diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2018-11-30 18:32:34 +0100 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2018-11-30 18:32:34 +0100 |
commit | 0f1f692375450338a36af308cbb538ffabd130f9 (patch) | |
tree | 073953b2000841a3c58382d1046f9683b528d5ac /kernel | |
parent | Merge tag 'drm-fixes-2018-11-30' of git://anongit.freedesktop.org/drm/drm (diff) | |
parent | sched, trace: Fix prev_state output in sched_switch tracepoint (diff) | |
download | linux-0f1f692375450338a36af308cbb538ffabd130f9.tar.xz linux-0f1f692375450338a36af308cbb538ffabd130f9.zip |
Merge tag 'trace-v4.20-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace
Pull tracing fixes from Steven Rostedt:
"While rewriting the function graph tracer, I discovered a design flaw
that was introduced by a patch that tried to fix one bug, but by doing
so created another bug.
As both bugs corrupt the output (but they do not crash the kernel), I
decided to fix the design such that it could have both bugs fixed. The
original fix, fixed time reporting of the function graph tracer when
doing a max_depth of one. This was code that can test how much the
kernel interferes with userspace. But in doing so, it could corrupt
the time keeping of the function profiler.
The issue is that the curr_ret_stack variable was being used for two
different meanings. One was to keep track of the stack pointer on the
ret_stack (shadow stack used by the function graph tracer), and the
other use case was the graph call depth. Although, the two may be
closely related, where they got updated was the issue that lead to the
two different bugs that required the two use cases to be updated
differently.
The big issue with this fix is that it requires changing each
architecture. The good news is, I was able to remove a lot of code
that was duplicated within the architectures and place it into a
single location. Then I could make the fix in one place.
I pushed this code into linux-next to let it settle over a week, and
before doing so, I cross compiled all the affected architectures to
make sure that they built fine.
In the mean time, I also pulled in a patch that fixes the sched_switch
previous tasks state output, that was not actually correct"
* tag 'trace-v4.20-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
sched, trace: Fix prev_state output in sched_switch tracepoint
function_graph: Have profiler use curr_ret_stack and not depth
function_graph: Reverse the order of pushing the ret_stack and the callback
function_graph: Move return callback before update of curr_ret_stack
function_graph: Use new curr_ret_depth to manage depth instead of curr_ret_stack
function_graph: Make ftrace_push_return_trace() static
sparc/function_graph: Simplify with function_graph_enter()
sh/function_graph: Simplify with function_graph_enter()
s390/function_graph: Simplify with function_graph_enter()
riscv/function_graph: Simplify with function_graph_enter()
powerpc/function_graph: Simplify with function_graph_enter()
parisc: function_graph: Simplify with function_graph_enter()
nds32: function_graph: Simplify with function_graph_enter()
MIPS: function_graph: Simplify with function_graph_enter()
microblaze: function_graph: Simplify with function_graph_enter()
arm64: function_graph: Simplify with function_graph_enter()
ARM: function_graph: Simplify with function_graph_enter()
x86/function_graph: Simplify with function_graph_enter()
function_graph: Create function_graph_enter() to consolidate architecture code
Diffstat (limited to 'kernel')
-rw-r--r-- | kernel/trace/ftrace.c | 7 | ||||
-rw-r--r-- | kernel/trace/trace_functions_graph.c | 49 |
2 files changed, 43 insertions, 13 deletions
diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index f536f601bd46..77734451cb05 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -817,7 +817,7 @@ function_profile_call(unsigned long ip, unsigned long parent_ip, #ifdef CONFIG_FUNCTION_GRAPH_TRACER static int profile_graph_entry(struct ftrace_graph_ent *trace) { - int index = trace->depth; + int index = current->curr_ret_stack; function_profile_call(trace->func, 0, NULL, NULL); @@ -852,7 +852,7 @@ static void profile_graph_return(struct ftrace_graph_ret *trace) if (!fgraph_graph_time) { int index; - index = trace->depth; + index = current->curr_ret_stack; /* Append this call time to the parent time to subtract */ if (index) @@ -6814,6 +6814,7 @@ static int alloc_retstack_tasklist(struct ftrace_ret_stack **ret_stack_list) atomic_set(&t->tracing_graph_pause, 0); atomic_set(&t->trace_overrun, 0); t->curr_ret_stack = -1; + t->curr_ret_depth = -1; /* Make sure the tasks see the -1 first: */ smp_wmb(); t->ret_stack = ret_stack_list[start++]; @@ -7038,6 +7039,7 @@ graph_init_task(struct task_struct *t, struct ftrace_ret_stack *ret_stack) void ftrace_graph_init_idle_task(struct task_struct *t, int cpu) { t->curr_ret_stack = -1; + t->curr_ret_depth = -1; /* * The idle task has no parent, it either has its own * stack or no stack at all. @@ -7068,6 +7070,7 @@ void ftrace_graph_init_task(struct task_struct *t) /* Make sure we do not use the parent ret_stack */ t->ret_stack = NULL; t->curr_ret_stack = -1; + t->curr_ret_depth = -1; if (ftrace_graph_active) { struct ftrace_ret_stack *ret_stack; diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index 169b3c44ee97..2561460d7baf 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -118,8 +118,8 @@ print_graph_duration(struct trace_array *tr, unsigned long long duration, struct trace_seq *s, u32 flags); /* Add a function return address to the trace stack on thread info.*/ -int -ftrace_push_return_trace(unsigned long ret, unsigned long func, int *depth, +static int +ftrace_push_return_trace(unsigned long ret, unsigned long func, unsigned long frame_pointer, unsigned long *retp) { unsigned long long calltime; @@ -177,9 +177,31 @@ ftrace_push_return_trace(unsigned long ret, unsigned long func, int *depth, #ifdef HAVE_FUNCTION_GRAPH_RET_ADDR_PTR current->ret_stack[index].retp = retp; #endif - *depth = current->curr_ret_stack; + return 0; +} + +int function_graph_enter(unsigned long ret, unsigned long func, + unsigned long frame_pointer, unsigned long *retp) +{ + struct ftrace_graph_ent trace; + + trace.func = func; + trace.depth = ++current->curr_ret_depth; + + if (ftrace_push_return_trace(ret, func, + frame_pointer, retp)) + goto out; + + /* Only trace if the calling function expects to */ + if (!ftrace_graph_entry(&trace)) + goto out_ret; return 0; + out_ret: + current->curr_ret_stack--; + out: + current->curr_ret_depth--; + return -EBUSY; } /* Retrieve a function return address to the trace stack on thread info.*/ @@ -241,7 +263,13 @@ ftrace_pop_return_trace(struct ftrace_graph_ret *trace, unsigned long *ret, trace->func = current->ret_stack[index].func; trace->calltime = current->ret_stack[index].calltime; trace->overrun = atomic_read(¤t->trace_overrun); - trace->depth = index; + trace->depth = current->curr_ret_depth--; + /* + * We still want to trace interrupts coming in if + * max_depth is set to 1. Make sure the decrement is + * seen before ftrace_graph_return. + */ + barrier(); } /* @@ -255,6 +283,12 @@ unsigned long ftrace_return_to_handler(unsigned long frame_pointer) ftrace_pop_return_trace(&trace, &ret, frame_pointer); trace.rettime = trace_clock_local(); + ftrace_graph_return(&trace); + /* + * The ftrace_graph_return() may still access the current + * ret_stack structure, we need to make sure the update of + * curr_ret_stack is after that. + */ barrier(); current->curr_ret_stack--; /* @@ -267,13 +301,6 @@ unsigned long ftrace_return_to_handler(unsigned long frame_pointer) return ret; } - /* - * The trace should run after decrementing the ret counter - * in case an interrupt were to come in. We don't want to - * lose the interrupt if max_depth is set. - */ - ftrace_graph_return(&trace); - if (unlikely(!ret)) { ftrace_graph_stop(); WARN_ON(1); |