Revert "Support -D_FORTIFY_SOURCE=3 by using __builtin_dynamic_object_size."

This reverts commit 0bd292567a.

It isn't guaranteed anywhere that __builtin_dynamic_object_size can
always deduce the size of every object passed to it so systemd
can end up using either malloc_usable_size or
__builtin_dynamic_object_size when pointers are passed around,
which in turn can lead to actual segfaults like the one mentioned in
https://github.com/systemd/systemd/issues/23619.

Apparently __builtin_object_size can return different results for
pointers referring to the same memory as well but somehow it hasn't
caused any issues yet. Looks like this whole
malloc_usable_size/FORTIFY_SOURCE stuff should be revisited.

Closes https://github.com/systemd/systemd/issues/23619 and
https://github.com/systemd/systemd/issues/23150.

Reopens https://github.com/systemd/systemd/issues/22801
This commit is contained in:
Evgeny Vereshchagin
2022-06-05 19:13:17 +00:00
parent 36cb69fc43
commit 2cfb790391
+3 -13
View File
@@ -174,23 +174,13 @@ void* greedy_realloc0(void **p, size_t need, size_t size);
* is compatible with _FORTIFY_SOURCES. If _FORTIFY_SOURCES is used many memory operations will take the
* object size as returned by __builtin_object_size() into account. Hence, let's return the smaller size of
* malloc_usable_size() and __builtin_object_size() here, so that we definitely operate in safe territory by
* both the compiler's and libc's standards. Note that _FORTIFY_SOURCES=3 handles also dynamically allocated
* objects and thus it's safer using __builtin_dynamic_object_size if _FORTIFY_SOURCES=3 is used (#22801).
* Moreover, when NULL is passed malloc_usable_size() is documented to return zero, and
* both the compiler's and libc's standards. Note that __builtin_object_size() evaluates to SIZE_MAX if the
* size cannot be determined, hence the MIN() expression should be safe with dynamically sized memory,
* too. Moreover, when NULL is passed malloc_usable_size() is documented to return zero, and
* __builtin_object_size() returns SIZE_MAX too, hence we also return a sensible value of 0 in this corner
* case. */
#if defined __has_builtin
# if __has_builtin(__builtin_dynamic_object_size)
# define MALLOC_SIZEOF_SAFE(x) \
MIN(malloc_usable_size(x), __builtin_dynamic_object_size(x, 0))
# endif
#endif
#ifndef MALLOC_SIZEOF_SAFE
#define MALLOC_SIZEOF_SAFE(x) \
MIN(malloc_usable_size(x), __builtin_object_size(x, 0))
#endif
/* Inspired by ELEMENTSOF() but operates on malloc()'ed memory areas: typesafely returns the number of items
* that fit into the specified memory block */