- Kernel - mailweb.openeuler.org

[PATCH OLK-6.6 v2] kabi: reserve space for bpf related structures
by Pu Lehui 24 Jan '24

24 Jan '24

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8UOP9 -------------------------------- reserve space for bpf related structures. Signed-off-by: Pu Lehui <pulehui(a)huawei.com> --- include/linux/bpf.h | 123 ++++++++++++++++++++++++++++++ include/linux/bpf_local_storage.h | 11 +++ include/linux/bpf_verifier.h | 40 ++++++++++ include/uapi/linux/bpf.h | 22 ++++++ 4 files changed, 196 insertions(+) diff --git a/include/linux/bpf.h b/include/linux/bpf.h index 392f581af2ce..3051926802bf 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -29,6 +29,7 @@ #include <linux/rcupdate_trace.h> #include <linux/static_call.h> #include <linux/memcontrol.h> +#include <linux/kabi.h> struct bpf_verifier_env; struct bpf_verifier_log; @@ -168,6 +169,11 @@ struct bpf_map_ops { /* bpf_iter info used to open a seq_file */ const struct bpf_iter_seq_info *iter_seq_info; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; enum { @@ -288,6 +294,11 @@ struct bpf_map { bool bypass_spec_v1; bool frozen; /* write-once; write-protected by freeze_mutex */ s64 __percpu *elem_count; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; static inline const char *btf_field_type_name(enum btf_field_type type) @@ -664,6 +675,13 @@ enum bpf_type_flag { /* DYNPTR points to xdp_buff */ DYNPTR_TYPE_XDP = BIT(16 + BPF_BASE_TYPE_BITS), +#ifdef CONFIG_KABI_RESERVE + BPF_TYPE_FLAG_RESERVE_1 = BIT(17 + BPF_BASE_TYPE_BITS), + BPF_TYPE_FLAG_RESERVE_2 = BIT(18 + BPF_BASE_TYPE_BITS), + BPF_TYPE_FLAG_RESERVE_3 = BIT(19 + BPF_BASE_TYPE_BITS), + BPF_TYPE_FLAG_RESERVE_4 = BIT(20 + BPF_BASE_TYPE_BITS), +#endif + __BPF_TYPE_FLAG_MAX, __BPF_TYPE_LAST_FLAG = __BPF_TYPE_FLAG_MAX - 1, }; @@ -714,6 +732,12 @@ enum bpf_arg_type { ARG_PTR_TO_TIMER, /* pointer to bpf_timer */ ARG_PTR_TO_KPTR, /* pointer to referenced kptr */ ARG_PTR_TO_DYNPTR, /* pointer to bpf_dynptr. See bpf_type_flag for dynptr type */ + +#ifdef CONFIG_KABI_RESERVE + BPF_ARG_TYPE_RESERVE_1, + BPF_ARG_TYPE_RESERVE_2, +#endif + __BPF_ARG_TYPE_MAX, /* Extended arg_types. */ @@ -730,6 +754,11 @@ enum bpf_arg_type { /* Pointer to valid memory of size known at compile time. */ ARG_PTR_TO_FIXED_SIZE_MEM = MEM_FIXED_SIZE | ARG_PTR_TO_MEM, +#ifdef CONFIG_KABI_RESERVE + BPF_ARG_TYPE_RESERVE_3, + BPF_ARG_TYPE_RESERVE_4, +#endif + /* This must be the last entry. Its purpose is to ensure the enum is * wide enough to hold the higher bits reserved for bpf_type_flag. */ @@ -748,6 +777,12 @@ enum bpf_return_type { RET_PTR_TO_MEM, /* returns a pointer to memory */ RET_PTR_TO_MEM_OR_BTF_ID, /* returns a pointer to a valid memory or a btf_id */ RET_PTR_TO_BTF_ID, /* returns a pointer to a btf_id */ + +#ifdef CONFIG_KABI_RESERVE + BPF_RET_TYPE_RESERVE_1, + BPF_RET_TYPE_RESERVE_2, +#endif + __BPF_RET_TYPE_MAX, /* Extended ret_types. */ @@ -760,6 +795,11 @@ enum bpf_return_type { RET_PTR_TO_BTF_ID_OR_NULL = PTR_MAYBE_NULL | RET_PTR_TO_BTF_ID, RET_PTR_TO_BTF_ID_TRUSTED = PTR_TRUSTED | RET_PTR_TO_BTF_ID, +#ifdef CONFIG_KABI_RESERVE + BPF_RET_TYPE_RESERVE_3, + BPF_RET_TYPE_RESERVE_4, +#endif + /* This must be the last entry. Its purpose is to ensure the enum is * wide enough to hold the higher bits reserved for bpf_type_flag. */ @@ -807,6 +847,11 @@ struct bpf_func_proto { }; int *ret_btf_id; /* return value btf_id */ bool (*allowed)(const struct bpf_prog *prog); + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; /* bpf_context is intentionally undefined structure. Pointer to bpf_context is @@ -866,6 +911,12 @@ enum bpf_reg_type { PTR_TO_BUF, /* reg points to a read/write buffer */ PTR_TO_FUNC, /* reg points to a bpf program function */ CONST_PTR_TO_DYNPTR, /* reg points to a const struct bpf_dynptr */ + +#ifdef CONFIG_KABI_RESERVE + BPF_REG_TYPE_RESERVE_1, + BPF_REG_TYPE_RESERVE_2, +#endif + __BPF_REG_TYPE_MAX, /* Extended reg_types. */ @@ -875,6 +926,11 @@ enum bpf_reg_type { PTR_TO_TCP_SOCK_OR_NULL = PTR_MAYBE_NULL | PTR_TO_TCP_SOCK, PTR_TO_BTF_ID_OR_NULL = PTR_MAYBE_NULL | PTR_TO_BTF_ID, +#ifdef CONFIG_KABI_RESERVE + BPF_REG_TYPE_RESERVE_3, + BPF_REG_TYPE_RESERVE_4, +#endif + /* This must be the last entry. Its purpose is to ensure the enum is * wide enough to hold the higher bits reserved for bpf_type_flag. */ @@ -942,6 +998,11 @@ struct bpf_verifier_ops { int (*btf_struct_access)(struct bpf_verifier_log *log, const struct bpf_reg_state *reg, int off, int size); + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; struct bpf_prog_offload_ops { @@ -1001,6 +1062,9 @@ struct btf_func_model { u8 nr_args; u8 arg_size[MAX_BPF_FUNC_ARGS]; u8 arg_flags[MAX_BPF_FUNC_ARGS]; + + KABI_RESERVE(1) + KABI_RESERVE(2) }; /* Restore arguments before returning from trampoline to let original function @@ -1121,6 +1185,11 @@ struct bpf_tramp_image { struct rcu_head rcu; struct work_struct work; }; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; struct bpf_trampoline { @@ -1149,6 +1218,11 @@ struct bpf_trampoline { /* Executable image of trampoline */ struct bpf_tramp_image *cur_image; struct module *mod; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; struct bpf_attach_target_info { @@ -1180,6 +1254,11 @@ struct bpf_dispatcher { struct static_call_key *sc_key; void *sc_tramp; #endif + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; static __always_inline __nocfi unsigned int bpf_dispatcher_nop_func( @@ -1465,6 +1544,15 @@ struct bpf_prog_aux { struct work_struct work; struct rcu_head rcu; }; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) + KABI_RESERVE(5) + KABI_RESERVE(6) + KABI_RESERVE(7) + KABI_RESERVE(8) }; struct bpf_prog { @@ -1499,6 +1587,11 @@ struct bpf_prog { DECLARE_FLEX_ARRAY(struct sock_filter, insns); DECLARE_FLEX_ARRAY(struct bpf_insn, insnsi); }; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; struct bpf_array_aux { @@ -1507,6 +1600,11 @@ struct bpf_array_aux { struct bpf_map *map; struct mutex poke_mutex; struct work_struct work; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; struct bpf_link { @@ -1516,6 +1614,11 @@ struct bpf_link { const struct bpf_link_ops *ops; struct bpf_prog *prog; struct work_struct work; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; struct bpf_link_ops { @@ -1529,6 +1632,11 @@ struct bpf_link_ops { struct bpf_link_info *info); int (*update_map)(struct bpf_link *link, struct bpf_map *new_map, struct bpf_map *old_map); + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; struct bpf_tramp_link { @@ -1626,6 +1734,11 @@ struct bpf_struct_ops { struct btf_func_model func_models[BPF_STRUCT_OPS_MAX_NR_MEMBERS]; u32 type_id; u32 value_id; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; #if defined(CONFIG_BPF_JIT) && defined(CONFIG_BPF_SYSCALL) @@ -1727,6 +1840,11 @@ struct bpf_array { DECLARE_FLEX_ARRAY(void *, ptrs) __aligned(8); DECLARE_FLEX_ARRAY(void __percpu *, pptrs) __aligned(8); }; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; #define BPF_COMPLEXITY_LIMIT_INSNS 1000000 /* yes. 1M insns */ @@ -1822,6 +1940,11 @@ struct bpf_prog_array_item { struct bpf_cgroup_storage *cgroup_storage[MAX_BPF_CGROUP_STORAGE_TYPE]; u64 bpf_cookie; }; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; struct bpf_prog_array { diff --git a/include/linux/bpf_local_storage.h b/include/linux/bpf_local_storage.h index 173ec7f43ed1..c2ae8788a5cb 100644 --- a/include/linux/bpf_local_storage.h +++ b/include/linux/bpf_local_storage.h @@ -14,6 +14,7 @@ #include <linux/hash.h> #include <linux/types.h> #include <linux/bpf_mem_alloc.h> +#include <linux/kabi.h> #include <uapi/linux/btf.h> #define BPF_LOCAL_STORAGE_CACHE_SIZE 16 @@ -59,6 +60,11 @@ struct bpf_local_storage_map { struct bpf_mem_alloc selem_ma; struct bpf_mem_alloc storage_ma; bool bpf_ma; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; struct bpf_local_storage_data { @@ -94,6 +100,11 @@ struct bpf_local_storage { */ struct rcu_head rcu; raw_spinlock_t lock; /* Protect adding/removing from the "list" */ + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; /* U16_MAX is much more than enough for sk local storage diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h index b6e58dab8e27..d2b59c6d9d60 100644 --- a/include/linux/bpf_verifier.h +++ b/include/linux/bpf_verifier.h @@ -8,6 +8,7 @@ #include <linux/btf.h> /* for struct btf and btf_id() */ #include <linux/filter.h> /* for MAX_BPF_STACK */ #include <linux/tnum.h> +#include <linux/kabi.h> /* Maximum variable offset umax_value permitted when resolving memory accesses. * In practice this is far bigger than any realistic pointer offset; this limit @@ -225,6 +226,11 @@ struct bpf_reg_state { enum bpf_reg_liveness live; /* if (!precise && SCALAR_VALUE) min/max/tnum don't affect safety */ bool precise; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; enum bpf_stack_slot_type { @@ -273,6 +279,11 @@ struct bpf_reference_state { * exiting a callback function. */ int callback_ref; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; /* state of the program: @@ -306,6 +317,11 @@ struct bpf_func_state { struct bpf_reference_state *refs; int allocated_stack; struct bpf_stack_state *stack; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; struct bpf_idx_pair { @@ -383,6 +399,11 @@ struct bpf_verifier_state { */ struct bpf_idx_pair *jmp_history; u32 jmp_history_cnt; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; #define bpf_get_spilled_reg(slot, frame) \ @@ -490,6 +511,11 @@ struct bpf_insn_aux_data { * this instruction, regardless of any heuristics */ bool force_checkpoint; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; #define MAX_USED_MAPS 64 /* max number of maps accessed by one eBPF program */ @@ -541,6 +567,11 @@ struct bpf_subprog_info { bool tail_call_reachable; bool has_ld_abs; bool is_async_cb; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; struct bpf_verifier_env; @@ -640,6 +671,15 @@ struct bpf_verifier_env { * e.g., in reg_type_str() to generate reg_type string */ char tmp_str_buf[TMP_STR_BUF_LEN]; + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) + KABI_RESERVE(5) + KABI_RESERVE(6) + KABI_RESERVE(7) + KABI_RESERVE(8) }; __printf(2, 0) void bpf_verifier_vlog(struct bpf_verifier_log *log, diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 9dd0b85549b6..56c3fc8e4123 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -944,6 +944,11 @@ enum bpf_map_type { BPF_MAP_TYPE_BLOOM_FILTER, BPF_MAP_TYPE_USER_RINGBUF, BPF_MAP_TYPE_CGRP_STORAGE, + + BPF_MAP_TYPE_RESERVE_1, + BPF_MAP_TYPE_RESERVE_2, + BPF_MAP_TYPE_RESERVE_3, + BPF_MAP_TYPE_RESERVE_4, }; /* Note that tracing related programs such as @@ -989,6 +994,11 @@ enum bpf_prog_type { BPF_PROG_TYPE_SYSCALL, /* a program that can execute syscalls */ BPF_PROG_TYPE_NETFILTER, BPF_PROG_TYPE_SCHED, + + BPF_PROG_TYPE_RESERVE_1, + BPF_PROG_TYPE_RESERVE_2, + BPF_PROG_TYPE_RESERVE_3, + BPF_PROG_TYPE_RESERVE_4, }; enum bpf_attach_type { @@ -1042,6 +1052,12 @@ enum bpf_attach_type { BPF_TCX_EGRESS, BPF_TRACE_UPROBE_MULTI, BPF_SCHED, + + BPF_ATTACH_TYPE_RESERVE_1, + BPF_ATTACH_TYPE_RESERVE_2, + BPF_ATTACH_TYPE_RESERVE_3, + BPF_ATTACH_TYPE_RESERVE_4, + __MAX_BPF_ATTACH_TYPE }; @@ -1061,6 +1077,12 @@ enum bpf_link_type { BPF_LINK_TYPE_NETFILTER = 10, BPF_LINK_TYPE_TCX = 11, BPF_LINK_TYPE_UPROBE_MULTI = 12, + + BPF_LINK_TYPE_RESERVE_1 = 13, + BPF_LINK_TYPE_RESERVE_2 = 14, + BPF_LINK_TYPE_RESERVE_3 = 15, + BPF_LINK_TYPE_RESERVE_4 = 16, + MAX_BPF_LINK_TYPE, }; -- 2.34.1

1 0

[PATCH OLK-6.6 v1 0/2] Reserve syscall entries for kabi compatibility
by Wei Li 24 Jan '24

24 Jan '24

In this patch set, we: 1. Fix map_shadow_stack() syscall reserving issue 2. Reserve 15 syscall entries for kabi compatibility Sohil Mehta (1): arch: Reserve map_shadow_stack() syscall number for all architectures Wei Li (1): kabi: Reserve syscall entries for kabi compatibility arch/alpha/kernel/syscalls/syscall.tbl | 1 + arch/arm/tools/syscall.tbl | 16 ++++++++ arch/arm64/include/asm/unistd.h | 2 +- arch/arm64/include/asm/unistd32.h | 32 ++++++++++++++++ arch/m68k/kernel/syscalls/syscall.tbl | 1 + arch/microblaze/kernel/syscalls/syscall.tbl | 1 + arch/mips/kernel/syscalls/syscall_n32.tbl | 1 + arch/mips/kernel/syscalls/syscall_n64.tbl | 1 + arch/mips/kernel/syscalls/syscall_o32.tbl | 1 + arch/parisc/kernel/syscalls/syscall.tbl | 1 + arch/powerpc/kernel/syscalls/syscall.tbl | 16 ++++++++ arch/s390/kernel/syscalls/syscall.tbl | 1 + arch/sh/kernel/syscalls/syscall.tbl | 1 + arch/sparc/kernel/syscalls/syscall.tbl | 1 + arch/x86/entry/syscalls/syscall_32.tbl | 16 ++++++++ arch/x86/entry/syscalls/syscall_64.tbl | 15 ++++++++ arch/xtensa/kernel/syscalls/syscall.tbl | 1 + include/uapi/asm-generic/unistd.h | 36 +++++++++++++++++- tools/include/uapi/asm-generic/unistd.h | 37 +++++++++++++++++-- .../arch/powerpc/entry/syscalls/syscall.tbl | 16 ++++++++ .../arch/x86/entry/syscalls/syscall_64.tbl | 15 ++++++++ 21 files changed, 207 insertions(+), 5 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10] sched/fair: Fix qos_timer deadlock when cpuhp offline
by Zhao Wenhui 24 Jan '24

24 Jan '24

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8WTPH -------------------------------- When cpu hotplug offline, if qos_overload_timer_handler() concurrently running, It could trigger an ABBA deadlock. As qos_overload_timer_handler() requires rq lock, while the cpu hotplug attached firstly and waiting for the qos_timer handling, This can cause Hard LOCKUP like: [359230.788754] Call trace: [359230.788755] hrtimer_active+0x7c/0xec [359230.788757] hrtimer_cancel+0x3c/0x60 [359230.788758] unthrottle_qos_cfs_rqs+0xbc/0x110 [359230.788760] unthrottle_offline_cfs_rqs+0x40/0x150 [359230.788762] rq_offline_fair+0x60/0x70 [359230.788764] set_rq_offline.part.0+0x54/0xf4 [359230.788765] set_rq_offline+0x34/0x44 [359230.788767] rq_attach_root+0x1e8/0x260 [359230.788768] cpu_attach_domain+0x244/0x430 [359230.788770] detach_destroy_domains+0xbc/0x140 [359230.788772] partition_sched_domains_locked+0x23c/0x314 [359230.788774] rebuild_sched_domains_locked+0x1f0/0x270 [359230.788776] cpuset_hotplug_workfn+0x514/0x74c [359230.788777] process_one_work+0x34c/0x800 [359230.788779] worker_thread+0xa8/0x500 [359230.788780] kthread+0x1e0/0x220 [359230.788782] ret_from_fork+0x10/0x18 [359230.788783] Kernel panic - not syncing: Hard LOCKUP Fix it by switch to use __unthrottle_qos_cfs_rqs(), instead of unthrottle_qos_cfs_rqs() in unthrottle_offline_cfs_rqs, so that it will not trigger cancel_qos_timer() when cpu hotplug offline. Fixes: c62a5f1384b9 ("sched/qos: Add qos_tg_{throttle,unthrottle}_{up,down}") Signed-off-by: Zhao Wenhui <zhaowenhui8(a)huawei.com> --- kernel/sched/fair.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 16f6720a244c..f39e7547523c 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -131,6 +131,7 @@ static DEFINE_PER_CPU_SHARED_ALIGNED(struct hrtimer, qos_overload_timer); 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 */ +static int __unthrottle_qos_cfs_rqs(int cpu); static int unthrottle_qos_cfs_rqs(int cpu); static bool qos_smt_expelled(int this_cpu); #endif @@ -5733,7 +5734,7 @@ static void __maybe_unused unthrottle_offline_cfs_rqs(struct rq *rq) lockdep_assert_rq_held(rq); #ifdef CONFIG_QOS_SCHED - unthrottle_qos_cfs_rqs(cpu_of(rq)); + __unthrottle_qos_cfs_rqs(cpu_of(rq)); #endif rcu_read_lock(); -- 2.34.1

2 1

[PATCH openEuler-1.0-LTS] sched/fair: Fix qos_timer deadlock when cpuhp offline
by Zhao Wenhui 24 Jan '24

24 Jan '24

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8YHBT -------------------------------- When cpu hotplug offline, if qos_overload_timer_handler() concurrently running, It could trigger an ABBA deadlock. As qos_overload_timer_handler() requires rq lock, while the cpu hotplug attached firstly and waiting for the qos_timer handling, This can cause Hard LOCKUP like: [359230.788754] Call trace: [359230.788755] hrtimer_active+0x7c/0xec [359230.788757] hrtimer_cancel+0x3c/0x60 [359230.788758] unthrottle_qos_cfs_rqs+0xbc/0x110 [359230.788760] unthrottle_offline_cfs_rqs+0x40/0x150 [359230.788762] rq_offline_fair+0x60/0x70 [359230.788764] set_rq_offline.part.0+0x54/0xf4 [359230.788765] set_rq_offline+0x34/0x44 [359230.788767] rq_attach_root+0x1e8/0x260 [359230.788768] cpu_attach_domain+0x244/0x430 [359230.788770] detach_destroy_domains+0xbc/0x140 [359230.788772] partition_sched_domains_locked+0x23c/0x314 [359230.788774] rebuild_sched_domains_locked+0x1f0/0x270 [359230.788776] cpuset_hotplug_workfn+0x514/0x74c [359230.788777] process_one_work+0x34c/0x800 [359230.788779] worker_thread+0xa8/0x500 [359230.788780] kthread+0x1e0/0x220 [359230.788782] ret_from_fork+0x10/0x18 [359230.788783] Kernel panic - not syncing: Hard LOCKUP Fix it by switch to use __unthrottle_qos_cfs_rqs(), instead of unthrottle_qos_cfs_rqs() in unthrottle_offline_cfs_rqs, so that it will not trigger cancel_qos_timer() when cpu hotplug offline. Fixes: 453eaea60e92 ("sched/qos: Add qos_tg_{throttle,unthrottle}_{up,down}") Signed-off-by: Zhao Wenhui <zhaowenhui8(a)huawei.com> --- kernel/sched/fair.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index ce6bb4304949..6d0ec315f7be 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -116,7 +116,7 @@ static DEFINE_PER_CPU_SHARED_ALIGNED(struct hrtimer, qos_overload_timer); 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 */ -static int unthrottle_qos_cfs_rqs(int cpu); +static int __unthrottle_qos_cfs_rqs(int cpu); #endif #ifdef CONFIG_CFS_BANDWIDTH @@ -5218,7 +5218,7 @@ static void __maybe_unused unthrottle_offline_cfs_rqs(struct rq *rq) lockdep_assert_held(&rq->lock); #ifdef CONFIG_QOS_SCHED - unthrottle_qos_cfs_rqs(cpu_of(rq)); + __unthrottle_qos_cfs_rqs(cpu_of(rq)); #endif rcu_read_lock(); -- 2.34.1

2 1

[PATCH OLK-6.6] tlb: reserve fields for struct mmu_gather
by Jinjiang Tu 24 Jan '24

24 Jan '24

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8YJQK -------------------------------- reserve fields for struct mmu_gather. Signed-off-by: Jinjiang Tu <tujinjiang(a)huawei.com> --- include/asm-generic/tlb.h | 3 +++ 1 file changed, 3 insertions(+) diff --git a/include/asm-generic/tlb.h b/include/asm-generic/tlb.h index 129a3a759976..1a3191e844b7 100644 --- a/include/asm-generic/tlb.h +++ b/include/asm-generic/tlb.h @@ -14,6 +14,7 @@ #include <linux/mmu_notifier.h> #include <linux/swap.h> #include <linux/hugetlb_inline.h> +#include <linux/kabi.h> #include <asm/tlbflush.h> #include <asm/cacheflush.h> @@ -348,6 +349,8 @@ struct mmu_gather { unsigned int page_size; #endif #endif + KABI_RESERVE(1) + KABI_RESERVE(2) }; void tlb_flush_mmu(struct mmu_gather *tlb); -- 2.25.1

3 2

[PATCH OLK-6.6] arch/mm/fault: fix major fault accounting when retrying under per-VMA lock
by Tong Tiangen 24 Jan '24

24 Jan '24

From: Suren Baghdasaryan <surenb(a)google.com> mainline inclusion from mainline-v6.7 commit 46e714c729c8d1d8110bc0545d7ffe8a759c9dc0 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8YKL3 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- A test [1] in Android test suite started failing after [2] was merged. It turns out that after handling a major fault under per-VMA lock, the process major fault counter does not register that fault as major. Before [2] read faults would be done under mmap_lock, in which case FAULT_FLAG_TRIED flag is set before retrying. That in turn causes mm_account_fault() to account the fault as major once retry completes. With per-VMA locks we often retry because a fault can't be handled without locking the whole mm using mmap_lock. Therefore such retries do not set FAULT_FLAG_TRIED flag. This logic does not work after [2] because we can now handle read major faults under per-VMA lock and upon retry the fact there was a major fault gets lost. Fix this by setting FAULT_FLAG_TRIED after retrying under per-VMA lock if VM_FAULT_MAJOR was returned. Ideally we would use an additional VM_FAULT bit to indicate the reason for the retry (could not handle under per-VMA lock vs other reason) but this simpler solution seems to work, so keeping it simple. [1] https://cs.android.com/android/platform/superproject/+/master:test/vts-test… [2] https://lore.kernel.org/all/20231006195318.4087158-6-willy@infradead.org/ Link: https://lkml.kernel.org/r/20231226214610.109282-1-surenb@google.com Fixes: 12214eba1992 ("mm: handle read faults under the VMA lock") Signed-off-by: Suren Baghdasaryan <surenb(a)google.com> Cc: Matthew Wilcox <willy(a)infradead.org> Cc: Alexander Gordeev <agordeev(a)linux.ibm.com> Cc: Andy Lutomirski <luto(a)kernel.org> Cc: Catalin Marinas <catalin.marinas(a)arm.com> Cc: Christophe Leroy <christophe.leroy(a)csgroup.eu> Cc: Dave Hansen <dave.hansen(a)linux.intel.com> Cc: Gerald Schaefer <gerald.schaefer(a)linux.ibm.com> Cc: Michael Ellerman <mpe(a)ellerman.id.au> Cc: Palmer Dabbelt <palmer(a)dabbelt.com> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Will Deacon <will(a)kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> conflict: arch/s390/mm/fault.c Signed-off-by: Tong Tiangen <tongtiangen(a)huawei.com> --- arch/arm64/mm/fault.c | 2 ++ arch/powerpc/mm/fault.c | 2 ++ arch/riscv/mm/fault.c | 2 ++ arch/s390/mm/fault.c | 3 +++ arch/x86/mm/fault.c | 2 ++ 5 files changed, 11 insertions(+) diff --git a/arch/arm64/mm/fault.c b/arch/arm64/mm/fault.c index 171810268f6f..c376e58e8cf0 100644 --- a/arch/arm64/mm/fault.c +++ b/arch/arm64/mm/fault.c @@ -633,6 +633,8 @@ static int __kprobes do_page_fault(unsigned long far, unsigned long esr, goto done; } count_vm_vma_lock_event(VMA_LOCK_RETRY); + if (fault & VM_FAULT_MAJOR) + mm_flags |= FAULT_FLAG_TRIED; /* Quick path to respond to signals */ if (fault_signal_pending(fault, regs)) { diff --git a/arch/powerpc/mm/fault.c b/arch/powerpc/mm/fault.c index b1723094d464..ec23164ad768 100644 --- a/arch/powerpc/mm/fault.c +++ b/arch/powerpc/mm/fault.c @@ -496,6 +496,8 @@ static int ___do_page_fault(struct pt_regs *regs, unsigned long address, goto done; } count_vm_vma_lock_event(VMA_LOCK_RETRY); + if (fault & VM_FAULT_MAJOR) + flags |= FAULT_FLAG_TRIED; if (fault_signal_pending(fault, regs)) return user_mode(regs) ? 0 : SIGBUS; diff --git a/arch/riscv/mm/fault.c b/arch/riscv/mm/fault.c index 90d4ba36d1d0..081339ddf47e 100644 --- a/arch/riscv/mm/fault.c +++ b/arch/riscv/mm/fault.c @@ -304,6 +304,8 @@ void handle_page_fault(struct pt_regs *regs) goto done; } count_vm_vma_lock_event(VMA_LOCK_RETRY); + if (fault & VM_FAULT_MAJOR) + flags |= FAULT_FLAG_TRIED; if (fault_signal_pending(fault, regs)) { if (!user_mode(regs)) diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c index b678295931c3..f5463535013a 100644 --- a/arch/s390/mm/fault.c +++ b/arch/s390/mm/fault.c @@ -424,6 +424,9 @@ static inline vm_fault_t do_exception(struct pt_regs *regs, int access) goto out; } count_vm_vma_lock_event(VMA_LOCK_RETRY); + if (fault & VM_FAULT_MAJOR) + flags |= FAULT_FLAG_TRIED; + /* Quick path to respond to signals */ if (fault_signal_pending(fault, regs)) { fault = VM_FAULT_SIGNAL; diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c index ab778eac1952..679b09cfe241 100644 --- a/arch/x86/mm/fault.c +++ b/arch/x86/mm/fault.c @@ -1370,6 +1370,8 @@ void do_user_addr_fault(struct pt_regs *regs, goto done; } count_vm_vma_lock_event(VMA_LOCK_RETRY); + if (fault & VM_FAULT_MAJOR) + flags |= FAULT_FLAG_TRIED; /* Quick path to respond to signals */ if (fault_signal_pending(fault, regs)) { -- 2.25.1

2 1

[PATCH openEuler-1.0-LTS] fs:/dcache.c: fix negative dentry limit not complete problem
by Long Li 24 Jan '24

24 Jan '24

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8WPQW CVE: NA -------------------------------- Restrictions of negative dentry can avoid softlock when too many dentry in memory, but restrictions before do not always take effect. For example, removing files which has been created would not enter retain_dentry(), because the dentry of file maybe has been added to lru list of superblock, it caused that fast_dput() reutrn true in last dput(). So, add restriction logic for the file which has been added to the lru list in fast_dput(), it prevents the negative dentry exceeding the limit when deleting existing file. Factor out negative dentry limit logic into limit_negative_dentry(), make the code more clean. Fixes: 7ba5d5d0e8b5 ("fs/dcache.c: avoid softlock since too many negative dentry") Signed-off-by: Long Li <leo.lilong(a)huawei.com> --- fs/dcache.c | 58 +++++++++++++++++++++++++++++++++-------------------- 1 file changed, 36 insertions(+), 22 deletions(-) diff --git a/fs/dcache.c b/fs/dcache.c index ef1641398eda..e52e3abdf378 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -636,10 +636,36 @@ static inline struct dentry *lock_parent(struct dentry *dentry) return __lock_parent(dentry); } -static inline bool retain_dentry(struct dentry *dentry) +/* + * Return true if dentry is negative and exceed negative dentry limit. + */ +static inline bool limit_negative_dentry(struct dentry *dentry) { struct dentry *parent; + parent = dentry->d_parent; + if (unlikely(!parent)) + return false; + + WARN_ON((atomic_read(&parent->d_neg_dnum) < 0)); + if (!dentry->d_inode) { + if (!(dentry->d_flags & DCACHE_NEGATIVE_ACCOUNT)) { + unsigned int flags = READ_ONCE(dentry->d_flags); + + flags |= DCACHE_NEGATIVE_ACCOUNT; + WRITE_ONCE(dentry->d_flags, flags); + atomic_inc(&parent->d_neg_dnum); + } + + if (atomic_read(&parent->d_neg_dnum) >= NEG_DENTRY_LIMIT) + return true; + } + + return false; +} + +static inline bool retain_dentry(struct dentry *dentry) +{ WARN_ON(d_in_lookup(dentry)); /* Unreachable? Get rid of it */ @@ -654,27 +680,9 @@ static inline bool retain_dentry(struct dentry *dentry) return false; } - if (unlikely(!dentry->d_parent)) - goto noparent; - - parent = dentry->d_parent; - /* Return false if it's negative */ - WARN_ON((atomic_read(&parent->d_neg_dnum) < 0)); - if (!dentry->d_inode) { - if (!(dentry->d_flags & DCACHE_NEGATIVE_ACCOUNT)) { - unsigned flags = READ_ONCE(dentry->d_flags); - - flags |= DCACHE_NEGATIVE_ACCOUNT; - WRITE_ONCE(dentry->d_flags, flags); - atomic_inc(&parent->d_neg_dnum); - } - } - - if (!dentry->d_inode && - atomic_read(&parent->d_neg_dnum) >= NEG_DENTRY_LIMIT) + if (unlikely(limit_negative_dentry(dentry))) return false; -noparent: /* retain; LRU fodder */ dentry->d_lockref.count--; if (unlikely(!(dentry->d_flags & DCACHE_LRU_LIST))) @@ -808,8 +816,14 @@ static inline bool fast_dput(struct dentry *dentry) d_flags &= DCACHE_REFERENCED | DCACHE_LRU_LIST | DCACHE_DISCONNECTED; /* Nothing to do? Dropping the reference was all we needed? */ - if (d_flags == (DCACHE_REFERENCED | DCACHE_LRU_LIST) && !d_unhashed(dentry)) - return true; + if (d_flags == (DCACHE_REFERENCED | DCACHE_LRU_LIST) && !d_unhashed(dentry)) { + /* + * If dentry is negative and need to be limitted, it maybe need + * to be killed, so shouldn't fast put and retain in memory. + */ + if (likely(!limit_negative_dentry(dentry))) + return true; + } /* * Not the fast normal case? Get the lock. We've already decremented -- 2.31.1

2 1

[PATCH openEuler-1.0-LTS] sched/fair: Fix qos_timer deadlock when cpuhp offline
by Zhao Wenhui 24 Jan '24

24 Jan '24

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8YHBT -------------------------------- When cpu hotplug offline, if qos_overload_timer_handler() concurrently running, It could trigger an ABBA deadlock. As qos_overload_timer_handler() requires rq lock, while the cpu hotplug attached firstly and waiting for the qos_timer handling, This can cause Hard LOCKUP like: [359230.788754] Call trace: [359230.788755] hrtimer_active+0x7c/0xec [359230.788757] hrtimer_cancel+0x3c/0x60 [359230.788758] unthrottle_qos_cfs_rqs+0xbc/0x110 [359230.788760] unthrottle_offline_cfs_rqs+0x40/0x150 [359230.788762] rq_offline_fair+0x60/0x70 [359230.788764] set_rq_offline.part.0+0x54/0xf4 [359230.788765] set_rq_offline+0x34/0x44 [359230.788767] rq_attach_root+0x1e8/0x260 [359230.788768] cpu_attach_domain+0x244/0x430 [359230.788770] detach_destroy_domains+0xbc/0x140 [359230.788772] partition_sched_domains_locked+0x23c/0x314 [359230.788774] rebuild_sched_domains_locked+0x1f0/0x270 [359230.788776] cpuset_hotplug_workfn+0x514/0x74c [359230.788777] process_one_work+0x34c/0x800 [359230.788779] worker_thread+0xa8/0x500 [359230.788780] kthread+0x1e0/0x220 [359230.788782] ret_from_fork+0x10/0x18 [359230.788783] Kernel panic - not syncing: Hard LOCKUP Fix it by switch to use __unthrottle_qos_cfs_rqs(), instead of unthrottle_qos_cfs_rqs() in unthrottle_offline_cfs_rqs, so that it will not trigger cancel_qos_timer() when cpu hotplug offline. Fixes: 453eaea60e92 ("sched/qos: Add qos_tg_{throttle,unthrottle}_{up,down}") Signed-off-by: Zhao Wenhui <zhaowenhui8(a)huawei.com> --- kernel/sched/fair.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index ce6bb4304949..81e3106962e3 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -116,6 +116,7 @@ static DEFINE_PER_CPU_SHARED_ALIGNED(struct hrtimer, qos_overload_timer); 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 */ +static int __unthrottle_qos_cfs_rqs(int cpu); static int unthrottle_qos_cfs_rqs(int cpu); #endif @@ -5218,7 +5219,7 @@ static void __maybe_unused unthrottle_offline_cfs_rqs(struct rq *rq) lockdep_assert_held(&rq->lock); #ifdef CONFIG_QOS_SCHED - unthrottle_qos_cfs_rqs(cpu_of(rq)); + __unthrottle_qos_cfs_rqs(cpu_of(rq)); #endif rcu_read_lock(); -- 2.34.1

2 1

[PATCH OLK-6.6 v2 0/4] reserve space for arm64 related structures.
by Yuntao Liu 24 Jan '24

24 Jan '24

Yuntao Liu (4): kabi: reserve space for struct efi.h:efi kabi: reserve space for struct extable.h:exception_table_entry kabi: reserve space for struct fb.h:fb_deferred_io fb.h:fb_info kabi: reserve space for struct processor.h:thread_struct arch/arm64/include/asm/extable.h | 3 +++ arch/arm64/include/asm/processor.h | 9 +++++++++ include/linux/efi.h | 3 +++ include/linux/fb.h | 7 +++++++ 4 files changed, 22 insertions(+) -- 2.34.1

2 5

openEuler Kernel SIG双周例会
by openEuler conference 24 Jan '24

24 Jan '24

您好！ Kernel 邀请您参加 2024-01-26 14:00 召开的WeLink会议(自动录制) 会议主题：openEuler Kernel SIG双周例会会议内容： 1. 进展update 2. 议题征集中（新增议题可直接回复本邮件申报，或者填入会议看板）会议链接：https://bmeeting.huaweicloud.com:36443/#/j/967727869 会议纪要：https://etherpad.openeuler.org/p/Kernel-meetings 更多资讯尽在：https://www.openeuler.org/zh/ Hello! Kernel invites you to attend the WeLink conference(auto recording) will be held at 2024-01-26 14:00, The subject of the conference is openEuler Kernel SIG双周例会, Summary: 1. 进展update 2. 议题征集中（新增议题可直接回复本邮件申报，或者填入会议看板） You can join the meeting at https://bmeeting.huaweicloud.com:36443/#/j/967727869. Add topics at https://etherpad.openeuler.org/p/Kernel-meetings. More information: https://www.openeuler.org/en/

1 0