openbsd, branch OPENBSD_4_7_BASE A mirror of https://github.com/libressl/openbsd.git https://git.lua4.win/openbsd/atom?h=OPENBSD_4_7_BASE 2010-03-10T20:46:17+00:00 This commit was manufactured by cvs2git to create tag 'OPENBSD_4_7_BASE'. 2010-03-10T20:46:17+00:00 cvs2svn admin@example.com 2010-03-10T20:46:17+00:00 urn:sha1:b11b85914ed7b8e8f47cb4d0af86678bc41d4354 It's unsetenv() that doesn't like `=' in the argument, not putenv(). 2010-03-10T20:46:16+00:00 kili 2010-03-10T20:46:16+00:00 urn:sha1:a7a853e149f5a2468fd738638c4641dd59d6bc31 ok millert@ cherrypick patch from OpenSSL 0.9.8m: 2010-03-04T11:02:42+00:00 djm 2010-03-04T11:02:42+00:00 urn:sha1:9f75721dec306b453d31851a26bad908e9605c4f *) Always check bn_wexpend() return values for failure. (CVE-2009-3245) [Martin Olsson, Neel Mehta] Fix the atelnet() function, which was wrong in several ways. 2010-02-27T00:58:56+00:00 nicm 2010-02-27T00:58:56+00:00 urn:sha1:95900d7e9c7c406efdaf6098a2bf4193d1722156 Pointed out by obsd at happyjack.org, fix based on a diff from kili@. ok deraadt Fix the worst among the fatal block-nesting issues we have in our tree, 2010-02-23T23:00:52+00:00 schwarze 2010-02-23T23:00:52+00:00 urn:sha1:e1078872617912e5553bcd54683368efa7d54328 in the end simply changing Oo Xo Oo Oc Oc Xc to Oo Xo Oo Oc Xc Oc. I tried five different variants without the Xo/Xc, because there is no real reason why Xo/Xc should be needed here, and two additional variants with Xo/Xc that look more pretty than what i'm now committing. All seven alternatives work both with mandoc and groff 1.20.1, but all seven trigger various different bugs in our old groff, and we still need to remain compatible with our old groff right now. ok jmc@ sobrado@ Our groff does not handle all cases of nested displays properly, so 2010-02-18T13:39:21+00:00 schwarze 2010-02-18T13:39:21+00:00 urn:sha1:00c63ab59645edbe96c6197c108e64832eacbc71 resolve the contradiction in the manual whether it's allowed or not in favour of "not allowed" for now, shelving my diff to support nested displays in mandoc(1) for later consideration. Found by and unbreaking the build with mandoc(1). Feedback and OK jmc@ Sync with current state of the tree. 2010-02-14T12:30:29+00:00 miod 2010-02-14T12:30:29+00:00 urn:sha1:e203152f87011065f17482bff74724c310cfed87 Test mkstemps() too 2010-02-11T16:39:05+00:00 guenther 2010-02-11T16:39:05+00:00 urn:sha1:1d0b350817518f34a39a1200d3431eb899278839 Add a regression suite for mkstemp(), verifying that it neither overruns 2010-02-11T07:35:38+00:00 guenther 2010-02-11T07:35:38+00:00 urn:sha1:387b67884b39e3f399ad34aa12a4e897ce8c1c71 no underruns the supplied template buffer, that it can generate names that don't contain any X's, and that all open() calls that pass O_CREAT also pass O_EXCL Use size_t in appropriate places; fixes sorting of big arrays; 2010-02-08T11:04:07+00:00 otto 2010-02-08T11:04:07+00:00 urn:sha1:f6129285462a0eff1c43458954cfc5ac8b927444 after the diff was written, I made it similar to the freebsd fix of the same code; pr6287 ok millert@ guenther@