jca [Fri, 2 May 2014 09:50:32 +0000 (09:50 +0000)]
The page daemon no longer has a fixed PID. ok kettenis@ jmc@
andre [Fri, 2 May 2014 09:41:32 +0000 (09:41 +0000)]
Add missing $ on OpenBSD cvs tags, ok djm
miod [Fri, 2 May 2014 04:55:48 +0000 (04:55 +0000)]
No need to figure out the area spanned by .got in _dl_md_reloc(); only
_dl_md_reloc_boot() needs to compute this.
djm [Fri, 2 May 2014 03:27:54 +0000 (03:27 +0000)]
revert __bounded change; it causes way more problems for portable than
it solves; pointed out by dtucker@
djm [Fri, 2 May 2014 02:54:00 +0000 (02:54 +0000)]
use the test_helper fuzzer rather than the hand-rolled code that
predates it
djm [Fri, 2 May 2014 02:53:37 +0000 (02:53 +0000)]
make the fuzzer seed const to make it clear that it is never modified
miod [Thu, 1 May 2014 22:27:59 +0000 (22:27 +0000)]
Enable use of assembly code for AES, BN (Montgomery), SHA1, SHA256 and SHA512.
RC4 assembler code is not used, as it runs about 35% slower than the C code.
miod [Thu, 1 May 2014 22:26:21 +0000 (22:26 +0000)]
Do not output SOM-specific directives.
miod [Thu, 1 May 2014 22:25:13 +0000 (22:25 +0000)]
Remove unreferenced OPENSSL_instrument_bus and OPENSSL_instrument_bus2 routines.
miod [Thu, 1 May 2014 21:10:13 +0000 (21:10 +0000)]
Make the implicit `l' in `impicit' explicit.
miod [Thu, 1 May 2014 20:42:59 +0000 (20:42 +0000)]
Enable use of the assembly code for BN (Montgomery) and SHA1.
miod [Thu, 1 May 2014 20:41:56 +0000 (20:41 +0000)]
Fix include filename to get register name aliases under BSD
miod [Thu, 1 May 2014 20:40:55 +0000 (20:40 +0000)]
Pass -Werror in the !BN_LLONG !BN_UMULT_LOHI !BN_UMULT_HIGH case.
miod [Thu, 1 May 2014 20:15:06 +0000 (20:15 +0000)]
Enable use of the assembly code for AES, BN, SHA1, SHA256 and SHA512.
miod [Thu, 1 May 2014 19:06:46 +0000 (19:06 +0000)]
dead meat
jmc [Thu, 1 May 2014 18:59:15 +0000 (18:59 +0000)]
options first, then pager commands; also remove one non-standard Sh;
jmc [Thu, 1 May 2014 18:53:22 +0000 (18:53 +0000)]
STANDARDS is no longer relevant to this page, so remove it; while here,
zap some trailing whitespace introduced in recent update
jmc [Thu, 1 May 2014 18:50:36 +0000 (18:50 +0000)]
the beginning of removing more details from this page, since we now have
a separate more(1): it's a big page so i'm doing it in bits.
in this commit, remove more stuff from SYNOPSIS, remove some cleverness
from SYNOPSIS, and use consistent argument names;
jmc [Thu, 1 May 2014 18:46:01 +0000 (18:46 +0000)]
change some argument names in order to give some consistency between traceroute
and traceroute6; i also fudged the formatting to allow SYNOPSIS to display a
bit nicer, and for usage() to look like SYNOPSIS;
ok florian
jmc [Thu, 1 May 2014 18:41:59 +0000 (18:41 +0000)]
tweak previous; ok sthen
jmc [Thu, 1 May 2014 18:41:03 +0000 (18:41 +0000)]
update currency exchange rates;
deraadt [Thu, 1 May 2014 17:10:47 +0000 (17:10 +0000)]
sync
miod [Thu, 1 May 2014 17:08:57 +0000 (17:08 +0000)]
+ e_chacha.c, and bump minor
jsing [Thu, 1 May 2014 16:52:21 +0000 (16:52 +0000)]
Update with recently added objects.
ok miod@
jsing [Thu, 1 May 2014 16:50:36 +0000 (16:50 +0000)]
KNF.
deraadt [Thu, 1 May 2014 16:40:36 +0000 (16:40 +0000)]
Extend fread() and fwrite() to check for integer overflow, in which case
errno EOVERFLOW is returned and error is set on the FILE.
ok kettenis miod beck
jsing [Thu, 1 May 2014 16:07:13 +0000 (16:07 +0000)]
Nuke unused evptests.txt - the real one is over in regress.
ok miod@
jsing [Thu, 1 May 2014 16:06:24 +0000 (16:06 +0000)]
Provide an EVP implementation for ChaCha.
ok miod@
reyk [Thu, 1 May 2014 15:50:20 +0000 (15:50 +0000)]
Move RSA keys from "lka" to a new dedicated "ca" process because lka
is handling some async requests and shouldn't be busy with sync RSA.
ok gilles@
miod [Thu, 1 May 2014 15:04:31 +0000 (15:04 +0000)]
Fix perl arch directory name.
miod [Thu, 1 May 2014 15:00:19 +0000 (15:00 +0000)]
Revert 1.49 (bad merge with free dejavu)
deraadt [Thu, 1 May 2014 14:47:47 +0000 (14:47 +0000)]
sync
martynas [Thu, 1 May 2014 14:15:42 +0000 (14:15 +0000)]
x86-64 ABI requires arrays greater than 16 bytes to be aligned to
16byte boundary. However, GCC 16-byte aligns arrays of >=16 BITS,
not BYTES.
This diff improves bug detectability for code which has local arrays
of [16 .. 127] bits: in those cases SSP will now detect even 1-byte
overflows.
OK kettenis@. Tested in snaps for a week.
jasper [Thu, 1 May 2014 13:26:34 +0000 (13:26 +0000)]
fixup SoC name, as it's am335x, not am334x.
ok bmercer@
jsing [Thu, 1 May 2014 13:15:22 +0000 (13:15 +0000)]
Add ChaCha to libcrypto, based on djb's public domain implementation.
ok deraadt@
jasper [Thu, 1 May 2014 11:48:47 +0000 (11:48 +0000)]
- add 'sunxi' to list of miniroot files
- add list of supported 'sunxi' and 'imx' hardware
ok bmercer@
miod [Thu, 1 May 2014 11:29:18 +0000 (11:29 +0000)]
Add support for the french ANSSI FRP256v1 elliptic curve.
While not to be considered a good choice of elliptic curve (refer to
http://safecurves.cr.yp.to/ for more details), it is nevertheless deemed a
good decision to allow developers with requirements to use such a curve,
to be able to do this via a crypto library allowing for much better choices
to be made, without having to change (much of) their code to get better crypto.
ok beck@ deraadt@
miod [Thu, 1 May 2014 11:13:26 +0000 (11:13 +0000)]
Add Brainpool elliptic curves. From OpenSSL RT#2239 via ${DAYJOB}.
Be sure to rerun `make includes' after updating.
ok tedu@ beck@ deraadt@
miod [Thu, 1 May 2014 11:11:37 +0000 (11:11 +0000)]
Remove fips_md_init() macro indirection for digest algorithms, used by the
OpenSSL FIPS module to prevent forbidden digests to be allowed.
No functional change but readability.
ok deraadt@
miod [Thu, 1 May 2014 10:35:47 +0000 (10:35 +0000)]
No longer mention the tools/ directory, which content is irrelevant those days;
forgotten during previous cleanups.
jsg [Thu, 1 May 2014 10:25:44 +0000 (10:25 +0000)]
move pointer use to after a NULL pointer check
ok dlg@
jsg [Thu, 1 May 2014 07:35:57 +0000 (07:35 +0000)]
Correct a test for X509_get_notAfter() failing or returning
an unsupported time type when passing data to keynote.
Problem introduced by angelos in 1.41 though the code
has been reformatted a few times since then.
ok otto@ miod@
tedu [Thu, 1 May 2014 04:25:02 +0000 (04:25 +0000)]
with some random chance, swizzle the current page for the pool to avoid
fully deterministic behavior. ok deraadt
tedu [Thu, 1 May 2014 04:08:13 +0000 (04:08 +0000)]
nibbles aren't enough random, use bytes. does a better job of picking
a free chunk at random and may allow to increase delayed chunk array.
ok otto
sasano [Thu, 1 May 2014 03:45:08 +0000 (03:45 +0000)]
fixed some debug messages
ok by dcoppa@
sthen [Wed, 30 Apr 2014 23:28:05 +0000 (23:28 +0000)]
first cut at documenting the change to malloc doing a partial 'junk' by
default and the new 'j' option to disable this; ok jmc@
kettenis [Wed, 30 Apr 2014 19:25:14 +0000 (19:25 +0000)]
Make sure we flush discarded pages even if the number of hash buckets doesn't
change. From Pedro Martelletto via bitrig.
ok beck@, krw@
naddy [Wed, 30 Apr 2014 19:07:48 +0000 (19:07 +0000)]
UMAC can use our local fallback implementation of AES when OpenSSL isn't
available. Glue code straight from Ted Krovetz's original umac.c.
ok markus@
kettenis [Wed, 30 Apr 2014 16:07:31 +0000 (16:07 +0000)]
Assigning list pointers doesn't really work with doubly linked lists. Use
a remove-and-insert-all-items approach for now and remove the comments that
suggest manipulating list pointers. Pointed out by Pedro Martelletto.
ok beck@, krw@, mikeb@
krw [Wed, 30 Apr 2014 15:11:00 +0000 (15:11 +0000)]
Preserve intended chronological order of leases in
dhclient.leases.<if> files by writing out in reverse the
reverse-chronological-order TAILQ client->leases.
Found while trying to restore static lease functionality.
krw [Wed, 30 Apr 2014 14:28:48 +0000 (14:28 +0000)]
Write last disk sector, not last 512-byte block, when testing validity
of new partition limit.
Fixes growfs on 4K-sector disks.
Reported by and fix tested by David Vasek via misc@.
jsing [Wed, 30 Apr 2014 13:51:58 +0000 (13:51 +0000)]
Avoid a potential null pointer dereference by checking that we actually
managed to allocate a fragment, before trying to memcpy data into it.
ok miod@
jsing [Wed, 30 Apr 2014 13:40:02 +0000 (13:40 +0000)]
First pass at removing win64 support from the assembly generating Perl
scripts. We certainly do not need an identical copy of the win64
exception handler in each script (surely one copy would be sufficient).
ok miod@
gilles [Wed, 30 Apr 2014 12:49:54 +0000 (12:49 +0000)]
when doing opportunistic TLS, do not only downgrade during negotiation, but
also downgrade if a TLS error happens during the session.
ok eric@ who helped me with this
mpi [Wed, 30 Apr 2014 10:04:33 +0000 (10:04 +0000)]
Remove carp_seroute() it no longer does anything useful but mess with
your IPv4 routes.
mcbride@ says that this function has been introduced in order to be able
to reach the MASTER node from a BACKUP node using the CARP address. The
reasons are:
1) For troubleshooting, so I can ping or otherwise monitor the MASTER
host.
2) In some cases it's undesirable (or even not possible) to run
services on other IP addresses. For example, services that only allow
you to configure 1 listening IP, or services where you wish to avoid
users connecting to anything but the MASTER server.
Sadly this function becames a horrible hack. So if somebody thinks the
reasons explained before justify such logic, feel free to submit a
correct diff.
ok bluhm@, henning@, mikeb@
gilles [Wed, 30 Apr 2014 09:17:29 +0000 (09:17 +0000)]
when using maildir, do not create automatically create folders to match tag
in email address (ie: gilles+tag => ~/Maildir/.tag), instead use the folder
if it already exists and deliver to the mail Maildir otherwise.
ok eric@ and chl@
reyk [Wed, 30 Apr 2014 08:23:42 +0000 (08:23 +0000)]
The RSA engine (used by pony) has to wait for a response from the
privileged process (lka) and receive the imsgs in a while loop
synchronously. But the lka also sends other imsgs (DNS etc.) that can
still be queued up in the buffer when waiting for the RSA response.
This only happens under load with many concurrent connections. For
now, we just call the pony imsg handler for non-RSA imsgs that are
already in the buffer.
ok gilles@ eric@ blambert@
blambert [Wed, 30 Apr 2014 07:01:34 +0000 (07:01 +0000)]
replace vis-ified output with human-readable strings
for a few OIDs, with a framework in place to add more
as the mood strikes
ok reyk@ sthen@
sf [Wed, 30 Apr 2014 06:24:23 +0000 (06:24 +0000)]
format string fixes for (u)int64 vars
ok kettenis@
jmc [Wed, 30 Apr 2014 06:06:09 +0000 (06:06 +0000)]
tidy up SYNOPSIS, usage() and the options list; ok sthen
djm [Wed, 30 Apr 2014 05:32:00 +0000 (05:32 +0000)]
unit tests for new buffer API; including basic fuzz testing
djm [Wed, 30 Apr 2014 05:29:56 +0000 (05:29 +0000)]
New buffer API; the first installment of the conversion/replacement
of OpenSSH's internals to make them usable as a standalone library.
This includes a set of wrappers to make it compatible with the
existing buffer API so replacement can occur incrementally.
With and ok markus@
Thanks also to Ben Hawkes, David Tomaschik, Ivan Fratric, Matthew
Dempsky and Ron Bowes for a detailed review.
miod [Wed, 30 Apr 2014 04:20:31 +0000 (04:20 +0000)]
Enforce proper alignment of stack variables which may get accessed with
double-word load and store instructions. This used to work by chance, but
recent compiler changes no longer put us in the lucky situation.
tweaks kettenis@
tedu [Wed, 30 Apr 2014 01:05:50 +0000 (01:05 +0000)]
i suck at math. b - 8 = 3, not 4. detected by Antoine Pitrou. thanks.
deraadt [Wed, 30 Apr 2014 00:20:28 +0000 (00:20 +0000)]
sync
sperreault [Wed, 30 Apr 2014 00:16:47 +0000 (00:16 +0000)]
minor++ because of AI_ADDRCONFIG
prodded by naddy, confirmed by Mark Kettenis
nicm [Tue, 29 Apr 2014 22:31:22 +0000 (22:31 +0000)]
fcntl.h is still needed here.
mpi [Tue, 29 Apr 2014 21:51:18 +0000 (21:51 +0000)]
Get rid of the per-softc freelist of transfer descriptors and use a
per-driver pool(9) instead.
dcoppa [Tue, 29 Apr 2014 21:30:20 +0000 (21:30 +0000)]
tedu ~/.klogin
ok henning@ deraadt@
reyk [Tue, 29 Apr 2014 21:04:17 +0000 (21:04 +0000)]
For RSA private key privsep, only ever load the keys after forking the
separated process. This improves the previous because we don't trust
the PEM and BIO routines to cleanup the keys correctly.
ok gilles@
dtucker [Tue, 29 Apr 2014 20:36:51 +0000 (20:36 +0000)]
Don't attempt to append a nul quote char to the filename. Should prevent
fatal'ing with "el_insertstr failed" when there's a single quote char
somewhere in the string. bz#2238, ok markus@
dtucker [Tue, 29 Apr 2014 19:58:50 +0000 (19:58 +0000)]
Move nulling of variable next to where it's freed. ok markus@
reyk [Tue, 29 Apr 2014 19:13:13 +0000 (19:13 +0000)]
Implement RSA privilege separation for OpenSMTPD, based on my previous
implementation for relayd(8). The smtpd(8) pony processes (mta
client, smtp server) don't keep the private keys in memory but send
their private key operations as imsgs to the "lookup"/mta process.
It's worth mentioning that this prevents acidental private key leakage
as it could have been caused by "Heartbleed".
ok gilles@
markus [Tue, 29 Apr 2014 18:01:49 +0000 (18:01 +0000)]
make compiling against OpenSSL optional (make OPENSSL=no);
reduces algorithms to curve25519, aes-ctr, chacha,
ed25519;
allows us to explore further options; with and ok djm
jmc [Tue, 29 Apr 2014 17:35:29 +0000 (17:35 +0000)]
a bit more merge for traceroute6 stuff, and some consistency fixes;
help/ok florian
gilles [Tue, 29 Apr 2014 17:32:42 +0000 (17:32 +0000)]
when a session fails due to a TLS error in a smtp+tls:// connection, try
plain before giving up
ok eric@
beck [Tue, 29 Apr 2014 15:46:54 +0000 (15:46 +0000)]
Constrain bytes read/written to positive values.
ok miod@ tedu@
markus [Tue, 29 Apr 2014 15:39:33 +0000 (15:39 +0000)]
re-add our own aesctr implementation; ok djm@
florian [Tue, 29 Apr 2014 15:03:07 +0000 (15:03 +0000)]
Move traceroute6 to the attic, fully merged into traceroute.
mpi [Tue, 29 Apr 2014 14:11:23 +0000 (14:11 +0000)]
Finally plug the public xfer leak #1 in our USB stack.
Every call to usbd_abort_pipe() on an interrupt pipe would simply
reset the intrxfer pointer, which would prevent usbd_close_pipe()
to free it. Since we abort pipes in a lot of situations: when a
device is detached, when a USB-to-serial adapter is closed, when
an error occurs, when the machine is suspended, etc, this would
result in hundreds of leaked xfers in most of my machines.
xhci(4) is not affected, but you can't enable it right now since
the stack is not ready :)
While here put a KASSERT() to make sure drivers are only calling
the interrupt abort method for intrxfer, if that's not the case,
please let met know.
tedu [Tue, 29 Apr 2014 14:10:07 +0000 (14:10 +0000)]
nginx dudes unrolled SSL_OP_ALL by hand so we also have to delete
SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS from here to benefit. ok miod
mcbride [Tue, 29 Apr 2014 13:49:19 +0000 (13:49 +0000)]
regen
jmc [Tue, 29 Apr 2014 13:48:18 +0000 (13:48 +0000)]
smtpd is the default mailer now;
From: Jan Stary
ok gilles
mcbride [Tue, 29 Apr 2014 13:46:55 +0000 (13:46 +0000)]
Add Microsoft Wireless Mobile Mouse 3500 and Wireless Arc Mouse.
ok mpi
djm [Tue, 29 Apr 2014 13:10:30 +0000 (13:10 +0000)]
bz#1818 - don't send channel success/failre replies on channels that
have sent a close already; analysis and patch from Simon Tatham;
ok markus@
mpi [Tue, 29 Apr 2014 12:53:33 +0000 (12:53 +0000)]
Make sure to call uhidev_close() upon detach, plug another xfer leak.
mpi [Tue, 29 Apr 2014 12:47:33 +0000 (12:47 +0000)]
Plug an xfer leak.
mpi [Tue, 29 Apr 2014 12:45:29 +0000 (12:45 +0000)]
Get rid of the per-softc freelist of transfer descriptors and use a
per-driver pool(9) instead.
With inputs from mikeb@
reyk [Tue, 29 Apr 2014 12:18:27 +0000 (12:18 +0000)]
Remove unused arguments from ssl_smtp_init()
ok gilles@
millert [Tue, 29 Apr 2014 12:11:25 +0000 (12:11 +0000)]
Remove support for the obsolete (non-POSIX) "more -d" prompt.
This was a local change that was only enabled when the LESS_IS_MORE
environment variable was set and not when invoked as "more".
OK shadchin@ jmc@
mpi [Tue, 29 Apr 2014 11:58:29 +0000 (11:58 +0000)]
If you plan to write an obfuscated-by-design kernel / userland
interface, I suggest you have a look at the link-layer sockaddr
interface:
/*
* A Link-Level Sockaddr may specify the interface in one of two
* ways: either by means of a system-provided index number (computed
* anew and possibly differently on every reboot), or by a human-readable
* string such as "il0" (for managerial convenience).
[...]
*/
ifa_ifwithnet() was not only checking for the sdl_index in order
to get the corresponding ifp for AF_LINK sockaddr, it was also
iterating over all the addresses on your system! But in this
case, the `address' field of "struct sockaddr_dl" is an interface
name set by link_addr(3).
How can this work? Well because the kernel allocates an empty
`netmask' field for each interface's lladdr, so that you can
abuse a network comparison function to reimplement strcmp(3)...
So when the userland does not specify an interface index, try
harder to see if it passed an ifp name, but at least be explicit
and use ifunit().
Found the hard way by/ok sthen@
markus [Tue, 29 Apr 2014 11:51:13 +0000 (11:51 +0000)]
make sure the state machine only advances if the AUTH payload has
been verified; with & ok mikeb@
reyk [Tue, 29 Apr 2014 10:18:06 +0000 (10:18 +0000)]
use explicit_bzero() instead of memset() to clear out sensitive data.
ok gilles@
reyk [Tue, 29 Apr 2014 10:08:55 +0000 (10:08 +0000)]
It is only required to load the keys and certs into the same SSL
context once. Simplify the code path by moving the loading from three
different places into ssl_ctx_create():
ok gilles@
kettenis [Tue, 29 Apr 2014 09:55:28 +0000 (09:55 +0000)]
Replace 1 << PAGE_SHIFT with PAGE_SIZE.
ok beck@, miod@
dcoppa [Tue, 29 Apr 2014 07:52:06 +0000 (07:52 +0000)]
Fix typo: d_addrt -> daddr_t
ok bmercer@
jsg [Tue, 29 Apr 2014 07:44:19 +0000 (07:44 +0000)]
fix a test for invalid timezones
oks from a bunch of people
andre [Tue, 29 Apr 2014 07:23:40 +0000 (07:23 +0000)]
Remove device white-listing as matching method, relies on on usage id
versus supported sensors table filtering.
discussed with and ok mpi@
sperreault [Mon, 28 Apr 2014 21:38:59 +0000 (21:38 +0000)]
Implement AI_ADDRCONFIG
This is a getaddrinfo() flag that is defined thusly in RFC 3493:
If the AI_ADDRCONFIG flag is specified, IPv4 addresses shall be
returned only if an IPv4 address is configured on the local system,
and IPv6 addresses shall be returned only if an IPv6 address is
configured on the local system. The loopback address is not
considered for this case as valid as a configured address.
For example, when using the DNS, a query for AAAA records should
occur only if the node has at least one IPv6 address configured
(other than IPv6 loopback) and a query for A records should occur
only if the node has at least one IPv4 address configured (other
than the IPv4 loopback).
The flag is set by default when hints is NULL.
ok Eric Faurot, Jason McIntyre
miod [Mon, 28 Apr 2014 21:14:50 +0000 (21:14 +0000)]
Remove WIN32, WIN64 and MINGW32 tentacles.
Also check for _LP64 rather than __arch64__ (the former being more reliable
than __LP64__ or __arch64__) to tell 64-bit int platforms apart from 32-bit
int platforms.
Loosely based upon a diff from Martijn van Duren on tech@