Skip to content

Commit 719b68f

Browse files
authored
Merge pull request #1473 from martin-frbg/p2align
Replace .align with .p2aligns in dscal.c and the Nehalem microkernels as well
2 parents fe9f15f + 497f0c3 commit 719b68f

12 files changed

+16
-16
lines changed

kernel/x86_64/daxpy_microk_nehalem-2.c

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -39,7 +39,7 @@ static void daxpy_kernel_8( BLASLONG n, FLOAT *x, FLOAT *y, FLOAT *alpha)
3939
"movsd (%4), %%xmm0 \n\t" // alpha
4040
"shufpd $0, %%xmm0, %%xmm0 \n\t"
4141

42-
".align 16 \n\t"
42+
".p2align 4 \n\t"
4343
"1: \n\t"
4444
// "prefetcht0 192(%2,%0,8) \n\t"
4545
// "prefetcht0 192(%3,%0,8) \n\t"

kernel/x86_64/ddot_microk_nehalem-2.c

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -41,7 +41,7 @@ static void ddot_kernel_8( BLASLONG n, FLOAT *x, FLOAT *y, FLOAT *dot)
4141
"xorpd %%xmm6, %%xmm6 \n\t"
4242
"xorpd %%xmm7, %%xmm7 \n\t"
4343

44-
".align 16 \n\t"
44+
".p2align 4 \n\t"
4545
"1: \n\t"
4646

4747
"movups (%2,%0,8), %%xmm12 \n\t" // 2 * x

kernel/x86_64/dgemv_n_microk_nehalem-4.c

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -62,7 +62,7 @@ static void dgemv_kernel_4x4( BLASLONG n, FLOAT **ap, FLOAT *x, FLOAT *y, FLOAT
6262
"subq $4 , %1 \n\t"
6363
"jz 2f \n\t"
6464

65-
".align 16 \n\t"
65+
".p2align 4 \n\t"
6666
"1: \n\t"
6767

6868
"xorpd %%xmm4 , %%xmm4 \n\t"

kernel/x86_64/dscal.c

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -99,7 +99,7 @@ static void dscal_kernel_inc_8(BLASLONG n, FLOAT *alpha, FLOAT *x, BLASLONG inc_
9999

100100
"leaq (%1,%4,4), %2 \n\t"
101101

102-
".align 16 \n\t"
102+
".p2align 4 \n\t"
103103

104104
"1: \n\t"
105105
"movsd (%1) , %%xmm4 \n\t"

kernel/x86_64/dsymv_L_microk_nehalem-2.c

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -47,7 +47,7 @@ static void dsymv_kernel_4x4(BLASLONG from, BLASLONG to, FLOAT **a, FLOAT *x, FL
4747
"shufpd $0, %%xmm6, %%xmm6 \n\t"
4848
"shufpd $0, %%xmm7, %%xmm7 \n\t"
4949

50-
".align 16 \n\t"
50+
".p2align 4 \n\t"
5151
"1: \n\t"
5252
"movups (%4,%0,8), %%xmm12 \n\t" // 2 * a
5353
"movups (%2,%0,8), %%xmm8 \n\t" // 2 * x

kernel/x86_64/dsymv_U_microk_nehalem-2.c

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -50,7 +50,7 @@ static void dsymv_kernel_4x4(BLASLONG n, FLOAT *a0, FLOAT *a1, FLOAT *a2, FLOAT
5050

5151
"xorq %0,%0 \n\t"
5252

53-
".align 16 \n\t"
53+
".p2align 4 \n\t"
5454
"1: \n\t"
5555
"movups (%4,%0,8), %%xmm12 \n\t" // 2 * a
5656
"movups (%2,%0,8), %%xmm8 \n\t" // 2 * x

kernel/x86_64/saxpy_microk_nehalem-2.c

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -39,7 +39,7 @@ static void saxpy_kernel_16( BLASLONG n, FLOAT *x, FLOAT *y, FLOAT *alpha)
3939
"movss (%4), %%xmm0 \n\t" // alpha
4040
"shufps $0, %%xmm0, %%xmm0 \n\t"
4141

42-
".align 16 \n\t"
42+
".p2align 4 \n\t"
4343
"1: \n\t"
4444
// "prefetcht0 192(%2,%0,4) \n\t"
4545
// "prefetcht0 192(%3,%0,4) \n\t"

kernel/x86_64/sdot_microk_nehalem-2.c

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -41,7 +41,7 @@ static void sdot_kernel_16( BLASLONG n, FLOAT *x, FLOAT *y, FLOAT *dot)
4141
"xorps %%xmm6, %%xmm6 \n\t"
4242
"xorps %%xmm7, %%xmm7 \n\t"
4343

44-
".align 16 \n\t"
44+
".p2align 4 \n\t"
4545
"1: \n\t"
4646
"movups (%2,%0,4), %%xmm12 \n\t" // 4 * x
4747
"movups (%3,%0,4), %%xmm8 \n\t" // 4 * x

kernel/x86_64/sgemv_n_microk_nehalem-4.c

Lines changed: 5 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -59,18 +59,18 @@ static void sgemv_kernel_4x8( BLASLONG n, FLOAT **ap, FLOAT *x, FLOAT *y, BLASLO
5959
"shufps $0, %%xmm6 , %%xmm6 \n\t"
6060

6161

62-
".align 16 \n\t"
62+
".p2align 4 \n\t"
6363
"1: \n\t"
6464
"xorps %%xmm4 , %%xmm4 \n\t"
6565
"xorps %%xmm5 , %%xmm5 \n\t"
6666
"movups (%3,%0,4), %%xmm7 \n\t" // 4 * y
6767

68-
".align 2 \n\t"
68+
".p2align 1 \n\t"
6969
"movups (%4,%0,4), %%xmm8 \n\t"
7070
"movups (%5,%0,4), %%xmm9 \n\t"
7171
"movups (%6,%0,4), %%xmm10 \n\t"
7272
"movups (%7,%0,4), %%xmm11 \n\t"
73-
".align 2 \n\t"
73+
".p2align 1 \n\t"
7474
"mulps %%xmm12, %%xmm8 \n\t"
7575
"mulps %%xmm13, %%xmm9 \n\t"
7676
"mulps %%xmm14, %%xmm10 \n\t"
@@ -84,7 +84,7 @@ static void sgemv_kernel_4x8( BLASLONG n, FLOAT **ap, FLOAT *x, FLOAT *y, BLASLO
8484
"movups (%5,%8,4), %%xmm9 \n\t"
8585
"movups (%6,%8,4), %%xmm10 \n\t"
8686
"movups (%7,%8,4), %%xmm11 \n\t"
87-
".align 2 \n\t"
87+
".p2align 1 \n\t"
8888
"mulps %%xmm0 , %%xmm8 \n\t"
8989
"mulps %%xmm1 , %%xmm9 \n\t"
9090
"mulps %%xmm2 , %%xmm10 \n\t"
@@ -154,7 +154,7 @@ static void sgemv_kernel_4x4( BLASLONG n, FLOAT **ap, FLOAT *x, FLOAT *y, FLOAT
154154
"movss (%8), %%xmm6 \n\t" // alpha
155155
"shufps $0, %%xmm6 , %%xmm6 \n\t"
156156

157-
".align 16 \n\t"
157+
".p2align 4 \n\t"
158158
"1: \n\t"
159159
"xorps %%xmm4 , %%xmm4 \n\t"
160160
"movups (%3,%0,4), %%xmm7 \n\t" // 4 * y

kernel/x86_64/sgemv_t_microk_nehalem-4.c

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -40,7 +40,7 @@ static void sgemv_kernel_4x4( BLASLONG n, FLOAT **ap, FLOAT *x, FLOAT *y)
4040
"xorps %%xmm6 , %%xmm6 \n\t"
4141
"xorps %%xmm7 , %%xmm7 \n\t"
4242

43-
".align 16 \n\t"
43+
".p2align 4 \n\t"
4444
"1: \n\t"
4545

4646
"movups (%2,%0,4), %%xmm12 \n\t" // 4 * x

0 commit comments

Comments
 (0)