Conceptually, this is a squashed version of:
Revert "Address feedback"
This reverts commit 75551e07bd.
and
Revert "Add CRYPTO_thread_glock_new"
This reverts commit ed6b2c7938.
But there were some intervening commits that made neither revert apply
cleanly, so instead do it all as one shot.
The crypto global locks were an attempt to cope with the awkward
POSIX semantics for pthread_atfork(); its documentation (the "RATIONALE"
section) indicates that the expected usage is to have the prefork handler
lock all "global" locks, and the parent and child handlers release those
locks, to ensure that forking happens with a consistent (lock) state.
However, the set of functions available in the child process is limited
to async-signal-safe functions, and pthread_mutex_unlock() is not on
the list of async-signal-safe functions! The only synchronization
primitives that are async-signal-safe are the semaphore primitives,
which are not really appropriate for general-purpose usage.
However, the state consistency problem that the global locks were
attempting to solve is not actually a serious problem, particularly for
OpenSSL. That is, we can consider four cases of forking application
that might use OpenSSL:
(1) Single-threaded, does not call into OpenSSL in the child (e.g.,
the child calls exec() immediately)
For this class of process, no locking is needed at all, since there is
only ever a single thread of execution and the only reentrancy is due to
signal handlers (which are themselves limited to async-signal-safe
operation and should not be doing much work at all).
(2) Single-threaded, calls into OpenSSL after fork()
The application must ensure that it does not fork() with an unexpected
lock held (that is, one that would get unlocked in the parent but
accidentally remain locked in the child and cause deadlock). Since
OpenSSL does not expose any of its internal locks to the application
and the application is single-threaded, the OpenSSL internal locks
will be unlocked for the fork(), and the state will be consistent.
(OpenSSL will need to reseed its PRNG in the child, but that is
an orthogonal issue.) If the application makes use of locks from
libcrypto, proper handling for those locks is the responsibility of
the application, as for any other locking primitive that is available
for application programming.
(3) Multi-threaded, does not call into OpenSSL after fork()
As for (1), the OpenSSL state is only relevant in the parent, so
no particular fork()-related handling is needed. The internal locks
are relevant, but there is no interaction with the child to consider.
(4) Multi-threaded, calls into OpenSSL after fork()
This is the case where the pthread_atfork() hooks to ensure that all
global locks are in a known state across fork() would come into play,
per the above discussion. However, these "calls into OpenSSL after
fork()" are still subject to the restriction to async-signal-safe
functions. Since OpenSSL uses all sorts of locking and libc functions
that are not on the list of safe functions (e.g., malloc()), this
case is not currently usable and is unlikely to ever be usable,
independently of the locking situation. So, there is no need to
go through contortions to attempt to support this case in the one small
area of locking interaction with fork().
In light of the above analysis (thanks @davidben and @achernya), go
back to the simpler implementation that does not need to distinguish
"library-global" locks or to have complicated atfork handling for locks.
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
(Merged from https://github.com/openssl/openssl/pull/5089)
cryptilib.h is the second.
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4188)
Remove some incorrect copyright references.
Move copyright to standard place
Add OpenSSL copyright where missing.
Remove copyrighted file that we don't use any more
Remove Itanium assembler for RC4 and MD5 (assembler versions of old and
weak algorithms for an old chip)
Standardize apps/rehash copyright comment; approved by Timo
Put dual-copyright notice on mkcert
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3691)
Remove assertion when mmap() fails.
Only run the 1<<31 limit test on Linux
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3455)
Issue 1:
sh.bittable_size is a size_t but i is and int, which can result in
freelist == -1 if sh.bittable_size exceeds an int.
This seems to result in an OPENSSL_assert due to invalid allocation
size, so maybe that is "ok."
Worse, if sh.bittable_size is exactly 1<<31, then this becomes an
infinite loop (because 1<<31 is a negative int, so it can be shifted
right forever and sticks at -1).
Issue 2:
CRYPTO_secure_malloc_init() sets secure_mem_initialized=1 even when
sh_init() returns 0.
If sh_init() fails, we end up with secure_mem_initialized=1 but
sh.minsize=0. If you then call secure_malloc(), which then calls,
sh_malloc(), this then enters an infite loop since 0 << anything will
never be larger than size.
Issue 3:
That same sh_malloc loop will loop forever for a size greater
than size_t/2 because i will proceed (assuming sh.minsize=16):
i=16, 32, 64, ..., size_t/8, size_t/4, size_t/2, 0, 0, 0, 0, ....
This sequence will never be larger than "size".
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3449)
The sh_add_to_list function will overwrite subsequent slots in the free list
for small allocations. This causes a segmentation fault if the writes goes
off the end of the secure memory. I've not investigated if this problem
can overwrite memory without the segmentation fault, but it seems likely.
This fix limits the minsize to the sizeof of the SH_LIST structure (which
also has a side effect of properly aligning the pointers).
The alternative would be to return an error if minsize is too small.
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2657)
Fix some of the variables to be (s)size_t, so that more than 1GB of
secure memory can be allocated. The arena has to be a power of 2, and
2GB fails because it ends up being a negative 32-bit signed number.
The |too_late| flag is not strictly necessary; it is easy to figure
out if something is secure memory by looking at the arena. As before,
secure memory allocations will not fail, but now they can be freed
correctly. Once initialized, secure memory can still be used, even if
allocations occured before initialization.
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Commit 05c7b1631 ("Implement the use of heap manipulator implementions")
added 'file' and 'line' arguments to CRYPTO_free() and friends, but neglected
to fix up the !IMPLEMENTED case within CRYPTO_secure_free(). Add the missing
arguments there too.
Signed-off-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
- Make use of the functions given through CRYPTO_set_mem_functions().
- CRYPTO_free(), CRYPTO_clear_free() and CRYPTO_secure_free() now receive
__FILE__ and __LINE__.
- The API for CRYPTO_set_mem_functions() and CRYPTO_get_mem_functions()
is slightly changed, the implementation for free() now takes a couple
of extra arguments, taking __FILE__ and __LINE__.
- The CRYPTO_ memory functions will *always* receive __FILE__ and __LINE__
from the corresponding OPENSSL_ macros, regardless of if crypto-mdebug
has been enabled or not. The reason is that if someone swaps out the
malloc(), realloc() and free() implementations, we can't know if they
will use them or not.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Only two macros CRYPTO_MDEBUG and CRYPTO_MDEBUG_ABORT to control this.
If CRYPTO_MDEBUG is not set, #ifdef out the whole debug machinery.
(Thanks to Jakob Bohm for the suggestion!)
Make the "change wrapper functions" be the only paradigm.
Wrote documentation!
Format the 'set func' functions so their paramlists are legible.
Format some multi-line comments.
Remove ability to get/set the "memory debug" functions at runtme.
Remove MemCheck_* and CRYPTO_malloc_debug_init macros.
Add CRYPTO_mem_debug(int flag) function.
Add test/memleaktest.
Rename CRYPTO_malloc_init to OPENSSL_malloc_init; remove needless calls.
Reviewed-by: Richard Levitte <levitte@openssl.org>
