linux_dsm_epyc7002/tools/perf/util/callchain.h
Sukadev Bhattiprolu a60335ba32 perf tools powerpc: Adjust callchain based on DWARF debug info
When saving the callchain on Power, the kernel conservatively saves excess
entries in the callchain. A few of these entries are needed in some cases
but not others. We should use the DWARF debug information to determine
when the entries are  needed.

Eg: the value in the link register (LR) is needed only when it holds the
return address of a function. At other times it must be ignored.

If the unnecessary entries are not ignored, we end up with duplicate arcs
in the call-graphs.

Use the DWARF debug information to determine if any callchain entries
should be ignored when building call-graphs.

Callgraph before the patch:

    14.67%          2234  sprintft  libc-2.18.so       [.] __random
            |
            --- __random
               |
               |--61.12%-- __random
               |          |
               |          |--97.15%-- rand
               |          |          do_my_sprintf
               |          |          main
               |          |          generic_start_main.isra.0
               |          |          __libc_start_main
               |          |          0x0
               |          |
               |           --2.85%-- do_my_sprintf
               |                     main
               |                     generic_start_main.isra.0
               |                     __libc_start_main
               |                     0x0
               |
                --38.88%-- rand
                          |
                          |--94.01%-- rand
                          |          do_my_sprintf
                          |          main
                          |          generic_start_main.isra.0
                          |          __libc_start_main
                          |          0x0
                          |
                           --5.99%-- do_my_sprintf
                                     main
                                     generic_start_main.isra.0
                                     __libc_start_main
                                     0x0

Callgraph after the patch:

    14.67%          2234  sprintft  libc-2.18.so       [.] __random
            |
            --- __random
               |
               |--95.93%-- rand
               |          do_my_sprintf
               |          main
               |          generic_start_main.isra.0
               |          __libc_start_main
               |          0x0
               |
                --4.07%-- do_my_sprintf
                          main
                          generic_start_main.isra.0
                          __libc_start_main
                          0x0

TODO:	For split-debug info objects like glibc, we can only determine
	the call-frame-address only when both .eh_frame and .debug_info
	sections are available. We should be able to determin the CFA
	even without the .eh_frame section.

Fix suggested by Anton Blanchard.

Thanks to valuable input on DWARF debug information from Ulrich Weigand.

Reported-by: Maynard Johnson <maynard@us.ibm.com>
Tested-by: Maynard Johnson <maynard@us.ibm.com>
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Link: http://lkml.kernel.org/r/20140625154903.GA29607@us.ibm.com
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
2014-06-27 11:14:51 +02:00

193 lines
4.6 KiB
C

#ifndef __PERF_CALLCHAIN_H
#define __PERF_CALLCHAIN_H
#include "../perf.h"
#include <linux/list.h>
#include <linux/rbtree.h>
#include "event.h"
#include "symbol.h"
enum perf_call_graph_mode {
CALLCHAIN_NONE,
CALLCHAIN_FP,
CALLCHAIN_DWARF,
CALLCHAIN_MAX
};
enum chain_mode {
CHAIN_NONE,
CHAIN_FLAT,
CHAIN_GRAPH_ABS,
CHAIN_GRAPH_REL
};
enum chain_order {
ORDER_CALLER,
ORDER_CALLEE
};
struct callchain_node {
struct callchain_node *parent;
struct list_head val;
struct rb_node rb_node_in; /* to insert nodes in an rbtree */
struct rb_node rb_node; /* to sort nodes in an output tree */
struct rb_root rb_root_in; /* input tree of children */
struct rb_root rb_root; /* sorted output tree of children */
unsigned int val_nr;
u64 hit;
u64 children_hit;
};
struct callchain_root {
u64 max_depth;
struct callchain_node node;
};
struct callchain_param;
typedef void (*sort_chain_func_t)(struct rb_root *, struct callchain_root *,
u64, struct callchain_param *);
enum chain_key {
CCKEY_FUNCTION,
CCKEY_ADDRESS
};
struct callchain_param {
enum chain_mode mode;
u32 print_limit;
double min_percent;
sort_chain_func_t sort;
enum chain_order order;
enum chain_key key;
};
struct callchain_list {
u64 ip;
struct map_symbol ms;
struct list_head list;
};
/*
* A callchain cursor is a single linked list that
* let one feed a callchain progressively.
* It keeps persistent allocated entries to minimize
* allocations.
*/
struct callchain_cursor_node {
u64 ip;
struct map *map;
struct symbol *sym;
struct callchain_cursor_node *next;
};
struct callchain_cursor {
u64 nr;
struct callchain_cursor_node *first;
struct callchain_cursor_node **last;
u64 pos;
struct callchain_cursor_node *curr;
};
extern __thread struct callchain_cursor callchain_cursor;
static inline void callchain_init(struct callchain_root *root)
{
INIT_LIST_HEAD(&root->node.val);
root->node.parent = NULL;
root->node.hit = 0;
root->node.children_hit = 0;
root->node.rb_root_in = RB_ROOT;
root->max_depth = 0;
}
static inline u64 callchain_cumul_hits(struct callchain_node *node)
{
return node->hit + node->children_hit;
}
int callchain_register_param(struct callchain_param *param);
int callchain_append(struct callchain_root *root,
struct callchain_cursor *cursor,
u64 period);
int callchain_merge(struct callchain_cursor *cursor,
struct callchain_root *dst, struct callchain_root *src);
/*
* Initialize a cursor before adding entries inside, but keep
* the previously allocated entries as a cache.
*/
static inline void callchain_cursor_reset(struct callchain_cursor *cursor)
{
cursor->nr = 0;
cursor->last = &cursor->first;
}
int callchain_cursor_append(struct callchain_cursor *cursor, u64 ip,
struct map *map, struct symbol *sym);
/* Close a cursor writing session. Initialize for the reader */
static inline void callchain_cursor_commit(struct callchain_cursor *cursor)
{
cursor->curr = cursor->first;
cursor->pos = 0;
}
/* Cursor reading iteration helpers */
static inline struct callchain_cursor_node *
callchain_cursor_current(struct callchain_cursor *cursor)
{
if (cursor->pos == cursor->nr)
return NULL;
return cursor->curr;
}
static inline void callchain_cursor_advance(struct callchain_cursor *cursor)
{
cursor->curr = cursor->curr->next;
cursor->pos++;
}
struct option;
struct hist_entry;
int record_parse_callchain(const char *arg, struct record_opts *opts);
int record_parse_callchain_opt(const struct option *opt, const char *arg, int unset);
int record_callchain_opt(const struct option *opt, const char *arg, int unset);
int sample__resolve_callchain(struct perf_sample *sample, struct symbol **parent,
struct perf_evsel *evsel, struct addr_location *al,
int max_stack);
int hist_entry__append_callchain(struct hist_entry *he, struct perf_sample *sample);
int fill_callchain_info(struct addr_location *al, struct callchain_cursor_node *node,
bool hide_unresolved);
extern const char record_callchain_help[];
int parse_callchain_report_opt(const char *arg);
static inline void callchain_cursor_snapshot(struct callchain_cursor *dest,
struct callchain_cursor *src)
{
*dest = *src;
dest->first = src->curr;
dest->nr -= src->pos;
}
#ifdef HAVE_SKIP_CALLCHAIN_IDX
extern int arch_skip_callchain_idx(struct machine *machine,
struct thread *thread, struct ip_callchain *chain);
#else
static inline int arch_skip_callchain_idx(struct machine *machine __maybe_unused,
struct thread *thread __maybe_unused,
struct ip_callchain *chain __maybe_unused)
{
return -1;
}
#endif
#endif /* __PERF_CALLCHAIN_H */