Changeset 3e94658359c07330b5ea11ae8520fd41e676bba5

Timestamp:

04/21/08 03:45:43 (9 months ago)

Author:

krayouva <krayouva@…>

Parents:

52150679808ed6ea37f1eb4c564756a1894e43d6, efd6cf5568bba67ec0b7453c93aa3e0c862d864a

Children:

585583fabc2d8f758ad113f07b02b6e37e03ca63, 7ef4a33027485f8cdf5bf9ae4e417537317e80d8, 4640c75a8385ecb909a2b09d28eca23a6c4f6130

git-committer:

krayouva <krayouva@gmail.com> / 2008-04-20T13:45:43Z-0400

Message:

Merge branch 'master' of ssh://git@davebalmain.com/var/git/ferret

Files:

• TODO (modified) (2 diffs)
• c/Makefile (modified) (7 diffs)
• c/benchmark/benchmark.c (modified) (6 diffs)

TODO

@@ -25 +25 @@
   - Make BitVector run as fast as bitset from C++ STL. See;
       c/benchmark/bm_bitvector.c
+  - Add a symbol table for field names. This will mean that we won't need to
+    worry about mallocing and freeing field names which happens all over the
+    place.
+  - Divide the headers into public and private (the private headers to be
+    stored in the src directory).
+  - Group-by search. ie you should be able to pass a field to group search
+    results by
+  - Auto-loading of documents during search. ie actual documents get returned
+    instead of document numbers.
+  - update benchmark suite to use getrusage.u
 
 * Ruby bindings
@@ -64 +74 @@
 * user defined sorting
 * Fix highlighting to work for external fields
+* investigate faster string hashing method
 
 Done

c/Makefile

@@ -21 +21 @@
 
 CC       = gcc
-CINCS    = -Iinclude -I$(STEMMER_INC) -I$(BZLIB_INC)
+CINCS    = -Iinclude -I$(STEMMER_INC) -I$(BZLIB_INC) -Itest
 DEFS     = -DBZ_NO_STDIO -D_FILE_OFFSET_BITS=64 -DDEBUG -D_POSIX_C_SOURCE=2
 DEFS    += -DHAVE_GDB
@@ -28 +28 @@
 DEP_DIR  = .deps
 GCOV_DIR = .gcov
-VPATH    = test:src
+VPATH    = test:src:benchmark
 
 ###
@@ -62 +62 @@
 test_lang.o
 
+BENCH_OBJS = benchmark.o bm_bitvector.o bm_hash.o \
+             bm_micro_string.o bm_store.o         \
+
 BZLIB_SRCS = \
 bzlib.c blocksort.c compress.c crctable.c decompress.c huffman.c randtable.c
@@ -69 +72 @@
 STEMMER_OBJS = $(snowball_sources:%.c=$(STEMMER_DIR)/%.o)
 
-FRT_OBJS = $(OBJS) $(TEST_OBJS)
+FRT_OBJS = $(OBJS) $(TEST_OBJS) $(BENCH_OBJS)
 EXT_OBJS = $(BZLIB_OBJS) $(STEMMER_OBJS)
 
@@ -88 +91 @@
 testall: $(TEST_OBJS) libferret.a
         @echo Building task: $@ ...
-        $(CC) $(CFLAGS) $(LDFLAGS) $^ -o $@
+        @$(CC) $(CFLAGS) $(LDFLAGS) $^ -o $@
+
+benchall: $(BENCH_OBJS) libferret.a
+        @echo Building task: $@ ...
+        @$(CC) $(CFLAGS) $(LDFLAGS) $^ -o $@
+
+bench: benchall
+        @./benchall
 
 clean:
@@ -124 +134 @@
                  --gen-suppressions=yes       \
                  --workaround-gcc296-bugs=yes -v ./testall -q
-
-###
-# Benchmarking
-###
-
-# TODO
 
 ###
@@ -176 +180 @@
 ###
 
-.PHONY: clean all test gcov gprof valgrind gen_valsupp
+.PHONY: clean all test bench gcov gprof valgrind gen_valsupp

c/benchmark/benchmark.c

@@ -1 @@
-#include <sys/times.h>
 #include <sys/time.h>
+#include <sys/resource.h>
 #include <unistd.h>
-#include <limits.h>
 #include <string.h>
 #include "global.h"
@@ -9 +8 @@
 #include "word_list.h"
 
-static int hertz;
-
 static int bmtcmp(const void *p1, const void *p2)
 {
     BenchMarkTimes *bmt1 = *(BenchMarkTimes **)p1;
     BenchMarkTimes *bmt2 = *(BenchMarkTimes **)p2;
-    
+
     if (bmt1->rtime > bmt2->rtime) return 1;
     else if (bmt1->rtime < bmt2->rtime) return -1;
@@ -60 +57 @@
 }
 
+#define TVAL_TO_SEC(before, after) \
+  ((double)after.tv_sec  + ((double)after.tv_usec/1000000)) - \
+  ((double)before.tv_sec + ((double)before.tv_usec/1000000))
+
 static void bm_single_run(BenchMarkUnit *unit, BenchMarkTimes *bm_times)
 {
     struct timeval tv_before, tv_after;
-    struct tms tms_before, tms_after;
-    double before, after;
+    struct rusage ru_before, ru_after;
 
     if (gettimeofday(&tv_before, NULL) == -1)
         RAISE(FRT_UNSUPPORTED_ERROR, "gettimeofday failed\n");
-    (void)times(&tms_before);
+    getrusage(RUSAGE_SELF, &ru_before);
 
     unit->run();
@@ -74 +74 @@
     if (gettimeofday(&tv_after, NULL) == -1)
         RAISE(FRT_UNSUPPORTED_ERROR, "gettimeofday failed\n");
-    (void)times(&tms_after);
-    before = (double)tv_before.tv_sec + ((double)tv_before.tv_usec/1000000);
-    after = (double)tv_after.tv_sec + ((double)tv_after.tv_usec/1000000);
-    bm_times->rtime = after - before;
-    bm_times->utime =
-        ((double)(tms_after.tms_utime - tms_before.tms_utime))/hertz;
-    bm_times->stime =
-        ((double)(tms_after.tms_stime - tms_before.tms_stime))/hertz;
+    getrusage(RUSAGE_SELF, &ru_after);
+
+    bm_times->rtime = TVAL_TO_SEC(tv_before, tv_after);
+    bm_times->utime = TVAL_TO_SEC(ru_before.ru_utime, ru_after.ru_utime);
+    bm_times->stime = TVAL_TO_SEC(ru_before.ru_stime, ru_after.ru_stime);
 }
 
@@ -132 +129 @@
         DO_TEARDOWN(benchmark);
     }
-    
+
     /* get maximum unit name length for print out */
     for (unit = benchmark->head; unit; unit = unit->next) {
@@ -159 +156 @@
     (void)argc; (void)argv;
     benchmark.head = benchmark.tail = NULL;
-    /* set the clock speed used for calculating run times */
-    hertz = sysconf(_SC_CLK_TCK);
 
     for (i = 0; i < NELEMS(all_benchmarks); i++) {