Changeset: 5a5ec6733a9d for MonetDB
URL: https://dev.monetdb.org/hg/MonetDB/rev/5a5ec6733a9d
Modified Files:
monetdb5/optimizer/opt_mitosis.c
Branch: Sep2022
Log Message:
Use consistent terminology & fix spelling.
diffs (76 lines):
diff --git a/monetdb5/optimizer/opt_mitosis.c b/monetdb5/optimizer/opt_mitosis.c
--- a/monetdb5/optimizer/opt_mitosis.c
+++ b/monetdb5/optimizer/opt_mitosis.c
@@ -11,8 +11,8 @@
#include "mal_interpreter.h"
#include "gdk_utils.h"
-#define MIN_SLIZE_SIZE 100000 /* minimal record count per partition */
-#define MAX_SLICES2THREADS_RATIO 4 /* There should be at most this
multiple more of slices then threads */
+#define MIN_PART_SIZE 100000 /* minimal record count per partition */
+#define MAX_PARTS2THREADS_RATIO 4 /* There should be at most this
multiple more of partitions then threads */
str
@@ -25,7 +25,7 @@ OPTmitosisImplementation(Client cntxt, M
InstrPtr p, q, *old, target = 0;
size_t argsize = 6 * sizeof(lng), m = 0, memclaim;
/* estimate size per operator estimate: 4 args + 2 res*/
- int threads = GDKnr_threads ? GDKnr_threads : 1, maxslices = MAXSLICES;
+ int threads = GDKnr_threads ? GDKnr_threads : 1, maxparts = MAXSLICES;
str msg = MAL_SUCCEED;
/* if the user has associated limitation on the number of threads,
respect it in the
@@ -53,7 +53,7 @@ OPTmitosisImplementation(Client cntxt, M
/* mitosis/mergetable bailout conditions */
/* Crude protection against self join explosion */
if (p->retc == 2 && isMatJoinOp(p))
- maxslices = threads;
+ maxparts = threads;
nr_aggrs += (p->argc > 2 && getModuleId(p) == aggrRef);
nr_maps += (isMapOp(p));
@@ -169,7 +169,7 @@ OPTmitosisImplementation(Client cntxt, M
*/
/* respect the memory limit size set for the user
- * and determine the column slice size
+ * and determine the column part size
*/
if( cntxt->memorylimit)
m = (((size_t) cntxt->memorylimit) * 1048576) / argsize;
@@ -190,17 +190,17 @@ OPTmitosisImplementation(Client cntxt, M
* i.e., (threads*(rowcnt/pieces) <= m),
* i.e., (rowcnt/pieces <= m/threads),
* i.e., (pieces => rowcnt/(m/threads))
- * (assuming that (m > threads*MIN_SLIZE_SIZE)) */
+ * (assuming that (m > threads*MIN_PART_SIZE)) */
/* the number of pieces affects SF-100, going beyond 8x
increases
* the optimizer costs beyond the execution time
*/
pieces = ((int) ceil((double)rowcnt / (m / threads)));
if (pieces <= threads)
pieces = threads;
- } else if (rowcnt > MIN_SLIZE_SIZE) {
+ } else if (rowcnt > MIN_PART_SIZE) {
/* exploit parallelism, but ensure minimal partition size to
* limit overhead */
- pieces = MIN((int) ceil((double)rowcnt /
MIN_SLIZE_SIZE), MAX_SLICES2THREADS_RATIO * threads);
+ pieces = MIN((int) ceil((double)rowcnt /
MIN_PART_SIZE), MAX_PARTS2THREADS_RATIO * threads);
}
}
@@ -210,10 +210,10 @@ OPTmitosisImplementation(Client cntxt, M
if (pieces < threads)
pieces = (int) MIN((BUN) threads, rowcnt);
/* prevent plan explosion */
- if (pieces > maxslices)
- pieces = maxslices;
+ if (pieces > maxparts)
+ pieces = maxparts;
/* to enable experimentation we introduce the option to set
- * the number of parts required and/or the size of each chunk (in K)
+ * the number of partitions required and/or the size of each chunk (in
K)
*/
mito_parts = GDKgetenv_int("mito_parts", 0);
if (mito_parts > 0)
_______________________________________________
checkin-list mailing list -- [email protected]
To unsubscribe send an email to [email protected]
