lkml.org 
[lkml]   [2019]   [Apr]   [5]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [RFC PATCH] perf/x86: make perf callchain work without CONFIG_FRAME_POINTER

And you forgot to Cc Josh..

On Fri, Apr 05, 2019 at 01:25:45AM +0800, Kairui Song wrote:
> Currently perf callchain is not working properly with ORC unwinder,
> we'll get useless in kernel callchain like this:
>
> perf 6429 [000] 22.498450: kmem:mm_page_alloc: page=0x176a17 pfn=1534487 order=0 migratetype=0 gfp_flags=GFP_KERNEL
> ffffffffbe23e32e __alloc_pages_nodemask+0x22e (/lib/modules/5.1.0-rc3+/build/vmlinux)
> 7efdf7f7d3e8 __poll+0x18 (/usr/lib64/libc-2.28.so)
> 5651468729c1 [unknown] (/usr/bin/perf)
> 5651467ee82a main+0x69a (/usr/bin/perf)
> 7efdf7eaf413 __libc_start_main+0xf3 (/usr/lib64/libc-2.28.so)
> 5541f689495641d7 [unknown] ([unknown])
>
> Without CONFIG_FRAME_POINTER, bp is not reserved as frame pointer so
> can't get callers frame pointer, instead current frame pointer is
> returned when trying to fetch caller registers. The unwinder will error
> out early, and end the stacktrace early.
>
> So instead of let the unwinder start with the dumped register, we start
> it right where the unwinding started when the stacktrace is triggered by
> trace event directly. And skip until the frame pointer is reached.
>
> This makes the callchain get the full in kernel stacktrace again:
>
> perf 6503 [000] 1567.570191: kmem:mm_page_alloc: page=0x16c904 pfn=1493252 order=0 migratetype=0 gfp_flags=GFP_KERNEL
> ffffffffb523e2ae __alloc_pages_nodemask+0x22e (/lib/modules/5.1.0-rc3+/build/vmlinux)
> ffffffffb52383bd __get_free_pages+0xd (/lib/modules/5.1.0-rc3+/build/vmlinux)
> ffffffffb52fd28a __pollwait+0x8a (/lib/modules/5.1.0-rc3+/build/vmlinux)
> ffffffffb521426f perf_poll+0x2f (/lib/modules/5.1.0-rc3+/build/vmlinux)
> ffffffffb52fe3e2 do_sys_poll+0x252 (/lib/modules/5.1.0-rc3+/build/vmlinux)
> ffffffffb52ff027 __x64_sys_poll+0x37 (/lib/modules/5.1.0-rc3+/build/vmlinux)
> ffffffffb500418b do_syscall_64+0x5b (/lib/modules/5.1.0-rc3+/build/vmlinux)
> ffffffffb5a0008c entry_SYSCALL_64_after_hwframe+0x44 (/lib/modules/5.1.0-rc3+/build/vmlinux)
> 7f71e92d03e8 __poll+0x18 (/usr/lib64/libc-2.28.so)
> 55a22960d9c1 [unknown] (/usr/bin/perf)
> 55a22958982a main+0x69a (/usr/bin/perf)
> 7f71e9202413 __libc_start_main+0xf3 (/usr/lib64/libc-2.28.so)
> 5541f689495641d7 [unknown] ([unknown])
>
> ----
>
> Just found with ORC unwinder the perf callchain is unusable, and this
> seems fixes it well, any suggestion is welcome, thanks!

That whole .direct stuff is horrible crap.

> ---
> arch/x86/events/core.c | 34 ++++++++++++++++++++++++++++------
> include/linux/perf_event.h | 3 ++-
> kernel/bpf/stackmap.c | 4 ++--
> kernel/events/callchain.c | 13 +++++++++++--
> kernel/events/core.c | 2 +-
> 5 files changed, 44 insertions(+), 12 deletions(-)
>
> diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c
> index e2b1447192a8..3f3e110794ac 100644
> --- a/arch/x86/events/core.c
> +++ b/arch/x86/events/core.c
> @@ -2355,8 +2355,9 @@ void arch_perf_update_userpage(struct perf_event *event,
> cyc2ns_read_end();
> }
>
> -void
> -perf_callchain_kernel(struct perf_callchain_entry_ctx *entry, struct pt_regs *regs)
> +static void
> +__perf_callchain_kernel(struct perf_callchain_entry_ctx *entry, struct pt_regs *regs,
> + bool direct_call)
> {
> struct unwind_state state;
> unsigned long addr;
> @@ -2366,17 +2367,38 @@ perf_callchain_kernel(struct perf_callchain_entry_ctx *entry, struct pt_regs *re
> return;
> }
>
> - if (perf_callchain_store(entry, regs->ip))
> - return;
> + /*
> + * Without frame pointer, we can't get a reliable caller bp value.
> + * If this is called directly from a trace point, just start the
> + * unwind from here and skip until the frame is reached.
> + */
> + if (IS_ENABLED(CONFIG_FRAME_POINTER) || !direct_call) {
> + if (perf_callchain_store(entry, regs->ip))
> + return;
> + unwind_start(&state, current, regs, NULL);
> + } else {
> + unwind_start(&state, current, NULL, (unsigned long*)regs->bp);
> + }
>
> - for (unwind_start(&state, current, regs, NULL); !unwind_done(&state);
> - unwind_next_frame(&state)) {
> + for (; !unwind_done(&state); unwind_next_frame(&state)) {
> addr = unwind_get_return_address(&state);
> if (!addr || perf_callchain_store(entry, addr))
> return;
> }
> }
>
> +void
> +perf_callchain_kernel(struct perf_callchain_entry_ctx *entry, struct pt_regs *regs)
> +{
> + __perf_callchain_kernel(entry, regs, false);
> +}
> +
> +void
> +perf_callchain_kernel_direct(struct perf_callchain_entry_ctx *entry, struct pt_regs *regs)
> +{
> + __perf_callchain_kernel(entry, regs, true);
> +}
> +
> static inline int
> valid_user_frame(const void __user *fp, unsigned long size)
> {
> diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
> index e47ef764f613..b0e33ba36695 100644
> --- a/include/linux/perf_event.h
> +++ b/include/linux/perf_event.h
> @@ -1154,9 +1154,10 @@ DECLARE_PER_CPU(struct perf_callchain_entry, perf_callchain_entry);
>
> extern void perf_callchain_user(struct perf_callchain_entry_ctx *entry, struct pt_regs *regs);
> extern void perf_callchain_kernel(struct perf_callchain_entry_ctx *entry, struct pt_regs *regs);
> +extern void perf_callchain_kernel_direct(struct perf_callchain_entry_ctx *entry, struct pt_regs *regs);
> extern struct perf_callchain_entry *
> get_perf_callchain(struct pt_regs *regs, u32 init_nr, bool kernel, bool user,
> - u32 max_stack, bool crosstask, bool add_mark);
> + u32 max_stack, bool crosstask, bool add_mark, bool direct_call);
> extern struct perf_callchain_entry *perf_callchain(struct perf_event *event, struct pt_regs *regs);
> extern int get_callchain_buffers(int max_stack);
> extern void put_callchain_buffers(void);
> diff --git a/kernel/bpf/stackmap.c b/kernel/bpf/stackmap.c
> index 950ab2f28922..376d774da6b5 100644
> --- a/kernel/bpf/stackmap.c
> +++ b/kernel/bpf/stackmap.c
> @@ -368,7 +368,7 @@ BPF_CALL_3(bpf_get_stackid, struct pt_regs *, regs, struct bpf_map *, map,
> return -EINVAL;
>
> trace = get_perf_callchain(regs, init_nr, kernel, user,
> - sysctl_perf_event_max_stack, false, false);
> + sysctl_perf_event_max_stack, false, false, false);
>
> if (unlikely(!trace))
> /* couldn't fetch the stack trace */
> @@ -476,7 +476,7 @@ BPF_CALL_4(bpf_get_stack, struct pt_regs *, regs, void *, buf, u32, size,
> else
> init_nr = sysctl_perf_event_max_stack - num_elem;
> trace = get_perf_callchain(regs, init_nr, kernel, user,
> - sysctl_perf_event_max_stack, false, false);
> + sysctl_perf_event_max_stack, false, false, false);
> if (unlikely(!trace))
> goto err_fault;
>
> diff --git a/kernel/events/callchain.c b/kernel/events/callchain.c
> index c2b41a263166..4cb5861f57ee 100644
> --- a/kernel/events/callchain.c
> +++ b/kernel/events/callchain.c
> @@ -40,6 +40,12 @@ __weak void perf_callchain_kernel(struct perf_callchain_entry_ctx *entry,
> {
> }
>
> +__weak void perf_callchain_kernel_direct(struct perf_callchain_entry_ctx *entry,
> + struct pt_regs *regs)
> +{
> + perf_callchain_kernel(entry, regs);
> +}
> +
> __weak void perf_callchain_user(struct perf_callchain_entry_ctx *entry,
> struct pt_regs *regs)
> {
> @@ -176,7 +182,7 @@ put_callchain_entry(int rctx)
>
> struct perf_callchain_entry *
> get_perf_callchain(struct pt_regs *regs, u32 init_nr, bool kernel, bool user,
> - u32 max_stack, bool crosstask, bool add_mark)
> + u32 max_stack, bool crosstask, bool add_mark, bool direct_call)
> {
> struct perf_callchain_entry *entry;
> struct perf_callchain_entry_ctx ctx;
> @@ -198,7 +204,10 @@ get_perf_callchain(struct pt_regs *regs, u32 init_nr, bool kernel, bool user,
> if (kernel && !user_mode(regs)) {
> if (add_mark)
> perf_callchain_store_context(&ctx, PERF_CONTEXT_KERNEL);
> - perf_callchain_kernel(&ctx, regs);
> + if (direct_call)
> + perf_callchain_kernel_direct(&ctx, regs);
> + else
> + perf_callchain_kernel(&ctx, regs);
> }
>
> if (user) {
> diff --git a/kernel/events/core.c b/kernel/events/core.c
> index 72d06e302e99..af9fb9b03283 100644
> --- a/kernel/events/core.c
> +++ b/kernel/events/core.c
> @@ -6377,7 +6377,7 @@ perf_callchain(struct perf_event *event, struct pt_regs *regs)
> return &__empty_callchain;
>
> callchain = get_perf_callchain(regs, 0, kernel, user,
> - max_stack, crosstask, true);
> + max_stack, crosstask, true, true);
> return callchain ?: &__empty_callchain;
> }
>
> --
> 2.20.1
>

\
 
 \ /
  Last update: 2019-04-05 09:56    [W:0.069 / U:1.056 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site