Move FREF() inside fd_getfile().

ok visa@

Call FREF() right after fd_getfile() in getsock().

This ensure that all operations manipulating a 'struct file *' do so
with a properly refcounted element.

ok visa@, bluhm@

Call socreate() before falloc() in sys_socket().

This is similar to what we do in sys_socketpair() and will allow us
to grab the KERNEL_LOCK() only after having created a socket.

This time with cor

Call socreate() before falloc() in sys_socket().

This is similar to what we do in sys_socketpair() and will allow us
to grab the KERNEL_LOCK() only after having created a socket.

This time with correct non-blocking check.

Tested by landry@, previous diff ok tedu@

show more ...

Grab solock() inside soconnect2() instead of asserting for it to be held.

ok millert@

Remove almost unused `flags' argument of suser().

The account flag `ASU' will no longer be set but that makes suser()
mpsafe since it no longer mess with a per-process field.

No objection from mill

Remove almost unused `flags' argument of suser().

The account flag `ASU' will no longer be set but that makes suser()
mpsafe since it no longer mess with a per-process field.

No objection from millert@, ok tedu@, bluhm@

show more ...

Revert rev 1.163. Causes network issues in Firefox.

ok mpi@ who will investigate.

Call socreate() before falloc() in sys_socket().

This is similar to what we do in sys_socketpair() and will allow us
to grab the KERNEL_LOCK() only after having created a socket.

ok tedu@

Change `so_state' and `so_error' to unsigned int such that they can
be atomically read from any context.

ok bluhm@, visa@

Stop assuming <sys/file.h> will pull in fcntl.h when _KERNEL is defined.

ok millert@ sthen@

Change sosetopt() to no longer free the mbuf it receives and change
all the callers to call m_freem(9).

Support from deraadt@ and tedu@, ok visa@, bluhm@

Make sogetopt(9) caller responsible for allocating an MT_SOOPTS mbuf.

Move a blocking memory allocation out of the socket lock and create
a simpler alloc/free pattern to review. Now both m_get() an

Make sogetopt(9) caller responsible for allocating an MT_SOOPTS mbuf.

Move a blocking memory allocation out of the socket lock and create
a simpler alloc/free pattern to review. Now both m_get() and m_free()
are in the same place.

Discussed with bluhm@.

Encouragements from deraadt@ and tedu@, ok kettenis@, florian@, visa@

show more ...

Move the solock()/sounlock() dance outside of sobind().

ok phessler@, visa@, bluhm@

Move the socket lock "above" sosetopt(), sogetopt() and sosplice().

Protect the fields modifieds by sosetopt() and simplify the dance
with the stars.

ok bluhm@

Extend the scope of the socket lock to protect `so_state' in connect(2).

As a side effect, soconnect() and soconnect2() now expect a locked socket,
so update all the callers.

ok bluhm@

When receiving a struct sockaddr from userland, enforce that memory
for sa_len and sa_family is provided. This will make handling of
socket name mbufs within the kernel safer.
issue reported by Ilja

When receiving a struct sockaddr from userland, enforce that memory
for sa_len and sa_family is provided. This will make handling of
socket name mbufs within the kernel safer.
issue reported by Ilja Van Sprundel; OK claudio@

show more ...

Move KTRPOINT call up. The lenght variable i is getting aligned and so
uninitialised data can be dumped into the ktrace message.
Found by Ilja Van Sprundel
OK bluhm@

Do not call fo_ioctl() in syscall that do, or will, take the socket
lock.

Prevents a future lock recursion since soo_ioctl() will need to grab
the lock.

ok bluhm@, visa@

Do not grab the socket lock in doaccept() twice. Pass NOTE_SUBMIT
to KNOTE() as we are already holding the lock. Fixes "panic:
rw_enter: netlock locking against myself" reported by Gregor Best
and

Do not grab the socket lock in doaccept() twice. Pass NOTE_SUBMIT
to KNOTE() as we are already holding the lock. Fixes "panic:
rw_enter: netlock locking against myself" reported by Gregor Best
and reproduced with src/regress/lib/libtls/gotls.
OK millert@

show more ...

Reorder FREF() and FRELE() in a way that the the global variable
syslogf always points to a file object with increased reference
count. This makes the implementation independent from the fact
whethe

Reorder FREF() and FRELE() in a way that the the global variable
syslogf always points to a file object with increased reference
count. This makes the implementation independent from the fact
whether changing the reference counter may sleep.
pointed out by Mateusz Guzik; OK deraadt@

show more ...

Wrap the NET_LOCK() into a per-socket solock() that does nothing for
unix domain sockets.

This should prevent the multiple deadlock related to unix domain sockets.

Inputs from millert@ and bluhm@,

Wrap the NET_LOCK() into a per-socket solock() that does nothing for
unix domain sockets.

This should prevent the multiple deadlock related to unix domain sockets.

Inputs from millert@ and bluhm@, ok bluhm@

show more ...

Add a flags argument to falloc() that lets it optionally set the
close-on-exec flag on the newly allocated fd. Make falloc()'s
return arguments non-optional: assert that they're not NULL.

ok mpi@ m

Add a flags argument to falloc() that lets it optionally set the
close-on-exec flag on the newly allocated fd. Make falloc()'s
return arguments non-optional: assert that they're not NULL.

ok mpi@ millert@

show more ...

Allocate the mbuf before the netlock. While here, move the setting of
nflag closer to where its value is used.

ok mpi@

Allocate and connect sockets first, then upon success create file
descriptors so that the NET_LOCK is satisfied.
ok mpi

Enable the NET_LOCK(), take 2.

Recursions are currently known and marked a XXXSMP.

Please report any assert to bugs@

In accept(2) and accept4(2) allocate a new file descriptor before
checking if the socket head's queue is empty and possibly sleeping.

This way we avoid lock ordering problems as the NET_LOCK() and
f

In accept(2) and accept4(2) allocate a new file descriptor before
checking if the socket head's queue is empty and possibly sleeping.

This way we avoid lock ordering problems as the NET_LOCK() and
fdplock() won't be held at the same time.

Note that socketpair(2) and close(2) are the two remaining syscalls
holding these locks at the same time. They both respect the same
order: fdplock() then NET_LOCK().

Initial deadlock reported by kettenis@ and ajacoutot@.

ok bluhm@, guenther@, deraadt@

show more ...