From: Jan Stancek <jstancek@redhat.com>
To: DAN LI <li.dan@cn.fujitsu.com>
Cc: LTP list <ltp-list@lists.sourceforge.net>
Subject: Re: [LTP] [PATCH] mmap/mmap12.c: new case to test MAP_POPULATE of mmap
Date: Wed, 19 Jun 2013 08:00:17 -0400 (EDT) [thread overview]
Message-ID: <22953735.4884030.1371643217523.JavaMail.root@redhat.com> (raw)
In-Reply-To: <51C1553B.7090600@cn.fujitsu.com>
----- Original Message -----
> From: "DAN LI" <li.dan@cn.fujitsu.com>
> To: "LTP list" <ltp-list@lists.sourceforge.net>
> Sent: Wednesday, 19 June, 2013 8:52:43 AM
> Subject: [LTP] [PATCH] mmap/mmap12.c: new case to test MAP_POPULATE of mmap
>
>
> Create test case for MAP_POPULATE of mmap.
>
> Verify the following statement
> "For a file mapping,this causes read-ahead on the file.
> Later accesses to the mapping will not be blocked by page faults."
> by mmap-ing a file with specified size and checking change of memory size.
>
>
> Signed-off-by: DAN LI <li.dan@cn.fujitsu.com>
Hi,
I think this isn't very reliable when you are low on free memory.
I could get it to fail easily just by reading random data to fill cache:
# free -m
total used free shared buffers cached
Mem: 994 704 289 0 0 525
-/+ buffers/cache: 179 814
Swap: 2015 0 2015
# ./mmap12
mmap12 1 TFAIL : Seems read-ahead do not work
Dropping caches after you create test file might not be bad idea:
system("echo 1 > /proc/sys/vm/drop_caches");
because I think sysinfo() sees that as used memory.
And also making sure there is enough free mem (say 2x of testfile)
It's still approximate, but I'm not sure if there's better way,
considering other options:
reading /proc/pid/io for read_bytes
reading /proc/pid/pagemap and checking that those pages are present
Regards,
Jan
> ---
> testcases/kernel/syscalls/mmap/mmap12.c | 163
> ++++++++++++++++++++++++++++++++
> 1 file changed, 163 insertions(+)
> create mode 100644 testcases/kernel/syscalls/mmap/mmap12.c
>
> diff --git a/testcases/kernel/syscalls/mmap/mmap12.c
> b/testcases/kernel/syscalls/mmap/mmap12.c
> new file mode 100644
> index 0000000..4092806
> --- /dev/null
> +++ b/testcases/kernel/syscalls/mmap/mmap12.c
> @@ -0,0 +1,163 @@
> +/*
> + * Copyright (c) 2013 FNST, DAN LI <li.dan@cn.fujitsu.com>
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License as published by
> + * the Free Software Foundation; either version 2 of the License, or
> + * (at your option) any later version.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
> + * the GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; if not, write to the Free Software
> + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301
> USA
> + */
> +
> +/*
> + * Test Description:
> + * Verify MAP_POPULATE works fine.
> + * "For a file mapping, this causes read-ahead on the file.
> + * Later accesses to the mapping will not be blocked by page faults"
> + *
> + * Expected Result:
> + * mmap() with MAP_POPULATE should succeed returning the address of the
> + * mapped region and this file has been read into RAM, so RAM should
> + * is reduced by at least size of file.
> + */
> +#include <stdio.h>
> +#include <stdlib.h>
> +#include <sys/types.h>
> +#include <errno.h>
> +#include <unistd.h>
> +#include <fcntl.h>
> +#include <string.h>
> +#include <signal.h>
> +#include <stdint.h>
> +#include <sys/stat.h>
> +#include <sys/mman.h>
> +#include <sys/shm.h>
> +#include <sys/sysinfo.h>
> +
> +#include "test.h"
> +#include "usctest.h"
> +
> +#define TEMPFILE "mmapfile"
> +#define MMAPSIZE (400UL<<20)
> +#define FLUCTUATION (50UL<<20)
> +
> +char *TCID = "mmap12";
> +int TST_TOTAL = 1;
> +
> +static char *addr;
> +static int fildes;
> +static struct sysinfo info;
> +static unsigned long long freeram_before;
> +static unsigned long long freeram_after;
> +
> +static void loop_setup(void);
> +static void loop_cleanup(void);
> +static void setup(void);
> +static void cleanup(void);
> +
> +int main(int argc, char *argv[])
> +{
> + int lc;
> + int ret = 0;
> + char *msg;
> +
> + msg = parse_opts(argc, argv, NULL, NULL);
> + if (msg != NULL)
> + tst_brkm(TBROK, NULL, "OPTION PARSING ERROR - %s", msg);
> +
> + setup();
> +
> + for (lc = 0; TEST_LOOPING(lc); lc++) {
> +
> + tst_count = 0;
> +
> + loop_setup();
> +
> + errno = 0;
> + addr = mmap(NULL, MMAPSIZE, PROT_READ | PROT_WRITE,
> + MAP_FILE | MAP_PRIVATE | MAP_POPULATE, fildes, 0);
> +
> + if (addr == MAP_FAILED) {
> + tst_resm(TFAIL | TERRNO, "mmap of %s failed", TEMPFILE);
> + continue;
> + }
> +
> + if (STD_FUNCTIONAL_TEST) {
> + ret = sysinfo(&info);
> + if (ret != 0)
> + tst_brkm(TFAIL, cleanup, "sysinfo failed");
> + freeram_after =
> + (unsigned long long)info.freeram *
> + (unsigned long long)info.mem_unit;
> +
> + if ((freeram_before - freeram_after) >=
> + (MMAPSIZE - FLUCTUATION))
> + tst_resm(TPASS,
> + "Functionality of mmap() successful");
> + else
> + tst_resm(TFAIL,
> + "Seems read-ahead do not work");
> + }
> + loop_cleanup();
> + }
> +
> + cleanup();
> + tst_exit();
> +
> +}
> +
> +static void loop_setup(void)
> +{
> + int ret;
> +
> + fildes = open(TEMPFILE, O_RDWR | O_CREAT, 0666);
> + if (fildes < 0)
> + tst_brkm(TFAIL, cleanup, "opening %s failed", TEMPFILE);
> +
> +
> + if (ftruncate(fildes, MMAPSIZE) < 0)
> + tst_brkm(TFAIL | TERRNO, cleanup, "ftruncate file failed");
> +
> + ret = sysinfo(&info);
> + if (ret != 0)
> + tst_brkm(TFAIL, cleanup, "sysinfo failed");
> +
> + freeram_before =
> + (unsigned long long)info.freeram *
> + (unsigned long long)info.mem_unit;
> +}
> +
> +static void loop_cleanup(void)
> +{
> + if (munmap(addr, MMAPSIZE) != 0)
> + tst_brkm(TFAIL | TERRNO, NULL, "munmap failed");
> +
> + close(fildes);
> +}
> +
> +static void setup(void)
> +{
> + tst_sig(FORK, DEF_HANDLER, cleanup);
> +
> + if ((tst_kvercmp(2, 6, 23)) < 0)
> + tst_brkm(TCONF, NULL,
> + "This test can only run on kernels that are 2.6.23 and "
> + "higher");
> +
> + TEST_PAUSE;
> +
> + tst_tmpdir();
> +}
> +
> +static void cleanup(void)
> +{
> + TEST_CLEANUP;
> + tst_rmdir();
> +}
> --
> 1.8.1
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Windows:
>
> Build for Windows Store.
>
> http://p.sf.net/sfu/windows-dev2dev
> _______________________________________________
> Ltp-list mailing list
> Ltp-list@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/ltp-list
>
------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:
Build for Windows Store.
http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Ltp-list mailing list
Ltp-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ltp-list
next prev parent reply other threads:[~2013-06-19 12:00 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-06-19 6:52 [LTP] [PATCH] mmap/mmap12.c: new case to test MAP_POPULATE of mmap DAN LI
2013-06-19 10:05 ` Caspar Zhang
2013-06-19 10:59 ` chrubis
2013-06-19 11:21 ` chrubis
2013-06-19 12:00 ` Jan Stancek [this message]
2013-06-20 7:11 ` DAN LI
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=22953735.4884030.1371643217523.JavaMail.root@redhat.com \
--to=jstancek@redhat.com \
--cc=li.dan@cn.fujitsu.com \
--cc=ltp-list@lists.sourceforge.net \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox