Revision bd62a3563de0004936f6506fdc5921210cb3bf14 authored by Zalman Stern on 08 June 2023, 01:18:19 UTC, committed by GitHub on 08 June 2023, 01:18:19 UTC
* Completely rework how RISC-V vector intrinsics are called to avoid issues iwth single element vectors being confused with scalars and other conversions that can happen via using call_intrin. Allows using any size vector. Only downside is splitting large vectors no longer happens, but RISC-V allows an LMUL of 8, meaning a vector of up to 8 times the vector register size will compile so this is much less of an issue. Splitting larger vectors can be added. Should also allow fractionaly LMUL in all cases, but this is not verified. * Significant refactor/rewrite of RISC V vector intrinsics support. Should handle many more cases and be well on the way to handling arbitary vector widths within the LMUL range. More tests added to simd_op_check_riscv . Likely well setup to move SVE2 to a similar approach, perhaps without the full genearilty on vector lengths. (I.e. they may need to be quantized to vscale, or offer better performance in that case.) * Formatting fixes. * More formatting. * Don't try to convert void types to match expected vector type. * Backout comment change that is no longer relevant. * Fix failure in camera_pipe app. (Code to make sure vector types match was being presented with a scalar only mismatch. Changed it to ignore scalar to scalar cases.) Address review feedback. * One more review comment. * Comment fix. --------- Co-authored-by: Steven Johnson <srj@google.com>
1 parent 67eaff3
Makefile
include ../support/Makefile.inc
CXX ?= c++
TOP := $(abspath $(dir $(lastword $(MAKEFILE_LIST)))/../..)
.PHONY: all $(TOP)
all: run run-two
HALIDE_LIB := $(TOP)/$(LIBHALIDE_LDFLAGS)
$(HALIDE_LIB): $(TOP)
$(MAKE) -C $(TOP)
test_%: test_%.cpp
$(CXX) -std=c++17 -I ../../include/ $< -L ../../bin/ -lHalide $(HALIDE_SYSTEM_LIBS) -o $@ -g
avg_filter_uint32t.o avg_filter_uint32t.h avg_filter_float.o avg_filter_float.h: test_oglc_avg
LD_LIBRARY_PATH=../../bin DYLD_LIBRARY_PATH=../../bin HL_TARGET=arm-32-android-armv7s-openglcompute ./$<
avg_filter_uint32t_arm.o avg_filter_uint32t_arm.h avg_filter_float_arm.o avg_filter_float_arm.h: test_oglc_avg
LD_LIBRARY_PATH=../../bin DYLD_LIBRARY_PATH=../../bin HL_TARGET=arm-32-android-armv7s ./$< "_arm"
AVG_FILTER_SRC = jni/oglc_run.cpp \
avg_filter_uint32t.o avg_filter_uint32t.h \
avg_filter_uint32t_arm.o avg_filter_uint32t_arm.h \
avg_filter_float.o avg_filter_float.h \
avg_filter_float_arm.o avg_filter_float_arm.h
libs/armeabi-v7a/oglc_run: $(HALIDE_LIB) $(AVG_FILTER_SRC)
ndk-build libs/armeabi-v7a/oglc_run
two_kernels_filter.o two_kernels_filter.h: test_two_kernels
LD_LIBRARY_PATH=../../bin DYLD_LIBRARY_PATH=../../bin HL_TARGET=arm-32-android-armv7s-openglcompute ./$<
TWO_KERNELS_SRC = jni/oglc_two_kernels_run.cpp \
two_kernels_filter.o two_kernels_filter.h
libs/armeabi-v7a/oglc_two_kernels_run: $(HALIDE_LIB) $(TWO_KERNELS_SRC)
ndk-build libs/armeabi-v7a/oglc_two_kernels_run libs/armeabi-v7a/liboglc_two_kernels.so
jni-libs: $(HALIDE_LIB) $(AVG_FILTER_SRC) $(TWO_KERNELS_SRC)
ndk-build libs/armeabi-v7a/liboglc_two_kernels.so libs/armeabi-v7a/liboglc.so
deploy: libs/armeabi-v7a/oglc_run
adb push libs/armeabi-v7a/oglc_run /mnt/sdcard/
define RUN_STEPS
su
mkdir -p /data/tmp
rm -rf /data/tmp/oglc
mkdir /data/tmp/oglc
cd /data/tmp/oglc
pwd
cp /mnt/sdcard/oglc_run .
chmod 777 /data/tmp/oglc/oglc_run
LD_LIBRARY_PATH=. ./oglc_run
exit
exit
endef
export RUN_STEPS
run: deploy
adb logcat -c
sh -c 'echo "$$RUN_STEPS" | adb shell'
adb logcat -d | grep "I oglc"
echo "Done"
deploy-two: libs/armeabi-v7a/oglc_two_kernels_run
adb push libs/armeabi-v7a/oglc_two_kernels_run /mnt/sdcard/
define RUN_TWO_STEPS
su
mkdir /data/tmp
cd /data/tmp
pwd
cp /mnt/sdcard/oglc_two_kernels_run .
chmod 777 /data/tmp/oglc_two_kernels_run
LD_LIBRARY_PATH=. ./oglc_two_kernels_run
exit
exit
endef
export RUN_TWO_STEPS
run-two: deploy-two
adb logcat -c
sh -c 'echo "$$RUN_TWO_STEPS" | adb shell'
adb logcat -d | grep "I oglc"
echo "Done"
clean:
rm -f test_oglc_avg
rm -rf test_oglc_avg.dSYM/
rm -f avg_filter*
rm -f test_two_kernels
rm -rf test_two_kernels.dSYM/
rm -rf libs/
rm -rf obj/
rm -rf bin/
rm -rf gen/
Computing file changes ...