Skip to contentSkip to conversation

lib: add new cgroup test API

PATCH Superseded Li Wang 2 participants 15 messages Jun 1, 2020 – Jun 8, 2020 Library code

v2 of 4 Superseded Latest is v4 →

lore
Share & tools
lore

Li Wang

author
#0 lore raw

PATCH v2 1/4 lib: add new cgroup test API

Many of our LTP tests need Control Group in the configuration, this patch makes cgroup unified mounting at setup phase to be possible. The method?is extracted from mem.h with the purpose of?extendible for further developing, and trying?to compatible the current two versions of cgroup,

It's hard to make all APIs be strictly consistent because there are many differences between v1 and v2. But it?capsulate the detail of cgroup mounting in high-level functions, which will be easier to use cgroup without considering too much technical thing.? ?

Btw, test get passed on RHEL7(x86_64), RHEL8(ppc64le), Fedora32(x86_64).

---

Notes:
v1 --> v2
* export tst_cgroup*_path, tst_cgroup_ctl_knob string to user
* add a shared list to store the cgroup mount path
* add parameter for multiple cgroups mounting support
* add tst_cgroup_set_knob() function
* add tst_cgroup_mem_swapacct_enabled() function
* remove the tst_cgroup_move_current() from lib invoke

 doc/test-writing-guidelines.txt |  44 +++++
 include/tst_cgroup.h            |  38 ++++
 include/tst_test.h              |   1 +
 lib/newlib_tests/test21.c       |  97 +++++++++
 lib/tst_cgroup.c                | 338 ++++++++++++++++++++++++++++++++
 5 files changed, 518 insertions(+)
 create mode 100644 include/tst_cgroup.h
 create mode 100644 lib/newlib_tests/test21.c
 create mode 100644 lib/tst_cgroup.c
doc/test-writing-guidelines.txt +44 -0
index 93ca506d9..bc450f11a 100644
@@ -2053,6 +2053,50 @@ the prefix field of file pointed by path equals to the value passed to this func
 Also having a similar api pair TST_ASSERT_FILE_INT/STR(path, prefix, val) to assert
 the field value of file.
 
+2.2.36 Using Control Group
+^^^^^^^^^^^^^^^^^^^^^^^^^^
+Some of LTP tests need Control Group in their configuration, tst_cgroup.h provides
+APIs to make cgroup unified mounting at setup phase to be possible. The method?is
+extracted from mem.h with the purpose of?extendible for further developing, and
+trying to compatible the current two versions of cgroup.
+
+Considering there are many differences between cgroup v1 and v2, here we capsulate
+the detail of cgroup mounting in high-level functions, which will be easier to use
+cgroup without caring about too much technical thing.? ?
+
+Also, we do cgroup mount/umount work for the different hierarchy automatically.
+
+[source,c]
+-------------------------------------------------------------------------------
+#include "tst_test.h"
+
+static void run(void)
+{
+	...
+
+	tst_cgroup_move_current(PATH_TMP_CG1_MEM);
+	tst_cgroup_mem_set_maxbytes(PATH_TMP_CG1_MEM, MEMSIZE);
+
+	// do test under cgroup
+	...
+}
+
+static void setup(void)
+{
+	tst_cgroup_mount(TST_CGROUP_MEMCG, PATH_TMP_CG1_MEM);
+}
+
+static void cleanup(void)
+{
+	tst_cgroup_umount(PATH_TMP_CG1_MEM);
+}
+
+struct tst_test test = {
+	.test_all = run,
+	...
+};
+
+-------------------------------------------------------------------------------
 2.3 Writing a testcase in shell
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
include/tst_cgroup.h new file +38 -0
new file mode 100644
index 000000000..bd813b7ec
@@ -0,0 +1,38 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (c) 2020 Red Hat, Inc.
+ */
+
+#ifndef TST_CGROUP_H
+#define TST_CGROUP_H
+
+#define PATH_SYS_CGROUP		"/sys/fs/cgroup"
+#define PATH_TMP_CG1_MEM	"/tmp/cg1_mem"
+#define PATH_TMP_CG1_CST	"/tmp/cg1_cst"
+#define PATH_TMP_CGROUP2	"/tmp/cgroup2" /* cgroup v2 has only single hierarchy */
+
+enum tst_cgroup_ver {
+	TST_CGROUP_V1 = 1,
+	TST_CGROUP_V2 = 2,
+};
+
+enum tst_cgroup_ctrl {
+	TST_CGROUP_MEMCG = 1,
+	TST_CGROUP_CPUSET = 2,
+	/* add cgroup controller */
+};
+
+extern char tst_cgroup_mnt_path[PATH_MAX];
+extern char tst_cgroup_new_path[PATH_MAX];
+extern char tst_cgroup_ctl_knob[PATH_MAX];
+
+enum tst_cgroup_ver tst_cgroup_version(void);
+void tst_cgroup_mount(enum tst_cgroup_ctrl ctrl, const char *cgroup_dir);
+void tst_cgroup_umount(const char *cgroup_dir);
+void tst_cgroup_set_knob(const char *cgroup_dir, const char *knob, long value);
+void tst_cgroup_move_current(const char *cgroup_dir);
+void tst_cgroup_mem_set_maxbytes(const char *cgroup_dir, long memsz);
+int tst_cgroup_mem_swapacct_enabled(const char *cgroup_dir);
+void tst_cgroup_mem_set_maxswap(const char *cgroup_dir, long memsz);
+
+#endif /* TST_CGROUP_H */
include/tst_test.h +1 -0
index 5bedb4f6f..b84f7b9dd 100644
@@ -39,6 +39,7 @@
 #include "tst_capability.h"
 #include "tst_hugepage.h"
 #include "tst_assert.h"
+#include "tst_cgroup.h"
 
 /*
  * Reports testcase result.
lib/newlib_tests/test21.c new file +97 -0

97 lines not on this page · raw message

lib/tst_cgroup.c new file +338 -0

338 lines not on this page · raw message

Signature · 1 line
2.21.1
author
#1 lore raw

PATCH v2 1/4 lib: add new cgroup test API

Li Wang <[email protected]> wrote:

...

Quoted 8 lines
include/tst_cgroup.h new file +4 -0
+#define PATH_SYS_CGROUP                "/sys/fs/cgroup"
+#define PATH_TMP_CG1_MEM       "/tmp/cg1_mem"
+#define PATH_TMP_CG1_CST       "/tmp/cg1_cst"
+#define PATH_TMP_CGROUP2       "/tmp/cgroup2" /* cgroup v2 has only

single hierarchy */
+

Seems there is no need to distinguish CG1 and CG2 mount directory in tst_cgroup.h, because we have introduced the 'tst_cgroup_mnt_path' to mount cgroup unified, it will choose cgroup versions automatically.

So maybe we can change these above definitions to:

-#define PATH_SYS_CGROUP "/sys/fs/cgroup"

-#define PATH_TMP_CG1_MEM       "/tmp/cg1_mem"
-#define PATH_TMP_CG1_CST       "/tmp/cg1_cst"
-#define PATH_TMP_CGROUP2       "/tmp/cgroup2" /* cgroup v2 has only single

hierarchy */
+#define PATH_TMP_CGROUP_MEM "/tmp/cgroup_mem"
+#define PATH_TMP_CGROUP_CST "/tmp/cgroup_cst"

Signature · 5 lines
Regards,
Li Wang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20200601/8375be37/attachment.htm>
author
#2 lore raw

PATCH v2 2/4 mem: take use of new cgroup API

In reply to Li Wang

For issue #611:
https://github.com/linux-test-project/ltp/issues/611

---
 testcases/kernel/mem/cpuset/cpuset01.c | 19 +++++++------
 testcases/kernel/mem/include/mem.h     |  1 +
 testcases/kernel/mem/ksm/ksm02.c       |  8 +++---
 testcases/kernel/mem/ksm/ksm03.c       | 12 +++-----
 testcases/kernel/mem/ksm/ksm04.c       | 16 +++++------
 testcases/kernel/mem/lib/mem.c         |  8 ++----
 testcases/kernel/mem/oom/oom03.c       | 22 ++++-----------
 testcases/kernel/mem/oom/oom04.c       | 12 ++++----
 testcases/kernel/mem/oom/oom05.c       | 38 ++++++++------------------
 9 files changed, 52 insertions(+), 84 deletions(-)
testcases/kernel/mem/cpuset/cpuset01.c +10 -9
index 853f7fe55..00b79cf3d 100644
@@ -51,11 +51,11 @@ static void test_cpuset(void)
 	unsigned long nmask[MAXNODES / BITS_PER_LONG] = { 0 };
 	char mems[BUFSIZ], buf[BUFSIZ];
 
-	read_cpuset_files(CPATH, "cpus", buf);
-	write_cpuset_files(CPATH_NEW, "cpus", buf);
-	read_cpuset_files(CPATH, "mems", mems);
-	write_cpuset_files(CPATH_NEW, "mems", mems);
-	SAFE_FILE_PRINTF(CPATH_NEW "/tasks", "%d", getpid());
+	tst_cgroup_move_current(PATH_TMP_CG1_CST);
+	read_cpuset_files(tst_cgroup_mnt_path, "cpus", buf);
+	write_cpuset_files(tst_cgroup_new_path, "cpus", buf);
+	read_cpuset_files(tst_cgroup_mnt_path, "mems", mems);
+	write_cpuset_files(tst_cgroup_new_path, "mems", mems);
 
 	child = SAFE_FORK();
 	if (child == 0) {
@@ -70,9 +70,9 @@ static void test_cpuset(void)
 	}
 
 	snprintf(buf, BUFSIZ, "%d", nodes[0]);
-	write_cpuset_files(CPATH_NEW, "mems", buf);
+	write_cpuset_files(tst_cgroup_new_path, "mems", buf);
 	snprintf(buf, BUFSIZ, "%d", nodes[1]);
-	write_cpuset_files(CPATH_NEW, "mems", buf);
+	write_cpuset_files(tst_cgroup_new_path, "mems", buf);
 
 	SAFE_WAITPID(child, &status, WUNTRACED | WCONTINUED);
 	if (WEXITSTATUS(status) != 0) {
@@ -85,7 +85,7 @@ static void test_cpuset(void)
 
 static void setup(void)
 {
-	mount_mem("cpuset", "cpuset", NULL, CPATH, CPATH_NEW);
+	tst_cgroup_mount(TST_CGROUP_CPUSET, PATH_TMP_CG1_CST);
 	ncpus = count_cpu();
 	if (get_allowed_nodes_arr(NH_MEMS | NH_CPUS, &nnodes, &nodes) < 0)
 		tst_brk(TBROK | TERRNO, "get_allowed_nodes_arr");
@@ -95,7 +95,7 @@ static void setup(void)
 
 static void cleanup(void)
 {
-	umount_mem(CPATH, CPATH_NEW);
+	tst_cgroup_umount(PATH_TMP_CG1_CST);
 }
 
 static void sighandler(int signo LTP_ATTRIBUTE_UNUSED)
@@ -183,6 +183,7 @@ static long count_cpu(void)
 
 static struct tst_test test = {
 	.needs_root = 1,
+	.forks_child = 1,
 	.setup = setup,
 	.cleanup = cleanup,
 	.test_all = test_cpuset,
testcases/kernel/mem/include/mem.h +1 -0
index cce9c0497..f553651c2 100644
@@ -2,6 +2,7 @@
 #define _MEM_H
 #include "config.h"
 #include "tst_test.h"
+#include "tst_cgroup.h"
 #include "ksm_helper.h"
 
 #if defined(__powerpc__) || defined(__powerpc64__)
testcases/kernel/mem/ksm/ksm02.c +4 -4
index 3d4c19deb..d680db247 100644
@@ -66,6 +66,8 @@ static void verify_ksm(void)
 	unsigned long nmask[MAXNODES / BITS_PER_LONG] = { 0 };
 	unsigned int node;
 
+	tst_cgroup_move_current(PATH_TMP_CG1_CST);
+
 	node = get_a_numa_node();
 	set_node(nmask, node);
 
@@ -88,8 +90,7 @@ static void cleanup(void)
 		FILE_PRINTF(PATH_KSM "merge_across_nodes",
 				 "%d", merge_across_nodes);
 
-	if (cpuset_mounted)
-		umount_mem(CPATH, CPATH_NEW);
+	tst_cgroup_umount(PATH_TMP_CG1_CST);
 }
 
 static void setup(void)
@@ -105,8 +106,7 @@ static void setup(void)
 		SAFE_FILE_PRINTF(PATH_KSM "merge_across_nodes", "1");
 	}
 
-	mount_mem("cpuset", "cpuset", NULL, CPATH, CPATH_NEW);
-	cpuset_mounted = 1;
+	tst_cgroup_mount(TST_CGROUP_CPUSET, PATH_TMP_CG1_CST);
 }
 
 static struct tst_test test = {
testcases/kernel/mem/ksm/ksm03.c +4 -8
index 0e5131654..649203595 100644
@@ -59,11 +59,10 @@
 #include "mem.h"
 #include "ksm_common.h"
 
-static int memcg_mounted;
-
 static void verify_ksm(void)
 {
-	write_memcg();
+	tst_cgroup_move_current(PATH_TMP_CG1_MEM);
+	tst_cgroup_mem_set_maxbytes(PATH_TMP_CG1_MEM, TESTMEM);
 	create_same_memory(size, num, unit);
 }
 
@@ -79,8 +78,7 @@ static void setup(void)
 	}
 
 	parse_ksm_options(opt_sizestr, &size, opt_numstr, &num, opt_unitstr, &unit);
-	mount_mem("memcg", "cgroup", "memory", MEMCG_PATH, MEMCG_PATH_NEW);
-	memcg_mounted = 1;
+	tst_cgroup_mount(TST_CGROUP_MEMCG, PATH_TMP_CG1_MEM);
 }
 
 static void cleanup(void)
@@ -88,9 +86,7 @@ static void cleanup(void)
 	if (access(PATH_KSM "merge_across_nodes", F_OK) == 0)
 		FILE_PRINTF(PATH_KSM "merge_across_nodes",
 				 "%d", merge_across_nodes);
-
-	if (memcg_mounted)
-		umount_mem(MEMCG_PATH, MEMCG_PATH_NEW);
+	tst_cgroup_umount(PATH_TMP_CG1_MEM);
 }
 
 static struct tst_test test = {
testcases/kernel/mem/ksm/ksm04.c +7 -9
index e393dbd40..abd3e0978 100644
@@ -70,7 +70,8 @@ static void verify_ksm(void)
 	node = get_a_numa_node();
 	set_node(nmask, node);
 
-	write_memcg();
+	tst_cgroup_move_current(PATH_TMP_CG1_MEM);
+	tst_cgroup_mem_set_maxbytes(PATH_TMP_CG1_MEM, TESTMEM);
 
 	if (set_mempolicy(MPOL_BIND, nmask, MAXNODES) == -1) {
 		if (errno != ENOSYS)
@@ -81,6 +82,7 @@ static void verify_ksm(void)
 	}
 	create_same_memory(size, num, unit);
 
+	tst_cgroup_move_current(PATH_TMP_CG1_CST);
 	write_cpusets(node);
 	create_same_memory(size, num, unit);
 }
@@ -91,10 +93,8 @@ static void cleanup(void)
 		FILE_PRINTF(PATH_KSM "merge_across_nodes",
 				 "%d", merge_across_nodes);
 
-	if (cpuset_mounted)
-		umount_mem(CPATH, CPATH_NEW);
-	if (memcg_mounted)
-		umount_mem(MEMCG_PATH, MEMCG_PATH_NEW);
+	tst_cgroup_umount(PATH_TMP_CG1_MEM);
+	tst_cgroup_umount(PATH_TMP_CG1_CST);
 }
 
 static void setup(void)
@@ -110,10 +110,8 @@ static void setup(void)
 
 	parse_ksm_options(opt_sizestr, &size, opt_numstr, &num, opt_unitstr, &unit);
 
-	mount_mem("cpuset", "cpuset", NULL, CPATH, CPATH_NEW);
-	cpuset_mounted = 1;
-	mount_mem("memcg", "cgroup", "memory", MEMCG_PATH, MEMCG_PATH_NEW);
-	memcg_mounted = 1;
+	tst_cgroup_mount(TST_CGROUP_MEMCG, PATH_TMP_CG1_MEM);
+	tst_cgroup_mount(TST_CGROUP_CPUSET, PATH_TMP_CG1_CST);
 }
 
 static struct tst_test test = {
testcases/kernel/mem/lib/mem.c +3 -5
index eca4c61c8..45f155922 100644
@@ -692,7 +692,7 @@ void write_cpusets(long nd)
 	char cpus[BUFSIZ] = "";
 
 	snprintf(buf, BUFSIZ, "%ld", nd);
-	write_cpuset_files(CPATH_NEW, "mems", buf);
+	write_cpuset_files(tst_cgroup_new_path, "mems", buf);
 
 	gather_node_cpus(cpus, nd);
 	/*
@@ -701,14 +701,12 @@ void write_cpusets(long nd)
 	 * the value of cpuset.cpus.
 	 */
 	if (strlen(cpus) != 0) {
-		write_cpuset_files(CPATH_NEW, "cpus", cpus);
+		write_cpuset_files(tst_cgroup_new_path, "cpus", cpus);
 	} else {
 		tst_res(TINFO, "No CPUs in the node%ld; "
 				"using only CPU0", nd);
-		write_cpuset_files(CPATH_NEW, "cpus", "0");
+		write_cpuset_files(tst_cgroup_new_path, "cpus", "0");
 	}
-
-	SAFE_FILE_PRINTF(CPATH_NEW "/tasks", "%d", getpid());
 }
 
 void umount_mem(char *path, char *path_new)
testcases/kernel/mem/oom/oom03.c +6 -16
index ce0b34c31..150824003 100644
@@ -36,27 +36,19 @@
 
 #ifdef HAVE_NUMA_V2
 
-static int memcg_mounted;
-
 static void verify_oom(void)
 {
 #ifdef TST_ABI32
 	tst_brk(TCONF, "test is not designed for 32-bit system.");
 #endif
 
-	SAFE_FILE_PRINTF(MEMCG_PATH_NEW "/tasks", "%d", getpid());
-	SAFE_FILE_PRINTF(MEMCG_LIMIT, "%ld", TESTMEM);
+	tst_cgroup_move_current(PATH_TMP_CG1_MEM);
+	tst_cgroup_mem_set_maxbytes(PATH_TMP_CG1_MEM, TESTMEM);
 
 	testoom(0, 0, ENOMEM, 1);
 
-	if (access(MEMCG_SW_LIMIT, F_OK) == -1) {
-		if (errno == ENOENT)
-			tst_res(TCONF,
-				"memcg swap accounting is disabled");
-		else
-			tst_brk(TBROK | TERRNO, "access");
-	} else {
-		SAFE_FILE_PRINTF(MEMCG_SW_LIMIT, "%ld", TESTMEM);
+	if (tst_cgroup_mem_swapacct_enabled(PATH_TMP_CG1_MEM)) {
+		tst_cgroup_mem_set_maxswap(PATH_TMP_CG1_MEM, TESTMEM);
 		testoom(0, 1, ENOMEM, 1);
 	}
 
@@ -73,16 +65,14 @@ static void setup(void)
 {
 	overcommit = get_sys_tune("overcommit_memory");
 	set_sys_tune("overcommit_memory", 1, 1);
-	mount_mem("memcg", "cgroup", "memory", MEMCG_PATH, MEMCG_PATH_NEW);
-	memcg_mounted = 1;
+	tst_cgroup_mount(TST_CGROUP_MEMCG, PATH_TMP_CG1_MEM);
 }
 
 static void cleanup(void)
 {
 	if (overcommit != -1)
 		set_sys_tune("overcommit_memory", overcommit, 0);
-	if (memcg_mounted)
-		umount_mem(MEMCG_PATH, MEMCG_PATH_NEW);
+	tst_cgroup_umount(PATH_TMP_CG1_MEM);
 }
 
 static struct tst_test test = {
testcases/kernel/mem/oom/oom04.c +5 -7
index 7185ef973..bff3faf97 100644
@@ -36,14 +36,14 @@
 
 #ifdef HAVE_NUMA_V2
 
-static int cpuset_mounted;
-
 static void verify_oom(void)
 {
 #ifdef TST_ABI32
 	tst_brk(TCONF, "test is not designed for 32-bit system.");
 #endif
 
+	tst_cgroup_move_current(PATH_TMP_CG1_CST);
+
 	tst_res(TINFO, "OOM on CPUSET...");
 	testoom(0, 0, ENOMEM, 1);
 
@@ -53,7 +53,7 @@ static void verify_oom(void)
 		 * is in charge of cpuset.memory_migrate, we can write
 		 * 1 to cpuset.memory_migrate to enable the migration.
 		 */
-		write_cpuset_files(CPATH_NEW, "memory_migrate", "1");
+		write_cpuset_files(tst_cgroup_new_path, "memory_migrate", "1");
 		tst_res(TINFO, "OOM on CPUSET with mem migrate:");
 		testoom(0, 0, ENOMEM, 1);
 	}
@@ -69,8 +69,7 @@ static void setup(void)
 	overcommit = get_sys_tune("overcommit_memory");
 	set_sys_tune("overcommit_memory", 1, 1);
 
-	mount_mem("cpuset", "cpuset", NULL, CPATH, CPATH_NEW);
-	cpuset_mounted = 1;
+	tst_cgroup_mount(TST_CGROUP_CPUSET, PATH_TMP_CG1_CST);
 
 	/*
 	 * Some nodes do not contain memory, so use
@@ -89,8 +88,7 @@ static void cleanup(void)
 {
 	if (overcommit != -1)
 		set_sys_tune("overcommit_memory", overcommit, 0);
-	if (cpuset_mounted)
-		umount_mem(CPATH, CPATH_NEW);
+	tst_cgroup_umount(PATH_TMP_CG1_CST);
 }
 
 static struct tst_test test = {
testcases/kernel/mem/oom/oom05.c +12 -26
index db24df6de..aacbbc220 100644
@@ -36,18 +36,16 @@
 
 #ifdef HAVE_NUMA_V2
 
-static int memcg_mounted;
-static int cpuset_mounted;
-
 static void verify_oom(void)
 {
-	int swap_acc_on = 1;
-
 #ifdef TST_ABI32
 	tst_brk(TCONF, "test is not designed for 32-bit system.");
 #endif
 
 	tst_res(TINFO, "OOM on CPUSET & MEMCG...");
+	tst_cgroup_move_current(PATH_TMP_CG1_MEM);
+	tst_cgroup_move_current(PATH_TMP_CG1_CST);
+	tst_cgroup_mem_set_maxbytes(PATH_TMP_CG1_MEM, TESTMEM);
 	testoom(0, 0, ENOMEM, 1);
 
 	/*
@@ -56,29 +54,22 @@ static void verify_oom(void)
 	 * 1 to cpuset.memory_migrate to enable the migration.
 	 */
 	if (is_numa(NULL, NH_MEMS, 2)) {
-		write_cpuset_files(CPATH_NEW, "memory_migrate", "1");
+		tst_cgroup_move_current(PATH_TMP_CG1_CST);
+		write_cpuset_files(tst_cgroup_new_path, "memory_migrate", "1");
 		tst_res(TINFO, "OOM on CPUSET & MEMCG with "
 				"cpuset.memory_migrate=1");
 		testoom(0, 0, ENOMEM, 1);
 	}
 
-	if (access(MEMCG_SW_LIMIT, F_OK) == -1) {
-		if (errno == ENOENT) {
-			tst_res(TCONF, "memcg swap accounting is disabled");
-			swap_acc_on = 0;
-		} else
-			tst_brk(TBROK|TERRNO, "access");
-	}
-
-	if (swap_acc_on) {
+	if (tst_cgroup_mem_swapacct_enabled(PATH_TMP_CG1_MEM)) {
 		tst_res(TINFO, "OOM on CPUSET & MEMCG with "
 				"special memswap limitation:");
-		SAFE_FILE_PRINTF(MEMCG_SW_LIMIT, "%ld", TESTMEM);
+		tst_cgroup_mem_set_maxswap(PATH_TMP_CG1_MEM, TESTMEM);
 		testoom(0, 0, ENOMEM, 1);
 
 		tst_res(TINFO, "OOM on CPUSET & MEMCG with "
 				"disabled memswap limitation:");
-		SAFE_FILE_PRINTF(MEMCG_SW_LIMIT, "-1");
+		tst_cgroup_mem_set_maxswap(PATH_TMP_CG1_MEM, -1);
 		testoom(0, 0, ENOMEM, 1);
 	}
 }
@@ -93,11 +84,8 @@ void setup(void)
 	overcommit = get_sys_tune("overcommit_memory");
 	set_sys_tune("overcommit_memory", 1, 1);
 
-	mount_mem("memcg", "cgroup", "memory", MEMCG_PATH, MEMCG_PATH_NEW);
-	memcg_mounted = 1;
-	mount_mem("cpuset", "cpuset", NULL, CPATH, CPATH_NEW);
-	cpuset_mounted = 1;
-	write_memcg();
+	tst_cgroup_mount(TST_CGROUP_MEMCG, PATH_TMP_CG1_MEM);
+	tst_cgroup_mount(TST_CGROUP_CPUSET, PATH_TMP_CG1_CST);
 
 	/*
 	 * Some nodes do not contain memory, so use
@@ -116,10 +104,8 @@ void cleanup(void)
 {
 	if (overcommit != -1)
 		set_sys_tune("overcommit_memory", overcommit, 0);
-	if (cpuset_mounted)
-		umount_mem(CPATH, CPATH_NEW);
-	if (memcg_mounted)
-		umount_mem(MEMCG_PATH, MEMCG_PATH_NEW);
+	tst_cgroup_umount(PATH_TMP_CG1_MEM);
+	tst_cgroup_umount(PATH_TMP_CG1_CST);
 }
 
 static struct tst_test test = {
Signature · 1 line
2.21.1
author
#3 lore raw

PATCH v2 3/4 mem: remove the old cgroup API

In reply to Li Wang
---
 testcases/kernel/mem/include/mem.h | 16 +--------
 testcases/kernel/mem/lib/mem.c     | 58 ------------------------------
 2 files changed, 1 insertion(+), 73 deletions(-)
testcases/kernel/mem/include/mem.h +1 -15
index f553651c2..0e089fae2 100644
@@ -60,24 +60,10 @@ void test_ksm_merge_across_nodes(unsigned long nr_pages);
 void check_hugepage(void);
 void write_memcg(void);
 
-/* cpuset/memcg */
-
-#define CPATH			"/dev/cpuset"
-#define CPATH_NEW		CPATH "/1"
-#define MEMCG_PATH		"/dev/cgroup"
-#define MEMCG_PATH_NEW		MEMCG_PATH "/1"
-#define MEMCG_LIMIT		MEMCG_PATH_NEW "/memory.limit_in_bytes"
-#define MEMCG_SW_LIMIT		MEMCG_PATH_NEW "/memory.memsw.limit_in_bytes"
-#if HAVE_SYS_EVENTFD_H
-#define PATH_OOMCTRL		MEMCG_PATH_NEW "/memory.oom_control"
-#define PATH_EVTCTRL		MEMCG_PATH_NEW "/cgroup.event_control"
-#endif
-
+/* cpuset/memcg - include/tst_cgroup.h */
 void read_cpuset_files(char *prefix, char *filename, char *retbuf);
 void write_cpuset_files(char *prefix, char *filename, char *buf);
 void write_cpusets(long nd);
-void mount_mem(char *name, char *fs, char *options, char *path, char *path_new);
-void umount_mem(char *path, char *path_new);
 
 /* shared */
 unsigned int get_a_numa_node(void);
testcases/kernel/mem/lib/mem.c +0 -58

70 lines not on this page · raw message

Signature · 1 line
2.21.1
author
#4 lore raw

PATCH v2 4/4 mm: add cpuset01 to runtest file

In reply to Li Wang

This test was moved to stress test since commit ae8fa55a8, but stress runtest has been removed from commit e752f7c19674 :).

Here adding back it as a general memory test without '-I 3600'.

---
 runtest/mm | 2 ++
 1 file changed, 2 insertions(+)
runtest/mm +2 -0
index a09f39c1e..612a4d066 100644
@@ -73,6 +73,8 @@ ksm06 ksm06
 ksm06_1 ksm06 -n 10
 ksm06_2 ksm06 -n 10000
 
+cpuset01 cpuset01
+
 oom01 oom01
 oom02 oom02
 oom03 oom03
Signature · 1 line
2.21.1
#5 lore raw

PATCH v2 1/4 lib: add new cgroup test API

In reply to Li Wang

----- Original Message -----

doc/test-writing-guidelines.txt the prefix field of file pointed by path equals to the value passed to this func · line 2078

Quoted 11 lines
doc/test-writing-guidelines.txt +11 -0
+
+[source,c]
+-------------------------------------------------------------------------------
+#include "tst_test.h"
+
+static void run(void)
+{
+	...
+
+	tst_cgroup_move_current(PATH_TMP_CG1_MEM);
+	tst_cgroup_mem_set_maxbytes(PATH_TMP_CG1_MEM, MEMSIZE);

Goal for API is to hide differences between cgroup 1/2, but example above is passing cgroup specific directory.

My suggestion was to have directory parameter relative to cgroup mount, I didn't consider there would be need for mounting cgroup more than once from single process. Is there such need?

Since there's only one global 'tst_cgroup_mnt_path', is there need to have paths absolute? If we assume that single process will mount cgroup only once, then all paths could be relative to 'tst_cgroup_mnt_path', and test doesn't need to even use 'tst_cgroup_mnt_path'.

Quoted 4 lines
lib/tst_cgroup.c new file +4 -0
+
+static void tst_cgroup_set_path(const char *cgroup_dir)
+{
+	struct tst_cgroup_path *tst_cgroup_path, *a;

lib/tst_cgroup.c line 173

Quoted 10 lines
lib/tst_cgroup.c new file +10 -0
+
+	if (!cgroup_dir)
+		tst_brk(TBROK, "Invalid cgroup dir, plese check cgroup_dir");
+
+	sprintf(tst_cgroup_mnt_path, "%s", cgroup_dir);
+	sprintf(tst_cgroup_new_path, "%s/ltp_%d", cgroup_dir, rand());
+
+	/* To store cgroup path in the shared 'path' list */
+	tst_cgroup_path = SAFE_MMAP(NULL, (sizeof(struct tst_cgroup_path)),
+			PROT_READ | PROT_WRITE, MAP_ANONYMOUS | MAP_SHARED, -1, 0);

I'm not sure I understand what is the reason to have tst_cgroup_path. Is it expected, that mount and umount are called by different processes? It might be easier to define API as per-process and require same process to call mount and umount.

lib/tst_cgroup.c line 175

Quoted 2 lines
lib/tst_cgroup.c new file +2 -0
+	tst_cgroup_path->mnt_path = SAFE_MALLOC(strlen(tst_cgroup_mnt_path));
+	tst_cgroup_path->new_path = SAFE_MALLOC(strlen(tst_cgroup_new_path));

Pointers are in shared memory, but content they point to is not, so it's accessible only from process that called tst_cgroup_set_path().

Can you describe all different scenarios you wanted to support?

author
#6 lore raw

PATCH v2 1/4 lib: add new cgroup test API

Hi Jan,

Quoted 7 lines · Jan Stancek

----- Original Message -----

Quoted 3 lines

+
+[source,c]

+-------------------------------------------------------------------------------

Quoted 12 lines

doc/test-writing-guidelines.txt the prefix field of file pointed by path equals to the value passed to this func · line 2078

Quoted 8 lines
doc/test-writing-guidelines.txt +8 -0
+#include "tst_test.h"
+
+static void run(void)
+{
+     ...
+
+     tst_cgroup_move_current(PATH_TMP_CG1_MEM);
+     tst_cgroup_mem_set_maxbytes(PATH_TMP_CG1_MEM, MEMSIZE);

Goal for API is to hide differences between cgroup 1/2, but example above is passing cgroup specific directory.

Sorry for the misleading info, actually that's no different to use any directory whatever you pass via parameter, it will recognize cgroup version to be mounted intelligently.

Here as I commented in the last email. the specific dir are typos which inherent from patch v1. We should fix them.

Quoted 3 lines

My suggestion was to have directory parameter relative to cgroup mount,

Patch v2 is already achieved this even includes more functions(i.e mount many same cgroup).

Quoted 3 lines

I didn't consider there would be need for mounting cgroup more than once from single process. Is there such need?

Yes right, the test21.c mounting many cgroups in a single process is just a test
to verify these APIs works fine, it meaningless in real situations.

Quoted 7 lines

Since there's only one global 'tst_cgroup_mnt_path', is there need to have paths absolute? If we assume that single process will mount cgroup only once,
then all paths could be relative to 'tst_cgroup_mnt_path', and test doesn't need to even use 'tst_cgroup_mnt_path'.

No, the global 'tst_cgroup_mnt_path' is not only to pass mount directory but also for storing path when searching from get_cgroup_get_path().

Why we need this? Because, if a testcase(i.e oom05.c) needs more than one cgroup
subsystem(memory, cpuset) on RHEL7(cgroup-v1), it should mount two different
directories and do some knob setting.

Quoted 18 lines

lib/tst_cgroup.c

Quoted 8 lines
lib/tst_cgroup.c new file +7 -0
+
+static void tst_cgroup_set_path(const char *cgroup_dir)
+{
+     struct tst_cgroup_path *tst_cgroup_path, *a;
+
+     if (!cgroup_dir)
+             tst_brk(TBROK, "Invalid cgroup dir, plese check

cgroup_dir");

lib/tst_cgroup.c

Quoted 7 lines
lib/tst_cgroup.c new file +7 -0
+
+     sprintf(tst_cgroup_mnt_path, "%s", cgroup_dir);
+     sprintf(tst_cgroup_new_path, "%s/ltp_%d", cgroup_dir, rand());
+
+     /* To store cgroup path in the shared 'path' list */
+     tst_cgroup_path = SAFE_MMAP(NULL, (sizeof(struct tst_cgroup_path)),
+                     PROT_READ | PROT_WRITE, MAP_ANONYMOUS |

MAP_SHARED, -1, 0);

Quoted 4 lines

I'm not sure I understand what is the reason to have tst_cgroup_path. Is it expected,
that mount and umount are called by different processes? It might be easier

The shared 'tst_cgroup_path' is necessary especially for mounting different cgoups in setup(). Otherwise, it would be easy to get lost which directory for kind of cgroup type.

And the worth to say, the random directory name for same cgroup mounting is also on purpose, though we mount same(i.e memory) cgroup in two places it still belongs to the one hierarchy, and create the same name of the new directory will be hit an error in EEXIST.

static void tst_cgroup_set_path(const char *cgroup_dir)

{
    ...
    sprintf(tst_cgroup_mnt_path, "%s", cgroup_dir);
    sprintf(tst_cgroup_new_path, "%s/ltp_%d", cgroup_dir, rand());
}
Quoted 12 lines

to define API as per-process and require same process to call mount and umount.

Quoted 1 line

+ tst_cgroup_path->mnt_path =

SAFE_MALLOC(strlen(tst_cgroup_mnt_path));

Quoted 1 line

+ tst_cgroup_path->new_path =

SAFE_MALLOC(strlen(tst_cgroup_new_path));

Pointers are in shared memory, but content they point to is not, so it's accessible
only from process that called tst_cgroup_set_path().

Good catch. This should be also use shared memory.

Quoted 3 lines

Can you describe all different scenarios you wanted to support?

Sure, we have to consider many scenarios:

1. mount only one cgroup in a single process on the system support cgroup v1 or v2 (i.e ksm03.c)

#include "tst_cgroup.h"
#define PATH_CGROUP /tmp/cgroup

static void run(void)

{
    tst_cgroup_move_current(PATH_CGROUP);
    tst_cgroup_mem_set_maxbytes(PATH_CGROUP, 1024*1024*1024);
    // do your test
}

static void setup(void)

{
    tst_cgroup_mount(TST_CGROUP_MEMCG, PATH_CGROUP);
}

static void cleanup(void)

{
    tst_cgroup_umount(PATH_CGROUP);
}

static struct tst_test test = {
...
.test_all = run,
};

2. mount different cgroups in a single process on the system only support cgroup-v1 (i.e ksm04.c, oom05.c)

#include "tst_cgroup.h"
#define PATH_CGROUP1 /tmp/cgroup1
#define PATH_CGROUP2 /tmp/cgroup2

static void run(void)

{
    tst_cgroup_move_current(PATH_CGROUP1);
    tst_cgroup_move_current(PATH_CGROUP2);

    // we have to recognize the correct cgroup path if we
    // mount two or more cgroup subsystems in a single
    // process, in case we get lost in knob setting

    // the tst_cgroup_get_path() helps to find and get
    // correct path in tst_cgroup_mnt_path, tst_cgroup_new_path

    // that's also the reason why we need a shared list to
    // store many cgoup pathes. And, this is extendible for
    // mounting more cgroup subsystems in the future.

    tst_cgroup_mem_set_maxbytes(PATH_CGROUP1, 1024*1024*1024);
    // some cpuset configration
}

static void setup(void)

{
    tst_cgroup_mount(TST_CGROUP_MEMCG, PATH_CGROUP1);
    tst_cgroup_mount(TST_CGROUP_CPUSET, PATH_CGROUP2);
}

static void cleanup(void)

{
    tst_cgroup_umount(PATH_CGROUP1);
    tst_cgroup_umount(PATH_CGROUP2);
}

static struct tst_test test = {
...
.test_all = run,
};

3. mount different cgroups in different process on the system support v1 or v2

#include "tst_cgroup.h"
#define PATH_CGROUP1 /tmp/cgroup1
#define PATH_CGROUP2 /tmp/cgroup2

static void run(void)

{
    if (fork() == 0) {
        tst_cgroup_move_current(PATH_CGROUP1);
        tst_cgroup_mem_set_maxbytes(PATH_CGROUP1, 1024*1024*1024);
        // test code
    }

    tst_cgroup_move_current(PATH_CGROUP2);
    // some cpuset configuration
    // and test code
}

static void setup(void)

{
    // we do cgroup mount work unified in setup()
    // whatever the cgroup is being used in the parent
    // or children, there is also no need to care about the
    // details of cgroup v1 or v2.

    tst_cgroup_mount(TST_CGROUP_MEMCG, PATH_CGROUP1);
    tst_cgroup_mount(TST_CGROUP_CPUSET, PATH_CGROUP2);
}

static void cleanup(void)

{
    tst_cgroup_umount(PATH_CGROUP1);
    tst_cgroup_umount(PATH_CGROUP2);
}

static struct tst_test test = {
...
.test_all = run,
};

4. mount same cgroup in different process on the system support v1 or v2

#include "tst_cgroup.h"
#define PATH_CGROUP1 /tmp/cgroup1
#define PATH_CGROUP2 /tmp/cgroup2

static void run(void)

{
    if (fork() == 0) {
        tst_cgroup_move_current(PATH_CGROUP1);
        tst_cgroup_mem_set_maxbytes(PATH_CGROUP1, 1024*1024*1024);
        // test code1
    }

        tst_cgroup_move_current(PATH_CGROUP2);
        tst_cgroup_mem_set_maxbytes(PATH_CGROUP2, 2048*2048);
        // test code2
}

static void setup(void)

{
    // we mount two memory cgroup in the parent
    // but setting different value in corresponding
    // knob for different process to test more configration

    tst_cgroup_mount(TST_CGROUP_MEMCG, PATH_CGROUP1);
    tst_cgroup_mount(TST_CGROUP_MEMCG, PATH_CGROUP2);
}

static void cleanup(void)

{
    tst_cgroup_umount(PATH_CGROUP1);
    tst_cgroup_umount(PATH_CGROUP2);
}

static struct tst_test test = {
...
.test_all = run,
};

5. mount many cgroups on different process for future extendible work i.e.

// tst_cgroup_mount(TST_CGROUP_MEMCG, PATH_CGROUP1); // tst_cgroup_mount(TST_CGROUP_CPUSET, PATH_CGROUP2); // tst_cgroup_mount(TST_CGROUP_CPUACT, PATH_CGROUP3); // tst_cgroup_mount(TST_CGROUP_PIDSCG, PATH_CGROUP4); // tst_cgroup_mount(TST_CGROUP_HUGTLB, PATH_CGROUP5); // ...

Btw, I attach the patch v2.1 for better readable.

Signature · 12 lines
Regards,
Li Wang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20200602/9e622499/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-lib-add-new-cgroup-test-API.patch
Type: text/x-patch
Size: 15933 bytes
Desc: not available
URL: <http://lists.linux.it/pipermail/ltp/attachments/20200602/9e622499/attachment-0001.bin>
#7 lore raw

PATCH v2 1/4 lib: add new cgroup test API

Hi Li,

Quoted 5 lines

Why we need this? Because, if a testcase(i.e oom05.c) needs more than one cgroup
subsystem(memory, cpuset) on RHEL7(cgroup-v1), it should mount two different
directories and do some knob setting.

Mounting with different controllers is fine, I meant do we have a case for mounting same controller multiple times? We might have, because current design allows only for single directory (tst_cgroup_new_path), that's automatically created on mount. (This is your example 4)

Quoted 20 lines
Quoted 20 lines

lib/tst_cgroup.c

Quoted 10 lines
lib/tst_cgroup.c new file +7 -0
+
+static void tst_cgroup_set_path(const char *cgroup_dir)
+{
+     struct tst_cgroup_path *tst_cgroup_path, *a;
+
+     if (!cgroup_dir)
+             tst_brk(TBROK, "Invalid cgroup dir, plese check

cgroup_dir");

lib/tst_cgroup.c

Quoted 7 lines
lib/tst_cgroup.c new file +7 -0
+
+     sprintf(tst_cgroup_mnt_path, "%s", cgroup_dir);
+     sprintf(tst_cgroup_new_path, "%s/ltp_%d", cgroup_dir, rand());
+
+     /* To store cgroup path in the shared 'path' list */
+     tst_cgroup_path = SAFE_MMAP(NULL, (sizeof(struct tst_cgroup_path)),
+                     PROT_READ | PROT_WRITE, MAP_ANONYMOUS |

MAP_SHARED, -1, 0);

Quoted 8 lines
Quoted 4 lines

I'm not sure I understand what is the reason to have tst_cgroup_path. Is it expected,
that mount and umount are called by different processes? It might be easier

The shared 'tst_cgroup_path' is necessary especially for mounting different cgoups in setup(). Otherwise, it would be easy to get lost which directory for kind of cgroup type.

But why is it shared? Is cleanup going to run in different process context? Which one of your examples needs shared memory?

Quoted 11 lines

And the worth to say, the random directory name for same cgroup mounting is also on purpose, though we mount same(i.e memory) cgroup in two places it still belongs to the one hierarchy, and create the same name of the new directory will be hit an error in EEXIST.

static void tst_cgroup_set_path(const char *cgroup_dir)

{
   ...
   sprintf(tst_cgroup_mnt_path, "%s", cgroup_dir);
   sprintf(tst_cgroup_new_path, "%s/ltp_%d", cgroup_dir, rand());

I see why you are tracking this in list, but this exchange of state through global variables does seem bit unclear.

Could we leave "new_path" creation to testcase itself? It would give us more flexibility and we don't have to worry about name collisions. It also avoids need to mount same controller multiple times (example 4 in your reply).

Let's assume this is API:

#include "tst_cgroup.h"
#define MEM_MNT "/tmp/cgroup1"
#define CPUSET_MNT "/tmp/cgroup2"
#define DIR1 "ltp_test_blah_dir1"
#define DIR2 "ltp_test_blah_dir2"
#define DIR3 "ltp_test_blah_dir3"

static void run(void)

{
    if (fork() == 0) {
        tst_cgroup_move_current(MEM_MNT, DIR2);
        // do your test
        exit(0);
    }
    tst_cgroup_move_current(MEM_MNT, DIR1);
    // do your test
}

static void setup(void)

{
    tst_cgroup_mount(TST_CGROUP_MEMCG, MEM_MNT);
    tst_cgroup_mkdir(MEM_MNT, DIR1);
    tst_cgroup_mem_set_maxbytes(MEM_MNT, DIR1, 1*1024*1024);
    tst_cgroup_mkdir(MEM_MNT, DIR2);
    tst_cgroup_mem_set_maxbytes(MEM_MNT, DIR2, 2*1024*1024);
    
    tst_cgroup_mount(TST_CGROUP_CPUSET, CPUSET_MNT);
    tst_cgroup_mkdir(CPUSET_MNT, DIR3);
    tst_cgroup_move_current(CPUSET_MNT, DIR3);
}

static void cleanup(void)

{
    tst_cgroup_umount(MEM_MNT);
    tst_cgroup_umount(CPUSET_MNT);
}

static struct tst_test test = {
...
.test_all = run,
};

On library side we would have a list that tracks all mounts. And every mount record would have list that tracks all mkdir() operations, so we can cleanup anything that test creates. I think tracking per-process would be sufficient.

(without spinning v3) What are your thoughts?

Regards,
Jan

author
#8 lore raw

PATCH v2 1/4 lib: add new cgroup test API

Quoted 43 lines · Jan Stancek

Hi Li,

Quoted 5 lines

Why we need this? Because, if a testcase(i.e oom05.c) needs more than one cgroup
subsystem(memory, cpuset) on RHEL7(cgroup-v1), it should mount two different
directories and do some knob setting.

Mounting with different controllers is fine, I meant do we have a case for mounting
same controller multiple times? We might have, because current design allows
only for single directory (tst_cgroup_new_path), that's automatically created on mount.
(This is your example 4)

Quoted 11 lines
Quoted 11 lines

lib/tst_cgroup.c

Quoted 10 lines
lib/tst_cgroup.c new file +7 -0
+
+static void tst_cgroup_set_path(const char *cgroup_dir)
+{
+     struct tst_cgroup_path *tst_cgroup_path, *a;
+
+     if (!cgroup_dir)
+             tst_brk(TBROK, "Invalid cgroup dir, plese check

cgroup_dir");

Quoted 6 lines
Quoted 6 lines

lib/tst_cgroup.c

Quoted 6 lines
lib/tst_cgroup.c new file +6 -0
+
+     sprintf(tst_cgroup_mnt_path, "%s", cgroup_dir);
+     sprintf(tst_cgroup_new_path, "%s/ltp_%d", cgroup_dir, rand());
+
+     /* To store cgroup path in the shared 'path' list */
+     tst_cgroup_path = SAFE_MMAP(NULL, (sizeof(struct

tst_cgroup_path)),

Quoted 6 lines
Quoted 6 lines
Quoted 1 line

+ PROT_READ | PROT_WRITE, MAP_ANONYMOUS |

MAP_SHARED, -1, 0);

I'm not sure I understand what is the reason to have tst_cgroup_path. Is it expected,
that mount and umount are called by different processes? It might be

easier

Quoted 2 lines
Quoted 1 line
Quoted 7 lines
Quoted 3 lines

The shared 'tst_cgroup_path' is necessary especially for mounting different cgoups in setup(). Otherwise, it would be easy to get lost which directory for kind of cgroup type.

But why is it shared? Is cleanup going to run in different process context? Which one of your examples needs shared memory?

My original thought was that if children want to remove some paths from the list, so this shared memory was made sense. But in the signed-off patch v2, it seems we only do the deletion in cleanup(), so private memory is enough.

Another slight reason to keep using share memory is to avoid the children have a new copy of the list, that will allocate more memory in testing, but that's fine, I have no objection to using private memory.

Quoted 7 lines
Quoted 7 lines

And the worth to say, the random directory name for same cgroup mounting is also on purpose, though we mount same(i.e memory) cgroup in two places it still belongs to the one hierarchy, and create the same name of the new directory will be hit an error in EEXIST.

Quoted 9 lines
Quoted 5 lines

static void tst_cgroup_set_path(const char *cgroup_dir)

{
   ...
   sprintf(tst_cgroup_mnt_path, "%s", cgroup_dir);
   sprintf(tst_cgroup_new_path, "%s/ltp_%d", cgroup_dir, rand());

I see why you are tracking this in list, but this exchange of state through global variables does seem bit unclear.

Yes, that's a compromise to support more usage of the APIs.

Quoted 4 lines

Could we leave "new_path" creation to testcase itself? It would give us more flexibility and we don't have to worry about name collisions.

Hmm, more flexibility sometimes means more complicated to use, I don't want these APIs to become difficult for newbies of LTP.

Quoted 3 lines

It also avoids need to mount same controller multiple times (example 4 in your reply).

To be honest, most situations we have are examples 1 and 2, the others(ex3,4,5) might rarely happen in our test. So maybe leave too much work(i.e "new_path" creation) in testcase is not wise, and I don't want to make usage complicated just for ingratiating the rare situations.

Quoted 22 lines

Let's assume this is API:

#include "tst_cgroup.h"
#define MEM_MNT "/tmp/cgroup1"
#define CPUSET_MNT "/tmp/cgroup2"
#define DIR1 "ltp_test_blah_dir1"
#define DIR2 "ltp_test_blah_dir2"
#define DIR3 "ltp_test_blah_dir3"

static void run(void)

{
    if (fork() == 0) {
        tst_cgroup_move_current(MEM_MNT, DIR2);
        // do your test
        exit(0);
    }
    tst_cgroup_move_current(MEM_MNT, DIR1);
    // do your test
}

static void setup(void)

Quoted 10 lines
{
    tst_cgroup_mount(TST_CGROUP_MEMCG, MEM_MNT);
    tst_cgroup_mkdir(MEM_MNT, DIR1);
    tst_cgroup_mem_set_maxbytes(MEM_MNT, DIR1, 1*1024*1024);
    tst_cgroup_mkdir(MEM_MNT, DIR2);
    tst_cgroup_mem_set_maxbytes(MEM_MNT, DIR2, 2*1024*1024);

    tst_cgroup_mount(TST_CGROUP_CPUSET, CPUSET_MNT);
    tst_cgroup_mkdir(CPUSET_MNT, DIR3);
    tst_cgroup_move_current(CPUSET_MNT, DIR3);

}

Quoted 12 lines

static void cleanup(void)

{
    tst_cgroup_umount(MEM_MNT);
    tst_cgroup_umount(CPUSET_MNT);
}

static struct tst_test test = {
...
.test_all = run,
};

Quoted 6 lines

On library side we would have a list that tracks all mounts. And every mount
record would have list that tracks all mkdir() operations, so we can cleanup anything that test creates. I think tracking per-process would be sufficient.

To compare with my v2, this design leaves simple code in lib, and flexible in usage. The only disadvantage is that people need to define more macros and mkdir() by themself.

I have no strong preference to shift from v2 to this method, or maybe we can try
to combine together and do more optimize work in the lib side.

Quoted 7 lines

(without spinning v3) What are your thoughts?

Regards,
Jan

Signature · 5 lines
Regards,
Li Wang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20200603/83db11d8/attachment.htm>
#9 lore raw

PATCH v2 1/4 lib: add new cgroup test API

Quoted 11 lines
Quoted 11 lines
Quoted 10 lines

And the worth to say, the random directory name for same cgroup mounting is also on purpose, though we mount same(i.e memory) cgroup in two places it still belongs to the one hierarchy, and create the same name of the new directory will be hit an error in EEXIST.

static void tst_cgroup_set_path(const char *cgroup_dir)

{
   ...
   sprintf(tst_cgroup_mnt_path, "%s", cgroup_dir);
   sprintf(tst_cgroup_new_path, "%s/ltp_%d", cgroup_dir, rand());
Quoted 5 lines
Quoted 3 lines

I see why you are tracking this in list, but this exchange of state through global variables does seem bit unclear.

Yes, that's a compromise to support more usage of the APIs.

I don't get why global variables are necessary.
tst_cgroup_mnt_path is always same as cgroup_dir parameter passed to all functions. tst_cgroup_get_path() could return pointer to tst_cgroup_pathes->new_path,
(you just need you call umount before del_path)
tst_cgroup_ctl_knob is set before each use, so it could be local variable too.

Quoted 3 lines

I have no strong preference to shift from v2 to this method, or maybe we can try
to combine together and do more optimize work in the lib side.

Sounds like we could use 3rd opinion.

author
#10 lore raw

PATCH v2 1/4 lib: add new cgroup test API

Quoted 11 lines · Jan Stancek
Quoted 11 lines
Quoted 11 lines
Quoted 10 lines

And the worth to say, the random directory name for same cgroup mounting is also on purpose, though we mount same(i.e memory) cgroup in two places it still belongs to the one hierarchy, and create the same name of the new directory will be hit an error in EEXIST.

static void tst_cgroup_set_path(const char *cgroup_dir)

{
   ...
   sprintf(tst_cgroup_mnt_path, "%s", cgroup_dir);
   sprintf(tst_cgroup_new_path, "%s/ltp_%d", cgroup_dir, rand());
Quoted 9 lines
Quoted 1 line
Quoted 1 line

I see why you are tracking this in list, but this exchange of state

through

Quoted 4 lines
Quoted 2 lines

global variables does seem bit unclear.

Yes, that's a compromise to support more usage of the APIs.

I don't get why global variables are necessary.

The only reason to export them as global variables is to make the legacy read/write_cpuse_files() happy. So that I said it is a compromise.

$ git grep tst_cgroup_new_path
cpuset/cpuset01.c:      write_cpuset_files(tst_cgroup_new_path, "cpus",
buf);
cpuset/cpuset01.c:      write_cpuset_files(tst_cgroup_new_path, "mems",
mems);
cpuset/cpuset01.c:      write_cpuset_files(tst_cgroup_new_path, "mems",
buf);
cpuset/cpuset01.c:      write_cpuset_files(tst_cgroup_new_path, "mems",
buf);
lib/mem.c:      write_cpuset_files(tst_cgroup_new_path, "mems", buf);
lib/mem.c:              write_cpuset_files(tst_cgroup_new_path, "cpus",
cpus);
lib/mem.c:              write_cpuset_files(tst_cgroup_new_path, "cpus",
"0");
oom/oom04.c:            write_cpuset_files(tst_cgroup_new_path,
"memory_migrate", "1");
oom/oom05.c:            write_cpuset_files(tst_cgroup_new_path,
"memory_migrate", "1");
Quoted 7 lines

tst_cgroup_mnt_path is always same as cgroup_dir parameter passed to all functions.
tst_cgroup_get_path() could return pointer to tst_cgroup_pathes->new_path,
(you just need you call umount before del_path)
tst_cgroup_ctl_knob is set before each use, so it could be local variable too.

tst_cgroup_ctl_knob can be localized since no other places need it.

Quoted 7 lines
Quoted 3 lines

I have no strong preference to shift from v2 to this method, or maybe we can try
to combine together and do more optimize work in the lib side.

Sounds like we could use 3rd opinion.

Signature · 5 lines
Regards,
Li Wang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20200603/e64a2a62/attachment.htm>
#11 lore raw

PATCH v2 1/4 lib: add new cgroup test API

Quoted 16 lines · Li Wang
Quoted 2 lines

I don't get why global variables are necessary.

The only reason to export them as global variables is to make the legacy read/write_cpuse_files() happy. So that I said it is a compromise.

$ git grep tst_cgroup_new_path
cpuset/cpuset01.c:      write_cpuset_files(tst_cgroup_new_path, "cpus",
buf);
cpuset/cpuset01.c:      write_cpuset_files(tst_cgroup_new_path, "mems",
mems);
cpuset/cpuset01.c:      write_cpuset_files(tst_cgroup_new_path, "mems",
buf);
cpuset/cpuset01.c:      write_cpuset_files(tst_cgroup_new_path, "mems",
buf);
lib/mem.c:      write_cpuset_files(tst_cgroup_new_path, "mems", buf);
Quoted 8 lines
lib/mem.c:              write_cpuset_files(tst_cgroup_new_path, "cpus",
cpus);
lib/mem.c:              write_cpuset_files(tst_cgroup_new_path, "cpus",
"0");
oom/oom04.c:            write_cpuset_files(tst_cgroup_new_path,
"memory_migrate", "1");
oom/oom05.c:            write_cpuset_files(tst_cgroup_new_path,
"memory_migrate", "1");

What if we provided access to it via API? Would we still need these global variables?

char *tst_cgroup_get_path(const char *cgroup_mnt)

      // return ptr to tst_cgroup_paths->new_path

mount path is always known to test, because it passes it to tst_cgroup_mount(), so it just needs to find out "new path".

Would that satisfy the need of this legacy test?

author
#12 lore raw

PATCH v2 1/4 lib: add new cgroup test API

Quoted 26 lines · Jan Stancek
Quoted 16 lines · Li Wang
Quoted 2 lines

I don't get why global variables are necessary.

The only reason to export them as global variables is to make the legacy read/write_cpuse_files() happy. So that I said it is a compromise.

$ git grep tst_cgroup_new_path
cpuset/cpuset01.c:      write_cpuset_files(tst_cgroup_new_path, "cpus",
buf);
cpuset/cpuset01.c:      write_cpuset_files(tst_cgroup_new_path, "mems",
mems);
cpuset/cpuset01.c:      write_cpuset_files(tst_cgroup_new_path, "mems",
buf);
cpuset/cpuset01.c:      write_cpuset_files(tst_cgroup_new_path, "mems",
buf);
lib/mem.c:      write_cpuset_files(tst_cgroup_new_path, "mems", buf);
Quoted 8 lines
lib/mem.c:              write_cpuset_files(tst_cgroup_new_path, "cpus",
cpus);
lib/mem.c:              write_cpuset_files(tst_cgroup_new_path, "cpus",
"0");
oom/oom04.c:            write_cpuset_files(tst_cgroup_new_path,
"memory_migrate", "1");
oom/oom05.c:            write_cpuset_files(tst_cgroup_new_path,
"memory_migrate", "1");
Quoted 6 lines

What if we provided access to it via API? Would we still need these global variables?

char *tst_cgroup_get_path(const char *cgroup_mnt)

      // return ptr to tst_cgroup_paths->new_path

The series of list operating function are hiding in the library. My thought is
to make the list transparent to users.

In your method, we have to export the tst_cgroup_get_path() as an external function, it stills needs an extra local pointer in testcase to store the got new_path,
it doesn't seem tidier too.

Quoted 5 lines

mount path is always known to test, because it passes it to tst_cgroup_mount(),
so it just needs to find out "new path".

Would that satisfy the need of this legacy test?

How about moving the cpuset legacy code to the library as part of APIs? That'd
help to capsulate all the operation of cgroup control in lib, and people just need
to invoke the related function as what he/she wants.

+void tst_cgroup_cpuset_read_files(const char *cgroup_dir, const char *filename, char *buf);
+void tst_cgroup_cpuset_write_files(const char *cgroup_dir, const char *filename, const char *buf);

Then 'tst_cgroup_new_path' searching work will all located internally. And 'tst_cgroup_ctl_knob' can
be local variable too.

Any other comments? (attach the v3.1)

Signature · 12 lines
Regards,
Li Wang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20200608/dcd2e1ac/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-lib-add-new-cgroup-test-API.patch
Type: text/x-patch
Size: 17454 bytes
Desc: not available
URL: <http://lists.linux.it/pipermail/ltp/attachments/20200608/dcd2e1ac/attachment-0001.bin>
#13 lore raw

PATCH v2 1/4 lib: add new cgroup test API

----- Original Message -----

Quoted 35 lines · Jan Stancek
Quoted 27 lines · Li Wang
Quoted 16 lines
Quoted 2 lines

I don't get why global variables are necessary.

The only reason to export them as global variables is to make the legacy read/write_cpuse_files() happy. So that I said it is a compromise.

$ git grep tst_cgroup_new_path
cpuset/cpuset01.c:      write_cpuset_files(tst_cgroup_new_path, "cpus",
buf);
cpuset/cpuset01.c:      write_cpuset_files(tst_cgroup_new_path, "mems",
mems);
cpuset/cpuset01.c:      write_cpuset_files(tst_cgroup_new_path, "mems",
buf);
cpuset/cpuset01.c:      write_cpuset_files(tst_cgroup_new_path, "mems",
buf);
lib/mem.c:      write_cpuset_files(tst_cgroup_new_path, "mems", buf);
Quoted 8 lines
lib/mem.c:              write_cpuset_files(tst_cgroup_new_path, "cpus",
cpus);
lib/mem.c:              write_cpuset_files(tst_cgroup_new_path, "cpus",
"0");
oom/oom04.c:            write_cpuset_files(tst_cgroup_new_path,
"memory_migrate", "1");
oom/oom05.c:            write_cpuset_files(tst_cgroup_new_path,
"memory_migrate", "1");
Quoted 6 lines

What if we provided access to it via API? Would we still need these global variables?

char *tst_cgroup_get_path(const char *cgroup_mnt)

      // return ptr to tst_cgroup_paths->new_path
Quoted 8 lines

The series of list operating function are hiding in the library. My thought is
to make the list transparent to users.

In your method, we have to export the tst_cgroup_get_path() as an external function, it stills needs an extra local pointer in testcase to store the got new_path,
it doesn't seem tidier too.

But there would be clear connection between function and variable.
new_path = tst_cgroup_get_path(cgroup_dir);
vs.
tst_cgroup_get_path(cgroup_dir);
// fyi, tst_cgroup_new_path is updated as side-effect of call above // What other calls do update tst_cgroup_new_path? Have a look at implementation.

Quoted 20 lines
Quoted 7 lines

mount path is always known to test, because it passes it to tst_cgroup_mount(),
so it just needs to find out "new path".

Would that satisfy the need of this legacy test?

How about moving the cpuset legacy code to the library as part of APIs? That'd
help to capsulate all the operation of cgroup control in lib, and people just need
to invoke the related function as what he/she wants.

+void tst_cgroup_cpuset_read_files(const char *cgroup_dir, const char *filename, char *buf);
+void tst_cgroup_cpuset_write_files(const char *cgroup_dir, const char *filename, const char *buf);

Quoted 5 lines

Then 'tst_cgroup_new_path' searching work will all located internally. And 'tst_cgroup_ctl_knob' can
be local variable too.

Any other comments? (attach the v3.1)

That makes it somewhat better, since it's only concern of library code now. But since there are no tests using "tst_cgroup_new_path", does it still need to be global variable?

author
#14 lore raw

PATCH v2 1/4 lib: add new cgroup test API

Quoted 5 lines · Jan Stancek

...

But there would be clear connection between function and variable.
new_path = tst_cgroup_get_path(cgroup_dir);

Yes, that's easier to understand the code workflow.

Quoted 5 lines

vs.
tst_cgroup_get_path(cgroup_dir);
// fyi, tst_cgroup_new_path is updated as side-effect of call above // What other calls do update tst_cgroup_new_path? Have a look at
implementation.

...

Quoted 5 lines

That makes it somewhat better, since it's only concern of library code now. But since there are no tests using "tst_cgroup_new_path", does it still need to be global variable?

Sure, we could delete tst_cgorup_mnt/new_path variables, since we have tracking all mount paths in the list, so it will be easy to find new_path in each
function.

Hope v4 is the final version:).

Signature · 5 lines
Regards,
Li Wang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20200608/5ecc6f74/attachment.htm>