Signed-off-by: Helge Deller <deller@gmx.de>

Signed-off-by: Helge Deller <deller@gmx.de>

introduce new setsockopt() command:

setsockopt(sock, SOL_SOCKET, SO_ATTACH_BPF, &prog_fd, sizeof(prog_fd))

where prog_fd was received from syscall bpf(BPF_PROG_LOAD, attr, ...)
and attr->prog_type == BPF_PROG_TYPE_SOCKET_FILTER

setsockopt() calls bpf_prog_get() which increments refcnt of the program,
so it doesn't get unloaded while socket is using the program.

The same eBPF program can be attached to multiple sockets.

User task exit automatically closes socket which calls sk_filter_uncharge()
which decrements refcnt of eBPF program

Signed-off-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Conflicts:
	drivers/net/ieee802154/fakehard.c

A bug fix went into 'net' for ieee802154/fakehard.c, which is removed
in 'net-next'.

Add build fix into the merge from Stephen Rothwell in openvswitch, the
logging macros take a new initial 'log' argument, a new call was added
in 'net' so when we merge that in here we have to explicitly add the
new 'log' arg to it else the build fails.

Signed-off-by: David S. Miller <davem@davemloft.net>

Alternative to RPS/RFS is to use hardware support for multiple
queues.

Then split a set of million of sockets into worker threads, each
one using epoll() to manage events on its own socket pool.

Ideally, we want one thread per RX/TX queue/cpu, but we have no way to
know after accept() or connect() on which queue/cpu a socket is managed.

We normally use one cpu per RX queue (IRQ smp_affinity being properly
set), so remembering on socket structure which cpu delivered last packet
is enough to solve the problem.

After accept(), connect(), or even file descriptor passing around
processes, applications can use :

 int cpu;
 socklen_t len = sizeof(cpu);

 getsockopt(fd, SOL_SOCKET, SO_INCOMING_CPU, &cpu, &len);

And use this information to put the socket into the right silo
for optimal performance, as all networking stack should run
on the appropriate cpu, without need to send IPI (RPS/RFS).

Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The gcc compiler provide the predefined __LP64__ macro. Use that
instead.

Signed-off-by: Helge Deller <deller@gmx.de>

Switch over the msgctl, shmat, shmctl and semtimedop syscalls to use the compat
layer. The problem was found with the debian procenv package, which called
	shmctl(0, SHM_INFO, &info);
in which the shmctl syscall then overwrote parts of the surrounding areas on
the stack on which the info variable was stored and thus lead to a segfault
later on.

Additionally fix the definition of struct shminfo64 to use unsigned longs like
the other architectures. This has no impact on userspace since we only have a
32bit userspace up to now.

Signed-off-by: Helge Deller <deller@gmx.de>
Cc: John David Anglin <dave.anglin@bell.net>
Cc: <stable@vger.kernel.org> # v3.10+

Signed-off-by: Helge Deller <deller@gmx.de>

This patch reduces the value of SIGRTMIN on PARISC from 37 to 32, thus
increasing the number of available RT signals and bring it in sync with other
Linux architectures.

Historically we wanted to natively support HP-UX 32bit binaries with the
PA-RISC Linux port.  Because of that we carried the various available signals
from HP-UX (e.g. SIGEMT and SIGLOST) and folded them in between the native
Linux signals.  Although this was the right decision at that time, this
required us to increase SIGRTMIN to at least 37 which left us with 27 (64-37)
RT signals.

Those 27 RT signals haven't been a problem in the past, but with the upcoming
importance of systemd we now got the problem that systemd alloctes (hardcoded)
signals up to SIGRTMIN+29 which is beyond our NSIG of 64. Because of that we
have not been able to use systemd on the PARISC Linux port yet.

Of course we could ask the systemd developers to not use those hardcoded
values, but this change is very unlikely, esp. with PA-RISC being a niche
architecture.

The other possibility would be to increase NSIG to e.g. 128, but this would
mean to duplicate most of the existing Linux signal handling code into the
parisc specific Linux kernel tree which would most likely introduce lots of new
bugs beside the code duplication.

The third option is to drop some HP-UX signals and shuffle some other signals
around to bring SIGRTMIN to 32.  This is of course an ABI change, but testing
has shown that existing Linux installations are not visibly affected by this
change - most likely because we move those signals around which are rarely used
and move them to slots which haven't been used in Linux yet. In an existing
installation I was able to exchange either the Linux kernel or glibc (or both)
without affecting the boot process and installed applications.

Dropping the HP-UX signals isn't an issue either, since support for HP-UX was
basically dropped a few months back with Kernel 3.14 in commit
f5a408d53edef3af07ac7697b8bc54a755628450 already, when we changed EWOULDBLOCK
to be equal to EAGAIN.

So, even if this is an ABI change, it's better to change it now and thus bring
PARISC Linux in sync with other architectures to avoid other issues in the
future.

Signed-off-by: Helge Deller <deller@gmx.de>
Cc: Carlos O'Donell <carlos@systemhalted.org>
Cc: John David Anglin <dave.anglin@bell.net>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Aaro Koskinen <aaro.koskinen@iki.fi>
Cc: PARISC Linux Kernel Mailinglist <linux-parisc@vger.kernel.org>
Tested-by: Aaro Koskinen <aaro.koskinen@iki.fi>

We want those fixes in here as well.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>