Age | Commit message (Collapse) | Author | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
This function caused the failure of ERKALE serial test.
I replaced it with LAPACK source code.
|
|
I used a smaller threshold since the stack size is 1MB on windows.
|
|
When NUM_THREADS(MAX_CPU_NUNBERS) is very large ,e.g. 256.
typedef struct {
volatile BLASLONG working[MAX_CPU_NUMBER][CACHE_LINE_SIZE * DIVIDE_RATE];
} job_t;
job_t job[MAX_CPU_NUMBER];
The job array is equal 8MB.
Thus, We use malloc instead of stack allocation.
|
|
|
|
|
|
instead of laswp_oncopy in getrf.
|
|
|