[X86] AMD Zen 3: same-reg SSE XMM ANDNPD is a 1-cycle(!) dep-breaking zero-idiom

As confirmed by exegesis measurements, and ref docs.

diff --git a/llvm/lib/Target/X86/X86ScheduleZnver3.td b/llvm/lib/Target/X86/X86ScheduleZnver3.td
index 770cdc2..1b2bdfb 100644 (file)
--- a/llvm/lib/Target/X86/X86ScheduleZnver3.td
+++ b/llvm/lib/Target/X86/X86ScheduleZnver3.td
@@ -1554,7 +1554,7 @@ def : IsZeroIdiomFunction<[
 
   // SSE XMM Zero-idioms.
   DepBreakingClass<[ XORPSrr, XORPDrr,
-                     ANDNPSrr ], ZeroIdiomPredicate>,
+                     ANDNPSrr, ANDNPDrr ], ZeroIdiomPredicate>,
 
   // AVX XMM Zero-idioms.
   DepBreakingClass<[ VXORPSrr, VXORPDrr,

diff --git a/llvm/test/tools/llvm-mca/X86/Znver3/zero-idioms-sse-xmm.s b/llvm/test/tools/llvm-mca/X86/Znver3/zero-idioms-sse-xmm.s
index cb0307b..04b81bf 100644 (file)
--- a/llvm/test/tools/llvm-mca/X86/Znver3/zero-idioms-sse-xmm.s
+++ b/llvm/test/tools/llvm-mca/X86/Znver3/zero-idioms-sse-xmm.s
@@ -298,12 +298,12 @@ andnpd %xmm0, %xmm1
 
 # CHECK:      Iterations:        10000
 # CHECK-NEXT: Instructions:      20000
-# CHECK-NEXT: Total Cycles:      20003
+# CHECK-NEXT: Total Cycles:      5004
 # CHECK-NEXT: Total uOps:        20000
 
 # CHECK:      Dispatch Width:    6
-# CHECK-NEXT: uOps Per Cycle:    1.00
-# CHECK-NEXT: IPC:               1.00
+# CHECK-NEXT: uOps Per Cycle:    4.00
+# CHECK-NEXT: IPC:               4.00
 # CHECK-NEXT: Block RThroughput: 0.5
 
 # CHECK:      Instruction Info:
@@ -319,13 +319,13 @@ andnpd %xmm0, %xmm1
 # CHECK-NEXT:  1      1     0.25                        andnpd %xmm0, %xmm1
 
 # CHECK:      Register File statistics:
-# CHECK-NEXT: Total number of mappings created:    20000
-# CHECK-NEXT: Max number of mappings used:         66
+# CHECK-NEXT: Total number of mappings created:    10000
+# CHECK-NEXT: Max number of mappings used:         37
 
 # CHECK:      *  Register File #1 -- Zn3FpPRF:
 # CHECK-NEXT:    Number of physical registers:     160
-# CHECK-NEXT:    Total number of mappings created: 20000
-# CHECK-NEXT:    Max number of mappings used:      66
+# CHECK-NEXT:    Total number of mappings created: 10000
+# CHECK-NEXT:    Max number of mappings used:      37
 
 # CHECK:      *  Register File #2 -- Zn3IntegerPRF:
 # CHECK-NEXT:    Number of physical registers:     192
@@ -363,16 +363,16 @@ andnpd %xmm0, %xmm1
 
 # CHECK:      Resource pressure by instruction:
 # CHECK-NEXT: [0]    [1]    [2]    [3]    [4]    [5]    [6]    [7]    [8]    [9]    [10]   [11]   [12.0] [12.1] [13]   [14.0] [14.1] [14.2] [15.0] [15.1] [15.2] [16.0] [16.1] Instructions:
-# CHECK-NEXT:  -      -      -      -      -      -      -      -      -     0.50    -     0.50    -      -      -      -      -      -      -      -      -      -      -     andnpd  %xmm1, %xmm1
-# CHECK-NEXT:  -      -      -      -      -      -      -      -     0.50    -     0.50    -      -      -      -      -      -      -      -      -      -      -      -     andnpd  %xmm0, %xmm1
+# CHECK-NEXT:  -      -      -      -      -      -      -      -      -     0.50   0.25   0.25    -      -      -      -      -      -      -      -      -      -      -     andnpd  %xmm1, %xmm1
+# CHECK-NEXT:  -      -      -      -      -      -      -      -     0.50    -     0.25   0.25    -      -      -      -      -      -      -      -      -      -      -     andnpd  %xmm0, %xmm1
 
 # CHECK:      Timeline view:
-# CHECK-NEXT: Index     0123456
+# CHECK-NEXT: Index     01234
 
-# CHECK:      [0,0]     DeER ..   andnpd       %xmm1, %xmm1
-# CHECK-NEXT: [0,1]     D=eER..   andnpd       %xmm0, %xmm1
-# CHECK-NEXT: [1,0]     D==eER.   andnpd       %xmm1, %xmm1
-# CHECK-NEXT: [1,1]     D===eER   andnpd       %xmm0, %xmm1
+# CHECK:      [0,0]     DeER.   andnpd %xmm1, %xmm1
+# CHECK-NEXT: [0,1]     D=eER   andnpd %xmm0, %xmm1
+# CHECK-NEXT: [1,0]     DeE-R   andnpd %xmm1, %xmm1
+# CHECK-NEXT: [1,1]     D=eER   andnpd %xmm0, %xmm1
 
 # CHECK:      Average Wait times (based on the timeline view):
 # CHECK-NEXT: [0]: Executions
@@ -381,6 +381,6 @@ andnpd %xmm0, %xmm1
 # CHECK-NEXT: [3]: Average time elapsed from WB until retire stage
 
 # CHECK:            [0]    [1]    [2]    [3]
-# CHECK-NEXT: 0.     2     2.0    0.5    0.0       andnpd      %xmm1, %xmm1
-# CHECK-NEXT: 1.     2     3.0    0.0    0.0       andnpd      %xmm0, %xmm1
-# CHECK-NEXT:        2     2.5    0.3    0.0       <total>
+# CHECK-NEXT: 0.     2     1.0    1.0    0.5       andnpd      %xmm1, %xmm1
+# CHECK-NEXT: 1.     2     2.0    0.0    0.0       andnpd      %xmm0, %xmm1
+# CHECK-NEXT:        2     1.5    0.5    0.3       <total>