openbsd
10 years agospaces
deraadt [Tue, 5 Aug 2014 14:35:47 +0000 (14:35 +0000)]
spaces

10 years agohandle wsconsctl.conf and sysctl.conf just being examples.
deraadt [Tue, 5 Aug 2014 13:15:51 +0000 (13:15 +0000)]
handle wsconsctl.conf and sysctl.conf just being examples.
from ratchov

10 years agoAdd $OpenBSD$ tags.
jsing [Tue, 5 Aug 2014 12:46:16 +0000 (12:46 +0000)]
Add $OpenBSD$ tags.

10 years agoadd a config option to specify the chroot directory
jsg [Tue, 5 Aug 2014 09:24:21 +0000 (09:24 +0000)]
add a config option to specify the chroot directory
ok reyk@

10 years agoenable httpd; ok deraadt@
naddy [Mon, 4 Aug 2014 20:17:09 +0000 (20:17 +0000)]
enable httpd; ok deraadt@

10 years agoTemporarily move the default location of the SSL/TLS server key and
reyk [Mon, 4 Aug 2014 18:12:15 +0000 (18:12 +0000)]
Temporarily move the default location of the SSL/TLS server key and
certificate from /var/www/ to /var/www/conf/.  Don't get scared - this
will be changed soon!  They're currently located in the chroot
directory but will be moved outside as soon as we adopted some of the
key privsep from relayd in ressl/httpd.

10 years agoAdd HTTPS = on CGI variable.
reyk [Mon, 4 Aug 2014 18:00:06 +0000 (18:00 +0000)]
Add HTTPS = on CGI variable.

10 years agoAdd HTTPS server example.
reyk [Mon, 4 Aug 2014 17:50:48 +0000 (17:50 +0000)]
Add HTTPS server example.

10 years agoRedirect to https:// if SSL/TLS is enabled.
reyk [Mon, 4 Aug 2014 17:43:20 +0000 (17:43 +0000)]
Redirect to https:// if SSL/TLS is enabled.

10 years agoProxy commit for jsing@:
reyk [Mon, 4 Aug 2014 17:38:12 +0000 (17:38 +0000)]
Proxy commit for jsing@:
"Add TLS/SSL support to httpd, based on the recent ressl commits."

From jsing@
ok reyk@

10 years agomanpage tweaks about logging
reyk [Mon, 4 Aug 2014 17:12:44 +0000 (17:12 +0000)]
manpage tweaks about logging

10 years agoImplement ressl_accept_socket, which allocates a new server connection
jsing [Mon, 4 Aug 2014 16:34:11 +0000 (16:34 +0000)]
Implement ressl_accept_socket, which allocates a new server connection
context (if necessary) and handles the TLS/SSL handshake over the given
socket.

10 years agoReturn -1 on error (not 1).
jsing [Mon, 4 Aug 2014 16:19:50 +0000 (16:19 +0000)]
Return -1 on error (not 1).

10 years agoA ressl server needs different configuration from a ressl client - provide
jsing [Mon, 4 Aug 2014 16:18:42 +0000 (16:18 +0000)]
A ressl server needs different configuration from a ressl client - provide
a specific server configuration function and call this from
ressl_configure.

10 years agoChange grammar from "log [style]" to "log style [style]".
reyk [Mon, 4 Aug 2014 16:07:59 +0000 (16:07 +0000)]
Change grammar from "log [style]" to "log style [style]".

10 years agoProvide a function that returns a server connection context.
jsing [Mon, 4 Aug 2014 16:07:25 +0000 (16:07 +0000)]
Provide a function that returns a server connection context.

10 years agoProvide a utility function for loading a private/public keypair.
jsing [Mon, 4 Aug 2014 15:58:29 +0000 (15:58 +0000)]
Provide a utility function for loading a private/public keypair.

10 years agoPrint error message if the log files cannot be opened.
reyk [Mon, 4 Aug 2014 15:57:25 +0000 (15:57 +0000)]
Print error message if the log files cannot be opened.

10 years agoImprove ressl_{read,write} handling of non-blocking reads/writes.
jsing [Mon, 4 Aug 2014 15:55:26 +0000 (15:55 +0000)]
Improve ressl_{read,write} handling of non-blocking reads/writes.

10 years agoAdd initial support for log files in /var/www/logs/. Logging with
reyk [Mon, 4 Aug 2014 15:49:28 +0000 (15:49 +0000)]
Add initial support for log files in /var/www/logs/.  Logging with
syslog is still supported but disabled by default.

ok deraadt@

10 years agoFree the SSL context first and let the reference counting do its thing.
jsing [Mon, 4 Aug 2014 15:48:01 +0000 (15:48 +0000)]
Free the SSL context first and let the reference counting do its thing.

10 years agoImplement PATH_INFO and add DOCUMENT_ROOT.
reyk [Mon, 4 Aug 2014 14:49:24 +0000 (14:49 +0000)]
Implement PATH_INFO and add DOCUMENT_ROOT.
PATH_INFO was requested by naddy@ who successfully tested it with "cvsweb".

ok naddy@

10 years agofix small layout inconsistency
jasper [Mon, 4 Aug 2014 13:24:42 +0000 (13:24 +0000)]
fix small layout inconsistency

ok'd by many

10 years agohttpd doesn't support SSL/TLS yet, remove the remaining bits.
reyk [Mon, 4 Aug 2014 11:09:25 +0000 (11:09 +0000)]
httpd doesn't support SSL/TLS yet, remove the remaining bits.
The secrect plan is to add it later using the ressl wrapper library.

10 years agono need for param.h
deraadt [Mon, 4 Aug 2014 06:35:31 +0000 (06:35 +0000)]
no need for param.h

10 years agowhitespace
deraadt [Mon, 4 Aug 2014 06:35:10 +0000 (06:35 +0000)]
whitespace

10 years agoIn chacha_init(), allow for a NULL iv. Reported by znz on github.
miod [Mon, 4 Aug 2014 04:16:11 +0000 (04:16 +0000)]
In chacha_init(), allow for a NULL iv. Reported by znz on github.
ok guenther@ jsing@

10 years agoOnly allow GET and HEAD for static files or return 405.
reyk [Sun, 3 Aug 2014 22:47:25 +0000 (22:47 +0000)]
Only allow GET and HEAD for static files or return 405.

ok florian@

10 years agoAlso write log messages, like 404 Not Found, on error. This is a bit
reyk [Sun, 3 Aug 2014 22:38:12 +0000 (22:38 +0000)]
Also write log messages, like 404 Not Found, on error.  This is a bit
tricky because we couldn't guarantee a sane state after
server_response_http() so fail hard afterwards and close the connection.

ok doug@

10 years agoc-type functions / makros need a cast to unsigned char, not int
florian [Sun, 3 Aug 2014 22:06:51 +0000 (22:06 +0000)]
c-type functions / makros need a cast to unsigned char, not int
"feel free to commit" reyk@

10 years agoAllocate http_host instead of carrying a buffer in the descriptor.
reyk [Sun, 3 Aug 2014 21:33:27 +0000 (21:33 +0000)]
Allocate http_host instead of carrying a buffer in the descriptor.

10 years agospacing
reyk [Sun, 3 Aug 2014 20:43:03 +0000 (20:43 +0000)]
spacing

10 years agoDynamically pass HTTP request headers as protocol-specific HTTP_* CGI
reyk [Sun, 3 Aug 2014 20:39:40 +0000 (20:39 +0000)]
Dynamically pass HTTP request headers as protocol-specific HTTP_* CGI
meta-variables.

ok florian@

10 years agoAdd ral(4) at pci and cardbus to RAMDISK on macppc; ok deraadt
stsp [Sun, 3 Aug 2014 17:52:46 +0000 (17:52 +0000)]
Add ral(4) at pci and cardbus to RAMDISK on macppc; ok deraadt

10 years agoAdd ral(4) at cardbus to GENERIC on macppc; ok deraadt
stsp [Sun, 3 Aug 2014 17:52:10 +0000 (17:52 +0000)]
Add ral(4) at cardbus to GENERIC on macppc; ok deraadt

10 years agosmtpd got privsep'ed and the _smtpq user is now used to manage these
rpe [Sun, 3 Aug 2014 16:00:15 +0000 (16:00 +0000)]
smtpd got privsep'ed and the _smtpq user is now used to manage these
directories. Ensure proper ownership in case smtpd has been started
before this change.

brought up by matthieu@
OK deraadt@

10 years agoX509_NAME_get_text_by_NID() returns -1 on error so the type
jsg [Sun, 3 Aug 2014 14:35:30 +0000 (14:35 +0000)]
X509_NAME_get_text_by_NID() returns -1 on error so the type
the return value is stored in must be signed.  Fixes a test for error.

ok jsing@ guenther@

10 years agodon't return unitialised memory on error in ehci_alloc_sqtd()
jsg [Sun, 3 Aug 2014 14:30:28 +0000 (14:30 +0000)]
don't return unitialised memory on error in ehci_alloc_sqtd()

It seems mpi introduced a problem into ehci.c rev 1.162
two weeks ago.  An error check that used to return NULL
now jumps to the end of the function, but sqtd isn't
initialised at that point.

ok miod@ deraadt@

10 years agoWhen stsp changed bwi_encap() for 30 bit addresses in rev 1.106 code
jsg [Sun, 3 Aug 2014 14:23:59 +0000 (14:23 +0000)]
When stsp changed bwi_encap() for 30 bit addresses in rev 1.106 code
that used to always initialise the error variable is no longer run.

And at the end of bwi_encap() there is:

if (error)
m_freem(m);
return (error);

Fixing this prevents packet loss stsp was seeing.

ok stsp@ miod@ deraadt@

10 years agoAdd function to iterate all headers. No functional change.
reyk [Sun, 3 Aug 2014 12:26:19 +0000 (12:26 +0000)]
Add function to iterate all headers.  No functional change.

10 years agoThe first server example should be the "minimal default" to illustrate
reyk [Sun, 3 Aug 2014 11:51:40 +0000 (11:51 +0000)]
The first server example should be the "minimal default" to illustrate
that you don't have to push all kinds of buttons to run httpd.

10 years agoMore examples, include FastCGI for php and cgi-bin and logging.
reyk [Sun, 3 Aug 2014 11:28:58 +0000 (11:28 +0000)]
More examples, include FastCGI for php and cgi-bin and logging.

10 years agoSplit fastcgi socket path and document root option and add the
reyk [Sun, 3 Aug 2014 11:16:10 +0000 (11:16 +0000)]
Split fastcgi socket path and document root option and add the
SCRIPT_FILENAME CGI param with a prepended root.  This fixes php-fpm
that expects SCRIPT_FILENAME and also works with slowcgi if you
configure the root correctly.  For example, if SCRIPT_NAME and
REQUEST_URI are /php/index.php, root is /htdocs, SCRIPT_FILENAME will
be /htdocs/php/index.php.  As tested and discussed with florian@

10 years agoAdd missing log call for FastCGI requests.
reyk [Sun, 3 Aug 2014 10:38:42 +0000 (10:38 +0000)]
Add missing log call for FastCGI requests.

10 years agoAdd another log mode "connection" for a relayd(8)-style log entry after
reyk [Sun, 3 Aug 2014 10:26:43 +0000 (10:26 +0000)]
Add another log mode "connection" for a relayd(8)-style log entry after
each connection, not every request.  The code was already there and enabled
on debug, I just turned it into an alternative log format.

10 years agoPrefer getnameinfo() with NI_NUMERICHOST over inet_ntop because it is also
reyk [Sun, 3 Aug 2014 10:22:30 +0000 (10:22 +0000)]
Prefer getnameinfo() with NI_NUMERICHOST over inet_ntop because it is also
aware of the IPv6 scope Id.  We already have a function print_host() that
uses getnameinfo, so no need for the inet_ntop cases. Confirmed by florian@

10 years agoLocations now inherit access log settings from the server.
doug [Sat, 2 Aug 2014 21:21:47 +0000 (21:21 +0000)]
Locations now inherit access log settings from the server.

Add log to the server flags.

input/"Looks ok" reyk@

10 years agodon't leak fcgi fd
florian [Sat, 2 Aug 2014 17:42:24 +0000 (17:42 +0000)]
don't leak fcgi fd

10 years agoPadding of fcgi records is optional, but if we receive padding data we
florian [Sat, 2 Aug 2014 17:05:18 +0000 (17:05 +0000)]
Padding of fcgi records is optional, but if we receive padding data we
should read it.

10 years agoWe need to read from the fcgi bufferevent until it's empty because the
florian [Sat, 2 Aug 2014 11:59:04 +0000 (11:59 +0000)]
We need to read from the fcgi bufferevent until it's empty because the
event handler will not be called again if no new data arrives.
Debugged with and OK reyk@

10 years agoAllow to specify a FastCGI TCP socket on localhost (eg. :9000). Used
reyk [Sat, 2 Aug 2014 11:52:00 +0000 (11:52 +0000)]
Allow to specify a FastCGI TCP socket on localhost (eg. :9000).  Used
for debugging, you should prefer local UNIX sockets, but it helped to
find an issue that will be fixed with the next commit.

OK florian@

10 years ago'fastcgi socket "path"' is the correct syntax; update the manpage.
reyk [Sat, 2 Aug 2014 10:24:36 +0000 (10:24 +0000)]
'fastcgi socket "path"' is the correct syntax; update the manpage.
Found by jsg@

10 years agospacing
reyk [Sat, 2 Aug 2014 09:54:13 +0000 (09:54 +0000)]
spacing

10 years agoscandir(3)-based directory auto index didn't work on NFS because the
reyk [Sat, 2 Aug 2014 09:46:51 +0000 (09:46 +0000)]
scandir(3)-based directory auto index didn't work on NFS because the
file system is not filling in d_type properly.  Using st_mode from the
stat call fixes the problem, eg. S_ISDIR(st.st_mode) instead of
dp->d_type == DT_DIR.  Pointed out by pelikan@

10 years agoremove nasty unclosed Xo in previous; ok reyk
jmc [Sat, 2 Aug 2014 08:07:07 +0000 (08:07 +0000)]
remove nasty unclosed Xo in previous; ok reyk

10 years agoUse the log buffer to defer the logging until the connection is closed
reyk [Fri, 1 Aug 2014 22:24:05 +0000 (22:24 +0000)]
Use the log buffer to defer the logging until the connection is closed
or the request completed.  Turn the old log message into a debug message.

ok doug@

10 years agoremove the global "log updates/all" option that came from relayd.
reyk [Fri, 1 Aug 2014 21:59:56 +0000 (21:59 +0000)]
remove the global "log updates/all" option that came from relayd.

10 years agoAdd common and combined access logging to httpd.
doug [Fri, 1 Aug 2014 21:51:02 +0000 (21:51 +0000)]
Add common and combined access logging to httpd.

ok reyk@

10 years agoRewrite fcgi_add_param and hand over a lot more http headers etc. to
florian [Fri, 1 Aug 2014 18:26:32 +0000 (18:26 +0000)]
Rewrite fcgi_add_param and hand over a lot more http headers etc. to
the cgi script.
OK reyk@
"blanket OK" for changes in httpd for the time beeing from deraadt@

10 years agoCorrectly parse fcgi records if we don't get the whole record in one
florian [Fri, 1 Aug 2014 08:34:46 +0000 (08:34 +0000)]
Correctly parse fcgi records if we don't get the whole record in one
bufferevent_read().
Input/OK reyk@

10 years agoDon't return, just skip over non existing rc.conf or rc.conf.local file.
rpe [Fri, 1 Aug 2014 04:57:01 +0000 (04:57 +0000)]
Don't return, just skip over non existing rc.conf or rc.conf.local file.

OK ajacoutot@ halex@

10 years agoWhen attempting to rebuild a softraid volume, use the actual data offset
jsing [Fri, 1 Aug 2014 01:32:09 +0000 (01:32 +0000)]
When attempting to rebuild a softraid volume, use the actual data offset
from the volume metadata rather than the currently defined data offset.
This allows rebuilds to work correctly when the volume metadata has a
different data offset to that currently in use (for example, volumes
created prior to softraid gaining boot support).

Found the hard way by henning@

ok deraadt@

10 years agounbreak the build - e2fs_isave now needs the superblock.
pelikan [Thu, 31 Jul 2014 19:49:25 +0000 (19:49 +0000)]
unbreak the build - e2fs_isave now needs the superblock.

10 years agoalways use the little-endian copy of a superblock.
pelikan [Thu, 31 Jul 2014 19:11:38 +0000 (19:11 +0000)]
always use the little-endian copy of a superblock.

Damn those memcpy-wrapping macros!

"do it" deraadt

10 years agounbreak aac(4) by re-adding uvm_extern.h for ptoa()
jasper [Thu, 31 Jul 2014 18:14:46 +0000 (18:14 +0000)]
unbreak aac(4) by re-adding uvm_extern.h for ptoa()

ok deraadt@ sthen@

10 years agoOnly write the HTTP header for the first fastcgi chunk.
reyk [Thu, 31 Jul 2014 18:07:11 +0000 (18:07 +0000)]
Only write the HTTP header for the first fastcgi chunk.

10 years agosome fastcgi improvements:
reyk [Thu, 31 Jul 2014 17:55:09 +0000 (17:55 +0000)]
some fastcgi improvements:
- DPRINTF instead of log_info for internal debugging.
- submit QUERY_STRING, if it exists
- use a proper function to create an HTTP header.
- use server_file_error() to detect EOF and fastcgi stream errors.
- disable keep-alive/persist for now until we have a reliable way to
get the content length from the cgi response or support chunked
encoding.

"Cool, jep" florian@

10 years agouse proper on-disk inode size: no more, no less.
pelikan [Thu, 31 Jul 2014 17:37:52 +0000 (17:37 +0000)]
use proper on-disk inode size: no more, no less.

Reported by Roman Yakovlev, thanks!

"do it now" deraadt

10 years agoIn debug mode, properly sort and drop duplicates so that we don't end
ajacoutot [Thu, 31 Jul 2014 14:57:41 +0000 (14:57 +0000)]
In debug mode, properly sort and drop duplicates so that we don't end
up with a confusing output like:
multicast_host      >NO<
<...>
multicast_host      >YES<

Also properly evaluate values _after_ running _rc_quirks() because these
can modify flags.

ok robert@ halex@

10 years agoOne bufferevent can be shared by file and fcgi.
reyk [Thu, 31 Jul 2014 14:25:14 +0000 (14:25 +0000)]
One bufferevent can be shared by file and fcgi.

10 years agoAllow to specify a non-default fastcgi socket.
reyk [Thu, 31 Jul 2014 14:18:38 +0000 (14:18 +0000)]
Allow to specify a non-default fastcgi socket.

10 years agoRename the "docroot" variable to "path" because it will be used for
reyk [Thu, 31 Jul 2014 13:28:15 +0000 (13:28 +0000)]
Rename the "docroot" variable to "path" because it will be used for
either files or the fastcgi socket (and there's no need to use a union yet).

10 years agoAdd a configuration variable "fastcgi" to enable it per server or location.
reyk [Thu, 31 Jul 2014 09:34:57 +0000 (09:34 +0000)]
Add a configuration variable "fastcgi" to enable it per server or location.

10 years agoPut in first stab at fastcgi. Very early work in progress. Putting it
florian [Thu, 31 Jul 2014 09:23:53 +0000 (09:23 +0000)]
Put in first stab at fastcgi. Very early work in progress. Putting it
in now so that we can quickly work on it in tree. Requested by reyk@.
deraadt@ is OK with this according to reyk@.

10 years agosync
deraadt [Wed, 30 Jul 2014 17:45:13 +0000 (17:45 +0000)]
sync

10 years agounhook auxcpp
jsg [Wed, 30 Jul 2014 17:21:50 +0000 (17:21 +0000)]
unhook auxcpp

10 years agoProperly warn when an example changes and the corresponding file is found
ajacoutot [Wed, 30 Jul 2014 16:54:09 +0000 (16:54 +0000)]
Properly warn when an example changes and the corresponding file is found
under /etc.

issue reported by Nathanael Rensen
"fine" deraadt@

10 years agoswitch to tradcpp
jsg [Wed, 30 Jul 2014 16:46:17 +0000 (16:46 +0000)]
switch to tradcpp

10 years agodescend into tradcpp
jsg [Wed, 30 Jul 2014 16:34:24 +0000 (16:34 +0000)]
descend into tradcpp

10 years agoAdd tradcpp 0.4, a standalone traditional whitespace preserving cpp
jsg [Wed, 30 Jul 2014 16:33:11 +0000 (16:33 +0000)]
Add tradcpp 0.4, a standalone traditional whitespace preserving cpp
by David A. Holland of NetBSD.

10 years agoMake "location" work with name-based virtual servers.
reyk [Wed, 30 Jul 2014 13:49:48 +0000 (13:49 +0000)]
Make "location" work with name-based virtual servers.

10 years agoFix evil typo (multicast_hosts -> multicast_host).
ajacoutot [Wed, 30 Jul 2014 13:18:59 +0000 (13:18 +0000)]
Fix evil typo (multicast_hosts -> multicast_host).

10 years agoprepare for post 5.6 packages, recognize special case where timestamp exist.
espie [Wed, 30 Jul 2014 12:44:26 +0000 (12:44 +0000)]
prepare for post 5.6 packages, recognize special case where timestamp exist.
(specifically, this is a nop for 5.6, but it will allow changes to packages
without needing anything in pkg_add)

10 years agoAdd "location" keyword to specify path-specific configuration in
reyk [Wed, 30 Jul 2014 10:05:14 +0000 (10:05 +0000)]
Add "location" keyword to specify path-specific configuration in
servers, for example auto index for a sub-directory only.  Internally,
a "location" is just a special type of a "virtual" server.

10 years agoSmall fix and clarification
reyk [Wed, 30 Jul 2014 09:51:40 +0000 (09:51 +0000)]
Small fix and clarification

10 years agoReserve an extra file descriptor per connection instead of per
reyk [Wed, 30 Jul 2014 07:09:38 +0000 (07:09 +0000)]
Reserve an extra file descriptor per connection instead of per
request.  This fixes fd accounting with persistent connections and
reduces the complexity of the implementation.

ok benno@

10 years agoThe inflight decremented message should only be printed with DEBUG.
reyk [Tue, 29 Jul 2014 16:38:34 +0000 (16:38 +0000)]
The inflight decremented message should only be printed with DEBUG.

10 years agoAdd extended directory index options: "[no] index" and "[no] auto index".
reyk [Tue, 29 Jul 2014 16:17:28 +0000 (16:17 +0000)]
Add extended directory index options: "[no] index" and "[no] auto index".
The option "directory auto index" implements basic directory listing
and is turned off by default.

ok deraadt@

10 years agodisable POOL_DEBUG for release
deraadt [Tue, 29 Jul 2014 12:57:08 +0000 (12:57 +0000)]
disable POOL_DEBUG for release

10 years agomove to -release mode
deraadt [Tue, 29 Jul 2014 12:56:41 +0000 (12:56 +0000)]
move to -release mode

10 years agoRevert the checks about RTF_LOCAL routes.
mpi [Tue, 29 Jul 2014 12:18:41 +0000 (12:18 +0000)]
Revert the checks about RTF_LOCAL routes.

Even if in the end we would like to be more strict about what userland
can do with kernel-managed route entries, most of the tools out there
are not yet ready for this.  Since RTF_LOCAL routes are for the moment
just like RTF_LLINFO routes without expire timer, allow userland tools
to remove/modify them.  In case they are missing, the good old cloning
mechanism will recreate what you need.

bluhm@ and deraadt@ agree.

10 years agoMove configurable TCP options into struct server_config.
reyk [Tue, 29 Jul 2014 12:16:36 +0000 (12:16 +0000)]
Move configurable TCP options into struct server_config.

10 years agoFix a usage string; the proper spelling of 'alot' is 'a lot'.
blambert [Tue, 29 Jul 2014 08:51:49 +0000 (08:51 +0000)]
Fix a usage string; the proper spelling of 'alot' is 'a lot'.

ok bcook@

10 years agoupdate sets sizes
miod [Tue, 29 Jul 2014 02:57:09 +0000 (02:57 +0000)]
update sets sizes

10 years agoLast (known) msgbuf_write() vs EOF fix.
krw [Mon, 28 Jul 2014 22:18:31 +0000 (22:18 +0000)]
Last (known) msgbuf_write() vs EOF fix.

ok gilles@ deraadt@

10 years agosync
deraadt [Mon, 28 Jul 2014 20:35:32 +0000 (20:35 +0000)]
sync

10 years agoI/O ktrace of sendsyslog(2) did not work. As uiomove() adjusts
bluhm [Mon, 28 Jul 2014 20:30:01 +0000 (20:30 +0000)]
I/O ktrace of sendsyslog(2) did not work.  As uiomove() adjusts
iov_len to 0, we need a propper length calculation.  While there,
use -1 for the file descriptor because 0 is reserved for stdin.
OK deraadt@ guenther@

10 years agoOn SPARCbook systems, the ledma device node has a `cable-selection' property
miod [Mon, 28 Jul 2014 18:31:39 +0000 (18:31 +0000)]
On SPARCbook systems, the ledma device node has a `cable-selection' property
specifying which media the on-board interface uses. We already query it to
set up proper register values; extend this to be able to pass a default
media to the le(4) child.

This makes SPARCbook system default to AUI without needing for a manual media
change.

tested by sebastia@

10 years agoRemove SRP code. It contains a bug (this should not surprise anyone), but
tedu [Mon, 28 Jul 2014 17:57:18 +0000 (17:57 +0000)]
Remove SRP code. It contains a bug (this should not surprise anyone), but
the details are under embargo. The original plan was to wait for the
embargo to lift, but we've been waiting for quite some time, and there's no
indication of when or even if it will end. No sense in dragging this out
any longer.

The SRP code has never been enabled in OpenBSD, though I understand it is
in use by some other people. However, in light of this and other issues,
we're officially saying SRP is outside the scope of libressl. (For now.)

10 years agoFix memory exhaustion occurring on DHCP options with 0 length.
tobias [Mon, 28 Jul 2014 16:45:35 +0000 (16:45 +0000)]
Fix memory exhaustion occurring on DHCP options with 0 length.

halex@ and krw@ pointed out that a NULL check before free can go, too.

ok deraadt@, halex@, krw@