blendTextureLinearR8 inner loop

gistfile1.txt

+0x210	    movaps              -288(%rbp), %xmm0
+0x217	    maxps               -464(%rbp), %xmm0
+0x21e	    minps               -448(%rbp), %xmm0
+0x225	    movaps              -304(%rbp), %xmm2
+0x22c	    maxps               -496(%rbp), %xmm2
+0x233	    minps               -480(%rbp), %xmm2
+0x23a	    cvttps2dq           %xmm0, %xmm1
+0x23e	    cvttps2dq           %xmm2, %xmm0
+0x242	    movdqa              %xmm1, %xmm2
+0x246	    pcmpeqd             %xmm12, %xmm12
+0x24b	    pcmpgtd             %xmm12, %xmm2
+0x250	    pand                %xmm1, %xmm2
+0x254	    movdqa              %xmm1, %xmm3
+0x258	    psrad               $7, %xmm3
+0x25d	    movdqa              %xmm0, %xmm1
+0x261	    psrad               $7, %xmm1
+0x266	    movdqa              %xmm3, %xmm4
+0x26a	    pxor                %xmm14, %xmm14
+0x26f	    pmaxsw              %xmm14, %xmm4
+0x274	    addl                $-2, %r11d
+0x278	    movd                %r11d, %xmm5
+0x27d	    pshufd              $0, %xmm5, %xmm8
+0x283	    pminsw              %xmm8, %xmm4
+0x288	    movdqa              %xmm1, %xmm6
+0x28c	    pmaxsw              %xmm14, %xmm6
+0x291	    addl                $-1, %r12d
+0x295	    movd                %r12d, %xmm7
+0x29a	    pshufd              $0, %xmm7, %xmm7
+0x29f	    pminsw              %xmm7, %xmm6
+0x2a3	    movd                8(%rdi), %xmm5
+0x2a8	    pshufd              $0, %xmm5, %xmm5
+0x2ad	    pmulld              %xmm5, %xmm6
+0x2b2	    paddd               %xmm4, %xmm6
+0x2b6	    pcmpgtd             %xmm1, %xmm7
+0x2ba	    movdqa              %xmm0, %xmm1
+0x2be	    pcmpgtd             %xmm12, %xmm1
+0x2c3	    pand                %xmm5, %xmm1
+0x2c7	    pand                %xmm7, %xmm1
+0x2cb	    paddd               %xmm6, %xmm1
+0x2cf	    pcmpgtd             %xmm8, %xmm3
+0x2d4	    por                 %xmm3, %xmm2
+0x2d8	    pand                112222240(%rip), %xmm2
+0x2e0	    psubd               %xmm3, %xmm2
+0x2e4	    movq                (%rdi), %rax
+0x2e7	    movd                %xmm6, %esi
+0x2eb	    movslq              %esi, %rsi
+0x2ee	    movzwl              (%rax,%rsi), %esi
+0x2f2	    movd                %esi, %xmm3
+0x2f6	    pextrd              $1, %xmm6, %esi
+0x2fc	    movslq              %esi, %rsi
+0x2ff	    movzwl              (%rax,%rsi), %esi
+0x303	    movd                %esi, %xmm4
+0x307	    pextrd              $2, %xmm6, %esi
+0x30d	    punpcklwd           %xmm4, %xmm3
+0x311	    movslq              %esi, %rsi
+0x314	    movzwl              (%rax,%rsi), %esi
+0x318	    movd                %esi, %xmm4
+0x31c	    pextrd              $3, %xmm6, %esi
+0x322	    movslq              %esi, %rsi
+0x325	    movzwl              (%rax,%rsi), %esi
+0x329	    movd                %esi, %xmm5
+0x32d	    punpcklwd           %xmm5, %xmm4
+0x331	    movd                %xmm1, %esi
+0x335	    movslq              %esi, %rsi
+0x338	    movzwl              (%rax,%rsi), %esi
+0x33c	    movd                %esi, %xmm5
+0x340	    pextrd              $1, %xmm1, %esi
+0x346	    punpckldq           %xmm4, %xmm3
+0x34a	    movslq              %esi, %rsi
+0x34d	    movzwl              (%rax,%rsi), %esi
+0x351	    movd                %esi, %xmm4
+0x355	    punpcklwd           %xmm4, %xmm5
+0x359	    pextrd              $2, %xmm1, %esi
+0x35f	    movslq              %esi, %rsi
+0x362	    movzwl              (%rax,%rsi), %esi
+0x366	    movd                %esi, %xmm4
+0x36a	    pextrd              $3, %xmm1, %esi
+0x370	    movslq              %esi, %rsi
+0x373	    movzwl              (%rax,%rsi), %eax
+0x377	    movd                %eax, %xmm1
+0x37b	    punpcklwd           %xmm1, %xmm4
+0x37f	    packssdw            %xmm2, %xmm2
+0x383	    pshufb              112222004(%rip), %xmm0
+0x38c	    pand                112222012(%rip), %xmm0
+0x394	    pmovzxbw            %xmm3, %xmm1
+0x399	    punpckldq           %xmm4, %xmm5
+0x39d	    pmovzxbw            %xmm5, %xmm3
+0x3a2	    psubw               %xmm1, %xmm3
+0x3a6	    punpcklwd           %xmm0, %xmm0
+0x3aa	    pmullw              %xmm3, %xmm0
+0x3ae	    psraw               $7, %xmm0
+0x3b3	    paddw               %xmm1, %xmm0
+0x3b7	    pshufb              112222064(%rip), %xmm0
+0x3c0	    pshufd              $238, %xmm0, %xmm1
+0x3c5	    psubw               %xmm0, %xmm1
+0x3c9	    pmullw              %xmm2, %xmm1
+0x3cd	    psraw               $7, %xmm1
+0x3d2	    paddw               %xmm0, %xmm1
+0x3d6	    punpcklwd           %xmm1, %xmm1
+0x3da	    movdqa              %xmm1, %xmm0
+0x3de	    punpcklwd           %xmm1, %xmm0
+0x3e2	    punpckhwd           %xmm1, %xmm1
+0x3e6	    movdqa              -512(%rbp), %xmm2
+0x3ee	    pmullw              %xmm2, %xmm1
+0x3f2	    pmullw              %xmm10, %xmm0
+0x3f7	    paddw               %xmm10, %xmm0
+0x3fc	    paddw               %xmm2, %xmm1
+0x400	    psrlw               $8, %xmm1
+0x405	    psrlw               $8, %xmm0
+0x40a	    movdqu              -16(%rbx), %xmm2
+0x40f	    movdqa              %xmm2, %xmm12
+0x414	    punpckhbw           %xmm14, %xmm12
+0x419	    movdqa              %xmm2, %xmm8
+0x41e	    pmovzxbw            %xmm2, %xmm14
+0x424	    movzbl              %dl, %eax
+0x427	    movslq              (%r13,%rax,4), %rax
+0x42c	    addq                %r13, %rax
+0x42f	    jmpq                *%rax


+0x477b	    pshuflw             $255, %xmm0, %xmm2
+0x4780	    paddw               %xmm14, %xmm0
+0x4785	    pshufhw             $255, %xmm2, %xmm2
+0x478a	    pshuflw             $255, %xmm1, %xmm3
+0x478f	    paddw               %xmm12, %xmm1
+0x4794	    pshufhw             $255, %xmm3, %xmm3
+0x4799	    pmullw              %xmm12, %xmm3
+0x479e	    pmullw              %xmm14, %xmm2
+0x47a3	    paddw               %xmm14, %xmm2
+0x47a8	    paddw               %xmm12, %xmm3
+0x47ad	    psrlw               $8, %xmm3
+0x47b2	    psubw               %xmm3, %xmm1
+0x47b6	    psrlw               $8, %xmm2
+0x47bb	    psubw               %xmm2, %xmm0
+0x47bf	    packuswb            %xmm1, %xmm0
+0x47c3	    movdqu              %xmm0, -16(%rbx)
+0x47c8	    cmpq                %r8, %rbx
+0x47cb	    jb                  "int blendTextureLinearR8<true, glsl::sampler2D_impl*, unsigned short vector[16], unsigned int>(glsl::sampler2D_impl*, glsl::vec2, int, glsl::vec4_scalar const&, unsigned short vector[16], unsigned int*)+0x4936"
+0x47d1	    jmp                 "int blendTextureLinearR8<true, glsl::sampler2D_impl*, unsigned short vector[16], unsigned int>(glsl::sampler2D_impl*, glsl::vec2, int, glsl::vec4_scalar const&, unsigned short vector[16], unsigned int*)+0x4a60"

+0x4936	    addq                $16, %r14
+0x493a	    movaps              -304(%rbp), %xmm0
+0x4941	    addps               -432(%rbp), %xmm0
+0x4948	    movaps              %xmm0, -304(%rbp)
+0x494f	    movaps              -288(%rbp), %xmm0
+0x4956	    addps               -416(%rbp), %xmm0
+0x495d	    movaps              %xmm0, -288(%rbp)
+0x4964	    movl                12(%rdi), %r12d
+0x4968	    movl                16(%rdi), %r11d
+0x496c	    addq                $16, %rbx
+0x4970	    addq                $16, %rcx
+0x4974	    jmp                 "int blendTextureLinearR8<true, glsl::sampler2D_impl*, unsigned short vector[16], unsigned int>(glsl::sampler2D_impl*, glsl::vec2, int, glsl::vec4_scalar const&, unsigned short vector[16], unsigned int*)+0x210"