[COMMITED,2.22,2.21,2.20,BZ,#19682] S390: Do not use direct socket syscalls if build on kernels >= 4.3.

  Hi,

Beginning with Linux 4.3, the kernel headers contain direct
system call numbers __NR_socket etc. on s390x. On older kernels,
the socket-multiplexer syscall __NR_socketcall was used.

To enable these new syscalls, the patch
"S390: Call direct system calls for socket operations."
(https://sourceware.org/git/?p=glibc.git;a=commit;h=016495b818cb61df7d0d10e6db54074271b3e3a5)
was applied upstream.

If glibc 2.23 is configured with --enable-kernel=4.3 and newer,
the direct socket syscalls are used.
For older kernels, the socket-multiplexer syscall is used instead.

In glibc 2.22 and earlier, this patch is not applied.
If you build glibc on a kernel < 4.3, the socket-multiplexer
syscall is used. But if you build glibc on kernel >= 4.3, the
direct socket-syscalls are used. If you install this glibc on a
kernel < 4.3, all socket operations will fail.
See "Bug 19682 - s390x: Incorrect syscall definitions cause
breakage with Linux 4.3 headers"
(https://sourceware.org/bugzilla/show_bug.cgi?id=19682)
The configure switch --enable-kernel does not influence this
behaviour on older glibc-releases.

The solution is to remove the direct socket-syscalls in
sysdeps/unix/sysv/linux/s390/s390-64/syscalls.list
(this patch) on older glibc-releases as it was done by the
upstream patch, too. These entries were never used on s390x,
but the c-files in sysdeps/unix/sysv/linux/.
After this removal, the behaviour of the socket functions are
not changed compared to the original glibc release version
and the socket-multiplexer-syscall is always used.

After talking to the distribution maintainers, the decision
is to backport it to glibc 2.22, 2.21, 2.20 as done with this
commit.

Bye Stefan
commit 6f78924e17da075a9110b1d9dd3950ac63900843
Author: Stefan Liebler <stli@linux.vnet.ibm.com>
Date:   Wed Mar 2 16:16:49 2016 +0100

    S390: Do not use direct socket syscalls if build on kernels >= 4.3. [BZ #19682]
    
    Beginning with Linux 4.3, the kernel headers contain direct
    system call numbers __NR_socket etc. on s390x. On older kernels,
    the socket-multiplexer syscall __NR_socketcall was used.
    
    To enable these new syscalls, the patch
    "S390: Call direct system calls for socket operations."
    (https://sourceware.org/git/?p=glibc.git;a=commit;h=016495b818cb61df7d0d10e6db54074271b3e3a5)
    was applied upstream.
    
    If glibc 2.23 is configured with --enable-kernel=4.3 and newer,
    the direct socket syscalls are used.
    For older kernels, the socket-multiplexer syscall is used instead.
    
    In glibc 2.22 and earlier, this patch is not applied.
    If you build glibc on a kernel < 4.3, the socket-multiplexer
    syscall is used. But if you build glibc on kernel >= 4.3, the
    direct socket-syscalls are used. If you install this glibc on a
    kernel < 4.3, all socket operations will fail.
    See "Bug 19682 - s390x: Incorrect syscall definitions cause
    breakage with Linux 4.3 headers"
    (https://sourceware.org/bugzilla/show_bug.cgi?id=19682)
    The configure switch --enable-kernel does not influence this
    behaviour on older glibc-releases.
    
    The solution is to remove the direct socket-syscalls in
    sysdeps/unix/sysv/linux/s390/s390-64/syscalls.list
    (this patch) on older glibc-releases as it was done by the
    upstream patch, too. These entries were never used on s390x,
    but the c-files in sysdeps/unix/sysv/linux/.
    After this removal, the behaviour of the socket functions are
    not changed compared to the original glibc release version
    and the socket-multiplexer-syscall is always used.
commit 21291f5a77921518b4f62636ec96ede8d8ca6529
Author: Stefan Liebler <stli@linux.vnet.ibm.com>
Date:   Wed Mar 2 15:52:48 2016 +0100

    S390: Do not use direct socket syscalls if build on kernels >= 4.3. [BZ #19682]
    
    Beginning with Linux 4.3, the kernel headers contain direct
    system call numbers __NR_socket etc. on s390x. On older kernels,
    the socket-multiplexer syscall __NR_socketcall was used.
    
    To enable these new syscalls, the patch
    "S390: Call direct system calls for socket operations."
    (https://sourceware.org/git/?p=glibc.git;a=commit;h=016495b818cb61df7d0d10e6db54074271b3e3a5)
    was applied upstream.
    
    If glibc 2.23 is configured with --enable-kernel=4.3 and newer,
    the direct socket syscalls are used.
    For older kernels, the socket-multiplexer syscall is used instead.
    
    In glibc 2.22 and earlier, this patch is not applied.
    If you build glibc on a kernel < 4.3, the socket-multiplexer
    syscall is used. But if you build glibc on kernel >= 4.3, the
    direct socket-syscalls are used. If you install this glibc on a
    kernel < 4.3, all socket operations will fail.
    See "Bug 19682 - s390x: Incorrect syscall definitions cause
    breakage with Linux 4.3 headers"
    (https://sourceware.org/bugzilla/show_bug.cgi?id=19682)
    The configure switch --enable-kernel does not influence this
    behaviour on older glibc-releases.
    
    The solution is to remove the direct socket-syscalls in
    sysdeps/unix/sysv/linux/s390/s390-64/syscalls.list
    (this patch) on older glibc-releases as it was done by the
    upstream patch, too. These entries were never used on s390x,
    but the c-files in sysdeps/unix/sysv/linux/.
    After this removal, the behaviour of the socket functions are
    not changed compared to the original glibc release version
    and the socket-multiplexer-syscall is always used.

diff --git a/ChangeLog b/ChangeLog
index 3578f2e..f15298a 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,9 @@
+2016-03-02  Stefan Liebler  <stli@linux.vnet.ibm.com>
+
+	[BZ #19682]
+	* sysdeps/unix/sysv/linux/s390/s390-64/syscalls.list:
+	Remove socketcall syscalls.
+
 2016-02-15  Carlos O'Donell  <carlos@redhat.com>
 
 	[BZ #18665]
diff --git a/NEWS b/NEWS
index f1abb66..dd78048 100644
--- a/NEWS
+++ b/NEWS
@@ -9,7 +9,8 @@ Version 2.21.1
 
 * The following bugs are resolved with this release:
 
-  17269, 17905, 17949, 18007, 18032, 18240, 18287, 18694, 18887, 18985.
+  17269, 17905, 17949, 18007, 18032, 18240, 18287, 18694, 18887, 18985,
+  19682.
 
 * A stack-based buffer overflow was found in libresolv when invoked from
   libnss_dns, allowing specially crafted DNS responses to seize control
diff --git a/sysdeps/unix/sysv/linux/s390/s390-64/syscalls.list b/sysdeps/unix/sysv/linux/s390/s390-64/syscalls.list
index 5b8c102..9f03d26 100644
--- a/sysdeps/unix/sysv/linux/s390/s390-64/syscalls.list
+++ b/sysdeps/unix/sysv/linux/s390/s390-64/syscalls.list
@@ -12,22 +12,3 @@ shmget		-	shmget		i:iii	__shmget	shmget
 semop		-	semop		i:ipi	__semop		semop
 semget		-	semget		i:iii	__semget	semget
 semctl		-	semctl		i:iiii	__semctl	semctl
-
-# proper socket implementations:
-accept		-	accept		Ci:iBN	__libc_accept	__accept accept
-bind		-	bind		i:ipi	__bind		bind
-connect		-	connect		Ci:ipi	__libc_connect	__connect connect
-getpeername	-	getpeername	i:ipp	__getpeername	getpeername
-getsockname	-	getsockname	i:ipp	__getsockname	getsockname
-getsockopt	-	getsockopt	i:iiiBN	__getsockopt	getsockopt
-listen		-	listen		i:ii	__listen	listen
-recv		-	recv		Ci:ibni	__libc_recv	__recv recv
-recvfrom	-	recvfrom	Ci:ibniBN	__libc_recvfrom	__recvfrom recvfrom
-recvmsg		-	recvmsg		Ci:ipi	__libc_recvmsg	__recvmsg recvmsg
-send		-	send		Ci:ibni	__libc_send	__send send
-sendmsg		-	sendmsg		Ci:ipi	__libc_sendmsg	__sendmsg sendmsg
-sendto		-	sendto		Ci:ibnibn	__libc_sendto	__sendto sendto
-setsockopt	-	setsockopt	i:iiibn	__setsockopt	setsockopt
-shutdown	-	shutdown	i:ii	__shutdown	shutdown
-socket		-	socket		i:iii	__socket	socket
-socketpair	-	socketpair	i:iiif	__socketpair	socketpair
commit 5acee335e6a546354a71df0b5265e021a40c4c16
Author: Stefan Liebler <stli@linux.vnet.ibm.com>
Date:   Wed Mar 2 15:25:03 2016 +0100

    S390: Do not use direct socket syscalls if build on kernels >= 4.3. [BZ #19682]
    
    Beginning with Linux 4.3, the kernel headers contain direct
    system call numbers __NR_socket etc. on s390x. On older kernels,
    the socket-multiplexer syscall __NR_socketcall was used.
    
    To enable these new syscalls, the patch
    "S390: Call direct system calls for socket operations."
    (https://sourceware.org/git/?p=glibc.git;a=commit;h=016495b818cb61df7d0d10e6db54074271b3e3a5)
    was applied upstream.
    
    If glibc 2.23 is configured with --enable-kernel=4.3 and newer,
    the direct socket syscalls are used.
    For older kernels, the socket-multiplexer syscall is used instead.
    
    In glibc 2.22 and earlier, this patch is not applied.
    If you build glibc on a kernel < 4.3, the socket-multiplexer
    syscall is used. But if you build glibc on kernel >= 4.3, the
    direct socket-syscalls are used. If you install this glibc on a
    kernel < 4.3, all socket operations will fail.
    See "Bug 19682 - s390x: Incorrect syscall definitions cause
    breakage with Linux 4.3 headers"
    (https://sourceware.org/bugzilla/show_bug.cgi?id=19682)
    The configure switch --enable-kernel does not influence this
    behaviour on older glibc-releases.
    
    The solution is to remove the direct socket-syscalls in
    sysdeps/unix/sysv/linux/s390/s390-64/syscalls.list
    (this patch) on older glibc-releases as it was done by the
    upstream patch, too. These entries were never used on s390x,
    but the c-files in sysdeps/unix/sysv/linux/.
    After this removal, the behaviour of the socket functions are
    not changed compared to the original glibc release version
    and the socket-multiplexer-syscall is always used.

diff --git a/ChangeLog b/ChangeLog
index 2ed60d6..56642cb 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,9 @@
+2016-03-02  Stefan Liebler  <stli@linux.vnet.ibm.com>
+
+	[BZ #19682]
+	* sysdeps/unix/sysv/linux/s390/s390-64/syscalls.list:
+	Remove socketcall syscalls.
+
 2016-03-01  Aurelien Jarno  <aurelien@aurel32.net>
 
 	* sysdeps/alpha/fpu/libm-test-ulps: Update ULPs.
diff --git a/NEWS b/NEWS
index 81ceeae..3d52bec 100644
--- a/NEWS
+++ b/NEWS
@@ -24,7 +24,8 @@ Version 2.22.1
 * The following bugs are resolved with this release:
 
   17905, 18420, 18421, 18480, 18589, 18743, 18778, 18781, 18787, 18796,
-  18870, 18887, 18921, 18928, 18969, 18985, 19018, 19058, 19174, 19178.
+  18870, 18887, 18921, 18928, 18969, 18985, 19018, 19058, 19174, 19178,
+  19682.
 
 * The LD_POINTER_GUARD environment variable can no longer be used to
   disable the pointer guard feature.  It is always enabled.
diff --git a/sysdeps/unix/sysv/linux/s390/s390-64/syscalls.list b/sysdeps/unix/sysv/linux/s390/s390-64/syscalls.list
index 5b8c102..9f03d26 100644
--- a/sysdeps/unix/sysv/linux/s390/s390-64/syscalls.list
+++ b/sysdeps/unix/sysv/linux/s390/s390-64/syscalls.list
@@ -12,22 +12,3 @@ shmget		-	shmget		i:iii	__shmget	shmget
 semop		-	semop		i:ipi	__semop		semop
 semget		-	semget		i:iii	__semget	semget
 semctl		-	semctl		i:iiii	__semctl	semctl
-
-# proper socket implementations:
-accept		-	accept		Ci:iBN	__libc_accept	__accept accept
-bind		-	bind		i:ipi	__bind		bind
-connect		-	connect		Ci:ipi	__libc_connect	__connect connect
-getpeername	-	getpeername	i:ipp	__getpeername	getpeername
-getsockname	-	getsockname	i:ipp	__getsockname	getsockname
-getsockopt	-	getsockopt	i:iiiBN	__getsockopt	getsockopt
-listen		-	listen		i:ii	__listen	listen
-recv		-	recv		Ci:ibni	__libc_recv	__recv recv
-recvfrom	-	recvfrom	Ci:ibniBN	__libc_recvfrom	__recvfrom recvfrom
-recvmsg		-	recvmsg		Ci:ipi	__libc_recvmsg	__recvmsg recvmsg
-send		-	send		Ci:ibni	__libc_send	__send send
-sendmsg		-	sendmsg		Ci:ipi	__libc_sendmsg	__sendmsg sendmsg
-sendto		-	sendto		Ci:ibnibn	__libc_sendto	__sendto sendto
-setsockopt	-	setsockopt	i:iiibn	__setsockopt	setsockopt
-shutdown	-	shutdown	i:ii	__shutdown	shutdown
-socket		-	socket		i:iii	__socket	socket
-socketpair	-	socketpair	i:iiif	__socketpair	socketpair