Sync removal of setsid(), setpgid() and a few dup2() from relayd. They are
redundant since we call daemon() earlier.

ok bluhm@

Call daemon() only in parent and before proc_exec() to avoid orphaning child
processes. Synced from relayd.

ok kn@

Convert to use imsg_get_fd() since proc_forward_imsg() never forwards a
file descriptor just use -1 there.
OK tb@

proc_ispeer() is not used anywhere anymore so remove it everywhere.

ok florian@ bluhm@
ok for vmd mlarkin@

Do not setup pipes between SERVER processes, they don't talk to each
other. Since this generates a full mesh, the amount of filedescriptors
needed grows quadratically with the amount of configured pr

Do not setup pipes between SERVER processes, they don't talk to each
other. Since this generates a full mesh, the amount of filedescriptors
needed grows quadratically with the amount of configured prefork
processes.

Might fix an out of filedescriptor bug that beck is seeing.
OK benno

show more ...

Move TAILQ initialization to files where they are used.

These priv-sep daemons all follow a similar design and use TAILQs
for tracking control process connections. In most cases, the TAILQs
are init

Move TAILQ initialization to files where they are used.

These priv-sep daemons all follow a similar design and use TAILQs
for tracking control process connections. In most cases, the TAILQs
are initialized separate from where they are used. Since the scope
of use is generally confined to a specific control process file,
this commit also removes any extern definitions and exposing the
TAILQ structures to other compilation units.

ok bluhm@, tb@

show more ...

remove tautological condition
from Ross L Richardson <openbsd AT rlr DOT id DOT au>, Thanks
ok claudio@

During the fork+exec implementation, daemon(3) was moved after
proc_init(). As a consequence httpd(8) and relayd(8) child processes
did not detach from the terminal anymore. Dup /dev/null to the
st

During the fork+exec implementation, daemon(3) was moved after
proc_init(). As a consequence httpd(8) and relayd(8) child processes
did not detach from the terminal anymore. Dup /dev/null to the
stdio file descriptors in the children.
OK benno@

show more ...

use __func__ in log messages.
From Hiltjo Posthuma hiltjo -AT codemadness -DOT- org, thanks!
ok florian, claudio

Nuke some whitespace that keeps poking me in the eye as I try to
steal code.

Stop accessing verbose and debug variables from log.c directly.

This replaces log_verbose() and "extern int verbose" with the two functions
log_setverbose() and log_getverbose().

Pointed out by ben

Stop accessing verbose and debug variables from log.c directly.

This replaces log_verbose() and "extern int verbose" with the two functions
log_setverbose() and log_getverbose().

Pointed out by benno@
OK krw@ eric@ gilles@ (OK gilles@ for the snmpd bits as well)

show more ...

Prevent fd exhaustion in the parent when loading the listening server
sockets by sending the fd one-by-one. This allows to start httpd with
max 32 server instances and many server sockets without ch

Prevent fd exhaustion in the parent when loading the listening server
sockets by sending the fd one-by-one. This allows to start httpd with
max 32 server instances and many server sockets without changing the
default rlimits in any way.

OK rzalamena@

show more ...

Modify httpd(8)'s proc.c to use less file descriptors during the daemon
start up. To achieve this proc_init() initiates only the necessary pipes
between child and parent, allocate and distribute fds

Modify httpd(8)'s proc.c to use less file descriptors during the daemon
start up. To achieve this proc_init() initiates only the necessary pipes
between child and parent, allocate and distribute fds in proc_connect().

In case of configuration checks ('-n') we do nothing in proc_init() and
proc_connect().

ok reyk@

show more ...

Fix msgbuf_write() usage idiom and modify the treatment for socket close
to exit gracefully instead of fatal()ing.

ok reyk@

Add more context to fatal*() messages so it makes easier to debug proc.c
internals.

ok phessler@

Check if oldd == newd before dup2(), if that is the case we need to remove
the CLOEXEC flag ourselves.

ok bluhm@, deraadt@

Call setsid() to create a new session for the executed processes.

From deraadt@
OK rzalamena@

sync proc.c with vmd: add p_pw to specify a non-standard user for a process.

OK rzalamena@

sync proc.c from switchd, includes minor cast qual fix and removal of p_env.

Replace [RELAY|SERVER]_MAXPROC with the new PROC_MAX_INSTANCES
variable and limit it from 128 to 32 instances (the old value).
While here, move a few PROC_ defines around.

OK rzalamena@

Use DPRINTF instead of #ifdef DEBUG + log_debug().

Pointed out by benno@

proc.c tweaks: Rename proc_listento() to proc_accept() as it is the
receiving side of proc_connect(). Move some code from main into
proc_init(), the function is now called by parent and children, no

proc.c tweaks: Rename proc_listento() to proc_accept() as it is the
receiving side of proc_connect(). Move some code from main into
proc_init(), the function is now called by parent and children, not
just the parent and it is less copy + paste for other daemons.

OK florian@

show more ...

Don't print "lost child" if the child process exited okay. This is
the old behaviour and unbreaks the regress tests.

spacing

Adjust log message, use process title now that it works again