* Calculation of chunk range limits was ignoring num_cpu

bug introduced by me in #1262 - should fix #1388

* Calculation of range limits was ignoring num_cpu

bug introduced by me in #1262

* Calculation of chunk range limits was ignoring num_cpu

bug introduced by me in #1262

* Calculation of chunk range limits was ignoring num_cpu

bug introduced by me in #1262

* Calculation of chunk range limits was ignoring num_cpu

bug introduced by me in #1262

* Calculation of chunk range limits was ignoring num_cpu

bug introduced by me in #1262
This commit is contained in:
Martin Kroeker 2017-12-09 22:29:03 +01:00 committed by GitHub
parent 281a2b952f
commit 177b78c8b4
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
6 changed files with 14 additions and 14 deletions

View File

@ -230,10 +230,10 @@ int CNAME(BLASLONG m, BLASLONG n, BLASLONG ku, BLASLONG kl, FLOAT *alpha, FLOAT
#ifndef TRANSA
range_m[num_cpu] = num_cpu * ((m + 15) & ~15);
if (range_m[num_cpu] > m) range_m[num_cpu] = m;
if (range_m[num_cpu] > m * num_cpu) range_m[num_cpu] = m * num_cpu;
#else
range_m[num_cpu] = num_cpu * ((n + 15) & ~15);
if (range_m[num_cpu] > n) range_m[num_cpu] = n;
if (range_m[num_cpu] > n * num_cpu) range_m[num_cpu] = n * num_cpu;
#endif
queue[num_cpu].mode = mode;

View File

@ -246,7 +246,7 @@ int CNAME(BLASLONG n, BLASLONG k, FLOAT *alpha, FLOAT *a, BLASLONG lda, FLOAT *x
range_m[MAX_CPU_NUMBER - num_cpu - 1] = range_m[MAX_CPU_NUMBER - num_cpu] - width;
range_n[num_cpu] = num_cpu * (((n + 15) & ~15) + 16);
if (range_n[num_cpu] > n) range_n[num_cpu] = n;
if (range_n[num_cpu] > n * num_cpu) range_n[num_cpu] = n * num_cpu;
queue[num_cpu].mode = mode;
queue[num_cpu].routine = sbmv_kernel;
@ -286,7 +286,7 @@ int CNAME(BLASLONG n, BLASLONG k, FLOAT *alpha, FLOAT *a, BLASLONG lda, FLOAT *x
range_m[num_cpu + 1] = range_m[num_cpu] + width;
range_n[num_cpu] = num_cpu * (((n + 15) & ~15) + 16);
if (range_n[num_cpu] > n) range_n[num_cpu] = n;
if (range_n[num_cpu] > n * num_cpu) range_n[num_cpu] = n * num_cpu;
queue[num_cpu].mode = mode;
queue[num_cpu].routine = sbmv_kernel;
@ -318,7 +318,7 @@ int CNAME(BLASLONG n, BLASLONG k, FLOAT *alpha, FLOAT *a, BLASLONG lda, FLOAT *x
range_m[num_cpu + 1] = range_m[num_cpu] + width;
range_n[num_cpu] = num_cpu * ((n + 15) & ~15);
if (range_n[num_cpu] > n) range_n[num_cpu] = n;
if (range_n[num_cpu] > n * num_cpu) range_n[num_cpu] = n * num_cpu;
queue[num_cpu].mode = mode;
queue[num_cpu].routine = sbmv_kernel;

View File

@ -246,7 +246,7 @@ int CNAME(BLASLONG m, FLOAT *alpha, FLOAT *a, FLOAT *x, BLASLONG incx, FLOAT *y,
range_m[MAX_CPU_NUMBER - num_cpu - 1] = range_m[MAX_CPU_NUMBER - num_cpu] - width;
range_n[num_cpu] = num_cpu * (((m + 15) & ~15) + 16);
if (range_n[num_cpu] > m) range_n[num_cpu] = m;
if (range_n[num_cpu] > m * num_cpu) range_n[num_cpu] = m * num_cpu;
queue[num_cpu].mode = mode;
queue[num_cpu].routine = spmv_kernel;
@ -286,7 +286,7 @@ int CNAME(BLASLONG m, FLOAT *alpha, FLOAT *a, FLOAT *x, BLASLONG incx, FLOAT *y,
range_m[num_cpu + 1] = range_m[num_cpu] + width;
range_n[num_cpu] = num_cpu * (((m + 15) & ~15) + 16);
if (range_n[num_cpu] > m) range_n[num_cpu] = m;
if (range_n[num_cpu] > m * num_cpu) range_n[num_cpu] = m * num_cpu;
queue[num_cpu].mode = mode;
queue[num_cpu].routine = spmv_kernel;

View File

@ -177,7 +177,7 @@ int CNAME(BLASLONG m, FLOAT *alpha, FLOAT *a, BLASLONG lda, FLOAT *x, BLASLONG i
range_m[num_cpu + 1] = range_m[num_cpu] + width;
range_n[num_cpu] = num_cpu * (((m + 15) & ~15) + 16);
if (range_n[num_cpu] > m) range_n[num_cpu] = m;
if (range_n[num_cpu] > m * num_cpu) range_n[num_cpu] = m * num_cpu;
queue[MAX_CPU_NUMBER - num_cpu - 1].mode = mode;
queue[MAX_CPU_NUMBER - num_cpu - 1].routine = symv_kernel;
@ -226,7 +226,7 @@ int CNAME(BLASLONG m, FLOAT *alpha, FLOAT *a, BLASLONG lda, FLOAT *x, BLASLONG i
range_m[num_cpu + 1] = range_m[num_cpu] + width;
range_n[num_cpu] = num_cpu * (((m + 15) & ~15) + 16);
if (range_n[num_cpu] > m) range_n[num_cpu] = m;
if (range_n[num_cpu] > m * num_cpu) range_n[num_cpu] = m * num_cpu;
queue[num_cpu].mode = mode;
queue[num_cpu].routine = symv_kernel;

View File

@ -288,7 +288,7 @@ int CNAME(BLASLONG n, BLASLONG k, FLOAT *a, BLASLONG lda, FLOAT *x, BLASLONG inc
range_m[MAX_CPU_NUMBER - num_cpu - 1] = range_m[MAX_CPU_NUMBER - num_cpu] - width;
range_n[num_cpu] = num_cpu * (((n + 15) & ~15) + 16);
if (range_n[num_cpu] > n) range_n[num_cpu] = n;
if (range_n[num_cpu] > n * num_cpu) range_n[num_cpu] = n * num_cpu;
queue[num_cpu].mode = mode;
queue[num_cpu].routine = trmv_kernel;
@ -328,7 +328,7 @@ int CNAME(BLASLONG n, BLASLONG k, FLOAT *a, BLASLONG lda, FLOAT *x, BLASLONG inc
range_m[num_cpu + 1] = range_m[num_cpu] + width;
range_n[num_cpu] = num_cpu * (((n + 15) & ~15) + 16);
if (range_n[num_cpu] > n) range_n[num_cpu] = n;
if (range_n[num_cpu] > n * num_cpu) range_n[num_cpu] = n * num_cpu;
queue[num_cpu].mode = mode;
queue[num_cpu].routine = trmv_kernel;
@ -358,7 +358,7 @@ int CNAME(BLASLONG n, BLASLONG k, FLOAT *a, BLASLONG lda, FLOAT *x, BLASLONG inc
range_m[num_cpu + 1] = range_m[num_cpu] + width;
range_n[num_cpu] = num_cpu * (((n + 15) & ~15) + 16);
if (range_n[num_cpu] > n) range_n[num_cpu] = n;
if (range_n[num_cpu] > n * num_cpu) range_n[num_cpu] = n * num_cpu;
queue[num_cpu].mode = mode;
queue[num_cpu].routine = trmv_kernel;

View File

@ -303,7 +303,7 @@ int CNAME(BLASLONG m, FLOAT *a, FLOAT *x, BLASLONG incx, FLOAT *buffer, int nthr
range_m[MAX_CPU_NUMBER - num_cpu - 1] = range_m[MAX_CPU_NUMBER - num_cpu] - width;
range_n[num_cpu] = num_cpu * (((m + 15) & ~15) + 16);
if (range_n[num_cpu] > m) range_n[num_cpu] = m;
if (range_n[num_cpu] > m * num_cpu) range_n[num_cpu] = m * num_cpu;
queue[num_cpu].mode = mode;
queue[num_cpu].routine = tpmv_kernel;
@ -343,7 +343,7 @@ int CNAME(BLASLONG m, FLOAT *a, FLOAT *x, BLASLONG incx, FLOAT *buffer, int nthr
range_m[num_cpu + 1] = range_m[num_cpu] + width;
range_n[num_cpu] = num_cpu * (((m + 15) & ~15) + 16);
if (range_n[num_cpu] > m) range_n[num_cpu] = m;
if (range_n[num_cpu] > m * num_cpu) range_n[num_cpu] = m * num_cpu;
queue[num_cpu].mode = mode;
queue[num_cpu].routine = tpmv_kernel;