Commit Graph
16 Commits
Author SHA1 Message Date
boykopovar f38d29fd08 refactor(sdk): add canonical headers for remaining libraries 2026-10-24 23:17:42 +03:00
xXPBSXxandClaude Sonnet 5.5 ff2b4ba4cc fix(libkernel): report EPIPE for a stream send on a connection the peer closed on Windows
sosend_generic fails a send on a socket that can send no more with EPIPE
before it looks at so_error. Winsock reports the same send as WSAECONNABORTED
or WSAECONNRESET, which reached the guest as ECONNABORTED or ECONNRESET.
A stream send now maps both to EPIPE. The send-after-close checks of
guest_tcp, guest_socket_read_write and guest_socket_pair run on Windows too.

Co-Authored-By: Claude Sonnet 5.5 <[email protected]>
2026-10-22 19:09:29 +02:00
xXPBSXxandClaude Sonnet 5.5 c02a5d8034 fix(libkernel): keep the receive-after-SHUT_RD answer on Windows only
On Linux the host keeps data that was buffered before shutdown(SHUT_RD) and
guest_socket_pair expects to read it, so the shortcut that answers 0 for any
receive after SHUT_RD broke that test there. Limit the shortcut to Windows,
where the host cannot answer it, and run the post-shutdown receives of
guest_tcp on Windows only.

Co-Authored-By: Claude Sonnet 5.5 <[email protected]>
2026-10-21 20:01:18 +02:00
xXPBSXxandClaude Sonnet 5.5 ba8b881fa6 fix(libkernel): answer a receive after SHUT_RD or with no buffer without the host
soreceive_generic returns 0 for a socket whose receive side is shut and for a
zero-length receive on a connected stream socket (ENOTCONN when it is not
connected), without waiting. Winsock fails the first with WSAESHUTDOWN, which
reached the guest as EPIPE, and blocks the second until data arrives. recv and
recvfrom now answer both themselves.

Co-Authored-By: Claude Sonnet 5.5 <[email protected]>
2026-10-21 20:01:18 +02:00
xXPBSXxandClaude Sonnet 5.5 26f861dd27 feat(libkernel): accept MSG_WAITALL in recv, recvfrom and recvmsg on a stream socket
soreceive_generic keeps a MSG_WAITALL receive on a stream socket waiting until
the buffer is full, the peer closes or an error occurs. The three calls failed
every flag but MSG_PEEK and MSG_DONTWAIT with EOPNOTSUPP. A stream socket now
passes the flag to the host receive; a datagram socket ignores it, as a record
is delivered whole.

Co-Authored-By: Claude Sonnet 5.5 <[email protected]>
2026-10-20 15:12:33 +02:00
Hussein Kadhim 0438621d74 fix(libkernel): getsockname on an unbound socket returns the zero address on Windows
FreeBSD's in_getsockaddr and in6_getsockaddr succeed for a socket that has no
local address and return the wildcard address with port 0. Winsock fails
getsockname with WSAEINVAL until the socket is bound, which the guest saw as
EINVAL. Return the zero address of the socket's family for that error.
2026-10-13 17:01:30 +13:00
Hussein Kadhim df8f3097be fix(libkernel): accepted sockets inherit the listener's non-blocking mode
FreeBSD accept() copies SS_NBIO from the listening socket (ACCEPT4_INHERIT in
kern_accept4). The accepted guest socket was always recorded as blocking:
fcntl(F_GETFL) never showed O_NONBLOCK, and the host socket kept the host's own
inheritance (kept on Windows, dropped on Linux). Apply the listener's mode to
the accepted socket.
2026-10-12 18:46:10 +00:00
Dean edddd6a17b Merge pull request #4370 from milescoviello/fix/ipv6-only-sockets
fix(net): make new IPv6 sockets IPv6-only as FreeBSD does
2026-10-13 07:21:34 +13:00
milescovielloandClaude Opus 5.5 98f6639e9e fix(net): make new IPv6 sockets IPv6-only as FreeBSD does
On Linux, socket() and sceNetSocket() create AF_INET6 sockets in the
host's default dual-stack mode (net.ipv6.bindv6only is 0), so a wildcard
IPv6 bind also claims the IPv4 port. A title that keeps an IPv4 and an
IPv6 socket on the same port fails: with 0.0.0.0:P bound, binding [::]:P
returns EADDRINUSE (48), and an IPv6 socket bound to [::] also receives
IPv4 traffic from v4-mapped peers.

FreeBSD makes every new IPv6 socket IPv6-only. sys/netinet6/in6_proto.c:

	VNET_DEFINE(int, ip6_v6only) = 1;

sys/netinet/in_pcb.c, in_pcballoc():

	if (INP_SOCKAF(so) == AF_INET6) {
		inp->inp_vflag |= INP_IPV6PROTO;
		if (V_ip6_v6only)
			inp->inp_flags |= IN6P_IPV6_V6ONLY;
	}

and sys/netinet6/in6_pcb.c, in6_pcbbind(), looks for an IPv4 owner of
the port only when the flag is clear:

	if ((inp->inp_flags & IN6P_IPV6_V6ONLY) == 0 &&
	    IN6_IS_ADDR_UNSPECIFIED(&sin6->sin6_addr)) {

socket() and sceNetSocket() now set IPV6_V6ONLY on every IPv6 socket
they create. Windows already defaults to it.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-12 08:44:27 +00:00
Aarav 6197edea2b fix(libkernel): report copied socket address lengths 2026-10-12 10:21:17 +05:30
Hussein Kadhim 87e9c70157 fix(libkernel): recvfrom on a stream socket reports no sender address
FreeBSD's soreceive leaves the sender unset for a stream socket, so recvfrom
returns an address length of 0 (kern_recvit). The guest call formatted the empty
host address as a 28-byte AF_INET6 sockaddr and reported length 28. Report
length 0 when the host returns no AF_INET or AF_INET6 sender, as recvmsg does.
2026-10-11 19:48:51 +00:00
Zurisen 2356aba27b fix(net): report EPIPE for a send after shutdown on Windows
Winsock fails send, sendto and sendmsg on a socket shut down for writing
with WSAESHUTDOWN, which neither Windows error map handled, so libkernel
and libSceNet reported EIO. FreeBSD reports EPIPE there, and the Linux
maps already translate the host's EPIPE to 32. Map WSAESHUTDOWN to 32
in both.
2026-10-10 02:02:29 +02:00
065280d58d fix(libkernel): map EPIPE on Linux and test send with MSG_NOSIGNAL on a closed peer
Case and test by hikazey in the PR thread.

Co-authored-by: hikazey <[email protected]>
Co-Authored-By: Claude <[email protected]>
2026-10-08 12:36:58 +02:00
Scott SchusterandClaude Opus 5.5 a2f9a50985 fix(libkernel): accept MSG_NOSIGNAL in send and sendto
send, sendto and sendmsg rejected every flag with EOPNOTSUPP, so callers
that pass FreeBSD's MSG_NOSIGNAL (0x20000) could not send at all. Accept
it: Windows sockets never raise SIGPIPE, and on Linux it maps to the
native MSG_NOSIGNAL. Other flags still fail with EOPNOTSUPP.

DOOM (PPSA99666) sends its upload server replies with MSG_NOSIGNAL and
closed every connection without an answer.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-08 12:36:58 +02:00
Jakub Krasuski cefdea7d14 fix(libkernel): report EINVAL, EACCES, ENOBUFS, ETIMEDOUT and EINTR from Linux sockets
The Linux branch of NativeError lacked five codes the Windows branch
already translates, so they fell through to EIO. bind on a bound socket
and accept on a socket that is not listening set errno to 5 on Linux
and to 22 on Windows.

Assisted-by: Claude Opus 5.5
2026-10-05 21:41:12 +02:00
Praetorian0420 e3eeac4a81 Implement network control and socket functionality improvements
- Enhanced the network control library to check for host addresses and update connection states accordingly.
- Added support for socket operations including connect, listen, accept, send, and recv in the UDP and TCP socket implementations.
- Introduced error handling for various socket operations, ensuring proper feedback for unsupported actions.
- Created comprehensive tests for both TCP and UDP socket functionalities, validating the expected behavior of the network stack.
- Refactored existing code to improve clarity and maintainability, including the removal of unused functions and the addition of necessary error checks.
2026-09-30 23:29:01 -05:00