src/secure/lib/libcrypt/crypt-blowfish.c, branch releng/12.2

Make libcrypt thread-safe. Add crypt_r(3).

2016-08-10T15:16:28Z

glibc has a pretty nice function called crypt_r(3), which is nothing more than crypt(3), but thread-safe. It accomplishes this by introducing a 'struct crypt_data' structure that contains a buffer that is large enough to hold the resulting string. Let's go ahead and also add this function. It would be a shame if a useful function like this wouldn't be usable in multithreaded apps. Refactor crypt.c and all of the backends to no longer declare static arrays, but write their output in a provided buffer. There is no need to do any buffer length computation here, as we'll just need to ensure that 'struct crypt_data' is large enough, which it is. _PASSWORD_LEN is defined to 128 bytes, but in this case I'm picking 256, as this is going to be part of the actual ABI. Differential Revision: https://reviews.freebsd.org/D7306

Add compatibility with $2y$ bcrypt hashes

2015-06-16T23:57:29Z

crypt_blowfish and many implementations based on it (Apache, PHP, PostgreSQL) implemented $2y$ before OpenBSD went with $2b$. This changes marks them as equivalent. http://www.openwall.com/lists/announce/2011/07/17/1 This change is required for applications that use the base crypt() implementation (including nginx) to be able to validate $2y$ hashes Reviewed by: eadler Approved by: delphij MFC after: 1 week Relnotes: yes Sponsored by: ScaleEngine Inc. Differential Revision: https://reviews.freebsd.org/D2742

Switch using the new $2b$ format by default, when bcrypt is used.

2014-05-14T00:50:31Z

MFC after: 2 weeks Relnotes: default Blowfish crypt(3) format have been changed to $2b$.

Refresh our implementation of OpenBSD's Blowfish password format.

2014-02-25T23:03:48Z

Notable changes: - Support of $2b$ password format to address a problem where very long passwords (more than 256 characters, when an integer overflow would happen and cause the length to wrap at 256). - Updated pseudo code in comments to reflect the reality. - Removed our local shortcut of processing magic string and rely on the centralized and tigntened validation. - Diff reduction from upstream. For now we are still generating the older $02a$ format of password but we will migrate to the new format once the format is formally finalized. MFC after: 1 month

Return NULL on error rather than ":", per the crypt(3) man page.

2012-02-22T01:23:14Z

Discussed in: http://www.openwall.com/lists/oss-security/2011/11/15/3

Strip the private blowfish code down to only that which is

2003-06-02T19:17:24Z

required to make crypt(3) blowfish "$2a$..." hashes. Lint and warnsify.

No functional change, but big code cleanup. WARNS, lint(1) and style(9).

2002-03-06T17:18:09Z

__FBSDID() (second half of src/lib/libcrypt changes)

2001-10-23T10:23:32Z

Add OpenBSD-style blowfish password hashing. This makes one less

2001-03-11T16:05:43Z

gratuitous difference between us and our sister project. This was given to me _ages_ ago. May apologies to Paul for the length of time its taken me to commit. Obtained from: Niels Provos /OpenBSD Submitted by: Paul Herman