From nobody Tue Feb 10 19:16:33 2026 Received: from out-171.mta1.migadu.com (out-171.mta1.migadu.com [95.215.58.171]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 284301DA617 for ; Mon, 28 Apr 2025 03:36:58 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=95.215.58.171 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1745811421; cv=none; b=u53Vf4Rla0irDT9G8OQnpkwnWIb6jW41hYOkLfu6lNtGo+LgnWawQiPcU2o2pXeOXF13QXfoq4Csib9cwb+2RIpMdTd64jH4z62fig1/qy0Evd41HqjdMq9uvsmCYXRe0VR6I1OPtPFqKxew6NJ3u1Nlqgid3oUPb/RDwr+nRgU= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1745811421; c=relaxed/simple; bh=Y2/ZjSBqhtmstk1n4jW6jLq1ecahkiy7PdRXoDx12zI=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=GsXe5Lftawbs6ViOld4LX/6D6XVDqF6kbTwZLHdAh302WoszMvwLV2q4KHIvHXrJeIy2h6TLh3Biq09Tjx++MAHV0h3DiIghCcMW0mJ+RDftlGm60KJ1MrtOXmMDAN/M73uwVbJKbqzusQkiM3j89fKhqu7dxm8/u/t0FB1BN4Y= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev; spf=pass smtp.mailfrom=linux.dev; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b=b4qM1b0v; arc=none smtp.client-ip=95.215.58.171 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=linux.dev Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b="b4qM1b0v" X-Report-Abuse: Please report any abuse attempt to abuse@migadu.com and include these headers. DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linux.dev; s=key1; t=1745811417; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=jDg9cWJDtBRFaLTipYCw/9zZK5hSUuCOB7afaAZtBZU=; b=b4qM1b0vHDaVKMDQkiweM3oerzIpfr+kT4u23fRV2Aj7X2MD0OkhU9s8ac2wjFSF/r0x5x uyPxww42H+ZN2hHtpOstR51Iv3G7vORPb5dxhqBlHfHO1OTT0Tfxwa7Qg/GgBrFPVV2/p6 BhQtCsd6CgUO88vCscH04YN+7nozXf0= From: Roman Gushchin To: linux-kernel@vger.kernel.org Cc: Andrew Morton , Alexei Starovoitov , Johannes Weiner , Michal Hocko , Shakeel Butt , Suren Baghdasaryan , David Rientjes , Josh Don , Chuyi Zhou , cgroups@vger.kernel.org, linux-mm@kvack.org, bpf@vger.kernel.org, Roman Gushchin Subject: [PATCH rfc 08/12] bpf: selftests: bpf OOM handler test Date: Mon, 28 Apr 2025 03:36:13 +0000 Message-ID: <20250428033617.3797686-9-roman.gushchin@linux.dev> In-Reply-To: <20250428033617.3797686-1-roman.gushchin@linux.dev> References: <20250428033617.3797686-1-roman.gushchin@linux.dev> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Migadu-Flow: FLOW_OUT Content-Type: text/plain; charset="utf-8" Implement a pseudo-realistic test for the OOM handling functionality. The OOM handling policy which is implemented in bpf is to kill all tasks belonging to the biggest leaf cgroup, which doesn't contain unkillable tasks (tasks with oom_score_adj set to -1000). Pagecache size is excluded from the accounting. The test creates a hierarchy of memory cgroups, causes an OOM at the top level, checks that the expected process will be killed and checks memcg's oom statistics. Signed-off-by: Roman Gushchin --- tools/testing/selftests/bpf/prog_tests/oom.c | 227 +++++++++++++++++++ tools/testing/selftests/bpf/progs/test_oom.c | 103 +++++++++ 2 files changed, 330 insertions(+) create mode 100644 tools/testing/selftests/bpf/prog_tests/oom.c create mode 100644 tools/testing/selftests/bpf/progs/test_oom.c diff --git a/tools/testing/selftests/bpf/prog_tests/oom.c b/tools/testing/s= elftests/bpf/prog_tests/oom.c new file mode 100644 index 000000000000..224c25334385 --- /dev/null +++ b/tools/testing/selftests/bpf/prog_tests/oom.c @@ -0,0 +1,227 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define _GNU_SOURCE + +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include "cgroup_helpers.h" +#include "test_oom.skel.h" + +struct cgroup_desc { + const char *path; + int fd; + unsigned long long id; + int pid; + size_t target; + size_t max; + int oom_score_adj; + bool victim; +}; + +#define MB (1024 * 1024) +#define OOM_SCORE_ADJ_MIN (-1000) +#define OOM_SCORE_ADJ_MAX 1000 + +static struct cgroup_desc cgroups[] =3D { + { .path =3D "/oom_test", .max =3D 80 * MB}, + { .path =3D "/oom_test/cg1", .target =3D 10 * MB, + .oom_score_adj =3D OOM_SCORE_ADJ_MAX }, + { .path =3D "/oom_test/cg2", .target =3D 40 * MB, + .oom_score_adj =3D OOM_SCORE_ADJ_MIN }, + { .path =3D "/oom_test/cg3" }, + { .path =3D "/oom_test/cg3/cg4", .target =3D 30 * MB, + .victim =3D true }, + { .path =3D "/oom_test/cg3/cg5", .target =3D 20 * MB }, +}; + +static int spawn_task(struct cgroup_desc *desc) +{ + char *ptr; + int pid; + + pid =3D fork(); + if (pid < 0) + return pid; + + if (pid > 0) { + /* parent */ + desc->pid =3D pid; + return 0; + } + + /* child */ + if (desc->oom_score_adj) { + char buf[64]; + int fd =3D open("/proc/self/oom_score_adj", O_WRONLY); + + if (fd < 0) + return -1; + + snprintf(buf, sizeof(buf), "%d", desc->oom_score_adj); + write(fd, buf, sizeof(buf)); + close(fd); + } + + ptr =3D (char *)malloc(desc->target); + if (!ptr) + return -ENOMEM; + + memset(ptr, 'a', desc->target); + + while (1) + sleep(1000); + + return 0; +} + +static void setup_environment(void) +{ + int i, err; + + err =3D setup_cgroup_environment(); + if (!ASSERT_OK(err, "setup_cgroup_environment")) + goto cleanup; + + for (i =3D 0; i < ARRAY_SIZE(cgroups); i++) { + cgroups[i].fd =3D create_and_get_cgroup(cgroups[i].path); + if (!ASSERT_GE(cgroups[i].fd, 0, "create_and_get_cgroup")) + goto cleanup; + + cgroups[i].id =3D get_cgroup_id(cgroups[i].path); + if (!ASSERT_GT(cgroups[i].id, 0, "get_cgroup_id")) + goto cleanup; + + if (i =3D=3D 0) { + /* Freeze the top-level cgroup */ + err =3D write_cgroup_file(cgroups[i].path, "cgroup.freeze", "1"); + if (!ASSERT_OK(err, "freeze cgroup")) + goto cleanup; + } + + if (!cgroups[i].target) { + /* Recursively enable the memory controller */ + err =3D write_cgroup_file(cgroups[i].path, "cgroup.subtree_control", + "+memory"); + if (!ASSERT_OK(err, "enable memory controller")) + goto cleanup; + } + + if (cgroups[i].max) { + char buf[256]; + + snprintf(buf, sizeof(buf), "%lu", cgroups[i].max); + err =3D write_cgroup_file(cgroups[i].path, "memory.max", buf); + if (!ASSERT_OK(err, "set memory.max")) + goto cleanup; + + snprintf(buf, sizeof(buf), "0"); + write_cgroup_file(cgroups[i].path, "memory.swap.max", buf); + + } + + if (cgroups[i].target) { + char buf[256]; + + err =3D spawn_task(&cgroups[i]); + if (!ASSERT_OK(err, "spawn task")) + goto cleanup; + + snprintf(buf, sizeof(buf), "%d", cgroups[i].pid); + err =3D write_cgroup_file(cgroups[i].path, "cgroup.procs", buf); + if (!ASSERT_OK(err, "put child into a cgroup")) + goto cleanup; + } + } + + return; + +cleanup: + cleanup_cgroup_environment(); +} + +static int run_and_wait_for_oom(void) +{ + int ret =3D -1; + bool first =3D true; + char buf[4096] =3D {}; + size_t size; + + ret =3D write_cgroup_file(cgroups[0].path, "cgroup.freeze", "0"); + if (!ASSERT_OK(ret, "freeze cgroup")) + return -1; + + for (;;) { + int i, status; + pid_t pid =3D wait(&status); + + if (pid =3D=3D -1) { + if (errno =3D=3D EINTR) + continue; + /* ECHILD */ + break; + } + + if (!first) + continue; + + first =3D false; + + for (i =3D 0; i < ARRAY_SIZE(cgroups); i++) { + if (!ASSERT_OK(cgroups[i].victim !=3D + (pid =3D=3D cgroups[i].pid), + "correct process was killed")) { + ret =3D -1; + break; + } + + if (!cgroups[i].victim) + continue; + + size =3D read_cgroup_file(cgroups[i].path, + "memory.events", + buf, sizeof(buf)); + if (!ASSERT_OK(size <=3D 0, "read memory.events")) { + ret =3D -1; + break; + } + + if (!ASSERT_OK(strstr(buf, "oom_kill 1") =3D=3D NULL, + "oom_kill count check")) { + ret =3D -1; + break; + } + } + + for (i =3D 0; i < ARRAY_SIZE(cgroups); i++) + if (cgroups[i].pid && cgroups[i].pid !=3D pid) + kill(cgroups[i].pid, SIGKILL); + } + + return ret; +} + +void test_oom(void) +{ + struct test_oom *skel; + int err; + + skel =3D test_oom__open_and_load(); + err =3D test_oom__attach(skel); + if (!ASSERT_OK(err, "test_oom__attach")) + goto cleanup; + + setup_environment(); + + run_and_wait_for_oom(); + + cleanup_cgroup_environment(); +cleanup: + test_oom__destroy(skel); +} diff --git a/tools/testing/selftests/bpf/progs/test_oom.c b/tools/testing/s= elftests/bpf/progs/test_oom.c new file mode 100644 index 000000000000..a8224d7c3fed --- /dev/null +++ b/tools/testing/selftests/bpf/progs/test_oom.c @@ -0,0 +1,103 @@ +// SPDX-License-Identifier: GPL-2.0-only +#include "vmlinux.h" +#include +#include +#include "bpf_misc.h" +#include "bpf_experimental.h" + +char _license[] SEC("license") =3D "GPL"; + +#define OOM_SCORE_ADJ_MIN (-1000) + +void bpf_rcu_read_lock(void) __ksym; +void bpf_rcu_read_unlock(void) __ksym; +struct task_struct *bpf_task_acquire(struct task_struct *p) __ksym; +void bpf_task_release(struct task_struct *p) __ksym; +struct mem_cgroup *bpf_get_root_mem_cgroup(void) __ksym; +struct mem_cgroup *bpf_get_mem_cgroup(struct cgroup_subsys_state *css) __k= sym; +void bpf_put_mem_cgroup(struct mem_cgroup *memcg) __ksym; +int bpf_oom_kill_process(struct oom_control *oc, struct task_struct *task, + const char *message__str) __ksym; + +bool mem_cgroup_killable(struct mem_cgroup *memcg) +{ + struct task_struct *task; + bool ret =3D true; + + bpf_for_each(css_task, task, &memcg->css, CSS_TASK_ITER_PROCS) + if (task->signal->oom_score_adj =3D=3D OOM_SCORE_ADJ_MIN) + return false; + + return ret; +} + +/* + * Find the largest leaf cgroup (ignoring page cache) without unkillable t= asks + * and kill all belonging tasks. + */ +SEC("fmod_ret.s/bpf_handle_out_of_memory") +int BPF_PROG(test_bpf_out_of_memory, struct oom_control *oc) +{ + struct task_struct *task; + struct mem_cgroup *root_memcg =3D oc->memcg; + struct mem_cgroup *memcg, *victim =3D NULL; + struct cgroup_subsys_state *css_pos; + unsigned long usage, max_usage =3D 0; + unsigned long pagecache =3D 0; + int ret =3D 0; + + if (root_memcg) + root_memcg =3D bpf_get_mem_cgroup(&root_memcg->css); + else + root_memcg =3D bpf_get_root_mem_cgroup(); + + if (!root_memcg) + return 0; + + bpf_rcu_read_lock(); + bpf_for_each(css, css_pos, &root_memcg->css, BPF_CGROUP_ITER_DESCENDANTS_= POST) { + if (css_pos->cgroup->nr_descendants + css_pos->cgroup->nr_dying_descenda= nts) + continue; + + memcg =3D bpf_get_mem_cgroup(css_pos); + if (!memcg) + continue; + + usage =3D bpf_mem_cgroup_usage(memcg, false); + pagecache =3D bpf_mem_cgroup_page_state(memcg, NR_FILE_PAGES); + + if (usage > pagecache) + usage -=3D pagecache; + else + usage =3D 0; + + if ((usage > max_usage) && mem_cgroup_killable(memcg)) { + max_usage =3D usage; + if (victim) + bpf_put_mem_cgroup(victim); + victim =3D bpf_get_mem_cgroup(&memcg->css); + } + + bpf_put_mem_cgroup(memcg); + } + bpf_rcu_read_unlock(); + + if (!victim) + goto exit; + + bpf_for_each(css_task, task, &victim->css, CSS_TASK_ITER_PROCS) { + struct task_struct *t =3D bpf_task_acquire(task); + + if (t) { + bpf_oom_kill_process(oc, task, "bpf oom test"); + bpf_task_release(t); + ret =3D 1; + } + } + + bpf_put_mem_cgroup(victim); +exit: + bpf_put_mem_cgroup(root_memcg); + + return ret; +} --=20 2.49.0.901.g37484f566f-goog