summaryrefslogtreecommitdiff
path: root/gnu
AgeCommit message (Collapse)Author
2001-08-27allow for local tags that include numbers (like $XFree86$). FirstMatthieu Herrb
character must be pure alpha though. Ok deraadt@, tholo@.
2001-08-27fix description so it is aligned with other entries surrounding itBrad Smith
-- Ok'd by millert@
2001-08-21Update to sendmail 8.11.6, from the release notes:Todd C. Miller
SECURITY: Fix a possible memory access violation when specifying out-of-bounds debug parameters. Problem detected by Cade Cairns of SecurityFocus. Avoid leaking recipient information in unrelated DSNs. This could happen if a connection is aborted, several mails had been scheduled for delivery via that connection, and the timeout is reached such that several DSNs are sent next. Problem noted by Dileepan Moorkanat of Hewlett-Packard. Fix a possible segmentation violation when specifying too many wildcard operators in a rule. Problem detected by Werner Wiethege. Avoid a segmentation fault on non-matching Hesiod lookups. Problem noted by Russell McOrmond of flora.ca
2001-08-18Document rules specific to the FSF subtree.Marc Espie
Basically, not following these rules might mean deep shit later on.
2001-08-12Use INFTIM where possible.Federico G. Schwindt
2001-08-08Prevent logging strerror(0) on EOF; patch sent to sendmail-bugs@sendmail.ORGTodd C. Miller
2001-08-07Use mktemp(1) to avoid /tmp race found by terra@diku.dk.Todd C. Miller
Fix based on patch from mbing@nfr.net.
2001-08-07tree must buildHugh Graham
2001-08-01Add `newer' gas balign/p2align/2bytes/4bytes.Marc Espie
Checked by me on i386, miod@ on m68k (LOTS of thanks). fgs@ was last seen trying it on sparc a long time ago. I know that sparcs are slow, but this is ridiculous. Nobody else moved, so 8-P
2001-08-01Do not define 'unix' on powerpc. ok drahnPeter Valchev
2001-08-01update to sendmail 8.11.5Todd C. Miller
2001-07-28Revert accidentally committed changeThorsten Lockert
2001-07-27Use snprintf to avoid potential overflows; zen-parse@gmx.netTodd C. Miller
2001-07-26add generics_entire_domain and GENERICS_DOMAINTodd C. Miller
2001-07-25Fix compilation on mvme88k (until binutils are ready).Miod Vallat
2001-07-25remove local hackTodd C. Miller
2001-07-25Sync with reality:Todd C. Miller
o Use a genericstable instead of the local mailnametable hack o Add IPv6 support.
2001-07-21SHLIB_MAJOR/_MINOR not set in bsd.lib.mk anymore. Use major/minor.Kenneth R Westerback
2001-07-19__attribute__((__nonnull__), separate from printf.Marc Espie
2001-07-19even worse: link against nearly any kerberos library, and you need all ofTheo de Raadt
them. these swedes are insane!
2001-07-19well, with the most recent reorganization, it now turns out that anyTheo de Raadt
program which links -lkrb5 also must link -lkafs as well! someone please mail hin@ an alpha or a powerpc, so that he can become aware of how binutils ("the other linker we use") works.
2001-07-11remove mips leftovers; ok bradPeter Valchev
2001-07-09correct type on last arg to execl(); nordin@cse.ogi.eduTheo de Raadt
2001-07-08Fix an obscure bug: make sure aliases are referenced.Marc Espie
Otherwise, if a shared library uses an aliased symbol in another library, the alias will only be partially resolved (everything setup... except for the real symbol marked as referenced, so the relocation ends up being zero and causes core-dumps on runtimes). Caused quite a few crashes: Xfree's 4 GL (linked with libc_r) or kerberosV exhibited the problem...
2001-07-07swat size_t vs. int bug (benign)Todd C. Miller
2001-07-07Same fix as I did for bcTodd C. Miller
2001-07-07Silence 2 warnings. GNU code acts as if any non-gnu libc is stupid,Todd C. Miller
ours is not...
2001-07-02fix some incorrect cat page dependenciesTodd C. Miller
2001-07-01when you support kerberosV, you apparently need to also link in the 12Theo de Raadt
other libraries that it's tentacles lead to. This is just disgusting, and are going to pressure the KTH people to fix this. It's unmaintainable. kerberosV and kerberosIV even have cross dependencies. It's insane. patch developed at the same time by drahn and I
2001-06-27More mips leftovers to get rid of...Miod Vallat
2001-06-23print names in reference (.%A) as belowJun-ichiro itojun Hagino
first guy, ... first guy and second guy, ... first, second, and second guy, ... based on comment from "Brian J. Kifiak" <bk@rt.fm>.
2001-06-13alpha now defaults to shared libs.Artur Grabowski
espie@ ok.
2001-06-09Revert to older demangle. Not tested nor updated yet.Marc Espie
2001-06-09bye byeMarc Espie
2001-06-09bye, byeMarc Espie
2001-06-09Bye, byeMarc Espie
2001-06-09Automatic cvs merge. Dread what I will have to fix after this excuseMarc Espie
of a program is done...
2001-06-09Long delayed import of binutils-2.10.1. Turns out art@ needs some alphaMarc Espie
bug-fixes...
2001-06-09libiberty version bundled with binutils 2.11.Marc Espie
2001-06-09Synch with binutils 2.10.1, add dyn-string to libiberty.Marc Espie
2001-06-09Fill some field earlier, should allow better .init/dlopen interaction.Marc Espie
From NetBSD. fgsch@ agrees.
2001-06-07Restore debugging to att format.Marc Espie
Noticed and requested by aazubel@mail.ru and fgs@.
2001-06-02Remove a few extraneous subshells. Okay millert@Marc Espie
2001-05-30OSTYPE is openbsd, not unknownTodd C. Miller
2001-05-29sendmail 8.12.0 with $Id tags converted to $SendmailTodd C. Miller
2001-05-29Update to sendmail 8.11.4:Todd C. Miller
8.11.4 revamps signal handling within the MTA in order to reduce the likelihood of a race condition that can lead to heap corruption as described in Michal Zalewski's advisory. The problems discussed in the advisory are not currently known to be exploitable but we recommend upgrading to 8.11.4 in case a method is found to exploit the signal handling race condition. 8.11.4 also fixes other bugs found since the release of 8.11.3. See the RELEASE_NOTES file for more details.
2001-05-25Support KERBEROS5 / gssapi. (hin@ ok)Hakan Olsson
2001-05-24merge in perl 5.6.1 with our local changesTodd C. Miller
2001-05-24stock perl 5.6.1Todd C. Miller
2001-05-08Insure detection of prefixes is limited to whole words. This avoidsMarc Espie
recognizing /usr/libdata as /usr/lib. Ok millert@