- Kernel - mailweb.openeuler.org

[PATCH OLK-5.5 00/10] Add SDEI Watchdog Support
by Xiongfeng Wang 27 Nov '23

27 Nov '23

Xiongfeng Wang (10): firmware: arm_sdei: add interrupt binding api firmware: arm_sdei: make 'sdei_api_event_disable/enable' public lockup_detector: init lockup detector after all the init_calls watchdog: add nmi_watchdog support for arm64 based on SDEI sdei_watchdog: clear EOI of the secure timer before kdump sdei_watchdog: set secure timer period base on 'watchdog_thresh' sdei_watchdog: avoid possible false hardlockup init: only move down lockup_detector_init() when sdei_watchdog is enabled kprobes/arm64: Blacklist sdei watchdog callback functions openeuler_defconfig: Enable SDEI Watchdog arch/arm64/configs/openeuler_defconfig | 3 +- arch/arm64/kernel/Makefile | 1 + arch/arm64/kernel/machine_kexec.c | 10 ++ arch/arm64/kernel/watchdog_sdei.c | 151 +++++++++++++++++++++++++ drivers/firmware/arm_sdei.c | 26 ++++- include/linux/arm_sdei.h | 5 + include/linux/nmi.h | 8 ++ include/uapi/linux/arm_sdei.h | 2 + init/main.c | 7 +- kernel/watchdog.c | 3 + lib/Kconfig.debug | 9 ++ 11 files changed, 221 insertions(+), 4 deletions(-) create mode 100644 arch/arm64/kernel/watchdog_sdei.c -- 2.20.1

1 10

[PATCH OLK-5.10] ipmi_si: fix a memleak in try_smi_init()
by Yi Yang 27 Nov '23

27 Nov '23

stable inclusion from stable-v5.10.195 commit 5c5f02e16b919c8cb6024dc3778c8d8f1fb1f26b category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7GMMZ Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- commit 6cf1a126de2992b4efe1c3c4d398f8de4aed6e3f upstream. Kmemleak reported the following leak info in try_smi_init(): unreferenced object 0xffff00018ecf9400 (size 1024): comm "modprobe", pid 2707763, jiffies 4300851415 (age 773.308s) backtrace: [<000000004ca5b312>] __kmalloc+0x4b8/0x7b0 [<00000000953b1072>] try_smi_init+0x148/0x5dc [ipmi_si] [<000000006460d325>] 0xffff800081b10148 [<0000000039206ea5>] do_one_initcall+0x64/0x2a4 [<00000000601399ce>] do_init_module+0x50/0x300 [<000000003c12ba3c>] load_module+0x7a8/0x9e0 [<00000000c246fffe>] __se_sys_init_module+0x104/0x180 [<00000000eea99093>] __arm64_sys_init_module+0x24/0x30 [<0000000021b1ef87>] el0_svc_common.constprop.0+0x94/0x250 [<0000000070f4f8b7>] do_el0_svc+0x48/0xe0 [<000000005a05337f>] el0_svc+0x24/0x3c [<000000005eb248d6>] el0_sync_handler+0x160/0x164 [<0000000030a59039>] el0_sync+0x160/0x180 The problem was that when an error occurred before handlers registration and after allocating `new_smi->si_sm`, the variable wouldn't be freed in the error handling afterwards since `shutdown_smi()` hadn't been registered yet. Fix it by adding a `kfree()` in the error handling path in `try_smi_init()`. Cc: stable(a)vger.kernel.org # 4.19+ Fixes: 7960f18a5647 ("ipmi_si: Convert over to a shutdown handler") Signed-off-by: Yi Yang <yiyang13(a)huawei.com> Co-developed-by: GONG, Ruiqi <gongruiqi(a)huaweicloud.com> Signed-off-by: GONG, Ruiqi <gongruiqi(a)huaweicloud.com> Message-Id: <20230629123328.2402075-1-gongruiqi(a)huaweicloud.com> Signed-off-by: Corey Minyard <minyard(a)acm.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Yi Yang <yiyang13(a)huawei.com> --- drivers/char/ipmi/ipmi_si_intf.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/char/ipmi/ipmi_si_intf.c b/drivers/char/ipmi/ipmi_si_intf.c index 9e3e24c43168..227eb9e8e47d 100644 --- a/drivers/char/ipmi/ipmi_si_intf.c +++ b/drivers/char/ipmi/ipmi_si_intf.c @@ -2089,6 +2089,11 @@ static int try_smi_init(struct smi_info *new_smi) new_smi->io.io_cleanup = NULL; } + if (rv && new_smi->si_sm) { + kfree(new_smi->si_sm); + new_smi->si_sm = NULL; + } + return rv; } -- 2.25.1

1 0

[PATCH OLK-5.10] fs/dirty_pages: fix some errors in seq_read_dirty()
by Zizhi Wo 27 Nov '23

27 Nov '23

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8JUUC?from=project-issue -------------------------------- In seq_read_dirty(), the parameter passed into simple_read_from_buffer() will cause problem on a 32-bit system; The simple_read_from_buffer() function has a lot of redundancy checks inside. Replaced it with copy_to_user() and add the return value check. And the update of the "copied" character count in seq_read_dirty() is also incorrect, fixed it. Fixes: aeb96447becc ("fs/dirty_pages: dump the number of dirty pages for each inode") Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/dirty_pages.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/dirty_pages.c b/fs/dirty_pages.c index 21e2251b6619..0eb818b599ba 100644 --- a/fs/dirty_pages.c +++ b/fs/dirty_pages.c @@ -217,12 +217,12 @@ static ssize_t seq_read_dirty( } n = min(m->count - m->from, size); - err = simple_read_from_buffer(buf, n, - (loff_t *) &m->from, m->buf, m->count); - if (err < 0) { + n -= copy_to_user(buf, m->buf + m->from, n); + if (unlikely(!n)) { err = -EFAULT; goto done; } + m->from += n; copied += n; done: if (!copied) -- 2.39.2

2 1

[PATCH openEuler-1.0-LTS 0/2] fix CFS bandwidth vs. hrtimer self deadlock
by Yu Liao 27 Nov '23

27 Nov '23

Thomas Gleixner (1): hrtimers: Push pending hrtimers away from outgoing CPU earlier Yu Liao (1): cpu/hotplug: fix kabi breakage in enum cpuhp_state include/linux/hrtimer.h | 4 ++-- include/linux/smp.h | 1 + kernel/cpu.c | 17 +++++++++++++++-- kernel/smp.c | 8 ++++++++ kernel/time/hrtimer.c | 33 ++++++++++++--------------------- 5 files changed, 38 insertions(+), 25 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10] fs/dirty_pages: fix inode reference count leakage error in dump_dirtypages_sb()
by Zizhi Wo 27 Nov '23

27 Nov '23

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8JUUC?from=project-issue -------------------------------- Inode reference count leakage exists in dump_dirtypages_sb() function. Executing iput(inode) in advance to avoid this problem. Fixes: aeb96447becc ("fs/dirty_pages: dump the number of dirty pages for each inode") Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/dirty_pages.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/dirty_pages.c b/fs/dirty_pages.c index 0eb818b599ba..b0cf442c5730 100644 --- a/fs/dirty_pages.c +++ b/fs/dirty_pages.c @@ -159,6 +159,7 @@ static void dump_dirtypages_sb(struct super_block *sb, struct seq_file *m) if (m->size <= m->count) { seq_set_overflow(m); strncpy(m->buf+m->count-12, "terminated\n\0", 12); + iput(inode); goto done; } seq_printf(m, "FSType: %s, Dev ID: %u(%u:%u) ino %lu, dirty pages %lu, path %s\n", -- 2.39.2

2 1

[PATCH OLK-6.6 00/33] folio conversions for numa balance
by Peng Zhang 27 Nov '23

27 Nov '23

From: ZhangPeng <zhangpeng362(a)huawei.com> Includes 3 patch series "mm: migrate: more folio conversion and unification", "mm: convert numa balancing functions to use a folio" and "mm: convert page cpupid functions to folios" to implement folio conversions of numa balance. Kefeng Wang (33): mm: migrate: remove PageTransHuge check in numamigrate_isolate_page() mm: migrate: remove THP mapcount check in numamigrate_isolate_page() mm: migrate: convert numamigrate_isolate_page() to numamigrate_isolate_folio() mm: migrate: convert migrate_misplaced_page() to migrate_misplaced_folio() mm: migrate: use __folio_test_movable() mm: migrate: use a folio in add_page_for_migration() mm: migrate: remove PageHead() check for HugeTLB in add_page_for_migration() mm: migrate: remove isolated variable in add_page_for_migration() mm: memory: add vm_normal_folio_pmd() mm: huge_memory: use a folio in do_huge_pmd_numa_page() mm: memory: use a folio in do_numa_page() mm: memory: make numa_migrate_prep() to take a folio mm: mempolicy: make mpol_misplaced() to take a folio sched/numa, mm: make numa migrate functions to take a folio mm_types: add virtual and _last_cpupid into struct folio mm: add folio_last_cpupid() mm: memory: use folio_last_cpupid() in do_numa_page() mm: huge_memory: use folio_last_cpupid() in do_huge_pmd_numa_page() mm: huge_memory: use folio_last_cpupid() in __split_huge_page_tail() mm: remove page_cpupid_last() mm: add folio_xchg_access_time() sched/fair: use folio_xchg_access_time() in numa_hint_fault_latency() mm: mprotect: use a folio in change_pte_range() mm: huge_memory: use a folio in change_huge_pmd() mm: remove xchg_page_access_time() mm: add folio_xchg_last_cpupid() sched/fair: use folio_xchg_last_cpupid() in should_numa_migrate_memory() mm: migrate: use folio_xchg_last_cpupid() in folio_migrate_flags() mm: huge_memory: use folio_xchg_last_cpupid() in __split_huge_page_tail() mm: make finish_mkwrite_fault() static mm: convert wp_page_reuse() and finish_mkwrite_fault() to take a folio mm: use folio_xchg_last_cpupid() in wp_page_reuse() mm: remove page_cpupid_xchg_last() include/linux/mempolicy.h | 5 +- include/linux/migrate.h | 4 +- include/linux/mm.h | 32 ++++--- include/linux/mm_types.h | 22 ++++- include/linux/sched/numa_balancing.h | 6 +- kernel/sched/fair.c | 12 +-- mm/huge_memory.c | 44 +++++---- mm/internal.h | 2 +- mm/memory.c | 85 +++++++++-------- mm/mempolicy.c | 22 +++-- mm/migrate.c | 134 +++++++++++++-------------- mm/mmzone.c | 6 +- mm/mprotect.c | 16 ++-- 13 files changed, 205 insertions(+), 185 deletions(-) -- 2.25.1

2 71

[PATCH OLK-5.10 v3 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 27 Nov '23

27 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 690 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 872 insertions(+), 171 deletions(-) -- 2.34.1

1 20

[PATCH OLK-6.6 v2] anolis: bond: broadcast ARP or ND messages to all slaves
by Zhengchao Shao 27 Nov '23

27 Nov '23

From: Tony Lu <tonylu(a)linux.alibaba.com> anolis inclusion from devel-5.10-v5.10.134-12 commit b90e28f7170e1ae40c572f9f80a50bbdc8f8b99f category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8JTCU Reference: https://gitee.com/anolis/cloud-kernel/commit/b90e28f7170e1ae40c572f9f80a50b… --------------------------- OpenAnolis Bug Tracker:0000282 This is achieved by broadcasting ARP or ND packets to all of its slave devices on transmit side. The switch will take further actions based on proper configuration. A new sysctl knob "net.bonding.broadcast_arp_or_nd" is introduced which controls the behaviour of broadcasting. Signed-off-by: Tony Lu <tonylu(a)linux.alibaba.com> Acked-by: Dust Li <dust.li(a)linux.alibaba.com> Signed-off-by: Qiao Ma <mqaio(a)linux.alibaba.com> Reviewed-by: Shile Zhang <shile.zhang(a)linux.alibaba.com> Acked-by: Dust Li <dust.li(a)linux.alibaba.com> Signed-off-by: Wang Yufen <wangyufen(a)huawei.com> (cherry picked from commit 7c902772d8ee8840b6092b1de94ed0c17c9a8a50) Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- v2: modify email address error --- Documentation/networking/ip-sysctl.rst | 11 +++++++ drivers/net/bonding/Makefile | 2 +- drivers/net/bonding/bond_main.c | 40 ++++++++++++++++++++++++++ drivers/net/bonding/bond_sysctl.c | 32 +++++++++++++++++++++ include/net/bonding.h | 4 +++ 5 files changed, 88 insertions(+), 1 deletion(-) create mode 100644 drivers/net/bonding/bond_sysctl.c diff --git a/Documentation/networking/ip-sysctl.rst b/Documentation/networking/ip-sysctl.rst index a66054d0763a..12fa38639102 100644 --- a/Documentation/networking/ip-sysctl.rst +++ b/Documentation/networking/ip-sysctl.rst @@ -3206,3 +3206,14 @@ max_dgram_qlen - INTEGER Default: 10 + +``/proc/sys/net/bonding/*`` +======================== + +broadcast_arp_or_nd - INTEGER + Control broadcasting ARP or ND messages to all slaves + + 0: Not broadcasting + 1: Broadcasting + + Default: 0 diff --git a/drivers/net/bonding/Makefile b/drivers/net/bonding/Makefile index 30e8ae3da2da..9dd55d91185a 100644 --- a/drivers/net/bonding/Makefile +++ b/drivers/net/bonding/Makefile @@ -5,7 +5,7 @@ obj-$(CONFIG_BONDING) += bonding.o -bonding-objs := bond_main.o bond_3ad.o bond_alb.o bond_sysfs.o bond_sysfs_slave.o bond_debugfs.o bond_netlink.o bond_options.o +bonding-objs := bond_main.o bond_3ad.o bond_alb.o bond_sysfs.o bond_sysfs_slave.o bond_debugfs.o bond_netlink.o bond_options.o bond_sysctl.o proc-$(CONFIG_PROC_FS) += bond_procfs.o bonding-objs += $(proc-y) diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 51d47eda1c87..842a138a574d 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -5127,6 +5127,39 @@ int bond_update_slave_arr(struct bonding *bond, struct slave *skipslave) return ret; } +/* Check whether the skb is arp or nd msg */ +static inline bool skb_is_arp_or_nd(struct sk_buff *skb) +{ + switch (ntohs(skb->protocol)) { + case ETH_P_ARP: + return true; + case ETH_P_IPV6: + if (pskb_may_pull(skb, sizeof(struct ipv6hdr) + + sizeof(struct nd_msg))) { + struct ipv6hdr *hdr = ipv6_hdr(skb); + u8 nexthdr = hdr->nexthdr; + struct icmp6hdr *icmp6; + + if (nexthdr == IPPROTO_ICMPV6) { + icmp6 = icmp6_hdr(skb); + + if ((icmp6->icmp6_type == + NDISC_NEIGHBOUR_SOLICITATION || + icmp6->icmp6_type == + NDISC_NEIGHBOUR_ADVERTISEMENT) && + icmp6->icmp6_code == 0) { + return true; + } + } + } + } + + return false; +} + +static netdev_tx_t bond_xmit_broadcast(struct sk_buff *skb, + struct net_device *bond_dev); + static struct slave *bond_xmit_3ad_xor_slave_get(struct bonding *bond, struct sk_buff *skb, struct bond_up_slave *slaves) @@ -5171,6 +5204,10 @@ static netdev_tx_t bond_3ad_xor_xmit(struct sk_buff *skb, struct bond_up_slave *slaves; struct slave *slave; + /* Broadcast to all slaves. */ + if (sysctl_bond_broadcast_arp_or_nd && skb_is_arp_or_nd(skb)) + return bond_xmit_broadcast(skb, dev); + slaves = rcu_dereference(bond->usable_slaves); slave = bond_xmit_3ad_xor_slave_get(bond, skb, slaves); if (likely(slave)) @@ -6488,6 +6525,7 @@ static int __init bonding_init(void) goto err_link; bond_create_debugfs(); + bond_create_sysctl(); for (i = 0; i < max_bonds; i++) { res = bond_create(&init_net, NULL); @@ -6504,6 +6542,7 @@ static int __init bonding_init(void) return res; err: bond_destroy_debugfs(); + bond_destroy_sysctl(); bond_netlink_fini(); err_link: unregister_pernet_subsys(&bond_net_ops); @@ -6516,6 +6555,7 @@ static void __exit bonding_exit(void) unregister_netdevice_notifier(&bond_netdev_notifier); bond_destroy_debugfs(); + bond_destroy_sysctl(); bond_netlink_fini(); unregister_pernet_subsys(&bond_net_ops); diff --git a/drivers/net/bonding/bond_sysctl.c b/drivers/net/bonding/bond_sysctl.c new file mode 100644 index 000000000000..17404d37a2fd --- /dev/null +++ b/drivers/net/bonding/bond_sysctl.c @@ -0,0 +1,32 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <net/net_namespace.h> +#include <linux/sysctl.h> +#include <net/bonding.h> + +int sysctl_bond_broadcast_arp_or_nd __read_mostly; +EXPORT_SYMBOL(sysctl_bond_broadcast_arp_or_nd); + +struct ctl_table_header *bond_broadcast_arp_or_nd_table_header; + +static struct ctl_table bond_broadcast_arp_or_nd_table[] = { + { + .procname = "broadcast_arp_or_nd", + .data = &sysctl_bond_broadcast_arp_or_nd, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + {} +}; + +void bond_create_sysctl(void) +{ + bond_broadcast_arp_or_nd_table_header = + register_net_sysctl(&init_net, "net/bonding", + bond_broadcast_arp_or_nd_table); +} + +void bond_destroy_sysctl(void) +{ + unregister_net_sysctl_table(bond_broadcast_arp_or_nd_table_header); +} diff --git a/include/net/bonding.h b/include/net/bonding.h index 5b8b1b644a2d..0f065f150b81 100644 --- a/include/net/bonding.h +++ b/include/net/bonding.h @@ -115,6 +115,8 @@ static inline int is_netpoll_tx_blocked(struct net_device *dev) #define is_netpoll_tx_blocked(dev) (0) #endif +extern int sysctl_bond_broadcast_arp_or_nd; + struct bond_params { int mode; int xmit_policy; @@ -684,6 +686,8 @@ struct bond_vlan_tag *bond_verify_device_path(struct net_device *start_dev, int bond_update_slave_arr(struct bonding *bond, struct slave *skipslave); void bond_slave_arr_work_rearm(struct bonding *bond, unsigned long delay); void bond_work_init_all(struct bonding *bond); +void bond_create_sysctl(void); +void bond_destroy_sysctl(void); #ifdef CONFIG_PROC_FS void bond_create_proc_entry(struct bonding *bond); -- 2.34.1

2 1

[PATCH OLK-6.6] anolis: bond: broadcast ARP or ND messages to all slaves
by Zhengchao Shao 27 Nov '23

27 Nov '23

From: Tony Lu <tonylu(a)linux.alibaba.com> anolis inclusion from devel-5.10-v5.10.134-12 commit b90e28f7170e1ae40c572f9f80a50bbdc8f8b99f category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8JTCU Reference: https://gitee.com/anolis/cloud-kernel/commit/b90e28f7170e1ae40c572f9f80a50b… --------------------------- OpenAnolis Bug Tracker:0000282 This is achieved by broadcasting ARP or ND packets to all of its slave devices on transmit side. The switch will take further actions based on proper configuration. A new sysctl knob "net.bonding.broadcast_arp_or_nd" is introduced which controls the behaviour of broadcasting. Signed-off-by: Tony Lu <tonylu(a)linux.alibaba.com> Acked-by: Dust Li <dust.li(a)linux.alibaba.com> Signed-off-by: Qiao Ma <mqaio(a)linux.alibaba.com> Reviewed-by: Shile Zhang <shile.zhang(a)linux.alibaba.com> Acked-by: Dust Li <dust.li(a)linux.alibaba.com> Signed-off-by: Wang Yufen <wangyufen(a)huawei.com> (cherry picked from commit 7c902772d8ee8840b6092b1de94ed0c17c9a8a50) Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- Documentation/networking/ip-sysctl.rst | 11 +++++++ drivers/net/bonding/Makefile | 2 +- drivers/net/bonding/bond_main.c | 40 ++++++++++++++++++++++++++ drivers/net/bonding/bond_sysctl.c | 32 +++++++++++++++++++++ include/net/bonding.h | 4 +++ 5 files changed, 88 insertions(+), 1 deletion(-) create mode 100644 drivers/net/bonding/bond_sysctl.c diff --git a/Documentation/networking/ip-sysctl.rst b/Documentation/networking/ip-sysctl.rst index a66054d0763a..12fa38639102 100644 --- a/Documentation/networking/ip-sysctl.rst +++ b/Documentation/networking/ip-sysctl.rst @@ -3206,3 +3206,14 @@ max_dgram_qlen - INTEGER Default: 10 + +``/proc/sys/net/bonding/*`` +======================== + +broadcast_arp_or_nd - INTEGER + Control broadcasting ARP or ND messages to all slaves + + 0: Not broadcasting + 1: Broadcasting + + Default: 0 diff --git a/drivers/net/bonding/Makefile b/drivers/net/bonding/Makefile index 30e8ae3da2da..9dd55d91185a 100644 --- a/drivers/net/bonding/Makefile +++ b/drivers/net/bonding/Makefile @@ -5,7 +5,7 @@ obj-$(CONFIG_BONDING) += bonding.o -bonding-objs := bond_main.o bond_3ad.o bond_alb.o bond_sysfs.o bond_sysfs_slave.o bond_debugfs.o bond_netlink.o bond_options.o +bonding-objs := bond_main.o bond_3ad.o bond_alb.o bond_sysfs.o bond_sysfs_slave.o bond_debugfs.o bond_netlink.o bond_options.o bond_sysctl.o proc-$(CONFIG_PROC_FS) += bond_procfs.o bonding-objs += $(proc-y) diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 51d47eda1c87..842a138a574d 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -5127,6 +5127,39 @@ int bond_update_slave_arr(struct bonding *bond, struct slave *skipslave) return ret; } +/* Check whether the skb is arp or nd msg */ +static inline bool skb_is_arp_or_nd(struct sk_buff *skb) +{ + switch (ntohs(skb->protocol)) { + case ETH_P_ARP: + return true; + case ETH_P_IPV6: + if (pskb_may_pull(skb, sizeof(struct ipv6hdr) + + sizeof(struct nd_msg))) { + struct ipv6hdr *hdr = ipv6_hdr(skb); + u8 nexthdr = hdr->nexthdr; + struct icmp6hdr *icmp6; + + if (nexthdr == IPPROTO_ICMPV6) { + icmp6 = icmp6_hdr(skb); + + if ((icmp6->icmp6_type == + NDISC_NEIGHBOUR_SOLICITATION || + icmp6->icmp6_type == + NDISC_NEIGHBOUR_ADVERTISEMENT) && + icmp6->icmp6_code == 0) { + return true; + } + } + } + } + + return false; +} + +static netdev_tx_t bond_xmit_broadcast(struct sk_buff *skb, + struct net_device *bond_dev); + static struct slave *bond_xmit_3ad_xor_slave_get(struct bonding *bond, struct sk_buff *skb, struct bond_up_slave *slaves) @@ -5171,6 +5204,10 @@ static netdev_tx_t bond_3ad_xor_xmit(struct sk_buff *skb, struct bond_up_slave *slaves; struct slave *slave; + /* Broadcast to all slaves. */ + if (sysctl_bond_broadcast_arp_or_nd && skb_is_arp_or_nd(skb)) + return bond_xmit_broadcast(skb, dev); + slaves = rcu_dereference(bond->usable_slaves); slave = bond_xmit_3ad_xor_slave_get(bond, skb, slaves); if (likely(slave)) @@ -6488,6 +6525,7 @@ static int __init bonding_init(void) goto err_link; bond_create_debugfs(); + bond_create_sysctl(); for (i = 0; i < max_bonds; i++) { res = bond_create(&init_net, NULL); @@ -6504,6 +6542,7 @@ static int __init bonding_init(void) return res; err: bond_destroy_debugfs(); + bond_destroy_sysctl(); bond_netlink_fini(); err_link: unregister_pernet_subsys(&bond_net_ops); @@ -6516,6 +6555,7 @@ static void __exit bonding_exit(void) unregister_netdevice_notifier(&bond_netdev_notifier); bond_destroy_debugfs(); + bond_destroy_sysctl(); bond_netlink_fini(); unregister_pernet_subsys(&bond_net_ops); diff --git a/drivers/net/bonding/bond_sysctl.c b/drivers/net/bonding/bond_sysctl.c new file mode 100644 index 000000000000..17404d37a2fd --- /dev/null +++ b/drivers/net/bonding/bond_sysctl.c @@ -0,0 +1,32 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <net/net_namespace.h> +#include <linux/sysctl.h> +#include <net/bonding.h> + +int sysctl_bond_broadcast_arp_or_nd __read_mostly; +EXPORT_SYMBOL(sysctl_bond_broadcast_arp_or_nd); + +struct ctl_table_header *bond_broadcast_arp_or_nd_table_header; + +static struct ctl_table bond_broadcast_arp_or_nd_table[] = { + { + .procname = "broadcast_arp_or_nd", + .data = &sysctl_bond_broadcast_arp_or_nd, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + {} +}; + +void bond_create_sysctl(void) +{ + bond_broadcast_arp_or_nd_table_header = + register_net_sysctl(&init_net, "net/bonding", + bond_broadcast_arp_or_nd_table); +} + +void bond_destroy_sysctl(void) +{ + unregister_net_sysctl_table(bond_broadcast_arp_or_nd_table_header); +} diff --git a/include/net/bonding.h b/include/net/bonding.h index 5b8b1b644a2d..0f065f150b81 100644 --- a/include/net/bonding.h +++ b/include/net/bonding.h @@ -115,6 +115,8 @@ static inline int is_netpoll_tx_blocked(struct net_device *dev) #define is_netpoll_tx_blocked(dev) (0) #endif +extern int sysctl_bond_broadcast_arp_or_nd; + struct bond_params { int mode; int xmit_policy; @@ -684,6 +686,8 @@ struct bond_vlan_tag *bond_verify_device_path(struct net_device *start_dev, int bond_update_slave_arr(struct bonding *bond, struct slave *skipslave); void bond_slave_arr_work_rearm(struct bonding *bond, unsigned long delay); void bond_work_init_all(struct bonding *bond); +void bond_create_sysctl(void); +void bond_destroy_sysctl(void); #ifdef CONFIG_PROC_FS void bond_create_proc_entry(struct bonding *bond); -- 2.34.1

2 1

[PATCH OLK-5.10 v3 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 27 Nov '23

27 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 690 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 872 insertions(+), 171 deletions(-) -- 2.34.1

1 20