July 2022 - Kernel - mailweb.openeuler.org

[PATCH openEuler-22.09 v2 0/8] bpf: sched: Introduce Schduler BPF
by Zheng Zengkai 22 Jul '22

22 Jul '22

This patchset aims to start a discussion about potential applications of BPF to the scheduler. It also aims to land some very basic BPF infrastructure necessary to add new BPF hooks to the scheduler, a minimal set of useful helpers, corresponding libbpf changes, etc. [Testing] kernel options: CONFIG_BPF_SCHED=y CONFIG_BPF_SYSCALL=y CONFIG_BPF_EVENTS=y Test passed with below step: 1.cd tools/testing/selftests/bpf & make 2.run cmd: ./test_progs -t test_sched 3.it will show like this: # ./test_progs -t test_sched #113/1 sched_tgidpid_mode:OK #113/2 sched_cgid_mode:OK #113 test_sched:OK Summary: 1/2 PASSED, 0 SKIPPED, 0 FAILED Chen Hui (1): sched: Move some definition to sched.h Ren Zhijie (1): bpf: sched: add selftests for BPF_PROG_TYPE_SCHED Roman Gushchin (6): bpf: sched: basic infrastructure for scheduler bpf bpf: sched: add convenient helpers to identify sched entities bpf: sched: introduce bpf_sched_enable() sched: cfs: add bpf hooks to control wakeup and tick preemption libbpf: add support for scheduler bpf programs bpftool: recognize scheduler programs include/linux/bpf_sched.h | 50 +++++ include/linux/bpf_types.h | 4 + include/linux/sched_hook_defs.h | 5 + include/uapi/linux/bpf.h | 25 +++ init/Kconfig | 10 + kernel/bpf/btf.c | 1 + kernel/bpf/syscall.c | 20 ++ kernel/bpf/trampoline.c | 1 + kernel/bpf/verifier.c | 11 +- kernel/sched/Makefile | 1 + kernel/sched/bpf_sched.c | 138 ++++++++++++ kernel/sched/fair.c | 79 +++---- kernel/sched/sched.h | 52 +++++ scripts/bpf_helpers_doc.py | 2 + tools/bpf/bpftool/common.c | 1 + tools/bpf/bpftool/prog.c | 1 + tools/include/uapi/linux/bpf.h | 25 +++ tools/lib/bpf/bpf.c | 3 +- tools/lib/bpf/libbpf.c | 33 ++- tools/lib/bpf/libbpf.h | 4 + tools/lib/bpf/libbpf.map | 3 + .../selftests/bpf/prog_tests/test_sched.c | 160 ++++++++++++++ tools/testing/selftests/bpf/progs/sched.c | 203 ++++++++++++++++++ 23 files changed, 780 insertions(+), 52 deletions(-) create mode 100644 include/linux/bpf_sched.h create mode 100644 include/linux/sched_hook_defs.h create mode 100644 kernel/sched/bpf_sched.c create mode 100644 tools/testing/selftests/bpf/prog_tests/test_sched.c create mode 100644 tools/testing/selftests/bpf/progs/sched.c -- 2.20.1

1 0

[PATCH openEuler-22.09 v2 1/8] bpf: sched: basic infrastructure for scheduler bpf
by Zheng Zengkai 22 Jul '22

22 Jul '22

From: Roman Gushchin <guro(a)fb.com> maillist inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I5F6X6 CVE: NA Reference: https://lore.kernel.org/all/20210916162451.709260-1-guro@fb.com/ ------------------- This commit introduces basic definitions and infrastructure for scheduler bpf programs. It defines the BPF_PROG_TYPE_SCHED program type and the BPF_SCHED attachment type. The implementation is inspired by lsm bpf programs and is based on kretprobes. This will allow to add new hooks with a minimal changes to the kernel code and without any changes to libbpf/bpftool. It's very convenient as I anticipate a large number of private patches being used for a long time before (or if at all) reaching upstream. Sched programs are expected to return an int, which meaning will be context defined. This patch doesn't add any real scheduler hooks (only a stub), it will be done by following patches in the series. Scheduler bpf programs as now are very restricted in what they can do: only the bpf_printk() helper is available. The scheduler context can impose significant restrictions on what's safe and what's not. So let's extend their abilities on case by case basis when a need arise. Signed-off-by: Roman Gushchin <guro(a)fb.com> Signed-off-by: Chen Hui <judy.chenhui(a)huawei.com> Signed-off-by: Ren Zhijie <renzhijie2(a)huawei.com> --- include/linux/bpf_sched.h | 26 ++++++++++++++ include/linux/bpf_types.h | 4 +++ include/linux/sched_hook_defs.h | 2 ++ include/uapi/linux/bpf.h | 2 ++ init/Kconfig | 10 ++++++ kernel/bpf/btf.c | 1 + kernel/bpf/syscall.c | 9 +++++ kernel/bpf/trampoline.c | 1 + kernel/bpf/verifier.c | 11 +++++- kernel/sched/Makefile | 1 + kernel/sched/bpf_sched.c | 62 +++++++++++++++++++++++++++++++++ tools/include/uapi/linux/bpf.h | 2 ++ 12 files changed, 130 insertions(+), 1 deletion(-) create mode 100644 include/linux/bpf_sched.h create mode 100644 include/linux/sched_hook_defs.h create mode 100644 kernel/sched/bpf_sched.c diff --git a/include/linux/bpf_sched.h b/include/linux/bpf_sched.h new file mode 100644 index 000000000000..874393e6a6aa --- /dev/null +++ b/include/linux/bpf_sched.h @@ -0,0 +1,26 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _LINUX_BPF_SCHED_H +#define _LINUX_BPF_SCHED_H + +#include <linux/bpf.h> + +#ifdef CONFIG_BPF_SCHED + +#define BPF_SCHED_HOOK(RET, DEFAULT, NAME, ...) \ + RET bpf_sched_##NAME(__VA_ARGS__); +#include <linux/sched_hook_defs.h> +#undef BPF_SCHED_HOOK + +int bpf_sched_verify_prog(struct bpf_verifier_log *vlog, + const struct bpf_prog *prog); + +#else /* !CONFIG_BPF_SCHED */ + +static inline int bpf_sched_verify_prog(struct bpf_verifier_log *vlog, + const struct bpf_prog *prog) +{ + return -EOPNOTSUPP; +} + +#endif /* CONFIG_BPF_SCHED */ +#endif /* _LINUX_BPF_SCHED_H */ diff --git a/include/linux/bpf_types.h b/include/linux/bpf_types.h index a8137bb6dd3c..5732b485c539 100644 --- a/include/linux/bpf_types.h +++ b/include/linux/bpf_types.h @@ -77,6 +77,10 @@ BPF_PROG_TYPE(BPF_PROG_TYPE_LSM, lsm, void *, void *) #endif /* CONFIG_BPF_LSM */ #endif +#ifdef CONFIG_BPF_SCHED +BPF_PROG_TYPE(BPF_PROG_TYPE_SCHED, bpf_sched, + void *, void *) +#endif /* CONFIG_BPF_SCHED */ BPF_MAP_TYPE(BPF_MAP_TYPE_ARRAY, array_map_ops) BPF_MAP_TYPE(BPF_MAP_TYPE_PERCPU_ARRAY, percpu_array_map_ops) diff --git a/include/linux/sched_hook_defs.h b/include/linux/sched_hook_defs.h new file mode 100644 index 000000000000..14344004e335 --- /dev/null +++ b/include/linux/sched_hook_defs.h @@ -0,0 +1,2 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +BPF_SCHED_HOOK(int, 0, dummy, void) diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 8fae845d80e2..4f8f3f2113a7 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -199,6 +199,7 @@ enum bpf_prog_type { BPF_PROG_TYPE_EXT, BPF_PROG_TYPE_LSM, BPF_PROG_TYPE_SK_LOOKUP, + BPF_PROG_TYPE_SCHED, }; enum bpf_attach_type { @@ -240,6 +241,7 @@ enum bpf_attach_type { BPF_XDP_CPUMAP, BPF_SK_LOOKUP, BPF_XDP, + BPF_SCHED, __MAX_BPF_ATTACH_TYPE }; diff --git a/init/Kconfig b/init/Kconfig index 27c5ed16fef1..c3fa3240d87d 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -1749,6 +1749,16 @@ config BPF_LSM If you are unsure how to answer this question, answer N. +config BPF_SCHED + bool "SCHED Instrumentation with BPF" + depends on BPF_EVENTS + depends on BPF_SYSCALL + help + Enables instrumentation of the sched hooks with eBPF programs for + implementing dynamic scheduling Policies. + + If you are unsure how to answer this question, answer N. + config BPF_SYSCALL bool "Enable bpf() system call" select BPF diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c index fba28f17e61a..9a0a9895ec62 100644 --- a/kernel/bpf/btf.c +++ b/kernel/bpf/btf.c @@ -4479,6 +4479,7 @@ bool btf_ctx_access(int off, int size, enum bpf_access_type type, return true; t = btf_type_by_id(btf, t->type); break; + case BPF_SCHED: case BPF_MODIFY_RETURN: /* For now the BPF_MODIFY_RETURN can only be attached to * functions that return an int. diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 419dbc3d060e..ff65862ae5ce 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -1997,6 +1997,7 @@ bpf_prog_load_check_attach(enum bpf_prog_type prog_type, case BPF_PROG_TYPE_LSM: case BPF_PROG_TYPE_STRUCT_OPS: case BPF_PROG_TYPE_EXT: + case BPF_PROG_TYPE_SCHED: break; default: return -EINVAL; @@ -2108,6 +2109,7 @@ static bool is_perfmon_prog_type(enum bpf_prog_type prog_type) case BPF_PROG_TYPE_LSM: case BPF_PROG_TYPE_STRUCT_OPS: /* has access to struct sock */ case BPF_PROG_TYPE_EXT: /* extends any prog */ + case BPF_PROG_TYPE_SCHED: return true; default: return false; @@ -2608,6 +2610,12 @@ static int bpf_tracing_prog_attach(struct bpf_prog *prog, goto out_put_prog; } break; + case BPF_PROG_TYPE_SCHED: + if (prog->expected_attach_type != BPF_SCHED) { + err = -EINVAL; + goto out_put_prog; + } + break; default: err = -EINVAL; goto out_put_prog; @@ -2838,6 +2846,7 @@ static int bpf_raw_tracepoint_open(const union bpf_attr *attr) case BPF_PROG_TYPE_TRACING: case BPF_PROG_TYPE_EXT: case BPF_PROG_TYPE_LSM: + case BPF_PROG_TYPE_SCHED: if (attr->raw_tracepoint.name) { /* The attach point for this category of programs * should be specified via btf_id during program load. diff --git a/kernel/bpf/trampoline.c b/kernel/bpf/trampoline.c index 986dabc3d11f..cc6ba35a1d14 100644 --- a/kernel/bpf/trampoline.c +++ b/kernel/bpf/trampoline.c @@ -357,6 +357,7 @@ static enum bpf_tramp_prog_type bpf_attach_type_to_tramp(struct bpf_prog *prog) switch (prog->expected_attach_type) { case BPF_TRACE_FENTRY: return BPF_TRAMP_FENTRY; + case BPF_SCHED: case BPF_MODIFY_RETURN: return BPF_TRAMP_MODIFY_RETURN; case BPF_TRACE_FEXIT: diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 6423f1714a2f..d26104b258ba 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -22,6 +22,7 @@ #include <linux/error-injection.h> #include <linux/bpf_lsm.h> #include <linux/btf_ids.h> +#include <linux/bpf_sched.h> #include "disasm.h" @@ -12178,6 +12179,7 @@ int bpf_check_attach_target(struct bpf_verifier_log *log, case BPF_LSM_MAC: case BPF_TRACE_FENTRY: case BPF_TRACE_FEXIT: + case BPF_SCHED: if (!btf_type_is_func(t)) { bpf_log(log, "attach_btf_id %u is not a function\n", btf_id); @@ -12283,7 +12285,8 @@ static int check_attach_btf_id(struct bpf_verifier_env *env) if (prog->type != BPF_PROG_TYPE_TRACING && prog->type != BPF_PROG_TYPE_LSM && - prog->type != BPF_PROG_TYPE_EXT) + prog->type != BPF_PROG_TYPE_EXT && + prog->type != BPF_PROG_TYPE_SCHED) return 0; ret = bpf_check_attach_target(&env->log, prog, tgt_prog, btf_id, &tgt_info); @@ -12323,6 +12326,12 @@ static int check_attach_btf_id(struct bpf_verifier_env *env) return ret; } + if (prog->type == BPF_PROG_TYPE_SCHED) { + ret = bpf_sched_verify_prog(&env->log, prog); + if (ret < 0) + return ret; + } + key = bpf_trampoline_compute_key(tgt_prog, btf_id); tr = bpf_trampoline_get(key, &tgt_info); if (!tr) diff --git a/kernel/sched/Makefile b/kernel/sched/Makefile index 5fc9c9b70862..8ae9e39eb83a 100644 --- a/kernel/sched/Makefile +++ b/kernel/sched/Makefile @@ -36,3 +36,4 @@ obj-$(CONFIG_CPU_FREQ_GOV_SCHEDUTIL) += cpufreq_schedutil.o obj-$(CONFIG_MEMBARRIER) += membarrier.o obj-$(CONFIG_CPU_ISOLATION) += isolation.o obj-$(CONFIG_PSI) += psi.o +obj-$(CONFIG_BPF_SCHED) += bpf_sched.o \ No newline at end of file diff --git a/kernel/sched/bpf_sched.c b/kernel/sched/bpf_sched.c new file mode 100644 index 000000000000..2f05c186cfd0 --- /dev/null +++ b/kernel/sched/bpf_sched.c @@ -0,0 +1,62 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/bpf.h> +#include <linux/cgroup.h> +#include <linux/bpf_verifier.h> +#include <linux/bpf_sched.h> +#include <linux/btf_ids.h> +#include "sched.h" + +/* + * For every hook declare a nop function where a BPF program can be attached. + */ +#define BPF_SCHED_HOOK(RET, DEFAULT, NAME, ...) \ +noinline RET bpf_sched_##NAME(__VA_ARGS__) \ +{ \ + return DEFAULT; \ +} + +#include <linux/sched_hook_defs.h> +#undef BPF_SCHED_HOOK + +#define BPF_SCHED_HOOK(RET, DEFAULT, NAME, ...) BTF_ID(func, bpf_sched_##NAME) +BTF_SET_START(bpf_sched_hooks) +#include <linux/sched_hook_defs.h> +#undef BPF_SCHED_HOOK +BTF_SET_END(bpf_sched_hooks) + +int bpf_sched_verify_prog(struct bpf_verifier_log *vlog, + const struct bpf_prog *prog) +{ + if (!prog->gpl_compatible) { + bpf_log(vlog, + "sched programs must have a GPL compatible license\n"); + return -EINVAL; + } + + if (!btf_id_set_contains(&bpf_sched_hooks, prog->aux->attach_btf_id)) { + bpf_log(vlog, "attach_btf_id %u points to wrong type name %s\n", + prog->aux->attach_btf_id, prog->aux->attach_func_name); + return -EINVAL; + } + + return 0; +} + +static const struct bpf_func_proto * +bpf_sched_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) +{ + switch (func_id) { + case BPF_FUNC_trace_printk: + return bpf_get_trace_printk_proto(); + default: + return NULL; + } +} + +const struct bpf_prog_ops bpf_sched_prog_ops = { +}; + +const struct bpf_verifier_ops bpf_sched_verifier_ops = { + .get_func_proto = bpf_sched_func_proto, + .is_valid_access = btf_ctx_access, +}; diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h index a44cb5155825..7d7c04f698ec 100644 --- a/tools/include/uapi/linux/bpf.h +++ b/tools/include/uapi/linux/bpf.h @@ -199,6 +199,7 @@ enum bpf_prog_type { BPF_PROG_TYPE_EXT, BPF_PROG_TYPE_LSM, BPF_PROG_TYPE_SK_LOOKUP, + BPF_PROG_TYPE_SCHED, }; enum bpf_attach_type { @@ -240,6 +241,7 @@ enum bpf_attach_type { BPF_XDP_CPUMAP, BPF_SK_LOOKUP, BPF_XDP, + BPF_SCHED, __MAX_BPF_ATTACH_TYPE }; -- 2.20.1

1 7

[PATCH openEuler-22.09 1/7] bpf: sched: basic infrastructure for scheduler bpf
by Zheng Zengkai 22 Jul '22

22 Jul '22

From: Roman Gushchin <guro(a)fb.com> maillist inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I5F6X6 CVE: NA Reference: https://lore.kernel.org/all/20210916162451.709260-1-guro@fb.com/ ------------------- This commit introduces basic definitions and infrastructure for scheduler bpf programs. It defines the BPF_PROG_TYPE_SCHED program type and the BPF_SCHED attachment type. The implementation is inspired by lsm bpf programs and is based on kretprobes. This will allow to add new hooks with a minimal changes to the kernel code and without any changes to libbpf/bpftool. It's very convenient as I anticipate a large number of private patches being used for a long time before (or if at all) reaching upstream. Sched programs are expected to return an int, which meaning will be context defined. This patch doesn't add any real scheduler hooks (only a stub), it will be done by following patches in the series. Scheduler bpf programs as now are very restricted in what they can do: only the bpf_printk() helper is available. The scheduler context can impose significant restrictions on what's safe and what's not. So let's extend their abilities on case by case basis when a need arise. Signed-off-by: Roman Gushchin <guro(a)fb.com> Signed-off-by: Chen Hui <judy.chenhui(a)huawei.com> Signed-off-by: Ren Zhijie <renzhijie2(a)huawei.com> --- include/linux/bpf_sched.h | 26 ++++++++++++++ include/linux/bpf_types.h | 4 +++ include/linux/sched_hook_defs.h | 2 ++ include/uapi/linux/bpf.h | 2 ++ init/Kconfig | 10 ++++++ kernel/bpf/btf.c | 1 + kernel/bpf/syscall.c | 9 +++++ kernel/bpf/trampoline.c | 1 + kernel/bpf/verifier.c | 11 +++++- kernel/sched/Makefile | 1 + kernel/sched/bpf_sched.c | 62 +++++++++++++++++++++++++++++++++ tools/include/uapi/linux/bpf.h | 2 ++ 12 files changed, 130 insertions(+), 1 deletion(-) create mode 100644 include/linux/bpf_sched.h create mode 100644 include/linux/sched_hook_defs.h create mode 100644 kernel/sched/bpf_sched.c diff --git a/include/linux/bpf_sched.h b/include/linux/bpf_sched.h new file mode 100644 index 000000000000..874393e6a6aa --- /dev/null +++ b/include/linux/bpf_sched.h @@ -0,0 +1,26 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _LINUX_BPF_SCHED_H +#define _LINUX_BPF_SCHED_H + +#include <linux/bpf.h> + +#ifdef CONFIG_BPF_SCHED + +#define BPF_SCHED_HOOK(RET, DEFAULT, NAME, ...) \ + RET bpf_sched_##NAME(__VA_ARGS__); +#include <linux/sched_hook_defs.h> +#undef BPF_SCHED_HOOK + +int bpf_sched_verify_prog(struct bpf_verifier_log *vlog, + const struct bpf_prog *prog); + +#else /* !CONFIG_BPF_SCHED */ + +static inline int bpf_sched_verify_prog(struct bpf_verifier_log *vlog, + const struct bpf_prog *prog) +{ + return -EOPNOTSUPP; +} + +#endif /* CONFIG_BPF_SCHED */ +#endif /* _LINUX_BPF_SCHED_H */ diff --git a/include/linux/bpf_types.h b/include/linux/bpf_types.h index a8137bb6dd3c..5732b485c539 100644 --- a/include/linux/bpf_types.h +++ b/include/linux/bpf_types.h @@ -77,6 +77,10 @@ BPF_PROG_TYPE(BPF_PROG_TYPE_LSM, lsm, void *, void *) #endif /* CONFIG_BPF_LSM */ #endif +#ifdef CONFIG_BPF_SCHED +BPF_PROG_TYPE(BPF_PROG_TYPE_SCHED, bpf_sched, + void *, void *) +#endif /* CONFIG_BPF_SCHED */ BPF_MAP_TYPE(BPF_MAP_TYPE_ARRAY, array_map_ops) BPF_MAP_TYPE(BPF_MAP_TYPE_PERCPU_ARRAY, percpu_array_map_ops) diff --git a/include/linux/sched_hook_defs.h b/include/linux/sched_hook_defs.h new file mode 100644 index 000000000000..14344004e335 --- /dev/null +++ b/include/linux/sched_hook_defs.h @@ -0,0 +1,2 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +BPF_SCHED_HOOK(int, 0, dummy, void) diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 8fae845d80e2..4f8f3f2113a7 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -199,6 +199,7 @@ enum bpf_prog_type { BPF_PROG_TYPE_EXT, BPF_PROG_TYPE_LSM, BPF_PROG_TYPE_SK_LOOKUP, + BPF_PROG_TYPE_SCHED, }; enum bpf_attach_type { @@ -240,6 +241,7 @@ enum bpf_attach_type { BPF_XDP_CPUMAP, BPF_SK_LOOKUP, BPF_XDP, + BPF_SCHED, __MAX_BPF_ATTACH_TYPE }; diff --git a/init/Kconfig b/init/Kconfig index 27c5ed16fef1..c3fa3240d87d 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -1749,6 +1749,16 @@ config BPF_LSM If you are unsure how to answer this question, answer N. +config BPF_SCHED + bool "SCHED Instrumentation with BPF" + depends on BPF_EVENTS + depends on BPF_SYSCALL + help + Enables instrumentation of the sched hooks with eBPF programs for + implementing dynamic scheduling Policies. + + If you are unsure how to answer this question, answer N. + config BPF_SYSCALL bool "Enable bpf() system call" select BPF diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c index fba28f17e61a..9a0a9895ec62 100644 --- a/kernel/bpf/btf.c +++ b/kernel/bpf/btf.c @@ -4479,6 +4479,7 @@ bool btf_ctx_access(int off, int size, enum bpf_access_type type, return true; t = btf_type_by_id(btf, t->type); break; + case BPF_SCHED: case BPF_MODIFY_RETURN: /* For now the BPF_MODIFY_RETURN can only be attached to * functions that return an int. diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 419dbc3d060e..ff65862ae5ce 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -1997,6 +1997,7 @@ bpf_prog_load_check_attach(enum bpf_prog_type prog_type, case BPF_PROG_TYPE_LSM: case BPF_PROG_TYPE_STRUCT_OPS: case BPF_PROG_TYPE_EXT: + case BPF_PROG_TYPE_SCHED: break; default: return -EINVAL; @@ -2108,6 +2109,7 @@ static bool is_perfmon_prog_type(enum bpf_prog_type prog_type) case BPF_PROG_TYPE_LSM: case BPF_PROG_TYPE_STRUCT_OPS: /* has access to struct sock */ case BPF_PROG_TYPE_EXT: /* extends any prog */ + case BPF_PROG_TYPE_SCHED: return true; default: return false; @@ -2608,6 +2610,12 @@ static int bpf_tracing_prog_attach(struct bpf_prog *prog, goto out_put_prog; } break; + case BPF_PROG_TYPE_SCHED: + if (prog->expected_attach_type != BPF_SCHED) { + err = -EINVAL; + goto out_put_prog; + } + break; default: err = -EINVAL; goto out_put_prog; @@ -2838,6 +2846,7 @@ static int bpf_raw_tracepoint_open(const union bpf_attr *attr) case BPF_PROG_TYPE_TRACING: case BPF_PROG_TYPE_EXT: case BPF_PROG_TYPE_LSM: + case BPF_PROG_TYPE_SCHED: if (attr->raw_tracepoint.name) { /* The attach point for this category of programs * should be specified via btf_id during program load. diff --git a/kernel/bpf/trampoline.c b/kernel/bpf/trampoline.c index 986dabc3d11f..cc6ba35a1d14 100644 --- a/kernel/bpf/trampoline.c +++ b/kernel/bpf/trampoline.c @@ -357,6 +357,7 @@ static enum bpf_tramp_prog_type bpf_attach_type_to_tramp(struct bpf_prog *prog) switch (prog->expected_attach_type) { case BPF_TRACE_FENTRY: return BPF_TRAMP_FENTRY; + case BPF_SCHED: case BPF_MODIFY_RETURN: return BPF_TRAMP_MODIFY_RETURN; case BPF_TRACE_FEXIT: diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 6423f1714a2f..d26104b258ba 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -22,6 +22,7 @@ #include <linux/error-injection.h> #include <linux/bpf_lsm.h> #include <linux/btf_ids.h> +#include <linux/bpf_sched.h> #include "disasm.h" @@ -12178,6 +12179,7 @@ int bpf_check_attach_target(struct bpf_verifier_log *log, case BPF_LSM_MAC: case BPF_TRACE_FENTRY: case BPF_TRACE_FEXIT: + case BPF_SCHED: if (!btf_type_is_func(t)) { bpf_log(log, "attach_btf_id %u is not a function\n", btf_id); @@ -12283,7 +12285,8 @@ static int check_attach_btf_id(struct bpf_verifier_env *env) if (prog->type != BPF_PROG_TYPE_TRACING && prog->type != BPF_PROG_TYPE_LSM && - prog->type != BPF_PROG_TYPE_EXT) + prog->type != BPF_PROG_TYPE_EXT && + prog->type != BPF_PROG_TYPE_SCHED) return 0; ret = bpf_check_attach_target(&env->log, prog, tgt_prog, btf_id, &tgt_info); @@ -12323,6 +12326,12 @@ static int check_attach_btf_id(struct bpf_verifier_env *env) return ret; } + if (prog->type == BPF_PROG_TYPE_SCHED) { + ret = bpf_sched_verify_prog(&env->log, prog); + if (ret < 0) + return ret; + } + key = bpf_trampoline_compute_key(tgt_prog, btf_id); tr = bpf_trampoline_get(key, &tgt_info); if (!tr) diff --git a/kernel/sched/Makefile b/kernel/sched/Makefile index 5fc9c9b70862..8ae9e39eb83a 100644 --- a/kernel/sched/Makefile +++ b/kernel/sched/Makefile @@ -36,3 +36,4 @@ obj-$(CONFIG_CPU_FREQ_GOV_SCHEDUTIL) += cpufreq_schedutil.o obj-$(CONFIG_MEMBARRIER) += membarrier.o obj-$(CONFIG_CPU_ISOLATION) += isolation.o obj-$(CONFIG_PSI) += psi.o +obj-$(CONFIG_BPF_SCHED) += bpf_sched.o \ No newline at end of file diff --git a/kernel/sched/bpf_sched.c b/kernel/sched/bpf_sched.c new file mode 100644 index 000000000000..2f05c186cfd0 --- /dev/null +++ b/kernel/sched/bpf_sched.c @@ -0,0 +1,62 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/bpf.h> +#include <linux/cgroup.h> +#include <linux/bpf_verifier.h> +#include <linux/bpf_sched.h> +#include <linux/btf_ids.h> +#include "sched.h" + +/* + * For every hook declare a nop function where a BPF program can be attached. + */ +#define BPF_SCHED_HOOK(RET, DEFAULT, NAME, ...) \ +noinline RET bpf_sched_##NAME(__VA_ARGS__) \ +{ \ + return DEFAULT; \ +} + +#include <linux/sched_hook_defs.h> +#undef BPF_SCHED_HOOK + +#define BPF_SCHED_HOOK(RET, DEFAULT, NAME, ...) BTF_ID(func, bpf_sched_##NAME) +BTF_SET_START(bpf_sched_hooks) +#include <linux/sched_hook_defs.h> +#undef BPF_SCHED_HOOK +BTF_SET_END(bpf_sched_hooks) + +int bpf_sched_verify_prog(struct bpf_verifier_log *vlog, + const struct bpf_prog *prog) +{ + if (!prog->gpl_compatible) { + bpf_log(vlog, + "sched programs must have a GPL compatible license\n"); + return -EINVAL; + } + + if (!btf_id_set_contains(&bpf_sched_hooks, prog->aux->attach_btf_id)) { + bpf_log(vlog, "attach_btf_id %u points to wrong type name %s\n", + prog->aux->attach_btf_id, prog->aux->attach_func_name); + return -EINVAL; + } + + return 0; +} + +static const struct bpf_func_proto * +bpf_sched_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) +{ + switch (func_id) { + case BPF_FUNC_trace_printk: + return bpf_get_trace_printk_proto(); + default: + return NULL; + } +} + +const struct bpf_prog_ops bpf_sched_prog_ops = { +}; + +const struct bpf_verifier_ops bpf_sched_verifier_ops = { + .get_func_proto = bpf_sched_func_proto, + .is_valid_access = btf_ctx_access, +}; diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h index a44cb5155825..7d7c04f698ec 100644 --- a/tools/include/uapi/linux/bpf.h +++ b/tools/include/uapi/linux/bpf.h @@ -199,6 +199,7 @@ enum bpf_prog_type { BPF_PROG_TYPE_EXT, BPF_PROG_TYPE_LSM, BPF_PROG_TYPE_SK_LOOKUP, + BPF_PROG_TYPE_SCHED, }; enum bpf_attach_type { @@ -240,6 +241,7 @@ enum bpf_attach_type { BPF_XDP_CPUMAP, BPF_SK_LOOKUP, BPF_XDP, + BPF_SCHED, __MAX_BPF_ATTACH_TYPE }; -- 2.20.1

1 6

[PATCH openEuler-1.0-LTS 0/6] sched: Introduce dynamic affinity
by Yongqiang Liu 21 Jul '22

21 Jul '22

*** BLURB HERE *** Hui Tang (6): sched: Introduce dynamic affinity for cfs scheduler cpuset: Introduce new interface for scheduler dynamic affinity sched: Adjust wakeup cpu range according CPU util dynamicly sched: Adjust cpu range in load balance dynamicly sched: Add statistics for scheduler dynamic affinity config: enable CONFIG_QOS_SCHED_DYNAMIC_AFFINITY by default arch/arm64/configs/hulk_defconfig | 1 + arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/hulk_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + include/linux/sched.h | 36 ++++++ include/linux/sched/sysctl.h | 4 + init/Kconfig | 10 ++ init/init_task.c | 3 + kernel/cgroup/cpuset.c | 151 ++++++++++++++++++++++- kernel/fork.c | 13 ++ kernel/sched/core.c | 103 ++++++++++++++++ kernel/sched/debug.c | 7 ++ kernel/sched/fair.c | 159 +++++++++++++++++++++++++ kernel/sysctl.c | 11 ++ 14 files changed, 500 insertions(+), 1 deletion(-) -- 2.25.1

1 6

[PATCH openEuler-1.0-LTS 1/3] video: fbdev: sm712fb: Fix crash in smtcfb_read()
by Yongqiang Liu 21 Jul '22

21 Jul '22

From: Helge Deller <deller(a)gmx.de> stable inclusion from stable-v4.19.238 commit 1caa40af491dcfe17b3ae870a854388d8ea01984 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I5GKWU CVE: CVE-2022-2380 -------------------------------- Zheyu Ma reported this crash in the sm712fb driver when reading three bytes from the framebuffer: BUG: unable to handle page fault for address: ffffc90001ffffff RIP: 0010:smtcfb_read+0x230/0x3e0 Call Trace: vfs_read+0x198/0xa00 ? do_sys_openat2+0x27d/0x350 ? __fget_light+0x54/0x340 ksys_read+0xce/0x190 do_syscall_64+0x43/0x90 Fix it by removing the open-coded endianess fixup-code and by moving the pointer post decrement out the fb_readl() function. Reported-by: Zheyu Ma <zheyuma97(a)gmail.com> Signed-off-by: Helge Deller <deller(a)gmx.de> Tested-by: Zheyu Ma <zheyuma97(a)gmail.com> Cc: stable(a)vger.kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Xia Longlong <xialonglong1(a)huawei.com> Reviewed-by: Xiu Jianfeng <xiujianfeng(a)huawei.com> Reviewed-by: Kefeng Wang <wangkefeng.wang(a)huawei.com> Signed-off-by: Yongqiang Liu <liuyongqiang13(a)huawei.com> --- drivers/video/fbdev/sm712fb.c | 25 +++++++------------------ 1 file changed, 7 insertions(+), 18 deletions(-) diff --git a/drivers/video/fbdev/sm712fb.c b/drivers/video/fbdev/sm712fb.c index f1dcc6766d1e..0713b8444ecc 100644 --- a/drivers/video/fbdev/sm712fb.c +++ b/drivers/video/fbdev/sm712fb.c @@ -1047,7 +1047,7 @@ static ssize_t smtcfb_read(struct fb_info *info, char __user *buf, if (count + p > total_size) count = total_size - p; - buffer = kmalloc((count > PAGE_SIZE) ? PAGE_SIZE : count, GFP_KERNEL); + buffer = kmalloc(PAGE_SIZE, GFP_KERNEL); if (!buffer) return -ENOMEM; @@ -1059,24 +1059,13 @@ static ssize_t smtcfb_read(struct fb_info *info, char __user *buf, while (count) { c = (count > PAGE_SIZE) ? PAGE_SIZE : count; dst = buffer; - for (i = c >> 2; i--;) { - *dst = fb_readl(src++); - *dst = big_swap(*dst); - dst++; - } - if (c & 3) { - u8 *dst8 = (u8 *)dst; - u8 __iomem *src8 = (u8 __iomem *)src; + for (i = (c + 3) >> 2; i--;) { + u32 val; - for (i = c & 3; i--;) { - if (i & 1) { - *dst8++ = fb_readb(++src8); - } else { - *dst8++ = fb_readb(--src8); - src8 += 2; - } - } - src = (u32 __iomem *)src8; + val = fb_readl(src); + *dst = big_swap(val); + src++; + dst++; } if (copy_to_user(buf, buffer, c)) { -- 2.25.1

1 2

openEuler kernel SIG例会
by openEuler conference 21 Jul '22

21 Jul '22

您好！ Kernel SIG 邀请您参加 2022-07-22 14:00 召开的Zoom会议(自动录制) 会议主题：openEuler kernel SIG例会会议内容： 1、introduce scheduler BPF特性回合评审 2、支持在离线混部任务优先级负载均衡特性回合评审 3、kernel仓支持PR回合补丁介绍 4、kernel仓库瘦身讨论会议链接：https://us06web.zoom.us/j/81074273618?pwd=dTV6UmRETzR6Umd3amU0OFRXVlhndz09 会议纪要：https://etherpad.openeuler.org/p/Kernel-meetings 温馨提醒：建议接入会议后修改参会人的姓名，也可以使用您在gitee.com的ID 更多资讯尽在：https://openeuler.org/zh/ Hello! openEuler Kernel SIG invites you to attend the Zoom conference(auto recording) will be held at 2022-07-22 14:00, The subject of the conference is openEuler kernel SIG例会, Summary: 1、introduce scheduler BPF特性回合评审 2、支持在离线混部任务优先级负载均衡特性回合评审 3、kernel仓支持PR回合补丁介绍 4、kernel仓库瘦身讨论 You can join the meeting at https://us06web.zoom.us/j/81074273618?pwd=dTV6UmRETzR6Umd3amU0OFRXVlhndz09. Add topics at https://etherpad.openeuler.org/p/Kernel-meetings. Note: You are advised to change the participant name after joining the conference or use your ID at gitee.com. More information: https://openeuler.org/en/

1 0

[PATCH openEuler-1.0-LTS 1/2] block: don't delete queue kobject before its children
by Yongqiang Liu 20 Jul '22

20 Jul '22

From: Eric Biggers <ebiggers(a)google.com> stable inclusion from linux-4.19.238 commit b2001eb10f59363da930cdd6e086a2861986fa18 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I5EWKO CVE: NA -------------------------------- [ Upstream commit 0f69288253e9fc7c495047720e523b9f1aba5712 ] kobjects aren't supposed to be deleted before their child kobjects are deleted. Apparently this is usually benign; however, a WARN will be triggered if one of the child kobjects has a named attribute group: sysfs group 'modes' not found for kobject 'crypto' WARNING: CPU: 0 PID: 1 at fs/sysfs/group.c:278 sysfs_remove_group+0x72/0x80 ... Call Trace: sysfs_remove_groups+0x29/0x40 fs/sysfs/group.c:312 __kobject_del+0x20/0x80 lib/kobject.c:611 kobject_cleanup+0xa4/0x140 lib/kobject.c:696 kobject_release lib/kobject.c:736 [inline] kref_put include/linux/kref.h:65 [inline] kobject_put+0x53/0x70 lib/kobject.c:753 blk_crypto_sysfs_unregister+0x10/0x20 block/blk-crypto-sysfs.c:159 blk_unregister_queue+0xb0/0x110 block/blk-sysfs.c:962 del_gendisk+0x117/0x250 block/genhd.c:610 Fix this by moving the kobject_del() and the corresponding kobject_uevent() to the correct place. Fixes: 2c2086afc2b8 ("block: Protect less code with sysfs_lock in blk_{un,}register_queue()") Reviewed-by: Hannes Reinecke <hare(a)suse.de> Reviewed-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Reviewed-by: Bart Van Assche <bvanassche(a)acm.org> Signed-off-by: Eric Biggers <ebiggers(a)google.com> Reviewed-by: Christoph Hellwig <hch(a)lst.de> Link: https://lore.kernel.org/r/20220124215938.2769-3-ebiggers@kernel.org Signed-off-by: Jens Axboe <axboe(a)kernel.dk> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Conflict: block/blk-sysfs.c Signed-off-by: Zhang Wensheng <zhangwensheng5(a)huawei.com> Reviewed-by: Jason Yan <yanaijie(a)huawei.com> Signed-off-by: Yongqiang Liu <liuyongqiang13(a)huawei.com> --- block/blk-sysfs.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c index 6beca7743c11..ee963759ae10 100644 --- a/block/blk-sysfs.c +++ b/block/blk-sysfs.c @@ -1037,8 +1037,6 @@ void blk_unregister_queue(struct gendisk *disk) blk_mq_unregister_dev(disk_to_dev(disk), q); mutex_unlock(&q->sysfs_lock); - kobject_uevent(&q->kobj, KOBJ_REMOVE); - kobject_del(&q->kobj); blk_trace_remove_sysfs(disk_to_dev(disk)); mutex_lock(&q->sysfs_lock); @@ -1046,5 +1044,9 @@ void blk_unregister_queue(struct gendisk *disk) elv_unregister_queue(q); mutex_unlock(&q->sysfs_lock); + /* Now that we've deleted all child objects, we can delete the queue. */ + kobject_uevent(&q->kobj, KOBJ_REMOVE); + kobject_del(&q->kobj); + kobject_put(&disk_to_dev(disk)->kobj); } -- 2.25.1

1 1

[PATCH openEuler-1.0-LTS 1/2] etmem:fix kasan slab-out-of-bounds in do_swapcache_reclaim
by Laibin Qiu 20 Jul '22

20 Jul '22

From: liubo <liubo254(a)huawei.com> euleros inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I5GN7K CVE: NA -------------------------------- In the do_swapcache_reclaim interface, there is a slab-out-of-bounds kasan problem; The reason for the problem is that when list_for_each_entry_safe_reverse_from traverses the LRU linked list, it does not consider that next may be equal to the head address, which may lead to the head address being accessed as the page address, causing problems. In response to the above problems, add a judgment about whether pos is head. Signed-off-by: liubo <liubo254(a)huawei.com> Reviewed-by: Miaohe Lin <linmiaohe(a)huawei.com> Reviewed-by: wangkefeng <wangkefeng.wang(a)huawei.com> Signed-off-by: Laibin Qiu <qiulaibin(a)huawei.com> --- mm/vmscan.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/mm/vmscan.c b/mm/vmscan.c index 42f24473756b..011827f666cc 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -4555,7 +4555,6 @@ EXPORT_SYMBOL_GPL(get_page_from_vaddr); static int add_page_for_reclaim_swapcache(struct page *page, struct list_head *pagelist, struct lruvec *lruvec, enum lru_list lru) { - struct list_head *src = &lruvec->lists[lru]; struct page *head; /* If the page is mapped by more than one process, do not swap it */ @@ -4574,7 +4573,6 @@ static int add_page_for_reclaim_swapcache(struct page *page, reliable_lru_add(lru, head, -hpage_nr_pages(head)); break; case -EBUSY: - list_move(&head->lru, src); return -1; default: break; @@ -4744,7 +4742,7 @@ int do_swapcache_reclaim(unsigned long *swapcache_watermark, * check if pos page is been released or not in LRU list, if true, * cancel the subsequent page scanning of the current node. */ - if (!pos) { + if (!pos || &pos->lru == src) { spin_unlock_irq(&pgdat->lru_lock); continue; } -- 2.25.1

1 1

[PATCH openEuler-1.0-LTS 1/8] ext4: fix race condition between ext4_ioctl_setflags and ext4_fiemap
by Yongqiang Liu 20 Jul '22

20 Jul '22

From: Baokun Li <libaokun1(a)huawei.com> hulk inclusion category: bugfix bugzilla: 187222, https://gitee.com/openeuler/kernel/issues/I5H3KE CVE: NA -------------------------------- Hulk Robot reported a BUG: ================================================================== kernel BUG at fs/ext4/extents_status.c:762! invalid opcode: 0000 [#1] SMP KASAN PTI [...] Call Trace: ext4_cache_extents+0x238/0x2f0 ext4_find_extent+0x785/0xa40 ext4_fiemap+0x36d/0xe90 do_vfs_ioctl+0x6af/0x1200 [...] ================================================================== Above issue may happen as follows: ------------------------------------- cpu1 cpu2 _____________________|_____________________ do_vfs_ioctl ext4_ioctl ext4_ioctl_setflags ext4_ind_migrate do_vfs_ioctl ioctl_fiemap ext4_fiemap ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS) ext4_fill_fiemap_extents down_write(&EXT4_I(inode)->i_data_sem); ext4_ext_check_inode ext4_clear_inode_flag(inode, EXT4_INODE_EXTENTS) memset(ei->i_data, 0, sizeof(ei->i_data)) up_write(&EXT4_I(inode)->i_data_sem); down_read(&EXT4_I(inode)->i_data_sem); ext4_find_extent ext4_cache_extents ext4_es_cache_extent BUG_ON(end < lblk) We can easily reproduce this problem with the syzkaller testcase: ``` 02:37:07 executing program 3: r0 = openat(0xffffffffffffff9c, &(0x7f0000000040)='./file0\x00', 0x26e1, 0x0) ioctl$FS_IOC_FSSETXATTR(r0, 0x40086602, &(0x7f0000000080)={0x17e}) mkdirat(0xffffffffffffff9c, &(0x7f00000000c0)='./file1\x00', 0x1ff) r1 = openat(0xffffffffffffff9c, &(0x7f0000000100)='./file1\x00', 0x0, 0x0) ioctl$FS_IOC_FIEMAP(r1, 0xc020660b, &(0x7f0000000180)={0x0, 0x1, 0x0, 0xef3, 0x6, []}) (async, rerun: 32) ioctl$FS_IOC_FSSETXATTR(r1, 0x40086602, &(0x7f0000000140)={0x17e}) (rerun: 32) ``` To solve this issue, we use __generic_block_fiemap() instead of generic_block_fiemap() and add inode_lock_shared to avoid race condition. Reported-by: Hulk Robot <hulkci(a)huawei.com> Signed-off-by: Baokun Li <libaokun1(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Yongqiang Liu <liuyongqiang13(a)huawei.com> --- fs/ext4/extents.c | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-) diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index d08e6d4afba6..8e5ed3e315cd 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -5267,13 +5267,18 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, return error; } + inode_lock_shared(inode); /* fallback to generic here if not in extents fmt */ - if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) - return generic_block_fiemap(inode, fieinfo, start, len, + if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) { + error = __generic_block_fiemap(inode, fieinfo, start, len, ext4_get_block); + goto out_unlock; + } - if (fiemap_check_flags(fieinfo, EXT4_FIEMAP_FLAGS)) - return -EBADR; + if (fiemap_check_flags(fieinfo, EXT4_FIEMAP_FLAGS)) { + error = -EBADR; + goto out_unlock; + } if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR) { error = ext4_xattr_fiemap(inode, fieinfo); @@ -5294,6 +5299,8 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, error = ext4_fill_fiemap_extents(inode, start_blk, len_blks, fieinfo); } +out_unlock: + inode_unlock_shared(inode); return error; } -- 2.25.1

1 7

[PATCH openEuler-5.10 01/29] crypto: hisilicon/sec - don't sleep when in softirq
by Zheng Zengkai 19 Jul '22

19 Jul '22

From: Zhengchao Shao <shaozhengchao(a)huawei.com> mainline inclusion from mainline-v5.19-rc6 commit 02884a4f12de11f54d4ca67a07dd1f111d96fdbd category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I5GEVR CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/herbert/cryptodev-2.6.git/c… -------------------------------- When kunpeng920 encryption driver is used to deencrypt and decrypt packets during the softirq, it is not allowed to use mutex lock. The kernel will report the following error: BUG: scheduling while atomic: swapper/57/0/0x00000300 Call trace: dump_backtrace+0x0/0x1e4 show_stack+0x20/0x2c dump_stack+0xd8/0x140 __schedule_bug+0x68/0x80 __schedule+0x728/0x840 schedule+0x50/0xe0 schedule_preempt_disabled+0x18/0x24 __mutex_lock.constprop.0+0x594/0x5dc __mutex_lock_slowpath+0x1c/0x30 mutex_lock+0x50/0x60 sec_request_init+0x8c/0x1a0 [hisi_sec2] sec_process+0x28/0x1ac [hisi_sec2] sec_skcipher_crypto+0xf4/0x1d4 [hisi_sec2] sec_skcipher_encrypt+0x1c/0x30 [hisi_sec2] crypto_skcipher_encrypt+0x2c/0x40 crypto_authenc_encrypt+0xc8/0xfc [authenc] crypto_aead_encrypt+0x2c/0x40 echainiv_encrypt+0x144/0x1a0 [echainiv] crypto_aead_encrypt+0x2c/0x40 esp_output_tail+0x348/0x5c0 [esp4] esp_output+0x120/0x19c [esp4] xfrm_output_one+0x25c/0x4d4 xfrm_output_resume+0x6c/0x1fc xfrm_output+0xac/0x3c0 xfrm4_output+0x64/0x130 ip_build_and_send_pkt+0x158/0x20c tcp_v4_send_synack+0xdc/0x1f0 tcp_conn_request+0x7d0/0x994 tcp_v4_conn_request+0x58/0x6c tcp_v6_conn_request+0xf0/0x100 tcp_rcv_state_process+0x1cc/0xd60 tcp_v4_do_rcv+0x10c/0x250 tcp_v4_rcv+0xfc4/0x10a4 ip_protocol_deliver_rcu+0xf4/0x200 ip_local_deliver_finish+0x58/0x70 ip_local_deliver+0x68/0x120 ip_sublist_rcv_finish+0x70/0x94 ip_list_rcv_finish.constprop.0+0x17c/0x1d0 ip_sublist_rcv+0x40/0xb0 ip_list_rcv+0x140/0x1dc __netif_receive_skb_list_core+0x154/0x28c __netif_receive_skb_list+0x120/0x1a0 netif_receive_skb_list_internal+0xe4/0x1f0 napi_complete_done+0x70/0x1f0 gro_cell_poll+0x9c/0xb0 napi_poll+0xcc/0x264 net_rx_action+0xd4/0x21c __do_softirq+0x130/0x358 irq_exit+0x11c/0x13c __handle_domain_irq+0x88/0xf0 gic_handle_irq+0x78/0x2c0 el1_irq+0xb8/0x140 arch_cpu_idle+0x18/0x40 default_idle_call+0x5c/0x1c0 cpuidle_idle_call+0x174/0x1b0 do_idle+0xc8/0x160 cpu_startup_entry+0x30/0x11c secondary_start_kernel+0x158/0x1e4 softirq: huh, entered softirq 3 NET_RX 0000000093774ee4 with preempt_count 00000100, exited with fffffe00? Fixes: 416d82204df4 ("crypto: hisilicon - add HiSilicon SEC V2 driver") Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> Signed-off-by: Herbert Xu <herbert(a)gondor.apana.org.au> Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> Reviewed-by: Yue Haibing <yuehaibing(a)huawei.com> Reviewed-by: Wei Yongjun <weiyongjun1(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- drivers/crypto/hisilicon/sec2/sec.h | 2 +- drivers/crypto/hisilicon/sec2/sec_crypto.c | 20 ++++++++++---------- 2 files changed, 11 insertions(+), 11 deletions(-) diff --git a/drivers/crypto/hisilicon/sec2/sec.h b/drivers/crypto/hisilicon/sec2/sec.h index c2e9b01187a7..a44c8dba3cda 100644 --- a/drivers/crypto/hisilicon/sec2/sec.h +++ b/drivers/crypto/hisilicon/sec2/sec.h @@ -119,7 +119,7 @@ struct sec_qp_ctx { struct idr req_idr; struct sec_alg_res res[QM_Q_DEPTH]; struct sec_ctx *ctx; - struct mutex req_lock; + spinlock_t req_lock; struct list_head backlog; struct hisi_acc_sgl_pool *c_in_pool; struct hisi_acc_sgl_pool *c_out_pool; diff --git a/drivers/crypto/hisilicon/sec2/sec_crypto.c b/drivers/crypto/hisilicon/sec2/sec_crypto.c index 6eebe739893c..71dfa7db6394 100644 --- a/drivers/crypto/hisilicon/sec2/sec_crypto.c +++ b/drivers/crypto/hisilicon/sec2/sec_crypto.c @@ -127,11 +127,11 @@ static int sec_alloc_req_id(struct sec_req *req, struct sec_qp_ctx *qp_ctx) { int req_id; - mutex_lock(&qp_ctx->req_lock); + spin_lock_bh(&qp_ctx->req_lock); req_id = idr_alloc_cyclic(&qp_ctx->req_idr, NULL, 0, QM_Q_DEPTH, GFP_ATOMIC); - mutex_unlock(&qp_ctx->req_lock); + spin_unlock_bh(&qp_ctx->req_lock); if (unlikely(req_id < 0)) { dev_err(req->ctx->dev, "alloc req id fail!\n"); return req_id; @@ -156,9 +156,9 @@ static void sec_free_req_id(struct sec_req *req) qp_ctx->req_list[req_id] = NULL; req->qp_ctx = NULL; - mutex_lock(&qp_ctx->req_lock); + spin_lock_bh(&qp_ctx->req_lock); idr_remove(&qp_ctx->req_idr, req_id); - mutex_unlock(&qp_ctx->req_lock); + spin_unlock_bh(&qp_ctx->req_lock); } static u8 pre_parse_finished_bd(struct bd_status *status, void *resp) @@ -273,7 +273,7 @@ static int sec_bd_send(struct sec_ctx *ctx, struct sec_req *req) !(req->flag & CRYPTO_TFM_REQ_MAY_BACKLOG)) return -EBUSY; - mutex_lock(&qp_ctx->req_lock); + spin_lock_bh(&qp_ctx->req_lock); ret = hisi_qp_send(qp_ctx->qp, &req->sec_sqe); if (ctx->fake_req_limit <= @@ -281,10 +281,10 @@ static int sec_bd_send(struct sec_ctx *ctx, struct sec_req *req) list_add_tail(&req->backlog_head, &qp_ctx->backlog); atomic64_inc(&ctx->sec->debug.dfx.send_cnt); atomic64_inc(&ctx->sec->debug.dfx.send_busy_cnt); - mutex_unlock(&qp_ctx->req_lock); + spin_unlock_bh(&qp_ctx->req_lock); return -EBUSY; } - mutex_unlock(&qp_ctx->req_lock); + spin_unlock_bh(&qp_ctx->req_lock); if (unlikely(ret == -EBUSY)) return -ENOBUFS; @@ -487,7 +487,7 @@ static int sec_create_qp_ctx(struct hisi_qm *qm, struct sec_ctx *ctx, qp->req_cb = sec_req_cb; - mutex_init(&qp_ctx->req_lock); + spin_lock_init(&qp_ctx->req_lock); idr_init(&qp_ctx->req_idr); INIT_LIST_HEAD(&qp_ctx->backlog); @@ -1382,7 +1382,7 @@ static struct sec_req *sec_back_req_clear(struct sec_ctx *ctx, { struct sec_req *backlog_req = NULL; - mutex_lock(&qp_ctx->req_lock); + spin_lock_bh(&qp_ctx->req_lock); if (ctx->fake_req_limit >= atomic_read(&qp_ctx->qp->qp_status.used) && !list_empty(&qp_ctx->backlog)) { @@ -1390,7 +1390,7 @@ static struct sec_req *sec_back_req_clear(struct sec_ctx *ctx, typeof(*backlog_req), backlog_head); list_del(&backlog_req->backlog_head); } - mutex_unlock(&qp_ctx->req_lock); + spin_unlock_bh(&qp_ctx->req_lock); return backlog_req; } -- 2.20.1

1 28