summaryrefslogtreecommitdiff
path: root/lib
diff options
context:
space:
mode:
authorPhilip Guenther <guenther@cvs.openbsd.org>2015-08-25 19:58:05 +0000
committerPhilip Guenther <guenther@cvs.openbsd.org>2015-08-25 19:58:05 +0000
commit585254a8e1745efcb2cd0b199ba4f8754571bae1 (patch)
tree606314ff8aa401f29dbf20776e60cf9fe6132a63 /lib
parent3918f7593b9c68116dbdd13da3af5aa71536f945 (diff)
Document kbind(2)
Diffstat (limited to 'lib')
-rw-r--r--lib/libc/sys/Makefile.inc4
-rw-r--r--lib/libc/sys/kbind.2120
2 files changed, 122 insertions, 2 deletions
diff --git a/lib/libc/sys/Makefile.inc b/lib/libc/sys/Makefile.inc
index fe791fd83ff..d1513a50240 100644
--- a/lib/libc/sys/Makefile.inc
+++ b/lib/libc/sys/Makefile.inc
@@ -1,4 +1,4 @@
-# $OpenBSD: Makefile.inc,v 1.126 2015/07/19 02:37:29 deraadt Exp $
+# $OpenBSD: Makefile.inc,v 1.127 2015/08/25 19:58:04 guenther Exp $
# $NetBSD: Makefile.inc,v 1.35 1995/10/16 23:49:07 jtc Exp $
# @(#)Makefile.inc 8.1 (Berkeley) 6/17/93
@@ -206,7 +206,7 @@ MAN+= __get_tcb.2 __thrsigdivert.2 __thrsleep.2 _exit.2 accept.2 \
getpeername.2 getpgrp.2 getpid.2 getpriority.2 getrlimit.2 \
getrtable.2 getrusage.2 getsid.2 getsockname.2 getsockopt.2 \
getthrid.2 gettimeofday.2 getuid.2 intro.2 ioctl.2 issetugid.2 \
- kill.2 kqueue.2 ktrace.2 link.2 listen.2 lseek.2 madvise.2 \
+ kbind.2 kill.2 kqueue.2 ktrace.2 link.2 listen.2 lseek.2 madvise.2 \
mincore.2 minherit.2 mkdir.2 mkfifo.2 mknod.2 mlock.2 \
mlockall.2 mmap.2 mount.2 mprotect.2 mquery.2 msgctl.2 \
msgget.2 msgrcv.2 msgsnd.2 msync.2 munmap.2 nanosleep.2 \
diff --git a/lib/libc/sys/kbind.2 b/lib/libc/sys/kbind.2
new file mode 100644
index 00000000000..4a9fe239d6a
--- /dev/null
+++ b/lib/libc/sys/kbind.2
@@ -0,0 +1,120 @@
+.\" $OpenBSD: kbind.2,v 1.1 2015/08/25 19:58:04 guenther Exp $
+.\"
+.\" Copyright (c) 2015 Philip Guenther <guenther@openbsd.org>
+.\"
+.\" Permission to use, copy, modify, and distribute this software for any
+.\" purpose with or without fee is hereby granted, provided that the above
+.\" copyright notice and this permission notice appear in all copies.
+.\"
+.\" THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+.\" WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+.\" MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+.\" ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+.\" WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+.\" ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+.\" OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+.\"
+.Dd $Mdocdate: August 25 2015 $
+.Dt KBIND 2
+.Os
+.Sh NAME
+.Nm kbind
+.Nd update protected memory for lazy-binding
+.Sh SYNOPSIS
+.In sys/unistd.h
+.Bd -literal
+struct __kbind {
+ void *kb_addr;
+ size_t kb_size;
+};
+#define KBIND_BLOCK_MAX 2 /* powerpc, sparc, and sparc64 need 2 blocks */
+#define KBIND_DATA_MAX 24 /* sparc64 needs 6, four-byte words */
+.Ed
+.Pp
+.Ft int
+.Fn kbind "const struct __kbind *param" "size_t psize" "int64_t cookie"
+.Sh DESCRIPTION
+The
+.Nm
+syscall updates the contents of one or more blocks of the process's memory
+with the supplied replacement data.
+This is used to efficiently and securely perform lazy-binding.
+.Pp
+.Fa param
+points to an array of
+.Vt __kbind
+structures giving the addresses and lengths to update.
+The last
+.Vt __kbind
+structure is immediately followed in the same order by the source
+data for the blocks to copy.
+.Fa psize
+specifies the total length of the parameters: both the
+.Vt __kbind
+structures and the associated source data.
+There may be at most
+.Dv KBIND_BLOCK_MAX
+.Vt __kbind
+structures and each block may be at most
+.Dv KBIND_DATA_MAX
+bytes in size.
+.Pp
+.Nm
+updates memory
+.Do
+as if
+.Dc
+the thread temporarily made the memory writable with
+.Xr mprotect 2 .
+If the process does not have write access to the underlying memory object,
+.Nm
+will fail without making the requested change.
+.Pp
+.Nm
+is currently intended for use by
+.Xr ld.so 1
+only.
+It is therefore not provided as a function and two security checks
+are performed to bind it (pun intended) to its use in
+.Xr ld.so 1 :
+the first time
+.Nm
+is used, the kernel records both the text address of the call and
+the value of the
+.Fa cookie
+argument.
+If those values differ in a latter
+.Nm
+call, then the process is killed.
+.Sh RETURN VALUES
+.Rv -std
+.Sh ERRORS
+.Fn kbind
+will fail if:
+.Bl -tag -width Er
+.It Bq Er ENOMEM
+Cannot allocate memory when updating a copy-on-write page.
+.It Bq Er EINVAL
+The parameters are inconsistent or exceed
+.Nm
+limits.
+.It Bq Er EFAULT
+Part of the structures or additional data pointed to by
+.Fa param
+is outside the process's allocated address space,
+or the underlying memory object is not writable.
+.El
+.Sh SEE ALSO
+.Xr ld.so 1 ,
+.Xr mprotect 2
+.Sh STANDARDS
+The
+.Nm
+syscall is specific to the
+.Ox
+dynamic linker and should not be used in portable applications.
+.Sh HISTORY
+The
+.Nm
+syscall appeared in
+.Ox 5.8 .