From: DAN LI <li.dan@cn.fujitsu.com>
To: Jan Stancek <jstancek@redhat.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: Thu, 20 Jun 2013 15:11:37 +0800 [thread overview]
Message-ID: <51C2AB29.8030207@cn.fujitsu.com> (raw)
In-Reply-To: <22953735.4884030.1371643217523.JavaMail.root@redhat.com>
On 06/19/2013 08:00 PM, Jan Stancek wrote:
>
>
>
> ----- 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)
Maybe i could cut the length of the mapping, and as you suggested,
drop caches, check if there is enough free mem before testing,
to make it more reliable.
>
> 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
I'll try this first. :)
Regards,
DAN LI
>
> 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
prev parent reply other threads:[~2013-06-20 7:12 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
2013-06-20 7:11 ` DAN LI [this message]
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=51C2AB29.8030207@cn.fujitsu.com \
--to=li.dan@cn.fujitsu.com \
--cc=jstancek@redhat.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