linux-brain/scripts/gcc-plugins/structleak_plugin.c

265 lines
7.4 KiB
C
Raw Permalink Normal View History

/*
* Copyright 2013-2017 by PaX Team <pageexec@freemail.hu>
* Licensed under the GPL v2
*
* Note: the choice of the license means that the compilation process is
* NOT 'eligible' as defined by gcc's library exception to the GPL v3,
* but for the kernel it doesn't matter since it doesn't link against
* any of the gcc libraries
*
* gcc plugin to forcibly initialize certain local variables that could
* otherwise leak kernel stack to userland if they aren't properly initialized
* by later code
*
* Homepage: http://pax.grsecurity.net/
*
* Options:
* -fplugin-arg-structleak_plugin-disable
* -fplugin-arg-structleak_plugin-verbose
gcc-plugins: structleak: Generalize to all variable types This adjusts structleak to also work with non-struct types when they are passed by reference, since those variables may leak just like anything else. This is exposed via an improved set of Kconfig options. (This does mean structleak is slightly misnamed now.) Building with CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL should give the kernel complete initialization coverage of all stack variables passed by reference, including padding (see lib/test_stackinit.c). Using CONFIG_GCC_PLUGIN_STRUCTLEAK_VERBOSE to count added initializations under defconfig: ..._BYREF: 5945 added initializations ..._BYREF_ALL: 16606 added initializations There is virtually no change to text+data size (both have less than 0.05% growth): text data bss dec hex filename 19502103 5051456 1917000 26470559 193e89f vmlinux.stock 19513412 5051456 1908808 26473676 193f4cc vmlinux.byref 19516974 5047360 1900616 26464950 193d2b6 vmlinux.byref_all The measured performance difference is in the noise for hackbench and kernel build benchmarks: Stock: 5x hackbench -g 20 -l 1000 Mean: 10.649s Std Dev: 0.339 5x kernel build (4-way parallel) Mean: 261.98s Std Dev: 1.53 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF: 5x hackbench -g 20 -l 1000 Mean: 10.540s Std Dev: 0.233 5x kernel build (4-way parallel) Mean: 260.52s Std Dev: 1.31 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL: 5x hackbench -g 20 -l 1000 Mean: 10.320 Std Dev: 0.413 5x kernel build (4-way parallel) Mean: 260.10 Std Dev: 0.86 This does not yet solve missing padding initialization for structures on the stack that are never passed by reference (which should be a tiny minority). Hopefully this will be more easily addressed by upstream compiler fixes after clarifying the C11 padding initialization specification. Signed-off-by: Kees Cook <keescook@chromium.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2019-01-24 08:19:29 +09:00
* -fplugin-arg-structleak_plugin-byref
* -fplugin-arg-structleak_plugin-byref-all
*
* Usage:
* $ # for 4.5/4.6/C based 4.7
* $ gcc -I`gcc -print-file-name=plugin`/include -I`gcc -print-file-name=plugin`/include/c-family -fPIC -shared -O2 -o structleak_plugin.so structleak_plugin.c
* $ # for C++ based 4.7/4.8+
* $ g++ -I`g++ -print-file-name=plugin`/include -I`g++ -print-file-name=plugin`/include/c-family -fPIC -shared -O2 -o structleak_plugin.so structleak_plugin.c
* $ gcc -fplugin=./structleak_plugin.so test.c -O2
*
* TODO: eliminate redundant initializers
*/
#include "gcc-common.h"
/* unused C type flag in all versions 4.5-6 */
#define TYPE_USERSPACE(TYPE) TYPE_LANG_FLAG_5(TYPE)
__visible int plugin_is_GPL_compatible;
static struct plugin_info structleak_plugin_info = {
gcc-plugins: structleak: Generalize to all variable types This adjusts structleak to also work with non-struct types when they are passed by reference, since those variables may leak just like anything else. This is exposed via an improved set of Kconfig options. (This does mean structleak is slightly misnamed now.) Building with CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL should give the kernel complete initialization coverage of all stack variables passed by reference, including padding (see lib/test_stackinit.c). Using CONFIG_GCC_PLUGIN_STRUCTLEAK_VERBOSE to count added initializations under defconfig: ..._BYREF: 5945 added initializations ..._BYREF_ALL: 16606 added initializations There is virtually no change to text+data size (both have less than 0.05% growth): text data bss dec hex filename 19502103 5051456 1917000 26470559 193e89f vmlinux.stock 19513412 5051456 1908808 26473676 193f4cc vmlinux.byref 19516974 5047360 1900616 26464950 193d2b6 vmlinux.byref_all The measured performance difference is in the noise for hackbench and kernel build benchmarks: Stock: 5x hackbench -g 20 -l 1000 Mean: 10.649s Std Dev: 0.339 5x kernel build (4-way parallel) Mean: 261.98s Std Dev: 1.53 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF: 5x hackbench -g 20 -l 1000 Mean: 10.540s Std Dev: 0.233 5x kernel build (4-way parallel) Mean: 260.52s Std Dev: 1.31 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL: 5x hackbench -g 20 -l 1000 Mean: 10.320 Std Dev: 0.413 5x kernel build (4-way parallel) Mean: 260.10 Std Dev: 0.86 This does not yet solve missing padding initialization for structures on the stack that are never passed by reference (which should be a tiny minority). Hopefully this will be more easily addressed by upstream compiler fixes after clarifying the C11 padding initialization specification. Signed-off-by: Kees Cook <keescook@chromium.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2019-01-24 08:19:29 +09:00
.version = "20190125vanilla",
.help = "disable\tdo not activate plugin\n"
gcc-plugins: structleak: Generalize to all variable types This adjusts structleak to also work with non-struct types when they are passed by reference, since those variables may leak just like anything else. This is exposed via an improved set of Kconfig options. (This does mean structleak is slightly misnamed now.) Building with CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL should give the kernel complete initialization coverage of all stack variables passed by reference, including padding (see lib/test_stackinit.c). Using CONFIG_GCC_PLUGIN_STRUCTLEAK_VERBOSE to count added initializations under defconfig: ..._BYREF: 5945 added initializations ..._BYREF_ALL: 16606 added initializations There is virtually no change to text+data size (both have less than 0.05% growth): text data bss dec hex filename 19502103 5051456 1917000 26470559 193e89f vmlinux.stock 19513412 5051456 1908808 26473676 193f4cc vmlinux.byref 19516974 5047360 1900616 26464950 193d2b6 vmlinux.byref_all The measured performance difference is in the noise for hackbench and kernel build benchmarks: Stock: 5x hackbench -g 20 -l 1000 Mean: 10.649s Std Dev: 0.339 5x kernel build (4-way parallel) Mean: 261.98s Std Dev: 1.53 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF: 5x hackbench -g 20 -l 1000 Mean: 10.540s Std Dev: 0.233 5x kernel build (4-way parallel) Mean: 260.52s Std Dev: 1.31 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL: 5x hackbench -g 20 -l 1000 Mean: 10.320 Std Dev: 0.413 5x kernel build (4-way parallel) Mean: 260.10 Std Dev: 0.86 This does not yet solve missing padding initialization for structures on the stack that are never passed by reference (which should be a tiny minority). Hopefully this will be more easily addressed by upstream compiler fixes after clarifying the C11 padding initialization specification. Signed-off-by: Kees Cook <keescook@chromium.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2019-01-24 08:19:29 +09:00
"byref\tinit structs passed by reference\n"
"byref-all\tinit anything passed by reference\n"
"verbose\tprint all initialized variables\n",
};
gcc-plugins: structleak: Generalize to all variable types This adjusts structleak to also work with non-struct types when they are passed by reference, since those variables may leak just like anything else. This is exposed via an improved set of Kconfig options. (This does mean structleak is slightly misnamed now.) Building with CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL should give the kernel complete initialization coverage of all stack variables passed by reference, including padding (see lib/test_stackinit.c). Using CONFIG_GCC_PLUGIN_STRUCTLEAK_VERBOSE to count added initializations under defconfig: ..._BYREF: 5945 added initializations ..._BYREF_ALL: 16606 added initializations There is virtually no change to text+data size (both have less than 0.05% growth): text data bss dec hex filename 19502103 5051456 1917000 26470559 193e89f vmlinux.stock 19513412 5051456 1908808 26473676 193f4cc vmlinux.byref 19516974 5047360 1900616 26464950 193d2b6 vmlinux.byref_all The measured performance difference is in the noise for hackbench and kernel build benchmarks: Stock: 5x hackbench -g 20 -l 1000 Mean: 10.649s Std Dev: 0.339 5x kernel build (4-way parallel) Mean: 261.98s Std Dev: 1.53 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF: 5x hackbench -g 20 -l 1000 Mean: 10.540s Std Dev: 0.233 5x kernel build (4-way parallel) Mean: 260.52s Std Dev: 1.31 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL: 5x hackbench -g 20 -l 1000 Mean: 10.320 Std Dev: 0.413 5x kernel build (4-way parallel) Mean: 260.10 Std Dev: 0.86 This does not yet solve missing padding initialization for structures on the stack that are never passed by reference (which should be a tiny minority). Hopefully this will be more easily addressed by upstream compiler fixes after clarifying the C11 padding initialization specification. Signed-off-by: Kees Cook <keescook@chromium.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2019-01-24 08:19:29 +09:00
#define BYREF_STRUCT 1
#define BYREF_ALL 2
static bool verbose;
gcc-plugins: structleak: Generalize to all variable types This adjusts structleak to also work with non-struct types when they are passed by reference, since those variables may leak just like anything else. This is exposed via an improved set of Kconfig options. (This does mean structleak is slightly misnamed now.) Building with CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL should give the kernel complete initialization coverage of all stack variables passed by reference, including padding (see lib/test_stackinit.c). Using CONFIG_GCC_PLUGIN_STRUCTLEAK_VERBOSE to count added initializations under defconfig: ..._BYREF: 5945 added initializations ..._BYREF_ALL: 16606 added initializations There is virtually no change to text+data size (both have less than 0.05% growth): text data bss dec hex filename 19502103 5051456 1917000 26470559 193e89f vmlinux.stock 19513412 5051456 1908808 26473676 193f4cc vmlinux.byref 19516974 5047360 1900616 26464950 193d2b6 vmlinux.byref_all The measured performance difference is in the noise for hackbench and kernel build benchmarks: Stock: 5x hackbench -g 20 -l 1000 Mean: 10.649s Std Dev: 0.339 5x kernel build (4-way parallel) Mean: 261.98s Std Dev: 1.53 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF: 5x hackbench -g 20 -l 1000 Mean: 10.540s Std Dev: 0.233 5x kernel build (4-way parallel) Mean: 260.52s Std Dev: 1.31 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL: 5x hackbench -g 20 -l 1000 Mean: 10.320 Std Dev: 0.413 5x kernel build (4-way parallel) Mean: 260.10 Std Dev: 0.86 This does not yet solve missing padding initialization for structures on the stack that are never passed by reference (which should be a tiny minority). Hopefully this will be more easily addressed by upstream compiler fixes after clarifying the C11 padding initialization specification. Signed-off-by: Kees Cook <keescook@chromium.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2019-01-24 08:19:29 +09:00
static int byref;
static tree handle_user_attribute(tree *node, tree name, tree args, int flags, bool *no_add_attrs)
{
*no_add_attrs = true;
/* check for types? for now accept everything linux has to offer */
if (TREE_CODE(*node) != FIELD_DECL)
return NULL_TREE;
*no_add_attrs = false;
return NULL_TREE;
}
static struct attribute_spec user_attr = { };
static void register_attributes(void *event_data, void *data)
{
user_attr.name = "user";
user_attr.handler = handle_user_attribute;
#if BUILDING_GCC_VERSION >= 4007
user_attr.affects_type_identity = true;
#endif
register_attribute(&user_attr);
}
static tree get_field_type(tree field)
{
return strip_array_types(TREE_TYPE(field));
}
static bool is_userspace_type(tree type)
{
tree field;
for (field = TYPE_FIELDS(type); field; field = TREE_CHAIN(field)) {
tree fieldtype = get_field_type(field);
enum tree_code code = TREE_CODE(fieldtype);
if (code == RECORD_TYPE || code == UNION_TYPE)
if (is_userspace_type(fieldtype))
return true;
if (lookup_attribute("user", DECL_ATTRIBUTES(field)))
return true;
}
return false;
}
static void finish_type(void *event_data, void *data)
{
tree type = (tree)event_data;
if (type == NULL_TREE || type == error_mark_node)
return;
#if BUILDING_GCC_VERSION >= 5000
if (TREE_CODE(type) == ENUMERAL_TYPE)
return;
#endif
if (TYPE_USERSPACE(type))
return;
if (is_userspace_type(type))
TYPE_USERSPACE(type) = 1;
}
static void initialize(tree var)
{
basic_block bb;
gimple_stmt_iterator gsi;
tree initializer;
gimple init_stmt;
gcc-plugins: structleak: Generalize to all variable types This adjusts structleak to also work with non-struct types when they are passed by reference, since those variables may leak just like anything else. This is exposed via an improved set of Kconfig options. (This does mean structleak is slightly misnamed now.) Building with CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL should give the kernel complete initialization coverage of all stack variables passed by reference, including padding (see lib/test_stackinit.c). Using CONFIG_GCC_PLUGIN_STRUCTLEAK_VERBOSE to count added initializations under defconfig: ..._BYREF: 5945 added initializations ..._BYREF_ALL: 16606 added initializations There is virtually no change to text+data size (both have less than 0.05% growth): text data bss dec hex filename 19502103 5051456 1917000 26470559 193e89f vmlinux.stock 19513412 5051456 1908808 26473676 193f4cc vmlinux.byref 19516974 5047360 1900616 26464950 193d2b6 vmlinux.byref_all The measured performance difference is in the noise for hackbench and kernel build benchmarks: Stock: 5x hackbench -g 20 -l 1000 Mean: 10.649s Std Dev: 0.339 5x kernel build (4-way parallel) Mean: 261.98s Std Dev: 1.53 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF: 5x hackbench -g 20 -l 1000 Mean: 10.540s Std Dev: 0.233 5x kernel build (4-way parallel) Mean: 260.52s Std Dev: 1.31 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL: 5x hackbench -g 20 -l 1000 Mean: 10.320 Std Dev: 0.413 5x kernel build (4-way parallel) Mean: 260.10 Std Dev: 0.86 This does not yet solve missing padding initialization for structures on the stack that are never passed by reference (which should be a tiny minority). Hopefully this will be more easily addressed by upstream compiler fixes after clarifying the C11 padding initialization specification. Signed-off-by: Kees Cook <keescook@chromium.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2019-01-24 08:19:29 +09:00
tree type;
/* this is the original entry bb before the forced split */
bb = single_succ(ENTRY_BLOCK_PTR_FOR_FN(cfun));
/* first check if variable is already initialized, warn otherwise */
for (gsi = gsi_start_bb(bb); !gsi_end_p(gsi); gsi_next(&gsi)) {
gimple stmt = gsi_stmt(gsi);
tree rhs1;
/* we're looking for an assignment of a single rhs... */
if (!gimple_assign_single_p(stmt))
continue;
rhs1 = gimple_assign_rhs1(stmt);
#if BUILDING_GCC_VERSION >= 4007
/* ... of a non-clobbering expression... */
if (TREE_CLOBBER_P(rhs1))
continue;
#endif
/* ... to our variable... */
if (gimple_get_lhs(stmt) != var)
continue;
/* if it's an initializer then we're good */
if (TREE_CODE(rhs1) == CONSTRUCTOR)
return;
}
/* these aren't the 0days you're looking for */
if (verbose)
inform(DECL_SOURCE_LOCATION(var),
"%s variable will be forcibly initialized",
gcc-plugins: structleak: Generalize to all variable types This adjusts structleak to also work with non-struct types when they are passed by reference, since those variables may leak just like anything else. This is exposed via an improved set of Kconfig options. (This does mean structleak is slightly misnamed now.) Building with CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL should give the kernel complete initialization coverage of all stack variables passed by reference, including padding (see lib/test_stackinit.c). Using CONFIG_GCC_PLUGIN_STRUCTLEAK_VERBOSE to count added initializations under defconfig: ..._BYREF: 5945 added initializations ..._BYREF_ALL: 16606 added initializations There is virtually no change to text+data size (both have less than 0.05% growth): text data bss dec hex filename 19502103 5051456 1917000 26470559 193e89f vmlinux.stock 19513412 5051456 1908808 26473676 193f4cc vmlinux.byref 19516974 5047360 1900616 26464950 193d2b6 vmlinux.byref_all The measured performance difference is in the noise for hackbench and kernel build benchmarks: Stock: 5x hackbench -g 20 -l 1000 Mean: 10.649s Std Dev: 0.339 5x kernel build (4-way parallel) Mean: 261.98s Std Dev: 1.53 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF: 5x hackbench -g 20 -l 1000 Mean: 10.540s Std Dev: 0.233 5x kernel build (4-way parallel) Mean: 260.52s Std Dev: 1.31 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL: 5x hackbench -g 20 -l 1000 Mean: 10.320 Std Dev: 0.413 5x kernel build (4-way parallel) Mean: 260.10 Std Dev: 0.86 This does not yet solve missing padding initialization for structures on the stack that are never passed by reference (which should be a tiny minority). Hopefully this will be more easily addressed by upstream compiler fixes after clarifying the C11 padding initialization specification. Signed-off-by: Kees Cook <keescook@chromium.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2019-01-24 08:19:29 +09:00
(byref && TREE_ADDRESSABLE(var)) ? "byref"
: "userspace");
/* build the initializer expression */
gcc-plugins: structleak: Generalize to all variable types This adjusts structleak to also work with non-struct types when they are passed by reference, since those variables may leak just like anything else. This is exposed via an improved set of Kconfig options. (This does mean structleak is slightly misnamed now.) Building with CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL should give the kernel complete initialization coverage of all stack variables passed by reference, including padding (see lib/test_stackinit.c). Using CONFIG_GCC_PLUGIN_STRUCTLEAK_VERBOSE to count added initializations under defconfig: ..._BYREF: 5945 added initializations ..._BYREF_ALL: 16606 added initializations There is virtually no change to text+data size (both have less than 0.05% growth): text data bss dec hex filename 19502103 5051456 1917000 26470559 193e89f vmlinux.stock 19513412 5051456 1908808 26473676 193f4cc vmlinux.byref 19516974 5047360 1900616 26464950 193d2b6 vmlinux.byref_all The measured performance difference is in the noise for hackbench and kernel build benchmarks: Stock: 5x hackbench -g 20 -l 1000 Mean: 10.649s Std Dev: 0.339 5x kernel build (4-way parallel) Mean: 261.98s Std Dev: 1.53 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF: 5x hackbench -g 20 -l 1000 Mean: 10.540s Std Dev: 0.233 5x kernel build (4-way parallel) Mean: 260.52s Std Dev: 1.31 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL: 5x hackbench -g 20 -l 1000 Mean: 10.320 Std Dev: 0.413 5x kernel build (4-way parallel) Mean: 260.10 Std Dev: 0.86 This does not yet solve missing padding initialization for structures on the stack that are never passed by reference (which should be a tiny minority). Hopefully this will be more easily addressed by upstream compiler fixes after clarifying the C11 padding initialization specification. Signed-off-by: Kees Cook <keescook@chromium.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2019-01-24 08:19:29 +09:00
type = TREE_TYPE(var);
if (AGGREGATE_TYPE_P(type))
initializer = build_constructor(type, NULL);
else
initializer = fold_convert(type, integer_zero_node);
/* build the initializer stmt */
init_stmt = gimple_build_assign(var, initializer);
gsi = gsi_after_labels(single_succ(ENTRY_BLOCK_PTR_FOR_FN(cfun)));
gsi_insert_before(&gsi, init_stmt, GSI_NEW_STMT);
update_stmt(init_stmt);
}
static unsigned int structleak_execute(void)
{
basic_block bb;
unsigned int ret = 0;
tree var;
unsigned int i;
/* split the first bb where we can put the forced initializers */
gcc_assert(single_succ_p(ENTRY_BLOCK_PTR_FOR_FN(cfun)));
bb = single_succ(ENTRY_BLOCK_PTR_FOR_FN(cfun));
if (!single_pred_p(bb)) {
split_edge(single_succ_edge(ENTRY_BLOCK_PTR_FOR_FN(cfun)));
gcc_assert(single_succ_p(ENTRY_BLOCK_PTR_FOR_FN(cfun)));
}
/* enumerate all local variables and forcibly initialize our targets */
FOR_EACH_LOCAL_DECL(cfun, i, var) {
tree type = TREE_TYPE(var);
gcc_assert(DECL_P(var));
if (!auto_var_in_fn_p(var, current_function_decl))
continue;
gcc-plugins: structleak: Generalize to all variable types This adjusts structleak to also work with non-struct types when they are passed by reference, since those variables may leak just like anything else. This is exposed via an improved set of Kconfig options. (This does mean structleak is slightly misnamed now.) Building with CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL should give the kernel complete initialization coverage of all stack variables passed by reference, including padding (see lib/test_stackinit.c). Using CONFIG_GCC_PLUGIN_STRUCTLEAK_VERBOSE to count added initializations under defconfig: ..._BYREF: 5945 added initializations ..._BYREF_ALL: 16606 added initializations There is virtually no change to text+data size (both have less than 0.05% growth): text data bss dec hex filename 19502103 5051456 1917000 26470559 193e89f vmlinux.stock 19513412 5051456 1908808 26473676 193f4cc vmlinux.byref 19516974 5047360 1900616 26464950 193d2b6 vmlinux.byref_all The measured performance difference is in the noise for hackbench and kernel build benchmarks: Stock: 5x hackbench -g 20 -l 1000 Mean: 10.649s Std Dev: 0.339 5x kernel build (4-way parallel) Mean: 261.98s Std Dev: 1.53 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF: 5x hackbench -g 20 -l 1000 Mean: 10.540s Std Dev: 0.233 5x kernel build (4-way parallel) Mean: 260.52s Std Dev: 1.31 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL: 5x hackbench -g 20 -l 1000 Mean: 10.320 Std Dev: 0.413 5x kernel build (4-way parallel) Mean: 260.10 Std Dev: 0.86 This does not yet solve missing padding initialization for structures on the stack that are never passed by reference (which should be a tiny minority). Hopefully this will be more easily addressed by upstream compiler fixes after clarifying the C11 padding initialization specification. Signed-off-by: Kees Cook <keescook@chromium.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2019-01-24 08:19:29 +09:00
/* only care about structure types unless byref-all */
if (byref != BYREF_ALL && TREE_CODE(type) != RECORD_TYPE && TREE_CODE(type) != UNION_TYPE)
continue;
/* if the type is of interest, examine the variable */
if (TYPE_USERSPACE(type) ||
gcc-plugins: structleak: Generalize to all variable types This adjusts structleak to also work with non-struct types when they are passed by reference, since those variables may leak just like anything else. This is exposed via an improved set of Kconfig options. (This does mean structleak is slightly misnamed now.) Building with CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL should give the kernel complete initialization coverage of all stack variables passed by reference, including padding (see lib/test_stackinit.c). Using CONFIG_GCC_PLUGIN_STRUCTLEAK_VERBOSE to count added initializations under defconfig: ..._BYREF: 5945 added initializations ..._BYREF_ALL: 16606 added initializations There is virtually no change to text+data size (both have less than 0.05% growth): text data bss dec hex filename 19502103 5051456 1917000 26470559 193e89f vmlinux.stock 19513412 5051456 1908808 26473676 193f4cc vmlinux.byref 19516974 5047360 1900616 26464950 193d2b6 vmlinux.byref_all The measured performance difference is in the noise for hackbench and kernel build benchmarks: Stock: 5x hackbench -g 20 -l 1000 Mean: 10.649s Std Dev: 0.339 5x kernel build (4-way parallel) Mean: 261.98s Std Dev: 1.53 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF: 5x hackbench -g 20 -l 1000 Mean: 10.540s Std Dev: 0.233 5x kernel build (4-way parallel) Mean: 260.52s Std Dev: 1.31 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL: 5x hackbench -g 20 -l 1000 Mean: 10.320 Std Dev: 0.413 5x kernel build (4-way parallel) Mean: 260.10 Std Dev: 0.86 This does not yet solve missing padding initialization for structures on the stack that are never passed by reference (which should be a tiny minority). Hopefully this will be more easily addressed by upstream compiler fixes after clarifying the C11 padding initialization specification. Signed-off-by: Kees Cook <keescook@chromium.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2019-01-24 08:19:29 +09:00
(byref && TREE_ADDRESSABLE(var)))
initialize(var);
}
return ret;
}
#define PASS_NAME structleak
#define NO_GATE
#define PROPERTIES_REQUIRED PROP_cfg
#define TODO_FLAGS_FINISH TODO_verify_il | TODO_verify_ssa | TODO_verify_stmts | TODO_dump_func | TODO_remove_unused_locals | TODO_update_ssa | TODO_ggc_collect | TODO_verify_flow
#include "gcc-generate-gimple-pass.h"
__visible int plugin_init(struct plugin_name_args *plugin_info, struct plugin_gcc_version *version)
{
int i;
const char * const plugin_name = plugin_info->base_name;
const int argc = plugin_info->argc;
const struct plugin_argument * const argv = plugin_info->argv;
bool enable = true;
PASS_INFO(structleak, "early_optimizations", 1, PASS_POS_INSERT_BEFORE);
if (!plugin_default_version_check(version, &gcc_version)) {
error(G_("incompatible gcc/plugin versions"));
return 1;
}
if (strncmp(lang_hooks.name, "GNU C", 5) && !strncmp(lang_hooks.name, "GNU C+", 6)) {
inform(UNKNOWN_LOCATION, G_("%s supports C only, not %s"), plugin_name, lang_hooks.name);
enable = false;
}
for (i = 0; i < argc; ++i) {
if (!strcmp(argv[i].key, "disable")) {
enable = false;
continue;
}
if (!strcmp(argv[i].key, "verbose")) {
verbose = true;
continue;
}
gcc-plugins: structleak: Generalize to all variable types This adjusts structleak to also work with non-struct types when they are passed by reference, since those variables may leak just like anything else. This is exposed via an improved set of Kconfig options. (This does mean structleak is slightly misnamed now.) Building with CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL should give the kernel complete initialization coverage of all stack variables passed by reference, including padding (see lib/test_stackinit.c). Using CONFIG_GCC_PLUGIN_STRUCTLEAK_VERBOSE to count added initializations under defconfig: ..._BYREF: 5945 added initializations ..._BYREF_ALL: 16606 added initializations There is virtually no change to text+data size (both have less than 0.05% growth): text data bss dec hex filename 19502103 5051456 1917000 26470559 193e89f vmlinux.stock 19513412 5051456 1908808 26473676 193f4cc vmlinux.byref 19516974 5047360 1900616 26464950 193d2b6 vmlinux.byref_all The measured performance difference is in the noise for hackbench and kernel build benchmarks: Stock: 5x hackbench -g 20 -l 1000 Mean: 10.649s Std Dev: 0.339 5x kernel build (4-way parallel) Mean: 261.98s Std Dev: 1.53 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF: 5x hackbench -g 20 -l 1000 Mean: 10.540s Std Dev: 0.233 5x kernel build (4-way parallel) Mean: 260.52s Std Dev: 1.31 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL: 5x hackbench -g 20 -l 1000 Mean: 10.320 Std Dev: 0.413 5x kernel build (4-way parallel) Mean: 260.10 Std Dev: 0.86 This does not yet solve missing padding initialization for structures on the stack that are never passed by reference (which should be a tiny minority). Hopefully this will be more easily addressed by upstream compiler fixes after clarifying the C11 padding initialization specification. Signed-off-by: Kees Cook <keescook@chromium.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2019-01-24 08:19:29 +09:00
if (!strcmp(argv[i].key, "byref")) {
byref = BYREF_STRUCT;
continue;
}
if (!strcmp(argv[i].key, "byref-all")) {
gcc-plugins: structleak: Generalize to all variable types This adjusts structleak to also work with non-struct types when they are passed by reference, since those variables may leak just like anything else. This is exposed via an improved set of Kconfig options. (This does mean structleak is slightly misnamed now.) Building with CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL should give the kernel complete initialization coverage of all stack variables passed by reference, including padding (see lib/test_stackinit.c). Using CONFIG_GCC_PLUGIN_STRUCTLEAK_VERBOSE to count added initializations under defconfig: ..._BYREF: 5945 added initializations ..._BYREF_ALL: 16606 added initializations There is virtually no change to text+data size (both have less than 0.05% growth): text data bss dec hex filename 19502103 5051456 1917000 26470559 193e89f vmlinux.stock 19513412 5051456 1908808 26473676 193f4cc vmlinux.byref 19516974 5047360 1900616 26464950 193d2b6 vmlinux.byref_all The measured performance difference is in the noise for hackbench and kernel build benchmarks: Stock: 5x hackbench -g 20 -l 1000 Mean: 10.649s Std Dev: 0.339 5x kernel build (4-way parallel) Mean: 261.98s Std Dev: 1.53 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF: 5x hackbench -g 20 -l 1000 Mean: 10.540s Std Dev: 0.233 5x kernel build (4-way parallel) Mean: 260.52s Std Dev: 1.31 CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL: 5x hackbench -g 20 -l 1000 Mean: 10.320 Std Dev: 0.413 5x kernel build (4-way parallel) Mean: 260.10 Std Dev: 0.86 This does not yet solve missing padding initialization for structures on the stack that are never passed by reference (which should be a tiny minority). Hopefully this will be more easily addressed by upstream compiler fixes after clarifying the C11 padding initialization specification. Signed-off-by: Kees Cook <keescook@chromium.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2019-01-24 08:19:29 +09:00
byref = BYREF_ALL;
continue;
}
error(G_("unknown option '-fplugin-arg-%s-%s'"), plugin_name, argv[i].key);
}
register_callback(plugin_name, PLUGIN_INFO, NULL, &structleak_plugin_info);
if (enable) {
register_callback(plugin_name, PLUGIN_PASS_MANAGER_SETUP, NULL, &structleak_pass_info);
register_callback(plugin_name, PLUGIN_FINISH_TYPE, finish_type, NULL);
}
register_callback(plugin_name, PLUGIN_ATTRIBUTES, register_attributes, NULL);
return 0;
}