Merge pull request #126 from X0rg/master

Fix for Zen 2

.gitignore

@@ -39,3 +39,4 @@ libcpuid/x64
 cpuid_tool/x32
 cpuid_tool/x64
 *.vcxproj.user
+tests/convert_instlatx64

libcpuid/recog_amd.c

@@ -348,9 +348,18 @@ static void load_amd_features(struct cpu_raw_data_t* raw, struct cpu_id_t* data)
 	if (raw->ext_cpuid[0][0] >= 0x80000007)
 		match_features(matchtable_edx87, COUNT_OF(matchtable_edx87), raw->ext_cpuid[7][3], data);
 	if (raw->ext_cpuid[0][0] >= 0x8000001a) {
-		/* We have the extended info about SSE unit size */
+		/* We have the extended info about SSE unit size
+		Extracted from BKDG, about CPUID_Fn8000001A_EAX [Performance Optimization Identifiers] (Core::X86::Cpuid::PerfOptId):
+		- bit 2: FP256
+		- bit 1: MOVU
+		- bit 0: FP128 */
 		data->detection_hints[CPU_HINT_SSE_SIZE_AUTH] = 1;
-		data->sse_size = (raw->ext_cpuid[0x1a][0] & 1) ? 128 : 64;
+		if ((raw->ext_cpuid[0x1a][0] >> 2) & 1)
+			data->sse_size = 256;
+		else if ((raw->ext_cpuid[0x1a][0]) & 1)
+			data->sse_size = 128;
+		else
+			data->sse_size = 64;
 	}
 }
 

tests/amd/zen2/ryzen5-matisse.test

@@ -73,6 +73,6 @@ ext_cpuid[31]=0001000f 0000016f 000001fd 00000001
 64
 64
 -1
-64 (authoritative)
+256 (authoritative)
 Ryzen 5 (Matisse)
 fpu vme de pse tsc msr pae mce cx8 apic mtrr sep pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht pni pclmul monitor ssse3 cx16 sse4_1 sse4_2 syscall movbe popcnt aes xsave osxsave avx mmxext nx fxsr_opt rdtscp lm lahf_lm cmp_legacy svm abm misalignsse sse4a 3dnowprefetch osvw ibs skinit wdt ts ttp tm_amd hwpstate constant_tsc fma3 f16c rdrand cpb aperfmperf avx2 bmi1 bmi2

tests/amd/zen2/ryzen7-matisse.test

@@ -0,0 +1,102 @@
+basic_cpuid[0]=00000010 68747541 444d4163 69746e65
+basic_cpuid[0]=00000010 68747541 444d4163 69746e65
+basic_cpuid[1]=00870f10 00100800 7ed8320b 178bfbff
+basic_cpuid[2]=00000000 00000000 00000000 00000000
+basic_cpuid[3]=00000000 00000000 00000000 00000000
+basic_cpuid[5]=00000040 00000040 00000003 00000011
+basic_cpuid[6]=00000004 00000000 00000001 00000000
+basic_cpuid[7]=00000000 219c91a9 00400004 00000000
+basic_cpuid[8]=00000000 00000000 00000000 00000000
+basic_cpuid[9]=00000000 00000000 00000000 00000000
+basic_cpuid[10]=00000000 00000000 00000000 00000000
+basic_cpuid[11]=00000001 00000002 00000100 00000000
+basic_cpuid[11]=00000001 00000002 00000100 00000000
+basic_cpuid[11]=00000007 00000010 00000201 00000000
+basic_cpuid[11]=00000007 00000010 00000201 00000000
+basic_cpuid[12]=00000000 00000000 00000000 00000000
+basic_cpuid[13]=00000207 00000340 00000380 00000000
+basic_cpuid[13]=00000207 00000340 00000380 00000000
+basic_cpuid[13]=0000000f 00000340 00000000 00000000
+basic_cpuid[13]=0000000f 00000340 00000000 00000000
+basic_cpuid[13]=00000100 00000240 00000000 00000000
+basic_cpuid[13]=00000100 00000240 00000000 00000000
+basic_cpuid[13]=00000040 00000340 00000000 00000000
+basic_cpuid[13]=00000040 00000340 00000000 00000000
+basic_cpuid[14]=00000000 00000000 00000000 00000000
+basic_cpuid[15]=00000000 000000ff 00000000 00000002
+basic_cpuid[15]=00000000 000000ff 00000000 00000002
+basic_cpuid[15]=00000000 00000040 000000ff 00000007
+basic_cpuid[15]=00000000 00000040 000000ff 00000007
+basic_cpuid[16]=00000000 00000002 00000000 00000000
+basic_cpuid[16]=00000000 00000002 00000000 00000000
+basic_cpuid[16]=0000000f 00000000 00000004 0000000f
+basic_cpuid[16]=0000000f 00000000 00000004 0000000f
+ext_cpuid[0]=80000020 68747541 444d4163 69746e65
+ext_cpuid[0]=80000020 68747541 444d4163 69746e65
+ext_cpuid[1]=00870f10 20000000 75c237ff 2fd3fbff
+ext_cpuid[2]=20444d41 657a7952 2037206e 30303733
+ext_cpuid[2]=20444d41 657a7952 2037206e 30303733
+ext_cpuid[3]=2d382058 65726f43 6f725020 73736563
+ext_cpuid[3]=2d382058 65726f43 6f725020 73736563
+ext_cpuid[4]=2020726f 20202020 20202020 00202020
+ext_cpuid[4]=2020726f 20202020 20202020 00202020
+ext_cpuid[5]=ff40ff40 ff40ff40 20080140 20080140
+ext_cpuid[5]=ff40ff40 ff40ff40 20080140 20080140
+ext_cpuid[6]=48006400 68006400 02006140 01009140
+ext_cpuid[6]=48006400 68006400 02006140 01009140
+ext_cpuid[7]=00000000 0000001b 00000000 00006799
+ext_cpuid[8]=00003030 010eb757 0000700f 00010000
+ext_cpuid[9]=00000000 00000000 00000000 00000000
+ext_cpuid[10]=00000001 00008000 00000000 0013bcff
+ext_cpuid[11]=00000000 00000000 00000000 00000000
+ext_cpuid[12]=00000000 00000000 00000000 00000000
+ext_cpuid[13]=00000000 00000000 00000000 00000000
+ext_cpuid[14]=00000000 00000000 00000000 00000000
+ext_cpuid[15]=00000000 00000000 00000000 00000000
+ext_cpuid[16]=00000000 00000000 00000000 00000000
+ext_cpuid[17]=00000000 00000000 00000000 00000000
+ext_cpuid[18]=00000000 00000000 00000000 00000000
+ext_cpuid[19]=00000000 00000000 00000000 00000000
+ext_cpuid[20]=00000000 00000000 00000000 00000000
+ext_cpuid[21]=00000000 00000000 00000000 00000000
+ext_cpuid[22]=00000000 00000000 00000000 00000000
+ext_cpuid[23]=00000000 00000000 00000000 00000000
+ext_cpuid[24]=00000000 00000000 00000000 00000000
+ext_cpuid[25]=f040f040 00000000 00000000 00000000
+ext_cpuid[26]=00000006 00000000 00000000 00000000
+ext_cpuid[27]=000003ff 00000000 00000000 00000000
+ext_cpuid[28]=00000000 00000000 00000000 00000000
+ext_cpuid[29]=00004121 01c0003f 0000003f 00000000
+ext_cpuid[29]=00004121 01c0003f 0000003f 00000000
+ext_cpuid[29]=00004122 01c0003f 0000003f 00000000
+ext_cpuid[29]=00004122 01c0003f 0000003f 00000000
+ext_cpuid[29]=00004143 01c0003f 000003ff 00000002
+ext_cpuid[29]=00004143 01c0003f 000003ff 00000002
+ext_cpuid[29]=0001c163 03c0003f 00003fff 00000001
+ext_cpuid[29]=0001c163 03c0003f 00003fff 00000001
+ext_cpuid[30]=00000000 00000100 00000000 00000000
+ext_cpuid[31]=0001000f 0000016f 000001fd 00000001
+--------------------------------------------------------------------------------
+15
+1
+0
+23
+113
+8
+16
+32
+32
+512
+32768
+-1
+8
+8
+0
+-1
+64
+64
+64
+-1
+256 (authoritative)
+Ryzen 7 (Matisse)
+fpu vme de pse tsc msr pae mce cx8 apic mtrr sep pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht pni pclmul monitor ssse3 cx16 sse4_1 sse4_2 syscall movbe popcnt aes xsave osxsave avx mmxext nx fxsr_opt rdtscp lm lahf_lm cmp_legacy svm abm misalignsse sse4a 3dnowprefetch osvw ibs skinit wdt ts ttp tm_amd hwpstate constant_tsc fma3 f16c rdrand cpb aperfmperf avx2 bmi1 bmi2

tests/amd/zen2/ryzen9-matisse.test

@@ -0,0 +1,102 @@
+basic_cpuid[0]=00000010 68747541 444d4163 69746e65
+basic_cpuid[0]=00000010 68747541 444d4163 69746e65
+basic_cpuid[1]=00870f10 00180800 7ed8320b 178bfbff
+basic_cpuid[2]=00000000 00000000 00000000 00000000
+basic_cpuid[3]=00000000 00000000 00000000 00000000
+basic_cpuid[5]=00000040 00000040 00000003 00000011
+basic_cpuid[6]=00000004 00000000 00000001 00000000
+basic_cpuid[7]=00000000 219c91a9 00400004 00000000
+basic_cpuid[8]=00000000 00000000 00000000 00000000
+basic_cpuid[9]=00000000 00000000 00000000 00000000
+basic_cpuid[10]=00000000 00000000 00000000 00000000
+basic_cpuid[11]=00000001 00000002 00000100 00000000
+basic_cpuid[11]=00000001 00000002 00000100 00000000
+basic_cpuid[11]=00000007 00000018 00000201 00000000
+basic_cpuid[11]=00000007 00000018 00000201 00000000
+basic_cpuid[12]=00000000 00000000 00000000 00000000
+basic_cpuid[13]=00000207 00000340 00000380 00000000
+basic_cpuid[13]=00000207 00000340 00000380 00000000
+basic_cpuid[13]=0000000f 00000340 00000000 00000000
+basic_cpuid[13]=0000000f 00000340 00000000 00000000
+basic_cpuid[13]=00000100 00000240 00000000 00000000
+basic_cpuid[13]=00000100 00000240 00000000 00000000
+basic_cpuid[13]=00000040 00000340 00000000 00000000
+basic_cpuid[13]=00000040 00000340 00000000 00000000
+basic_cpuid[14]=00000000 00000000 00000000 00000000
+basic_cpuid[15]=00000000 000000ff 00000000 00000002
+basic_cpuid[15]=00000000 000000ff 00000000 00000002
+basic_cpuid[15]=00000000 00000040 000000ff 00000007
+basic_cpuid[15]=00000000 00000040 000000ff 00000007
+basic_cpuid[16]=00000000 00000002 00000000 00000000
+basic_cpuid[16]=00000000 00000002 00000000 00000000
+basic_cpuid[16]=0000000f 00000000 00000004 0000000f
+basic_cpuid[16]=0000000f 00000000 00000004 0000000f
+ext_cpuid[0]=80000020 68747541 444d4163 69746e65
+ext_cpuid[0]=80000020 68747541 444d4163 69746e65
+ext_cpuid[1]=00870f10 20000000 75c237ff 2fd3fbff
+ext_cpuid[2]=20444d41 657a7952 2039206e 30303933
+ext_cpuid[2]=20444d41 657a7952 2039206e 30303933
+ext_cpuid[3]=32312058 726f432d 72502065 7365636f
+ext_cpuid[3]=32312058 726f432d 72502065 7365636f
+ext_cpuid[4]=20726f73 20202020 20202020 00202020
+ext_cpuid[4]=20726f73 20202020 20202020 00202020
+ext_cpuid[5]=ff40ff40 ff40ff40 20080140 20080140
+ext_cpuid[5]=ff40ff40 ff40ff40 20080140 20080140
+ext_cpuid[6]=48006400 68006400 02006140 02009140
+ext_cpuid[6]=48006400 68006400 02006140 02009140
+ext_cpuid[7]=00000000 0000001b 00000000 00006799
+ext_cpuid[8]=00003030 010eb757 00007017 00010000
+ext_cpuid[9]=00000000 00000000 00000000 00000000
+ext_cpuid[10]=00000001 00008000 00000000 0013bcff
+ext_cpuid[11]=00000000 00000000 00000000 00000000
+ext_cpuid[12]=00000000 00000000 00000000 00000000
+ext_cpuid[13]=00000000 00000000 00000000 00000000
+ext_cpuid[14]=00000000 00000000 00000000 00000000
+ext_cpuid[15]=00000000 00000000 00000000 00000000
+ext_cpuid[16]=00000000 00000000 00000000 00000000
+ext_cpuid[17]=00000000 00000000 00000000 00000000
+ext_cpuid[18]=00000000 00000000 00000000 00000000
+ext_cpuid[19]=00000000 00000000 00000000 00000000
+ext_cpuid[20]=00000000 00000000 00000000 00000000
+ext_cpuid[21]=00000000 00000000 00000000 00000000
+ext_cpuid[22]=00000000 00000000 00000000 00000000
+ext_cpuid[23]=00000000 00000000 00000000 00000000
+ext_cpuid[24]=00000000 00000000 00000000 00000000
+ext_cpuid[25]=f040f040 00000000 00000000 00000000
+ext_cpuid[26]=00000006 00000000 00000000 00000000
+ext_cpuid[27]=000003ff 00000000 00000000 00000000
+ext_cpuid[28]=00000000 00000000 00000000 00000000
+ext_cpuid[29]=00004121 01c0003f 0000003f 00000000
+ext_cpuid[29]=00004121 01c0003f 0000003f 00000000
+ext_cpuid[29]=00004122 01c0003f 0000003f 00000000
+ext_cpuid[29]=00004122 01c0003f 0000003f 00000000
+ext_cpuid[29]=00004143 01c0003f 000003ff 00000002
+ext_cpuid[29]=00004143 01c0003f 000003ff 00000002
+ext_cpuid[29]=00014163 03c0003f 00003fff 00000001
+ext_cpuid[29]=00014163 03c0003f 00003fff 00000001
+ext_cpuid[30]=00000000 00000100 00000000 00000000
+ext_cpuid[31]=0001000f 0000016f 000001fd 00000001
+--------------------------------------------------------------------------------
+15
+1
+0
+23
+113
+12
+24
+32
+32
+512
+65536
+-1
+8
+8
+0
+-1
+64
+64
+64
+-1
+256 (authoritative)
+Ryzen 9 (Matisse)
+fpu vme de pse tsc msr pae mce cx8 apic mtrr sep pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht pni pclmul monitor ssse3 cx16 sse4_1 sse4_2 syscall movbe popcnt aes xsave osxsave avx mmxext nx fxsr_opt rdtscp lm lahf_lm cmp_legacy svm abm misalignsse sse4a 3dnowprefetch osvw ibs skinit wdt ts ttp tm_amd hwpstate constant_tsc fma3 f16c rdrand cpb aperfmperf avx2 bmi1 bmi2