Message ID | 20220908114659.102775-1-jolsa@kernel.org (mailing list archive) |
---|---|
State | Superseded |
Delegated to: | BPF |
Headers | show |
Series | [bpf-next] bpf: Prevent bpf program recursion for raw tracepoint probes | expand |
On 09/08, Jiri Olsa wrote: > We got report from sysbot [1] about warnings that were caused by > bpf program attached to contention_begin raw tracepoint triggering > the same tracepoint by using bpf_trace_printk helper that takes > trace_printk_lock lock. > Call Trace: > <TASK> > ? trace_event_raw_event_bpf_trace_printk+0x5f/0x90 > bpf_trace_printk+0x2b/0xe0 > bpf_prog_a9aec6167c091eef_prog+0x1f/0x24 > bpf_trace_run2+0x26/0x90 > native_queued_spin_lock_slowpath+0x1c6/0x2b0 > _raw_spin_lock_irqsave+0x44/0x50 > bpf_trace_printk+0x3f/0xe0 > bpf_prog_a9aec6167c091eef_prog+0x1f/0x24 > bpf_trace_run2+0x26/0x90 > native_queued_spin_lock_slowpath+0x1c6/0x2b0 > _raw_spin_lock_irqsave+0x44/0x50 > bpf_trace_printk+0x3f/0xe0 > bpf_prog_a9aec6167c091eef_prog+0x1f/0x24 > bpf_trace_run2+0x26/0x90 > native_queued_spin_lock_slowpath+0x1c6/0x2b0 > _raw_spin_lock_irqsave+0x44/0x50 > bpf_trace_printk+0x3f/0xe0 > bpf_prog_a9aec6167c091eef_prog+0x1f/0x24 > bpf_trace_run2+0x26/0x90 > native_queued_spin_lock_slowpath+0x1c6/0x2b0 > _raw_spin_lock_irqsave+0x44/0x50 > __unfreeze_partials+0x5b/0x160 > ... > The can be reproduced by attaching bpf program as raw tracepoint on > contention_begin tracepoint. The bpf prog calls bpf_trace_printk > helper. Then by running perf bench the spin lock code is forced to > take slowpath and call contention_begin tracepoint. > Fixing this by skipping execution of the bpf program if it's > already running, Using bpf prog 'active' field, which is being > currently used by trampoline programs for the same reason. Makes sense to me and seems to address Alexei's earlier point about bpf_prog_active. Reviewed-by: Stanislav Fomichev <sdf@google.com> > Reported-by: syzbot+2251879aa068ad9c960d@syzkaller.appspotmail.com > [1] https://lore.kernel.org/bpf/YxhFe3EwqchC%2FfYf@krava/T/#t > Signed-off-by: Jiri Olsa <jolsa@kernel.org> > --- > include/linux/bpf.h | 1 + > kernel/bpf/trampoline.c | 6 +++--- > kernel/trace/bpf_trace.c | 6 ++++++ > 3 files changed, 10 insertions(+), 3 deletions(-) > diff --git a/include/linux/bpf.h b/include/linux/bpf.h > index 48ae05099f36..4737bd0fcbb8 100644 > --- a/include/linux/bpf.h > +++ b/include/linux/bpf.h > @@ -2640,4 +2640,5 @@ static inline void bpf_cgroup_atype_get(u32 > attach_btf_id, int cgroup_atype) {} > static inline void bpf_cgroup_atype_put(int cgroup_atype) {} > #endif /* CONFIG_BPF_LSM */ > +void notrace bpf_prog_inc_misses_counter(struct bpf_prog *prog); > #endif /* _LINUX_BPF_H */ > diff --git a/kernel/bpf/trampoline.c b/kernel/bpf/trampoline.c > index ad76940b02cc..a098bdc33209 100644 > --- a/kernel/bpf/trampoline.c > +++ b/kernel/bpf/trampoline.c > @@ -863,7 +863,7 @@ static __always_inline u64 notrace > bpf_prog_start_time(void) > return start; > } > -static void notrace inc_misses_counter(struct bpf_prog *prog) > +void notrace bpf_prog_inc_misses_counter(struct bpf_prog *prog) > { > struct bpf_prog_stats *stats; > unsigned int flags; > @@ -896,7 +896,7 @@ u64 notrace __bpf_prog_enter(struct bpf_prog *prog, > struct bpf_tramp_run_ctx *ru > run_ctx->saved_run_ctx = bpf_set_run_ctx(&run_ctx->run_ctx); > if (unlikely(this_cpu_inc_return(*(prog->active)) != 1)) { > - inc_misses_counter(prog); > + bpf_prog_inc_misses_counter(prog); > return 0; > } > return bpf_prog_start_time(); > @@ -967,7 +967,7 @@ u64 notrace __bpf_prog_enter_sleepable(struct > bpf_prog *prog, struct bpf_tramp_r > might_fault(); > if (unlikely(this_cpu_inc_return(*(prog->active)) != 1)) { > - inc_misses_counter(prog); > + bpf_prog_inc_misses_counter(prog); > return 0; > } > diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c > index 68e5cdd24cef..c8cd1aa7b112 100644 > --- a/kernel/trace/bpf_trace.c > +++ b/kernel/trace/bpf_trace.c > @@ -2042,9 +2042,15 @@ static __always_inline > void __bpf_trace_run(struct bpf_prog *prog, u64 *args) > { > cant_sleep(); > + if (unlikely(this_cpu_inc_return(*(prog->active)) != 1)) { > + bpf_prog_inc_misses_counter(prog); > + goto out; > + } > rcu_read_lock(); > (void) bpf_prog_run(prog, args); > rcu_read_unlock(); > +out: > + this_cpu_dec(*(prog->active)); > } > #define UNPACK(...) __VA_ARGS__ > -- > 2.37.3
Hi Jiri, I love your patch! Yet something to improve: [auto build test ERROR on bpf-next/master] url: https://github.com/intel-lab-lkp/linux/commits/Jiri-Olsa/bpf-Prevent-bpf-program-recursion-for-raw-tracepoint-probes/20220908-194832 base: https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next.git master config: arm64-buildonly-randconfig-r002-20220907 (https://download.01.org/0day-ci/archive/20220909/202209091236.avgRKOSj-lkp@intel.com/config) compiler: aarch64-linux-gcc (GCC) 12.1.0 reproduce (this is a W=1 build): wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross # https://github.com/intel-lab-lkp/linux/commit/f68b567cfb6572c20e431242a440cc5f01452485 git remote add linux-review https://github.com/intel-lab-lkp/linux git fetch --no-tags linux-review Jiri-Olsa/bpf-Prevent-bpf-program-recursion-for-raw-tracepoint-probes/20220908-194832 git checkout f68b567cfb6572c20e431242a440cc5f01452485 # save the config file mkdir build_dir && cp config build_dir/.config COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-12.1.0 make.cross W=1 O=build_dir ARCH=arm64 SHELL=/bin/bash If you fix the issue, kindly add following tag where applicable Reported-by: kernel test robot <lkp@intel.com> All errors (new ones prefixed by >>): aarch64-linux-ld: Unexpected GOT/PLT entries detected! aarch64-linux-ld: Unexpected run-time procedure linkages detected! aarch64-linux-ld: kernel/trace/bpf_trace.o: in function `__bpf_trace_run': >> kernel/trace/bpf_trace.c:2046: undefined reference to `bpf_prog_inc_misses_counter' >> aarch64-linux-ld: kernel/trace/bpf_trace.c:2046: undefined reference to `bpf_prog_inc_misses_counter' >> aarch64-linux-ld: kernel/trace/bpf_trace.c:2046: undefined reference to `bpf_prog_inc_misses_counter' >> aarch64-linux-ld: kernel/trace/bpf_trace.c:2046: undefined reference to `bpf_prog_inc_misses_counter' >> aarch64-linux-ld: kernel/trace/bpf_trace.c:2046: undefined reference to `bpf_prog_inc_misses_counter' aarch64-linux-ld: kernel/trace/bpf_trace.o:kernel/trace/bpf_trace.c:2046: more undefined references to `bpf_prog_inc_misses_counter' follow vim +2046 kernel/trace/bpf_trace.c 2040 2041 static __always_inline 2042 void __bpf_trace_run(struct bpf_prog *prog, u64 *args) 2043 { 2044 cant_sleep(); 2045 if (unlikely(this_cpu_inc_return(*(prog->active)) != 1)) { > 2046 bpf_prog_inc_misses_counter(prog); 2047 goto out; 2048 } 2049 rcu_read_lock(); 2050 (void) bpf_prog_run(prog, args); 2051 rcu_read_unlock(); 2052 out: 2053 this_cpu_dec(*(prog->active)); 2054 } 2055
Hi Jiri, I love your patch! Yet something to improve: [auto build test ERROR on bpf-next/master] url: https://github.com/intel-lab-lkp/linux/commits/Jiri-Olsa/bpf-Prevent-bpf-program-recursion-for-raw-tracepoint-probes/20220908-194832 base: https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next.git master config: x86_64-randconfig-c022 (https://download.01.org/0day-ci/archive/20220909/202209091544.TU8KWEUM-lkp@intel.com/config) compiler: gcc-11 (Debian 11.3.0-5) 11.3.0 reproduce (this is a W=1 build): # https://github.com/intel-lab-lkp/linux/commit/f68b567cfb6572c20e431242a440cc5f01452485 git remote add linux-review https://github.com/intel-lab-lkp/linux git fetch --no-tags linux-review Jiri-Olsa/bpf-Prevent-bpf-program-recursion-for-raw-tracepoint-probes/20220908-194832 git checkout f68b567cfb6572c20e431242a440cc5f01452485 # save the config file mkdir build_dir && cp config build_dir/.config make W=1 O=build_dir ARCH=x86_64 SHELL=/bin/bash If you fix the issue, kindly add following tag where applicable Reported-by: kernel test robot <lkp@intel.com> All errors (new ones prefixed by >>): ld: kernel/trace/bpf_trace.o: in function `__bpf_trace_run': kernel/trace/bpf_trace.c:2046: undefined reference to `bpf_prog_inc_misses_counter' >> ld: kernel/trace/bpf_trace.c:2046: undefined reference to `bpf_prog_inc_misses_counter' >> ld: kernel/trace/bpf_trace.c:2046: undefined reference to `bpf_prog_inc_misses_counter' >> ld: kernel/trace/bpf_trace.c:2046: undefined reference to `bpf_prog_inc_misses_counter' >> ld: kernel/trace/bpf_trace.c:2046: undefined reference to `bpf_prog_inc_misses_counter' ld: kernel/trace/bpf_trace.o:kernel/trace/bpf_trace.c:2046: more undefined references to `bpf_prog_inc_misses_counter' follow
On Fri, Sep 09, 2022 at 03:27:57PM +0800, kernel test robot wrote: > Hi Jiri, > > I love your patch! Yet something to improve: > > [auto build test ERROR on bpf-next/master] > > url: https://github.com/intel-lab-lkp/linux/commits/Jiri-Olsa/bpf-Prevent-bpf-program-recursion-for-raw-tracepoint-probes/20220908-194832 > base: https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next.git master > config: x86_64-randconfig-c022 (https://download.01.org/0day-ci/archive/20220909/202209091544.TU8KWEUM-lkp@intel.com/config) > compiler: gcc-11 (Debian 11.3.0-5) 11.3.0 > reproduce (this is a W=1 build): > # https://github.com/intel-lab-lkp/linux/commit/f68b567cfb6572c20e431242a440cc5f01452485 > git remote add linux-review https://github.com/intel-lab-lkp/linux > git fetch --no-tags linux-review Jiri-Olsa/bpf-Prevent-bpf-program-recursion-for-raw-tracepoint-probes/20220908-194832 > git checkout f68b567cfb6572c20e431242a440cc5f01452485 > # save the config file > mkdir build_dir && cp config build_dir/.config > make W=1 O=build_dir ARCH=x86_64 SHELL=/bin/bash > > If you fix the issue, kindly add following tag where applicable > Reported-by: kernel test robot <lkp@intel.com> > > All errors (new ones prefixed by >>): > > ld: kernel/trace/bpf_trace.o: in function `__bpf_trace_run': > kernel/trace/bpf_trace.c:2046: undefined reference to `bpf_prog_inc_misses_counter' > >> ld: kernel/trace/bpf_trace.c:2046: undefined reference to `bpf_prog_inc_misses_counter' > >> ld: kernel/trace/bpf_trace.c:2046: undefined reference to `bpf_prog_inc_misses_counter' > >> ld: kernel/trace/bpf_trace.c:2046: undefined reference to `bpf_prog_inc_misses_counter' > >> ld: kernel/trace/bpf_trace.c:2046: undefined reference to `bpf_prog_inc_misses_counter' > ld: kernel/trace/bpf_trace.o:kernel/trace/bpf_trace.c:2046: more undefined references to `bpf_prog_inc_misses_counter' follow ah right, trampoline.o is for JIT config only, will move bpf_prog_inc_misses_counter to some common place jirka > > -- > 0-DAY CI Kernel Test Service > https://01.org/lkp
diff --git a/include/linux/bpf.h b/include/linux/bpf.h index 48ae05099f36..4737bd0fcbb8 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -2640,4 +2640,5 @@ static inline void bpf_cgroup_atype_get(u32 attach_btf_id, int cgroup_atype) {} static inline void bpf_cgroup_atype_put(int cgroup_atype) {} #endif /* CONFIG_BPF_LSM */ +void notrace bpf_prog_inc_misses_counter(struct bpf_prog *prog); #endif /* _LINUX_BPF_H */ diff --git a/kernel/bpf/trampoline.c b/kernel/bpf/trampoline.c index ad76940b02cc..a098bdc33209 100644 --- a/kernel/bpf/trampoline.c +++ b/kernel/bpf/trampoline.c @@ -863,7 +863,7 @@ static __always_inline u64 notrace bpf_prog_start_time(void) return start; } -static void notrace inc_misses_counter(struct bpf_prog *prog) +void notrace bpf_prog_inc_misses_counter(struct bpf_prog *prog) { struct bpf_prog_stats *stats; unsigned int flags; @@ -896,7 +896,7 @@ u64 notrace __bpf_prog_enter(struct bpf_prog *prog, struct bpf_tramp_run_ctx *ru run_ctx->saved_run_ctx = bpf_set_run_ctx(&run_ctx->run_ctx); if (unlikely(this_cpu_inc_return(*(prog->active)) != 1)) { - inc_misses_counter(prog); + bpf_prog_inc_misses_counter(prog); return 0; } return bpf_prog_start_time(); @@ -967,7 +967,7 @@ u64 notrace __bpf_prog_enter_sleepable(struct bpf_prog *prog, struct bpf_tramp_r might_fault(); if (unlikely(this_cpu_inc_return(*(prog->active)) != 1)) { - inc_misses_counter(prog); + bpf_prog_inc_misses_counter(prog); return 0; } diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index 68e5cdd24cef..c8cd1aa7b112 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -2042,9 +2042,15 @@ static __always_inline void __bpf_trace_run(struct bpf_prog *prog, u64 *args) { cant_sleep(); + if (unlikely(this_cpu_inc_return(*(prog->active)) != 1)) { + bpf_prog_inc_misses_counter(prog); + goto out; + } rcu_read_lock(); (void) bpf_prog_run(prog, args); rcu_read_unlock(); +out: + this_cpu_dec(*(prog->active)); } #define UNPACK(...) __VA_ARGS__
We got report from sysbot [1] about warnings that were caused by bpf program attached to contention_begin raw tracepoint triggering the same tracepoint by using bpf_trace_printk helper that takes trace_printk_lock lock. Call Trace: <TASK> ? trace_event_raw_event_bpf_trace_printk+0x5f/0x90 bpf_trace_printk+0x2b/0xe0 bpf_prog_a9aec6167c091eef_prog+0x1f/0x24 bpf_trace_run2+0x26/0x90 native_queued_spin_lock_slowpath+0x1c6/0x2b0 _raw_spin_lock_irqsave+0x44/0x50 bpf_trace_printk+0x3f/0xe0 bpf_prog_a9aec6167c091eef_prog+0x1f/0x24 bpf_trace_run2+0x26/0x90 native_queued_spin_lock_slowpath+0x1c6/0x2b0 _raw_spin_lock_irqsave+0x44/0x50 bpf_trace_printk+0x3f/0xe0 bpf_prog_a9aec6167c091eef_prog+0x1f/0x24 bpf_trace_run2+0x26/0x90 native_queued_spin_lock_slowpath+0x1c6/0x2b0 _raw_spin_lock_irqsave+0x44/0x50 bpf_trace_printk+0x3f/0xe0 bpf_prog_a9aec6167c091eef_prog+0x1f/0x24 bpf_trace_run2+0x26/0x90 native_queued_spin_lock_slowpath+0x1c6/0x2b0 _raw_spin_lock_irqsave+0x44/0x50 __unfreeze_partials+0x5b/0x160 ... The can be reproduced by attaching bpf program as raw tracepoint on contention_begin tracepoint. The bpf prog calls bpf_trace_printk helper. Then by running perf bench the spin lock code is forced to take slowpath and call contention_begin tracepoint. Fixing this by skipping execution of the bpf program if it's already running, Using bpf prog 'active' field, which is being currently used by trampoline programs for the same reason. Reported-by: syzbot+2251879aa068ad9c960d@syzkaller.appspotmail.com [1] https://lore.kernel.org/bpf/YxhFe3EwqchC%2FfYf@krava/T/#t Signed-off-by: Jiri Olsa <jolsa@kernel.org> --- include/linux/bpf.h | 1 + kernel/bpf/trampoline.c | 6 +++--- kernel/trace/bpf_trace.c | 6 ++++++ 3 files changed, 10 insertions(+), 3 deletions(-)