diff options
| author | Martin Thomson <mt@lowentropy.net> | 2019-06-25 12:25:40 +1000 |
|---|---|---|
| committer | Martin Thomson <mt@lowentropy.net> | 2019-06-25 12:25:40 +1000 |
| commit | af023085c60058746b5f15e7042bba598f318197 (patch) | |
| tree | b2e916b3fbd95df0f9e0174677c055860c26035d /lib/ssl | |
| parent | 20ded1736ed1abf64ca66de0c70c7c4385549c2a (diff) | |
| download | nss-hg-af023085c60058746b5f15e7042bba598f318197.tar.gz | |
Bug 1556591 - Eliminate races in uses of PK11_SetWrapKey, r=kjacobs
These functions are terrible, but in the spirit of ABI compatibility, I
can't really change their signature. And it isn't important enough to fix them,
so this introduces a quick hack.
This is wrapped up in the mess that is Bug 1556588, but I'm proceeding here
under the assumption that unconditional use of the lock is safe and that a
non-thread-safe module will explode in many other ways.
Differential Revision: https://phabricator.services.mozilla.com/D34650
Diffstat (limited to 'lib/ssl')
| -rw-r--r-- | lib/ssl/ssl3con.c | 12 |
1 files changed, 12 insertions, 0 deletions
diff --git a/lib/ssl/ssl3con.c b/lib/ssl/ssl3con.c index d9f4b5234..67f162955 100644 --- a/lib/ssl/ssl3con.c +++ b/lib/ssl/ssl3con.c @@ -11378,7 +11378,19 @@ ssl3_CacheWrappedSecret(sslSocket *ss, sslSessionID *sid, wrappingKey = PK11_KeyGen(symKeySlot, mechanism, NULL, keyLength, pwArg); if (wrappingKey) { + /* The thread safety characteristics of PK11_[SG]etWrapKey is + * abominable. This protects against races in calling + * PK11_SetWrapKey by dropping and re-acquiring the canonical + * value once it is set. The mutex in PK11_[SG]etWrapKey will + * ensure that races produce the same value in the end. */ PK11_SetWrapKey(symKeySlot, wrapKeyIndex, wrappingKey); + PK11_FreeSymKey(wrappingKey); + wrappingKey = PK11_GetWrapKey(symKeySlot, wrapKeyIndex, + CKM_INVALID_MECHANISM, incarnation, pwArg); + if (!wrappingKey) { + PK11_FreeSlot(symKeySlot); + return SECFailure; + } } } } else { |