sysvipc: Set ipc_perm mode as mode_t (BZ#18231)azanella/bz18231

This patch sets the mode field in ipc_perm as mode_t for all architectures,
as POSIX specification [1].  The changes required are as follow:

  1. It moves the ipc_perm definition out of ipc.h to its own header
     ipc_perm.h.  It also allows consolidate the IPC_* definition on
     only one header.

  2. The generic implementation follow the kernel ipc64_perm size so the
     syscall can be made directly without temporary buffer copy.  However,
     since glibc defines the MODE field as mode_t, it omits the __PAD1 field
     (since glibc does not export mode_t as 16-bit for any architecture).

     It is a two-fold improvement:

     2.1. New implementation which follow Linux UAPI will not need to
	  provide an arch-specific ipc-perm.h header neither wrongly
          use the wrong 16-bit definition from previous default ipc.h
	  (as csky did).

     2.1. It allows consolidate ipc_perm definition for architectures that
          already provide mode_t as 32-bit.

  3. All kernel ABIs for the supported architectures already provides the
     expected padding for mode type extension to 32-bit.  However, some
     architectures the padding has the wrong placement, so it requires
     the ipc control routines (msgctl, semctl, and shmctl) to adjust the
     mode field accordingly.  Currently they are armeb, microblaze, m68k,
     s390, and sheb.

     A new assume is added, __ASSUME_SYSVIPC_SUPPORT_MODE32, which the
     required architecture undefine.

  4. For the architecture that undefined __ASSUME_SYSVIPC_SUPPORT_MODE32,
     it also requires compat symbols that do not adjust the mode field.

Checked on arm-linux-gnueabihf, aarch64-linux-gnu, powerpc64le-linux-gnu,
and x86_64-linux-gnu. I also checked the sysvipc tests on hppa-linux-gnu,
sh4-linux-gnu, s390x-linux-gnu, and s390-linux-gnu.

	[BZ #18231]
	* sysdeps/unix/sysv/linux/Makefile (sysdep_headers): Add
	bits/ipc-perm.h.
	* sysdeps/unix/sysv/linux/aarch64/bits/ipc.h: Remove file.
	* sysdeps/unix/sysv/linux/alpha/bits/ipc.h: Likewise.
	* sysdeps/unix/sysv/linux/hppa/bits/ipc.h: Likewise.
	* sysdeps/unix/sysv/linux/ia64/bits/ipc.h: Likewise.
	* sysdeps/unix/sysv/linux/mips/bits/ipc.h: Likewise.
	* sysdeps/unix/sysv/linux/powerpc/bits/ipc.h: Likewise.
	* sysdeps/unix/sysv/linux/s390/bits/ipc.h: Likewise.
	* sysdeps/unix/sysv/linux/sparc/bits/ipc.h: Likewise.
	* sysdeps/unix/sysv/linux/arm/kernel-features.h
	[__BYTE_ORDER == __BIG_ENDIAN] (__ASSUME_SYSVIPC_SUPPORT_MODE32):
	Undefine.
	* sysdeps/sysv/linux/microblaze/kernel-features.h: Likewise.
	* sysdeps/unix/sysv/linux/s390/kernel-features.h
	[!__s390x__] (__ASSUME_SYSVIPC_SUPPORT_MODE32): Define.
	* sysdeps/unix/sysv/linux/sh/kernel-features.h
	(__ASSUME_SYSVIPC_SUPPORT_MODE32): Define.
	* sysdeps/unix/sysv/linux/m68k/kernel-features.h: Likewise.
	* sysdeps/unix/sysv/linux/bits/ipc-perm.h: New file.
	* sysdeps/unix/sysv/linux/powerpc/bits/ipc-perm.h: Likewise.
	* sysdeps/unix/sysv/linux/sparc/bits/ipc-perm.h: Likewise.
	* sysdeps/unix/sysv/linux/bits/ipc.h (ipc_perm): Move to
	bits/ipc-perm.h.
	* sysdeps/unix/sysv/linux/hppa/bits/ipc-perm.h: New file.
	* sysdeps/unix/sysv/linux/kernel-features.h
	(__ASSUME_SYSVIPC_SUPPORT_MODE32): Define.
	* sysdeps/unix/sysv/linux/msgctl.c (DEFAULT_VERSION): Define as
	2.30 if __ASSUME_SYSVIPC_SUPPORT_MODE32 is not defined.
	(msgctl_syscall, __msgctl_mode16): New symbol.
	(__new_msgctl): Handle if __ASSUME_SYSVIPC_SUPPORT_MODE32 is not
	defined.
	* sysdeps/unix/sysv/linux/semctl.c: Likewise.
	* sysdeps/unix/sysv/linux/shmctl.c: Likewise.
	* sysdeps/unix/sysv/linux/arm/be/libc.abilist (GLIBC_2.30): Add
	msgctl, semctl, and shmctl.
	* sysdeps/sysv/linux/microblaze/be/libc.abilist: Likewise.
	* conform/data/sys/ipc.h-data: Only xfail {struct ipc_perm} mode_t
	mode for Hurd.
	* sysdeps/unix/sysv/linux/m68k/coldfire/libc.abilist: Likewise.
	* sysdeps/unix/sysv/linux/m68k/m680x0/libc.abilist: Likewise.
	* sysdeps/unix/sysv/linux/s390/s390-32/libc.abilist: Likewise.
	* sysdeps/unix/sysv/linux/sh/be/libc.abilist: Likewise.

[1] http://pubs.opengroup.org/onlinepubs/9699919799/basedefs/sys_ipc.h.html

1 files changed, 8 insertions, 0 deletions

diff --git a/sysdeps/unix/sysv/linux/kernel-features.h b/sysdeps/unix/sysv/linux/kernel-features.h
index 1518bb5228..91152158f1 100644
--- a/sysdeps/unix/sysv/linux/kernel-features.h
+++ b/sysdeps/unix/sysv/linux/kernel-features.h
@@ -81,6 +81,14 @@
    either support ipc syscall and/or all the ipc correspondent syscalls.  */
 #define __ASSUME_DIRECT_SYSVIPC_SYSCALLS	1
 
+/* All supported architectures reserve a 32-bit for MODE field in sysvipc
+   ipc_perm.  However, some kernel ABI interfaces still expect a 16-bit
+   field.  This is only an issue if arch-defined IPC_PERM padding is on a
+   wrong position regarding endianness.  In this case, the IPC control
+   routines (msgctl, semctl, and semtctl) requires to shift the value to
+   correct place.
+   The ABIs that requires it define __ASSUME_SYSVIPC_BROKEN_MODE_T.  */
+
 /* Support for p{read,write}v2 was added in 4.6.  However Linux default
    implementation does not assume the __ASSUME_* and instead use a fallback
    implementation based on p{read,write}v and returning an error for