BSF (R64, R64) - Port Usage

With blocking instructions for port '1':

Code:

   0:	66 0f 58 c1          	addpd  xmm0,xmm1
   4:	66 0f 58 d1          	addpd  xmm2,xmm1
   8:	66 0f 58 d9          	addpd  xmm3,xmm1
   c:	66 0f 58 e1          	addpd  xmm4,xmm1
  10:	66 0f 58 e9          	addpd  xmm5,xmm1
  14:	66 0f 58 f1          	addpd  xmm6,xmm1
  18:	66 0f 58 f9          	addpd  xmm7,xmm1
  1c:	66 44 0f 58 c1       	addpd  xmm8,xmm1
  21:	66 44 0f 58 c9       	addpd  xmm9,xmm1
  26:	66 44 0f 58 d1       	addpd  xmm10,xmm1
  2b:	4d 0f bc c1          	bsf    r8,r9

Init:

MOV RAX, 0x4000000040000000;
MOV [R14+0], RAX;
MOV [R14+8], RAX;
MOVUPD XMM0, [R14];
MOVUPD XMM1, [R14];
MOVUPD XMM2, [R14];
MOVUPD XMM3, [R14];
MOVUPD XMM4, [R14];
MOVUPD XMM5, [R14];
MOVUPD XMM6, [R14];
MOVUPD XMM7, [R14];
MOVUPD XMM8, [R14];
MOVUPD XMM9, [R14];
MOVUPD XMM10, [R14];
MOVUPD XMM11, [R14];
MOVUPD XMM12, [R14]

Show nanoBench command
Results:
- Instructions retired: 11.0
- Core cycles: 11.54
- Reference cycles: 11.52
- UOPS_PORT_1: 10.98

⇨ One μop that can only use port '1'

With blocking instructions for ports {'0', '5'}:

Code:

   0:	41 0f 6e c2          	movd   mm0,r10d
   4:	41 0f 6e ca          	movd   mm1,r10d
   8:	41 0f 6e d2          	movd   mm2,r10d
   c:	41 0f 6e da          	movd   mm3,r10d
  10:	41 0f 6e e2          	movd   mm4,r10d
  14:	41 0f 6e ea          	movd   mm5,r10d
  18:	41 0f 6e f2          	movd   mm6,r10d
  1c:	41 0f 6e fa          	movd   mm7,r10d
  20:	41 0f 6e c2          	movd   mm0,r10d
  24:	41 0f 6e ca          	movd   mm1,r10d
  28:	4d 0f bc c1          	bsf    r8,r9

Show nanoBench command
Results:
- Instructions retired: 11.0
- Core cycles: 5.01
- Reference cycles: 5.0
- UOPS_PORT_0: 5.02
- UOPS_PORT_5: 4.96

With blocking instructions for ports {'0', '1', '5'}:

Code:

   0:	66 0f 55 c1          	andnpd xmm0,xmm1
   4:	66 0f 55 d1          	andnpd xmm2,xmm1
   8:	66 0f 55 d9          	andnpd xmm3,xmm1
   c:	66 0f 55 e1          	andnpd xmm4,xmm1
  10:	66 0f 55 e9          	andnpd xmm5,xmm1
  14:	66 0f 55 f1          	andnpd xmm6,xmm1
  18:	66 0f 55 f9          	andnpd xmm7,xmm1
  1c:	66 44 0f 55 c1       	andnpd xmm8,xmm1
  21:	66 44 0f 55 c9       	andnpd xmm9,xmm1
  26:	66 44 0f 55 d1       	andnpd xmm10,xmm1
  2b:	4d 0f bc c1          	bsf    r8,r9

Init:

PXOR XMM0, XMM0;
PXOR XMM1, XMM1;
PXOR XMM2, XMM2;
PXOR XMM3, XMM3;
PXOR XMM4, XMM4;
PXOR XMM5, XMM5;
PXOR XMM6, XMM6;
PXOR XMM7, XMM7;
PXOR XMM8, XMM8;
PXOR XMM9, XMM9;
PXOR XMM10, XMM10;
PXOR XMM11, XMM11;
PXOR XMM12, XMM12

Show nanoBench command
Results:
- Instructions retired: 11.0
- Core cycles: 10.99
- Reference cycles: 11.01
- UOPS_PORT_0: 4.0
- UOPS_PORT_1: 5.0
- UOPS_PORT_5: 3.0

⇨ One μop that can only use ports {'0', '1', '5'}