added config option "memory deduplication (ksm) = yes/no" to control memory deduplication with ksm #38

4 files changed, 38 insertions, 9 deletions

diff --git a/src/common.c b/src/common.c
index 2df676f902..63fed58511 100755
--- a/src/common.c
+++ b/src/common.c
@@ -14,8 +14,10 @@
 
 #include "log.h"
 #include "common.h"
+#include "appconfig.h"
 
 char *global_host_prefix = "";
+int enable_ksm = 1;
 
 /*
 // http://stackoverflow.com/questions/7666509/hash-function-for-string
@@ -107,7 +109,7 @@ char *trim(char *s)
 	return s;
 }
 
-void *mymmap(const char *filename, unsigned long size, int flags)
+void *mymmap(const char *filename, unsigned long size, int flags, int ksm)
 {
 	int fd;
 	void *mem = NULL;
@@ -117,14 +119,35 @@ void *mymmap(const char *filename, unsigned long size, int flags)
 	if(fd != -1) {
 		if(lseek(fd, size, SEEK_SET) == (long)size) {
 			if(write(fd, "", 1) == 1) {
-
 				if(ftruncate(fd, size))
 					error("Cannot truncate file '%s' to size %ld. Will use the larger file.", filename, size);
 
-				mem = mmap(NULL, size, PROT_READ|PROT_WRITE, flags, fd, 0);
-				if(mem) {
-					if(madvise(mem, size, MADV_SEQUENTIAL|MADV_DONTFORK|MADV_WILLNEED) != 0)
-						error("Cannot advise the kernel about the memory usage of file '%s'.", filename);
+				if(flags & MAP_SHARED || !enable_ksm || !ksm) {
+					mem = mmap(NULL, size, PROT_READ|PROT_WRITE, flags, fd, 0);
+					if(mem) {
+						int advise = MADV_SEQUENTIAL|MADV_DONTFORK;
+						if(flags & MAP_SHARED) advise |= MADV_WILLNEED;
+
+						if(madvise(mem, size, advise) != 0)
+							error("Cannot advise the kernel about the memory usage of file '%s'.", filename);
+					}
+				}
+				else {
+					mem = mmap(NULL, size, PROT_READ|PROT_WRITE, flags|MAP_ANONYMOUS, -1, 0);
+					if(mem) {
+						if(lseek(fd, 0, SEEK_SET) == 0) {
+							if(read(fd, mem, size) != size)
+								error("Cannot read from file '%s'", filename);
+						}
+						else
+							error("Cannot seek to beginning of file '%s'.", filename);
+
+						// don't use MADV_SEQUENTIAL|MADV_DONTFORK, they disable MADV_MERGEABLE
+						if(madvise(mem, size, MADV_MERGEABLE) != 0)
+							error("Cannot advise the kernel about the memory usage of file '%s'.", filename);
+					}
+					else
+						error("Cannot allocate PRIVATE ANONYMOUS memory for KSM for file '%s'.", filename);
 				}
 			}
 			else error("Cannot write to file '%s' at position %ld.", filename, size);
diff --git a/src/common.h b/src/common.h
index 46af8bc867..ba00af74fd 100755
--- a/src/common.h
+++ b/src/common.h
@@ -17,12 +17,13 @@ extern void strreverse(char* begin, char* end);
 extern char *mystrsep(char **ptr, char *s);
 extern char *trim(char *s);
 
-extern void *mymmap(const char *filename, unsigned long size, int flags);
+extern void *mymmap(const char *filename, unsigned long size, int flags, int ksm);
 extern int savememory(const char *filename, void *mem, unsigned long size);
 
 extern int fd_is_valid(int fd);
 
 extern char *global_host_prefix;
+extern int enable_ksm;
 
 /* Number of ticks per second */
 #define HZ		hz
diff --git a/src/main.c b/src/main.c
index 11d69cdb15..1126c43866 100755
--- a/src/main.c
+++ b/src/main.c
@@ -289,6 +289,11 @@ int main(int argc, char **argv)
 
 		// --------------------------------------------------------------------
 
+		enable_ksm = config_get_boolean("global", "memory deduplication (ksm)", enable_ksm);
+
+		// --------------------------------------------------------------------
+
+
 		global_host_prefix = config_get("global", "host access prefix", "");
 		setenv("NETDATA_HOST_PREFIX", global_host_prefix, 1);
 
diff --git a/src/rrd.c b/src/rrd.c
index 3b35b4aade..9604337df3 100755
--- a/src/rrd.c
+++ b/src/rrd.c
@@ -353,7 +353,7 @@ RRDSET *rrdset_create(const char *type, const char *id, const char *name, const
 	debug(D_RRD_CALLS, "Creating RRD_STATS for '%s.%s'.", type, id);
 
 	snprintf(fullfilename, FILENAME_MAX, "%s/main.db", cache_dir);
-	if(rrd_memory_mode != RRD_MEMORY_MODE_RAM) st = (RRDSET *)mymmap(fullfilename, size, ((rrd_memory_mode == RRD_MEMORY_MODE_MAP)?MAP_SHARED:MAP_PRIVATE));
+	if(rrd_memory_mode != RRD_MEMORY_MODE_RAM) st = (RRDSET *)mymmap(fullfilename, size, ((rrd_memory_mode == RRD_MEMORY_MODE_MAP)?MAP_SHARED:MAP_PRIVATE), 0);
 	if(st) {
 		if(strcmp(st->magic, RRDSET_MAGIC) != 0) {
 			errno = 0;
@@ -468,7 +468,7 @@ RRDDIM *rrddim_add(RRDSET *st, const char *id, const char *name, long multiplier
 
 	rrdset_strncpy_name(filename, id, FILENAME_MAX);
 	snprintf(fullfilename, FILENAME_MAX, "%s/%s.db", st->cache_dir, filename);
-	if(rrd_memory_mode != RRD_MEMORY_MODE_RAM) rd = (RRDDIM *)mymmap(fullfilename, size, ((rrd_memory_mode == RRD_MEMORY_MODE_MAP)?MAP_SHARED:MAP_PRIVATE));
+	if(rrd_memory_mode != RRD_MEMORY_MODE_RAM) rd = (RRDDIM *)mymmap(fullfilename, size, ((rrd_memory_mode == RRD_MEMORY_MODE_MAP)?MAP_SHARED:MAP_PRIVATE), 1);
 	if(rd) {
 		struct timeval now;
 		gettimeofday(&now, NULL);