net_server: handle EINTR on select()

Change-Id: I90af101cef72b33ecb6decec911d03f592c1ec7c
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3729
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
Reviewed-by: Axel Dörfler <axeld@pinc-software.de>

Fix memory leaks

* Delete/free as appropriate

Most cases involve not freeing memory after encountering error or
unexpected situation (identified via cppcheck static analysis)

Change-Id: I90ba2fca518b00d2dfa9ec1ddbcebe1920a34b7c
Reviewed-on: https://review.haiku-os.org/c/haiku/+/1038
Reviewed-by: waddlesplash <waddlesplash@gmail.com>

userland: Pass more sizes with ioctls where we can.

No "functional" change intended.

net_server: warning on delete of address of a reference.

* fixes #12244.

net_server: Print message on fork() error.

* Also removed an extra check for the name in _Update(); _ToService()
will already return an error in that case.

net_server: Clients can now check if a service is running.

* Using the new kMsgIsServiceRunning command.

BNetworkSettings: made service stuff writable.

* And added the ability to disable a service without losing its
configuration.

BNetworkSettings: made interface settings writable.

* It also makes use of the new driver settings converter mechanism.

Moved more functionality from net_server to NetworkSettings.

* Added (still incomplete) helper classes for the settings messages.
* The net_server now uses these classes for its interfaces, and services.
* Renamed service_address to service_connection, as that better matches
what it is used for.

Moved net_server's settings to libbnetapi.

* As experimental API class BNetworkSettings.
* Added add/remove methods for interfaces, and services, too.
* Moved the conversion of the wireless networks into the settings class,
too, so that it only gives out converted ones (but accepts both
variants).

Revert "Move getifaddrs to libnetwork again."

This reverts commit 31ea76548a64b232ed10cb444bf84ca1f7e40b0f.

Adrien, please try again without clobbering the otherwise nice
BNetworkInterface API!

Conflicts:
src/kits/network/getifaddrs.cpp

Move getifaddrs to libnetwork again.

* BNetworkInterfaceAddress is moved to libnetwork. It is modified to not
use BNetworkAddress (which is in libbnetapi) and instead use sockaddr
and sockaddr_storage directly. All callers are adjusted to this.
* Some support code is shared between BNetworkInterface and
BNetworkInterfaceAddress, move it to libnetwork but in the BPrivate
namespace.

include strings.h where appriopriate...

instead or additionally to string.h, in preparation for functions move.
* moves str[n]casecmp() functions and others to strings.h.
* strings.h doesn't include string.h anymore.
* this solves #10949

Fixed up network stack and drivers for x86_64.

* Various compilation fixes.
* Fixes to the FreeBSD compatibility layer (from comparing the x86-
specific bits with the equivalent amd64 sources in FreeBSD).
* Compile all the Ethernet drivers except for sis900 and wb840, these
require a bit more work to fix (will file a ticket soon). Tested
ipro1000 and rtl81xx, no issues.

Port value in servent struct is in network byte order.
This was making net_server listening wrong service port. Telnet & FTP
works again in non stand-alone mode.
This close #7108.



git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@40241 a95241bf-73f2-0310-859d-f6bbb57e9c96

* Now uses the new network API instead of directly using socket ioctls.
This makes the code much cleaner and simpler.
* Completely untested, though (will do that next).


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@39595 a95241bf-73f2-0310-859d-f6bbb57e9c96

* Fixed warnings.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32320 a95241bf-73f2-0310-859d-f6bbb57e9c96

* Added "stand_alone" setting to the services - this way you can use the
net_server to launch services that work like a daemon theirselves (like sshd).
A service like this is also killed (via SIGTERM) when the netserver stops the
service (with sshd, you have to pass the '-D' option to make this work).
* If a boolean is given without value, it now defaults to true.
* The "launch" argument now gets its arguments properly parsed, and added, too.
* Services::_Update() did accidently set fUpdate on the wrong service, causing
services that were not changed to be stopped.
* More helpful output.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32319 a95241bf-73f2-0310-859d-f6bbb57e9c96

* Minor cleanup.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30940 a95241bf-73f2-0310-859d-f6bbb57e9c96

Fixing all headers/namespaces/name lookup problems that hinder a GCC 4.3.2
build. I sure hope that this doesn't break the build for anyone else.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@28992 a95241bf-73f2-0310-859d-f6bbb57e9c96

bonefish+axeld:
* the net_server leaked file descriptors for every opened session.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23987 a95241bf-73f2-0310-859d-f6bbb57e9c96

* Now a service is quit when it is removed from the settings file.
* When deleting the Services handler, it will now also stop all running services,
and close the command pipe (since it's only quit when the net_server quits, that
wasn't much of a problem, though).


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@21566 a95241bf-73f2-0310-859d-f6bbb57e9c96

* Now uses BPathMonitor instead of the node monitor; this fixes several issues with
changes to settings files that were ignored (for example, watching "services" did
only work if "interfaces" existed).
* On services update, Services::_Update() accidently compared the pointers of the
service objects, instead of the objects themselves.
* Implemented comparison of address changes in service objects; IOW when you change
the addresses a service should bind itself to, but leave the rest of the service
unchanged, this will now be detected as well.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@21555 a95241bf-73f2-0310-859d-f6bbb57e9c96

* _Update() was called too early in the constructor of class Service; that caused
incorrect values in fSet, and fMinSocket which also resulted in select() never
to return (in would only work for settings you changed after the net_server
was run).
* Minor cleanup.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@21551 a95241bf-73f2-0310-859d-f6bbb57e9c96

Put the new server into its own session.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@19664 a95241bf-73f2-0310-859d-f6bbb57e9c96

Actually launching services should work now:
* ServiceSocketMap now maps directly to service_address objects which now have
gotten a pointer to their owner.
* now uses listen(), and accept() for SOCK_STREAM sockets
* fMinSocket could be wrong.
* use FD_CLOEXEC on all sockets, and the communication pipe.
* removed unused service::type and service::protocol.
* if no family is specified, we now default to AF_INET.
* uses fork()/exec() to actually spawn the servers to not clobber the net_server's
input/output streams.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@19663 a95241bf-73f2-0310-859d-f6bbb57e9c96

Work-in-progress on the "services" implementation: it now reads the configuration
message, parses it, and stops/starts the services as needed - it doesn't launch
any actual servers yet, though.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@19645 a95241bf-73f2-0310-859d-f6bbb57e9c96

gcc4 build fix


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@19632 a95241bf-73f2-0310-859d-f6bbb57e9c96

* Added "services" to the network settings, and refactored the settings class a bit
to be easier extendable with more classes.
* Fixed a bug in the settings that caused updated settings to be appended to the
existing ones (missing BMessage::MakeEmpty()).
* Started services system, doesn't do anything useful yet (inetd replacement).
* Fixed the bug that caused the loopback default netmask to be incorrectly chosen;
removed the temporary fix Bruno introduced before. The same bug has been in
ifconfig where I copied the code from (but wrote it there, too) :-).


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@19609 a95241bf-73f2-0310-859d-f6bbb57e9c96

net_server: Print message on fork() error.

* Also removed an extra check for the name in _Update(); _ToService()
will already return an error in that case.

net_server: Clients can now check if a service is running.

* Using the new kMsgIsServiceRunning command.

BNetworkSettings: made service stuff writable.

* And added the ability to disable a service without losing its
configuration.

BNetworkSettings: made interface settings writable.

* It also makes use of the new driver settings converter mechanism.

Moved more functionality from net_server to NetworkSettings.

* Added (still incomplete) helper classes for the settings messages.
* The net_server now uses these classes for its interfaces, and services.
* Renamed service_address to service_connection, as that better matches
what it is used for.

Moved net_server's settings to libbnetapi.

* As experimental API class BNetworkSettings.
* Added add/remove methods for interfaces, and services, too.
* Moved the conversion of the wireless networks into the settings class,
too, so that it only gives out converted ones (but accepts both
variants).

Revert "Move getifaddrs to libnetwork again."

This reverts commit 31ea76548a64b232ed10cb444bf84ca1f7e40b0f.

Adrien, please try again without clobbering the otherwise nice
BNetworkInterface API!

Conflicts:
src/kits/network/getifaddrs.cpp

Move getifaddrs to libnetwork again.

* BNetworkInterfaceAddress is moved to libnetwork. It is modified to not
use BNetworkAddress (which is in libbnetapi) and instead use sockaddr
and sockaddr_storage directly. All callers are adjusted to this.
* Some support code is shared between BNetworkInterface and
BNetworkInterfaceAddress, move it to libnetwork but in the BPrivate
namespace.

include strings.h where appriopriate...

instead or additionally to string.h, in preparation for functions move.
* moves str[n]casecmp() functions and others to strings.h.
* strings.h doesn't include string.h anymore.
* this solves #10949

Fixed up network stack and drivers for x86_64.

* Various compilation fixes.
* Fixes to the FreeBSD compatibility layer (from comparing the x86-
specific bits with the equivalent amd64 sources in FreeBSD).
* Compile all the Ethernet drivers except for sis900 and wb840, these
require a bit more work to fix (will file a ticket soon). Tested
ipro1000 and rtl81xx, no issues.

Port value in servent struct is in network byte order.
This was making net_server listening wrong service port. Telnet & FTP
works again in non stand-alone mode.
This close #7108.



git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@40241 a95241bf-73f2-0310-859d-f6bbb57e9c96

* Now uses the new network API instead of directly using socket ioctls.
This makes the code much cleaner and simpler.
* Completely untested, though (will do that next).


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@39595 a95241bf-73f2-0310-859d-f6bbb57e9c96

* Fixed warnings.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32320 a95241bf-73f2-0310-859d-f6bbb57e9c96

* Added "stand_alone" setting to the services - this way you can use the
net_server to launch services that work like a daemon theirselves (like sshd).
A service like this is also killed (via SIGTERM) when the netserver stops the
service (with sshd, you have to pass the '-D' option to make this work).
* If a boolean is given without value, it now defaults to true.
* The "launch" argument now gets its arguments properly parsed, and added, too.
* Services::_Update() did accidently set fUpdate on the wrong service, causing
services that were not changed to be stopped.
* More helpful output.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32319 a95241bf-73f2-0310-859d-f6bbb57e9c96

* Minor cleanup.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30940 a95241bf-73f2-0310-859d-f6bbb57e9c96

Fixing all headers/namespaces/name lookup problems that hinder a GCC 4.3.2
build. I sure hope that this doesn't break the build for anyone else.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@28992 a95241bf-73f2-0310-859d-f6bbb57e9c96

bonefish+axeld:
* the net_server leaked file descriptors for every opened session.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23987 a95241bf-73f2-0310-859d-f6bbb57e9c96

* Now a service is quit when it is removed from the settings file.
* When deleting the Services handler, it will now also stop all running services,
and close the command pipe (since it's only quit when the net_server quits, that
wasn't much of a problem, though).


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@21566 a95241bf-73f2-0310-859d-f6bbb57e9c96

* Now uses BPathMonitor instead of the node monitor; this fixes several issues with
changes to settings files that were ignored (for example, watching "services" did
only work if "interfaces" existed).
* On services update, Services::_Update() accidently compared the pointers of the
service objects, instead of the objects themselves.
* Implemented comparison of address changes in service objects; IOW when you change
the addresses a service should bind itself to, but leave the rest of the service
unchanged, this will now be detected as well.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@21555 a95241bf-73f2-0310-859d-f6bbb57e9c96

* _Update() was called too early in the constructor of class Service; that caused
incorrect values in fSet, and fMinSocket which also resulted in select() never
to return (in would only work for settings you changed after the net_server
was run).
* Minor cleanup.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@21551 a95241bf-73f2-0310-859d-f6bbb57e9c96

Put the new server into its own session.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@19664 a95241bf-73f2-0310-859d-f6bbb57e9c96

Actually launching services should work now:
* ServiceSocketMap now maps directly to service_address objects which now have
gotten a pointer to their owner.
* now uses listen(), and accept() for SOCK_STREAM sockets
* fMinSocket could be wrong.
* use FD_CLOEXEC on all sockets, and the communication pipe.
* removed unused service::type and service::protocol.
* if no family is specified, we now default to AF_INET.
* uses fork()/exec() to actually spawn the servers to not clobber the net_server's
input/output streams.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@19663 a95241bf-73f2-0310-859d-f6bbb57e9c96

Work-in-progress on the "services" implementation: it now reads the configuration
message, parses it, and stops/starts the services as needed - it doesn't launch
any actual servers yet, though.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@19645 a95241bf-73f2-0310-859d-f6bbb57e9c96

gcc4 build fix


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@19632 a95241bf-73f2-0310-859d-f6bbb57e9c96

* Added "services" to the network settings, and refactored the settings class a bit
to be easier extendable with more classes.
* Fixed a bug in the settings that caused updated settings to be appended to the
existing ones (missing BMessage::MakeEmpty()).
* Started services system, doesn't do anything useful yet (inetd replacement).
* Fixed the bug that caused the loopback default netmask to be incorrectly chosen;
removed the temporary fix Bruno introduced before. The same bug has been in
ifconfig where I copied the code from (but wrote it there, too) :-).


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@19609 a95241bf-73f2-0310-859d-f6bbb57e9c96