summaryrefslogtreecommitdiffstats
path: root/libclc
diff options
context:
space:
mode:
authorTom Stellard <thomas.stellard@amd.com>2013-06-26 18:22:11 +0000
committerTom Stellard <thomas.stellard@amd.com>2013-06-26 18:22:11 +0000
commit51441f80c55fa7719d11b4dff73d4e9132729234 (patch)
tree3844be2caaeacd41cd6cd16037355a2ddd8fa94e /libclc
parent66ecbc7c18f4f923b41ead5f2177ff5f64b5df57 (diff)
downloadbcm5719-llvm-51441f80c55fa7719d11b4dff73d4e9132729234.tar.gz
bcm5719-llvm-51441f80c55fa7719d11b4dff73d4e9132729234.zip
libclc: Initial vstore implementation
Assumes that the target supports byte-addressable stores. Completely unoptimized. Patch by: Aaron Watry llvm-svn: 185007
Diffstat (limited to 'libclc')
-rw-r--r--libclc/generic/include/clc/clc.h1
-rw-r--r--libclc/generic/include/clc/shared/vstore.h36
-rw-r--r--libclc/generic/lib/SOURCES1
-rw-r--r--libclc/generic/lib/shared/vstore.cl56
4 files changed, 94 insertions, 0 deletions
diff --git a/libclc/generic/include/clc/clc.h b/libclc/generic/include/clc/clc.h
index 7937003cb9f..10d30e032fa 100644
--- a/libclc/generic/include/clc/clc.h
+++ b/libclc/generic/include/clc/clc.h
@@ -72,6 +72,7 @@
#include <clc/shared/max.h>
#include <clc/shared/min.h>
#include <clc/shared/vload.h>
+#include <clc/shared/vstore.h>
/* 6.11.5 Geometric Functions */
#include <clc/geometric/cross.h>
diff --git a/libclc/generic/include/clc/shared/vstore.h b/libclc/generic/include/clc/shared/vstore.h
new file mode 100644
index 00000000000..1f784f82fec
--- /dev/null
+++ b/libclc/generic/include/clc/shared/vstore.h
@@ -0,0 +1,36 @@
+#define _CLC_VSTORE_DECL(PRIM_TYPE, VEC_TYPE, WIDTH, ADDR_SPACE) \
+ _CLC_OVERLOAD _CLC_DECL void vstore##WIDTH(VEC_TYPE vec, size_t offset, ADDR_SPACE PRIM_TYPE *out);
+
+#define _CLC_VECTOR_VSTORE_DECL(PRIM_TYPE, ADDR_SPACE) \
+ _CLC_VSTORE_DECL(PRIM_TYPE, PRIM_TYPE##2, 2, ADDR_SPACE) \
+ _CLC_VSTORE_DECL(PRIM_TYPE, PRIM_TYPE##3, 3, ADDR_SPACE) \
+ _CLC_VSTORE_DECL(PRIM_TYPE, PRIM_TYPE##4, 4, ADDR_SPACE) \
+ _CLC_VSTORE_DECL(PRIM_TYPE, PRIM_TYPE##8, 8, ADDR_SPACE) \
+ _CLC_VSTORE_DECL(PRIM_TYPE, PRIM_TYPE##16, 16, ADDR_SPACE)
+
+#define _CLC_VECTOR_VSTORE_PRIM1(PRIM_TYPE) \
+ _CLC_VECTOR_VSTORE_DECL(PRIM_TYPE, __private) \
+ _CLC_VECTOR_VSTORE_DECL(PRIM_TYPE, __local) \
+ _CLC_VECTOR_VSTORE_DECL(PRIM_TYPE, __global) \
+
+#define _CLC_VECTOR_VSTORE_PRIM() \
+ _CLC_VECTOR_VSTORE_PRIM1(char) \
+ _CLC_VECTOR_VSTORE_PRIM1(uchar) \
+ _CLC_VECTOR_VSTORE_PRIM1(short) \
+ _CLC_VECTOR_VSTORE_PRIM1(ushort) \
+ _CLC_VECTOR_VSTORE_PRIM1(int) \
+ _CLC_VECTOR_VSTORE_PRIM1(uint) \
+ _CLC_VECTOR_VSTORE_PRIM1(long) \
+ _CLC_VECTOR_VSTORE_PRIM1(ulong) \
+ _CLC_VECTOR_VSTORE_PRIM1(float) \
+
+#ifdef cl_khr_fp64
+#define _CLC_VECTOR_VSTORE() \
+ _CLC_VECTOR_VSTORE_PRIM1(double) \
+ _CLC_VECTOR_VSTORE_PRIM()
+#else
+#define _CLC_VECTOR_VSTORE() \
+ _CLC_VECTOR_VSTORE_PRIM()
+#endif
+
+_CLC_VECTOR_VSTORE()
diff --git a/libclc/generic/lib/SOURCES b/libclc/generic/lib/SOURCES
index 5d9e3faef63..50cc9bd522c 100644
--- a/libclc/generic/lib/SOURCES
+++ b/libclc/generic/lib/SOURCES
@@ -24,5 +24,6 @@ shared/clamp.cl
shared/max.cl
shared/min.cl
shared/vload.cl
+shared/vstore.cl
workitem/get_global_id.cl
workitem/get_global_size.cl
diff --git a/libclc/generic/lib/shared/vstore.cl b/libclc/generic/lib/shared/vstore.cl
new file mode 100644
index 00000000000..e88ccc53243
--- /dev/null
+++ b/libclc/generic/lib/shared/vstore.cl
@@ -0,0 +1,56 @@
+#include <clc/clc.h>
+
+#pragma OPENCL EXTENSION cl_khr_byte_addressable_store : enable
+
+#define VSTORE_VECTORIZE(PRIM_TYPE, ADDR_SPACE) \
+ _CLC_OVERLOAD _CLC_DEF void vstore2(PRIM_TYPE##2 vec, size_t offset, ADDR_SPACE PRIM_TYPE *mem) { \
+ mem[offset] = vec.s0; \
+ mem[offset+1] = vec.s1; \
+ } \
+\
+ _CLC_OVERLOAD _CLC_DEF void vstore3(PRIM_TYPE##3 vec, size_t offset, ADDR_SPACE PRIM_TYPE *mem) { \
+ mem[offset] = vec.s0; \
+ mem[offset+1] = vec.s1; \
+ mem[offset+2] = vec.s2; \
+ } \
+\
+ _CLC_OVERLOAD _CLC_DEF void vstore4(PRIM_TYPE##4 vec, size_t offset, ADDR_SPACE PRIM_TYPE *mem) { \
+ mem[offset] = vec.s0; \
+ mem[offset+1] = vec.s1; \
+ mem[offset+2] = vec.s2; \
+ mem[offset+3] = vec.s3; \
+ } \
+\
+ _CLC_OVERLOAD _CLC_DEF void vstore8(PRIM_TYPE##8 vec, size_t offset, ADDR_SPACE PRIM_TYPE *mem) { \
+ vstore4(vec.lo, offset, mem); \
+ vstore4(vec.hi, offset+4, mem); \
+ } \
+\
+ _CLC_OVERLOAD _CLC_DEF void vstore16(PRIM_TYPE##16 vec, size_t offset, ADDR_SPACE PRIM_TYPE *mem) { \
+ vstore8(vec.lo, offset, mem); \
+ vstore8(vec.hi, offset+8, mem); \
+ } \
+
+#define VSTORE_ADDR_SPACES(SCALAR_GENTYPE) \
+ VSTORE_VECTORIZE(SCALAR_GENTYPE, __private) \
+ VSTORE_VECTORIZE(SCALAR_GENTYPE, __local) \
+ VSTORE_VECTORIZE(SCALAR_GENTYPE, __global) \
+
+#define VSTORE_TYPES() \
+ VSTORE_ADDR_SPACES(char) \
+ VSTORE_ADDR_SPACES(uchar) \
+ VSTORE_ADDR_SPACES(short) \
+ VSTORE_ADDR_SPACES(ushort) \
+ VSTORE_ADDR_SPACES(int) \
+ VSTORE_ADDR_SPACES(uint) \
+ VSTORE_ADDR_SPACES(long) \
+ VSTORE_ADDR_SPACES(ulong) \
+ VSTORE_ADDR_SPACES(float) \
+
+VSTORE_TYPES()
+
+#ifdef cl_khr_fp64
+#pragma OPENCL EXTENSION cl_khr_fp64 : enable
+ VSTORE_ADDR_SPACES(double)
+#endif
+
OpenPOWER on IntegriCloud