9e6d8dd4e3
sendmsg/recvmsg with SCM_RIGHTS. Item 1 of docs/plan/foundations.md and
mechanism 1 of docs/plan/hybrid.md, which are the same primitive seen from two
directions: Wayland's wl_shm_create_pool passes a file descriptor with no
integer-shaped alternative, and handing a driver process the device it is
allowed to touch is the same act. Both plans stopped here.
Linux's structures byte for byte -- msghdr, cmsghdr, the CMSG_ macros -- because
the programs that need this are not ours. libwayland and toybox both call
recvmsg with Linux's headers and would not notice a friendlier door.
The delivery rule is the whole design. A batch of descriptors is pinned to the
stream position of the first byte of the message that carried it, and that has
two halves: a reader arriving at that position is handed them, and a reader that
has not reached it is stopped SHORT of it. Leave out the second half and a
program asking for 4096 bytes swallows three messages and is handed one
message's descriptors, with nothing in the result saying which. socktest checks
exactly that: two one-byte messages, an eight-byte read, one byte back.
The kernel holds a reference to every descriptor in flight, so a sender may
close its own copy the moment sendmsg returns. Giving that reference back is the
part that leaks if it is not written down, and there are three ways a message
ends:
- received with room installed in the receiver's table
- received without room, or by
plain read(2) closed; MSG_CTRUNC for the first case
- never received, socket closed closed by sockclose
The middle row looks like a choice and is not: holding kernel references for a
receiver that has proved it is not looking is a leak with extra steps.
No garbage collector. A socket descriptor sent over that same socket and never
received holds a reference to the socket that holds the reference to it, and
neither close reaches zero. Linux carries several hundred lines for this.
Nothing on either plan passes a socket over a socket -- Wayland passes memory,
the driver work passes devices -- so it is a bounded leak that is documented
rather than a silent one.
Fourteen checks in cmd/socktest, and every one moves data through the descriptor
rather than inspecting it. The descriptor passed is always a pipe end, because a
pipe announces its own lifetime: a read of a pipe whose writers are all gone
returns 0, which turns "was the kernel's reference released" -- otherwise
invisible from userspace -- into a value a test can compare.
Two bugs fell out, neither in the new code:
- sockpair() never set the socket's domain, so every socketpair() socket
claimed domain 0. Invisible while the data paths asked only whether a socket
was AF_INET or AF_NETLINK, since zero answers "neither" correctly.
SCM_RIGHTS asks "is this AF_UNIX", and zero is wrong for that one.
- pipe_poll() reports POLLHUP but not POLLIN at end of file, so a non-blocking
read of a pipe whose writers had all gone returned EAGAIN forever instead of
0 -- a program that polls and reads spins waiting for an end that already
happened. sockreadable() always had the equivalent condition; the pipe path
did not. Found because the test helper for "is this descriptor closed" is a
non-blocking read of a pipe.
projects/schcom/README.md said SCM_RIGHTS was done for a long time while nothing
in the kernel implemented it, and an entire GUI plan was drawn on that line.
It is true now, which is not the same as it having been right, and the README
says so.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VQ7Qx8E1q1M1bbCq8QLHS2
300 lines
10 KiB
SQL
300 lines
10 KiB
SQL
// Segregated Kernel -- central syscall table. ONE line per syscall.
|
|
//
|
|
// Format: SYSCALL(nr, linux_nr, name)
|
|
// nr SchOS native syscall number
|
|
// linux_nr Linux x86_64 syscall number, or NO_LINUX if none
|
|
// name implemented as sys_<name>() in kernel/sys/sys_*.c
|
|
//
|
|
// To add a syscall:
|
|
// 1. Add ONE line here.
|
|
// 2. Write the body with SYSCALL_DEFINE(<name>) in a kernel/sys/sys_*.c.
|
|
//
|
|
// Everything else is generated from this file: the SYS_* constants
|
|
// (include/syscall.h), the dispatch table (kernel/sys/syscall.c), and the
|
|
// Linux ABI translation map (kernel/sys/linux_abi.c). The Linux number used
|
|
// to live in a second hand-maintained table in linux_abi.c; it does not any
|
|
// more, so a syscall can no longer be half-registered.
|
|
//
|
|
// A syscall reachable under more than one Linux number (e.g. stat/newstat)
|
|
// gets extra SYSCALL_ALIAS(linux_nr, name) lines.
|
|
|
|
|
|
// --- process ---
|
|
SYSCALL( 1, 57, fork)
|
|
SYSCALL( 2, 60, exit)
|
|
SYSCALL( 3, NO_LINUX, wait)
|
|
SYSCALL( 6, 62, kill)
|
|
SYSCALL( 7, 59, exec)
|
|
SYSCALL( 11, 39, getpid)
|
|
SYSCALL( 13, NO_LINUX, sleep)
|
|
SYSCALL( 14, NO_LINUX, uptime)
|
|
SYSCALL( 29, NO_LINUX, waitpid)
|
|
SYSCALL( 30, 112, setsid)
|
|
SYSCALL( 31, NO_LINUX, reboot)
|
|
SYSCALL( 32, NO_LINUX, shutdown)
|
|
SYSCALL( 66, 56, clone)
|
|
SYSCALL( 67, 231, exit_group)
|
|
SYSCALL( 68, 61, wait4)
|
|
SYSCALL( 83, 105, setuid)
|
|
SYSCALL( 84, 106, setgid)
|
|
SYSCALL( 85, 124, getsid)
|
|
SYSCALL( 89, 110, getppid)
|
|
SYSCALL( 91, 121, getpgid)
|
|
SYSCALL( 94, 186, gettid)
|
|
SYSCALL( 92, 109, setpgid)
|
|
SYSCALL( 93, 111, getpgrp)
|
|
|
|
// --- identity ---
|
|
SYSCALL( 47, 102, getuid)
|
|
SYSCALL( 48, 107, geteuid)
|
|
SYSCALL( 69, 104, getgid)
|
|
SYSCALL( 70, 108, getegid)
|
|
SYSCALL( 71, 95, umask)
|
|
|
|
// --- memory ---
|
|
SYSCALL( 12, NO_LINUX, sbrk)
|
|
SYSCALL( 35, 9, mmap)
|
|
SYSCALL( 36, 11, munmap)
|
|
// Shared memory. These carried NO_LINUX while they were ENOSYS stubs, which
|
|
// meant a musl-linked program calling shmget() never reached them at all --
|
|
// it got "bad syscall" from the ABI translation rather than an answer. Now
|
|
// that they do something, they need the numbers programs actually use.
|
|
SYSCALL( 37, 29, shmget)
|
|
SYSCALL( 38, 30, shmat)
|
|
SYSCALL( 39, 67, shmdt)
|
|
SYSCALL( 59, 10, mprotect)
|
|
SYSCALL( 74, 12, brk)
|
|
|
|
// --- filesystem ---
|
|
SYSCALL( 4, 22, pipe)
|
|
SYSCALL( 5, 0, read)
|
|
SYSCALL( 8, 5, fstat)
|
|
SYSCALL( 9, 80, chdir)
|
|
SYSCALL( 10, 32, dup)
|
|
SYSCALL( 15, 2, open)
|
|
SYSCALL( 16, 1, write)
|
|
SYSCALL( 17, 133, mknod)
|
|
SYSCALL( 18, 87, unlink)
|
|
SYSCALL( 19, 86, link)
|
|
SYSCALL( 20, 83, mkdir)
|
|
SYSCALL( 21, 3, close)
|
|
SYSCALL( 24, 8, lseek)
|
|
SYSCALL( 33, 33, dup2)
|
|
SYSCALL( 34, 72, fcntl)
|
|
// `delete` -- the recycle bin. It hides a file instead of removing it, so the
|
|
// blocks are never freed and a program that deletes a temp file gets no space
|
|
// back. Inherited from the third-party filesystem in kernel/fs/fs.c and
|
|
// orphaned by the decision to keep only Linux syscall numbering, so nothing
|
|
// reaches it. Its effect on read and write has been removed; the entry stays
|
|
// only because sys_delete() is still compiled, and both die with fs.c.
|
|
SYSCALL( 51, NO_LINUX, delete)
|
|
SYSCALL( 52, NO_LINUX, show)
|
|
SYSCALL( 53, NO_LINUX, showstat)
|
|
SYSCALL( 54, 77, ftruncate)
|
|
SYSCALL( 55, 76, truncate)
|
|
SYSCALL( 56, 82, rename)
|
|
SYSCALL( 57, 88, symlink)
|
|
SYSCALL( 58, 89, readlink)
|
|
SYSCALL( 61, 79, getcwd)
|
|
SYSCALL( 73, 19, readv)
|
|
SYSCALL( 72, 20, writev)
|
|
SYSCALL( 76, 4, stat)
|
|
SYSCALL( 77, 6, lstat)
|
|
SYSCALL( 78, 21, access)
|
|
SYSCALL( 79, 257, openat)
|
|
SYSCALL( 80, 81, fchdir)
|
|
SYSCALL( 86, 292, dup3)
|
|
SYSCALL( 87, 293, pipe2)
|
|
|
|
// --- device / I/O ---
|
|
SYSCALL( 22, NO_LINUX, vga_sync)
|
|
SYSCALL( 23, NO_LINUX, getmouse)
|
|
SYSCALL( 45, 23, select)
|
|
SYSCALL( 46, 7, poll)
|
|
SYSCALL( 60, 16, ioctl)
|
|
|
|
// --- sockets ---
|
|
//
|
|
// These carried NO_LINUX while they were ENOSYS stubs, which meant a
|
|
// musl-linked program calling socket() never reached them at all -- the ABI
|
|
// translation answered "bad syscall" before the stub could answer ENOSYS.
|
|
// Now that they do something they need the numbers programs actually use.
|
|
//
|
|
// `shutdown` is deliberately not the name here: SchOS already has a native
|
|
// shutdown() that powers the machine off (syscall 32), and two syscalls
|
|
// cannot share a sys_ function name. Linux 48 maps to sockshutdown instead.
|
|
SYSCALL( 40, 41, socket)
|
|
SYSCALL( 41, 49, bind)
|
|
SYSCALL( 42, 50, listen)
|
|
SYSCALL( 43, 43, accept)
|
|
SYSCALL( 44, 42, connect)
|
|
SYSCALL(187, 44, sendto)
|
|
SYSCALL(188, 45, recvfrom)
|
|
SYSCALL(189, 48, sockshutdown)
|
|
SYSCALL(190, 53, socketpair)
|
|
SYSCALL(191, 54, setsockopt)
|
|
SYSCALL(192, 55, getsockopt)
|
|
// accept4() differs from accept() only by a flags argument carrying
|
|
// SOCK_CLOEXEC/SOCK_NONBLOCK. It was an alias for accept, which silently
|
|
// dropped both -- and it cannot be an alias now that they are honoured: accept
|
|
// takes three arguments, so reading a fourth register from an accept() caller
|
|
// reads whatever it last left there and invents a flag out of stack litter.
|
|
SYSCALL(198, 288, accept4)
|
|
// sendmsg/recvmsg exist for the one thing only they can do: carry a descriptor
|
|
// from one process to another (SCM_RIGHTS). docs/plan/foundations.md item 1.
|
|
SYSCALL(199, 46, sendmsg)
|
|
SYSCALL(200, 47, recvmsg)
|
|
|
|
// --- time ---
|
|
SYSCALL( 28, 37, alarm)
|
|
SYSCALL( 49, 96, gettimeofday)
|
|
SYSCALL( 50, 35, nanosleep)
|
|
SYSCALL( 88, 228, clock_gettime)
|
|
|
|
// --- signals ---
|
|
SYSCALL( 25, NO_LINUX, sigfillset)
|
|
SYSCALL( 26, 14, sigprocmask)
|
|
SYSCALL( 27, NO_LINUX, sigwait)
|
|
SYSCALL( 90, 13, rt_sigaction)
|
|
|
|
// --- POSIX / Linux-ABI compat ---
|
|
SYSCALL( 62, 158, arch_prctl)
|
|
SYSCALL( 95, 63, uname)
|
|
SYSCALL( 96, 217, getdents64)
|
|
SYSCALL( 63, 218, set_tid_address)
|
|
SYSCALL( 64, 273, set_robust_list)
|
|
SYSCALL( 65, 202, futex)
|
|
SYSCALL( 75, 302, prlimit)
|
|
SYSCALL( 81, NO_LINUX, enable_linux_abi)
|
|
SYSCALL( 82, 97, getrlimit)
|
|
|
|
// --- signal extensions ---
|
|
SYSCALL( 97, 15, rt_sigreturn)
|
|
SYSCALL( 98, 131, sigaltstack)
|
|
SYSCALL( 99, 127, rt_sigpending)
|
|
SYSCALL(100, 128, rt_sigtimedwait)
|
|
SYSCALL(101, 130, rt_sigsuspend)
|
|
|
|
// --- process extensions ---
|
|
SYSCALL(102, 157, prctl)
|
|
SYSCALL(103, 234, tgkill)
|
|
SYSCALL(104, 200, tkill)
|
|
SYSCALL(105, 34, pause)
|
|
SYSCALL(106, NO_LINUX, getitimer)
|
|
SYSCALL(107, NO_LINUX, setitimer)
|
|
SYSCALL(108, 98, getrusage)
|
|
SYSCALL(109, 100, times)
|
|
SYSCALL(110, 99, sysinfo)
|
|
SYSCALL(111, 24, sched_yield)
|
|
SYSCALL(112, 204, sched_getaffinity)
|
|
SYSCALL(113, 203, sched_setaffinity)
|
|
SYSCALL(114, 247, waitid)
|
|
|
|
// --- identity extensions ---
|
|
SYSCALL(115, 113, setreuid)
|
|
SYSCALL(116, 114, setregid)
|
|
SYSCALL(117, 118, getresuid)
|
|
SYSCALL(118, 120, getresgid)
|
|
SYSCALL(119, 117, setresuid)
|
|
SYSCALL(120, 119, setresgid)
|
|
SYSCALL(121, 115, getgroups)
|
|
SYSCALL(122, 116, setgroups)
|
|
SYSCALL(123, 122, setfsuid)
|
|
SYSCALL(124, 123, setfsgid)
|
|
|
|
// --- filesystem extensions ---
|
|
SYSCALL(125, 90, chmod)
|
|
SYSCALL(126, 91, fchmod)
|
|
SYSCALL(127, 92, chown)
|
|
SYSCALL(128, 93, fchown)
|
|
SYSCALL(129, 94, lchown)
|
|
SYSCALL(130, 84, rmdir)
|
|
SYSCALL(131, NO_LINUX, fsync)
|
|
SYSCALL(132, NO_LINUX, fdatasync)
|
|
SYSCALL(133, 162, sync)
|
|
SYSCALL(134, 17, pread)
|
|
SYSCALL(135, 18, pwrite)
|
|
SYSCALL(148, 132, utime)
|
|
SYSCALL(149, 235, utimes)
|
|
SYSCALL(150, 261, utimensat)
|
|
SYSCALL(151, 187, sendfile)
|
|
SYSCALL(152, NO_LINUX, preadv2)
|
|
SYSCALL(153, NO_LINUX, pwritev2)
|
|
SYSCALL(154, 285, fallocate)
|
|
SYSCALL(155, 137, statfs)
|
|
SYSCALL(156, 138, fstatfs)
|
|
SYSCALL(157, 161, chroot)
|
|
|
|
// --- attributes, and the index over them (SchFS phase 5) ---
|
|
//
|
|
// Linux's numbers for the xattr calls, so a program that already knows how to
|
|
// set an extended attribute needs no changes. schfs_query has no Linux
|
|
// equivalent -- it is the one genuinely new call -- so it is native-only.
|
|
SYSCALL(193, 188, setxattr)
|
|
SYSCALL(194, 191, getxattr)
|
|
SYSCALL(195, 197, removexattr)
|
|
SYSCALL(196, NO_LINUX, schfs_index)
|
|
SYSCALL(197, NO_LINUX, schfs_query)
|
|
|
|
// --- memory extensions ---
|
|
SYSCALL(136, 28, madvise)
|
|
SYSCALL(137, NO_LINUX, mremap)
|
|
SYSCALL(138, NO_LINUX, msync)
|
|
SYSCALL(139, NO_LINUX, mincore)
|
|
SYSCALL(140, 149, mlock)
|
|
SYSCALL(141, 150, munlock)
|
|
SYSCALL(142, 151, mlockall)
|
|
SYSCALL(143, 152, munlockall)
|
|
SYSCALL(144, 221, fadvise64)
|
|
SYSCALL(145, 160, setrlimit)
|
|
SYSCALL(146, NO_LINUX, getpriority)
|
|
SYSCALL(147, NO_LINUX, setpriority)
|
|
|
|
// --- at-syscalls ---
|
|
SYSCALL(158, 258, mkdirat)
|
|
SYSCALL(159, 259, mknodat)
|
|
SYSCALL(160, 263, unlinkat)
|
|
SYSCALL(161, 264, renameat)
|
|
SYSCALL(162, 265, linkat)
|
|
SYSCALL(163, 266, symlinkat)
|
|
SYSCALL(164, 267, readlinkat)
|
|
SYSCALL(165, 268, fchmodat)
|
|
SYSCALL(166, 269, faccessat)
|
|
SYSCALL(167, 262, fstatat)
|
|
SYSCALL(168, 260, fchownat)
|
|
SYSCALL(169, 316, renameat2)
|
|
|
|
// --- time extensions ---
|
|
SYSCALL(170, 201, time)
|
|
SYSCALL(171, 229, clock_getres)
|
|
SYSCALL(172, 230, clock_nanosleep)
|
|
|
|
// --- misc ---
|
|
SYSCALL(173, 103, syslog)
|
|
SYSCALL(174, 318, getrandom)
|
|
SYSCALL(175, 270, pselect6)
|
|
SYSCALL(176, 271, ppoll)
|
|
|
|
// --- epoll / eventfd / signalfd / timerfd ---
|
|
SYSCALL(177, 291, epoll_create1)
|
|
SYSCALL(178, 233, epoll_ctl)
|
|
SYSCALL(179, 232, epoll_wait)
|
|
SYSCALL(180, 282, signalfd4)
|
|
SYSCALL(181, 284, eventfd2)
|
|
SYSCALL(182, 283, timerfd_create)
|
|
SYSCALL(183, 286, timerfd_settime)
|
|
SYSCALL(184, 287, timerfd_gettime)
|
|
|
|
// --- mounts ---
|
|
SYSCALL(185, 165, mount)
|
|
SYSCALL(186, 166, umount2)
|
|
|
|
|
|
// --- additional Linux numbers routed to an existing handler ---
|
|
SYSCALL_ALIAS( 58, fork)
|
|
SYSCALL_ALIAS( 280, utimensat)
|
|
SYSCALL_ALIAS( 281, epoll_wait)
|
|
SYSCALL_ALIAS( 289, signalfd4)
|
|
SYSCALL_ALIAS( 290, eventfd2)
|
|
SYSCALL_ALIAS( 295, readv)
|
|
SYSCALL_ALIAS( 296, writev)
|