- Kernel - mailweb.openeuler.org

[PATCH OLK-5.10 0/5] Add network relationship for NUMA isolation and consolidation
by Liu Jian 24 May '24

24 May '24

Add network relationship for NUMA isolation and consolidation. Eric Dumazet (1): net: initialize net->net_cookie at netns setup Liu Jian (3): net: fix kabi breakage in struct net net: add one bpf prog type for network numa relationship net: add some bpf hooks in tcp stack for network numa relationship Martynas Pumputis (1): net: retrieve netns cookie via getsocketopt arch/alpha/include/uapi/asm/socket.h | 2 + arch/mips/include/uapi/asm/socket.h | 2 + arch/parisc/include/uapi/asm/socket.h | 2 + arch/sparc/include/uapi/asm/socket.h | 2 + include/linux/bpf_types.h | 4 + include/linux/filter.h | 67 ++++++ include/linux/skbuff.h | 4 + include/net/net_namespace.h | 8 +- include/net/net_rship.h | 329 ++++++++++++++++++++++++++ include/net/sock.h | 4 + include/uapi/asm-generic/socket.h | 2 + include/uapi/linux/bpf.h | 23 ++ init/Kconfig | 1 + kernel/bpf/syscall.c | 19 ++ net/Kconfig | 6 + net/core/dev.c | 4 + net/core/filter.c | 218 ++++++++++++++++- net/core/net_namespace.c | 19 +- net/core/skbuff.c | 17 +- net/core/sock.c | 28 +++ net/core/sysctl_net_core.c | 18 ++ net/ipv4/tcp.c | 6 + net/ipv4/tcp_output.c | 3 + tools/bpf/bpftool/prog.c | 5 + tools/include/uapi/linux/bpf.h | 27 +++ tools/lib/bpf/libbpf.c | 8 + tools/lib/bpf/libbpf_probes.c | 1 + 27 files changed, 804 insertions(+), 25 deletions(-) create mode 100644 include/net/net_rship.h -- 2.34.1

2 6

[PATCH openEuler-22.03-LTS-SP1] media: vidtv: psi: Add check for kstrdup
by Cai Xinchen 24 May '24

24 May '24

From: Jiasheng Jiang <jiasheng(a)iscas.ac.cn> stable inclusion from stable-v5.10.201 commit 3387490c89b10aeb4e71d78b65dbc9ba4b2385b9 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I9RFIR CVE: CVE-2023-52844 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- [ Upstream commit 76a2c5df6ca8bd8ada45e953b8c72b746f42918d ] Add check for the return value of kstrdup() and return the error if it fails in order to avoid NULL pointer dereference. Fixes: 7a7899f6f58e ("media: vidtv: psi: Implement an Event Information Table (EIT)") Fixes: c2f78f0cb294 ("media: vidtv: psi: add a Network Information Table (NIT)") Fixes: f90cf6079bf6 ("media: vidtv: add a bridge driver") Signed-off-by: Jiasheng Jiang <jiasheng(a)iscas.ac.cn> Signed-off-by: Hans Verkuil <hverkuil-cisco(a)xs4all.nl> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Cai Xinchen <caixinchen1(a)huawei.com> --- drivers/media/test-drivers/vidtv/vidtv_psi.c | 45 +++++++++++++++++--- 1 file changed, 40 insertions(+), 5 deletions(-) diff --git a/drivers/media/test-drivers/vidtv/vidtv_psi.c b/drivers/media/test-drivers/vidtv/vidtv_psi.c index 1724bb485e67..1726e76f0106 100644 --- a/drivers/media/test-drivers/vidtv/vidtv_psi.c +++ b/drivers/media/test-drivers/vidtv/vidtv_psi.c @@ -308,16 +308,29 @@ struct vidtv_psi_desc_service *vidtv_psi_service_desc_init(struct vidtv_psi_desc desc->service_name_len = service_name_len; - if (service_name && service_name_len) + if (service_name && service_name_len) { desc->service_name = kstrdup(service_name, GFP_KERNEL); + if (!desc->service_name) + goto free_desc; + } desc->provider_name_len = provider_name_len; - if (provider_name && provider_name_len) + if (provider_name && provider_name_len) { desc->provider_name = kstrdup(provider_name, GFP_KERNEL); + if (!desc->provider_name) + goto free_desc_service_name; + } vidtv_psi_desc_chain(head, (struct vidtv_psi_desc *)desc); return desc; + +free_desc_service_name: + if (service_name && service_name_len) + kfree(desc->service_name); +free_desc: + kfree(desc); + return NULL; } struct vidtv_psi_desc_registration @@ -362,8 +375,13 @@ struct vidtv_psi_desc_network_name desc->length = network_name_len; - if (network_name && network_name_len) + if (network_name && network_name_len) { desc->network_name = kstrdup(network_name, GFP_KERNEL); + if (!desc->network_name) { + kfree(desc); + return NULL; + } + } vidtv_psi_desc_chain(head, (struct vidtv_psi_desc *)desc); return desc; @@ -449,15 +467,32 @@ struct vidtv_psi_desc_short_event iso_language_code = "eng"; desc->iso_language_code = kstrdup(iso_language_code, GFP_KERNEL); + if (!desc->iso_language_code) + goto free_desc; - if (event_name && event_name_len) + if (event_name && event_name_len) { desc->event_name = kstrdup(event_name, GFP_KERNEL); + if (!desc->event_name) + goto free_desc_language_code; + } - if (text && text_len) + if (text && text_len) { desc->text = kstrdup(text, GFP_KERNEL); + if (!desc->text) + goto free_desc_event_name; + } vidtv_psi_desc_chain(head, (struct vidtv_psi_desc *)desc); return desc; + +free_desc_event_name: + if (event_name && event_name_len) + kfree(desc->event_name); +free_desc_language_code: + kfree(desc->iso_language_code); +free_desc: + kfree(desc); + return NULL; } struct vidtv_psi_desc *vidtv_psi_desc_clone(struct vidtv_psi_desc *desc) -- 2.34.1

2 1

[PATCH openEuler-22.03-LTS-SP2] media: vidtv: psi: Add check for kstrdup
by Cai Xinchen 24 May '24

24 May '24

From: Jiasheng Jiang <jiasheng(a)iscas.ac.cn> stable inclusion from stable-v5.10.201 commit 3387490c89b10aeb4e71d78b65dbc9ba4b2385b9 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I9RFIR CVE: CVE-2023-52844 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- [ Upstream commit 76a2c5df6ca8bd8ada45e953b8c72b746f42918d ] Add check for the return value of kstrdup() and return the error if it fails in order to avoid NULL pointer dereference. Fixes: 7a7899f6f58e ("media: vidtv: psi: Implement an Event Information Table (EIT)") Fixes: c2f78f0cb294 ("media: vidtv: psi: add a Network Information Table (NIT)") Fixes: f90cf6079bf6 ("media: vidtv: add a bridge driver") Signed-off-by: Jiasheng Jiang <jiasheng(a)iscas.ac.cn> Signed-off-by: Hans Verkuil <hverkuil-cisco(a)xs4all.nl> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Cai Xinchen <caixinchen1(a)huawei.com> --- drivers/media/test-drivers/vidtv/vidtv_psi.c | 45 +++++++++++++++++--- 1 file changed, 40 insertions(+), 5 deletions(-) diff --git a/drivers/media/test-drivers/vidtv/vidtv_psi.c b/drivers/media/test-drivers/vidtv/vidtv_psi.c index 1724bb485e67..1726e76f0106 100644 --- a/drivers/media/test-drivers/vidtv/vidtv_psi.c +++ b/drivers/media/test-drivers/vidtv/vidtv_psi.c @@ -308,16 +308,29 @@ struct vidtv_psi_desc_service *vidtv_psi_service_desc_init(struct vidtv_psi_desc desc->service_name_len = service_name_len; - if (service_name && service_name_len) + if (service_name && service_name_len) { desc->service_name = kstrdup(service_name, GFP_KERNEL); + if (!desc->service_name) + goto free_desc; + } desc->provider_name_len = provider_name_len; - if (provider_name && provider_name_len) + if (provider_name && provider_name_len) { desc->provider_name = kstrdup(provider_name, GFP_KERNEL); + if (!desc->provider_name) + goto free_desc_service_name; + } vidtv_psi_desc_chain(head, (struct vidtv_psi_desc *)desc); return desc; + +free_desc_service_name: + if (service_name && service_name_len) + kfree(desc->service_name); +free_desc: + kfree(desc); + return NULL; } struct vidtv_psi_desc_registration @@ -362,8 +375,13 @@ struct vidtv_psi_desc_network_name desc->length = network_name_len; - if (network_name && network_name_len) + if (network_name && network_name_len) { desc->network_name = kstrdup(network_name, GFP_KERNEL); + if (!desc->network_name) { + kfree(desc); + return NULL; + } + } vidtv_psi_desc_chain(head, (struct vidtv_psi_desc *)desc); return desc; @@ -449,15 +467,32 @@ struct vidtv_psi_desc_short_event iso_language_code = "eng"; desc->iso_language_code = kstrdup(iso_language_code, GFP_KERNEL); + if (!desc->iso_language_code) + goto free_desc; - if (event_name && event_name_len) + if (event_name && event_name_len) { desc->event_name = kstrdup(event_name, GFP_KERNEL); + if (!desc->event_name) + goto free_desc_language_code; + } - if (text && text_len) + if (text && text_len) { desc->text = kstrdup(text, GFP_KERNEL); + if (!desc->text) + goto free_desc_event_name; + } vidtv_psi_desc_chain(head, (struct vidtv_psi_desc *)desc); return desc; + +free_desc_event_name: + if (event_name && event_name_len) + kfree(desc->event_name); +free_desc_language_code: + kfree(desc->iso_language_code); +free_desc: + kfree(desc); + return NULL; } struct vidtv_psi_desc *vidtv_psi_desc_clone(struct vidtv_psi_desc *desc) -- 2.34.1

2 1

[PATCH openEuler-22.03-LTS] media: vidtv: psi: Add check for kstrdup
by Cai Xinchen 24 May '24

24 May '24

From: Jiasheng Jiang <jiasheng(a)iscas.ac.cn> stable inclusion from stable-v5.10.201 commit 3387490c89b10aeb4e71d78b65dbc9ba4b2385b9 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I9RFIR CVE: CVE-2023-52844 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- [ Upstream commit 76a2c5df6ca8bd8ada45e953b8c72b746f42918d ] Add check for the return value of kstrdup() and return the error if it fails in order to avoid NULL pointer dereference. Fixes: 7a7899f6f58e ("media: vidtv: psi: Implement an Event Information Table (EIT)") Fixes: c2f78f0cb294 ("media: vidtv: psi: add a Network Information Table (NIT)") Fixes: f90cf6079bf6 ("media: vidtv: add a bridge driver") Signed-off-by: Jiasheng Jiang <jiasheng(a)iscas.ac.cn> Signed-off-by: Hans Verkuil <hverkuil-cisco(a)xs4all.nl> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Cai Xinchen <caixinchen1(a)huawei.com> --- drivers/media/test-drivers/vidtv/vidtv_psi.c | 45 +++++++++++++++++--- 1 file changed, 40 insertions(+), 5 deletions(-) diff --git a/drivers/media/test-drivers/vidtv/vidtv_psi.c b/drivers/media/test-drivers/vidtv/vidtv_psi.c index 1724bb485e67..1726e76f0106 100644 --- a/drivers/media/test-drivers/vidtv/vidtv_psi.c +++ b/drivers/media/test-drivers/vidtv/vidtv_psi.c @@ -308,16 +308,29 @@ struct vidtv_psi_desc_service *vidtv_psi_service_desc_init(struct vidtv_psi_desc desc->service_name_len = service_name_len; - if (service_name && service_name_len) + if (service_name && service_name_len) { desc->service_name = kstrdup(service_name, GFP_KERNEL); + if (!desc->service_name) + goto free_desc; + } desc->provider_name_len = provider_name_len; - if (provider_name && provider_name_len) + if (provider_name && provider_name_len) { desc->provider_name = kstrdup(provider_name, GFP_KERNEL); + if (!desc->provider_name) + goto free_desc_service_name; + } vidtv_psi_desc_chain(head, (struct vidtv_psi_desc *)desc); return desc; + +free_desc_service_name: + if (service_name && service_name_len) + kfree(desc->service_name); +free_desc: + kfree(desc); + return NULL; } struct vidtv_psi_desc_registration @@ -362,8 +375,13 @@ struct vidtv_psi_desc_network_name desc->length = network_name_len; - if (network_name && network_name_len) + if (network_name && network_name_len) { desc->network_name = kstrdup(network_name, GFP_KERNEL); + if (!desc->network_name) { + kfree(desc); + return NULL; + } + } vidtv_psi_desc_chain(head, (struct vidtv_psi_desc *)desc); return desc; @@ -449,15 +467,32 @@ struct vidtv_psi_desc_short_event iso_language_code = "eng"; desc->iso_language_code = kstrdup(iso_language_code, GFP_KERNEL); + if (!desc->iso_language_code) + goto free_desc; - if (event_name && event_name_len) + if (event_name && event_name_len) { desc->event_name = kstrdup(event_name, GFP_KERNEL); + if (!desc->event_name) + goto free_desc_language_code; + } - if (text && text_len) + if (text && text_len) { desc->text = kstrdup(text, GFP_KERNEL); + if (!desc->text) + goto free_desc_event_name; + } vidtv_psi_desc_chain(head, (struct vidtv_psi_desc *)desc); return desc; + +free_desc_event_name: + if (event_name && event_name_len) + kfree(desc->event_name); +free_desc_language_code: + kfree(desc->iso_language_code); +free_desc: + kfree(desc); + return NULL; } struct vidtv_psi_desc *vidtv_psi_desc_clone(struct vidtv_psi_desc *desc) -- 2.34.1

2 1

[PATCH OLK-5.10 v3 0/2] sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_GROUP_SCHED
by Yipeng Zou 24 May '24

24 May '24

It's typo error, should be FAIR_GROUP_SCHED but not FAIR_CGROUP_SCHED. Yipeng Zou (2): Revert "sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_CGROUP_SCHED" sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_GROUP_SCHED init/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) -- 2.34.1

2 3

[PATCH OLK-5.10 v3 0/2] sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_GROUP_SCHED
by Yipeng Zou 24 May '24

24 May '24

It's typo error, should be FAIR_GROUP_SCHED but not FAIR_CGROUP_SCHED. Yipeng Zou (2): Revert "sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_CGROUP_SCHED" sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_GROUP_SCHED init/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) -- 2.34.1

2 3

[PATCH openEuler-1.0-LTS] sched/rt: Fix rt_runtime leaks with cpu hotplug and RT_RUNTIME_SHARE
by Zhao Wenhui 24 May '24

24 May '24

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I9RZC8 -------------------------------- When using cgroup rt_bandwidth with RT_RUNTIME_SHARE, if there are cpu hotplug and cpu.rt_runtime_us changing concurrently, the warning in __disable_runtime may occur: [ 991.697692] WARNING: CPU: 0 PID: 49573 at kernel/sched/rt.c:802 rq_offline_rt+0x24d/0x260 [ 991.697795] CPU: 0 PID: 49573 Comm: kworker/1:0 Kdump: loaded Not tainted 6.9.0-rc1+ #4 [ 991.697800] Workqueue: events cpuset_hotplug_workfn [ 991.697803] RIP: 0010:rq_offline_rt+0x24d/0x260 [ 991.697825] Call Trace: [ 991.697827] <TASK> [ 991.697858] set_rq_offline.part.125+0x2d/0x70 [ 991.697864] rq_attach_root+0xda/0x110 [ 991.697867] cpu_attach_domain+0x433/0x860 [ 991.697880] partition_sched_domains_locked+0x2a8/0x3a0 [ 991.697885] rebuild_sched_domains_locked+0x608/0x800 [ 991.697895] rebuild_sched_domains+0x1b/0x30 [ 991.697897] cpuset_hotplug_workfn+0x4b6/0x1160 [ 991.697909] process_scheduled_works+0xad/0x430 [ 991.697917] worker_thread+0x105/0x270 [ 991.697922] kthread+0xde/0x110 [ 991.697928] ret_from_fork+0x2d/0x50 [ 991.697935] ret_from_fork_asm+0x11/0x20 [ 991.697940] </TASK> [ 991.697941] ---[ end trace 0000000000000000 ]--- That's how it happens: CPU0 CPU1 ----- ----- set_rq_offline(rq) __disable_runtime(rq) (1) tg_set_rt_bandwidth (2) do_balance_runtime (3) set_rq_online(rq) __enable_runtime(rq) (4) In step(1) rt_rq->rt_runtime is set to RUNTIME_INF, and this rtrq's runtime is not supposed to change until its rq gets online. However, in step(2) tg_set_rt_bandwidth can set rt_rq->rt_runtime to rt_bandwidth.rt_runtime. Then, in step(3) rtrq's runtime is not RUNTIME_INF, so others can borrow rt_runtime from it. Finally, in step(4) the rq gets online, so its rtrq's runtime is set to rt_bandwidth.rt_runtime again, and Since then the total rt_runtime in the domain is increased by this way. After these steps, when offline cpu, rebuilding the sched_domain will offline all rq, and the last rq will find the rt_runtime is increased but nowhere to return. To fix this, we can add a state RUNTIME_DISABLED, which means the runtime is disabled and should not be used. When the rq is offline, we can set its rtrq's rt_runtime to RUNTIME_DISABLED, and when rq online, reset it. And in tg_set_rt_bandwidth and do_balance_runtime, never change a disabled rt_runtime. Fixes: 7def2be1dc67 ("sched: fix hotplug cpus on ia64") Closes: https://lore.kernel.org/all/47b4a790-9a27-2fc5-f2aa-f9981c6da015@huawei.com/ Co-developed-by: Hui Tang <tanghui20(a)huawei.com> Signed-off-by: Hui Tang <tanghui20(a)huawei.com> Signed-off-by: Zhao Wenhui <zhaowenhui8(a)huawei.com> --- kernel/sched/rt.c | 15 +++++++++------ kernel/sched/sched.h | 5 +++++ 2 files changed, 14 insertions(+), 6 deletions(-) diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c index 58364f489529..9093fa9f8fdc 100644 --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -664,7 +664,8 @@ static void do_balance_runtime(struct rt_rq *rt_rq) * or __disable_runtime() below sets a specific rq to inf to * indicate its been disabled and disalow stealing. */ - if (iter->rt_runtime == RUNTIME_INF) + if (iter->rt_runtime == RUNTIME_INF || + iter->rt_runtime == RUNTIME_DISABLED) goto next; /* @@ -735,7 +736,9 @@ static void __disable_runtime(struct rq *rq) /* * Can't reclaim from ourselves or disabled runqueues. */ - if (iter == rt_rq || iter->rt_runtime == RUNTIME_INF) + if (iter == rt_rq || + iter->rt_runtime == RUNTIME_INF || + iter->rt_runtime == RUNTIME_DISABLED) continue; raw_spin_lock(&iter->rt_runtime_lock); @@ -761,10 +764,9 @@ static void __disable_runtime(struct rq *rq) WARN_ON_ONCE(want); balanced: /* - * Disable all the borrow logic by pretending we have inf - * runtime - in which case borrowing doesn't make sense. + * Disable all the borrow logic by marking runtime disabled. */ - rt_rq->rt_runtime = RUNTIME_INF; + rt_rq->rt_runtime = RUNTIME_DISABLED; rt_rq->rt_throttled = 0; raw_spin_unlock(&rt_rq->rt_runtime_lock); raw_spin_unlock(&rt_b->rt_runtime_lock); @@ -2555,7 +2557,8 @@ static int tg_set_rt_bandwidth(struct task_group *tg, struct rt_rq *rt_rq = tg->rt_rq[i]; raw_spin_lock(&rt_rq->rt_runtime_lock); - rt_rq->rt_runtime = rt_runtime; + if (rt_rq->rt_runtime != RUNTIME_DISABLED) + rt_rq->rt_runtime = rt_runtime; raw_spin_unlock(&rt_rq->rt_runtime_lock); } raw_spin_unlock_irq(&tg->rt_bandwidth.rt_runtime_lock); diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index 4dd0e4de0aab..80e9d254ab7c 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -163,6 +163,11 @@ static inline void cpu_load_update_active(struct rq *this_rq) { } */ #define RUNTIME_INF ((u64)~0ULL) +/* + * Single value that denotes runtime is disabled, and it should not be used. + */ +#define RUNTIME_DISABLED (-2ULL) + static inline int idle_policy(int policy) { return policy == SCHED_IDLE; -- 2.34.1

2 1

[PATCH OLK-5.10 v2 0/2] sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_GROUP_SCHED
by Yipeng Zou 24 May '24

24 May '24

It's typo error, should be FAIR_GROUP_SCHED but not FAIR_CGROUP_SCHED. Yipeng Zou (2): Revert "sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_CGROUP_SCHED" sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_GROUP_SCHED init/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) -- 2.34.1

2 3

[PATCH openEuler-1.0-LTS] sched/rt: Fix rt_runtime leaks with cpu hotplug and RT_RUNTIME_SHARE
by Zhao Wenhui 24 May '24

24 May '24

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I9RZC8 -------------------------------- When using cgroup rt_bandwidth with RT_RUNTIME_SHARE, if there are cpu hotplug and cpu.rt_runtime_us changing concurrently, the warning in __disable_runtime may occur: [ 991.697692] WARNING: CPU: 0 PID: 49573 at kernel/sched/rt.c:802 rq_offline_rt+0x24d/0x260 [ 991.697795] CPU: 0 PID: 49573 Comm: kworker/1:0 Kdump: loaded Not tainted 6.9.0-rc1+ #4 [ 991.697800] Workqueue: events cpuset_hotplug_workfn [ 991.697803] RIP: 0010:rq_offline_rt+0x24d/0x260 [ 991.697825] Call Trace: [ 991.697827] <TASK> [ 991.697858] set_rq_offline.part.125+0x2d/0x70 [ 991.697864] rq_attach_root+0xda/0x110 [ 991.697867] cpu_attach_domain+0x433/0x860 [ 991.697880] partition_sched_domains_locked+0x2a8/0x3a0 [ 991.697885] rebuild_sched_domains_locked+0x608/0x800 [ 991.697895] rebuild_sched_domains+0x1b/0x30 [ 991.697897] cpuset_hotplug_workfn+0x4b6/0x1160 [ 991.697909] process_scheduled_works+0xad/0x430 [ 991.697917] worker_thread+0x105/0x270 [ 991.697922] kthread+0xde/0x110 [ 991.697928] ret_from_fork+0x2d/0x50 [ 991.697935] ret_from_fork_asm+0x11/0x20 [ 991.697940] </TASK> [ 991.697941] ---[ end trace 0000000000000000 ]--- That's how it happens: CPU0 CPU1 ----- ----- set_rq_offline(rq) __disable_runtime(rq) (1) tg_set_rt_bandwidth (2) do_balance_runtime (3) set_rq_online(rq) __enable_runtime(rq) (4) In step(1) rt_rq->rt_runtime is set to RUNTIME_INF, and this rtrq's runtime is not supposed to change until its rq gets online. However, in step(2) tg_set_rt_bandwidth can set rt_rq->rt_runtime to rt_bandwidth.rt_runtime. Then, in step(3) rtrq's runtime is not RUNTIME_INF, so others can borrow rt_runtime from it. Finally, in step(4) the rq gets online, so its rtrq's runtime is set to rt_bandwidth.rt_runtime again, and Since then the total rt_runtime in the domain is increased by this way. After these steps, when offline cpu, rebuilding the sched_domain will offline all rq, and the last rq will find the rt_runtime is increased but nowhere to return. To fix this, we can add a state RUNTIME_DISABLED, which means the runtime is disabled and should not be used. When the rq is offline, we can set its rtrq's rt_runtime to RUNTIME_DISABLED, and when rq online, reset it. And in tg_set_rt_bandwidth and do_balance_runtime, never change a disabled rt_runtime. Fixes: 7def2be1dc67 ("sched: fix hotplug cpus on ia64") Closes: https://lore.kernel.org/all/47b4a790-9a27-2fc5-f2aa-f9981c6da015@huawei.com/ Co-developed-by: Hui Tang <tanghui20(a)huawei.com> Signed-off-by: Hui Tang <tanghui20(a)huawei.com> Signed-off-by: Zhao Wenhui <zhaowenhui8(a)huawei.com> --- kernel/sched/fair.c.rej | 52 +++++++++++++++++++++++++++++++++++++++++ kernel/sched/rt.c | 3 ++- kernel/sched/rt.c.rej | 11 +++++++++ 3 files changed, 65 insertions(+), 1 deletion(-) create mode 100644 kernel/sched/fair.c.rej create mode 100644 kernel/sched/rt.c.rej diff --git a/kernel/sched/fair.c.rej b/kernel/sched/fair.c.rej new file mode 100644 index 000000000000..7275501f1bf3 --- /dev/null +++ b/kernel/sched/fair.c.rej @@ -0,0 +1,52 @@ +diff a/kernel/sched/fair.c b/kernel/sched/fair.c (rejected hunks) +@@ -161,9 +161,12 @@ int __weak arch_asym_cpu_priority(int cpu) + #endif + + #ifdef CONFIG_QOS_SCHED +- ++struct qos_overload_checker { ++ struct hrtimer qos_overload_timer; ++ unsigned int cpu; ++}; + static DEFINE_PER_CPU_SHARED_ALIGNED(struct list_head, qos_throttled_cfs_rq); +-static DEFINE_PER_CPU_SHARED_ALIGNED(struct hrtimer, qos_overload_timer); ++static DEFINE_PER_CPU_SHARED_ALIGNED(struct qos_overload_checker, qos_overload_checker); + static DEFINE_PER_CPU(int, qos_cpu_overload); + unsigned int sysctl_overload_detect_period = 5000; /* in ms */ + unsigned int sysctl_offline_wait_interval = 100; /* in ms */ +@@ -7951,14 +7954,17 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p, int wake_ + } + + #ifdef CONFIG_QOS_SCHED ++static struct hrtimer *qos_timer(int cpu) { ++ return &(per_cpu_ptr(&qos_overload_checker, cpu)->qos_overload_timer); ++} + static inline bool qos_timer_is_activated(int cpu) + { +- return hrtimer_active(per_cpu_ptr(&qos_overload_timer, cpu)); ++ return hrtimer_active(qos_timer(cpu)); + } + + static inline void cancel_qos_timer(int cpu) + { +- hrtimer_cancel(per_cpu_ptr(&qos_overload_timer, cpu)); ++ hrtimer_cancel(qos_timer(cpu)); + } + + static inline bool is_offline_task(struct task_struct *p) +@@ -8183,10 +8189,13 @@ void sched_qos_offline_wait(void) + static enum hrtimer_restart qos_overload_timer_handler(struct hrtimer *timer) + { + struct rq_flags rf; +- struct rq *rq = this_rq(); ++ struct qos_overload_checker *checker = container_of(timer, ++ struct qos_overload_checker, qos_overload_timer); ++ int cpu = checker->cpu; ++ struct rq *rq = cpu_rq(cpu); + + rq_lock_irqsave(rq, &rf); +- __unthrottle_qos_cfs_rqs(smp_processor_id()); ++ __unthrottle_qos_cfs_rqs(cpu); + __this_cpu_write(qos_cpu_overload, 1); + + /* Determine whether we need to wake up potentially idle CPU. */ diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c index 58364f489529..bb1f76264686 100644 --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -664,7 +664,8 @@ static void do_balance_runtime(struct rt_rq *rt_rq) * or __disable_runtime() below sets a specific rq to inf to * indicate its been disabled and disalow stealing. */ - if (iter->rt_runtime == RUNTIME_INF) + if (iter->rt_runtime == RUNTIME_INF || + iter->rt_runtime == RUNTIME_DISABLED) goto next; /* diff --git a/kernel/sched/rt.c.rej b/kernel/sched/rt.c.rej new file mode 100644 index 000000000000..710818859059 --- /dev/null +++ b/kernel/sched/rt.c.rej @@ -0,0 +1,11 @@ +diff a/kernel/sched/rt.c b/kernel/sched/rt.c (rejected hunks) +@@ -704,7 +704,8 @@ static void do_balance_runtime(struct rt_rq *rt_rq) + * or __disable_runtime() below sets a specific rq to inf to + * indicate its been disabled and disallow stealing. + */ +- if (iter->rt_runtime == RUNTIME_INF) ++ if (iter->rt_runtime == RUNTIME_INF || ++ iter->rt_runtime == RUNTIME_DISABLED) + goto next; + + /* -- 2.34.1

2 1

[PATCH OLK-5.10 v2 0/2] sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_GROUP_SCHED
by Yipeng Zou 24 May '24

24 May '24

It's typo error, should be FAIR_GROUP_SCHED but not FAIR_CGROUP_SCHED. Yipeng Zou (2): Revert "sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_CGROUP_SCHED" sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_GROUP_SCHED init/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) -- 2.34.1

2 3