Skip to contentSkip to conversation

mm/thp: add two new case

PATCH Superseded Zhouping Liu 2 participants 7 messages Apr 15, 2013 1 message not in the archive

v4 of 5 Superseded Latest is v5 →

lore
Share & tools
lore
#0 lore raw

v4 cover mm/thp: add two new case

The patchset is desinged to test THP functionality.

when one process allocate hugepage aligned anonymouse pages, kernel thread 'khugepaged' controlled by sysfs knobs /sys/kernel/mm/transparent_hugepage/* will scan them, and make them as transparent hugepage if they are suited, you can find out how many transparent hugepages are there in one process from /proc/<pid>/smaps, among the file contents, 'AnonHugePages' entry stand for transparent hugepage.

v1 to v2:
accepted Wanlong's suggestion, added some comment for '10s' added 'static' key word for single-file-only variant.

v2 to v3:
added some comments for low memory system.

v3 to v4:
Wanglong point out that the value of khugepaged/defrag is string type, not a numeric before applying the commit e27e6151b154 in v2.6.39-rc4, and I checked the code again, found the current cases thp0* don't be effected by khugepaged/defrag, so I removed the checking of khugepaged/defrag from khugepaged_scan_done().

Zhouping Liu (3):
mm/thp: new case thp04.c
lib/mem: introduce a new function set_global_mempolicy() mm/thp: add new case thp05

 runtest/mm                         |   6 ++
 testcases/kernel/mem/include/mem.h |  12 +++
 testcases/kernel/mem/lib/mem.c     | 198 ++++++++++++++++++++++++++++++++++++-
 testcases/kernel/mem/thp/thp04.c   | 136 +++++++++++++++++++++++++
 testcases/kernel/mem/thp/thp05.c   | 152 ++++++++++++++++++++++++++++
 5 files changed, 503 insertions(+), 1 deletion(-)
 create mode 100644 testcases/kernel/mem/thp/thp04.c
 create mode 100644 testcases/kernel/mem/thp/thp05.c
Signature · 13 lines
1.7.11.7


------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list
#1 lore raw

PATCH v4 1/3 mm/thp: new case thp04.c

The case is desinged to test THP functionality.

when one process allocate hugepage aligned anonymouse pages, kernel thread 'khugepaged' controlled by sysfs knobs /sys/kernel/mm/transparent_hugepage/* will scan them, and make them as transparent hugepage if they are suited, you can find out how many transparent hugepages are there in one process from /proc/<pid>/smaps, among the file contents, 'AnonHugePages' entry stand for transparent hugepage.

---
 runtest/mm                         |   3 +
 testcases/kernel/mem/include/mem.h |  11 +++
 testcases/kernel/mem/lib/mem.c     | 187 +++++++++++++++++++++++++++++++++++++
 testcases/kernel/mem/thp/thp04.c   | 136 +++++++++++++++++++++++++++
 4 files changed, 337 insertions(+)
 create mode 100644 testcases/kernel/mem/thp/thp04.c
runtest/mm +3 -0
index 56b83f8..3fbb20f 100644
@@ -84,6 +84,9 @@ swapping01 swapping01 -i 5
 thp01 thp01 -I 120
 thp02 thp02
 thp03 thp03
+thp04_1 thp04
+thp04_2 thp04 -n 10 -N 20
+thp04_3 thp04 -n 1 -N 300
 
 vma01 vma01
 vma02 vma02
testcases/kernel/mem/include/mem.h +11 -0
index fdf558e..ccae47d 100644
@@ -32,6 +32,17 @@ void testoom(int mempolicy, int lite);
 
 #define PATH_KSM		"/sys/kernel/mm/ksm/"
 
+/* THP */
+
+#define PATH_THP		"/sys/kernel/mm/transparent_hugepage/"
+#define PATH_KHPD		PATH_THP "khugepaged/"
+
+int opt_nr_child, opt_nr_thps;
+char *opt_nr_child_str, *opt_nr_thps_str;
+void test_transparent_hugepage(int nr_child, int nr_thps, int hg_aligned);
+void check_thp_options(int *nr_child, int *nr_thps);
+void thp_usage(void);
+
 /* HUGETLB */
 
 #define PATH_SHMMAX		"/proc/sys/kernel/shmmax"
testcases/kernel/mem/lib/mem.c +187 -0

193 lines not on this page · raw message

testcases/kernel/mem/thp/thp04.c new file +136 -0

136 lines not on this page · raw message

Signature · 13 lines
1.7.11.7


------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list
#2 lore raw

PATCH v4 2/3 lib/mem: introduce a new function set_global_mempolicy()

Fork out a new function set_global_mempolicy() from the previous testoom() func, which will be useful for other func.

---
 testcases/kernel/mem/lib/mem.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
testcases/kernel/mem/lib/mem.c +6 -1
index 3f9bbfd..35e8bc2 100644
@@ -91,7 +91,7 @@ void oom(int testcase, int lite)
 	}
 }
 
-void testoom(int mempolicy, int lite)
+static void set_global_mempolicy(int mempolicy)
 {
 #if HAVE_NUMA_H && HAVE_LINUX_MEMPOLICY_H && HAVE_NUMAIF_H \
 	&& HAVE_MPOL_CONSTANTS
@@ -133,6 +133,11 @@ void testoom(int mempolicy, int lite)
 			tst_brkm(TBROK|TERRNO, cleanup, "set_mempolicy");
 	}
 #endif
+}
+
+void testoom(int mempolicy, int lite)
+{
+	set_global_mempolicy(mempolicy);
 
 	tst_resm(TINFO, "start normal OOM testing.");
 	oom(NORMAL, lite);
Signature · 13 lines
1.7.11.7


------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list
#3 lore raw

PATCH v4 3/3 mm/thp: add new case thp05

added new case thp05, which is used to test transparent hugepage under mempolicy.

---
 runtest/mm                         |   3 +
 testcases/kernel/mem/include/mem.h |   3 +-
 testcases/kernel/mem/lib/mem.c     |   6 +-
 testcases/kernel/mem/thp/thp04.c   |   2 +-
 testcases/kernel/mem/thp/thp05.c   | 152 +++++++++++++++++++++++++++++++++++++
 5 files changed, 163 insertions(+), 3 deletions(-)
 create mode 100644 testcases/kernel/mem/thp/thp05.c
runtest/mm +3 -0
index 3fbb20f..7c7abf1 100644
@@ -87,6 +87,9 @@ thp03 thp03
 thp04_1 thp04
 thp04_2 thp04 -n 10 -N 20
 thp04_3 thp04 -n 1 -N 300
+thp05_1 thp05
+thp05_2 thp05 -n 10 -N 20
+thp05_3 thp05 -n 1 -N 300
 
 vma01 vma01
 vma02 vma02
testcases/kernel/mem/include/mem.h +2 -1
index ccae47d..6b020af 100644
@@ -39,7 +39,8 @@ void testoom(int mempolicy, int lite);
 
 int opt_nr_child, opt_nr_thps;
 char *opt_nr_child_str, *opt_nr_thps_str;
-void test_transparent_hugepage(int nr_child, int nr_thps, int hg_aligned);
+void test_transparent_hugepage(int nr_child, int nr_thps,
+			       int hg_aligned, int mempolicy);
 void check_thp_options(int *nr_child, int *nr_thps);
 void thp_usage(void);
 
testcases/kernel/mem/lib/mem.c +5 -1
index 35e8bc2..0361eba 100644
@@ -617,12 +617,16 @@ static void verify_thp_size(int *child, int nr_child, int nr_thps)
 	}
 }
 
-void test_transparent_hugepage(int nr_child, int nr_thps, int hg_aligned)
+void test_transparent_hugepage(int nr_child, int nr_thps,
+			       int hg_aligned, int mempolicy)
 {
 	unsigned long hugepagesize, memfree;
 	int i, *pid, ret, status;
 	char path[BUFSIZ];
 
+	if (mempolicy)
+		set_global_mempolicy(mempolicy);
+
 	memfree = read_meminfo("MemFree:");
 	tst_resm(TINFO, "The current MemFree is %luMB", memfree / KB);
 	if (memfree < MB)
testcases/kernel/mem/thp/thp04.c +1 -1
index 0f6c553..ff41f75 100644
@@ -83,7 +83,7 @@ int main(int argc, char *argv[])
 	for (lc = 0; TEST_LOOPING(lc); lc++) {
 		tst_count = 0;
 
-		test_transparent_hugepage(nr_child, nr_thps, 1);
+		test_transparent_hugepage(nr_child, nr_thps, 1, 0);
 	}
 
 	cleanup();
testcases/kernel/mem/thp/thp05.c new file +152 -0

152 lines not on this page · raw message

Signature · 13 lines
1.7.11.7


------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list
#4 lore raw

Re: [LTP] [PATCH v4 3/3] mm/thp: add new case thp05

Hi!

testcases/kernel/mem/thp/thp05.c line 125

Quoted 10 lines
testcases/kernel/mem/thp/thp05.c new file +10 -0
+	snprintf(path, BUFSIZ, PATH_KHPD "scan_sleep_millisecs");
+	SAFE_FILE_SCANF(NULL, path, "%d", &pre_thp_scan_sleep_millisecs);
+	SAFE_FILE_PRINTF(cleanup, path, "%d", 0);
+
+	snprintf(path, BUFSIZ, PATH_KHPD "alloc_sleep_millisecs");
+	SAFE_FILE_SCANF(NULL, path, "%d", &pre_thp_alloc_sleep_millisecs);
+	SAFE_FILE_PRINTF(NULL, path, "%d", 0);
+
+	snprintf(path, BUFSIZ, PATH_THP "enabled");
+	write_file(path, "always");

Same as the previous one. Remove the useless snprintf() and pass the strings directly.

Signature · 13 lines
Cyril Hrubis
[email protected]

------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list
#5 lore raw

Re: [LTP] [PATCH v4 1/3] mm/thp: new case thp04.c

In reply to Zhouping Liu

Hi!

testcases/kernel/mem/lib/mem.c void ksm_usage(void) · line 555

Quoted 11 lines
testcases/kernel/mem/lib/mem.c +11 -0
+static void khugepaged_scan_done(void)
+{
+	int changing = 1, count = 0;
+	long old_pages_collapsed, old_max_ptes_none, old_pages_to_scan;
+	long pages_collapsed = 0, max_ptes_none = 0, pages_to_scan = 0;
+
+	while (changing) {
+		/*
+		 * as 'khugepaged' run 100% during testing, so 10s
+		 * is an enough for us to recognize if 'khugepaged'
+		 * finish scanning proceses' anonymouse hugepages
                                                ^
                                             anonymous
Quoted 16 lines
testcases/kernel/mem/lib/mem.c +16 -0
+		 * or not.
+		 */
+		sleep(10);
+		count++;
+
+		SAFE_FILE_SCANF(cleanup, PATH_KHPD "pages_collapsed",
+			       "%ld", &pages_collapsed);
+		SAFE_FILE_SCANF(cleanup, PATH_KHPD "max_ptes_none",
+			       "%ld", &max_ptes_none);
+		SAFE_FILE_SCANF(cleanup, PATH_KHPD "pages_to_scan",
+			       "%ld", &pages_to_scan);
+
+		if (pages_collapsed != old_pages_collapsed ||
+		    max_ptes_none != old_max_ptes_none ||
+		    pages_to_scan != old_pages_to_scan) {
+			old_pages_collapsed = pages_collapsed;

testcases/kernel/mem/lib/mem.c void ksm_usage(void) · line 581

Quoted 10 lines
testcases/kernel/mem/lib/mem.c +10 -0
+			old_max_ptes_none = max_ptes_none;
+			old_pages_to_scan = pages_to_scan;
+		} else {
+			changing = 0;
+		}
+	}
+
+	tst_resm(TINFO, "khugepaged daemon takes %ds to scan all thp pages",
+		 count * 10);
+}

Ah, so it acutally does polling, this is fine.

How long scanning takes. If it is about 10 seconds, we should poll faster so that the testcase has chance to finish reasonably fast. (if it takes 10 seconds and we miss the first windows it will sleep
doing noting for another ten seconds)

Quoted 33 lines
testcases/kernel/mem/lib/mem.c +33 -0
+static void verify_thp_size(int *child, int nr_child, int nr_thps)
+{
+	FILE *fp;
+	char path[BUFSIZ], buf[BUFSIZ], line[BUFSIZ];
+	int i, ret;
+	long expect_thps; /* the amount of per child's transparent hugepages */
+	long val, actual_thps;
+	long hugepagesize;
+
+	hugepagesize = read_meminfo("Hugepagesize:");
+	expect_thps = nr_thps * hugepagesize;
+
+	for (i = 0; i < nr_child; i++) {
+		actual_thps = 0;
+
+		snprintf(path, BUFSIZ, "/proc/%d/smaps", child[i]);
+		fp = fopen(path, "r");
+		while (fgets(line, BUFSIZ, fp) != NULL) {
+			ret = sscanf(line, "%64s %ld", buf, &val);
+			if (ret == 2 && val != 0) {
+				if (strcmp(buf, "AnonHugePages:") == 0)
+					actual_thps += val;
+			}
+		}
+
+		if (actual_thps != expect_thps)
+			tst_resm(TFAIL, "child[%d] got %ldKB thps - expect %ld"
+				"KB thps", getpid(), actual_thps, expect_thps);
+		fclose(fp);
+	}
+}
+
+void test_transparent_hugepage(int nr_child, int nr_thps, int hg_aligned)

testcases/kernel/mem/lib/mem.c void ksm_usage(void) · line 625

Quoted 10 lines
testcases/kernel/mem/lib/mem.c +10 -0
+{
+	unsigned long hugepagesize, memfree;
+	int i, *pid, ret, status;
+	char path[BUFSIZ];
+
+	memfree = read_meminfo("MemFree:");
+	tst_resm(TINFO, "The current MemFree is %luMB", memfree / KB);
+	if (memfree < MB)
+		tst_resm(TWARN, "The system has low memory, which maybe "
+				"cause the case failed");

Hmm, TWARN is most likely wrong here, as it gets propagated into the test return value. If it is known to fail on low memory, I would just exit with TCONF and "Not enough memory for testing" message. Comments anybody?

testcases/kernel/mem/lib/mem.c void ksm_usage(void) · line 631

Quoted 5 lines
testcases/kernel/mem/lib/mem.c +5 -0
+	hugepagesize = read_meminfo("Hugepagesize:");
+
+	pid = malloc(nr_child * sizeof(int));
+	if (pid == NULL)
+		tst_brkm(TBROK | TERRNO, cleanup, "malloc");

This is really small nit, but I would have named this array as 'pids'.

Quoted 24 lines
testcases/kernel/mem/lib/mem.c +24 -0
+	for (i = 0; i < nr_child; i++) {
+		switch (pid[i] = fork()) {
+		case -1:
+			tst_brkm(TBROK | TERRNO, cleanup, "fork");
+
+		case 0:
+			ret = alloc_transparent_hugepages(nr_thps, hg_aligned);
+			exit(ret);
+		}
+	}
+
+	tst_resm(TINFO, "Stop all children...");
+	for (i = 0; i < nr_child; i++) {
+		if (waitpid(pid[i], &status, WUNTRACED) == -1)
+			tst_brkm(TBROK|TERRNO, cleanup, "waitpid");
+		if (!WIFSTOPPED(status))
+			tst_brkm(TBROK, cleanup,
+				 "child[%d] was not stoppted", pid[i]);
+	}
+
+	tst_resm(TINFO, "Start to scan all transparent hugepages...");
+	khugepaged_scan_done();
+
+	tst_resm(TINFO, "Start to verify transparent hugepage size...");

testcases/kernel/mem/lib/mem.c void ksm_usage(void) · line 666

Quoted 10 lines
testcases/kernel/mem/lib/mem.c +10 -0
+	verify_thp_size(pid, nr_child, nr_thps);
+
+	tst_resm(TINFO, "Wake up all children...");
+	for (i = 0; i < nr_child; i++) {
+		if (kill(pid[i], SIGCONT) == -1)
+			tst_brkm(TBROK | TERRNO, cleanup,
+				 "signal continue child[%d]", pid[i]);
+	}
+
+	/* wait all children finish himself task */
                                       ^
				    their tasks
Quoted 126 lines
testcases/kernel/mem/lib/mem.c +24 -0
+	for (i = 0; i < nr_child; i++) {
+		if (waitpid(pid[i], &status, 0) == -1)
+			tst_brkm(TBROK|TERRNO, cleanup, "waitpid %d", pid[i]);
+
+		if (!WIFEXITED(status) || WEXITSTATUS(status) != 0)
+			tst_resm(TFAIL, "the child[%d] unexpectedly failed:"
+				 " %d", pid[i], status);
+	}
+}
+
+void check_thp_options(int *nr_child, int *nr_thps)
+{
+	if (opt_nr_child)
+		*nr_child = SAFE_STRTOL(NULL, opt_nr_child_str, 0, LONG_MAX);
+	if (opt_nr_thps)
+		*nr_thps = SAFE_STRTOL(NULL, opt_nr_thps_str, 0, LONG_MAX);
+}
+
+void thp_usage(void)
+{
+	printf("  -n      Number of processes\n");
+	printf("  -N      Number of transparent hugepages\n");
+}
+
 /* cpuset/memcg */
 
 static void gather_node_cpus(char *cpus, long nd)
testcases/kernel/mem/thp/thp04.c new file +93 -0
new file mode 100644
index 0000000..0f6c553
@@ -0,0 +1,136 @@
+/*
+ * Copyright (C) 2013 Linux Test Project
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it would be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+ *
+ * Further, this software is distributed without any warranty that it
+ * is free of the rightful claim of any third person regarding
+ * infringement or the like.  Any license provided herein, whether
+ * implied or otherwise, applies only to this software file.  Patent
+ * licenses, if any, provided herein do not apply to combinations of
+ * this program with other software, or any other product whatsoever.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA.
+ */
+
+/*
+ * The case is designed to test the functionality of transparent
+ * hugepage - THP
+ *
+ * when one process allocate hugepage aligned anonymouse pages,
+ * kernel thread 'khugepaged' controlled by sysfs knobs
+ * /sys/kernel/mm/transparent_hugepage/ will scan them, and make
+ * them as transparent hugepage if they are suited, you can find out
+ * how many transparent hugepages are there in one process from
+ * /proc/<pid>/smaps, among the file contents, 'AnonHugePages' entry
+ * stand for transparent hugepage.
+ */
+
+#include <sys/types.h>
+#include <sys/mman.h>
+#include <sys/stat.h>
+#include <sys/wait.h>
+#include <errno.h>
+#include <fcntl.h>
+#include <signal.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <unistd.h>
+#include "test.h"
+#include "usctest.h"
+#include "mem.h"
+
+char *TCID = "thp04";
+int TST_TOTAL = 1;
+
+option_t thp_options[] = {
+	{"n:", &opt_nr_child, &opt_nr_child_str},
+	{"N:", &opt_nr_thps, &opt_nr_thps_str},
+	{NULL, NULL, NULL}
+};
+
+static int pre_thp_scan_sleep_millisecs;
+static int pre_thp_alloc_sleep_millisecs;
+static char pre_thp_enabled[BUFSIZ];
+
+int main(int argc, char *argv[])
+{
+	int lc;
+	char *msg;
+	int nr_child = 2, nr_thps = 64;
+
+	msg = parse_opts(argc, argv, thp_options, thp_usage);
+	if (msg != NULL)
+		tst_brkm(TBROK, tst_exit, "OPTION PARSING ERROR - %s", msg);
+	check_thp_options(&nr_child, &nr_thps);
+
+	setup();
+
+	tst_resm(TINFO, "Start to test transparent hugepage...");
+	tst_resm(TINFO, "There are %d children allocating %d "
+			"transparent hugepages", nr_child, nr_thps);
+
+	for (lc = 0; TEST_LOOPING(lc); lc++) {
+		tst_count = 0;
+
+		test_transparent_hugepage(nr_child, nr_thps, 1);
+	}
+
+	cleanup();
+	tst_exit();
+}
+
+void setup(void)

testcases/kernel/mem/thp/thp04.c line 103

Quoted 10 lines
testcases/kernel/mem/thp/thp04.c new file +10 -0
+{
+	char path[BUFSIZ];
+
+	tst_require_root(NULL);
+
+	if (access(PATH_THP, F_OK) == -1)
+		tst_brkm(TCONF, NULL, "THP is not enabled");
+
+	snprintf(path, BUFSIZ, PATH_KHPD "scan_sleep_millisecs");
+	SAFE_FILE_SCANF(NULL, path, "%d", &pre_thp_scan_sleep_millisecs);

The PATH_KPHD and "scan_sleep_millisecs" are both static strings and are merged before they are passed to the snprintf() at compile time, which results in printf format string __without__ any characters to substitute. This is really weird.

testcases/kernel/mem/thp/thp04.c line 105

Quoted 2 lines
testcases/kernel/mem/thp/thp04.c new file +2 -0
+	/* set 0 to khugepaged/scan_sleep_millisecs to run khugepaged 100% */
+	SAFE_FILE_PRINTF(cleanup, path, "%d", 0);

I'm confused here, this is the only function in the setup that takes a pointer to the cleanup defined bellow.

testcases/kernel/mem/thp/thp04.c line 115

Quoted 9 lines
testcases/kernel/mem/thp/thp04.c new file +9 -0
+	snprintf(path, BUFSIZ, PATH_KHPD "alloc_sleep_millisecs");
+	SAFE_FILE_SCANF(NULL, path, "%d", &pre_thp_alloc_sleep_millisecs);
+	/*
+	 * set 0 to khugepaged/alloc_sleep_millisecs to make sure khugepaged
+	 * don't stop if there's a hugepage allcation failure.
+	 */
+	SAFE_FILE_PRINTF(NULL, path, "%d", 0);
+
+	snprintf(path, BUFSIZ, PATH_THP "enabled");

Here as well.

testcases/kernel/mem/thp/thp04.c line 116

Quoted 1 line
testcases/kernel/mem/thp/thp04.c new file +1 -0
+	write_file(path, "always");

Where this came from? Does the function check for failures? Why not to use SAFE_FILE_PRINTF() as the rest of the file?

Quoted 7 lines
testcases/kernel/mem/thp/thp04.c new file +7 -0
+	tst_sig(FORK, DEF_HANDLER, NULL);
+	TEST_PAUSE;
+}
+
+void cleanup(void)
+{
+	char path[BUFSIZ];

testcases/kernel/mem/thp/thp04.c line 134

Quoted 10 lines
testcases/kernel/mem/thp/thp04.c new file +10 -0
+
+	snprintf(path, BUFSIZ, PATH_KHPD "scan_sleep_millisecs");
+	SAFE_FILE_PRINTF(NULL, path, "%d", pre_thp_scan_sleep_millisecs);
+
+	snprintf(path, BUFSIZ, PATH_KHPD "alloc_sleep_millisecs");
+	SAFE_FILE_PRINTF(NULL, path, "%d", pre_thp_alloc_sleep_millisecs);
+
+	snprintf(path, BUFSIZ, PATH_THP "enabled");
+	write_file(path, pre_thp_enabled);
+

Same here.

Quoted 16 lines
testcases/kernel/mem/thp/thp04.c new file +2 -0
+	TEST_CLEANUP;
+}
Signature · 13 lines
1.7.11.7


------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list
Signature · 13 lines
Cyril Hrubis
[email protected]

------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

A message that is not in the archive (sent off-list or not yet received): [email protected]

#7 lore raw

Re: [LTP] [PATCH v4 1/3] mm/thp: new case thp04.c

Hi!

Quoted 28 lines
Quoted 28 lines
Quoted 16 lines
testcases/kernel/mem/lib/mem.c +16 -0
+		 * or not.
+		 */
+		sleep(10);
+		count++;
+
+		SAFE_FILE_SCANF(cleanup, PATH_KHPD "pages_collapsed",
+			       "%ld", &pages_collapsed);
+		SAFE_FILE_SCANF(cleanup, PATH_KHPD "max_ptes_none",
+			       "%ld", &max_ptes_none);
+		SAFE_FILE_SCANF(cleanup, PATH_KHPD "pages_to_scan",
+			       "%ld", &pages_to_scan);
+
+		if (pages_collapsed != old_pages_collapsed ||
+		    max_ptes_none != old_max_ptes_none ||
+		    pages_to_scan != old_pages_to_scan) {
+			old_pages_collapsed = pages_collapsed;

testcases/kernel/mem/lib/mem.c void ksm_usage(void) · line 581

Quoted 10 lines
testcases/kernel/mem/lib/mem.c +10 -0
+			old_max_ptes_none = max_ptes_none;
+			old_pages_to_scan = pages_to_scan;
+		} else {
+			changing = 0;
+		}
+	}
+
+	tst_resm(TINFO, "khugepaged daemon takes %ds to scan all thp pages",
+		 count * 10);
+}

Ah, so it acutally does polling, this is fine.

Quoted 6 lines
Quoted 4 lines

How long scanning takes. If it is about 10 seconds, we should poll faster so that the testcase has chance to finish reasonably fast. (if it takes 10 seconds and we miss the first windows it will sleep
doing noting for another ten seconds)

so how about 5 seconds? I will test the value on different systems.

If the scan takes 10 seconds to finish I would go for 2 seconds sleep as that can increase the duration maximally by 20%, which is not as bad as 100% or 50%. The overhead of opening and reading a file should be really small.

Quoted 15 lines
Quoted 3 lines

testcases/kernel/mem/thp/thp04.c line 116

Quoted 1 line
testcases/kernel/mem/thp/thp04.c new file +1 -0
+	write_file(path, "always");

Where this came from? Does the function check for failures? Why not to use SAFE_FILE_PRINTF() as the rest of the file?

yes, there's some confused issue here.

but I have tried to use SAFE_FILE_PRINTF, but it didn't work fine for transparent_hugepage/enabled,
if # cat /sys/kernel/mm/transparent_hugepage/enabled
always [madvise] never
SAFE_FILE_PRINTF only get 'always', expected 'madvise'
if # cat /sys/kernel/mm/transparent_hugepage/enabled [always] madvise never
SAFE_FILE_PRINTF get '[always]', expected 'always'

Quoted 2 lines

I didn't find the reason why SAFE_FILE_PRINTF couldn't get the expected value?

That is strange, the only difference is that SAFE_FILE_PRINTF() uses FILE interface while write_file() uses file descriptor directly. Otherwise they should do exactly the same in this case. If they don't something is broken.

Do you have a snipped of code that misbehaves this way?

Signature · 13 lines
Cyril Hrubis
[email protected]

------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list