adding support for global symbol config

one global tree to rule them all
- only one tree is searched during the plt entry/exit
- symbols are added during the bind audit callback

8 files changed, 251 insertions, 44 deletions

diff --git a/ChangeLog b/ChangeLog
index 27022f2..1bae726 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,9 @@
+2010-01-31 Jiri Olsa <olsajiri@gmail.com>
+	* adding support for global symbol config
+	  one global symbol tree to rule them all
+	    - only one tree is searched during the plt entry/exit
+	    - symbols are added during the bind audit callback
+
 2010-10-17 Jiri Olsa <olsajiri@gmail.com>
 	* Artur Skawina <art.08.09@gmail.com>
 		- enhancing names check with *-logic for
diff --git a/src/Makefile b/src/Makefile
index 796028a..eb16064 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -28,7 +28,8 @@ AUDIT_OBJS=\
 	src/fifo.o \
 	src/output.o \
 	src/objsearch.o \
-	src/stack.o
+	src/stack.o \
+	src/symbol.o
 
 ifeq ($(CONFIG_ARCH_HAVE_ARGS),y)
 AUDIT_OBJS+=\
diff --git a/src/args.c b/src/args.c
index 8672156..a7b745e 100644
--- a/src/args.c
+++ b/src/args.c
@@ -749,25 +749,6 @@ int lt_args_init(struct lt_config_shared *cfg)
 	return ret;
 }
 
-static struct lt_args_sym* getsym(struct lt_config_shared *cfg, char *sym)
-{
-	struct lt_args_sym *a;
-	ENTRY e, *ep;
-
-        PRINT_VERBOSE(cfg, 1, "request for <%s>\n", sym);
-
-	e.key = sym;
-	hsearch_r(e, FIND, &ep, &cfg->args_tab);
-
-	if (!ep)
-		return NULL;
-
-	a = (struct lt_args_sym*) ep->data;
-
-        PRINT_VERBOSE(cfg, 1, "found %p <%s>\n", a, a->name);
-	return a;
-}
-
 static int getstr_addenum(struct lt_config_shared *cfg, struct lt_arg *arg,
 			char *argbuf, int alen, long val)
 {
@@ -1096,12 +1077,32 @@ struct lt_args_include* lt_args_buf_get(void)
 	return inc;
 }
 
-int lt_args_sym_entry(struct lt_config_shared *cfg, char *sym, La_regs *regs, 
-			char **argbuf, char **argdbuf)
+struct lt_args_sym* lt_args_sym_get(struct lt_config_shared *cfg,
+				    const char *sym)
+{
+	struct lt_args_sym *a;
+	ENTRY e, *ep;
+
+	PRINT_VERBOSE(cfg, 1, "request for <%s>\n", sym);
+
+	e.key = (char*) sym;
+	hsearch_r(e, FIND, &ep, &cfg->args_tab);
+
+	if (!ep)
+		return NULL;
+
+	a = (struct lt_args_sym*) ep->data;
+
+	PRINT_VERBOSE(cfg, 1, "found %p <%s>\n", a, a->name);
+	return a;
+}
+
+int lt_args_sym_entry(struct lt_config_shared *cfg, struct lt_symbol *sym,
+			La_regs *regs, char **argbuf, char **argdbuf)
 {
-	struct lt_args_sym *asym;
+	struct lt_args_sym *asym = sym ? sym->args : NULL;
 
-	if (NULL == (asym = getsym(cfg, sym)))
+	if (!asym)
 		return -1;
 
 	return getargs(cfg, asym, regs, argbuf, argdbuf);
@@ -1146,12 +1147,13 @@ static int getargs_ret(struct lt_config_shared *cfg, struct lt_args_sym *asym,
 	return lt_stack_process_ret(cfg, asym, regs, &data);
 }
 
-int lt_args_sym_exit(struct lt_config_shared *cfg, char *sym, La_regs *inregs, La_retval *outregs, 
+int lt_args_sym_exit(struct lt_config_shared *cfg, struct lt_symbol *sym,
+			La_regs *inregs, La_retval *outregs,
 			char **argbuf, char **argdbuf)
 {
-	struct lt_args_sym *asym;
+	struct lt_args_sym *asym = sym ? sym->args : NULL;
 
-	if (NULL == (asym = getsym(cfg, sym)))
+	if (!asym)
 		return -1;
 
 	return getargs_ret(cfg, asym, outregs, argbuf, argdbuf);
diff --git a/src/args.h b/src/args.h
index b7fa427..e0f4846 100644
--- a/src/args.h
+++ b/src/args.h
@@ -23,6 +23,7 @@
 #define ARGS_H
 
 struct lt_config_shared;
+struct lt_symbol;
 
 enum {
 	LT_ARGS_DTYPE_POD = 1,
@@ -126,10 +127,13 @@ struct lt_args_data {
 
 /* arguments */
 int lt_args_init(struct lt_config_shared *cfg);
-int lt_args_sym_entry(struct lt_config_shared *cfg, char *sym, La_regs *regs,
+struct lt_args_sym* lt_args_sym_get(struct lt_config_shared *cfg,
+					const char *sym);
+int lt_args_sym_entry(struct lt_config_shared *cfg, struct lt_symbol *sym,
+			La_regs *regs, char **argbuf, char **argdbuf);
+int lt_args_sym_exit(struct lt_config_shared *cfg, struct lt_symbol *sym,
+			La_regs *inregs, La_retval *outregs,
 			char **argbuf, char **argdbuf);
-int lt_args_sym_exit(struct lt_config_shared *cfg, char *sym, La_regs *inregs,
-			La_retval *outregs, char **argbuf, char **argdbuf);
 int lt_args_add_enum(struct lt_config_shared *cfg, char *name,
 			struct lt_list_head *h);
 struct lt_enum_elem* lt_args_get_enum(struct lt_config_shared *cfg, char *name, char *val);
diff --git a/src/audit-init.c b/src/audit-init.c
index 686c56d..fad89ca 100644
--- a/src/audit-init.c
+++ b/src/audit-init.c
@@ -242,6 +242,10 @@ int audit_init(int argc, char **argv, char **env)
 	if (lt_sh(&cfg, not_follow_fork))
 		lt_sh(&cfg, pid) = getpid();
 
+	/* enable global symbols if needed */
+	lt_sh(&cfg, global_symbols) = lt_sh(&cfg, args_enabled);
+	PRINT_VERBOSE(&cfg, 1, "global_symbols %d\n", lt_sh(&cfg, global_symbols));
+
 	cfg.init_ok = 1;
 	return 0;
 }
diff --git a/src/audit.c b/src/audit.c
index 2313c7c..d26a539 100644
--- a/src/audit.c
+++ b/src/audit.c
@@ -70,12 +70,13 @@ static int check_flow_below(const char *symname, int in)
 	return ret;
 }
 
-static int sym_entry(const char *symname, char *lib_from, char *lib_to, 
-			La_regs *regs)
+static int sym_entry(const char *symname, void *ptr,
+		     char *lib_from, char *lib_to, La_regs *regs)
 {
 	int argret = -1;
 	char *argbuf = "", *argdbuf = "";
 	struct timeval tv;
+	struct lt_symbol *sym = NULL;
 
 	PRINT_VERBOSE(&cfg, 2, "%s@%s\n", symname, lib_to);
 
@@ -85,9 +86,12 @@ static int sym_entry(const char *symname, char *lib_from, char *lib_to,
 	if (lt_sh(&cfg, timestamp) || lt_sh(&cfg, counts))
 		gettimeofday(&tv, NULL);
 
+	if (lt_sh(&cfg, global_symbols))
+		sym = lt_symbol_get(cfg.sh, ptr, symname);
+
 #ifdef CONFIG_ARCH_HAVE_ARGS
 	argret = lt_sh(&cfg, args_enabled) ?
-		lt_args_sym_entry(cfg.sh, (char*) symname, regs, &argbuf, &argdbuf) : -1;
+		lt_args_sym_entry(cfg.sh, sym, regs, &argbuf, &argdbuf) : -1;
 #endif
 
 	if (lt_sh(&cfg, pipe)) {
@@ -119,12 +123,14 @@ static int sym_entry(const char *symname, char *lib_from, char *lib_to,
 	return 0;
 }
 
-static int sym_exit(const char *symname, char *lib_from, char *lib_to, 
-			const La_regs *inregs, La_retval *outregs)
+static int sym_exit(const char *symname, void *ptr,
+			 char *lib_from, char *lib_to,
+			 const La_regs *inregs, La_retval *outregs)
 {
 	int argret = -1;
 	char *argbuf = "", *argdbuf = "";
 	struct timeval tv;
+	struct lt_symbol *sym = NULL;
 
 	PRINT_VERBOSE(&cfg, 2, "%s@%s\n", symname, lib_to);
 
@@ -134,9 +140,12 @@ static int sym_exit(const char *symname, char *lib_from, char *lib_to,
 	if (lt_sh(&cfg, timestamp) || lt_sh(&cfg, counts))
 		gettimeofday(&tv, NULL);
 
+	if (lt_sh(&cfg, global_symbols))
+		sym = lt_symbol_get(cfg.sh, ptr, symname);
+
 #ifdef CONFIG_ARCH_HAVE_ARGS
 	argret = lt_sh(&cfg, args_enabled) ?
-		lt_args_sym_exit(cfg.sh, (char*) symname,
+		lt_args_sym_exit(cfg.sh, sym,
 			(La_regs*) inregs, outregs, &argbuf, &argdbuf) : -1;
 #endif
 
@@ -225,7 +234,7 @@ unsigned int la_objopen(struct link_map *l, Lmid_t a, uintptr_t *cookie)
 	return 0;
 }
 
-static unsigned int la_symbind(const char *symname)
+static unsigned int la_symbind(ElfW(Sym) *sym, const char *symname)
 {
 	unsigned int flags = 0;
 
@@ -240,6 +249,10 @@ static unsigned int la_symbind(const char *symname)
 			flags = LA_SYMB_NOPLTENTER|LA_SYMB_NOPLTEXIT;
 	}
 
+	/* we are interested in this symbol */
+	if (!(flags & LA_SYMB_NOPLTENTER))
+		lt_symbol_bind(cfg.sh, (void*) sym->st_value, symname);
+
 	return flags;
 }
 
@@ -267,19 +280,21 @@ unsigned int la_objclose(uintptr_t *__cookie)
 	return 0;
 }
 
+#if __ELF_NATIVE_CLASS == 32
 uintptr_t la_symbind32(Elf32_Sym *sym, unsigned int ndx, uintptr_t *refcook,
 		uintptr_t *defcook, unsigned int *flags, const char *symname)
 {
-	*flags = la_symbind(symname);
+	*flags = la_symbind(sym, symname);
 	return sym->st_value;
 }
-
+#elif __ELF_NATIVE_CLASS == 64
 uintptr_t la_symbind64(Elf64_Sym *sym, unsigned int ndx, uintptr_t *refcook,
 		uintptr_t *defcook, unsigned int *flags, const char *symname)
 {
-	*flags = la_symbind(symname);
+	*flags = la_symbind(sym, symname);
 	return sym->st_value;
 }
+#endif
 
 ElfW(Addr)
 pltenter(ElfW(Sym) *sym, unsigned int ndx, uintptr_t *refcook,
@@ -295,8 +310,10 @@ pltenter(ElfW(Sym) *sym, unsigned int ndx, uintptr_t *refcook,
 
 		CHECK_PID(sym->st_value);
 
-		sym_entry(symname, lr ? lr->l_name : NULL,
-				   ld ? ld->l_name : NULL, regs);
+		sym_entry(symname, (void*) sym->st_value,
+			  lr ? lr->l_name : NULL,
+			  ld ? ld->l_name : NULL,
+			  regs);
 
 	} while(0);
 
@@ -318,8 +335,10 @@ unsigned int pltexit(ElfW(Sym) *sym, unsigned int ndx, uintptr_t *refcook,
 
 		CHECK_PID(0);
 
-		sym_exit(symname, lr ? lr->l_name : NULL,
-				  ld ? ld->l_name : NULL, inregs, outregs);
+		sym_exit(symname, (void*) sym->st_value,
+			 lr ? lr->l_name : NULL,
+			 ld ? ld->l_name : NULL,
+			 inregs, outregs);
 
 	} while(0);
 
diff --git a/src/config.h b/src/config.h
index 29c9c4f..c7f88ee 100644
--- a/src/config.h
+++ b/src/config.h
@@ -118,6 +118,7 @@ struct lt_config_shared {
 	int not_follow_fork;
 	int framesize_check;
 	unsigned int framesize;
+	int global_symbols;
 
 	/* for 'not_follow_fork' */
 	pid_t pid;
@@ -260,6 +261,15 @@ struct lt_thread {
 	struct lt_thread *next;
 };
 
+struct lt_symbol {
+	struct lt_args_sym *args;
+
+	/* symbol name */
+	const char *name;
+	/* symbol address */
+	void *ptr;
+};
+
 /* ctl */
 int main_ctl(int argc, char **argv);
 
@@ -307,6 +317,12 @@ char* lt_objsearch(struct lt_config_audit *cfg, const char *name,
 /* stack */
 int lt_stack_framesize(struct lt_config_audit *cfg, La_regs *regs);
 
+/* symbol */
+struct lt_symbol* lt_symbol_bind(struct lt_config_shared *cfg,
+				void *ptr, const char *name);
+struct lt_symbol* lt_symbol_get(struct lt_config_shared *cfg,
+				void *ptr, const char *name);
+
 #define PRINT(fmt, args...) \
 do { \
 	char lpbuf[1024]; \
@@ -316,6 +332,7 @@ do { \
 		__LINE__, \
 		fmt); \
 	printf(lpbuf, ## args); \
+	fflush(NULL); \
 } while(0)
 
 #define PRINT_VERBOSE(cfg, cond, fmt, args...) \
diff --git a/src/symbol.c b/src/symbol.c
new file mode 100644
index 0000000..c4e25b9
--- /dev/null
+++ b/src/symbol.c
@@ -0,0 +1,154 @@
+/*
+  Copyright (C) 2011 Jiri Olsa <olsajiri@gmail.com>
+
+  This file is part of the latrace.
+
+  The latrace is free software: you can redistribute it and/or modify
+  it under the terms of the GNU General Public License as published by
+  the Free Software Foundation, either version 3 of the License, or
+  (at your option) any later version.
+
+  The latrace is distributed in the hope that it will be useful,
+  but WITHOUT ANY WARRANTY; without even the implied warranty of
+  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+  GNU General Public License for more details.
+
+  You should have received a copy of the GNU General Public License
+  along with the latrace (file COPYING).  If not, see
+  <http://www.gnu.org/licenses/>.
+*/
+
+
+#include <strings.h>
+#include <stdlib.h>
+#include <string.h>
+
+#include "config.h"
+
+static void *root = NULL;
+extern struct lt_config_audit cfg;
+
+static int compare(const void *a, const void *b)
+{
+	const struct lt_symbol *sym_a = a;
+	const struct lt_symbol *sym_b = b;
+
+	PRINT_VERBOSE(&cfg, 1, "a %p, b %p\n", sym_a, sym_b);
+	PRINT_VERBOSE(&cfg, 1, "a ptr %p, b ptr  %p\n",
+			sym_a ? sym_a->ptr : NULL,
+			sym_b ? sym_b->ptr : NULL);
+	PRINT_VERBOSE(&cfg, 1, "a name %s, b name %s\n",
+			sym_a ? sym_a->name : "",
+			sym_b ? sym_b->name: "");
+
+	if (!sym_a || !sym_b)
+		return 1;
+
+/* XXX There's a glibc bug/feature causing the symbol to
+ * have different value in plt entry/exit.. so using name
+ * check instead.. so far.. ;) */
+
+#define SEARCH_NAME
+#undef SEARCH_PTR
+
+#ifdef SEARCH_NAME
+	return strcmp(sym_a->name, sym_b->name);
+#endif
+
+#if SEARCH_PTR
+	PRINT_VERBOSE(&cfg, 1, "%s(%p) %s(%p)\n",
+			sym_a->name, sym_a->ptr,
+			sym_b->name, sym_b->ptr);
+
+	if (sym_a->ptr > sym_b->ptr)
+		return 1;
+
+	if (sym_a->ptr < sym_b->ptr)
+		return -1;
+
+	return 0;
+#endif
+}
+
+static int symbol_init(struct lt_config_shared *cfg,
+		       struct lt_symbol *sym, const char *name)
+{
+	struct lt_args_sym* a;
+
+	if (lt_sh(cfg, args_enabled)) {
+		a = lt_args_sym_get(cfg, name);
+		if (!a)
+			return -1;
+	}
+
+	sym->args = a;
+
+	PRINT_VERBOSE(cfg, 1, "ok name %s, ptr %p, sym %p\n",
+			name, sym->ptr, sym);
+	return 0;
+}
+
+struct lt_symbol* lt_symbol_bind(struct lt_config_shared *cfg,
+				void *ptr, const char *name)
+{
+	static struct lt_symbol *sym = NULL;
+	struct lt_symbol *s = NULL;
+	void *val;
+
+	if (!sym) {
+		sym = malloc(sizeof(*sym));
+		if (!sym)
+			return NULL;
+	}
+
+	bzero(sym, sizeof(*sym));
+	sym->ptr  = ptr;
+	sym->name = name;
+
+	PRINT_VERBOSE(cfg, 1, "checking %s(%p)\n", name, ptr);
+
+	val = tsearch((void *) sym, &root, compare);
+	if (!val)
+		return NULL;
+
+	s = (*(void**) val);
+
+	/* symbol already in */
+	if (s != sym) {
+		PRINT_VERBOSE(cfg, 1, "found %s, ptr %p, sym %p\n",
+				name, sym->ptr, sym);
+		return s;
+	}
+
+	PRINT_VERBOSE(cfg, 1, "added %s, ptr %p, sym %p\n",
+			name, sym->ptr, sym);
+
+	/* not interesting symbol */
+	if (symbol_init(cfg, sym, name))
+		return NULL;
+
+	/* symbol properly added */
+	sym = NULL;
+	return s;
+}
+
+struct lt_symbol* lt_symbol_get(struct lt_config_shared *cfg,
+				void *ptr, const char *name)
+{
+	void *val;
+	struct lt_symbol *s;
+	struct lt_symbol sym = {
+		.ptr  = ptr,
+		.name = name,
+	};
+
+	val = tfind(&sym, &root, compare);
+
+	if (!val)
+		s = NULL;
+	else
+		s = *(struct lt_symbol**) val;
+
+	PRINT_VERBOSE(cfg, 1, "found %p '%s'\n", s, s ? s->name : "");
+	return s;
+}