[EagleBearerYS]

What

sceNetSetsockopt and sceNetGetsockopt answered EOPNOTSUPP for 0x1105 and 0x1106, the SCE send and receive timeouts (an int in microseconds); only FreeBSD's 0x1005 and 0x1006 were known. A title that set a receive timeout this way lost it, and a later sceNetRecv on an idle socket blocked for good:

	main	This branch
sceNetSetsockopt(s, 0xffff, 0x1106, 200000)	0x8041012d (errno 45)	0
sceNetGetsockopt(s, 0xffff, 0x1106)	0x8041012d (errno 45)	0, value 200000
sceNetSetsockopt / sceNetGetsockopt with 0x1105	errno 45 / errno 45	0 / 200000
sceNetRecv on an empty UDP socket afterwards	still blocked after 3 s	0x80410123 (EAGAIN) after 200 ms

Both options now set and read the timeouts that 0x1005 and 0x1006 already do; 0x1005 and 0x1006 are unchanged.

Values: shadPS4 net_types.h (ORBIS_NET_SO_SNDTIMEO/RCVTIMEO, read as microseconds in net.cpp), SharpProspero Socket.cs (SoSndTimeo/SoRcvTimeo, "in microseconds") and KytyPS5 network.cpp (0x1105 to SO_SNDTIMEO). Not measured on a console. PPSA34501 sets 0x1105 = 10000 on its UDP game sockets (seen in #2745, closed by the intake pause). #4371 changes the same function for libkernel's FreeBSD-numbered path; the two merge cleanly.

Tested

Windows 11 (10.0.26300), WinLibs MinGW-w64 GCC 15.2.0 r7 (UCRT), Release, -DBUILD_TESTING=ON, on main at 2a5eca1:

• With main's Export.cpp, the new CheckMicrosecondTimeouts in guest_sce_net aborts (exit 0xc0000409); guest_sce_net passes on main itself, IPv6 checks included, so the abort is the new check. With the change it passes, 20 runs in a row.
• All 128 tests registered in core/libs/prx/libc/CMakeLists.txt and core/libs/prx/libkernel/CMakeLists.txt that build on Windows pass, and so do the network tests (guest_datagram_recv, guest_datagram_recvfrom, guest_datagram_sceNetRecv, guest_datagram_sceNetRecvfrom, guest_udp, guest_resolver, guest_posix_filesystem).

Linux (Ubuntu 24.04, GCC 13.3, Release), base 33ae282: the new CheckMicrosecondTimeouts in guest_sce_net aborts on main at sceNetSetsockopt(..., 0x1106, ...) == 0 (errno 45) and passes with the change, 20/20; the libSceNet, libc and libkernel tests are unchanged. Again on WSL2 (Ubuntu 24.04, GCC 13.3.0, Release with CI's configure flags), merged into main at c952c2b, IPv6 checks included: it aborts with main's Export.cpp and passes with the change, 20 runs in a row.

tools/check_conventions.py --base origin/main passes.

Checklist

• Based on current main; no other open PR implements the same functions
• One topic per PR; follow-ups go in a new PR
• No comments in code except technical debt
• Unimplemented paths throw (NotImplemented_nid_no_patch); silent stubs are listed in TechnicalDebt
• No notes, investigation .md files or images added to the repository (attach them to this PR)
• New third-party code is a submodule built from source
• Generic behaviour, not specific to one title
• Depends on: none
• AI-assisted: yes

[adriadam10]

Checked on current main: guest_sce_net and the libSceNet/libs ctest scope pass; the new CheckMicrosecondTimeouts fails with main's Export.cpp. 0x1105/0x1106 map to the same int-microsecond timeouts as 0x1005/0x1006 in set and get, matching shadPS4, SharpProspero and KytyPS5. Merging once the full Build is green.