Age | Commit message (Collapse) | Author | |
---|---|---|---|
2002-12-17 | Make SysV-style shared memory and semaphore limits sysctl'able. | Todd C. Miller | |
Instead of allocating a static amount of memory for the data structures via valloc() in allocsys(), allocate things dynamically using pool(9) when possible and malloc(9) when not. The various members of struct seminfo and struct shminfo are in kern.seminfo and kern.shminfo respectively (not all members of kern.seminfo are changable). The data structures used still leave something to be desired but things are not made worse in that respect by this commit. | |||
2002-12-15 | Casting to unsigned int in isfoo() causes problems on alpha and | Todd C. Miller | |
sparc64. Change cast back to unsigned char but do a bitwise AND with 0xff to avoid any sign extension weirdness and to make it impossible for us to overflow _C_ctype_. The bitwise AND is probably not needed and may be removed later if this does not trigger compiler bugs. | |||
2002-12-15 | more writeable -> writable by torh | Henning Brauer | |
2002-12-14 | Instead of doing "if (c == EOF) return 0;" use the hook ('?') operator | Todd C. Miller | |
instead to make these one-liners. Works around a compiler bug on vax that affects both the libc and inline versions identically. | |||
2002-12-13 | Less stupid check for 7-bit ascii in toupper/tolower | Todd C. Miller | |
Remove useless check for EOF in isascii | |||
2002-12-13 | Cast to unsigned int, not unsigned char, since what is being cast is an | Todd C. Miller | |
int, not a char. | |||
2002-12-11 | Convert ctype.h macros into inline functions. This fixes the issues we | Todd C. Miller | |
currently have with the macro versions and makes the ctype.h versions 100% identical to what is in libc. Discussed with pjanzen@ and OK'd by deraadt@. | |||
2002-11-25 | Mention the net.inet.tcp.ecn sysctl. | Wilbern Cobb | |
ok kjc@ | |||
2002-11-24 | Fix comment: s/trailing NULL/trailing NUL/, no code changes. | Chad Loder | |
2002-11-24 | Fix instance where syslog(3) after closelog(3) could access freed log | Chad Loder | |
tag name. OK millert@, deraadt@ | |||
2002-11-22 | Initialize status to zero before calling waitpid(). That way, if | Todd C. Miller | |
there is nothing to be waited for (for instance if SIGCHLD is being ignored) we don't check WIFEXITED(garbage off the stack). | |||
2002-11-22 | more snprintf; ok millert | Theo de Raadt | |
2002-11-21 | shadow -> _shadow | Todd C. Miller | |
2002-11-21 | Add a "shadow" group and make the shadow passwd db readable by that | Todd C. Miller | |
group. This changes getpw* to always try the shadow db first and then fall back to the db w/o password hashes. In the future, /usr/libexec/auth/login_passwd (and others) will be setgid shadow instead of setuid root. OK deraadt@ If you track -current you should do the following: o add group shadow to /etc/group o chgrp shadow /etc/spwd.db o chmod 640 /etc/spwd.db o rebuild and install src/usr.sbin/pwd_mkdb You do not need to rebuild libc yet, but it would't hurt to do so. | |||
2002-10-30 | Revert this va_copy addition, it cannot work on macppc (which is the | Dale Rahn | |
only arch where va_copy is currently required), current version of va_copy uses alloca, but since the results of the va_copy are not used here, only after return, stack trashing will occur. This will have to be revised again for gcc 3.2 support on powerpc. | |||
2002-10-24 | Use va_copy() for varargs assignemnt. va_copy() is standard with C99 | Dale Rahn | |
and gcc3.2 requires this on powerpc. | |||
2002-10-16 | Replace "if (value != NULL && value == as->FOO)" with just | Todd C. Miller | |
"if (value == as->FOO)" -- NULL needs no special handling. Pointed out by Paul Borman. | |||
2002-10-16 | Implement isfdtype(3) as per the POSIX.1g draft; requested by David Hill | Todd C. Miller | |
2002-10-15 | Fix typo introduce in last commit | Todd C. Miller | |
2002-10-15 | cast NULL varargs sentinel to char * so it is 64 bit on alpha & sparc64 | Todd C. Miller | |
2002-10-15 | Document the fact that calling auth_setitem() with a NULL value | Todd C. Miller | |
clears (frees) the corresponding value in the auth session. | |||
2002-10-15 | If auth_setitem() is called with the current value (ie: the pointer | Todd C. Miller | |
is the same as the private value) then just return 0 as there is nothing to do. This fixes a potentially nasty problem where the caller could grab the username or style from the auth session via auth_getitem() and then call auth_verify() with those values. auth_setitem() would eventually get called which would make a private copy and free the old values in the auth session. After all this, the stashed username and/or style pointers would point to freed memory. | |||
2002-10-15 | Note that when calling auth_verify() with a style and/or username, | Todd C. Miller | |
the old values in the auth session will get freed and the new values added. | |||
2002-10-15 | Cast NULL sentinel in auth_call() to a pointer | Todd C. Miller | |
2002-10-13 | Move the clearing of as->state in auth_challenge() to after the sanity | Todd C. Miller | |
checks. The check for "as != NULL" is useless if you have already dereferenced "as". Also fix some comment typos. From Moritz Jodeit. | |||
2002-10-09 | Use strtoll() and strtoull() instead of strtoq() and strtouq() now | Todd C. Miller | |
that we have them as per the original BSDi code. | |||
2002-08-04 | If fork() fails, log/report the error *before* we close our fd's just | Todd C. Miller | |
in case close(2) were to set errno. | |||
2002-07-30 | malloc paranoia; it is unlikely that any filesystem will support | Todd C. Miller | |
enough directory entries to cause a problem but it is good form anyway. deraadt@ OK. | |||
2002-07-30 | be even more careful with strlcpy() | Theo de Raadt | |
2002-07-18 | typo | Todd C. Miller | |
2002-07-14 | indent, and double free fix; millert ok | Theo de Raadt | |
2002-07-12 | Add a missing check for fts_alloc() returning NULL; Chad Loder | Todd C. Miller | |
2002-07-08 | Use qsort() instead of mergesort() since the latter can fail due | Todd C. Miller | |
to malloc(). opendir() requires a stable sort so we rig the compare routine to never return 0. From Lars J. Buitinck | |||
2002-07-06 | Remove kernel support for NTP. ok deraadt@ and tholo@ | Thomas Nordin | |
2002-07-06 | KNF | Theo de Raadt | |
2002-07-06 | KNF | Theo de Raadt | |
2002-07-06 | cope better if realloc fails; better fix than cloder@acm.org | Theo de Raadt | |
2002-07-05 | more refinement | Theo de Raadt | |
2002-07-02 | missing priority argument in example; ok millert@ | Kevin Steves | |
2002-07-01 | need string.h | Theo de Raadt | |
2002-07-01 | Disc -> Disk for EDQUOT; now the error and log messages that appear when | Paul Janzen | |
quotas are exceeded are consistent. deraadt@ ok | |||
2002-07-01 | Fix strnvis(): don't truncate unnecessarily, set the return value correctly | Paul Janzen | |
if we had to truncate, and don't NUL-terminate if size == 0. | |||
2002-07-01 | indent | Theo de Raadt | |
2002-06-28 | Add RPP_STDIN flag which acts as the converse of RPP_REQUIRE_TTY. | Todd C. Miller | |
Based on a patch from Brett Eldridge. | |||
2002-06-28 | Lies, damn lies and getcap(3). The description for cgetclose() was | Todd C. Miller | |
completely wrong and has apparently caused people to believe that cgetclose() frees memory allocated by cgetent() which is *not* the case. | |||
2002-06-27 | %u | Theo de Raadt | |
2002-06-27 | KNF | Theo de Raadt | |
2002-06-27 | %u and uid/gid | Theo de Raadt | |
2002-06-23 | uid_t is unsigned | Theo de Raadt | |
2002-06-22 | use strtok_r() instead of strtok(); millert ok | Theo de Raadt | |