linux.git/net/phonet, branch v2.6.39

Phonet: fix aligned-mode pipe socket buffer header reserve

2011-03-15T21:55:49+00:00

When the pipe uses aligned-mode data packets, we must reserve 4 bytes
instead of 3 for the pipe protocol header. Otherwise the Phonet header
would not be aligned, resulting in potentially corrupted headers with
later unaligned memory writes.

Signed-off-by: Rémi Denis-Courmont 
Signed-off-by: David S. Miller

Phonet: kill the ST-Ericsson pipe controller Kconfig

2011-03-09T19:59:33+00:00

This is now a run-time choice so that a single kernel can support both
old and new generation ISI modems. Support for manually enabling the
pipe flow is removed as it did not work properly, does not fit well
with the socket API, and I am not aware of any use at the moment.

Signed-off-by: Rémi Denis-Courmont 
Signed-off-by: David S. Miller

Phonet: support active connection without pipe controller on modem

2011-03-09T19:59:33+00:00

This provides support for newer ISI modems with no need for the
earlier experimental compile-time alternative choice. With this,
we can now use the same kernel and userspace with both types of
modems.

This also avoids confusing two different and incompatible state
machines, actively connected vs accepted sockets, and adds
connection response error handling (processing "SYN/RST" of sorts).

Signed-off-by: Rémi Denis-Courmont 
Signed-off-by: David S. Miller

Phonet: provide pipe socket option to retrieve the pipe identifier

2011-03-09T19:59:32+00:00

User-space sometimes needs this information. In particular, the GPRS
context or the AT commands pipe setups may use the pipe handle as a
reference.

This removes the settable pipe handle with CONFIG_PHONET_PIPECTRLR.
It did not handle error cases correctly. Furthermore, the kernel
*could* implement a smart scheme for allocating handles (if ever
needed), but userspace really cannot.

Signed-off-by: Rémi Denis-Courmont 
Signed-off-by: David S. Miller

Phonet: allocate sock from accept syscall rather than soft IRQ

2011-03-09T19:59:32+00:00

This moves most of the accept logic to process context like other
socket stacks do. Then we can use a few more common socket helpers
and simplify a bit.

Signed-off-by: Rémi Denis-Courmont 
Signed-off-by: David S. Miller

Phonet: factor common code to send control messages

2011-03-09T19:59:31+00:00

With the addition of the pipe controller, there is now quite a bit
of repetitive code for small signaling messages. Lets factor it.

Signed-off-by: Rémi Denis-Courmont 
Signed-off-by: David S. Miller

Phonet: correct pipe backlog callback return values

2011-03-09T19:59:30+00:00

In some cases, the Phonet pipe backlog callbacks returned negative
errno instead of NET_RX_* values.

In other cases, NET_RX_DROP was returned for invalid packets, even
though it seems only intended for buffering problems (not for
deliberately discarded packets).

Signed-off-by: Rémi Denis-Courmont 
Signed-off-by: David S. Miller

Phonet: return an error when packet TX fails

2011-03-09T19:59:30+00:00

Phonet assumes that packets are never dropped. We try our best to
avoid this situation. But lets return ENOBUFS if queueing to the
network device fails so that the caller knows things went wrong.

Signed-off-by: Rémi Denis-Courmont 
Signed-off-by: David S. Miller

Phonet: fix NULL dereference on TX path with implicit source

2011-03-09T19:59:29+00:00

The previous Phonet patch series introduced per-socket implicit
destination (i.e. connect()). In that case, the destination
socket address is NULL in the transmit function.
However commit a8059512b120362b15424f152b2548fe8b11bd0c
("Phonet: implement per-socket destination/peer address")
is incomplete and would trigger a NULL dereference.
(Fortunately, the code is not in released kernel, and in fact
 currently not reachable.)

Signed-off-by: Rémi Denis-Courmont 
Signed-off-by: David S. Miller

phonet: Protect pipe_do_remove() with appropriate ifdefs.

2011-02-25T19:23:22+00:00

It is only used when CONFIG_PHONET_PIPECTRLR is not set.

Signed-off-by: David S. Miller