Id: Ic26154fcf47dc7bc70cec43f0daf023fb83dfd78
sCpuCore.cpp
|
| b0abb140ac51b93d1a85aadaa63fe057f2d29850 |
12-Mar-2015 |
David Gross <dgross@google.com> |
Pass RsExpandKernelDriverInfo not RsExpandKernelParams.
Which is to say: retire RsExpandKernelParams and pass RsExpandKernelDriverInfo
directly to kernel wrapper functions instead.
Requires related change in frameworks/compile/libbcc.
Change-Id: I453f45ec18f389e88e27fcfa57ddf245d077cb98
sCpuCore.cpp
sCpuCore.h
sCpuCoreRuntime.h
sCpuIntrinsic3DLUT.cpp
sCpuIntrinsicBlend.cpp
sCpuIntrinsicBlur.cpp
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsicConvolve3x3.cpp
sCpuIntrinsicConvolve5x5.cpp
sCpuIntrinsicHistogram.cpp
sCpuIntrinsicLUT.cpp
sCpuIntrinsicResize.cpp
sCpuIntrinsicYuvToRGB.cpp
sCpuScript.h
sCpuScriptGroup.cpp
sCpuScriptGroup.h
sCpuScriptGroup2.cpp
sCpuScriptGroup2.h
|
| 958d8b23ac969d13ea3da0a2d9a355f5951afa8c |
12-Mar-2015 |
Yang Ni <yangni@google.com> |
Merge "Handle FP precision in kernel fusion"
|
| dced5c96bc53c45a1aac782ea9bd738b0d50fd09 |
12-Mar-2015 |
David Gross <dgross@google.com> |
Add new APIs rsGetDim[XYZ]() and add RSTest test cases to exercise them.
Test UT_kernel3d does not pass reliably, so it is disabled for now.
Tests will not compile without related
platform/frameworks/compile/libbcc and
platform/frameworks/compile/slang changes.
Bug: 18964628
Change-Id: I39660400953306ceb47c8b5e9665e8f0084775dd
sCpuCore.h
sCpuCoreRuntime.h
sCpuScriptGroup2.cpp
sCpuScriptGroup2.h
|
| edf4ea312cc3f7dd4373f8db5aaf9325ff054c8e |
11-Mar-2015 |
Yang Ni <yangni@google.com> |
Handle FP precision in kernel fusion
b/19098612
Find the proper core lib for full-precision and relaxed-precions
FP. Path both to bcc while compiling a script group, and let bcc
pick the right one based on the precision of the merged module.
Change-Id: I2a641387f0990463887594729a935a5c3f0f856f
sCpuScriptGroup2.cpp
|
| 062c287f573ecc06c38ee4295e5627e12c52ac3d |
21-Feb-2015 |
Yang Ni <yangni@google.com> |
Runtime support for Script Group as single module
Also made two other cleanups:
- Changed KernelID/InvokeID into IDBase in class Closure
Rather than having two fields in class Closure, one of type ScriptKernelID and
the other InovkeID, use a single field of the common base class IDBase. Added a
boolean field to indicate whether it is kernel or invoke, since -fno-rtti is on.
- Removed user pointer from CPU closure
Change-Id: I5553f86b2e58325f85649078d48685a38f12d62f
sCpuExecutable.cpp
sCpuExecutable.h
sCpuScript.cpp
sCpuScript.h
sCpuScriptGroup2.cpp
sCpuScriptGroup2.h
|
| 506cb4f10c3496864b111b335aac43e154e8f646 |
04-Mar-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
am ad1930cf: Merge "Add build fingerprint and ensure match on .so load"
* commit 'ad1930cf38d6a0a0e97f6ccacd7ecae96504f831':
Add build fingerprint and ensure match on .so load
|
| ad1930cf38d6a0a0e97f6ccacd7ecae96504f831 |
04-Mar-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
Merge "Add build fingerprint and ensure match on .so load"
|
| aa6757ffc1b23d771566439c3179fdbc1e5ba569 |
14-Feb-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
Add build fingerprint and ensure match on .so load
bug 19216507
If ro.debuggable system property is set,
- Compute the Adler32 checksum of bitcode, compiler arguments, bcc,
libclcore files.
- On shared object load, check that there is an embedded checksum value
in .rs.info and it matches the value computed in the previous step.
The .so is not loaded if the checksums don't match.
- If invoking bcc, pass the checksum as an argument.
If ro.debuggable is not set, a constant checksum is added to .rs.info.
But it is never checked.
Change-Id: I0a9d16b34148b748daf744007a8515511714a5a5
ndroid.mk
sCpuExecutable.cpp
sCpuExecutable.h
sCpuScript.cpp
sCpuScript.h
|
| 8bce1b68212210f7568590177635bfc71edd4368 |
24-Feb-2015 |
Yabin Cui <yabinc@google.com> |
am 3e5318a3: Merge "Move use of tempnam to mkstemp."
* commit '3e5318a36be470ba7a8c5cf82bbe069178733b11':
Move use of tempnam to mkstemp.
|
| 3e5318a36be470ba7a8c5cf82bbe069178733b11 |
24-Feb-2015 |
Yabin Cui <yabinc@google.com> |
Merge "Move use of tempnam to mkstemp."
|
| 433558f0f9abbf07770db288183a15fd261cace2 |
24-Feb-2015 |
Yabin Cui <yabinc@google.com> |
Move use of tempnam to mkstemp.
Bug: 19340053
Change-Id: I1734e085beef395ebcc6453af1efaa19f7409e22
sCpuScriptGroup2.cpp
|
| 8186f5ca9bf4bda113edeece558b4049732b3ed5 |
20-Feb-2015 |
Yang Ni <yangni@google.com> |
am 405710bf: Merge "Moved class ScriptExecutable to its own module"
* commit '405710bfdbb988764fe1dd65ac13352b8858848c':
Moved class ScriptExecutable to its own module
|
| 2abfcc6d129fe3defddef4540aa95cc445c03a7a |
18-Feb-2015 |
Yang Ni <yangni@google.com> |
Moved class ScriptExecutable to its own module
b/19283946
This move also included the related class SharedObjectUtils.
Change-Id: Iefd587b659c4ad99340c330955b41cd92df45563
ndroid.mk
sCpuExecutable.cpp
sCpuExecutable.h
sCpuScript.cpp
sCpuScript.h
sCpuScriptGroup2.cpp
|
| e9d5ce1d5cc056e5d5d2685a68ebf1318ba93559 |
20-Feb-2015 |
Yang Ni <yangni@google.com> |
am e15ea72b: Merge "Remove STL references from header files"
* commit 'e15ea72be26441644c21d536f18525702a24d28a':
Remove STL references from header files
|
| 5b15e9248ea9740d3651f1aff4bd9eae21cfd2c7 |
20-Feb-2015 |
Yang Ni <yangni@google.com> |
am e9c79f66: Merge "Revert "Replaced android::Vector with std::vector.""
* commit 'e9c79f66478170e4e45e49c71f5e9324842916cf':
Revert "Replaced android::Vector with std::vector."
|
| c31585b8ca865bf2b35abc79c8a8ee42de27bee8 |
15-Feb-2015 |
Yang Ni <yangni@google.com> |
Remove STL references from header files
b/19148482
Replaced std::string with android::String8, and std::pair with
android::renderscript::Pair, in header files.
Change-Id: I20ac1a3b7a019e5eee12fac46ad6adf64d850459
sCpuCore.h
sCpuScript.cpp
sCpuScript.h
|
| e9c79f66478170e4e45e49c71f5e9324842916cf |
19-Feb-2015 |
Yang Ni <yangni@google.com> |
Merge "Revert "Replaced android::Vector with std::vector.""
|
| 6673c80dc61df9dbc576034ee83f07a077c57c54 |
19-Feb-2015 |
Bill Yi <byi@google.com> |
Merge commit '5d459b53fbc1ad6e52940c358d009393945c4f45' into HEAD
|
| 5c061d10fe1e5d4ae20715b0ca4e92fba12f82b8 |
19-Feb-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
Merge "Move bcc::getCommandLine to fw/rs"
|
| 508b1afdff080ccb8546021649559fc5c720aecd |
19-Feb-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
Move bcc::getCommandLine to fw/rs
libbcc had a utility function to concatenate string arguments. This
patch moves that function to frameworks/rs. The function will no longer
be needed in libbcc once we move all the checksum/cache invalidation
logic to frameworks/rs.
Change-Id: I2b1c15c2763d90d7b099e41257bb72f200680943
sCpuScript.cpp
|
| 64c682b65cd04ac83b51251b40dca14423df351a |
09-Jan-2015 |
Tim Murray <timmurray@google.com> |
Add BLAS to supported intrinsics.
Change-Id: I8e776b2ffdbac09a73924035eee2eca0a12facb3
ndroid.mk
sCpuCore.cpp
sCpuIntrinsicBLAS.cpp
|
| b8353c5943f4038fd7f08db3d958390ce9418798 |
15-Feb-2015 |
Yang Ni <yangni@google.com> |
Revert "Replaced android::Vector with std::vector."
b/19148482
This reverts commit 93d6bc872b7d9fba63abfa7513d56b38d9c3d371.
I also made some changes to Make it work with HEAD.
Change-Id: I5e516976ec1d85ffe1cf388c01795348a9441982
sCpuScriptGroup.cpp
|
| 9e5a1bdb972e296c83b4afd579ad21e7524ecd07 |
13-Feb-2015 |
Yang Ni <yangni@google.com> |
am 1ce0dfd8: Merge "Reformat and remove STL reference from header"
* commit '1ce0dfd82311d4922dbffdde302599410fc25c5d':
Reformat and remove STL reference from header
|
| ff2bb54ebf593b1d19d3a2e4cfa70a8ea4432c0d |
02-Feb-2015 |
Yang Ni <yangni@google.com> |
Reformat and remove STL reference from header
b/19148482
Reformmated according to Android C++ style guidelines.
Removed STL references in headers.
Change-Id: I6d82b8fe5ac868067b6d9ebe797125feb97e5641
sCpuScriptGroup2.cpp
sCpuScriptGroup2.h
|
| 2dc2c22d95197716317deb576e68a69eba80dbfe |
05-Feb-2015 |
Yang Ni <yangni@google.com> |
am 841ba73b: Merge "Runtime support for invocables in ScriptGroups"
* commit '841ba73b843babb8ca0117e9801b6db6ba6cd394':
Runtime support for invocables in ScriptGroups
|
| 28e2a30f118a52d87399bf1a961f3ea4f440fce8 |
05-Feb-2015 |
Miao Wang <miaowang@google.com> |
am fe4f02f8: Merge "Make rsContext be able to have info about the path of native libs for 64bit Renderscript compat lib."
* commit 'fe4f02f81d7c461835be97eebbde744885846e99':
Make rsContext be able to have info about the path of native libs for 64bit Renderscript compat lib.
|
| eb9aa675754c49f613c6ad71d41472b30f38b007 |
27-Jan-2015 |
Yang Ni <yangni@google.com> |
Runtime support for invocables in ScriptGroups
This also includes support for InvokeID
Change-Id: I5b59df166ea30b309b8dd9623825ac0e72d03856
sCpuScript.cpp
sCpuScriptGroup2.cpp
sCpuScriptGroup2.h
|
| fe4f02f81d7c461835be97eebbde744885846e99 |
05-Feb-2015 |
Miao Wang <miaowang@google.com> |
Merge "Make rsContext be able to have info about the path of native libs for 64bit Renderscript compat lib."
|
| f3213d7fd648da98bb3b03204eaf90f03c31926b |
14-Jan-2015 |
Miao Wang <miaowang@google.com> |
Make rsContext be able to have info about the path of native libs
for 64bit Renderscript compat lib.
- Native lib path stored in Java and queried from JNI.
- Then set to Context during context creation.
Change-Id: I3c997cf849efb7b9a0b0ac35d5b62289d70e7434
sCpuScript.cpp
sCpuScript.h
|
| 0371d55d837206f0d94c269a0928e2e2d257cbe0 |
03-Feb-2015 |
Yang Ni <yangni@google.com> |
am cd32123c: Merge "Runtime support for compiler kernel fusion."
* commit 'cd32123c0268d67d5509774576c24e299ea69fd7':
Runtime support for compiler kernel fusion.
|
| da0f069871343119251d6b0586be356dc2146a62 |
12-Jan-2015 |
Yang Ni <yangni@google.com> |
Runtime support for compiler kernel fusion.
The runtime will start a seperate process to call the new bcc to fuse kernels.
Change-Id: Ia73ea917a126a5055ec97f13d90a5feaafd6a2f5
sCpuScript.cpp
sCpuScript.h
sCpuScriptGroup2.cpp
sCpuScriptGroup2.h
|
| eea118cd0f7eb18c45d37f88178ecbd821ec97e8 |
02-Feb-2015 |
Yang Ni <yangni@google.com> |
am f5e39bba: Merge "Remove STL references from header files"
* commit 'f5e39bba1d0e7be4adfa7364d92ba3f541420fdd':
Remove STL references from header files
|
| e8f9fba78f0cb79fa8773373a635e30382113a75 |
30-Jan-2015 |
Yang Ni <yangni@google.com> |
Remove STL references from header files
b/19148482
Change-Id: I2138664d8ef51d15da23be69ac7d224f30224f42
sCpuScript.cpp
sCpuScript.h
|
| 9031b4b30126c62985d0a9f057fd93ff96145f74 |
29-Jan-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
am 104c694d: Merge "Mark scripts as threadable or not"
* commit '104c694d67a224adda7b673de04b763bc63d6c3f':
Mark scripts as threadable or not
|
| 68173de001c5d41c74828a426651e07cfd9e8710 |
28-Jan-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
Mark scripts as threadable or not
bug 19095896
Read embedded flag in .rs.info and use it to mark Scripts as threadable.
In compat lib path, scripts are always threadable.
Change-Id: I85a3753ffae788dbdec5fd130c20581790c4ee07
sCpuScript.cpp
sCpuScript.h
|
| 9a9146656a8355b2b5fb6b308809760e8f03e613 |
27-Jan-2015 |
Jason Sams <jsams@google.com> |
am f9630165: Merge "add array launch support."
* commit 'f963016579f1d244b932788e7ae4276e1a51944c':
add array launch support.
|
| bf2111d3b3de310932099514f06924e48fa1d7b2 |
27-Jan-2015 |
Jason Sams <jsams@google.com> |
add array launch support.
Change-Id: I66cd89b5b44eafa92f391708a06464cd7cdde3ed
sCpuCore.cpp
sCpuCore.h
sCpuIntrinsic.cpp
sCpuScript.cpp
sCpuScript.h
sCpuScriptGroup.cpp
sCpuScriptGroup2.cpp
|
| e47202dde0f0fce8b16a1cb202ebbd4d2b4917b2 |
26-Jan-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
am a309ef08: Merge "Read pragma keys and values from .rs.info"
* commit 'a309ef08865aeccd5575b7e4f727936bb6c18aab':
Read pragma keys and values from .rs.info
|
| a309ef08865aeccd5575b7e4f727936bb6c18aab |
26-Jan-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
Merge "Read pragma keys and values from .rs.info"
|
| 577194ac9c2bf10f31e564de91371764b265929a |
23-Jan-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
Read pragma keys and values from .rs.info
bug 18322681
Pragmas were not embedded into shared objects in the compat lib path
because they aren't used. Along with the sibling change in libbcc, this
patch stores and retrieves pragmas from the .rs.info varible embedded in
shared object files.
Change-Id: Iebbf9a79e715a5a4a1ee0962a08386fa78516c6b
sCpuScript.cpp
sCpuScript.h
|
| dd403b1ebf2961efdf53d73db2dabaf182e2ae7d |
24-Jan-2015 |
Tim Murray <timmurray@google.com> |
am 1ad7abaf: Merge "Always recompile a script when using the debug context."
* commit '1ad7abafe1f5672c81fa4509ce0fda9a3cfdb376':
Always recompile a script when using the debug context.
|
| bf96a520e162b503b293f65ca9ca9be66d4a0036 |
24-Jan-2015 |
Tim Murray <timmurray@google.com> |
Always recompile a script when using the debug context.
Change-Id: Icea9834805702c20c250d2dec48cdf8f11a0acfb
sCpuScript.cpp
|
| e9e6f97ab2cfe6ab5618d0766b389043c70ddeb0 |
23-Jan-2015 |
Yang Ni <yangni@google.com> |
am e35c5288: Merge "Made shared lib routines public."
* commit 'e35c5288c3fb145f8d423b59bdf96df473e055eb':
Made shared lib routines public.
|
| 1c44cb6b1133730cd18929d637e71e9326c70bd0 |
22-Jan-2015 |
Yang Ni <yangni@google.com> |
Made shared lib routines public.
So that we can reuse the shared library create and loading routines
for script groups.
Change-Id: I9f35d8efc4be686c92141ab770f78f35ea3a02c1
sCpuScript.cpp
sCpuScript.h
|
| 14398352d56ce20b59605485489aa037fa0f20d3 |
22-Jan-2015 |
Yang Ni <yangni@google.com> |
am b46b67af: Merge "Created a new class to represent executable."
* commit 'b46b67af3053200e88259a3b0ad1275689e1b679':
Created a new class to represent executable.
|
| d9bae689c1b8c3f2ed1a5f2b374dc9393584b8dd |
21-Jan-2015 |
Yang Ni <yangni@google.com> |
Created a new class to represent executable.
The new class represents executables created from shared objects,
which allows query exported fields, for each, etc.
This allows this functionality to be shared with other part of the
system besides CPU script, e.g., script groups.
Change-Id: I5223c329bdb70085eb1dce7aab5ec91626f579f4
sCpuScript.cpp
sCpuScript.h
|
| bb0b068bb8e0b1e23844f7e6199c43b157ec4328 |
22-Jan-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
am 24c1f31c: Merge "Do not check for .so in /system in native mode"
* commit '24c1f31c6f03ffffdb08ee2c1c309edbc15b035f':
Do not check for .so in /system in native mode
|
| f0558cce29611fdba9fdd50d39785f53d95fe1d5 |
21-Jan-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
Do not check for .so in /system in native mode
Bug: 18322681
In native mode, the .so file will be present only in the app specific
cache directory. Any match in /system/lib is merely due to name clash
with another app. Shared libraries in /system/lib are relevant only in
compatibility mode.
Change-Id: I2713820c76b63bd69f726e50f0a39974661c05f1
sCpuScript.cpp
|
| e02d1253d7c326ce266a0e693d77aec30f4c5c48 |
21-Jan-2015 |
Stephen Hines <srhines@google.com> |
am e3aa5ab8: Merge "Remove linkloader from frameworks/rs."
* commit 'e3aa5ab8ddba9931e2db6796a5b93e3faed24f41':
Remove linkloader from frameworks/rs.
|
| 45e753a46e587c69b3b0d0c5138e88715a24a29a |
20-Jan-2015 |
Stephen Hines <srhines@google.com> |
Remove linkloader from frameworks/rs.
Bug: 18322681
Now that we have switched completely to the new shared library object
loading path, we can safely remove the legacy linkloader path.
In frameworks/rs, this removes the actual linkloader code, as well as
all helper calls into linkloader. This change also coalesces code paths
between the support library version of RS and the native version of RS,
since they both now depend on a similar shared library loader. A missing
call to dlclose() on Script teardown is also added on the native library
path.
Change-Id: Ie5cc152d93f5e75383f7c21a4523579cfae8823f
inkloader/android/librsloader.cpp
inkloader/android/librsloader.h
inkloader/android/test-librsloader.c
inkloader/include/ELF.h
inkloader/include/ELFHeader.h
inkloader/include/ELFObject.h
inkloader/include/ELFReloc.h
inkloader/include/ELFSection.h
inkloader/include/ELFSectionBits.h
inkloader/include/ELFSectionHeader.h
inkloader/include/ELFSectionHeaderTable.h
inkloader/include/ELFSectionNoBits.h
inkloader/include/ELFSectionProgBits.h
inkloader/include/ELFSectionRelTable.h
inkloader/include/ELFSectionStrTab.h
inkloader/include/ELFSectionSymTab.h
inkloader/include/ELFSymbol.h
inkloader/include/ELFTypes.h
inkloader/include/GOT.h
inkloader/include/MemChunk.h
inkloader/include/StubLayout.h
inkloader/include/impl/ELFHeader.hxx
inkloader/include/impl/ELFObject.hxx
inkloader/include/impl/ELFReloc.hxx
inkloader/include/impl/ELFSection.hxx
inkloader/include/impl/ELFSectionBits.hxx
inkloader/include/impl/ELFSectionHeader.hxx
inkloader/include/impl/ELFSectionHeaderTable.hxx
inkloader/include/impl/ELFSectionNoBits.hxx
inkloader/include/impl/ELFSectionProgBits.hxx
inkloader/include/impl/ELFSectionRelTable.hxx
inkloader/include/impl/ELFSectionStrTab.hxx
inkloader/include/impl/ELFSectionSymTab.hxx
inkloader/include/impl/ELFSymbol.hxx
inkloader/include/mmanWindows.h
inkloader/lib/ELFHeader.cpp
inkloader/lib/ELFSectionHeader.cpp
inkloader/lib/ELFSymbol.cpp
inkloader/lib/ELFTypes.cpp
inkloader/lib/GOT.cpp
inkloader/lib/MemChunk.cpp
inkloader/lib/StubLayout.cpp
inkloader/lib/mmanWindows.cpp
inkloader/main.cpp
inkloader/utils/flush_cpu_cache.h
inkloader/utils/helper.cpp
inkloader/utils/helper.h
inkloader/utils/raw_ostream.cpp
inkloader/utils/raw_ostream.h
inkloader/utils/rsl_assert.cpp
inkloader/utils/rsl_assert.h
inkloader/utils/serialize.h
inkloader/utils/traits.h
sCpuIntrinsicColorMatrix.cpp
sCpuScript.cpp
sCpuScript.h
sd_cpu.h
|
| 50d707b073cf0c3d73c3dca7e050dcabf295ea84 |
21-Jan-2015 |
Jason Sams <jsams@google.com> |
am 365357ac: Merge "Cleanup of ForEachParams in cpu ref"
* commit '365357ac91d52c385e5af38ca6fe24d853f0ddb1':
Cleanup of ForEachParams in cpu ref
|
| c0d68470b978a79ce024fde56f23ea3690603ccd |
20-Jan-2015 |
Jason Sams <jsams@google.com> |
Cleanup of ForEachParams in cpu ref
Change-Id: I8cc51915b2a605c240d98e3010619b741a13bae2
sCpuCore.cpp
sCpuCore.h
sCpuScript.cpp
|
| e9d9dab68add83b93b2493b277adc1168711334c |
16-Jan-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
am 5c6002cc: Merge "Turn on .so path by default"
* commit '5c6002cc19cc3f1beffd204fd6cbca8bd75c0012':
Turn on .so path by default
|
| b5215a52df4402a8a513c3dc65d06b9bd46e879d |
16-Jan-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
Turn on .so path by default
Bug: 18322681
If there are problems in the .so loading path, you can turn it off by
having is_skip_linkloader() return false.
Change-Id: I31ae2d8a65f70005f4d749f9ecbe55ae6f153d8c
sCpuScript.cpp
|
| 94c0d6903ad66299bc3967db7c3a68df2e30d7ea |
15-Jan-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
am 3e3fbed5: Merge "Skip linkloader, use shared object files"
* commit '3e3fbed52a47a9e3118f9df6bab6d45859138520':
Skip linkloader, use shared object files
|
| 22d158c258f6ae55fb2666331c8f5de94b77451c |
15-Jan-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
am e8151ab5: Merge "Fix ilogb\'s return type"
* commit 'e8151ab5a9cbcc875cdce64b84287650418922e3':
Fix ilogb's return type
|
| dc0d8f7c0f1f43f25c34fbc04656ad578f6e953b |
03-Dec-2014 |
Pirama Arumuga Nainar <pirama@google.com> |
Skip linkloader, use shared object files
Bug: 18322681
- In rsCpuScript, if property rs.skip.linkloader is set, look for a .so
file in the cache directory and load it. If it is not available, use
bcc to generate relocatable object file and link it to a .so using
ld.mc. Use the embedded symbols in .rs.info and follow steps similar
to the compatibility library to invoke script functions or access
script variables.
- Add rs* symbols like rsGetAllocation to libRSCpuRef (ala
libRSSupport). Do necessary changes to argument types to get mangled
names correct.
- Make 64-bit version of rsSetObject take two pointers instead of a
pointer and a large object. rsIsObject takes a pointer instead of a
large object. Otherwise, we get failures in x86_64 due to calling
convention mismatch. To match the function names in the shared object
path, define these functions as 'extern "C"' with their mangled names.
- Add stubbed Math functions from rsCpuRuntimeMath and
rsCpuRuntimeMathFuncs into libRSCpuRef.so.
- Coalesce separate #ifdef paths in libRSCpuRef. Function parameters
for runtime callbacks and bcc plugin are needed in the
non-compatibilty path, but take default NULL arguments. This patch
introduces these parameters into the compatibility path as well, and
passes default NULL arguments.
Change-Id: I8a853350e39d30b4d852c30e4b5da5a75a2f2820
ndroid.mk
sCpuCore.cpp
sCpuCore.h
sCpuRuntimeMath.cpp
sCpuRuntimeMathFuncs.cpp
sCpuScript.cpp
sCpuScript.h
sd_cpu.h
|
| 6fdd0601e45aaf7fd6eaa3ec7db0e981d4f8c5dd |
13-Jan-2015 |
Pirama Arumuga Nainar <pirama@google.com> |
Fix ilogb's return type
Fix ilobg's return type in compat mode - was float instead of an int.
Change-Id: I523248fd1e2a867f3f8dbb0772771ef51f76ff53
sCpuRuntimeMath.cpp
sCpuRuntimeMathFuncs.cpp
|
| 650644594921f109c081127d45b6afaedf6df7aa |
09-Jan-2015 |
Tim Murray <timmurray@google.com> |
am db5c7eef: Merge "Enable autovectorization in CPU driver."
* commit 'db5c7eef1b1ae60e94932ce19628b342c1ed2727':
Enable autovectorization in CPU driver.
|
| 687cfe85c3cffb244a7e5daeb1231f441a1ed54d |
08-Jan-2015 |
Tim Murray <timmurray@google.com> |
Enable autovectorization in CPU driver.
Change-Id: Ibc10c61dee8d071256a9ebd6c4c47cb7dc617284
sCpuScript.cpp
|
| 3f28e7d2c69a7f89d3af6b9eeb10018581929e9e |
08-Jan-2015 |
Yang Ni <yangni@google.com> |
am 015ba466: Merge "New Script Group API: runtime and cpu driver support."
* commit '015ba466dff2c05653b21afded83d311d35cfa7c':
New Script Group API: runtime and cpu driver support.
|
| 015ba466dff2c05653b21afded83d311d35cfa7c |
08-Jan-2015 |
Yang Ni <yangni@google.com> |
Merge "New Script Group API: runtime and cpu driver support."
|
| 5b9ad7995e7f0e3414067de4fa7131188a685a2e |
07-Jan-2015 |
Tim Murray <timmurray@google.com> |
am 1ca22d17: Merge "Use integrated-as on Darwin"
* commit '1ca22d17f714c2ebae5f0b06ba6fa7dee55bb0b9':
Use integrated-as on Darwin
|
| 1ca22d17f714c2ebae5f0b06ba6fa7dee55bb0b9 |
07-Jan-2015 |
Tim Murray <timmurray@google.com> |
Merge "Use integrated-as on Darwin"
|
| 04fe99fa8f4daef507e52e04eb8c6c2a37a4827c |
17-Dec-2014 |
Tim Murray <timmurray@google.com> |
Use integrated-as on Darwin
Change-Id: I2a888713285fa5b435c6a54a3a4a7bb33130458d
ndroid.mk
|
| 1ffd86b448d78366190c540f98f8b6d641cdb6cf |
07-Jan-2015 |
Yang Ni <yangni@google.com> |
New Script Group API: runtime and cpu driver support.
Change-Id: I9c612cf8874aabaf0ca7d1640567464c71ed3070
ndroid.mk
sCpuCore.cpp
sCpuCore.h
sCpuScriptGroup.cpp
sCpuScriptGroup.h
sCpuScriptGroup2.cpp
sCpuScriptGroup2.h
sd_cpu.h
|
| b6370d17eaa77cc365a471b4410e4fbde9d6a159 |
06-Jan-2015 |
Jason Sams <jsams@google.com> |
am 9cc5708a: Merge "Align ScriptInvokeV parameters to 16 bytes for x86_64"
* commit '9cc5708a0bf7a81551ccad8ab5c5abe23ec1b853':
Align ScriptInvokeV parameters to 16 bytes for x86_64
|
| 9cc5708a0bf7a81551ccad8ab5c5abe23ec1b853 |
06-Jan-2015 |
Jason Sams <jsams@google.com> |
Merge "Align ScriptInvokeV parameters to 16 bytes for x86_64"
|
| dff96fef1c2ecb7770342964008670dba00f360c |
18-Dec-2014 |
Stephen Hines <srhines@google.com> |
am c837d22b: Merge "BCCContext is actually dead code."
* commit 'c837d22b93b0e0f5854a523d14956fc4a43a26a8':
BCCContext is actually dead code.
|
| 584e58bb75a4d742e3a9dfcfea36eba59b38dbd9 |
17-Dec-2014 |
Stephen Hines <srhines@google.com> |
BCCContext is actually dead code.
This should have been removed as part of the switchover to standalone bcc.
Change-Id: Icbb9a4c064dd9f002f9acfa1ac7748745501de2b
sCpuScript.cpp
sCpuScript.h
|
| eaba5a3ca215729258dcf9ac6f0bb5f88c78f998 |
12-Dec-2014 |
Yong Chen <yong.a.chen@intel.com> |
Align ScriptInvokeV parameters to 16 bytes for x86_64
Change-Id: I0c9aea220b12f364c0a1a90232fbaa40ad90aeec
Signed-off-by: Yong Chen <yong.a.chen@intel.com>
sCpuScript.cpp
|
| 8c5f5fdd8e00dd0fb364eec505d6b78a28f2fa4c |
10-Dec-2014 |
Tim Murray <timmurray@google.com> |
am 79d1df90: Merge "Enable A53 erratum workaround in bcc."
* commit '79d1df9063189cb2cad3643f78423dbf8f99bbe7':
Enable A53 erratum workaround in bcc.
|
| 79d1df9063189cb2cad3643f78423dbf8f99bbe7 |
10-Dec-2014 |
Tim Murray <timmurray@google.com> |
Merge "Enable A53 erratum workaround in bcc."
|
| 358ffb84f56929cd2d61f6429a790a1606eaf865 |
09-Dec-2014 |
Tim Murray <timmurray@google.com> |
Enable A53 erratum workaround in bcc.
bug 18683922
Change-Id: I33636dfa74605211c4898cc08fcee6e0f77978e5
ndroid.mk
sCpuScript.cpp
|
| 4d61270b1f08cbaf3fca940047bb30617831fece |
09-Dec-2014 |
Elliott Hughes <enh@google.com> |
am 4041272a: Merge "Remove unused #include."
* commit '4041272ab5cebc81f71c0f75a952547a0138ef2a':
Remove unused #include.
|
| f75a0f094a6116aaeda1b2154e1b6f9a44f5bf34 |
09-Dec-2014 |
Elliott Hughes <enh@google.com> |
Remove unused #include.
Bug: 18556103
Change-Id: I5af65a0211bf6dddefeb35f7b156590c2517ef21
sCpuIntrinsics_neon_ColorMatrix.S
|
| fdceadb811ec22c69f879ea0d0108be3d287708b |
18-Nov-2014 |
Miao Wang <miaowang@google.com> |
am 4283f579: Fix rounding bug of IntrinsicConvolve in cpu_ref
* commit '4283f579c424f07bc07c7f075398053eed3f8281':
Fix rounding bug of IntrinsicConvolve in cpu_ref
|
| 4283f579c424f07bc07c7f075398053eed3f8281 |
17-Nov-2014 |
Miao Wang <miaowang@google.com> |
Fix rounding bug of IntrinsicConvolve in cpu_ref
bug 18121051
Change-Id: Ica3e32731a7f78c8869b84e1e257216a52f79f8a
sCpuIntrinsicConvolve3x3.cpp
sCpuIntrinsicConvolve5x5.cpp
|
| ca3331c4984482c764561b82e00f451293722cf7 |
14-Nov-2014 |
Miao Wang <miaowang@google.com> |
Merge "Add Float32 support for IntrinsicResize."
|
| 31a3d42110a82c7c357fa9e6e8871f045d545585 |
14-Nov-2014 |
Stephen Hines <srhines@google.com> |
Merge "[MIPS] Add MIPS64 R6 relocations"
|
| d75cc84d23f372e55f7173c727129cd5de00748b |
14-Nov-2014 |
Miao Wang <miaowang@google.com> |
Add Float32 support for IntrinsicResize.
Change-Id: I1b1da92f58a11d9cb6870f4d117fe4ed80207a3b
sCpuIntrinsicResize.cpp
|
| a2bd5e85ddb7c4cc439f4b4646dafa21558ea5c7 |
11-Nov-2014 |
Miao Wang <miaowang@google.com> |
Fix off-by-one bug of resize, when scale = 1;
Use a better way to do scaling.
bug 18296081
Change-Id: I3858d6d0f1cd0166d9862e9013386b1d844bd3f9
sCpuIntrinsicResize.cpp
|
| 51d016c72e44c9c6da47c53665033ce8816ab463 |
11-Nov-2014 |
Miao Wang <miaowang@google.com> |
Fix off-by-one bug of resize, when scale = 1;
Use a better way to do scaling.
bug 18296081
Change-Id: I3858d6d0f1cd0166d9862e9013386b1d844bd3f9
sCpuIntrinsicResize.cpp
|
| ef05d4666eb87a924c8883e193fd505245101414 |
24-Oct-2014 |
Miao Wang <miaowang@google.com> |
Fix the rounding bug of resize test and impl; Add mirror test.
Change-Id: Ibfcc942ae3126d0186556f80dc5af17f77060b04
sCpuIntrinsicResize.cpp
|
| 77d57a305f4134e78ebc91869011c4009988104e |
24-Oct-2014 |
Jason Sams <jsams@google.com> |
Fix query for CPU count.
Some devices report fewer processors online versus
configured. Always get the configured (higher) number
bug 18108290
Change-Id: Ic6202e05ad8c4686dd79795f880baf5429674d70
sCpuCore.cpp
|
| 32f9d04ae7b5f680c0921b3f9d4cdbf1665532b3 |
23-Oct-2014 |
Jason Sams <jsams@google.com> |
Enable 64 bit asm for int->int ColorMatrix
This brings 64 bit perf close to 32bit for the common
color matrix paths.
bug 17923388
Change-Id: I96c2c28a07864bf71c6c4a0186fb76f4a946316f
sCpuIntrinsicColorMatrix.cpp
|
| bfc23288e830aa3689d24f803561d98174c524e3 |
22-Oct-2014 |
Jason Sams <jsams@google.com> |
Enable ASM path for blur intrinsic when clipped
Fix minor clipping bug in ASM code where the start offset was
added twice.
Change-Id: I6d831478b4a7da8460e70015151dbadf16bd7096
sCpuIntrinsicBlur.cpp
sCpuIntrinsics_advsimd_Blur.S
sCpuIntrinsics_neon_Blur.S
|
| 319babf8106540bf26ac5e76a85416846bae8d3b |
09-Oct-2014 |
Jason Sams <jsams@google.com> |
Fix 64bit YUV intrinsic performance
bug 17923388
Change-Id: I9ee9ebf05eb5d3b661d8199864f92c14f0d0191d
sCpuIntrinsicYuvToRGB.cpp
|
| c886b1f69987151f29b3d5d27959c8b5aae4ea19 |
24-Oct-2014 |
Miao Wang <miaowang@google.com> |
Merge "Fix the rounding bug of resize test and impl; Add mirror test." into lmp-mr1-dev
|
| da68eff288c8b403bc41c1fa856ca09d73f35e2f |
24-Oct-2014 |
Miao Wang <miaowang@google.com> |
Fix the rounding bug of resize test and impl; Add mirror test.
Change-Id: Ibfcc942ae3126d0186556f80dc5af17f77060b04
sCpuIntrinsicResize.cpp
|
| 43641e457d1c8aff9e8bc9febdded5261053abc6 |
24-Oct-2014 |
Jason Sams <jsams@google.com> |
Fix query for CPU count.
Some devices report fewer processors online versus
configured. Always get the configured (higher) number
bug 18108290
Change-Id: Ic6202e05ad8c4686dd79795f880baf5429674d70
sCpuCore.cpp
|
| ff983194dbb0504f4a341487aa20bb08027a52b6 |
23-Oct-2014 |
Jason Sams <jsams@google.com> |
Enable 64 bit asm for int->int ColorMatrix
This brings 64 bit perf close to 32bit for the common
color matrix paths.
bug 17923388
Change-Id: I96c2c28a07864bf71c6c4a0186fb76f4a946316f
sCpuIntrinsicColorMatrix.cpp
|
| 4c8f2477285848ab0a4f33ad854de9398d332e8c |
22-Oct-2014 |
Jason Sams <jsams@google.com> |
Enable ASM path for blur intrinsic when clipped
Fix minor clipping bug in ASM code where the start offset was
added twice.
Change-Id: I6d831478b4a7da8460e70015151dbadf16bd7096
sCpuIntrinsicBlur.cpp
sCpuIntrinsics_advsimd_Blur.S
sCpuIntrinsics_neon_Blur.S
|
| 5eb463f2c28cf161fe31b8078908a7b01198516f |
20-Sep-2014 |
Simon Hosie <simon.hosie@arm.com> |
Subrectangle bug fixes to ARM Blur assembly.
Correctly sweep the ends of the initial convolution window where that window
overlaps one or both sides of the source image.
Change-Id: I4e3d2bfa7eb22ce29af2615bf324a69561f7e4d6
sCpuIntrinsics_advsimd_Blur.S
sCpuIntrinsics_neon_Blur.S
|
| 9107db900edec122b8f6f49a81ab444d8a6c9c42 |
20-Sep-2014 |
Simon Hosie <simon.hosie@arm.com> |
Minor fixes to AArch64 Blur assembly.
Use the correct register for a low-weighted tap in uchar1 case, and use the
correct clipped radius calculation for the right-hand edge.
Change-Id: Ib6ecd8b115c8898bb641958ab0beab11a8fccc36
sCpuIntrinsics_advsimd_Blur.S
|
| c10c49404a35384be5fe44bd04b6eed1ffc6a98c |
22-Oct-2014 |
Bill Yi <byi@google.com> |
Merge commit '6a95b944f23e7c6a8fc3a6fe42350df4fad7d27d' into HEAD
|
| b4d28c50d522e9e396b6aad7cf63debe51fd00a3 |
22-Oct-2014 |
Jason Sams <jsams@google.com> |
Merge "Subrectangle bug fixes to ARM Blur assembly."
|
| 81954efe716f4a14ffb8fd1e0bd16681d64744fc |
22-Oct-2014 |
Jason Sams <jsams@google.com> |
Merge "Minor fixes to AArch64 Blur assembly."
|
| d675c704622b7291f6d7245cb50ea4462ec4943d |
22-Oct-2014 |
Jason Sams <jsams@google.com> |
Merge "Expand AArch64 YuvToRGB to use more registers."
|
| 772b306f2d7ca994fe35d7ed4a2e0318a2d9fbf7 |
09-Oct-2014 |
Jason Sams <jsams@google.com> |
Fix 64bit YUV intrinsic performance
bug 17923388
Change-Id: I9ee9ebf05eb5d3b661d8199864f92c14f0d0191d
sCpuIntrinsicYuvToRGB.cpp
|
| e8814f7c80f84f08e60150e70b1a4e6a11b588bd |
19-Jun-2014 |
Simon Hosie <simon.hosie@arm.com> |
Expand AArch64 YuvToRGB to use more registers.
bug 17923388
Change-Id: Ib8ebea344ff863bb4c6e13f11efae3dd50f7c0fa
sCpuIntrinsics_advsimd_YuvToRGB.S
|
| c5e6b521e0782158adc2acba4c8691ade44578fb |
21-Oct-2014 |
Jason Sams <jsams@google.com> |
Merge "Tail-store fixes to AArch64 YuvToRGB assembly." into lmp-mr1-dev
|
| ccd2125b76fb04fd4676e9282d6f9d72e87950e6 |
17-Oct-2014 |
Petar Jovanovic <petar.jovanovic@imgtec.com> |
[MIPS] Add MIPS64 R6 relocations
This patch adds 6 MIPS64 R6 relocations:
- R_MIPS_PC21_S2
- R_MIPS_PC26_S2
- R_MIPS_PC18_S3
- R_MIPS_PC19_S2
- R_MIPS_PCHI16
- R_MIPS_PCLO16
It also adds missing R_MIPS_PC16 and R_MIPS_32.
Change-Id: I89601a93e4c1b9b40e81edc3c84474377fb98eb6
inkloader/include/impl/ELFObject.hxx
|
| c8b78bcfa73890ddf5fdc9a2f9364403c8b71d87 |
10-Oct-2014 |
Tim Murray <timmurray@google.com> |
am b26ead53: Merge "Fix missing relocation entries and libclcore entry points." into lmp-dev
* commit 'b26ead533bb9d4e5e78f9a95fb543a577e7f95f2':
Fix missing relocation entries and libclcore entry points.
|
| b26ead533bb9d4e5e78f9a95fb543a577e7f95f2 |
10-Oct-2014 |
Tim Murray <timmurray@google.com> |
Merge "Fix missing relocation entries and libclcore entry points." into lmp-dev
|
| cbdb6480528fc374144df1ae9c604c9100d33268 |
09-Oct-2014 |
Tim Murray <timmurray@google.com> |
Fix missing relocation entries and libclcore entry points.
This fixes support for rsSetElementAt_ulong from pre-21 SDKs.
bug 17934731
Change-Id: I1598f612257b8faeec871df6b2f126395b0e7d82
sCpuRuntimeStubs.cpp
|
| 09d7d5993e75e22ade6b37bc98c42d30a81ec6af |
09-Oct-2014 |
Jason Sams <jsams@google.com> |
am c214fe59: Fix incorrect error check for mmap
* commit 'c214fe59fc48740ed003a3cde4e5a60517c5d5ce':
Fix incorrect error check for mmap
|
| 9732e859ff5d1911915eb83411c9b1ae991c7523 |
20-Sep-2014 |
Simon Hosie <simon.hosie@arm.com> |
Tail-store fixes to AArch64 YuvToRGB assembly.
bug 17923388
Change-Id: I692cf3b38b0cf57404024170874070d5f2b95480
sCpuIntrinsics_advsimd_YuvToRGB.S
|
| 3e8d240af4e4b77d1a4903548c3757cbc15420a5 |
09-Oct-2014 |
Jason Sams <jsams@google.com> |
am 7d38b3a9: Merge "Tail-store fixes to AArch64 YuvToRGB assembly."
* commit '7d38b3a9b0fcad32cbac6f6431d7d73f552dabac':
Tail-store fixes to AArch64 YuvToRGB assembly.
|
| 476e523301f35c30aa1871cf8f8adea41d8977bb |
09-Oct-2014 |
Stephen Hines <srhines@google.com> |
am cac74c05: Merge "Fix script instancing issue for RS compatibility library."
* commit 'cac74c05b41acd5b5d64bedee9e214a7803f2362':
Fix script instancing issue for RS compatibility library.
|
| 7d38b3a9b0fcad32cbac6f6431d7d73f552dabac |
09-Oct-2014 |
Jason Sams <jsams@google.com> |
Merge "Tail-store fixes to AArch64 YuvToRGB assembly."
|
| c15aabdc69839824b4bae6648aee1604c31dbdf8 |
09-Oct-2014 |
Jason Sams <jsams@google.com> |
Merge "Revert "RS: Add VP9 LoopFilter Intrinsic""
|
| c214fe59fc48740ed003a3cde4e5a60517c5d5ce |
09-Oct-2014 |
Jason Sams <jsams@google.com> |
Fix incorrect error check for mmap
bug 17909809
Change-Id: I8fd0d5c489ff8b110d657cd63d3cae164d4067b9
sCpuIntrinsicColorMatrix.cpp
|
| cac74c05b41acd5b5d64bedee9e214a7803f2362 |
09-Oct-2014 |
Stephen Hines <srhines@google.com> |
Merge "Fix script instancing issue for RS compatibility library."
|
| fe0922575f26af84ee33429626f36049410cb7b6 |
09-Oct-2014 |
Jason Sams <jsams@google.com> |
Revert "RS: Add VP9 LoopFilter Intrinsic"
This reverts commit 6fc3e12b8912458cb4adcfd32e2f53d76b0cc737.
Change-Id: I4eb50620548805344bd45669fda1af81128195f5
ndroid.mk
sCpuCore.cpp
sCpuIntrinsicLoopFilter.cpp
p9_loopfilter_16_neon.S
p9_loopfilter_neon.S
p9_mb_lpf_neon.S
|
| a1adf77cbfe83f989dbfb103bf24f258d61b6b80 |
08-Oct-2014 |
Chih-Hung Hsieh <chh@google.com> |
am 99573dde: Merge "Add -no-integrated-as at local level."
* commit '99573dde2a5b68dc69e00da5220a0d7898bc5e9c':
Add -no-integrated-as at local level.
|
| 12f143e33cfc5ccead1ab3209944f648b1a8d349 |
03-Oct-2014 |
Chih-Hung Hsieh <chh@google.com> |
Add -no-integrated-as at local level.
Later we will enable integrated-as as default at the global level.
BUG: 17820427
Change-Id: I6d7c28c7188deab5d507cb6836adac460e0affe7
ndroid.mk
|
| b6bab6173430178d131f95f62e3b29627b9eb313 |
02-Oct-2014 |
Stephen Hines <srhines@google.com> |
Define calculatedValue for all paths.
Change-Id: I0e9318a4537e41a406abdbef9ec296bba72fdd70
(cherry picked from commit d176f116564d942a26039ff2f91e343c55ccc888)
inkloader/include/impl/ELFObject.hxx
|
| d176f116564d942a26039ff2f91e343c55ccc888 |
02-Oct-2014 |
Stephen Hines <srhines@google.com> |
Define calculatedValue for all paths.
Change-Id: I0e9318a4537e41a406abdbef9ec296bba72fdd70
inkloader/include/impl/ELFObject.hxx
|
| fca65171c3dfbaeced40b8a38387a3c12582cc7f |
02-Oct-2014 |
Stephen Hines <srhines@google.com> |
am 13f65fea: Merge "Add missing <memory> include for std::unique_ptr." into lmp-dev
* commit '13f65fea3830b84a0476d26f750896175bc832c1':
Add missing <memory> include for std::unique_ptr.
|
| 13f65fea3830b84a0476d26f750896175bc832c1 |
02-Oct-2014 |
Stephen Hines <srhines@google.com> |
Merge "Add missing <memory> include for std::unique_ptr." into lmp-dev
|
| 7d774858f7b0459dc90158341a8f7eb4b395b61c |
01-Oct-2014 |
Stephen Hines <srhines@google.com> |
Fix script instancing issue for RS compatibility library.
Bug: 17734273
This switches us from using symlinked files to using explicit copies. The
latest Android linker actually checks the inode for the loaded shared object,
so a symlink is now insufficient to have it count as a distinct object to
load.
Change-Id: I729993696ef8d656761fe571703ab96d4bec940b
sCpuScript.cpp
|
| 28c034238b8891398da625b070279c34185b3494 |
20-Sep-2014 |
Simon Hosie <simon.hosie@arm.com> |
Subrectangle bug fixes to ARM Blur assembly.
Correctly sweep the ends of the initial convolution window where that window
overlaps one or both sides of the source image.
Change-Id: I4e3d2bfa7eb22ce29af2615bf324a69561f7e4d6
sCpuIntrinsics_advsimd_Blur.S
sCpuIntrinsics_neon_Blur.S
|
| 6267c335745f32fb0d898335930da6b0904be577 |
20-Sep-2014 |
Simon Hosie <simon.hosie@arm.com> |
Minor fixes to AArch64 Blur assembly.
Use the correct register for a low-weighted tap in uchar1 case, and use the
correct clipped radius calculation for the right-hand edge.
Change-Id: Ib6ecd8b115c8898bb641958ab0beab11a8fccc36
sCpuIntrinsics_advsimd_Blur.S
|
| db683cd11b6cfd3d9edc208908332f87176a22d9 |
30-Sep-2014 |
Jason Sams <jsams@google.com> |
resolved conflicts for merge of 79a9dacb to lmp-dev-plus-aosp
Change-Id: Id4a5aafcd67afa6962bd9c042ce81fb6170c8c12
|
| 79a9dacb328ac37a2ec29b06b2ee6740fe8fadf0 |
30-Sep-2014 |
Jason Sams <jsams@google.com> |
Revert "RS: Add VP9 LoopFilter Intrinsic"
This reverts commit 6fc3e12b8912458cb4adcfd32e2f53d76b0cc737.
Conflicts:
cpu_ref/Android.mk
cpu_ref/rsCpuCore.cpp
cpu_ref/rsCpuIntrinsicLoopFilter.cpp
rsDefines.h
bug 17720646
Change-Id: Ie297cfbfa04e61944ac1340797ac81c23b2f2ea8
ndroid.mk
sCpuCore.cpp
sCpuIntrinsicLoopFilter.cpp
p9_loopfilter_16_neon.S
p9_loopfilter_neon.S
p9_mb_lpf_neon.S
|
| 340f54c0eb3cd71bc12b993f98cd94e1a2a7e6dc |
30-Sep-2014 |
Stephen Hines <srhines@google.com> |
Add missing <memory> include for std::unique_ptr.
Bug: 14416410
Change-Id: I96055b34bf497bceead776ec956a0b5b0dd817c3
inkloader/include/ELFHeader.h
inkloader/include/ELFObject.h
|
| b91add7e84a36e677d453919c209e551623c13b1 |
19-Jun-2014 |
Simon Hosie <simon.hosie@arm.com> |
Expand AArch64 YuvToRGB to use more registers.
Change-Id: Ib8ebea344ff863bb4c6e13f11efae3dd50f7c0fa
sCpuIntrinsics_advsimd_YuvToRGB.S
|
| 48fab7f9adae67c18b3d221a48ebe43e2fa77289 |
20-Sep-2014 |
Simon Hosie <simon.hosie@arm.com> |
Tail-store fixes to AArch64 YuvToRGB assembly.
Change-Id: I692cf3b38b0cf57404024170874070d5f2b95480
sCpuIntrinsics_advsimd_YuvToRGB.S
|
| 9b2711a80e755cefcf3f006d2f7f5f2665658578 |
20-Sep-2014 |
Stephen Hines <srhines@google.com> |
am ee7d3f48: Merge "Fix wrong machine identifier for AArch64 in linkloader."
* commit 'ee7d3f48fab7fd938885b18497cbf79247ddce71':
Fix wrong machine identifier for AArch64 in linkloader.
|
| 4b6eea2ec76fd3aa8096358092d828c6b4f0a45e |
20-Sep-2014 |
Stephen Hines <srhines@google.com> |
am 88f462c6: Merge "Implement getMaxNumStubs() for x86 and x86_64."
* commit '88f462c66c2f91e35ec24f14175162af485f9b01':
Implement getMaxNumStubs() for x86 and x86_64.
|
| 9adc78bbef17208ba11d5bc4b15e9677c985d87f |
20-Sep-2014 |
Stephen Hines <srhines@google.com> |
Fix wrong machine identifier for AArch64 in linkloader.
This doesn't change functionality at all, since the only "machine" variant
with special handling is EM_MIPS. Thus, reconfiguring as EM_AARCH64 instead
of EM_ARM is safe (and correct).
Change-Id: I6249682155957d20253ae48d2564e5b31cf21351
inkloader/include/impl/ELFObject.hxx
|
| 88f462c66c2f91e35ec24f14175162af485f9b01 |
20-Sep-2014 |
Stephen Hines <srhines@google.com> |
Merge "Implement getMaxNumStubs() for x86 and x86_64."
|
| 5e62c3a34aedaf6a91022da352df20647623935b |
17-Sep-2014 |
Xiaofei Wan <xiaofei.wan@intel.com> |
Implement getMaxNumStubs() for x86 and x86_64.
Change-Id: I5e7a446d17d3bdf4e93fd8fdaf8fd3ef9c3de571
Signed-off-by: Xiaofei Wan <xiaofei.wan@intel.com>
inkloader/include/impl/ELFSectionRelTable.hxx
|
| a0574689b5c8d7cafcc8ceedb14bb5776a3286d6 |
16-Sep-2014 |
Yong Chen <yong.a.chen@intel.com> |
am 76887149: Fix two bugs in cpu_ref
* commit '7688714916905f29362071ce2eb9e296ca469838':
Fix two bugs in cpu_ref
|
| 7688714916905f29362071ce2eb9e296ca469838 |
15-Sep-2014 |
Yong Chen <yong.a.chen@intel.com> |
Fix two bugs in cpu_ref
- Make correct parameters for x86 IntrinsicConvolve5x5
- Use correct intrinsic ID for loopFilter
Change-Id: I5180b9e2a80a6e023299963b19bf8b4dcd4ebd4b
Signed-off-by: Yong Chen <yong.a.chen@intel.com>
sCpuIntrinsicConvolve5x5.cpp
sCpuIntrinsicLoopFilter.cpp
|
| 45d29c41b1b9805991dcd8557f6d1b70977f5428 |
15-Sep-2014 |
Yong Chen <yong.a.chen@intel.com> |
Fix two bugs in cpu_ref
- Make correct parameters for x86 IntrinsicConvolve5x5
- Use correct intrinsic ID for loopFilter
Change-Id: I5180b9e2a80a6e023299963b19bf8b4dcd4ebd4b
Signed-off-by: Yong Chen <yong.a.chen@intel.com>
sCpuIntrinsicConvolve5x5.cpp
sCpuIntrinsicLoopFilter.cpp
|
| 4df913a7efe2f62b2bb9550d0c98bdc9ce360eea |
11-Sep-2014 |
Stephen Hines <srhines@google.com> |
am 97df6f6b: Merge "[MIPS] Add GOT right below the text section"
* commit '97df6f6b0f354b8b08f24c2561546d53ce983775':
[MIPS] Add GOT right below the text section
|
| 97df6f6b0f354b8b08f24c2561546d53ce983775 |
11-Sep-2014 |
Stephen Hines <srhines@google.com> |
Merge "[MIPS] Add GOT right below the text section"
|
| 33164686a7ac88d4eda38201be4127937e9c12b0 |
01-Sep-2014 |
Yong Chen <yong.a.chen@intel.com> |
Fix bugs for x86 platform
- Fix incorrect input/output pointer for blur intrinsic
- Avoid the negative value for length function in bc lib
Signed-off-by: Yong Chen <yong.a.chen@intel.com>
(cherry picked from commit a0cdfe05905a5a4d2e494665809c8af9d040c116)
Change-Id: I3c6e082dffdfef122999c93d1c235e99af8c62ba
sCpuIntrinsicBlur.cpp
|
| eac3e9b89366d1381e2c567b358a6e4a7fa390e8 |
04-Sep-2014 |
Petar Jovanovic <petar.jovanovic@imgtec.com> |
[MIPS] Add GOT right below the text section
Make sure that GOT is close to the text section, so the function can reach
it easily. This is important for MIPS64 for which there is an assumption
that GOT can be reached if 32-bit value is added to $t9 (the start of the
function for PIC). The MIPS64 code looks like this:
// lui $v0, %hi(%neg(%gp_rel(fname)))
// daddu $v1, $v0, $t9
// daddiu $globalbasereg, $v1, %lo(%neg(%gp_rel(fname)))
This patch prevents possible issue in which a function could not reach GOT.
Change-Id: I024f1f9483ffcd312bb6fcbc61145a171e7096e9
inkloader/include/GOT.h
inkloader/include/impl/ELFSectionProgBits.hxx
inkloader/lib/GOT.cpp
|
| bf2963268010b1e17cba00674a73287bf12a1be9 |
04-Sep-2014 |
Stephen Hines <srhines@google.com> |
am 2993761f: Merge "Fix bugs for x86 platform"
* commit '2993761fda2146755533886495691936a78871ac':
Fix bugs for x86 platform
|
| a0cdfe05905a5a4d2e494665809c8af9d040c116 |
01-Sep-2014 |
Yong Chen <yong.a.chen@intel.com> |
Fix bugs for x86 platform
- Fix incorrect input/output pointer for blur intrinsic
- Avoid the negative value for length function in bc lib
Change-Id: If173662b63d7aba906a75bb50856d8d50c04639f
Signed-off-by: Yong Chen <yong.a.chen@intel.com>
sCpuIntrinsicBlur.cpp
|
| bff7da6ff43ba594dddeaf15bf4ad4d029eb3897 |
23-Aug-2014 |
Dan Albert <danalbert@google.com> |
am da2f0714: Merge "Fix x86 build for SSSE3."
* commit 'da2f071483b148bba216d40070d87f92785b527f':
Fix x86 build for SSSE3.
|
| ebf0eb95cba9579af7cb67205b94b286f221c4ed |
22-Aug-2014 |
Dan Albert <danalbert@google.com> |
Fix x86 build for SSSE3.
These functions were previously defined in a C file. The file was
renamed to C++, but the extern "C"s were not removed.
Change-Id: I62ec9ba7c73d24c221bbe3f7580cfd7f09bf701c
sCpuIntrinsicBlend.cpp
sCpuIntrinsicBlur.cpp
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsics_x86.cpp
|
| 61db21cef81cbc8439b34a8f1d32952339ffc190 |
22-Aug-2014 |
Stephen Hines <srhines@google.com> |
resolved conflicts for merge of 81cfa2c5 to lmp-dev-plus-aosp
Change-Id: Ia3a7f084693825d4e46a4864788332a7d039d38f
|
| 4c513c12f0f620c336efce7b92b8f26aae39ffdd |
22-Aug-2014 |
Jason Sams <jsams@google.com> |
resolved conflicts for merge of 5f95a8ab to lmp-dev-plus-aosp
Change-Id: I3044361ad6d5af09906e4a32e47efc3f29486514
|
| d25fef7232a939faaffcdb83a1be28285313c38e |
22-Aug-2014 |
Jason Sams <jsams@google.com> |
Fix blur and 3dlut with clipped output
bug 17157250
Change-Id: I388a255380fbdd9f6b5d1c7cb9f14df6f035ae48
sCpuIntrinsic3DLUT.cpp
sCpuIntrinsicBlur.cpp
|
| 4592e82e77fdd8b6bc1cbc99c6809907f5dcddbe |
21-Aug-2014 |
Jason Sams <jsams@google.com> |
Fix build break from bad merge.
Underlying structure changed in AOSP since LMP was branched.
Change-Id: I7793f9b03e5febfd9bc8965f7976ecaa9c6e2275
sCpuIntrinsicColorMatrix.cpp
|
| ef48ce24c2fac16326ea9719cae7e930633c52dc |
21-Aug-2014 |
Jason Sams <jsams@google.com> |
resolved conflicts for merge of adbc54f3 to lmp-dev-plus-aosp
Change-Id: I04f438a6fc4cd374a821c32102cc758d9a30d731
|
| 7e3243389f49a0bf673c35f9a82e4bbfe211877b |
21-Aug-2014 |
Stephen Hines <srhines@google.com> |
am 4bb773ef: Merge "Replace android::String8 with std::string"
* commit '4bb773ef75eeff5960852b733bcca5834020ee9e':
Replace android::String8 with std::string
|
| 44bef6fba6244292b751387f3d6c31cca96c28ad |
12-Aug-2014 |
Chris Wailes <chriswailes@google.com> |
Replace NULL macros with nullptr literals.
Change-Id: I918c40879aa547438f77e7d1a95fa2aa33bec398
inkloader/android/librsloader.cpp
inkloader/include/ELFHeader.h
inkloader/include/ELFObject.h
inkloader/include/ELFSectionBits.h
inkloader/include/ELFSectionProgBits.h
inkloader/lib/MemChunk.cpp
inkloader/lib/StubLayout.cpp
inkloader/main.cpp
inkloader/utils/serialize.h
sCpuCore.cpp
sCpuCore.h
sCpuIntrinsicBlur.cpp
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsicHistogram.cpp
sCpuIntrinsicLoopFilter.cpp
sCpuIntrinsicYuvToRGB.cpp
sCpuRuntimeMath.cpp
sCpuRuntimeStubs.cpp
sCpuScript.cpp
sCpuScript.h
sCpuScriptGroup.cpp
sd_cpu.h
|
| de52a834dbcb2a3196948e7b9f67d395493ea9a4 |
21-Aug-2014 |
Jason Sams <jsams@google.com> |
Fix two intrinsic clipping bugs.
bug 17157250
Change-Id: I65c945da6bd90733333a5472c1a85b5d3e3e4d6b
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsicConvolve5x5.cpp
|
| 569b09b34243f71dd6d81ce637a92a0a707adbe4 |
21-Aug-2014 |
Stephen Hines <srhines@google.com> |
am 3fa79f84: Merge changes I4c6abd96,I9bcf36d0
* commit '3fa79f84fcce0b5f59498980a2bb8e355cf53cd4':
Replaced android::Vector with std::vector.
Adjust to handle the change to std::vector in libbcc.
|
| 237599760fbd1e98622c957ffd6c17279895be59 |
20-Aug-2014 |
Stephen Hines <srhines@google.com> |
am db0a8d1d: Merge "Rename rsCpuIntrinsics_x86 to a C++ source file."
* commit 'db0a8d1de452ad4c69120aa26ab1d604cad2c67d':
Rename rsCpuIntrinsics_x86 to a C++ source file.
|
| 4905166b56d74c97a83eee152bbbf1fa690ab132 |
19-Aug-2014 |
Stephen Hines <srhines@google.com> |
am cdb78457: Merge "Remove the instep parameter."
* commit 'cdb784579cee7094ad35ecc0cd3b45245eb1c382':
Remove the instep parameter.
|
| 6847e73314e13aa02231268cca245a81eb0539ca |
12-Aug-2014 |
Chris Wailes <chriswailes@google.com> |
Replace android::String8 with std::string
Change-Id: I5b2b6d3e38afd0b040f0a584613745206bf01ba0
sCpuScript.cpp
|
| 3fa79f84fcce0b5f59498980a2bb8e355cf53cd4 |
20-Aug-2014 |
Stephen Hines <srhines@google.com> |
Merge changes I4c6abd96,I9bcf36d0
* changes:
Replaced android::Vector with std::vector.
Adjust to handle the change to std::vector in libbcc.
|
| 7de1920270dabef99538bc82a344632e04487fa9 |
20-Aug-2014 |
Stephen Hines <srhines@google.com> |
Rename rsCpuIntrinsics_x86 to a C++ source file.
Change-Id: Ib2897d71180259796be780efda63abe14a209e6f
ndroid.mk
sCpuIntrinsics_x86.c
sCpuIntrinsics_x86.cpp
|
| 93d6bc872b7d9fba63abfa7513d56b38d9c3d371 |
29-Jul-2014 |
Chris Wailes <chriswailes@google.com> |
Replaced android::Vector with std::vector.
Change-Id: I4c6abd964dc4b1412ec2e592fc8e835fecfe53f6
sCpuScript.cpp
sCpuScriptGroup.cpp
|
| 70d4971878444fe7cf4263998b0c3f46e2b6c71f |
08-Aug-2014 |
Chris Wailes <chriswailes@google.com> |
Adjust to handle the change to std::vector in libbcc.
Change-Id: I9bcf36d088d423246312d97e2a87a05e90f6617d
sCpuScript.cpp
|
| 9ed79105cc6a8dbfaf959875249f36022cc2c798 |
26-Jul-2014 |
Chris Wailes <chriswailes@google.com> |
Remove the instep parameter.
This patch removes the instep parameter from calls to expanded kernels and
from the CPU reference implementation intrinsics.
Change-Id: I059db548a57702c576963f6b17a002b2ee393cdb
sCpuCore.cpp
sCpuIntrinsic3DLUT.cpp
sCpuIntrinsicBlend.cpp
sCpuIntrinsicBlur.cpp
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsicConvolve3x3.cpp
sCpuIntrinsicConvolve5x5.cpp
sCpuIntrinsicHistogram.cpp
sCpuIntrinsicLUT.cpp
sCpuIntrinsicLoopFilter.cpp
sCpuIntrinsicResize.cpp
sCpuIntrinsicYuvToRGB.cpp
sCpuScript.h
sCpuScriptGroup.cpp
sCpuScriptGroup.h
|
| 26f26c7f30604712794cdce967f8724931078b23 |
18-Aug-2014 |
Stephen Hines <srhines@google.com> |
am 8d63899d: Merge "Collapse code paths for single- and multi-input kernels."
* commit '8d63899d90e3048832dc3c6c5ea5ef434658c8d5':
Collapse code paths for single- and multi-input kernels.
|
| f37121300217d3b39ab66dd9c8881bcbcad932df |
17-Jul-2014 |
Chris Wailes <chriswailes@google.com> |
Collapse code paths for single- and multi-input kernels.
This patch simplifies the RenderScript driver and CPU reference implementation
by removing the distinction between sing- and multi-input kernels in many
places. The distinction is maintained in some places due to the need to
maintain backwards compatibility. This permits the deletion of some functions
and struct members that are no longer needed. Several related functions were
also cleaned up.
Change-Id: Id70a223ea5e3aa2b0b935b2b7f9af933339ae8a4
ndroid.mk
sCpuCore.cpp
sCpuCore.h
sCpuIntrinsic.cpp
sCpuIntrinsic.h
sCpuIntrinsic3DLUT.cpp
sCpuIntrinsicBlend.cpp
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsicHistogram.cpp
sCpuIntrinsicLUT.cpp
sCpuIntrinsicResize.cpp
sCpuScript.cpp
sCpuScript.h
sCpuScriptGroup.cpp
sd_cpu.h
|
| 385e32592394cf796d33102b8b7034ae81fdc13e |
13-Aug-2014 |
Stephen Hines <srhines@google.com> |
am cd8df40f: Merge "Revert "Collapse code paths for single- and multi-input kernels.""
* commit 'cd8df40f07d88f896bea05ed06a20d1a4d5e9728':
Revert "Collapse code paths for single- and multi-input kernels."
|
| 38f10a1eee245153b843dd78712ecaff6e89c412 |
13-Aug-2014 |
Stephen Hines <srhines@google.com> |
am f2dd1651: Merge "Collapse code paths for single- and multi-input kernels."
* commit 'f2dd1651ea50bf176d5e6580cc0589f7a3e89995':
Collapse code paths for single- and multi-input kernels.
|
| 818cfa034e257c7bb48356257f5cb67334e19aa6 |
17-Jul-2014 |
Chris Wailes <chriswailes@google.com> |
Collapse code paths for single- and multi-input kernels.
This patch simplifies the RenderScript driver and CPU reference implementation
by removing the distinction between sing- and multi-input kernels in many
places. The distinction is maintained in some places due to the need to
maintain backwards compatibility. This permits the deletion of some functions
and struct members that are no longer needed. Several related functions were
also cleaned up.
Change-Id: I77e4b155cc7ca1581b05bf901c70ae53a9ff0b12
ndroid.mk
sCpuCore.cpp
sCpuCore.h
sCpuIntrinsic.cpp
sCpuIntrinsic.h
sCpuIntrinsic3DLUT.cpp
sCpuIntrinsicBlend.cpp
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsicHistogram.cpp
sCpuIntrinsicLUT.cpp
sCpuIntrinsicResize.cpp
sCpuScript.cpp
sCpuScript.h
sCpuScriptGroup.cpp
sd_cpu.h
|
| bc140a9945120ec7d48824864acf0595dd31f7e2 |
13-Aug-2014 |
Stephen Hines <srhines@google.com> |
am fb0a2749: Merge "Add support for MIPS64 in linkloader"
* commit 'fb0a274983ae8bfb07aff8c292305389789d6e92':
Add support for MIPS64 in linkloader
|
| 4b2bea3dc20865f3a198797702e19912a6a2171c |
13-Aug-2014 |
Stephen Hines <srhines@google.com> |
Revert "Collapse code paths for single- and multi-input kernels."
This reverts commit 818cfa034e257c7bb48356257f5cb67334e19aa6.
Change-Id: I59f39f52e6c8f60bb01cbcb8ccf2215eaf46a57f
ndroid.mk
sCpuCore.cpp
sCpuCore.h
sCpuIntrinsic.cpp
sCpuIntrinsic.h
sCpuIntrinsic3DLUT.cpp
sCpuIntrinsicBlend.cpp
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsicHistogram.cpp
sCpuIntrinsicLUT.cpp
sCpuIntrinsicResize.cpp
sCpuScript.cpp
sCpuScript.h
sCpuScriptGroup.cpp
sd_cpu.h
|
| fb0a274983ae8bfb07aff8c292305389789d6e92 |
13-Aug-2014 |
Stephen Hines <srhines@google.com> |
Merge "Add support for MIPS64 in linkloader"
|
| 6a45ddb32f391060aa05da6ff09c4814d450586e |
06-Aug-2014 |
Tim Murray <timmurray@google.com> |
Bugfixes for AArch64.
- Disable use of runtime pointer until we can set it correctly in invokes.
- Disable YUV and ColorMatrix ASIMD intrinsics
- Modify runtime stubs to use large objects correctly
- Fix calling convention in allocation.ll
- Add appropriate rsr functions for compat lib and large objects
bug 16846318
Change-Id: I7f6a4ff66fd1b3ad40b08cfc9ecdda7d53e95c7d
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsicYuvToRGB.cpp
|
| 05ef73f2d934f1083cc3b8aeb33fe21de9d6e88f |
05-Aug-2014 |
Jason Sams <jsams@google.com> |
Fix isObject when in 64bit mode.
This also cleans up the passing of args to SetObject and ClearObject.
Fix related issues in 64bit teardown.
Change-Id: I90d1b93c45ff5bc84957d5441ab48a4e2f6feb20
sCpuScript.cpp
|
| 89226c21a902990642d4c1a9b8a71b89825e3e01 |
12-Aug-2014 |
Tim Murray <timmurray@google.com> |
am 1aa9dfc0: Bugfixes for AArch64.
* commit '1aa9dfc002f6b763d34d75d9f47abb4aa70584a2':
Bugfixes for AArch64.
|
| 1aa9dfc002f6b763d34d75d9f47abb4aa70584a2 |
06-Aug-2014 |
Tim Murray <timmurray@google.com> |
Bugfixes for AArch64.
- Disable use of runtime pointer until we can set it correctly in invokes.
- Disable YUV and ColorMatrix ASIMD intrinsics
- Modify runtime stubs to use large objects correctly
- Fix calling convention in allocation.ll
- Add appropriate rsr functions for compat lib and large objects
bug 16846318
Change-Id: I7f6a4ff66fd1b3ad40b08cfc9ecdda7d53e95c7d
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsicYuvToRGB.cpp
|
| 59886ba8f9bd4375bca1b82fe245696dfb383716 |
28-Jul-2014 |
Tim Murray <timmurray@google.com> |
Fix relocation issue for AArch64.
Change-Id: Ie10563b6a1ea894c45076480b123dcc2a854b5a5
inkloader/include/impl/ELFObject.hxx
|
| 5b0df2a1e2b329cbc3da9d4eed3bc469368c4009 |
06-Aug-2014 |
Jason Sams <jsams@google.com> |
am 262d77c2: Merge "Fix isObject when in 64bit mode." into lmp-dev
* commit '262d77c269ab55966d5803e4f795b3b60a548a37':
Fix isObject when in 64bit mode.
|
| f29edf80510d483b8663e9d8b859959e37361aee |
05-Aug-2014 |
Jason Sams <jsams@google.com> |
Fix isObject when in 64bit mode.
This also cleans up the passing of args to SetObject and ClearObject.
Fix related issues in 64bit teardown.
Change-Id: I90d1b93c45ff5bc84957d5441ab48a4e2f6feb20
sCpuScript.cpp
|
| 39f577243c8ce69a24a7d7ab50ba33b1800ad74e |
30-Jul-2014 |
Tim Murray <timmurray@google.com> |
am ad0544fd: Merge "Fix relocation issue for AArch64."
* commit 'ad0544fdf918e64cec05d1c98588880f10b09220':
Fix relocation issue for AArch64.
|
| ad0544fdf918e64cec05d1c98588880f10b09220 |
30-Jul-2014 |
Tim Murray <timmurray@google.com> |
Merge "Fix relocation issue for AArch64."
|
| 4e658279f46c556b5d18955aaba695991210400b |
29-Jul-2014 |
Yang Ni <yangni@google.com> |
am 99192979: Validate input/output for a ScriptGroup.
* commit '99192979b3ef356998a438f5bd2e45fede6b457f':
Validate input/output for a ScriptGroup.
|
| 5f6f16fe612e1c30732becc66205500994889ac9 |
25-Jul-2014 |
Yang Ni <yangni@google.com> |
Validate input/output for a ScriptGroup.
If a script group requires input or output allocations, make sure
they are set before starting the execution of the script group.
If any allocation is missing, skip the execution and set a bad
value error with a message.
Change-Id: I7ee2da96aca3e6c9820225e6b2c39cb9378cb42f
sCpuScriptGroup.cpp
|
| 99192979b3ef356998a438f5bd2e45fede6b457f |
25-Jul-2014 |
Yang Ni <yangni@google.com> |
Validate input/output for a ScriptGroup.
If a script group requires input or output allocations, make sure
they are set before starting the execution of the script group.
If any allocation is missing, skip the execution and set a bad
value error with a message.
Change-Id: I7ee2da96aca3e6c9820225e6b2c39cb9378cb42f
sCpuScriptGroup.cpp
|
| 9aaebe5695c0b14c1ff33a696a8074e8736acef9 |
28-Jul-2014 |
Tim Murray <timmurray@google.com> |
Fix relocation issue for AArch64.
Change-Id: Ie10563b6a1ea894c45076480b123dcc2a854b5a5
inkloader/include/impl/ELFObject.hxx
|
| cfd173917c39e172ed84be26dce229bc8b5cccbb |
25-Jul-2014 |
Stephen Hines <srhines@google.com> |
am 28c8f3a8: Merge "Split the RsForEachStubParamStruct in two."
* commit '28c8f3a8f4e92ef9956c5a01f09347eff1e59417':
Split the RsForEachStubParamStruct in two.
|
| d282991c78d1fd79a222c5f114e623da255ba8a0 |
17-Jul-2014 |
Stephen Hines <srhines@google.com> |
Update RS for LLVM rebase to r212749.
This is mostly s/LLVM::OwningPtr/std::unique_ptr/.
Change-Id: I393f801e0afb867aa262c004844c2542d6d12da4
inkloader/android/librsloader.cpp
inkloader/include/ELFHeader.h
inkloader/include/ELFObject.h
inkloader/include/ELFReloc.h
inkloader/include/ELFSection.h
inkloader/include/ELFSectionBits.h
inkloader/include/ELFSectionHeader.h
inkloader/include/ELFSectionHeaderTable.h
inkloader/include/ELFSymbol.h
inkloader/include/impl/ELFObject.hxx
inkloader/include/impl/ELFReloc.hxx
inkloader/include/impl/ELFSectionHeader.hxx
inkloader/include/impl/ELFSectionHeaderTable.hxx
inkloader/include/impl/ELFSectionNoBits.hxx
inkloader/include/impl/ELFSectionProgBits.hxx
inkloader/include/impl/ELFSectionRelTable.hxx
inkloader/include/impl/ELFSectionStrTab.hxx
inkloader/include/impl/ELFSectionSymTab.hxx
inkloader/include/impl/ELFSymbol.hxx
inkloader/main.cpp
|
| 8386869e7f0b4bd6d8f75cf9839202435a63d393 |
25-Jul-2014 |
Dave Allison <dallison@google.com> |
am d315b98b: Merge "Fix linker veneer for aarch64 in renderscript" into lmp-dev
* commit 'd315b98bdf02789476ed3a0d1f35e210b917ff0a':
Fix linker veneer for aarch64 in renderscript
|
| 4e42468106fb8d5c63e2a4d09f1ccafdad61f893 |
24-Jul-2014 |
Dave Allison <dallison@google.com> |
Fix linker veneer for aarch64 in renderscript
Fix to ldr instruction encoding for loading target address.
Also adds a debug print for future debug (default is off)
Bug: 16031597
Change-Id: Ibe47226d9f95e7982dcbbf559389e927c48aed3d
inkloader/include/impl/ELFObject.hxx
inkloader/lib/StubLayout.cpp
|
| 28c8f3a8f4e92ef9956c5a01f09347eff1e59417 |
25-Jul-2014 |
Stephen Hines <srhines@google.com> |
Merge "Split the RsForEachStubParamStruct in two."
|
| d315b98bdf02789476ed3a0d1f35e210b917ff0a |
25-Jul-2014 |
Dave Allison <dallison@google.com> |
Merge "Fix linker veneer for aarch64 in renderscript" into lmp-dev
|
| 8817e3ebd67635810000de8898afee615a674760 |
24-Jul-2014 |
Dave Allison <dallison@google.com> |
Fix linker veneer for aarch64 in renderscript
Fix to ldr instruction encoding for loading target address.
Also adds a debug print for future debug (default is off)
Bug: 16031597
Change-Id: Ibe47226d9f95e7982dcbbf559389e927c48aed3d
inkloader/include/impl/ELFObject.hxx
inkloader/lib/StubLayout.cpp
|
| 34b2068c973ae65962c61fa11cf785a69bb5e233 |
24-Jul-2014 |
Stephen Hines <srhines@google.com> |
am fb8fe820: Merge changes I2a12d44b,I393f801e
* commit 'fb8fe8207eeb8b93429c7816606d27cf8066ab9f':
Pass -Bsymbolic to the linker from clang when compiling against stlport_static.
Update RS for LLVM rebase to r212749.
|
| 5600ebccedd4ae86bd6e41459e33bd2ae3992261 |
06-May-2014 |
Simon Hosie <simon.hosie@arm.com> |
Arbitrary sub-rectangle handling for ColorMatrix.
AArch64 code for odd-length cases, and correct pointer offsets.
Change-Id: I28049a768a1e3e65611898904fa42bd295208871
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsics_advsimd_ColorMatrix.S
|
| abe41a94aeda9bc1c3d544e992534f5367a4874a |
23-Jul-2014 |
Jason Sams <jsams@google.com> |
am 4a7c5fc6: Merge "Arbitrary sub-rectangle handling for ColorMatrix."
* commit '4a7c5fc6025329cf8d31c44b5ea6ff1f955e71f1':
Arbitrary sub-rectangle handling for ColorMatrix.
|
| fb8fe8207eeb8b93429c7816606d27cf8066ab9f |
24-Jul-2014 |
Stephen Hines <srhines@google.com> |
Merge changes I2a12d44b,I393f801e
* changes:
Pass -Bsymbolic to the linker from clang when compiling against stlport_static.
Update RS for LLVM rebase to r212749.
|
| 4a7c5fc6025329cf8d31c44b5ea6ff1f955e71f1 |
23-Jul-2014 |
Jason Sams <jsams@google.com> |
Merge "Arbitrary sub-rectangle handling for ColorMatrix."
|
| c29686a4d4bd0ea8c0341d3ddbec796fa905f3a7 |
17-Jul-2014 |
Stephen Hines <srhines@google.com> |
Update RS for LLVM rebase to r212749.
This is mostly s/LLVM::OwningPtr/std::unique_ptr/.
Change-Id: I393f801e0afb867aa262c004844c2542d6d12da4
inkloader/android/librsloader.cpp
inkloader/include/ELFHeader.h
inkloader/include/ELFObject.h
inkloader/include/ELFReloc.h
inkloader/include/ELFSection.h
inkloader/include/ELFSectionBits.h
inkloader/include/ELFSectionHeader.h
inkloader/include/ELFSectionHeaderTable.h
inkloader/include/ELFSymbol.h
inkloader/include/impl/ELFObject.hxx
inkloader/include/impl/ELFReloc.hxx
inkloader/include/impl/ELFSectionHeader.hxx
inkloader/include/impl/ELFSectionHeaderTable.hxx
inkloader/include/impl/ELFSectionNoBits.hxx
inkloader/include/impl/ELFSectionProgBits.hxx
inkloader/include/impl/ELFSectionRelTable.hxx
inkloader/include/impl/ELFSectionStrTab.hxx
inkloader/include/impl/ELFSectionSymTab.hxx
inkloader/include/impl/ELFSymbol.hxx
inkloader/main.cpp
|
| b9226587245293eefb8959d10fee05ebc2658efd |
17-Jul-2014 |
Jean-Luc Brouillet <jeanluc@google.com> |
am b43d5d8c: Merge "Deprecate rs_fp_imprecise"
* commit 'b43d5d8cec78dfdfa2040d0dbd4f436d0d6a1bba':
Deprecate rs_fp_imprecise
|
| 7ebb02bf54cc3e6b6424c8da642ed75b460cc40d |
10-Jul-2014 |
Jean-Luc Brouillet <jeanluc@google.com> |
Deprecate rs_fp_imprecise
Change-Id: I063e356e939393e6fa4f39d056cb526aeaf8ee70
sCpuScript.cpp
|
| 80ef693674f69c0343c41564e30f80e7fb513b60 |
08-Jul-2014 |
Chris Wailes <chriswailes@google.com> |
Split the RsForEachStubParamStruct in two.
This patch splits the RsForEachStubParamStruct into two smaller structs, one
used specifically by the driver and the other by the expanded kernels. Doing
so makes it clearer what data is used where. In addition, fewer data are
copied between memory locations during kernel invocation.
Several fields that were not being used were removed from the structs.
Change-Id: I7788ef754add44463b17a6b571c7cde6e73b9712
sCpuCore.cpp
sCpuCore.h
sCpuIntrinsic3DLUT.cpp
sCpuIntrinsicBlend.cpp
sCpuIntrinsicBlur.cpp
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsicConvolve3x3.cpp
sCpuIntrinsicConvolve5x5.cpp
sCpuIntrinsicHistogram.cpp
sCpuIntrinsicLUT.cpp
sCpuIntrinsicLoopFilter.cpp
sCpuIntrinsicResize.cpp
sCpuIntrinsicYuvToRGB.cpp
sCpuScript.h
sCpuScriptGroup.cpp
sCpuScriptGroup.h
|
| a3a08243577a5d03239434e3f072af4c59f466e2 |
10-Jul-2014 |
Petar Jovanovic <petar.jovanovic@imgtec.com> |
Add support for MIPS64 in linkloader
Add relocateMIPS64() and handle the necessary relocations for the 64-bit
architecture.
Change-Id: Ib6372185dad349fce038a1801cf533e7dae8b27f
inkloader/include/ELFObject.h
inkloader/include/ELFReloc.h
inkloader/include/GOT.h
inkloader/include/impl/ELFObject.hxx
inkloader/lib/GOT.cpp
|
| 058a7a4f5bfda35e75cbd87af2be58e874f254c4 |
12-Jul-2014 |
Tim Murray <timmurray@google.com> |
am 8c8eb23d: Merge "Add 64bit large object support"
* commit '8c8eb23d31768bca01e2fd69c3931aeda31f3329':
Add 64bit large object support
|
| 35180b6231025f81e42f6ff5ec761f24a91d30ee |
12-Jul-2014 |
Stephen Hines <srhines@google.com> |
am 60506d94: Merge "Fix build break for size_t vs. uint32_t difference."
* commit '60506d94c399dea8b6325000be27b07fed67852a':
Fix build break for size_t vs. uint32_t difference.
|
| c562a29b2c221c22792cdbb7deeadca3fc0a6edd |
12-Jul-2014 |
Stephen Hines <srhines@google.com> |
am db69bf01: Merge "Adds support for multi-input kernels to Frameworks/RS."
* commit 'db69bf0117c44d8996a3d179cc3b7a98bf8726ea':
Adds support for multi-input kernels to Frameworks/RS.
|
| eaff410b9ba00927aeaad92dbbe72a45ff9e00aa |
12-Jul-2014 |
Stephen Hines <srhines@google.com> |
am 11007c74: Merge "Moved RsForEachStubParamStruct out of rs_hal.h and into rsCpuCore.h."
* commit '11007c7488436bf4213d62e2785e25111eb97bf7':
Moved RsForEachStubParamStruct out of rs_hal.h and into rsCpuCore.h.
|
| f4d3836c83097f57b62c235910fbf490c56ddf56 |
10-Jul-2014 |
Jean-Luc Brouillet <jeanluc@google.com> |
Deprecate rs_fp_imprecise
Change-Id: I063e356e939393e6fa4f39d056cb526aeaf8ee70
sCpuScript.cpp
|
| 6c6f7539377ec779a9ea36bfcc5859ad41f6e677 |
09-Jul-2014 |
Tim Murray <timmurray@google.com> |
am 8c8eb23d: Merge "Add 64bit large object support"
* commit '8c8eb23d31768bca01e2fd69c3931aeda31f3329':
Add 64bit large object support
|
| 191f131b123e20a217bcc57a3b2b626cc39f6bac |
09-Jul-2014 |
Stephen Hines <srhines@google.com> |
am 60506d94: Merge "Fix build break for size_t vs. uint32_t difference."
* commit '60506d94c399dea8b6325000be27b07fed67852a':
Fix build break for size_t vs. uint32_t difference.
|
| 09e1490af01ecd12cf19ee1b507cea860814470b |
08-Jul-2014 |
Stephen Hines <srhines@google.com> |
am db69bf01: Merge "Adds support for multi-input kernels to Frameworks/RS."
* commit 'db69bf0117c44d8996a3d179cc3b7a98bf8726ea':
Adds support for multi-input kernels to Frameworks/RS.
|
| 7f57593db8e1c9a59c3e028c377321f652f9fad0 |
08-Jul-2014 |
Stephen Hines <srhines@google.com> |
am 11007c74: Merge "Moved RsForEachStubParamStruct out of rs_hal.h and into rsCpuCore.h."
* commit '11007c7488436bf4213d62e2785e25111eb97bf7':
Moved RsForEachStubParamStruct out of rs_hal.h and into rsCpuCore.h.
|
| 4b3c34e6833e39bc89c2128002806b654b8e623d |
11-Jun-2014 |
Chris Wailes <chriswailes@google.com> |
Adds support for multi-input kernels to Frameworks/RS.
This patch modifies Frameworks/RS in the following ways:
* Adjusted the data-layout of the C/C++ version of RsForEachStubParamStruct to
accommodate a pointer to an array of input allocations and a pointer to an
array of stride sizes for each of these allocatoins.
* Adds a new code path for Java code to pass multiple allocations to a RS
kernel.
* Packs base pointers and step values for multi-input kernels into the new
RsForEachStubParamStruct members.
Change-Id: I46d2834c37075b2a2407fd8b010546818a4540d1
sCpuCore.cpp
sCpuCore.h
sCpuIntrinsic.cpp
sCpuIntrinsic.h
sCpuScript.cpp
sCpuScript.h
sd_cpu.h
|
| cb51798a0379409c0f9927c44bbcdd772ed7ec18 |
08-Jul-2014 |
Chris Wailes <chriswailes@google.com> |
Moved RsForEachStubParamStruct out of rs_hal.h and into rsCpuCore.h.
Change-Id: Iea673ccd9db5077abef373457f51c6b0d09cea92
sCpuCore.h
|
| c5a20170784a6a44ee1de1a754ca8c7175b78a6d |
09-Jul-2014 |
Stephen Hines <srhines@google.com> |
Fix build break for size_t vs. uint32_t difference.
Change-Id: I11b9592214c4fa57ef62f42fd086a5a3df33abbf
sCpuIntrinsic.cpp
sCpuIntrinsic.h
|
| a36c50a6ab87f4c9049318d4c6c8ec7b0a1e6e12 |
17-Jun-2014 |
Jason Sams <jsams@google.com> |
Add 64bit large object support
Add HAL hooks for filling in LO fields.
Change-Id: Ib9bdee714a062de519d2a9708c1016404a1b11a4
Conflicts:
driver/rsdBcc.cpp
driver/rsdCore.cpp
rsAllocation.h
rs_hal.h
sCpuScript.cpp
|
| c2c00e81713f6a0d6fde711898e94826167795ef |
27-Jun-2014 |
Jean-Luc Brouillet <jeanluc@google.com> |
am 32e093ad: Merge "Use build fingerprint and compile command for caching."
* commit '32e093ad58e9b858a6a3b747f262c5f0757a2e22':
Use build fingerprint and compile command for caching.
|
| 1c38baa7565a61f4c69169aa2753b644586afd00 |
26-Jun-2014 |
Stephen Hines <srhines@google.com> |
am a3f7905a: Merge "Add in a missing unlockMutex() on an error path."
* commit 'a3f7905addd8ef7302efbfa7c0c3ce89a88e7660':
Add in a missing unlockMutex() on an error path.
|
| f94e8db4232979b2fc93e8a77c42bfa57d3da56a |
26-Jun-2014 |
Stephen Hines <srhines@google.com> |
Add in a missing unlockMutex() on an error path.
Change-Id: I223e64381190184e37a92ce6478340cb7a323e29
sCpuScript.cpp
|
| 1b70788fd4571cd6e94db2219c930b5b59081969 |
26-Jun-2014 |
Jason Sams <jsams@google.com> |
am 51905042: Merge "Use remainder of AArch64 register file in Blur."
* commit '519050428c14962951e1fbde020fe2cf2fd64fbe':
Use remainder of AArch64 register file in Blur.
|
| afb3114606d96c247ad91cb2bd15dc7ed6e1579a |
25-Jun-2014 |
Stephen Hines <srhines@google.com> |
am 03836b04: Merge "Switch the dimensions array to use uint32_t instead of size_t."
* commit '03836b040d3bf53a93b8577060a4eb0c11639a5e':
Switch the dimensions array to use uint32_t instead of size_t.
|
| 519050428c14962951e1fbde020fe2cf2fd64fbe |
26-Jun-2014 |
Jason Sams <jsams@google.com> |
Merge "Use remainder of AArch64 register file in Blur."
|
| 32e093ad58e9b858a6a3b747f262c5f0757a2e22 |
27-Jun-2014 |
Jean-Luc Brouillet <jeanluc@google.com> |
Merge "Use build fingerprint and compile command for caching."
|
| 40e35cdbe217ec8bf9fc3c69873c7d62fc14158f |
26-Jun-2014 |
Jean-Luc Brouillet <jeanluc@google.com> |
Use build fingerprint and compile command for caching.
If either of those have changed, we need to invalidate the
compiled bit code.
Change-Id: Ic0c392ea5bfff5bf6dc8511740306895b1b12c82
sCpuScript.cpp
sCpuScript.h
|
| 4bea0d3b51fcdd9976af72c553a4a1d492016ca2 |
04-Jun-2014 |
Simon Hosie <simon.hosie@arm.com> |
Use remainder of AArch64 register file in Blur.
A lot of load/store can be avoided by using the rest of the register file,
here, so take advantage of that.
Change-Id: Ifaa2071d73ddb4f1f49f7de04f29001b5621ef7a
sCpuIntrinsics_advsimd_Blur.S
|
| ac8d146a41f18afad5314ac8af440d6aedbe20bf |
25-Jun-2014 |
Stephen Hines <srhines@google.com> |
Switch the dimensions array to use uint32_t instead of size_t.
size_t isn't safe, since we pack/unpack the array as a 32-bit int array, but
that is the wrong type for 64-bit. Switching to uint32_t is better, since we
only support 1 dimension today, and won't need many more than that even for
complex cases in the future.
Change-Id: Ie0dda264a9398b0e385e0f9ee0a91cda08325dbc
sCpuIntrinsic.cpp
sCpuIntrinsic.h
sCpuScript.cpp
sCpuScript.h
sd_cpu.h
|
| 18f3f6fc011d6d73a4dddd24d628aa4d83839c3a |
20-Jun-2014 |
Jean-Luc Brouillet <jeanluc@google.com> |
am 5557f339: Merge "Move the lib paths out of RSInfo."
* commit '5557f3391d1fe27269290fb65498d06f7ed42425':
Move the lib paths out of RSInfo.
|
| 9ab5094dd32352b33e251e540934f6e814c5fa5b |
19-Jun-2014 |
Jean-Luc Brouillet <jeanluc@google.com> |
Move the lib paths out of RSInfo.
Change-Id: Ic65874459509043fef261f7e0122aade4e557c57
sCpuScript.cpp
sCpuScript.h
|
| 60b4af5728f5ea84aa3bb0d4ce556d91c3c9bed4 |
18-Jun-2014 |
Andrew Hsieh <andrewhsieh@google.com> |
am 8b0ab0a6: Merge "Fix Windows SDK build."
* commit '8b0ab0a6c33816485acf4a9a1864d5e4fd645ce2':
Fix Windows SDK build.
|
| 8b0ab0a6c33816485acf4a9a1864d5e4fd645ce2 |
18-Jun-2014 |
Andrew Hsieh <andrewhsieh@google.com> |
Merge "Fix Windows SDK build."
|
| 1409b09faf025bf48edf061afee683c22ad320e8 |
17-Jun-2014 |
Jason Sams <jsams@google.com> |
am d94d3b91: Merge "Use relative addresses in Blend function table."
* commit 'd94d3b91b3307610adbf48d11d4ab5e0f04e5ec5':
Use relative addresses in Blend function table.
|
| d94d3b91b3307610adbf48d11d4ab5e0f04e5ec5 |
17-Jun-2014 |
Jason Sams <jsams@google.com> |
Merge "Use relative addresses in Blend function table."
|
| 8d50174ff0e80bc3846f9dfd6f7d154fc4949db4 |
26-May-2014 |
Logan Chien <tzuhsiang.chien@gmail.com> |
Fix Windows SDK build.
* Use gcc instead of clang in order to get proper exception support.
* Use libstdc++ instead of libc++ since gcc does not support _Atomic.
* Fix various 32/64 typing in linkloader.
Change-Id: I34abcf1e9fa98633e0fd35b4e1b71a0d5389008a
ndroid.mk
inkloader/android/librsloader.cpp
inkloader/include/impl/ELFObject.hxx
|
| 35e1335de1ae815513af5b0b93c5794b653e25ac |
14-Jun-2014 |
Tim Murray <timmurray@google.com> |
am b26cd991: Merge "Disable NEON libclcore for AArch64."
* commit 'b26cd991dc7f1eea42f1688d1db303ed7775d043':
Disable NEON libclcore for AArch64.
|
| 53b0d778806fb21fb653f5346595ae95dc51b5e1 |
13-Jun-2014 |
Stephen Hines <srhines@google.com> |
am f6328b8d: Merge "Resolve R_X86_64_PC32 relocation type in RS."
* commit 'f6328b8db3b5de20fb27812157db2ff8bda4669c':
Resolve R_X86_64_PC32 relocation type in RS.
|
| 3a5382c2514ecf11454f169415f9c35a6c00968f |
13-Jun-2014 |
Stephen Hines <srhines@google.com> |
am 1109ae5c: Merge "Add stub for x86 & x86_64."
* commit '1109ae5cbbebb4922aba0ae84cd859c6695a6184':
Add stub for x86 & x86_64.
|
| e43e66fdbbdb461655499ec035ab5720b6520dab |
12-Jun-2014 |
Dave Allison <dallison@google.com> |
am 9d7e8d94: Merge "Fix CALL26/JUMP26 relocations for AARCH64"
* commit '9d7e8d948791e5f7d87ba71f92ef2d6d4e012caa':
Fix CALL26/JUMP26 relocations for AARCH64
|
| b26cd991dc7f1eea42f1688d1db303ed7775d043 |
14-Jun-2014 |
Tim Murray <timmurray@google.com> |
Merge "Disable NEON libclcore for AArch64."
|
| bee48d79ba974e3dfbb782f9cce5c8d554f488e7 |
13-Jun-2014 |
Tim Murray <timmurray@google.com> |
Disable NEON libclcore for AArch64.
On ARMv8/AArch64, we can use the standard libclcore.bc for any precision,
as ASIMD is full IEEE 754 precision.
Change-Id: Iecff433b20f40a7f2272e497ad2e47a2f4b63ef3
ndroid.mk
sCpuScript.cpp
|
| 1109ae5cbbebb4922aba0ae84cd859c6695a6184 |
13-Jun-2014 |
Stephen Hines <srhines@google.com> |
Merge "Add stub for x86 & x86_64."
|
| 72151aac81a3cc92b9777aaa9b80f31a4ce9fcf7 |
12-Jun-2014 |
Xiaofei Wan <xiaofei.wan@intel.com> |
Add stub for x86 & x86_64.
Add stub support for x86, this will be used in x86 & x86_64 RS relocation.
Change-Id: Id5d1cc0679b43348a4f66707c21c6bdf85a3a7e4
Signed-off-by: Xiaofei Wan <xiaofei.wan@intel.com>
inkloader/include/ELFSectionProgBits.h
inkloader/include/StubLayout.h
inkloader/lib/StubLayout.cpp
|
| 77867c3e8dbad29699db85099551004d9af560bf |
12-Jun-2014 |
Dave Allison <dallison@google.com> |
Fix CALL26/JUMP26 relocations for AARCH64
This fixes the code to relocate CALL26 and JUMP26 AARCH64 relocations. It will
jump directly to the target address if it is in range, otherwise it will use
a linker veneer (stub).
Change-Id: I816de505b95a5797e1ba3bf1023dfc24484e5da8
inkloader/include/impl/ELFObject.hxx
inkloader/include/impl/ELFSectionProgBits.hxx
|
| 29dec29b96f513ceaa06ec1a8c422579a0a3a320 |
13-Jun-2014 |
Xiaofei Wan <xiaofei.wan@intel.com> |
Resolve R_X86_64_PC32 relocation type in RS.
Add stub when offset is out of range in x86_64
Change-Id: I7ad4879545971cd3ce59b8634c943bf4ee723fd8
Signed-off-by: Xiaofei Wan <xiaofei.wan@intel.com>
inkloader/include/impl/ELFObject.hxx
|
| 4b2ea46bdc7676d1a0328709435ae57f0d7d7289 |
11-Jun-2014 |
Jason Sams <jsams@google.com> |
am d45b4439: Merge "Convert LUT loop to use scalar data types."
* commit 'd45b443972955574f0a36c46855760eefd63e66b':
Convert LUT loop to use scalar data types.
|
| ed157c86dca0f671ea11d57a1b884cd15f76cdb2 |
11-Jun-2014 |
Jason Sams <jsams@google.com> |
Convert LUT loop to use scalar data types.
Change-Id: I53f6d87a4d7093ee0dc20371dcf4cdb8396522fd
sCpuIntrinsicLUT.cpp
|
| 1c9a5e1e15ccef0ea769750a3d7353cd5fbe4733 |
10-Jun-2014 |
Jean-Luc Brouillet <jeanluc@google.com> |
am a979674d: Merge "Remove FAKE_ARM64_BUILD now that we have 64 build working."
* commit 'a979674d5063a947b2d8f340e306f51b4a0632fb':
Remove FAKE_ARM64_BUILD now that we have 64 build working.
|
| a979674d5063a947b2d8f340e306f51b4a0632fb |
10-Jun-2014 |
Jean-Luc Brouillet <jeanluc@google.com> |
Merge "Remove FAKE_ARM64_BUILD now that we have 64 build working."
|
| f4d216ee0f0ef18edbde208a61dc4e2fe4c8f050 |
10-Jun-2014 |
Jean-Luc Brouillet <jeanluc@google.com> |
Remove FAKE_ARM64_BUILD now that we have 64 build working.
Change-Id: Ieb6627adcd1d097fe709559e182efcfc6841d13e
sCpuScript.cpp
sCpuScript.h
|
| 0d53201417f03300689ec7b9ccf6a7d66aaeed52 |
10-Jun-2014 |
Dave Allison <dallison@google.com> |
am 45927b48: Merge "Add R_AARCH64_ADD_ABS_LO12_NC relocation to RS linkloader."
* commit '45927b4889bdc69e7403f90cd3e9e9e1bfb445b1':
Add R_AARCH64_ADD_ABS_LO12_NC relocation to RS linkloader.
|
| d25f7a0a19446fa3d0c20fcf375e11e3fb5d1a0a |
09-Jun-2014 |
Dave Allison <dallison@google.com> |
Add R_AARCH64_ADD_ABS_LO12_NC relocation to RS linkloader.
Add the missing relocation.
Also add R_AARCH64_LDST8_ABS_LO12_NC as I noticed this was missing too.
Bug: 15448747
Change-Id: I6e79017e8ef659eb8913347c4eacb5ff0ef008a6
inkloader/include/impl/ELFObject.hxx
|
| f07408eb1df70fad7d842067e2991db95815d286 |
04-Jun-2014 |
Stephen Hines <srhines@google.com> |
am c2c41ff8: Merge "[MIPS64] Enable mips64 build"
* commit 'c2c41ff87105cbc67ad55daf9339ba2a54bf341c':
[MIPS64] Enable mips64 build
|
| e8e408ada85b77e0a7314ef8e734aa9fc981a237 |
14-May-2014 |
Duane Sand <duane.sand@imgtec.com> |
[MIPS64] Enable mips64 build
Change-Id: Iad5141ff83be3497c577c63177516f899775e271
ndroid.mk
|
| 6e7e258316f72be95039278e88e3bc1daea1668f |
06-May-2014 |
Simon Hosie <simon.hosie@arm.com> |
Arbitrary sub-rectangle handling for ColorMatrix.
AArch64 code for odd-length cases, and correct pointer offsets.
Change-Id: I28049a768a1e3e65611898904fa42bd295208871
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsics_advsimd_ColorMatrix.S
|
| 4650eba1dfd0dd6d33aca105adceee7aeff29cfe |
03-Jun-2014 |
Tim Murray <timmurray@google.com> |
am 5e0109a2: am f34ba456: Merge "Modify script loading behavior."
* commit '5e0109a20a4c6c950c0e4938a204acb4dec82d7e':
Modify script loading behavior.
|
| 29809d1f95d4cd4cbc6b2f9384b3321759691e13 |
28-May-2014 |
Tim Murray <timmurray@google.com> |
Modify script loading behavior.
This moves away from RSInfo and toward bcinfo. This also disables all caching for the time being.
Change-Id: Id7c717258d51e0b45540a78596f1bc9c94468efb
sCpuScript.cpp
sCpuScript.h
|
| 9cca69a233b87786ef27dd547fca7cea2d402faf |
31-May-2014 |
Tim Murray <timmurray@google.com> |
am 482bd65e: am 822db82c: Merge "Update runtime stubs."
* commit '482bd65e9bbc3d4cf012e72158e0142fb468e856':
Update runtime stubs.
|
| 6a9cc720f106358880eb33f2ea48fe5f5d6d0c37 |
31-May-2014 |
Tim Murray <timmurray@google.com> |
Update runtime stubs.
Adds exp2f.
Change-Id: I97c83af4b3be2cdadea2c2e9243da4772d5b4e69
sCpuRuntimeMath.cpp
|
| 2b54b28c8235fb762fbafd88a3ef7857053987b3 |
23-May-2014 |
Simon Hosie <simon.hosie@arm.com> |
Use relative addresses in Blend function table.
Change-Id: I6399ab11daf6b13626116ef653ad9cf572a94ff5
sCpuIntrinsics_advsimd_Blend.S
|
| f6deeee12e6cfdcac4c0f5321095625226c44fdf |
27-May-2014 |
Tim Murray <timmurray@google.com> |
am 7a09ae78: am bebfcc59: Merge "Update runtime stubs. "
* commit '7a09ae78e200ea6cdf4f1f8280adcca68062872c':
Update runtime stubs.
|
| 60fe47db884673ace2b41c6a037a376bbd0fd670 |
27-May-2014 |
Tim Murray <timmurray@google.com> |
Update runtime stubs.
Adds rsDebug for doubles. Also adds additional powf lookup.
Change-Id: I0f0a815a2b143102bcbf6d274d1ca4a0d897b651
sCpuRuntimeMath.cpp
sCpuRuntimeStubs.cpp
|
| e923f9a978362b1d4c0cd17aacb4a830c49e60b7 |
23-May-2014 |
Jason Sams <jsams@google.com> |
am ef32455d: am 64c12679: Merge "Enable ARM64 intrinsics."
* commit 'ef32455d351ed3306a696e74742defbc0a4e9527':
Enable ARM64 intrinsics.
|
| 51be0e996482f47b40234e641b3310737a203c72 |
23-May-2014 |
Stephen Hines <srhines@google.com> |
am 60500c8b: am 529f0e00: Merge "Switch to the separate RS loading path."
* commit '60500c8be57d05bf24601d8e15c6ce83cf1a8968':
Switch to the separate RS loading path.
|
| 64c126798d1210b15a103ef98c09861f60a9d7de |
22-May-2014 |
Jason Sams <jsams@google.com> |
Merge "Enable ARM64 intrinsics."
|
| 529f0e0026d4e9c497cbaabd4712533951a6bd9d |
22-May-2014 |
Stephen Hines <srhines@google.com> |
Merge "Switch to the separate RS loading path."
|
| 074424a4ac5b093331df2c92e7a5bcbfff136b71 |
22-May-2014 |
Jason Sams <jsams@google.com> |
Enable ARM64 intrinsics.
This also moves ARM intrinsic ifdefs behing ARCH_ARM_USE_INTRINSICS instead of ARCH_ARM_HAVE_VFP.
Change-Id: I48d3d55c77feb931e22288828247e281db43d32b
ndroid.mk
sCpuCore.cpp
sCpuIntrinsic3DLUT.cpp
sCpuIntrinsicBlend.cpp
sCpuIntrinsicBlur.cpp
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsicConvolve3x3.cpp
sCpuIntrinsicConvolve5x5.cpp
sCpuIntrinsicLoopFilter.cpp
sCpuIntrinsicYuvToRGB.cpp
|
| 25e3af55a43faddced1a9931574dfdc3cc8ad8fd |
22-May-2014 |
Stephen Hines <srhines@google.com> |
Switch to the separate RS loading path.
This change goes with a similar change to libbcc that separates compilation
from loading. We create our own symbol resolvers in the driver.
Change-Id: Ifdeed588d5935c49a1e19bdc46d0a8f0b9252e00
sCpuScript.cpp
sCpuScript.h
|
| daeeba27842fc77e99652e93a1a1cbf6f14044b2 |
22-May-2014 |
Dave Allison <dallison@google.com> |
am ca48a48d: am f981663b: Merge "Add aarch64 relocations"
* commit 'ca48a48de6529529fe9e5ecd7465acfd6780f721':
Add aarch64 relocations
|
| 723ba16bac04e65c147742fa08ae2b87da3c0fd5 |
20-May-2014 |
Dave Allison <dallison@google.com> |
Add aarch64 relocations
This adds missing aarch64 relocations for the RS linker.
Bug: 15020895
Change-Id: Ibcd8537426c57247ac6f7cbc358bfc9b78546249
inkloader/include/impl/ELFObject.hxx
|
| cdd409931b5802ba1284abcad06a4ce714a68522 |
19-May-2014 |
Stephen Hines <srhines@google.com> |
am c348888c: am 57fa42d4: Merge "Switch to standalone bcc compilation path (again)."
* commit 'c348888c5a24a52c02417840110b0e06cea6617d':
Switch to standalone bcc compilation path (again).
|
| d69b0eb7f8c180b401b45376144ee05af078b13a |
15-May-2014 |
Stephen Hines <srhines@google.com> |
Switch to standalone bcc compilation path (again).
Bug: 7342767
Change-Id: I6d876742cbd4dbbc7fac1d559a7c54da9e118889
sCpuScript.cpp
|
| 7e47dca8875d31953a6e3205954b885866d5533a |
14-May-2014 |
Jason Sams <jsams@google.com> |
am 82325ad0: am 53a408b7: Merge "YuvToRGB sub-rectangle handling."
* commit '82325ad020e458b896a3d80ed3fdc9485b247902':
YuvToRGB sub-rectangle handling.
|
| 53a408b7cfb5992eeb159ec305f16dd112ad3080 |
13-May-2014 |
Jason Sams <jsams@google.com> |
Merge "YuvToRGB sub-rectangle handling."
|
| 2d28948a13f296ad9937893ccfaabfdb91019bb9 |
13-May-2014 |
Jason Sams <jsams@google.com> |
am bdf8dc3a: am bbbd923d: Merge "Fix some empty matrix cases in ColorMatrix."
* commit 'bdf8dc3a0dfe6501da17b8c75ac8031717e90aa6':
Fix some empty matrix cases in ColorMatrix.
|
| bbbd923d622eb0b9db27157629955f4f64e924a9 |
13-May-2014 |
Jason Sams <jsams@google.com> |
Merge "Fix some empty matrix cases in ColorMatrix."
|
| 43bd992337fb8efde6f7bf02d68b858720760a60 |
13-May-2014 |
Jason Sams <jsams@google.com> |
am 7eac4da7: am 3849dae1: Merge "Handle 3DLUT odd-length cases in assembly."
* commit '7eac4da73a5bb7721da9328cfa82e339501d5913':
Handle 3DLUT odd-length cases in assembly.
|
| 3849dae10c6af1abc36de4ac856879f2728415bc |
13-May-2014 |
Jason Sams <jsams@google.com> |
Merge "Handle 3DLUT odd-length cases in assembly."
|
| f0e571f6d5a352ce547bfee67f670bbde7b665fe |
09-May-2014 |
Stephen Hines <srhines@google.com> |
am 6a194e62: am 8711f9da: Merge "Use the proper triple when compiling with bcc."
* commit '6a194e6206250b738558ea725cf3129a6a90cc32':
Use the proper triple when compiling with bcc.
|
| 8711f9dad12664fd981eec0db6bed5da8ef4e6cc |
09-May-2014 |
Stephen Hines <srhines@google.com> |
Merge "Use the proper triple when compiling with bcc."
|
| 9daf5e00a26395934285055c011fbf94a4665447 |
08-May-2014 |
Jason Sams <jsams@google.com> |
am 7da83646: am 5d1e327e: Merge "Strict clamping when packing to u8."
* commit '7da8364655bed45823c22d5d809b5dfcff596585':
Strict clamping when packing to u8.
|
| 5d1e327e51838ebc3baeb35db2d2d00603b7a86a |
08-May-2014 |
Jason Sams <jsams@google.com> |
Merge "Strict clamping when packing to u8."
|
| 82e0a6779cb1006921153081fb33374197e7c9de |
06-May-2014 |
Stephen Hines <srhines@google.com> |
Use the proper triple when compiling with bcc.
Change-Id: I6cbd14e1e196012583929dd383803903dc80e9ff
ndroid.mk
sCpuScript.cpp
|
| f7805ffb93d7e0daf2164549a75a8ced74c7f376 |
08-May-2014 |
Ying Wang <wangying@google.com> |
am 0b048ffd: am 5ffc8d2f: Merge "Remove unused LOCAL_LDLIBS."
* commit '0b048ffd29c8c1e7836b2d50fe03e020a7d8235b':
Remove unused LOCAL_LDLIBS.
|
| af353bfc2d857f8c800d3feecff3353e0c5f41e9 |
08-May-2014 |
Ying Wang <wangying@google.com> |
Remove unused LOCAL_LDLIBS.
Change-Id: Ic0b3b5dc136f4b537136936fa3f0c36a71f2c4ba
ndroid.mk
|
| 1d9c887c58d115975e01c9d500595f503803dc8c |
02-May-2014 |
Simon Hosie <simon.hosie@arm.com> |
YuvToRGB sub-rectangle handling.
Fix some difficult edge cases when processing only a portion of the image.
Also fix a register-marshalling bug in AArch64 assembly.
Change-Id: I8cd67f394fb42b216b2c3c7401e90eb2b86fca3d
sCpuIntrinsicYuvToRGB.cpp
sCpuIntrinsics_advsimd_YuvToRGB.S
sCpuIntrinsics_neon_YuvToRGB.S
|
| 07e4665c04a71462e6cfc1c2bb2300a9ed111e60 |
10-Apr-2014 |
Simon Hosie <simon.hosie@arm.com> |
Handle 3DLUT odd-length cases in assembly.
Change-Id: I43802cad1a8ae74e369791a88a9644dc389519af
sCpuIntrinsic3DLUT.cpp
sCpuIntrinsics_advsimd_3DLUT.S
sCpuIntrinsics_neon_3DLUT.S
|
| 42488d249e0c5b2f12973d576261bda74b54c85f |
16-Apr-2014 |
Simon Hosie <simon.hosie@arm.com> |
Strict clamping when packing to u8.
Change-Id: I511ec4b7f0985c9404a5d1daabcf97fe380c62af
sCpuIntrinsics_neon_ColorMatrix.S
|
| c7c255e86b2cbd36e4da94632c49b3c8b4f74031 |
08-Mar-2014 |
Simon Hosie <simon.hosie@arm.com> |
Fix some empty matrix cases in ColorMatrix.
Change-Id: I77ee44844472c647f12e80e2d68c11e2b272e595
sCpuIntrinsicColorMatrix.cpp
|
| 98950317aadf5e19c613644ce7a0e7ee293ac30e |
06-May-2014 |
Jason Sams <jsams@google.com> |
am be75fc2d: am a5b0367f: Merge "Switch to using LLVM intrinsics for conversions"
* commit 'be75fc2da17e719631329c0338565db9d0dda780':
Switch to using LLVM intrinsics for conversions
|
| 70404fa9781f13a8ca07d621f74f3810daab9704 |
05-May-2014 |
Jason Sams <jsams@google.com> |
Switch to using LLVM intrinsics for conversions
Replaces the hand written .ll files for conversion.
Change-Id: Ie8c6c8354847ddc84217f163060a9ee909496327
sCpuIntrinsicInlines.h
|
| b5cf2b46ed356ac8c5e2674fbf8afa4ec7a17b2e |
02-May-2014 |
Stephen Hines <srhines@google.com> |
am 5fe9a5a1: am 345db92b: Merge "Revert "Switch to standalone bcc compilation path.""
* commit '5fe9a5a1aa10afc7a9e7e9a4ca90a33eb6069a7d':
Revert "Switch to standalone bcc compilation path."
|
| 748a43678b33a7d814f4466f2846b8c955a9e1e8 |
02-May-2014 |
Stephen Hines <srhines@google.com> |
Revert "Switch to standalone bcc compilation path."
The current standalone bcc on the device needs to be able to select from
multiple possible target triples before we enable it by default.
Bug: 14470052
This reverts commit 99281b5bc2a81a42fa490ddbc1496621c2d84486.
Change-Id: I68f98b3c4f78d708aa5f544d36570e3db9c3e7d1
sCpuScript.cpp
|
| dcefcec86f9b99ed944ad118ecac89b598a3460b |
01-May-2014 |
Jean-Luc Brouillet <jeanluc@google.com> |
am 5470b9fb: am 2d08f007: Merge "Fix comments in gaussian blur"
* commit '5470b9fb37bf149a39cd409476b8f093cf00e7d7':
Fix comments in gaussian blur
|
| 2d08f007e16273691bbfeba45d43257dafee1dc6 |
01-May-2014 |
Jean-Luc Brouillet <jeanluc@google.com> |
Merge "Fix comments in gaussian blur"
|
| e09b71b3efd482b2fc45e0e742257e19a2964245 |
30-Apr-2014 |
Stephen Hines <srhines@google.com> |
am 0b66b5f7: am df531391: Merge "Switch to standalone bcc compilation path."
* commit '0b66b5f727d541ba44aa28d7362d1748c830b3bc':
Switch to standalone bcc compilation path.
|
| 212304e1395963e5dc79496e5133718c337dd2c0 |
30-Apr-2014 |
Stephen Hines <srhines@google.com> |
am 912029f8: am 0ddfb1eb: Merge "Configure standalone bcc compiler to work with plugin libraries."
* commit '912029f88ab9a1c9fa8f96c429e0fafac33b3f34':
Configure standalone bcc compiler to work with plugin libraries.
|
| df53139188b08a70f1a94fb9f476cbc3364f68c8 |
30-Apr-2014 |
Stephen Hines <srhines@google.com> |
Merge "Switch to standalone bcc compilation path."
|
| 0ddfb1ebd5ef0630e7351a202248285ada097d5f |
30-Apr-2014 |
Stephen Hines <srhines@google.com> |
Merge "Configure standalone bcc compiler to work with plugin libraries."
|
| 8b7117dfbcca8a74dfd83aaff7f82d2d7e3abd8f |
30-Apr-2014 |
Jean-Luc Brouillet <jeanluc@google.com> |
Fix comments in gaussian blur
Change-Id: Ifae435f9b2e65d6064412d09558b0c6ea32aea25
sCpuIntrinsicBlur.cpp
|
| f4c981e1abbbf877ac418637b4b05998da71b463 |
30-Apr-2014 |
Stephen Hines <srhines@google.com> |
am 90e7e571: am 41718a0d: Merge "Switch to __builtin___clear_cache(), since __clear_cache is deprecated."
* commit '90e7e571c48cd7bb6821f475241a421bbc06d193':
Switch to __builtin___clear_cache(), since __clear_cache is deprecated.
|
| 41718a0de646f7bc389b26ea871c62ed3494fb94 |
30-Apr-2014 |
Stephen Hines <srhines@google.com> |
Merge "Switch to __builtin___clear_cache(), since __clear_cache is deprecated."
|
| 99281b5bc2a81a42fa490ddbc1496621c2d84486 |
04-Feb-2014 |
Stephen Hines <srhines@google.com> |
Switch to standalone bcc compilation path.
Bug: 7342767
Change-Id: Idabdc75cb7daefa44177b23efe6ef8c3df4e908a
sCpuScript.cpp
|
| 005113297b19ed256b6db9d6bc293ed9266899fc |
31-Jan-2014 |
Stephen Hines <srhines@google.com> |
Configure standalone bcc compiler to work with plugin libraries.
Bug: 7342767
This change adds support (hidden behind the EXTERNAL_BCC_COMPILER ifdef)
for loading plugin libraries via the external bcc toolchain. The external
bcc compiler loads the named library and will then invoke a customized
rsCompilerDriverInit() from that library.
Change-Id: I07c2ea68be54c2255d36926fd37e395db790ef8f
ndroid.mk
sCpuCore.cpp
sCpuCore.h
sCpuScript.cpp
sCpuScript.h
sd_cpu.h
|
| dc061e48d5b17d14cfd18c170b4331e67d857475 |
30-Apr-2014 |
Stephen Hines <srhines@google.com> |
Switch to __builtin___clear_cache(), since __clear_cache is deprecated.
Change-Id: Ib6ab398b39c72365dfa81c467bf6f56326f4097a
inkloader/utils/flush_cpu_cache.h
|
| 7b0cddeae03a7ba85e681e43cbf4c3083b6883b5 |
30-Apr-2014 |
Jason Sams <jsams@google.com> |
am e96419bd: am 5d56c90e: Merge "Add comments for fast-path RS kernel"
* commit 'e96419bd36a15340ac69634a0f177e02f30fef2d':
Add comments for fast-path RS kernel
|
| e7fd0788c9735acf76e3d251013d0aa2d7a900c4 |
30-Apr-2014 |
Jason Sams <jsams@google.com> |
am 0a637ca4: am f1304740: Merge changes Iafa1fe3d,I32dba6b3
* commit '0a637ca4db4ef739863662c203f4a002320d19e2':
Mark convert_(long/double) as API 21
Fix performance regression in color matrix.
|
| 858d0352934596aa46fe97a70f30d4d837f6fc7f |
30-Apr-2014 |
Jason Sams <jsams@google.com> |
Add comments for fast-path RS kernel
Change-Id: I4b7450383d7b4c788fdfd5555d79dc94836b9fee
sCpuIntrinsicColorMatrix.cpp
|
| 98dd4bb2b1b08f04dd5034fe0c69daa15f6cc6da |
30-Apr-2014 |
Jason Sams <jsams@google.com> |
Fix performance regression in color matrix.
Previous bug fix which added "len &= 3" was buggy.
Change-Id: I32dba6b340ced35a7686243c5d9c468b5ade1ad6
sCpuIntrinsicColorMatrix.cpp
|
| 065665487fe9f9fb2d2f4e52f99e6387845b02d2 |
25-Apr-2014 |
Stephen Hines <srhines@google.com> |
am a6d32f10: Merge "Update fw/rs for Clang 3.5."
* commit 'a6d32f109788d3a3275648664ea8b4452ecf7bdd':
Update fw/rs for Clang 3.5.
|
| 8c24cd62187fb963cb1b76d2e91ae8c247c7f6fb |
11-Apr-2014 |
Tim Murray <timmurray@google.com> |
Update fw/rs for Clang 3.5.
Change-Id: I0a9a1b1e61cb435f0102936ad9ffd51599c49e88
ndroid.mk
|
| d56fa5280ec42b2609d2b0baa96a2b7006c4a611 |
24-Apr-2014 |
Jason Sams <jsams@google.com> |
resolved conflicts for merge of d3fe4992 to klp-modular-dev-plus-aosp
Change-Id: I7204f824ea46d2d1fff9c5269526e586c0bf4e77
|
| 1fb4e815d77966fa38c65107e4c34f22f7fe0065 |
24-Apr-2014 |
Jason Sams <jsams@google.com> |
am 0a6fea1a: Merge "Improve RS intrinsics performance."
* commit '0a6fea1ac38508b4e48f0da6d1024e211e900e84':
Improve RS intrinsics performance.
|
| d3fe4992b47848deb9a2876951aeb0bb1c62ad3f |
24-Apr-2014 |
Jason Sams <jsams@google.com> |
Merge "Revert "Add VP9 inter-frame prediction intrinsic""
|
| 0a6fea1ac38508b4e48f0da6d1024e211e900e84 |
24-Apr-2014 |
Jason Sams <jsams@google.com> |
Merge "Improve RS intrinsics performance."
|
| ee0f4835e065ef08a6283e3f86cdc671a5a156c7 |
24-Apr-2014 |
Jason Sams <jsams@google.com> |
Revert "Add VP9 inter-frame prediction intrinsic"
This reverts commit 60498fe9679ea25a260a503d6dfd27cbc0a0c079.
Change-Id: I4d8bb284793874a08c0cc991c0e04ecc104e1e0f
Conflicts:
cpu_ref/Android.mk
cpu_ref/rsCpuCore.cpp
rsDefines.h
ndroid.mk
onvolve/convolve.c
onvolve/convolve8_avg_neon.s
onvolve/convolve8_neon.s
onvolve/convolve_avg_neon.s
onvolve/convolve_copy_neon.s
onvolve/convolve_neon.c
onvolve/vp9_common.h
onvolve/vp9_filter.h
sCpuConvolve.h
sCpuCore.cpp
sCpuIntrinsicInterPred.cpp
sCpuIntrinsicInterPred.h
|
| e660532259b1baf2c112e693360a42dc7bce4336 |
22-Apr-2014 |
Stephen Hines <srhines@google.com> |
am 70fe08dd: Merge "Skip the .ARM.exidx section, since it is currently unused."
* commit '70fe08dd27dc8fa682275a7057e8baafa6aa9369':
Skip the .ARM.exidx section, since it is currently unused.
|
| 0715b383c2d121de12af5bdcba5183edd753f523 |
21-Apr-2014 |
Stephen Hines <srhines@google.com> |
Skip the .ARM.exidx section, since it is currently unused.
Change-Id: I454e4cf311a3eed5cc3b1f4398bdbc7976a0d430
inkloader/include/impl/ELFObject.hxx
|
| 7b7060c61e4182b29186849c5a857ea5f0898e56 |
21-Apr-2014 |
Rose, James <james.rose@intel.com> |
Improve RS intrinsics performance.
Renderscript CPU performance for intrinsics cases is not good for x86 platforms.
In many cases it is significantly slower even with SIMD Intrinsics. In current x86 implementation
it is using full 32 bit multiplies which aren't well supported on current Atom platforms.
This patch uses 16 bit multiply with 32 bit add pmaddwd instruction where appropriate.
It also adds atom specificoptimizations to improve RS intrinsics performance.
Change-Id: Ifc01b5a6d6f7430d2dc218f1618b9df3fb7937fe
Signed-off-by: Xiaofei Wan <xiaofei.wan@intel.com>
ndroid.mk
sCpuCore.cpp
sCpuIntrinsicBlend.cpp
sCpuIntrinsicBlur.cpp
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsicConvolve3x3.cpp
sCpuIntrinsicConvolve5x5.cpp
sCpuIntrinsics_x86.c
|
| e7c71b42e03d596a5e2f49b9cb76c479c4956cd9 |
17-Apr-2014 |
Tim Murray <timmurray@google.com> |
am 48360016: Merge "Fix ARM64 build."
* commit '4836001604783fb7e3ce0ffd9a33afaf34865bba':
Fix ARM64 build.
|
| 810aa7a47f3dfe2b207e755121e801ae88bda0fa |
17-Apr-2014 |
Jason Sams <jsams@google.com> |
resolved conflicts for merge of 0d6043ca to klp-modular-dev-plus-aosp
Change-Id: Ic0b72a5ae43d78ed224d2aeda25a83a79f9f26ec
|
| 39ab94aafb7f0916a7f6e345ee1fa0f5ff3bbacd |
17-Apr-2014 |
Jason Sams <jsams@google.com> |
Bicubic resize intrinsic
Change-Id: Ie869484505c3e25e8ea57ff208b9e052ee8dca7b
ndroid.mk
sCpuCore.cpp
sCpuIntrinsicResize.cpp
|
| 0d6043caef208ee6c661eb17bcb376abfe90cd48 |
17-Apr-2014 |
Jason Sams <jsams@google.com> |
Bicubic resize intrinsic [DO NOT MERGE]
Change-Id: I49fec0d33f0aab9b32438e4e97db18ea036989a7
ndroid.mk
sCpuCore.cpp
sCpuIntrinsicResize.cpp
|
| 6cdb84d219baf95287e223c8d078c1852fa1a29a |
17-Apr-2014 |
Tim Murray <timmurray@google.com> |
Fix ARM64 build.
Change-Id: Ic947b16baaea22a0a6ad02da938ea48b986b2f0e
inkloader/include/impl/ELFObject.hxx
|
| db2ffd2e9c7b7949905621e55a73d3431cf4fc90 |
04-Apr-2014 |
WeiTang <wei.a.tang@intel.com> |
Enable linker loader support for 64bit elf object
Change-Id: I6425c427d6a9eddd30e44abf595601180ca52eba
Conflicts:
cpu_ref/linkloader/include/impl/ELFObject.hxx
inkloader/android/librsloader.cpp
inkloader/include/impl/ELFObject.hxx
|
| cc141397f74536754b9ca6d847f7385b38f92b50 |
16-Apr-2014 |
Tim Murray <timmurray@google.com> |
Merge "Resolve x86_64 relocations in RS."
|
| c726af7297d4818264999ced4078bdae728b9a04 |
11-Apr-2014 |
Dave Allison <dallison@google.com> |
Merge "Add support for AARCH64 relocations to the RS linkloader."
|
| 7c4d899920c9aecb5e64da2f667a4a24496bbbe5 |
08-Apr-2014 |
Jason Sams <jsams@google.com> |
Fix potential overrun by optimized code.
Change-Id: I0435eb04d51c3f95a9db4af2dbd7da0919dfe366
sCpuIntrinsic3DLUT.cpp
|
| 863b8c326f138f570f2d055e1a14100d00fb8983 |
08-Apr-2014 |
Jason Sams <jsams@google.com> |
Fix color matrix bug.
Change-Id: I8087572249f73165136cec3520e8d1c29eea002f
sCpuIntrinsicColorMatrix.cpp
|
| adab849c946f3b18bcc6da06ad1a4efcec7cff3c |
01-Apr-2014 |
Matthieu Delahaye <matthieu@multicorewareinc.com> |
RS: Fix VP9 LoopFilter to handle 4K video
One of the buffer allocated within the intrinsic was not
wide enough to accomodate space needed for 4K videos.
Change-Id: I3079e8a4c175a25ea427cefc80958ee660674e0b
sCpuIntrinsicLoopFilter.cpp
|
| 30e2a4c8d848bc788cc5600e6b929b4f761f3f03 |
28-Mar-2014 |
Dave Allison <dallison@google.com> |
Add support for AARCH64 relocations to the RS linkloader.
This is still a WIP until we can verify what relocations are actually used by
the 64 bit compiler. This covers ABS, PREL, CALL and JUMP relocations only for
now.
This has not been tested but does build.
Change-Id: Iaa85c49e96a63e8c4eef225b8cab619a1cdb861f
inkloader/include/ELFObject.h
inkloader/include/ELFSectionProgBits.h
inkloader/include/StubLayout.h
inkloader/include/impl/ELFObject.hxx
inkloader/include/impl/ELFSectionRelTable.hxx
inkloader/lib/ELFHeader.cpp
inkloader/lib/StubLayout.cpp
|
| 4e5c414252846e96d2e353515134387d86150c68 |
16-Mar-2014 |
Simon Hosie <simon.hosie@arm.com> |
AArch64 ports of Convolve3x3 and Convolve5x5.
Change-Id: I68d9e0ddda128f07d4383d20e9bad161f0c6965b
ndroid.mk
sCpuIntrinsics_advsimd_Convolve.S
sCpuIntrinsics_neon.S
sCpuIntrinsics_neon_Convolve.S
|
| 5dcaaa5f50926bebf6877e254c521faa7e2593e3 |
16-Mar-2014 |
Simon Hosie <simon.hosie@arm.com> |
Optimisations to 3DLUT assembly.
Process more pixels at once to try to keep the register file fuller and more
tightly packed and allow more concurrency. Implementations in AArch32 and
AArch64 assembly.
Change-Id: I683078ff02155cc14bacce35bce3d3fe06857095
ndroid.mk
sCpuIntrinsic3DLUT.cpp
sCpuIntrinsics_advsimd_3DLUT.S
sCpuIntrinsics_neon.S
sCpuIntrinsics_neon_3DLUT.S
|
| ea76eb386a2d851d50be69ebeb7ae593f84a5be9 |
16-Mar-2014 |
Simon Hosie <simon.hosie@arm.com> |
Make Blur AArch64 assembly position-independent.
Change-Id: I426fba9fff3ac165f5be5f78e2458dbc3b59ab02
sCpuIntrinsics_advsimd_Blur.S
|
| 0462a39371659d1eeed5eb48dd6d507760301c22 |
08-Mar-2014 |
Simon Hosie <simon.hosie@arm.com> |
Add AArch64 assembly for ColorMatrix.
Change-Id: I2fcc57aceea08243d5132287f6de053b846c5fe7
ndroid.mk
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsics_advsimd_ColorMatrix.S
sCpuIntrinsics_neon_ColorMatrix.S
|
| 7e4f65d7477617a2e19a371e88f888f554f7736a |
16-Mar-2014 |
Simon Hosie <simon.hosie@arm.com> |
Avoid Clang's integrated AArch64 assembler.
Change-Id: I9babbce1e7b04bb1f8af29886d7cafb16fd8f71c
ndroid.mk
|
| 6dc86b492191f9062e912afd948e08362201f332 |
25-Mar-2014 |
Jason Sams <jsams@google.com> |
Merge "RS: Add VP9 LoopFilter Intrinsic"
|
| 41de3047284411afde00707e0a98fe2949bc04ec |
20-Mar-2014 |
Xiaofei Wan <xiaofei.wan@intel.com> |
Resolve x86_64 relocations in RS.
Change-Id: I7e32aa150bd04d6b48afd656291a85b0ca5c182d
Signed-off-by: Xiaofei Wan <xiaofei.wan@intel.com>
inkloader/include/impl/ELFObject.hxx
|
| 3cdd1fbfa0ad068b1b584643453883770425722c |
19-Mar-2014 |
Ian Rogers <irogers@google.com> |
Define FAKE_ARM64_BUILD for all 64bit targets.
Change-Id: I105ce21b11453b89c8544b907ebab97340ae02d3
ndroid.mk
|
| 0e780a7ce6794e53e475573dbcb8af01fe57e708 |
19-Mar-2014 |
Narayan Kamath <narayan@google.com> |
Merge "Move renderscript over to __clear_cache (compiler-rt)."
|
| 6fc3e12b8912458cb4adcfd32e2f53d76b0cc737 |
04-Mar-2014 |
Matthieu Delahaye <matthieu@multicorewareinc.com> |
RS: Add VP9 LoopFilter Intrinsic
Change-Id: I5caa46da2c825a95cc1ed35a1cdbcd6da0ffce88
ndroid.mk
sCpuCore.cpp
sCpuIntrinsicLoopFilter.cpp
p9_loopfilter_16_neon.S
p9_loopfilter_neon.S
p9_mb_lpf_neon.S
|
| 00dbeacfd62bdecd5fce9426c4795aec8618753b |
18-Mar-2014 |
Jason Sams <jsams@google.com> |
Merge "Revert "RS: Add VP9 LoopFilter Intrinsic""
|
| 933bdc9b648995ab68da746c6daa2206eec02b0f |
18-Mar-2014 |
Jason Sams <jsams@google.com> |
Revert "RS: Add VP9 LoopFilter Intrinsic"
This has build errors with the x86 SDK.
This reverts commit 64048e720cf940cb0f7f6f9a4ab4f061918a1fd9.
Change-Id: Ia712a46abd06e2a580853c863bfa53410b7f99e9
ndroid.mk
sCpuCore.cpp
sCpuIntrinsicLoopFilter.cpp
p9_loopfilter_16_neon.S
p9_loopfilter_neon.S
p9_mb_lpf_neon.S
|
| 21dbc8ba61dd6a5852b1346c14bd29373326c240 |
18-Mar-2014 |
Jason Sams <jsams@google.com> |
Merge "Solve four separate memory leaks related to rsdHalInit"
|
| 3028980f677cb19d94add01f6c0a9f54efd39306 |
18-Mar-2014 |
Jason Sams <jsams@google.com> |
Merge "RS: Add VP9 LoopFilter Intrinsic"
|
| 64048e720cf940cb0f7f6f9a4ab4f061918a1fd9 |
04-Mar-2014 |
Matthieu Delahaye <matthieu@multicorewareinc.com> |
RS: Add VP9 LoopFilter Intrinsic
Change-Id: If1ac77774c74b5513ce7a2db4ef31888a351a9c5
ndroid.mk
sCpuCore.cpp
sCpuIntrinsicLoopFilter.cpp
p9_loopfilter_16_neon.S
p9_loopfilter_neon.S
p9_mb_lpf_neon.S
|
| 07ef704308b514272ed2f5c3e6a2f4c055550158 |
19-Feb-2014 |
Jens Gulin <jens.gulin@sonymobile.com> |
Solve four separate memory leaks related to rsdHalInit
Three of the items are local to RsdCpuReferenceImpl and now freed in
destructor after all threads are stopped.
Last one is the RsdHal item itself where the pointer for some reason
was explicitly cleared but not freed. There is no reference counting
but it should be ok to free in Shutdown.
Change-Id: I7832e412d12f4bd7cc728481ae0c782fa57b57e4
sCpuCore.cpp
|
| 1e2aedbef554a10a16296d3b529327fffcb10e0d |
14-Mar-2014 |
Jason Sams <jsams@google.com> |
Revert "RS: Add VP9 LoopFilter Intrinsic"
This reverts commit e4749f3a5a6a6041ef2894162edce5115b307db0.
Change-Id: I45ccdacb1706abd4df7f635c5e64dcb1ee4b876d
ndroid.mk
sCpuCore.cpp
sCpuIntrinsicLoopFilter.cpp
p9_loopfilter_16_neon.S
p9_loopfilter_neon.S
p9_mb_lpf_neon.S
|
| e195a3f57ace3b66d313a6ee88c6e93d5c9d87f4 |
13-Mar-2014 |
Tim Murray <timmurray@google.com> |
Make RS compile on ARM64.
bug 13280327
Change-Id: I79080c448ca17d0cba8df075737c22b24ce9a085
ndroid.mk
sCpuIntrinsicColorMatrix.cpp
sCpuScript.cpp
sCpuScript.h
|
| 72f5f8c0e7833d9a4f0288700256982dc7127e23 |
11-Mar-2014 |
Narayan Kamath <narayan@google.com> |
Move renderscript over to __clear_cache (compiler-rt).
bug: 12965705
Change-Id: Ia9917a4aff377fd43a4e71c71a886639763eb250
inkloader/utils/flush_cpu_cache.h
sCpuIntrinsicColorMatrix.cpp
|
| e4749f3a5a6a6041ef2894162edce5115b307db0 |
04-Mar-2014 |
Matthieu Delahaye <matthieu@multicorewareinc.com> |
RS: Add VP9 LoopFilter Intrinsic
Change-Id: Ia49e56c7e21fee1601a0418bd105ef6429c336ca
ndroid.mk
sCpuCore.cpp
sCpuIntrinsicLoopFilter.cpp
p9_loopfilter_16_neon.S
p9_loopfilter_neon.S
p9_mb_lpf_neon.S
|
| 83f304cb26008d3f4da154cec19c3a12fa2e6c74 |
06-Mar-2014 |
Jason Sams <jsams@google.com> |
Fix build issues with external patch.
Change-Id: Ib5ea4338df179eb27e4ce9958ef42df1e3ac3eb1
ndroid.mk
sCpuCore.cpp
|
| 60498fe9679ea25a260a503d6dfd27cbc0a0c079 |
18-Feb-2014 |
Matthieu Delahaye <matthieu@multicorewareinc.com> |
Add VP9 inter-frame prediction intrinsic
Change-Id: If8985a6200fb6d34083eff711ccdf2f1b3c374e6
ndroid.mk
onvolve/convolve.c
onvolve/convolve8_avg_neon.s
onvolve/convolve8_neon.s
onvolve/convolve_avg_neon.s
onvolve/convolve_copy_neon.s
onvolve/convolve_neon.c
onvolve/vp9_common.h
onvolve/vp9_filter.h
sCpuConvolve.h
sCpuCore.cpp
sCpuIntrinsicInterPred.cpp
sCpuIntrinsicInterPred.h
|
| 02170e63931c2d93f57912efbf5d7706578a8438 |
05-Mar-2014 |
Ying Wang <wangying@google.com> |
Use arch-specific flags.
To fix multilib build.
Change-Id: Ifaaea96662de6d3d96a3448754dfe720c4e60deb
ndroid.mk
|
| 446788007efe0a673d0366284026adfa17b36fed |
20-Feb-2014 |
Simon Hosie <simon.hosie@arm.com> |
Optimisations to blur intrinsic.
Try to keep all data in-register whereever possible, and use only a minimal
circular buffer on the stack when necessary. Implementations in AArch32 and
AArch64 NEON.
Change-Id: If3dd4932a94ee1cadde46e298b8f6bf14b6c2bdc
ndroid.mk
sCpuIntrinsicBlur.cpp
sCpuIntrinsics_advsimd_Blur.S
sCpuIntrinsics_neon.S
sCpuIntrinsics_neon_Blur.S
|
| ccd7a46d0c0052209bf3ab8657f40622065d1d1f |
01-Feb-2014 |
Simon Hosie <simon.hosie@arm.com> |
Optimise YuvToRGB using 16-bit arithmetic.
Reimplement YuvToRGB intrinsic using 16-bit SIMD arithmetic to increase
throughput. Implementations in AArch32 and AArch64 NEON.
Change-Id: Idd43e383f5147c33b0b546fa822c970de432c19d
ndroid.mk
sCpuIntrinsicYuvToRGB.cpp
sCpuIntrinsics_advsimd_YuvToRGB.S
sCpuIntrinsics_neon.S
sCpuIntrinsics_neon_YuvToRGB.S
|
| 5d06919bc8019322180ea34768a7a4137fa64d11 |
20-Feb-2014 |
Simon Hosie <simon.hosie@arm.com> |
Revised NEON implementations of blend.
Reimplement blend intrinsic using wider memory accesses and a few more 8-bit
operations where possible. Implementations in AArch32 and AArch64 NEON.
Change-Id: I5e56010376b1db1628a911cf09d97baf5af289b3
ndroid.mk
sCpuIntrinsicBlend.cpp
sCpuIntrinsics_advsimd_Blend.S
sCpuIntrinsics_neon.S
sCpuIntrinsics_neon_Blend.S
|
| 95ad60aad51227e277d57b3669496f129b88add3 |
20-Feb-2014 |
Elliott Hughes <enh@google.com> |
Merge "Make renderscript immune from changes to bionic implementation details."
|
| eab7cd3de8639f8dfc01dcab59b2a472e4d7dbd1 |
20-Feb-2014 |
Elliott Hughes <enh@google.com> |
Make renderscript immune from changes to bionic implementation details.
This is a copy of what these macros looked like at the time you started
using them, which is presumably what you intended, though you may prefer
a different code alignment in future.
Change-Id: I8c061e2255df6e0c2f722063779488d9d2dfaab5
sCpuIntrinsics_neon.S
sCpuIntrinsics_neon_ColorMatrix.S
|
| 165aa63b99b77abc8c486679a9f90104ce9fff73 |
07-Feb-2014 |
Colin Cross <ccross@android.com> |
rs: support multilib builds
Generate source files into $(local-generated-sources-dir)
Restrict which architectures try to build using LOCAL_MODULE_TARGET_ARCH
Change-Id: Ibb1049c248ed4bcec32b81771f1f9fbcce66a390
ndroid.mk
|
| f8852d0494a260c583795a96a2a06c49b86a9b10 |
30-Jan-2014 |
Ian Rogers <irogers@google.com> |
Work-around 64bit build issues.
Log errors and abort as a temporary work-around for 64bit support.
Change-Id: I7f6b483d671189bd12ae0ef79515fcdd871eba17
inkloader/android/librsloader.cpp
inkloader/include/impl/ELFObject.hxx
|
| 8994abbe699bb05fa70cff101becc925db6b2c26 |
05-Dec-2013 |
synergy dev <synergye@codefi.re> |
cpu_ref: fix c++11 narrowing violations
the best fix for these is to use static casting
Change-Id: I00344826a33e5387638ae5edf4c6b85c92f43769
sCpuIntrinsic3DLUT.cpp
sCpuIntrinsicYuvToRGB.cpp
|
| 4bc1a0ce2124624045f5b19d027175d3598670fa |
13-Dec-2013 |
Stephen Hines <srhines@google.com> |
Merge "Add missing functions and tests for support library."
|
| d9b0f02db23455d7bd2f11fda871b2af9120f85e |
13-Dec-2013 |
Elliott Hughes <enh@google.com> |
Stop using the PLD macro.
Every ARM processor we run on has the pld instruction.
Change-Id: I05512bf20a73590698c714d6921db4e44e971606
sCpuIntrinsics_neon.S
|
| cadee38f6940e5584d3ec3398a21bd2a494361e2 |
12-Dec-2013 |
Stephen Hines <srhines@google.com> |
Add missing functions and tests for support library.
Missing functions:
rsFrac(float)
rsMatrixLoadRotate(rs_matrix4x4 *, float, float, float, float)
rsMatrixLoadScale(rs_matrix4x4 *, float, float, float)
rsMatrixLoadTranslate(rs_matrix4x4 *, float, float, float)
rsMatrixRotate(rs_matrix4x4 *, float, float, float, float)
rsMatrixScale(rs_matrix4x4 *, float, float, float)
rsMatrixTranslate(rs_matrix4x4 *, float, float, float)
rsMatrixLoadOrtho(rs_matrix4x4 *, float, float, float, float, float, float)
rsMatrixLoadFrustum(rs_matrix4x4 *, float, float, float, float, float, float)
rsMatrixLoadPerspective(rs_matrix4x4 *, float, float, float, float)
rsMatrixInverse(rs_matrix4x4 *)
rsMatrixInverseTranspose(rs_matrix4x4 *)
rsMatrixTranspose(rs_matrix4x4 *)
rsMatrixTranspose(rs_matrix3x3 *)
rsMatrixTranspose(rs_matrix2x2 *)
rsUptimeNanos()
rsGetDt()
Headers tested in this CL:
rs_element.rsh
rs_math.rsh
rs_matrix.rsh
rs_object.rsh
rs_sampler.rsh
rs_time.rsh
Change-Id: I6cb1972a44f66dc7fed8fe3cf0329720d166af81
sCpuRuntimeMath.cpp
|
| c859de1556dc9ac5c6f5647298bb5c685c5d2bb0 |
02-Dec-2013 |
Jason Sams <jsams@google.com> |
Fix bug in blur when processing odd widths
Change-Id: I10537cb623cf230636f191fd2b84b050d0f42700
sCpuIntrinsicBlur.cpp
|
| 6de1d8375526502b468232d77fce3e957c705137 |
14-Nov-2013 |
Tim Murray <timmurray@google.com> |
Fix histogram intrinsic.
bug 11676706
Change-Id: I12ac5ed5f8528574619feb9737737d0111ad0be6
sCpuIntrinsicHistogram.cpp
|
| ee48c0bbf290a73e2cd4710b70d62fc203dac0dc |
31-Oct-2013 |
Stephen Hines <srhines@google.com> |
Create a cache directory for our symlinked files if necessary.
Bug: 11519886
Change-Id: I93d3c194a15e4842455c31406095e90bced882ee
sCpuScript.cpp
|
| 75adb8213f045bf3ffbc5deb1350b36d486e228a |
22-Oct-2013 |
Jason Sams <jsams@google.com> |
Fix crash running blur on 4K images
bug 11258120
The temporary buffer could be misaligned due to realloc only
aligning to 8 bytes.
Fix issues with with Histogram and 2/3 vector sizes.
Change-Id: I271f4635ead5ff0a7e7f89bb5cfcbc2bba626e22
sCpuIntrinsicBlur.cpp
sCpuIntrinsicHistogram.cpp
|
| 4da42506a08ed7fdb61615b3524f111df939fc6e |
03-Oct-2013 |
Stephen Hines <srhines@google.com> |
Remove references/use of ARCH_X86_HAVE_* in frameworks/rs.
Bug: 11048298
These SSE-related defines are unnecessary because all x86 Android devices
have these features. We switch all the checks to __i386__ instead.
This also fixes an issue where non-ARM bitcode files were being created with
the ARM defines present. This is problematic when we then try to do things
that would be arch-specific (i.e. the presence of SSE-enabled functions for
clamp(), ...).
This also cleans up the different target architecture specified in the x86
bitcode files so that they match the platform target (i686-unknown-linux).
Change-Id: I4776bbdce360de26e8a00e05d2cb19341d94a173
ndroid.mk
sCpuScript.cpp
|
| 4a1495f3404b939459aa817f06ea9325ca1a4625 |
30-Sep-2013 |
Jason Sams <jsams@google.com> |
The NEON yuv asm code may overrun the buffer
by up to four bytes while reading U & V. Adjust the cutoff
for the boundary to allow for this.
bug 10826418
Change-Id: I93252a5f823cfd0dbf65fbfc194b4848cf44557a
sCpuIntrinsicYuvToRGB.cpp
|
| 0052f8dcb5ebb5c9205a1d25445629fb5a772380 |
20-Sep-2013 |
Jason Sams <jsams@google.com> |
Fix for YUV bugs
Two issues, YUV intrinsic would repeat u&v for the edge padding
During the support for flexible YUV, u&v got swapped in the accessor.
bug 10826418
Change-Id: I1cf6e27c2114807d6aece4f9dd44257d4f4aa477
sCpuIntrinsicYuvToRGB.cpp
|
| f5e947e74cea50b4687500970237c41ff698d9da |
12-Sep-2013 |
Jason Sams <jsams@google.com> |
Merge "Fix YUV intrinsic with camera input where frames are NULL" into klp-dev
|
| ac67f6d87fddf45cda71cce31770aa41cc435fbd |
12-Sep-2013 |
Jason Sams <jsams@google.com> |
Merge "Fix issues with U8->U8 when mixed with add or lane conversion" into klp-dev
|
| e99f3e21a0d88ca59eb9281699e8212a3a4205a3 |
12-Sep-2013 |
Jason Sams <jsams@google.com> |
Fix YUV intrinsic with camera input where
frames are NULL
bug 10710716
Change-Id: I9e69a8a51ccfc15798a09760249a951561dcd544
sCpuIntrinsicYuvToRGB.cpp
|
| ec3cd2dff915d476ce0d7cdbb20c0497635a700f |
12-Sep-2013 |
Jason Sams <jsams@google.com> |
Fix issues with U8->U8 when mixed with add or lane conversion
bug 10427746
Change-Id: Iea14223c754d0b2256d752d9211a0c45109c3a02
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsics_neon_ColorMatrix.S
|
| a23b201166cfab963488fc565a8b14d618b0f8d9 |
12-Sep-2013 |
Jason Sams <jsams@google.com> |
Merge "Fix ColorMatrix performance and correctness issues." into klp-dev
|
| 9e4a96af136dab5b21a37580d17cbcb89872114e |
12-Sep-2013 |
Jason Sams <jsams@google.com> |
Fix ColorMatrix performance and correctness issues.
bug 10427746
Change-Id: Ie42753a551badf4de3144f16fa0e407eca5b9c74
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsics_neon_ColorMatrix.S
|
| c78839b5bbcffae7d64a5a1c9aa60c9a4c5d3918 |
11-Sep-2013 |
Stephen Hines <srhines@google.com> |
Detect use of RS objects in kernels to disable a ScriptGroup optimization.
Bug: 10394865
This change disables a ScriptGroup optimization when we can't guarantee
that there is no implicit ordering between two kernels. In this case, it
is possible to communicate between kernels using the same bound global
RS object (like Allocation). A subsequent kernel in the ScriptGroup could
accidentally pick up stale data, leading to incorrect results.
Now, we disable this optimization whenever we see binds/sets of variables
that are potentiall RS object types for a given Script/kernel. This is
overly conservative, but sufficient for now.
This change also fixes a small issue with preLaunch/postLaunch missing for
the default ScriptGroup execution case.
Change-Id: I0d19d200cc8dc397d68008a4df6ea423b1e4d04f
sCpuScriptGroup.cpp
|
| 17e3cdc24776d8fdbf1ce16287b9b4dcd516708f |
10-Sep-2013 |
Jason Sams <jsams@google.com> |
Fix ScriptGroup performance regression
bug 10151545
Change-Id: Ica4a30c6fe8718f7fdbff0b446885d9ac7083769
sCpuIntrinsicColorMatrix.cpp
sCpuScript.cpp
sCpuScript.h
sCpuScriptGroup.cpp
|
| 61656a7c6fc13421679d0a1cdf8b5b861e286892 |
04-Sep-2013 |
Jason Sams <jsams@google.com> |
Cleanup type offsets which cannot be calculated for flexible YUV.
Support flexible YUV
bug 10567550
Change-Id: I4f6e5a8d86eeee635605460f1751208f3320969b
(cherry picked from commit a75372759e288be3fb8835735a830b1f7d1a4c42)
sCpuIntrinsicYuvToRGB.cpp
sCpuIntrinsics_neon.S
|
| 927edc2b99b73d1f236320663ed7ed4cfeee2f91 |
31-Aug-2013 |
Tim Murray <timmurray@google.com> |
am 869d1d0b: am aa782376: Fix YUV intrinsic with compat lib.
* commit '869d1d0b7eb90c78ea4e621315895d178c6fd72a':
Fix YUV intrinsic with compat lib.
|
| aa782376e05c47141bc3f90966d609cfc69ebc39 |
30-Aug-2013 |
Tim Murray <timmurray@google.com> |
Fix YUV intrinsic with compat lib.
Change-Id: I05bfd1d4de24e3e9eadf092ec519d54ed411aa89
sCpuIntrinsicYuvToRGB.cpp
|
| 2b0d8e60d7daeffa7c9a5e11cfbfc9e5e04933b7 |
30-Aug-2013 |
Jason Sams <jsams@google.com> |
Float color matrix fixes
Fix histogram to use update API names.
bug 10427746
Change-Id: Idc9d6025380b2344d14f2aeb4da66bff47169021
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsics_neon_ColorMatrix.S
|
| 68c817211a6fe87bebed83d38a05fff32cc24a7e |
22-Aug-2013 |
Jason Sams <jsams@google.com> |
color matrix bugs
bug 10427746
Change-Id: Ie652fe5180702a2eceb3a7077bc1821ba21c1ee4
sCpuIntrinsicColorMatrix.cpp
|
| 5e37175676a1089fc7c40ac8b705799dccabcc30 |
16-Aug-2013 |
Stephen Hines <srhines@google.com> |
am fd47ac29: am 43cfc0cb: Remove barriers to using NDK v8 includes.
* commit 'fd47ac29478ed01a581b3326d898ad848858a314':
Remove barriers to using NDK v8 includes.
|
| fd5136767dee529cfcab02bb766b7d369e20ba58 |
16-Aug-2013 |
Stephen Hines <srhines@google.com> |
am 6699b978: am 11418c87: Provide a tgammaf() wrapper for compatibility on Gingerbread.
* commit '6699b9783ec6ed29b62c4f7bdb491fec1c51eaeb':
Provide a tgammaf() wrapper for compatibility on Gingerbread.
|
| 43cfc0cbe6e6e8f585a0ae5f1d9cc2859ab1dda7 |
15-Aug-2013 |
Stephen Hines <srhines@google.com> |
Remove barriers to using NDK v8 includes.
Bug: 10315692
Removed CLOCK_BOOTTIME, added some additional guards around cutils includes,
and explicitly included errno.h when we use ETIMEDOUT.
Change-Id: I0a764beea2f8ff7387a9afe985f00b5f03c68725
sCpuRuntimeMath.cpp
|
| 11418c87254f0cbffa910fe8f105b7da92452487 |
15-Aug-2013 |
Stephen Hines <srhines@google.com> |
Provide a tgammaf() wrapper for compatibility on Gingerbread.
Bug: 10315692
GB devices don't have tgammaf(float) and only have tgamma(double). We
provide a simple wrapper to convert between the two for the compatibility
library.
Change-Id: Id8566450405482e876148d62ffbe46fa62049049
sCpuRuntimeMath.cpp
|
| 8e0edc869aa8d7dbe8d629785b834a5e67d31a4d |
10-Aug-2013 |
Tim Murray <timmurray@google.com> |
Merge "fix SDK build breakage with NEON" into klp-dev
|
| 56b0b63c8d2ad0f880820db133a2e97ac56a942a |
09-Aug-2013 |
Tim Murray <timmurray@google.com> |
fix SDK build breakage with NEON
Change-Id: I823ab46a24a9db24e1b8a800284a5987747ae2ee
Conflicts:
cpu_ref/Android.mk
ndroid.mk
|
| a65de10aabdee0794d0e9c96db944e990166ef0d |
09-Aug-2013 |
Jason Sams <jsams@google.com> |
Float color matrix support.
Change-Id: I106b0cb1ad511aa549e4c90d79b3e84b55da45e4
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsics_neon_ColorMatrix.S
|
| 3870c6666e2e7d372daa6e0610f0f83695c954ec |
09-Aug-2013 |
Tim Murray <timmurray@google.com> |
fix SDK build breakage in mr2-dev and NEON
Change-Id: I823ab46a24a9db24e1b8a800284a5987747ae2ee
ndroid.mk
|
| 5cb36d9b36617f6b0493602ef61d620dc8f7e0ae |
09-Aug-2013 |
Jason Sams <jsams@google.com> |
Merge commit 'b10a68c3' into manualmerge
Conflicts:
cpu_ref/rsCpuIntrinsicColorMatrix.cpp
Change-Id: Ibc2f1514f8858d99f08380f698bc9ae533c69212
|
| e6b3d13534186d08f14d0753c26c535825ffdec8 |
08-Aug-2013 |
Stephen Hines <srhines@google.com> |
Move mmanWindows.* to proper locations.
Change-Id: If7044224de65735e257cf90d0c3bf655c31dd4d5
inkloader/include/mmanWindows.h
inkloader/lib/mmanWindows.cpp
|
| ddceab9a001f07a3395226c5e06e3b420720af0f |
07-Aug-2013 |
Jason Sams <jsams@google.com> |
Refactor hal to remove cpuConsumer from drivers
This CL should minimize build breaks due to
BufferQueue changes in the future.
Change-Id: I565a6eae5cc25603741fef32f2cfcb31a32eb757
sCpuScript.cpp
|
| f5ef8df639ba6363aa5d546e57ce872d04144cb6 |
06-Aug-2013 |
Jason Sams <jsams@google.com> |
Neon detection for RS SDK compat lib.
Change-Id: I3887158c7ec97ba116c28dc7b1d0c789b81fae60
sCpuCore.cpp
sCpuCore.h
sCpuIntrinsic3DLUT.cpp
sCpuIntrinsicBlend.cpp
sCpuIntrinsicBlur.cpp
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsicConvolve3x3.cpp
sCpuIntrinsicConvolve5x5.cpp
sCpuIntrinsicYuvToRGB.cpp
|
| 2644c42ddc16ef2515fc2e566ad2f60f1a1b47c6 |
03-Aug-2013 |
Raphael Moll <raphael@google.com> |
am 9b21cf51: am 1033b597: Fix Windows build issues for librsloader.
* commit '9b21cf5166b9ce3ac01c98d5122198dbde972b24':
Fix Windows build issues for librsloader.
|
| 6e8baf5350d41d87fdfb31b18facb4566307dedf |
31-Jul-2013 |
Jason Sams <jsams@google.com> |
Merge "Color Matrix improvements."
|
| 9b2b9efa1a1f9b0ec8c20601216f8dc5698c75f5 |
30-Jul-2013 |
Jason Sams <jsams@google.com> |
Color Matrix improvements.
Change-Id: I4594ea43a0a2b298a9ad66bd5e63d8b829d4f620
ndroid.mk
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsics_neon.S
sCpuIntrinsics_neon_ColorMatrix.S
|
| 35cfed80c9f8ae98af389bb934da21a5adfe8278 |
30-Jul-2013 |
Stephen Hines <srhines@google.com> |
Merge "Fix Script instancing issue for support library."
|
| c2c11cc9037d5ddd55282c6dab82db542b398d9e |
19-Jul-2013 |
Stephen Hines <srhines@google.com> |
Fix Script instancing issue for support library.
Our support library relies on opening shared library versions of Scripts
on older devices. This has the unfortunate side effect of sharing global
variables across Script instances. In traditional RenderScript, the actual
Scripts are always instanced, and thus have distinct storage/values for
every global (and static global).
This change fixes the instancing issue by keeping track of loaded libraries
(via a global set, since dlopen is tracked/shared per-process). If we detect
a collision, we create a new randomly named symlink of the library to actually
load from. The dlopen() function only looks at the library name passed to it,
so this allows us to cheaply create a duplicate instance. Once the symlinked
library is actually loaded, we destroy it.
A new unit test has also been added to RSTest_CompatLib to verify instance
behavior. This is a modified version of the same test we have present in CTS
for traditional RenderScript instance verification.
Change-Id: I6a0bca2b65d125b54793d98f0f73ae92e0280af7
sCpuScript.cpp
|
| 6ba6c664f773a53d7b6ddaf2c216d11404fd18fc |
13-Jul-2013 |
Stephen Hines <srhines@google.com> |
Move linkloader into cpu_ref/ instead of driver/.
This library is actually part of the reference implementation and not part of
the direct instantiation in driver/. Having this code live in driver/ (for
partners to modify/replace) is misleading.
Change-Id: Ifba4575ff3bcdd09a3839384cfb26c2b0bf9d5fa
inkloader/android/librsloader.cpp
inkloader/android/librsloader.h
inkloader/android/test-librsloader.c
inkloader/include/ELF.h
inkloader/include/ELFHeader.h
inkloader/include/ELFObject.h
inkloader/include/ELFReloc.h
inkloader/include/ELFSection.h
inkloader/include/ELFSectionBits.h
inkloader/include/ELFSectionHeader.h
inkloader/include/ELFSectionHeaderTable.h
inkloader/include/ELFSectionNoBits.h
inkloader/include/ELFSectionProgBits.h
inkloader/include/ELFSectionRelTable.h
inkloader/include/ELFSectionStrTab.h
inkloader/include/ELFSectionSymTab.h
inkloader/include/ELFSymbol.h
inkloader/include/ELFTypes.h
inkloader/include/GOT.h
inkloader/include/MemChunk.h
inkloader/include/StubLayout.h
inkloader/include/impl/ELFHeader.hxx
inkloader/include/impl/ELFObject.hxx
inkloader/include/impl/ELFReloc.hxx
inkloader/include/impl/ELFSection.hxx
inkloader/include/impl/ELFSectionBits.hxx
inkloader/include/impl/ELFSectionHeader.hxx
inkloader/include/impl/ELFSectionHeaderTable.hxx
inkloader/include/impl/ELFSectionNoBits.hxx
inkloader/include/impl/ELFSectionProgBits.hxx
inkloader/include/impl/ELFSectionRelTable.hxx
inkloader/include/impl/ELFSectionStrTab.hxx
inkloader/include/impl/ELFSectionSymTab.hxx
inkloader/include/impl/ELFSymbol.hxx
inkloader/lib/ELFHeader.cpp
inkloader/lib/ELFSectionHeader.cpp
inkloader/lib/ELFSymbol.cpp
inkloader/lib/ELFTypes.cpp
inkloader/lib/GOT.cpp
inkloader/lib/MemChunk.cpp
inkloader/lib/StubLayout.cpp
inkloader/main.cpp
inkloader/utils/flush_cpu_cache.h
inkloader/utils/helper.cpp
inkloader/utils/helper.h
inkloader/utils/raw_ostream.cpp
inkloader/utils/raw_ostream.h
inkloader/utils/rsl_assert.cpp
inkloader/utils/rsl_assert.h
inkloader/utils/serialize.h
inkloader/utils/traits.h
|
| 140a7acade66ab5d1f3dc55803a3a65a71f3f86c |
11-Jul-2013 |
Stephen Hines <srhines@google.com> |
resolved conflicts for merge of 5376c9bf to master
Change-Id: I51507da10f8d7116a2aa29446a00a43d397a37c8
|
| b0934b67b95cc27e2358c2aa4db5f7c1067c8f9b |
04-Jul-2013 |
Stephen Hines <srhines@google.com> |
Remove libutils and fix rsDebug for RS support library.
Bug: 9664050
Our bitcode runtime library translates vector rsDebug() calls into passing
their parameters via pointers. The previous version of libRSSupport.so was
being created with non-pointer versions of these routines accidentally.
This change also fixes a missing permission issue for ImageProcessing2, so
that the compatibility library can be verified.
This change also removes the use of libutils by switching the implementation of
String8/Vector in the compatibility library to internal types backed by
libstlport_static.
Change-Id: I20da75e8c19a82a42dc2bceaba1937d21372db84
sCpuCore.cpp
sCpuRuntimeStubs.cpp
sCpuScript.cpp
sCpuScriptGroup.cpp
|
| 34b0d3119567b992f0f876a2dffc0161bdcef3e6 |
27-Jun-2013 |
Jason Sams <jsams@google.com> |
Implement all formats for convolve 5x5.
Change-Id: I93456429e909beffa2b76bc3f7f46bd306c5941d
sCpuIntrinsicConvolve5x5.cpp
|
| 4768dade83376d58dbee788f2a705a63918ee8bc |
28-Jun-2013 |
Stephen Hines <srhines@google.com> |
Merge "Provide path to execute standalone bcc compiler."
|
| 3b35d775a777c36a178ce3fc97ff1e169aab3f1e |
26-Jun-2013 |
Jason Sams <jsams@google.com> |
Implement U1, U2, F1, F2, F4 convolve 3x3
Change-Id: Ib6ffd75ba19cf09a710f39fcd07400aae12d60ca
sCpuIntrinsicConvolve3x3.cpp
sCpuIntrinsicInlines.h
|
| b58d9adf52f0216c0281c749023bdd5b52236875 |
20-Jun-2013 |
Stephen Hines <srhines@google.com> |
Provide path to execute standalone bcc compiler.
Bug: 7342767
Change-Id: Ife9d9fee77416832445dcf30f0e3eba3dd173bad
ndroid.mk
sCpuScript.cpp
|
| 389ae9ac71245bc213b8148f2982464074a4b158 |
19-Jun-2013 |
Stephen Hines <srhines@google.com> |
Merge "Add the name of the forEach functions to the tracing output"
|
| b68ba7e6f3d1e6440c1409e85a6c3af14fc39ee6 |
19-Jun-2013 |
Jason Sams <jsams@google.com> |
Performance fix
Add support for processing 2 and 3 channel.
Change-Id: I124e1e92a65a296de1c3a839d9555f5eeeeb75c6
sCpuIntrinsicHistogram.cpp
|
| 47935ac8e3dd619f758abd576605a564e33bb59c |
17-Jun-2013 |
Tobias Grosser <grosser@google.com> |
Add the name of the forEach functions to the tracing output
Instead of showing up as 'runForEach', kernels are now e.g. printed as
'runForEach_createAntLookup'.
The String8 is carefully constructed here, such that we only pay
the string processing overhead in case tracing is actually enabled.
Change-Id: Id18319e271a02cbe888bcb9fe806794007d00fca
sCpuScript.cpp
|
| 2282e2816ac5f5de53f9bd4f3ecbdfd6d756d120 |
18-Jun-2013 |
Jason Sams <jsams@google.com> |
add histogram intrinsic
Change-Id: I42c297bfe116ea29cf015680fcc2143ff4cc95d2
ndroid.mk
sCpuCore.cpp
sCpuIntrinsic.cpp
sCpuIntrinsic.h
sCpuIntrinsicHistogram.cpp
|
| d89c234c749386a4dc6e362e719cb230853f18e4 |
13-Jun-2013 |
Tim Murray <timmurray@google.com> |
am 4accc1fe: am 606e5004: Fix Vine crash.
* commit '4accc1fe021321ecfe504897ed45121f11e6bd4b':
Fix Vine crash.
|
| 606e50048307530335b0885364ba888da436229b |
13-Jun-2013 |
Tim Murray <timmurray@google.com> |
Fix Vine crash.
Legacy YUV path was not computing the uv row coordinates correctly.
bug 9272491
Change-Id: I3af68fb8be191373205870573027ed38f059203e
sCpuIntrinsicYuvToRGB.cpp
|
| 8aff55716d056aaec7a896843abf80ba0f56e062 |
12-Jun-2013 |
Jason Sams <jsams@google.com> |
am 06bd91ec: Fix crash in Vine
* commit '06bd91ecb6ba59c5c19b2ad1325b30bd9344832d':
Fix crash in Vine
|
| 06bd91ecb6ba59c5c19b2ad1325b30bd9344832d |
12-Jun-2013 |
Jason Sams <jsams@google.com> |
Fix crash in Vine
bug 9272491
Legacy YUV formats could trigger null pointer using YUV intrinsic.
Change-Id: I2eca7254a22c757c50e302a190e9aa5c05437534
sCpuIntrinsicYuvToRGB.cpp
|
| ba17ae494add84056bbf3275b68e40e62a643db0 |
06-Jun-2013 |
Stephen Hines <srhines@google.com> |
Switch to separate build/load steps for Scripts.
Bug: 7342767
Change-Id: I94f81cb4dc917dd2cfc10c7718e9263454571db4
sCpuScript.cpp
|
| b7d9c80c98fc96aa7c638e3124be24f13a6436b2 |
30-Apr-2013 |
Stephen Hines <srhines@google.com> |
Provide a mechanism for adjusting RSCompilerDriver after construction.
We add a simple callback to the reference implementation of libRSDriver.so,
such that additional BCC flags can be toggled/adjusted before doing any actual
CPU compilation.
Change-Id: Iaf253b7d967d0382937369b1c5dae2d23a99e8be
sCpuCore.cpp
sCpuCore.h
sCpuScript.cpp
sd_cpu.h
|
| 70cc3f31ebe6dd61e173d9e0a120a2e4932b14c5 |
22-Apr-2013 |
Jason Sams <jsams@google.com> |
Merge "Fix YUV intrinsic" into jb-mr2-dev
|
| 6b58909686d7e439dac6dbb48c7e2236cd33c68d |
19-Apr-2013 |
Jason Sams <jsams@google.com> |
Fix YUV intrinsic
One format of YUV was not correctly handled by the built
in YUV intrinsic. This CL adds support. The scope is confined to
the YUV instrinsic and cannot impact other paths.
bug 8566866
Change-Id: I9a96b2117b0676213f4906b0ca12416693b9c0e4
sCpuIntrinsicYuvToRGB.cpp
sCpuIntrinsics_neon.S
|
| f47e8b4b86bf194e65398032f3f5f47a6da89f3f |
18-Apr-2013 |
Stephen Hines <srhines@google.com> |
Mark use of a debug context with the RSCompilerDriver.
Bug: 7343201
Change-Id: Idb003a3314247481c38aa70f9991f000135996b3
sCpuScript.cpp
|
| cca3d6ca444bef3b6d75431ec19bd07bfe40a733 |
15-Apr-2013 |
Stephen Hines <srhines@google.com> |
Add support for DEBUG context type.
Bug: 7343201
This change adds switching for the debug runtime (libclcore_debug.bc) when
running under a DEBUG RS context. It also fixes a small naming issue for
rsSetElementAt_int().
Change-Id: I8b574c9ce915a440375530325ba7de2b46707cfd
sCpuScript.cpp
|
| 886e51b2ef45ce123379c8a11579851c5580d957 |
12-Apr-2013 |
Tim Murray <timmurray@google.com> |
Merge "Add support for synchronous get()." into jb-mr2-dev
|
| 9c64239ebbfa4170190ede812e69150035e008e0 |
11-Apr-2013 |
Tim Murray <timmurray@google.com> |
Add support for synchronous get().
Change-Id: Ic94f1e36f7cffaaeda11fa9f0bb70e441fe1e535
sCpuScript.cpp
sCpuScript.h
sd_cpu.h
|
| a66d000ed939eeb014f65f3ff6a7da06961aaeb6 |
11-Apr-2013 |
Stephen Hines <srhines@google.com> |
Merge "Remove invalid logging on missing root.expand() functions." into jb-mr2-dev
|
| b2411816414d33e5c3dd1488be1702b3ef67b250 |
10-Apr-2013 |
Ying Wang <wangying@google.com> |
Add liblog
Bug: 8580410
Change-Id: Icc2c8db9304e1115b977b2cd1992803bf75618d7
ndroid.mk
|
| ef7481e2f0a4ad7b32bb626245e4207cabe171dc |
10-Apr-2013 |
Stephen Hines <srhines@google.com> |
Remove invalid logging on missing root.expand() functions.
Bug: 8566872
This log message is confusing and unncessary. Due to legacy compatibility,
slot 0 is always reserved for root.expand(). We know that root.expand() may
not exist for many user scripts, since we have supported proper names for
kernels for several releases now.
Change-Id: Id869496aa83f98ce64d6bd784314d788e25182e0
sCpuScript.cpp
|
| 941a61732cb044d0d9e4153dd3ef39d3055673e7 |
03-Apr-2013 |
Jason Sams <jsams@google.com> |
Cleanup debug and WAR vector type packing.
Change-Id: I4ff9619981dd1aa247b1bde511ec6a5bc1985954
sCpuRuntimeStubs.cpp
|
| 0782188b07ceeca03a45b26873bec0ccfd412373 |
02-Apr-2013 |
Jason Sams <jsams@google.com> |
Fix corruption when neon is disabled.
Change-Id: I5b7b143f78bb87c9f26265c45d727ed1b3063a73
sCpuIntrinsicBlur.cpp
|
| 3aeab4f0a628cad27c3f18ec1cff6f154ae01b62 |
02-Apr-2013 |
Jason Sams <jsams@google.com> |
Delete sampler optimization.
Calling overhead makes this mostly pointless.
Change-Id: I899778db3e6b06f46a5d5ca9af34fc17cf105b14
ndroid.mk
sCpuSample_neon.S
|
| 1d476620399d54774e4fd386c1d23cc583d49522 |
30-Mar-2013 |
Stephen Hines <srhines@google.com> |
Add callback to allow replacement of runtime support library.
Change-Id: I84ec56dfb29a0158015ebf31b3a73ac5bf34ef98
sCpuCore.cpp
sCpuCore.h
sCpuScript.cpp
sd_cpu.h
|
| b93cb42b4b8511d244ec8df4559fdc5c0b2de1bf |
28-Mar-2013 |
Stephen Hines <srhines@google.com> |
Add rsRand(float, float) support back to compat lib.
Change-Id: I0acaa409908f83c690c95b47979f820bc2dac9f5
sCpuRuntimeMath.cpp
sCpuRuntimeMathFuncs.cpp
|
| d6f1f46b2929ed56c73ba32357cde31b82972fbe |
26-Mar-2013 |
Tim Murray <timmurray@google.com> |
Add some missing compat lib functions and fix a bug.
Change-Id: I2950e1b4f785bcc1f0b6154ce822fe17832b3f4f
sCpuRuntimeMath.cpp
sCpuRuntimeMathFuncs.cpp
|
| a8cd40f1c75d0af0e2276c9f169eb0a1ba034adf |
26-Mar-2013 |
Tim Murray <timmurray@google.com> |
Merge "Start making RS 64-bit clean." into jb-mr2-dev
|
| 06b0f7d3b345ae1c36aed6686e79e403650f8baa |
22-Mar-2013 |
Jason Sams <jsams@google.com> |
start refactoring rsSample
Change-Id: I3bc2209991dcb2a13a2629e8ecfe6955ae83a9a8
ndroid.mk
sCpuSample_neon.S
|
| 099bc262f862cdeb547cf8a78fe9e0e92560f437 |
21-Mar-2013 |
Tim Murray <timmurray@google.com> |
Start making RS 64-bit clean.
Change-Id: Ie40ad9a1d2b59094c86eb7e40b358e60120ce213
sCpuIntrinsicBlur.cpp
|
| 962e720b3d1c27bcfec90374ff393584b99577b3 |
19-Mar-2013 |
Tim Murray <timmurray@google.com> |
Merge "Add x86 server support." into jb-mr2-dev
|
| 0b575de8ed0b628d84d256f5846500b0385979bd |
15-Mar-2013 |
Tim Murray <timmurray@google.com> |
Add x86 server support.
Change-Id: I674acaf15b67afa48bc736f72942a11e2e38e940
sCpuCore.cpp
sCpuIntrinsic3DLUT.cpp
sCpuIntrinsicBlend.cpp
sCpuIntrinsicConvolve5x5.cpp
sCpuIntrinsicYuvToRGB.cpp
sCpuRuntimeMath.cpp
sCpuRuntimeStubs.cpp
sCpuScript.cpp
sCpuScriptGroup.cpp
|
| 8ca358a2abe7e0dba23993e0fc8d64b8b55bd9ca |
19-Mar-2013 |
Jason Sams <jsams@google.com> |
Fix bug reporting CPU count.
Change-Id: Ib76a17c3239dc5b52624a567b40cace16f412327
sCpuCore.cpp
|
| 110f181b7966212a36ef18016f9b81c7322d0a2f |
15-Mar-2013 |
Jason Sams <jsams@google.com> |
Reduce diffs with compat lib.
Change-Id: Ibfa2c7c7da56d158f424ac343e364feacd0a21da
sCpuRuntimeMath.cpp
sCpuScript.cpp
sCpuScript.h
sd_cpu.h
|
| 029e82e1f688dfd132c19b1b94d2cad60d3ceaac |
12-Mar-2013 |
Jason Sams <jsams@google.com> |
Move simple functions to on device lib.
Change-Id: I690673f98244c490302cb25145cfcc8dd4125637
sCpuRuntimeMath.cpp
|
| cadfac411e6690e39de36c4f9e94deb9b7d2d08e |
07-Mar-2013 |
Jason Sams <jsams@google.com> |
Sync with compat lib.
Change-Id: Id8ace103814cf126f0d157100d1d4a12cc0b8664
sCpuCore.cpp
sCpuCore.h
sCpuScript.cpp
sCpuScript.h
sd_cpu.h
|
| d533c4c66510f2b83b9397607756479d79baae10 |
06-Mar-2013 |
Stephen Hines <srhines@google.com> |
This cl provides a proper WAR for thumb codegen.
Bug: 8326082
Change-Id: Ia471c2990461c1693e99e834624b856a5af4cac7
sCpuIntrinsic3DLUT.cpp
|
| 5edd18e4307e3c223b5db8a6cc5ca309a3a69c2a |
06-Mar-2013 |
Stephen Hines <srhines@google.com> |
WAR for non-neon ARM crash.
Change-Id: I3eb059874a84581d9bfbb22f6731f524ea1715c6
sCpuIntrinsic3DLUT.cpp
|
| 3a25fdd3786c1a08b783d8a83ef94b756347ff5c |
23-Feb-2013 |
Tim Murray <timmurray@google.com> |
Cleanup lots of things related to IO_OUTPUT and error checking.
Change-Id: Ic6802dd0ba9d3edc8c53f99002cdd905214a515c
sCpuScript.cpp
|
| b6d2d2a6bc1cee3e04b7e632a96134bdb56f38ee |
23-Feb-2013 |
Jason Sams <jsams@google.com> |
Fix corner cases with small blurs.
Change-Id: I7d1d42c3903ca3defd1737bd6e2557e2f5998fe5
sCpuIntrinsicBlur.cpp
|
| a31850150be87c472c71857c17ea5bc6676c1d8f |
19-Feb-2013 |
Stephen Hines <srhines@google.com> |
Merge "Support LinkRuntimeCallback() with RS compiler."
|
| bc0ca6ba4e31239bf77060578d0bdf1a10e04168 |
16-Feb-2013 |
Jason Sams <jsams@google.com> |
Support typed YUV allocations
Change-Id: I844051ee1be1462e497d238f5460f301be1aaa1d
sCpuIntrinsicYuvToRGB.cpp
|
| 385e9d7a982341252a337ae94ef199933e7583ec |
15-Feb-2013 |
Stephen Hines <srhines@google.com> |
Remove unused functions that insert __aeabi_idiv0() dependency.
We no longer use/call these functions, since we pick up modsi3() and the
variants directly from libcompiler_rt instead. By having them here, we are
accidentally pulling in additional dependencies on unnecessary functions.
Change-Id: Ia7aa29e443d77ec852987a13dc8fdbb4fa80af1d
sCpuRuntimeStubs.cpp
|
| f218bf115af4ae4fd79adbb8842608b308a4cf07 |
13-Feb-2013 |
Stephen Hines <srhines@google.com> |
Support LinkRuntimeCallback() with RS compiler.
Change-Id: I28ada4e7c462cb9673de6886d934dce855fac339
sCpuCore.cpp
sCpuCore.h
sCpuScript.cpp
sCpuScript.h
sd_cpu.h
|
| d4ecb17adc9b099351f2ca1779a74f5283f20a3d |
07-Feb-2013 |
Tim Murray <timmurray@google.com> |
Add API entry points for clipped kernels and fix Z-dimension clipping.
Change-Id: I43074cb8556d9b28d3e549930f5ac68387f8d823
sCpuScript.cpp
|
| 887922659e80b83b925631a3a459362d67db666e |
05-Feb-2013 |
Jason Sams <jsams@google.com> |
am 1f80b2c8: am 902868b7: Merge "YUV(NV21) to RGBA function NEON optimizations."
# By Vassilis Laganakos
# Via Android Git Automerger (1) and others
* commit '1f80b2c8b2b52c836c4039f716f702f96f213b0d':
YUV(NV21) to RGBA function NEON optimizations.
|
| ce0351debba8dadd1a7af2b3e926de6d787b49af |
26-Jan-2013 |
Jason Sams <jsams@google.com> |
Fix intrinsic bugs.
Change-Id: I027e5dcd8e538e52a21941facc5b93db2a6eac8c
sCpuIntrinsicBlur.cpp
sCpuIntrinsicConvolve5x5.cpp
sCpuIntrinsics_neon.S
|
| 2913f381a554c28abb44f49eddd1ee4c68a72578 |
15-Jan-2013 |
Stephen Hines <srhines@google.com> |
Align all allocations to a 16-byte boundary.
This change also fixes an issue in the Blur intrinsic, where we mis-cast a
float array to float4 (and thus encountered some new alignment errors with
the updated LLVM).
Change-Id: I3955b38f156c35f4d160652c75ab416bae09b2c8
sCpuIntrinsicBlur.cpp
|
| 5e3fb0b3cfadcb44a74cf4b6ec9ec65c11ba811e |
10-Jan-2013 |
Stephen Hines <srhines@google.com> |
Add parentheses for latest Clang warnings.
Change-Id: I425b7e13578d91a294397d37f4e66d9c1a6ff501
sCpuIntrinsic3DLUT.cpp
|
| 7c4b888f2147edf99690b6af75470774ff31c43b |
04-Jan-2013 |
Jason Sams <jsams@google.com> |
Functional 3D LUT intrinsic.
1600x1000 takes ~23ms on manta.
Change-Id: I142d6dedded66df05aa5f49e3da409a34c6e1b6e
ndroid.mk
sCpuCore.cpp
sCpuIntrinsic3DLUT.cpp
sCpuIntrinsicInlines.h
sCpuIntrinsics_neon.S
|
| 93eacc7ce0aad4314b4cb41a281f59ce54bb3286 |
18-Dec-2012 |
Jason Sams <jsams@google.com> |
Cleanup diff with compat.
Change-Id: Ieedcdec03ccb05cac68cbebbd398e07cb885f4d3
sCpuScriptGroup.cpp
sCpuScriptGroup.h
|
| 74121b55311b1a31ce41e6e2814ef5ad4ded1864 |
13-Dec-2012 |
Jason Sams <jsams@google.com> |
Remove unused functions.
Change-Id: I962ea4327c11d10b35e93bb71a1367a7ee41464f
sCpuRuntimeMath.cpp
|
| 241b999530999c7613edb742778c81fb4bf0e91e |
05-Dec-2012 |
Tim Murray <timmurray@google.com> |
Merge "enable synchronous mode (functional)"
|
| 4d252d6e807b89764dad123ac845df298c52ca97 |
29-Nov-2012 |
Tim Murray <timmurray@google.com> |
enable synchronous mode (functional)
Change-Id: I613610013e7e4d1623620ab94d2d25d8a1bd82b3
Bug: 5972398
sCpuCore.cpp
|
| c44d6706868749abe37780fc28b2cc627ddcf269 |
29-Nov-2012 |
Jason Sams <jsams@google.com> |
Fix uchar blur with neon offsets
fix hardcoded width limit in blur.
Change-Id: I4f56b8454181110133c4a637b66f1df3c79da33e
sCpuCore.h
sCpuIntrinsicBlur.cpp
sCpuIntrinsics_neon.S
|
| 7079cd8ddb0874d431446aea305b8c3b10e4f0ba |
28-Nov-2012 |
Jason Sams <jsams@google.com> |
Fix uchar blur performance.
Fix bug with misaligned access on odd width images.
Change-Id: Ie9f40121cf8eddbafdb2149814fcff28fff62036
sCpuIntrinsicBlur.cpp
sCpuIntrinsics_neon.S
|
| c905efd76fdcc1b8846b229bf7d991d185a7b4b7 |
27-Nov-2012 |
Jason Sams <jsams@google.com> |
Cleanup pass + implement blur uchar
Change-Id: Ib7f1c5218663b468a3c11daa2c3373ae132145ac
Conflicts:
cpu_ref/rsCpuIntrinsicBlend.cpp
sCpuCore.cpp
sCpuIntrinsic.cpp
sCpuIntrinsic.h
sCpuIntrinsicBlend.cpp
sCpuIntrinsicBlur.cpp
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsicConvolve3x3.cpp
sCpuIntrinsicConvolve5x5.cpp
sCpuIntrinsicLUT.cpp
sCpuIntrinsicYuvToRGB.cpp
sCpuIntrinsics_neon.S
|
| 4cca49b13db92b13ca07c1d330ad450d1b10f507 |
21-Nov-2012 |
Tim Murray <timmurray@google.com> |
Fix off-by-one error in convolve3x3.
Bug: 7599424
Change-Id: If668bd1ce2306b53efd20a09d16104cbf9fd4b77
sCpuIntrinsicConvolve3x3.cpp
|
| 709a0978ae141198018ca9769f8d96292a8928e6 |
16-Nov-2012 |
Jason Sams <jsams@google.com> |
Separate CPU driver impl from reference driver.
Change-Id: Ifb484edda665959b81d7b1f890d108bfa20a535d
ndroid.mk
sCpuCore.cpp
sCpuCore.h
sCpuIntrinsic.cpp
sCpuIntrinsic.h
sCpuIntrinsicBlend.cpp
sCpuIntrinsicBlur.cpp
sCpuIntrinsicColorMatrix.cpp
sCpuIntrinsicConvolve3x3.cpp
sCpuIntrinsicConvolve5x5.cpp
sCpuIntrinsicInlines.h
sCpuIntrinsicLUT.cpp
sCpuIntrinsicYuvToRGB.cpp
sCpuIntrinsics_neon.S
sCpuRuntimeMath.cpp
sCpuRuntimeStubs.cpp
sCpuScript.cpp
sCpuScript.h
sCpuScriptGroup.cpp
sCpuScriptGroup.h
sd_cpu.h
|