This patch adds a check for bad length in set key flow. This solves the
Oops reported by Thomas Backlund, Joonwoo Park and Ian Schram.
It also adds some debug printing that can be useful.
Assaf Krauss [Thu, 12 Jun 2008 01:47:06 +0000 (09:47 +0800)]
iwlwifi: Fix mode changes (ad-hoc <--> managed)
This fix allows to move between modes (ad-hoc to managed, and vice versa).
Since mode changes can only be done while driver is down, check for ibss
support can only be made when the channel is set (afte the driver goes up).
Assaf Krauss [Thu, 12 Jun 2008 01:47:05 +0000 (09:47 +0800)]
iwlwifi enabling IBSS (Ad-Hoc) mode
This patch enables ibss mode. It consists of two changes upon entering ibss
mode:
1. Removing the redundant line which clears the driver's station table.
This line creates a discrepancy between the driver and the FW's station
table. This prevented the generation of beacons.
2. Assigning a default value to priv's assoc_id. Normally given by an AP in
STA mode, this field is used as an indication for association. Being 0,
it prevented normal TX flow.
3. Remove a redundant ADD_STA command that cause uCode error.
4. Delay the set_mode until after the uCode is ready.
iwlwifi: add possibility to disable tx_power calibration
This patch adds the possibility to disable the tx_power calibration.
In 5000 HW, this calibration is implemented in uCode, hence, it is
disabled in driver by default.
This patch makes some clean up in setup/cancel_deferred_work.
iwl_setup_deferred_work does the work that is common to 4965
and 5000, then it calls to HW specific handlers.
This patch also removes uneeded work_struct from iwl_priv.
This patch removes the possibility not to compile the run time
calibrations. It also renames priv->sensitivity_work to
priv->run_time_calib_work, and moves bg_run_time_calib to iwl4965_base
since it is common to both: 4965 and 5000.
Tony Vroon [Wed, 11 Jun 2008 20:23:56 +0000 (16:23 -0400)]
mac80211: implement EU regulatory domain
Implement missing EU regulatory domain for mac80211. Based on the
information in IEEE 802.11-2007 (specifically pages 1142, 1143 & 1148)
and ETSI 301 893 (V1.4.1).
With thanks to Johannes Berg.
Ron Rindjunsky [Mon, 9 Jun 2008 19:54:35 +0000 (22:54 +0300)]
iwlwifi: removing IWL4965_HT config
This patch removes CONFIG_IWL4965_HT #ifdefs for iwl 4965 and 5000. 11n
feature is stable in those drivers and its mode of operation is determined
in mac80211, so this dependency is not needed any more.
Rework the alignment check in rt2x00dev.c to be independent of any
potential alignment measures that may be taken before. Just check
whether the payload is aligned based on the pointer addresses.
Note: This is preparatory for the dynamically mapped skb buffers
for the PCI drivers, as these need 4-byte alignment instead of the
currently enforced offset by 2 bytes.
rt2x00: Fix double usage of skb->cb in USB RX path.
It is not safe to use the skb->cb area for both the rxd and
skb_frame_desc data at the same time, while they occupy an overlapping
piece of memory. This can lead to hard to trace crashes as pointers
within skb_frame_desc are pointing into nowhere, or the rxd data is
overwritten with non-sense.
Fix it by copying the rxd to a small buffer on the stack.
rt2x00: Centralize RX packet alignment handling in rt2x00lib.
When rt2x00pci will be switched over to dynamically mapped skb's
instead of statically allocated DMA buffers, it no longer can handle
alignment of RX packets in a copy step, and needs to implement the
same scheme as rt2x00usb does.
In order to make the patch on dynamically mapped skb's smaller,
already centralize the alignment handling into rt2x00lib. This allows
us to move more code in rt2x00lib, and thus remove code duplication
between rt2x00usb and rt2x00pci.
Ivo van Doorn [Fri, 6 Jun 2008 20:53:14 +0000 (22:53 +0200)]
rt2x00: Don't kick TX queue after each frame
TX queues shouldn't be kicked after each frame that is put into the
queue. This could cause problems during RTS and CTS-to-self as well
as with fragmentation. In all those cases you want all frames to be
send out in a single burst. Off course we shouldn't let the queue fill
up entirely, thus we introduce a 10% threshold which, when reached,
will force the frames to be send out regardless of the frame.
In addition we should prevent queues to become full in such a way
that the tx() handler can fail. Instead of stopping the queue when
it is full, we should stop it when it is below the threshold.
Ivo van Doorn [Fri, 6 Jun 2008 20:47:39 +0000 (22:47 +0200)]
rt2x00: Implement rt2x00usb_kick_tx_queue()
rt2x00usb_kick_tx_queue() will loop over all entries
within the INDEX_DONE->INDEX range and kick each entry
which is pending to be kicked. This makes the kick_tx_queue
approach work the same as with the PCI drivers which
will allow for more code generalisation into rt2x00lib.
Ivo van Doorn [Tue, 3 Jun 2008 18:29:05 +0000 (20:29 +0200)]
rt2x00: Fix queue initialization
qid should be initialized to QID_BEACON and QID_ATIM
for the beacon and atim quue. This makes checking for
a particular queue much saner, and it shouldn't harm,
because the only places where the value is send to
the hardware, we are allowed to send any value we
want since it is only used as argument in the
TX done register.
Ivo van Doorn [Tue, 3 Jun 2008 20:45:35 +0000 (22:45 +0200)]
rt2x00: Calculate register offset during compile time
By using __ffs() the register offsets were always calculated
at run-time which all FIELD32/FIELD16 definitions were builtin
constants. This means we can heavily optimize the register handling
by allowing GCC to do all the work during compilation.
Add some compile_ffs() macros to perform the calculation at
compile time. After this each rt2x00 module size is reduced
by ~2500 bytes. And the stack size of several functions is reduced
as well which further limits the number of rt2x00 results in
'make checkstack'.
v2: Merge GertJan's bugfix of patch [1/11] directly into this patch
instead of providing it as seperate patch.
v3: Add extra parentheses when bitshifting __x
The commit 3f8cb098859bbea29d7b3765a3102e4a6bf81b85
(drivers/net/lib8390: fix warning, trim trailing whitespace) removed
ei_local from ei_tx_err() and ei_rx_overrun() resulting in the following
build errors on m68k and sh:
Using /scratch1/tony/next as source for kernel
GEN /scratch1/tony/next_out/Makefile
CHK include/linux/version.h
CHK include/linux/utsrelease.h
CALL /scratch1/tony/next/scripts/checksyscalls.sh
CHK include/linux/compile.h
CC [M] drivers/net/zorro8390.o
In file included from /scratch1/tony/next/drivers/net/zorro8390.c:47:
drivers/net/lib8390.c: In function 'ei_tx_err':
drivers/net/lib8390.c:556: error: 'ei_local' undeclared (first use in this function)
drivers/net/lib8390.c:556: error: (Each undeclared identifier is reported only once
drivers/net/lib8390.c:556: error: for each function it appears in.)
drivers/net/lib8390.c: In function 'ei_rx_overrun':
drivers/net/lib8390.c:823: error: 'ei_local' undeclared (first use in this function)
make[3]: *** [drivers/net/zorro8390.o] Error 1
make[2]: *** [drivers/net] Error 2
make[1]: *** [drivers] Error 2
make: *** [sub-make] Error 2
The problem is that ei_inb_p() is using various #defines (from
drivers/net/8390.h) that use EI_SHIFT, which in some drivers on some
architectures use ei_local. Tag ei_local as "__maybe_unused" to keep it
around and keep the warnings the original commit is trying to silence
... silenced.
This driver is for a number of different Option devices. Originally
written by Option and Andrew Bird, but cleaned up massivly for
acceptance into mainline by me and others.
Adrian Bunk [Mon, 9 Jun 2008 22:23:47 +0000 (01:23 +0300)]
drivers/net/macsonic.c: make functions static
This patch makes the following needlessly global functions static:
- macsonic_init()
- mac_onboard_sonic_ethernet_addr()
- mac_onboard_sonic_probe()
- mac_nubus_sonic_ethernet_addr()
- macsonic_ident()
- mac_nubus_sonic_probe()
Magnus Damm [Mon, 9 Jun 2008 23:33:54 +0000 (16:33 -0700)]
smc911x: pass along private data and use iomem
This patch contains changes needed for platform data support:
- Move smc911x_local structure to header file
- Pass along smc911x_local structure pointer to macros
- Keep register base address in smc911x_local structure
- Remove unused ioaddr variables
Magnus Damm [Mon, 9 Jun 2008 23:33:53 +0000 (16:33 -0700)]
smc911x: fix 16-bit I/O operations
This patch fixes the following issues related to 16-bit support:
- Remove unused 16-bit PXA DMA implementation.
- Remove unused SMC_inw() and SMC_outw() functions.
- Fix 16-bit SMC_outl to use writew() instead of writel().
Benjamin Thery [Wed, 28 May 2008 12:51:24 +0000 (14:51 +0200)]
ipv6 netns: Address labels per namespace
This pacth makes IPv6 address labels per network namespace.
It keeps the global label tables, ip6addrlbl_table, but
adds a 'net' member to each ip6addrlbl_entry.
This new member is taken into account when matching labels.
Changelog
=========
* v1: Initial version
* v2:
* Minize the penalty when network namespaces are not configured:
* the 'net' member is added only if CONFIG_NET_NS is
defined. This saves space when network namespaces are not
configured.
* 'net' value is retrieved with the inlined function
ip6addrlbl_net() that always return &init_net when
CONFIG_NET_NS is not defined.
* 'net' member in ip6addrlbl_entry renamed to the less generic
'lbl_net' name (helps code search).
Harvey Harrison [Tue, 10 Jun 2008 19:48:45 +0000 (12:48 -0700)]
bluetooth: hci_bcsp.c small cleanups/api users
Use bitrev16 from lib/bitrev.c.
Use the get_unaligned_be16 to get the crc from the packet, create a
small helper function for this.
Fix a shadowed variable sparse warning:
drivers/bluetooth/hci_bcsp.c:218:26: warning: symbol 'hdr' shadows an earlier one
drivers/bluetooth/hci_bcsp.c:187:5: originally declared here
- No need to perform data_len = 0 in the switch command, since data_len
is initialized to 0 in the beginning of the ipq_build_packet_message()
method.
- {ip,ip6}_queue: We can reach nlmsg_failure only from one place; skb is
sure to be NULL when getting there; since skb is NULL, there is no need
to check this fact and call kfree_skb().
Rami Rosen [Mon, 9 Jun 2008 23:00:22 +0000 (16:00 -0700)]
netfilter: nf_conntrack: remove unnecessary function declaration
This patch removes nf_ct_ipv4_ct_gather_frags() method declaration from
include/net/netfilter/ipv4/nf_conntrack_ipv4.h, since it is unused in
the Linux kernel.
netfilter: ctnetlink: include conntrack status in destroy event message
When a conntrack is destroyed, the connection status does not get
exported to netlink. I don't see a reason for not doing so. This patch
exports the status on all conntrack events.
Currently the last packet of a connection isn't accounted when its causing
abnormal termination.
Introduces nf_ct_kill_acct() which increments the accounting counters on
conntrack kill. The new function was necessary, because there are calls
to nf_ct_kill() which don't need accounting:
nf_conntrack_proto_tcp.c line ~847:
Kills ct and returns NF_REPEAT. We don't want to count twice.
nf_conntrack_proto_tcp.c line ~880:
Kills ct and returns NF_DROP. I think we don't want to count dropped
packets.
nf_conntrack_netlink.c line ~824:
As far as I can see ctnetlink_del_conntrack() is used to destroy a
conntrack on behalf of the user. There is an sk_buff, but I don't think
this is an actual packet. Incrementing counters here is therefore not
desired.
Pekka Enberg [Mon, 9 Jun 2008 22:58:39 +0000 (15:58 -0700)]
netfilter: nf_conntrack_extend: use krealloc() in nf_conntrack_extend.c V2
The ksize() API is going away because it is being abused and it doesn't even
work consistenly across different allocators. Therefore, convert
net/netfilter/nf_conntrack_extend.c to use krealloc().
James Morris [Mon, 9 Jun 2008 22:57:24 +0000 (15:57 -0700)]
netfilter: ip_tables: add iptables security table for mandatory access control rules
The following patch implements a new "security" table for iptables, so
that MAC (SELinux etc.) networking rules can be managed separately to
standard DAC rules.
This is to help with distro integration of the new secmark-based
network controls, per various previous discussions.
The need for a separate table arises from the fact that existing tools
and usage of iptables will likely clash with centralized MAC policy
management.
The SECMARK and CONNSECMARK targets will still be valid in the mangle
table to prevent breakage of existing users.
projects / linux.git / log