ProcSyms: deduplicate symbol names (#598)

diff --git a/src/cc/bcc_syms.cc b/src/cc/bcc_syms.cc
index b13cb88..be60485 100644 (file)
--- a/src/cc/bcc_syms.cc
+++ b/src/cc/bcc_syms.cc
@@ -132,7 +132,8 @@ bool ProcSyms::resolve_name(const char *module, const char *name,
 int ProcSyms::Module::_add_symbol(const char *symname, uint64_t start,
                                   uint64_t end, int flags, void *p) {
   Module *m = static_cast<Module *>(p);
-  m->syms_.emplace_back(symname, start, end, flags);
+  auto res = m->symnames_.emplace(symname);
+  m->syms_.emplace_back(&*(res.first), start, end, flags);
   return 0;
 }
 
@@ -160,7 +161,7 @@ bool ProcSyms::Module::find_name(const char *symname, uint64_t *addr) {
   load_sym_table();
 
   for (Symbol &s : syms_) {
-    if (s.name == symname) {
+    if (*(s.name) == symname) {
       *addr = is_so() ? start_ + s.start : s.start;
       return true;
     }
@@ -176,7 +177,7 @@ bool ProcSyms::Module::find_addr(uint64_t addr, struct bcc_symbol *sym) {
   sym->module = name_.c_str();
   sym->offset = offset;
 
-  auto it = std::upper_bound(syms_.begin(), syms_.end(), Symbol("", offset, 0));
+  auto it = std::upper_bound(syms_.begin(), syms_.end(), Symbol(nullptr, offset, 0));
   if (it != syms_.begin())
     --it;
   else
@@ -184,7 +185,7 @@ bool ProcSyms::Module::find_addr(uint64_t addr, struct bcc_symbol *sym) {
 
   if (it != syms_.end()
       && offset >= it->start && offset < it->start + it->size) {
-    sym->name = it->name.c_str();
+    sym->name = it->name->c_str();
     sym->offset = (offset - it->start);
     return true;
   }

diff --git a/src/cc/syms.h b/src/cc/syms.h
index eb34f2c..22bc30c 100644 (file)
--- a/src/cc/syms.h
+++ b/src/cc/syms.h
@@ -18,6 +18,7 @@
 #include <algorithm>
 #include <string>
 #include <unordered_map>
+#include <unordered_set>
 #include <vector>
 
 #include <sys/types.h>
@@ -63,9 +64,9 @@ public:
 
 class ProcSyms : SymbolCache {
   struct Symbol {
-    Symbol(const char *name, uint64_t start, uint64_t size, int flags = 0)
+    Symbol(const std::string *name, uint64_t start, uint64_t size, int flags = 0)
         : name(name), start(start), size(size), flags(flags) {}
-    std::string name;
+    const std::string *name;
     uint64_t start;
     uint64_t size;
     int flags;
@@ -81,6 +82,7 @@ class ProcSyms : SymbolCache {
     std::string name_;
     uint64_t start_;
     uint64_t end_;
+    std::unordered_set<std::string> symnames_;
     std::vector<Symbol> syms_;
 
     void load_sym_table();