On Tue, Aug 23, 2016 at 8:41 AM, Robert Haas <robertmh...@gmail.com> wrote:
> We could test to see how much it slows things down. But it
> may be worth paying the cost even if it ends up being kinda expensive.
Here are some numbers from a Xeon E7-8830 @ 2.13GHz running Linux 3.10
running the attached program. It's fairly noisy and I didn't run
super long tests with many repeats, but the general theme is visible.
If you're actually going to USE the memory, it's only a small extra
cost to have reserved seats. But if there's a strong chance you'll
never access most of the memory, you might call it expensive.
Segment size 1MB:
base = shm_open + ftruncate + mmap + munmap + close = 5us
base + fallocate = 38us
base + memset = 332us
base + fallocate + memset = 346us
Segment size 1GB:
base = shm_open + ftruncate + mmap + munmap + close = 10032us
base + fallocate = 30774us
base + memset = 602925us
base + fallocate + memset = 655433us
--
Thomas Munro
http://www.enterprisedb.com
#include <fcntl.h>
#include <stdbool.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/mman.h>
#include <unistd.h>
#define SEGMENT_NAME "/my_test_segment"
int
main(int argc, char *argv[])
{
int loops, i;
off_t size;
bool fallocatep;
bool memsetp;
bool hugep;
void *mem;
if (argc != 6)
{
fprintf(stderr, "Usage: %s <loops> <size_in_mb> <fallocate?> <memset?> <huge?>\n", argv[0]);
return EXIT_FAILURE;
}
loops = atoi(argv[1]);
size = atoi(argv[2]) * 1024 * 1024;
fallocatep = atoi(argv[3]) != 0;
memsetp = atoi(argv[4]) != 0;
hugep = atoi(argv[5]) != 0;
for (i = 0; i < loops; ++i)
{
int fd;
fd = shm_open(SEGMENT_NAME, O_CREAT | O_RDWR, S_IWUSR | S_IRUSR);
if (fd < 0)
{
perror("shm_open");
goto cleanup;
}
if (ftruncate(fd, size))
{
perror("ftruncate");
goto cleanup;
}
if (fallocatep && fallocate(fd, 0, 0, size))
{
perror("fallocate");
goto cleanup;
}
mem = mmap(0, size, PROT_READ | PROT_WRITE, MAP_SHARED | (hugep ? MAP_HUGETLB : 0), fd, 0);
if (mem == NULL)
{
fprintf(stderr, "mmap failed");
goto cleanup;
}
if (memsetp)
memset(mem, 0, size);
munmap(mem, size);
close(fd);
}
shm_unlink(SEGMENT_NAME);
return EXIT_SUCCESS;
cleanup:
shm_unlink(SEGMENT_NAME);
return EXIT_FAILURE;
}
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers
