summaryrefslogtreecommitdiff
path: root/usr.sbin
AgeCommit message (Collapse)Author
2014-01-13fix another chmod on permanent tempfile to respect umask.Marc Espie
2014-01-13have Ustar extract owners for symlinks, which isn't really important forMarc Espie
pkg_add, since it skips links and restores owners from the plist anyways.
2014-01-13say something if we sign an empty repositoryMarc Espie
tweak mode on tmp signed file
2014-01-13new signify options. from and ok espieTed Unangst
2014-01-12revert back to 1.97Peter Hessler
There is a memory leak when using internal GZip, so switch back to the external gzip for now. OK espie@
2014-01-12plen is unsignedTheo de Raadt
2014-01-12allow prev release keys for now, transition 5.4 -> 5.5 kindof requiresMarc Espie
it.
2014-01-11typos, from Markus Lude, thx!Marc Espie
2014-01-11remove extraneous D, from Markus LudeStuart Henderson
2014-01-11simplify code: always extract, then install, so that initial installationsMarc Espie
and updates are more similar.
2014-01-11a bit of spring cleanup in advance: scrape old stuff that's not reallyMarc Espie
used.
2014-01-11Create cleaner & less noisy makefiles, now that we've been usingTheo de Raadt
this for 20 years. We don't need to see the splatter as much anymore.
2014-01-11just a little TLCTheo de Raadt
2014-01-11Let tcpdump detect bad ICMPv6 checksums with the -v flag.Lawrence Teo
Tested on amd64, i386, loongson, and macppc. OK florian@
2014-01-11Let tcpdump detect bad ICMP checksums with the -v flag.Lawrence Teo
Tested on amd64, i386, loongson, and macppc. OK florian@
2014-01-11Make icmp_print() accept the length variable, which is the length of theLawrence Teo
packet without the IP header. This is needed by the next commit that will allow tcpdump to detect bad ICMP checksums. Related functions like {tcp,udp,icmp6}_print() already accept this length variable, so this change makes icmp_print() consistent with them as well. This commit makes no functional change to tcpdump itself. OK florian@
2014-01-10improve ntpctl usage so that the manual page does not need to be readTheo de Raadt
every time ok jmc
2014-01-10do not list sha1 and sha256 in SEE ALSO, since md5 is already there, andJason McIntyre
they're one and the same page now;
2014-01-10likewise. this is uselessMarc Espie
2014-01-10by popular demand, remove excessive paranoiaMarc Espie
2014-01-10signify silent by default, don't bother working around stdout.Marc Espie
2014-01-10Check the return values of the strdup() calls.Lawrence Teo
OK deraadt@
2014-01-09bump copyright to 2014Marc Espie
2014-01-09if we're re-signing, check old sig first. shouldn't sign stuff we don'tMarc Espie
trust.
2014-01-09with the intermediate gunzip gone, this is a simple pipe. No need forMarc Espie
separate handles to tweak.
2014-01-09-C was useless with signify, report @signer identity instead (prependMarc Espie
"reportedly" to make tedu happy :) )
2014-01-09@signer makes sense only for signify, so move it there.Marc Espie
2014-01-09simplify a bit: pass the first SIGNER for @signer.Marc Espie
2014-01-09document SIGNERMarc Espie
2014-01-09switch to internal gzip/gunzip.Marc Espie
reduces the number of external processes and the complexity of the code. tested on a few select arches by tobiasu, naddy. If it breaks somewhere, tough. This one is simple to revert.
2014-01-09tweak signing yet again. Have pkg_create automatically add signingMarc Espie
identities every time, and make matching identities mandatory. e.g., pkg_create and pkg_add must have matching -DSIGNER. by default, signer is derived from uname -r and role (pkg_add/fw_update), e.g., 54pkg, 54fw...
2014-01-09Use destination bound rather than the source bound for out_line andMartynas Venckus
out_name. OK millert@.
2014-01-08zap trailing whitespace;Jason McIntyre
2014-01-08ctype cleanup; ok gillesTheo de Raadt
2014-01-08synch with signify(1) paranoiaMarc Espie
2014-01-07Asa Yeamans points out that we need to CMSG_SPACE[] for the 2nd messageTheo de Raadt
as well. ok guenther millert
2014-01-07severely restrict what characters can be in a signer, to makeMarc Espie
certain it's not possible to go ../ from /etc/signify
2014-01-07even if Ustar handles fifo and stuff gracefully, pkg contents should be moreMarc Espie
limited...
2014-01-07use IF_NAMESIZE as size of buffer for interface nameGleydson Soares
OK mpi@ deraadt@
2014-01-07scrape support for old +COMMENT file.Marc Espie
8 years of backward compatibility is long enough... :)
2014-01-07recognize special file +METAMarc Espie
2014-01-07tedu some very old checks: we no longer have INSTALL/UNINSTALL scripts,Marc Espie
so we don't care whether /var/db/pkg is mounted noexec.
2014-01-07let's verify all checksums.Marc Espie
I forgot about special files, since they've been extracted early, but they can be checked nonetheless. (prevents people from tampering with DESC and the like)
2014-01-07integrate signer logic and fix resign with -DSIGNERMarc Espie
2014-01-07document new stuffMarc Espie
2014-01-07if HISTORY_DIR is defined, create a new "permanent" file during build.Marc Espie
this file records the sha256 of all elements in the archive, and keeps track of changes (quite simply: sha256 that were already present end up at the end of the file). Start recording these *now*, so that we have enough info to shuffle packages later.
2014-01-07simplify set_modes, there's no need to stat(2) stuff, chown -1 takes careMarc Espie
of things.
2014-01-07more stuff than cannot end in .libs nor partial- packagesMarc Espie
2014-01-07... and let sign_list(\@ARGV... work too)Marc Espie
2014-01-07allow processing of package lists in parallel.Marc Espie
Turns out re-gzipping a signed package is really expensive, so turn it into several jobs with a -j option.