Test added: cpu/spu_fpu

.gitignore

@@ -18,4 +18,5 @@
 
 # Uninteresting Source files
 tests/rsx/rsx_fp_dynamic/*.cpp
-tests/rsx/rsx_fp_static/*.cpp
+tests/rsx/rsx_fp_static/*.cpp
+tests/cpu/spu*/*.ppu.cpp

tests/cpu/spu_fpu/spu_fpu.spu.cpp

@@ -0,0 +1,154 @@
+#include <cell.h>
+#include <spu_printf.h>
+#include <spu_internals.h>
+#include <math.h>
+#include <float.h>
+
+
+// Loops with floats
+#define ITERATE1f(name, instruction) \
+	for (unsigned int i=0; i<sizeof(testFloats)/sizeof(float); i++) { \
+		volatile vec_float4 v1 = (vec_float4)(testFloats[i]); \
+		volatile vec_float4 r  = (vec_float4)instruction((qword)v1); \
+		spu_printf(name"([%02d]) -> %08x %08x %08x %08x\n",i,r[0],r[1],r[2],r[3]); \
+	}
+
+#define ITERATE2f(name, instruction) \
+	for (unsigned int i=0; i<sizeof(testFloats)/sizeof(float); i++) \
+		for (unsigned int j=0; j<sizeof(testFloats)/sizeof(float); j++) { \
+			volatile vec_float4 v1 = (vec_float4)(testFloats[i]); \
+			volatile vec_float4 v2 = (vec_float4)(testFloats[j],testFloats[j],0.0,0.0); \
+			volatile vec_float4 r  = (vec_float4)instruction((qword)v1,(qword)v2); \
+			spu_printf(name"([%02d],[%02d]) -> %08x %08x %08x %08x\n",i,j,r[0],r[1],r[2],r[3]); \
+		}
+
+#define ITERATE3f(name, instruction) \
+	for (unsigned int i=0; i<sizeof(testFloats)/sizeof(float); i++) \
+		for (unsigned int j=0; j<sizeof(testFloats)/sizeof(float); j++) \
+			for (unsigned int k=0; k<sizeof(testFloats)/sizeof(float); k++) { \
+				volatile vec_float4 v1 = (vec_float4)(testFloats[i]); \
+				volatile vec_float4 v2 = (vec_float4)(testFloats[j],testFloats[j],0.0,0.0); \
+				volatile vec_float4 v3 = (vec_float4)(testFloats[k],0.0,testFloats[k],0.0); \
+				volatile vec_float4 r  = (vec_float4)instruction((qword)v1,(qword)v2,(qword)v3); \
+				spu_printf(name"([%02d],[%02d],[%02d]) -> %08x %08x %08x %08x\n",i,j,k,r[0],r[1],r[2],r[3]); \
+			}
+
+// Loops with doubles
+#define ITERATE1d(name, instruction) \
+	for (unsigned int i=0; i<sizeof(testDoubles)/sizeof(double); i++) { \
+		volatile vec_double2 v1 = (vec_double2)(testDoubles[i]); \
+		volatile vec_double2 r  = (vec_double2)instruction((qword)v1); \
+		spu_printf(name"([%02d]) -> %016llx %016llx\n",i,r[0],r[1]); \
+	}
+
+#define ITERATE2d(name, instruction) \
+	for (unsigned int i=0; i<sizeof(testDoubles)/sizeof(double); i++) \
+		for (unsigned int j=0; j<sizeof(testDoubles)/sizeof(double); j++) { \
+			volatile vec_double2 v1 = (vec_double2)(testDoubles[i]); \
+			volatile vec_double2 v2 = (vec_double2)(testDoubles[j],0.0); \
+			volatile vec_double2 r  = (vec_double2)instruction((qword)v1,(qword)v2); \
+			spu_printf(name"([%02d],[%02d]) -> %016llx %016llx\n",i,j,r[0],r[1]); \
+		}
+
+#define ITERATE3d(name, instruction) \
+	for (unsigned int i=0; i<sizeof(testDoubles)/sizeof(double); i++) \
+		for (unsigned int j=0; j<sizeof(testDoubles)/sizeof(double); j++) \
+			for (unsigned int k=0; k<sizeof(testDoubles)/sizeof(double); k++) { \
+				volatile vec_double2 v1 = (vec_double2)(testDoubles[i]); \
+				volatile vec_double2 v2 = (vec_double2)(testDoubles[j]); \
+				volatile vec_double2 v3 = (vec_double2)(testDoubles[k],0.0); \
+				volatile vec_double2 r  = (vec_double2)instruction((qword)v1,(qword)v2,(qword)v3); \
+				spu_printf(name"([%02d],[%02d],[%02d]) -> %016llx %016llx\n",i,j,k,r[0],r[1]); \
+			}
+
+// Conversion loops
+#define ITERATE1iimm(name, instruction, imm) \
+	for (unsigned int i=0; i<sizeof(testInts)/sizeof(int); i++) { \
+		volatile vec_int4 v1 = (vec_int4)(testInts[i]); \
+		volatile vec_float4 r  = (vec_float4)instruction((qword)v1, imm); \
+		spu_printf(name"([%02d]) -> %08x %08x %08x %08x\n",i,r[0],r[1],r[2],r[3]); \
+	}
+
+#define ITERATE1fimm(name, instruction, imm) \
+	for (unsigned int i=0; i<sizeof(testFloats)/sizeof(float); i++) { \
+		volatile vec_float4 v1 = (vec_float4)(testFloats[i]); \
+		volatile vec_int4 r  = (vec_int4)instruction((qword)v1, imm); \
+		spu_printf(name"([%02d]) -> %08x %08x %08x %08x\n",i,r[0],r[1],r[2],r[3]); \
+	}
+
+
+// Test values
+int testInts[] = {
+	0, 1, 2, 3, 4, -1, -2, -3, -4,
+	0x00033333, 0x00fffff0, 0x10000000, 0x12345678,
+	0x1fffffff, 0x42385722, 0x70000000, 0x72233411,
+	0x7fffffff, 0x80000000, 0x80000001, 0x8fffffff,
+};
+
+float testFloats[] = {
+	0.0, -0.0,  1.0, -1.0,  1.5, -1.5,
+	1.6, -1.6,  1.4, -1.4,  2.0, -2.0,
+	4.0, -10000000.4, 20000000.0, -20000.5, 20000.6,
+	FLT_MIN, FLT_MAX, INFINITY, -INFINITY, NAN,
+};
+
+double testDoubles[] = {
+	0.0, -0.0,  1.0, -1.0,  1.5, -1.5,
+	1.6, -1.6,  1.4, -1.4,  2.0, -2.0,
+	4.0, -10000000.4, 20000000.0, -20000.5, 20000.6,
+	DBL_MIN, DBL_MAX, INFINITY, -INFINITY, NAN,
+};
+
+
+int main()
+{
+	/////////////////////////////////
+	// FLOATING-POINT INSTRUCTIONS //
+	/////////////////////////////////
+
+	// Single Precision instructions
+	ITERATE2f("fa      ", si_fa);
+	ITERATE2f("fs      ", si_fs);
+	ITERATE2f("fm      ", si_fm);
+	ITERATE3f("fma     ", si_fma);
+	ITERATE3f("fnms    ", si_fnms);
+	ITERATE3f("fms     ", si_fms);
+	ITERATE1f("frest   ", si_frest);
+	ITERATE1f("frsqest ", si_frsqest);
+	ITERATE2f("fi      ", si_fi);
+	ITERATE1f("fesd    ", si_fesd);
+	ITERATE2f("fceq    ", si_fceq);
+	ITERATE2f("fcmeq   ", si_fcmeq);
+	ITERATE2f("fcgt    ", si_fcgt);
+	ITERATE2f("fcmgt   ", si_fcmgt);
+
+	// Double Precision instructions (without eDP instructions: dfceq, dfcmeq, dfcgt, dfcmgt, dftsv)
+	ITERATE2d("dfa     ", si_dfa);
+	ITERATE2d("dfs     ", si_dfs);
+	ITERATE2d("dfm     ", si_dfm);
+	ITERATE3d("dfma    ", si_dfma);
+	ITERATE3d("dfnms   ", si_dfnms);
+	ITERATE3d("dfms    ", si_dfms);
+	ITERATE3d("dfnma   ", si_dfnma);
+	ITERATE1d("frds    ", si_frds);
+
+	// Conversion instructions
+	ITERATE1iimm("csflt   ", si_csflt, 0x00);
+	ITERATE1iimm("csflt   ", si_csflt, 0x01);
+	ITERATE1iimm("csflt   ", si_csflt, 0x04);
+	ITERATE1iimm("csflt   ", si_csflt, 0x7f);
+	ITERATE1fimm("cflts   ", si_cflts, 0x00);
+	ITERATE1fimm("cflts   ", si_cflts, 0x01);
+	ITERATE1fimm("cflts   ", si_cflts, 0x04);
+	ITERATE1fimm("cflts   ", si_cflts, 0x7f);
+	ITERATE1iimm("cuflt   ", si_cuflt, 0x00);
+	ITERATE1iimm("cuflt   ", si_cuflt, 0x01);
+	ITERATE1iimm("cuflt   ", si_cuflt, 0x04);
+	ITERATE1iimm("cuflt   ", si_cuflt, 0x7f);
+	ITERATE1fimm("cfltu   ", si_cfltu, 0x00);
+	ITERATE1fimm("cfltu   ", si_cfltu, 0x01);
+	ITERATE1fimm("cfltu   ", si_cfltu, 0x04);
+	ITERATE1fimm("cfltu   ", si_cfltu, 0x7f);
+
+	// TODO: fscrwr, fscrrd
+}