summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2002-02-16nanosleepArtur Grabowski
2002-02-16Test nanosleep(2).Artur Grabowski
2002-02-16add some (disabled) tests for interop with ssh.com serves.Markus Friedl
2002-02-16Some disk names are longer than 4. Cope with that.Thierry Deval
ok deraadt@
2002-02-15krb5_get_err_text() does not like context==NULL; he@nordu.net via google; ok ↵Markus Friedl
provos@
2002-02-15fork and start up a listener to discard bytes instead ofMarco S Hyman
relying on the discard service being enabled. Listener listens on localhost, only
2002-02-15split do_child(), ok mouring@Markus Friedl
2002-02-15syncTheo de Raadt
2002-02-15fix htdocs pages, changed with apache 1.3.23 update, to install the rightBob Beck
ones and not install the old ones.
2002-02-15more compact .out fileTodd T. Fries
2002-02-15syncTodd T. Fries
2002-02-15tweaks for vax; I must find a way of over-riding default target's, until then,Todd T. Fries
this feels alot like distrib/sets .. mi global and md for everybody that's gotta be different ..
2002-02-15Improve comment. ok art@Thomas Nordin
2002-02-15Don't cast nonexistent return value from splx to (void). ok art@Thomas Nordin
2002-02-15backwards testTheo de Raadt
2002-02-15fix the possible 4G overflow issuesMichael Shalayeff
2002-02-15slight snprintf paranoiaTheo de Raadt
2002-02-15undo latest commit for now, it breaks nanosleepPeter Valchev
2002-02-15Make rescheduling to a later time faster. ok art@Thomas Nordin
2002-02-15Don't get confused in nanosleep(2) when the time changes.Artur Grabowski
There is no need to use 'time' when we can use 'mono_time'. (oh and now it's faster too)
2002-02-15pf only uses seconds for time measuring. There is no need to call microtimeArtur Grabowski
on every packet. Use time.tv_sec to get seconds. In the places where it seemed to matter, make sure that time doesn't change under our feet. And it's really unnecessary to do a test on every packet when the test will only fire once every 10 seconds. That's a real waste of time, that's what we have timeouts for. ok frantzen@
2002-02-15Fix the __asm__ statements and remove the workarounds.Artur Grabowski
From NetBSD.
2002-02-15Clean up arpinput (based on similiar changes in NetBSD), also be sure toJason Wright
setup the arpcom when we're allowing lazy matches on bridged interfaces.
2002-02-15we don't need a listening sshd for most of our tests, justMarkus Friedl
connect client and server w/o TCP: ssh -o 'proxycommand sshd -i -f $OBJ/sshd_proxy'
2002-02-15Fix memory leak in get_proxy_env().Mike Pechkin
deraadt@ ok
2002-02-15delete agent keys, tooMarkus Friedl
2002-02-15bit of strcpy and sprintf cullingTheo de Raadt
2002-02-15Much simpler timeout calculation.Artur Grabowski
2002-02-15complete paranoiaTheo de Raadt
2002-02-15KNFTheo de Raadt
2002-02-15Xr tvtohz.Artur Grabowski
2002-02-15Document tvtohz.Artur Grabowski
2002-02-15Add a tvtohz function. Like hzto, but doesn't subtract the current time.Artur Grabowski
2002-02-15my guess at what will make a full distributionTheo de Raadt
2002-02-15vax support, bit hackish, will do for now; input from hughTheo de Raadt
2002-02-15 make pressing return in dired mode open the file we're currently at.Vincent Labrecque
ok deraadt@
2002-02-15chmod 644 authorized_keysMarkus Friedl
2002-02-15Correct BNF, unary port operators are optional and default to =,Daniel Hartmeier
"pass ... to any port = ssh" and "pass ... to any port ssh" are equivalent.
2002-02-14Reorder struct pf_pdesc members, saves 8 bytes.Daniel Hartmeier
2002-02-14hide some more implementation details of cipher.[ch] and prepares for moveMarkus Friedl
to EVP, ok deraadt@
2002-02-14increase the SSH v2 window size to 4 packets. comsumes a littleMarkus Friedl
bit more memory for slow receivers but increases througput.
2002-02-14serveral tests for ssh-keygen/ssh/sshd/ssh-add/ssh-agentMarkus Friedl
note that these tests will start sshd on 127.0.0.1:4242 and will fail if the realpath of /usr/src/usr.sbin/ssh/obj/ contains a group or world writeable directory.
2002-02-14Some fixes and additions, from Denis Afonin.Daniel Hartmeier
2002-02-14 remove the strcpy call, plus some KNFVincent Labrecque
ok deraadt@
2002-02-14d_makename now takes a length parameter so we can remove theVincent Labrecque
strcpy call.
2002-02-14No more RFORK_FDSAHRE option, always allow FORK_SHAREFILES.Artur Grabowski
2002-02-14The RFORK_FDSHARE option is going away. It's default and safe now.Artur Grabowski
2002-02-14More FREF/FRELE fixesPeter Valchev
2002-02-14 some programs now need the errno definition and sys/params, so put them inVincent Labrecque
there.
2002-02-14M4 generated MAKEDEV still premature.Hugh Graham