summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2002-03-02Add machine specific commands to the bootblocks, to let the user displayMiod Vallat
and change the console and keyboard settings; and devise a manual page for it, with lots of loot borrowed from the i386 version. These commands have been extensively tested on 712, 715, and 720, although your mileage may vary and scary warnings may appear in the manual page in the future. ok mickey@
2002-03-02Add a strcmp() prototype, which will be needed shortly.Miod Vallat
2002-03-02Merge the useful bits of the 4.4BSD src/share/dict/README andTodd C. Miller
src/usr.bin/spell/README files into something (hopefully) coherent.
2002-03-02document tcp.ackonpushNiels Provos
2002-03-02disable immediate ack on TH_PUSH. make behaviour sysctl tuneable.Niels Provos
from netbsd; also fix a bug where setting TF_ACKNOW didn't actually result in an ack.
2002-03-02kill #ifdef REMOTE stuff that only obfuscates issues.Marc Espie
ok millert@, miod@
2002-03-01reassignMichael Shalayeff
2002-03-01fix a few irq assignmentsMichael Shalayeff
2002-03-01remove tcp_fasttimo and convert delayed acks to the timeout(9) API instead.Niels Provos
adapated from netbsd. okay angelos@
2002-03-01Sort -df, not sort -fTodd C. Miller
2002-03-01spell command from 4.4BSD, now freed by Caldera.Todd C. Miller
I have made some significant changes: o spell no longer uses hashed files. The algorithm for generating the hashed files breaks down for large files. Instead, I use the same approach as look(1) and do binary searches of mmap(2)ed files which has acceptable performance. A side effect of this is that spell can now search multiple word lists. o The spell binary has been renamed to "spellprog" so as to not conflict with the spell shell script. This is consistent with other operating systems. o The putsuf() and putword() functions have been renamed to trypref() and tryword() respectively for consistency with modern versions and for general sanity. o the spell front-end shell script has been completely rewritten o The code has been ANSIfied and re-style(9)d spell currently must be compiled with -fwriteable-strings to make the -b option work. This should be fixed.
2002-03-01From NetBSD, noticed by Alexander YurchenkoChris Cappuccio
"In hpt_pci_intr(): There is interrupt pending once IDEDMA_CTL_ACT is turned off. It seems that IDEDMA_CTL_INTR is asserted before DMA transfer is complete, leading to race condition in case of interrupt sharing. Discovered reading the FreeBSD code."
2002-03-01Change a config(8) reference to a more appropriate boot_config(8) reference.Miod Vallat
2002-03-01disable undo until it is fixed to be (1) correct and (2) not crash like it ↵Theo de Raadt
does now
2002-03-01Update to reflect recent changes in DH group handling. Remove BUGSHakan Olsson
section.
2002-03-01Change DH group handling in the pre-generated parts of theHakan Olsson
configuration. Add a -GRP{1,2,5} component to transform and suite names to directly specify which group to use. If no group is specified, use DH group 2 (MODP_1024). Earlier transforms and suites using the MD5 hash defaulted to DH group 1, this is no longer true. niklas@ ok.
2002-03-01this is needed tooMarc Espie
2002-03-01undo the 'delay hostname lookup' changeMarkus Friedl
match.c must not use compress.c (via canonhost.c/packet.c) thanks to wilfried@
2002-03-01Kill hand-made memory allocation code, that is definitely buggy.Marc Espie
Replace with simple wrapper around malloc, at least this works, and it's easier to debug anyways.
2002-03-01fp doesn't necessarily have to be set when we do the early FRELE.Artur Grabowski
2002-03-01sync with KAMEKenjiro Cho
minor man fixes from NetBSD.
2002-03-01sync with KAMEKenjiro Cho
fix the argument for an unknown hash policy: argv -> *argv
2002-03-01mac68k syncMiod Vallat
2002-03-01sync with KAMEKenjiro Cho
move the function prototype of altq_etherclassify() from netinet/if_ether.h to altq/if_altq.h.
2002-03-01document cryptodevallowsoftNiels Provos
2002-03-01sysctl for cryptodevallowsoft - /dev/crypto will allow the software engineNiels Provos
to be used if there are no hardware cards. mostly for debugging and regression.
2002-03-01remove CRYPTO_BUF_CONTIG and convert to handle iovs. okay deraadt@Niels Provos
2002-03-01helper functions for uio (cuio_apply and cuio_getptr) okay deraadt@Niels Provos
2002-03-01#!/bin/sh not cshMiod Vallat
2002-03-01Fix an err() that should really be errx()Todd C. Miller
2002-03-01Add a big CAVEATS section that explains the volatile environment whenArtur Grabowski
shutdown hooks are run.
2002-02-28Remove the ifconfig line again, it's in pflog(4), which is now linked,Daniel Hartmeier
and in pflogd context (started from rc), the interface is up already.
2002-02-28Document ethernet layer expressions.Daniel Hartmeier
2002-02-28Up the pflog0 interface. alphabetize xrefs and add pflog(4)Kjell Wooding
ok dhartmei@
2002-02-28forgot viper_hpaMichael Shalayeff
2002-02-28change the text following the preprocessor directives '#else' andDan Harnett
'#endif' into comments. ok millert@
2002-02-28hide proto from locoreMichael Shalayeff
2002-02-28All the -m options were not being displayed, use .Oo and .Oc toTodd C. Miller
avoid number of arguments limit. Also, file is optional; Brian Poole
2002-02-28merge mod_ssl 2.8.7Brad Smith
-- Ok'd by: beck@
2002-02-28log user not allowed details, from dwd@bell-labs.com; ok markus@Kevin Steves
2002-02-28-u0 DNS for user@hostKevin Steves
2002-02-28no, that error message was bad.Bob Beck
2002-02-28DenyUsers allows user@host pattern alsoKevin Steves
2002-02-28Some new servers return the hostname as type XBob Beck
this ensures we check against the right thing and don't reject leases when we shouldn't.
2002-02-28delay hostname lookup until we see a ``@'' in DenyUsers and AllowUsersKevin Steves
for sshd -u0; ok markus@
2002-02-28Check in some new vnode man page from FreeBSD. Thanks to Peter WernerConstantine Sapuntzakis
2002-02-28remove more 0 byte files.Brad Smith
2002-02-28remove 0 byte files. these files should have been removed over 14 monthsBrad Smith
ago.
2002-02-28use red-black tree for lookup_entry. the red-black tree case forNiels Provos
map_findspace is still broken on alpha. this will make debugging easier. okay millert@
2002-02-28another bugTodd C. Miller