Merge pull request #1364 from palmtenor/multi_section_so

Fix shared library loaded into multiple sections

src/cc/bcc_proc.c

@@ -81,7 +81,8 @@ int bcc_procutils_each_module(int pid, bcc_procutils_modulecb callback,
   FILE *procmap;
   int ret;
 
-  snprintf(procmap_filename, sizeof(procmap_filename), "/proc/%ld/maps", (long)pid);
+  snprintf(procmap_filename, sizeof(procmap_filename), "/proc/%ld/maps",
+           (long)pid);
   procmap = fopen(procmap_filename, "r");
 
   if (!procmap)
@@ -90,10 +91,10 @@ int bcc_procutils_each_module(int pid, bcc_procutils_modulecb callback,
   do {
     char endline[4096];
     char perm[8], dev[8];
-    long long begin, end, size, inode;
+    long long begin, end, offset, inode;
 
     ret = fscanf(procmap, "%llx-%llx %s %llx %s %lld", &begin, &end, perm,
-                 &size, dev, &inode);
+                 &offset, dev, &inode);
 
     if (!fgets(endline, sizeof(endline), procmap))
       break;
@@ -105,10 +106,12 @@ int bcc_procutils_each_module(int pid, bcc_procutils_modulecb callback,
       if (newline)
         newline[0] = '\0';
 
-      while (isspace(mapname[0])) mapname++;
+      while (isspace(mapname[0]))
+        mapname++;
 
       if (strchr(perm, 'x') && bcc_mapping_is_file_backed(mapname)) {
-        if (callback(mapname, (uint64_t)begin, (uint64_t)end, true, payload) < 0)
+        if (callback(mapname, (uint64_t)begin, (uint64_t)end, (uint64_t)offset,
+                     true, payload) < 0)
           break;
       }
     }
@@ -122,12 +125,15 @@ int bcc_procutils_each_module(int pid, bcc_procutils_modulecb callback,
   // Try perf-<PID>.map path with process's mount namespace, chroot and NSPID,
   // in case it is generated by the process itself.
   if (bcc_perf_map_path(map_path, sizeof(map_path), pid))
-    callback(map_path, 0, -1, true, payload);
+    if (callback(map_path, 0, -1, 0, true, payload) < 0)
+      return 0;
   // Try perf-<PID>.map path with global root and PID, in case it is generated
   // by other Process. Avoid checking mount namespace for this.
   int res = snprintf(map_path, 4096, "/tmp/perf-%d.map", pid);
   if (res > 0 && res < 4096)
-    callback(map_path, 0, -1, false, payload);
+    if (callback(map_path, 0, -1, 0, false, payload) < 0)
+      return 0;
+
   return 0;
 }
 

src/cc/bcc_proc.h

@@ -24,14 +24,21 @@ extern "C" {
 
 #include <stdint.h>
 
-// Module name, start address, end address, whether to check mount namespace, payload
-typedef int (*bcc_procutils_modulecb)(const char *, uint64_t, uint64_t, bool, void *);
+// Module name, start address, end address, file_offset,
+// whether to check mount namespace, payload
+// Callback returning a negative value indicates to stop the iteration
+typedef int (*bcc_procutils_modulecb)(const char *, uint64_t, uint64_t,
+                                      uint64_t, bool, void *);
 // Symbol name, address, payload
 typedef void (*bcc_procutils_ksymcb)(const char *, uint64_t, void *);
 
 char *bcc_procutils_which_so(const char *libname, int pid);
 char *bcc_procutils_which(const char *binpath);
 int bcc_mapping_is_file_backed(const char *mapname);
+// Iterate over all executable memory mapping sections of a Process.
+// All anonymous and non-file-backed mapping sections, namely those
+// listed in bcc_mapping_is_file_backed, will be ignored.
+// Returns -1 on error, and 0 on success
 int bcc_procutils_each_module(int pid, bcc_procutils_modulecb callback,
                               void *payload);
 int bcc_procutils_each_ksym(bcc_procutils_ksymcb callback, void *payload);

src/cc/bcc_syms.cc

@@ -14,7 +14,6 @@
  * limitations under the License.
  */
 
-#include <algorithm>
 #include <cxxabi.h>
 #include <cstring>
 #include <fcntl.h>
@@ -114,7 +113,7 @@ ProcSyms::ProcSyms(int pid, struct bcc_symbol_option *option)
 int ProcSyms::_add_load_sections(uint64_t v_addr, uint64_t mem_sz,
                                  uint64_t file_offset, void *payload) {
   auto module = static_cast<Module *>(payload);
-  module->add_range(v_addr, v_addr + mem_sz);
+  module->ranges_.emplace_back(v_addr, v_addr + mem_sz, file_offset);
   return 0;
 }
 
@@ -148,7 +147,7 @@ void ProcSyms::refresh() {
 }
 
 int ProcSyms::_add_module(const char *modname, uint64_t start, uint64_t end,
-                          bool check_mount_ns, void *payload) {
+                          uint64_t offset, bool check_mount_ns, void *payload) {
   ProcSyms *ps = static_cast<ProcSyms *>(payload);
   auto it = std::find_if(
       ps->modules_.begin(), ps->modules_.end(),
@@ -162,7 +161,11 @@ int ProcSyms::_add_module(const char *modname, uint64_t start, uint64_t end,
     else
       return 0;
   }
-  it->add_range(start, end);
+  it->ranges_.emplace_back(start, end, offset);
+  // perf-PID map is added last. We try both inside the Process's mount
+  // namespace + chroot, and in global /tmp. Make sure we only add one.
+  if (it->type_ == ModuleType::PERF_MAP)
+    return -1;
 
   return 0;
 }
@@ -275,21 +278,12 @@ void ProcSyms::Module::load_sym_table() {
   std::sort(syms_.begin(), syms_.end());
 }
 
-void ProcSyms::Module::add_range(uint64_t st, uint64_t en) {
-  if (!ranges_.empty()) {
-    Range &last = ranges_.back();
-    if (st >= last.start && st <= last.end) {
-      last.end = std::max(en, last.end);
-      return;
-    }
-  }
-  ranges_.emplace_back(st, en);
-}
-
 bool ProcSyms::Module::contains(uint64_t addr, uint64_t &offset) const {
   for (const auto &range : ranges_)
     if (addr >= range.start && addr < range.end) {
-      offset = type_ == ModuleType::SO ? addr - range.start : addr;
+      offset = (type_ == ModuleType::SO)
+                   ? (addr - range.start + range.file_offset)
+                   : addr;
       return true;
     }
   return false;
@@ -401,13 +395,15 @@ void bcc_symcache_refresh(void *resolver) {
 struct mod_st {
   const char *name;
   uint64_t start;
+  uint64_t file_offset;
 };
 
-static int _find_module(const char *modname, uint64_t start, uint64_t end, bool,
-                        void *p) {
+static int _find_module(const char *modname, uint64_t start, uint64_t end,
+                        uint64_t offset, bool, void *p) {
   struct mod_st *mod = (struct mod_st *)p;
   if (!strcmp(modname, mod->name)) {
     mod->start = start;
+    mod->file_offset = offset;
     return -1;
   }
   return 0;
@@ -420,7 +416,7 @@ int bcc_resolve_global_addr(int pid, const char *module, const uint64_t address,
       mod.start == 0x0)
     return -1;
 
-  *global = mod.start + address;
+  *global = mod.start + mod.file_offset + address;
   return 0;
 }
 

src/cc/syms.h

@@ -92,10 +92,12 @@ class ProcSyms : SymbolCache {
     struct Range {
       uint64_t start;
       uint64_t end;
-      Range(uint64_t s, uint64_t e) : start(s), end(e) {}
+      uint64_t file_offset;
+      Range(uint64_t s, uint64_t e, uint64_t f)
+          : start(s), end(e), file_offset(f) {}
     };
 
-    Module(const char *name, ProcMountNS* mount_ns,
+    Module(const char *name, ProcMountNS *mount_ns,
            struct bcc_symbol_option *option);
     bool init();
 
@@ -111,7 +113,6 @@ class ProcSyms : SymbolCache {
 
     void load_sym_table();
 
-    void add_range(uint64_t st, uint64_t en);
     bool contains(uint64_t addr, uint64_t &offset) const;
     uint64_t start() const { return ranges_.begin()->start; }
 
@@ -130,7 +131,8 @@ class ProcSyms : SymbolCache {
 
   static int _add_load_sections(uint64_t v_addr, uint64_t mem_sz,
                                 uint64_t file_offset, void *payload);
-  static int _add_module(const char *, uint64_t, uint64_t, bool, void *);
+  static int _add_module(const char *, uint64_t, uint64_t, uint64_t, bool,
+                         void *);
   void load_exe();
   void load_modules();
 

src/cc/usdt.h

@@ -207,7 +207,8 @@ class Context {
 
   static void _each_probe(const char *binpath, const struct bcc_elf_usdt *probe,
                           void *p);
-  static int _each_module(const char *modpath, uint64_t, uint64_t, bool, void *p);
+  static int _each_module(const char *modpath, uint64_t, uint64_t, uint64_t,
+                          bool, void *p);
 
   void add_probe(const char *binpath, const struct bcc_elf_usdt *probe);
   std::string resolve_bin_path(const std::string &bin_path);

src/cc/usdt/usdt.cc

@@ -217,7 +217,8 @@ void Context::_each_probe(const char *binpath, const struct bcc_elf_usdt *probe,
   ctx->add_probe(binpath, probe);
 }
 
-int Context::_each_module(const char *modpath, uint64_t, uint64_t, bool, void *p) {
+int Context::_each_module(const char *modpath, uint64_t, uint64_t, uint64_t,
+                          bool, void *p) {
   Context *ctx = static_cast<Context *>(p);
   // Modules may be reported multiple times if they contain more than one
   // executable region. We are going to parse the ELF on disk anyway, so we