26 Mar, 2008
40 commits
-
Signed-off-by: Al Viro
Signed-off-by: Jeff Garzik -
missed read*/write* plus a bunch of wrong-sized ones...
Signed-off-by: Al Viro
Signed-off-by: Jeff Garzik -
isa_bus_to_virt() is the wrong thing to do here; it happens
to work on i386, but only by accident. What we want is
normal ioremap/readb/etc. set - it's all in iomem.Signed-off-by: Al Viro
Signed-off-by: Jeff Garzik -
take iscp-based testing into helper, kill the loop, stop
wanking with reassignments of priv->iscpSigned-off-by: Al Viro
Signed-off-by: Jeff Garzik -
* initialize spinlock once
* check586() used to be done before we'd allocated ->priv; these days
it's there from the very beginning, so we don't have to play with
private copy. Consequently, we don't need to mess with reinitializing
->base, etc. afterwards.Signed-off-by: Al Viro
Signed-off-by: Jeff Garzik -
... and store the virt address where we map the ->mem_addr, while we
are at it.Signed-off-by: Al Viro
Signed-off-by: Jeff Garzik -
Passing ISA bus address explicitly cast to char * only to cast it back to
unsigned long is dumb; so's passing it at all when it's always dev->mem_start...Signed-off-by: Al Viro
Signed-off-by: Jeff Garzik -
unsigned short != u8, it's u16...
Signed-off-by: Al Viro
Signed-off-by: Jeff Garzik -
Driver is still broken, though; partially from Alan's checkpatch-induced
fun, partially from layers of ancient mess ;-)By the end of the series... hell, might be even worth trying to stick
such card into old alpha or ppc with an ISA slot and see if it work -
would be for the first time ever in case of alpha and for the first
time since at least 2.5.3 in case of ppc...Signed-off-by: Al Viro
Signed-off-by: Jeff Garzik -
ixgb can remove irq_sem by auditing all the call sites to make sure
that each of them makes sure the adapter is in the correct state
before re-enabling interrupts. after doing this to all of our other
drivers it is becoming easier.Signed-off-by: Jesse Brandeburg
Signed-off-by: Auke Kok
Signed-off-by: Jeff Garzik -
irq_sem was just a hack to prevent interrupts from being enabled
unexpectedly in deep call paths. Simply finding those call paths and
fixing them by hand results in a driver that behaves as we expect and
doesn't need the atomic at all.Signed-off-by: Jesse Brandeburg
Signed-off-by: Auke Kok
Signed-off-by: Jeff Garzik -
irq_sem can safely be removed by auditing all irq.*able sites to
make sure that interrupts don't get enabled unexpectedly when the
interface is down.Signed-off-by: Jesse Brandeburg
Signed-off-by: Auke Kok
Signed-off-by: Jeff Garzik -
This function is no longer used now that 82573 uses the eerd
read method as well. Thanks to Adrian Bunk for pointing this out.Signed-off-by: Auke Kok
Signed-off-by: Jeff Garzik -
Signed-off-by: Auke Kok
Signed-off-by: Jeff Garzik -
> send me a patch for e1000 and for ixgb and I'll happily apply those :)
boolean_t to bool
TRUE to true
FALSE to falseSigned-off-by: Joe Perches
Signed-off-by: Auke Kok
Signed-off-by: Jeff Garzik -
In order to remove the irq_sem code we need to implement strict
adapter state checking to prevent accidental double up or downs
or resets. This code is largely copied from e1000/e1000e.Signed-off-by: Jesse Brandeburg
Signed-off-by: Auke Kok
Signed-off-by: Jeff Garzik -
On Thu, 2008-03-06 at 10:07 -0800, Kok, Auke wrote:
> send me a patch for e1000 and for ixgb and I'll happily apply those :)boolean_t to bool
TRUE to true
FALSE to false
comment typo ahread to aheadSigned-off-by: Joe Perches
Signed-off-by: Auke Kok
Signed-off-by: Jeff Garzik -
rt73usb and rt2500usb used in_atomic to determine
if a configuration step should be rescheduled or not.
Since in_atomic() is not a valid method to determine
if sleeping is allowed we should fix the way this is handled
by adding a new flag to rt2x00.In addition mark LED class support for the drivers broken
since that also uses the broken in_atomic() method but
so far no solution exists to have LED triggers work only
in scheduled context.Signed-off-by: Ivo van Doorn
Signed-off-by: John W. Linville -
It has been observed on rt2500pci hardware that some
frames received with signal 0x0C do not have the OFDM
flag set.Signals can have 2 meanings:
1) The PLCP value
2) The bitrate * 10For rt2500pci (1) is for frames received with a OFDM rate,
and (2) is for frames received with a CCK rate.
But 0x0C is a invalid bitrate value but is a valid PLCP
value for 54Mbs (obvious OFDM rate).
This means that it is possible that the hardware does not
set the OFDM bit correctly under all circumstances.
This results in rt2x00 failing to detect the rate and
mac80211 triggering a WARN_ON() and dropping the frame.To bypass this, print a warning when such a frame is received,
and reset the rate to the lowest supported rate for the current band.Signed-off-by: Ivo van Doorn
Signed-off-by: John W. Linville -
From version 2.20 of the PS3 system software, the hypervisor allows
the guest OSes to specify separate cipher for group and pairwise.Signed-off-by: Masakazu Mokuno
Signed-off-by: John W. Linville -
There's no need to compile the N-PHY support code, when the
N-PHY support is disabled in Kconfig.Signed-off-by: Michael Buesch
Signed-off-by: John W. Linville -
This patch completes the changes regarding the packaging of user
parameters.Signed-off-by: Assaf Krauss
Signed-off-by: Tomas Winkler
Signed-off-by: John W. Linville -
a) a bunch of printks in error-handling assums that ->status is
big-endian.
b) bitfields trouble
c) missing annotationsNB: a bunch of structs is declared packed for no good reason, AFAICS.
Signed-off-by: Al Viro
Acked-by: Reinette Chatre
Signed-off-by: John W. Linville -
The rxdone_entry_desc structure contains 3 fields
which are always 1 or 0. We can safe 8 bytes by
replacing them with a single dev_flags fields which
contain the flags for those settings.Additionally we can remove the OFDM flag since it
is no longer used since the introduction of the
SIGNAL_PLCP flag.Signed-off-by: Ivo van Doorn
Signed-off-by: John W. Linville -
Default WMM params have to be set according to beacon/probe response
information prior to authentication (or IBSS start/join); beacon queue
is configured only in IBSS. This does not affect the use of 'real' WMM
params as reported by AP.Signed-off-by: Vladimir Koutny
Acked-by: Johannes Berg
Signed-off-by: John W. Linville -
Postpone calling ieee80211_hw_config if hardware scanning is active.
This is similar to solution for software scanning where channel setting
is delayed until scan complete.Signed-off-by: Mohamed Abbas
Signed-off-by: Reinette Chatre
Acked-by: Johannes Berg
Signed-off-by: John W. Linville -
This patch adds a clean tear down for all block ack sessions if interface
goes down or if a deauthentication is done.Signed-off-by: Ron Rindjunsky
Signed-off-by: Tomas Winkler
Signed-off-by: John W. Linville -
This patch also fixes the Rx timer's comments
Signed-off-by: Ron Rindjunsky
Signed-off-by: Tomas Winkler
Signed-off-by: John W. Linville -
This patch fixes a wrong debug print when receiving delba
Signed-off-by: Ron Rindjunsky
Signed-off-by: Tomas Winkler
Signed-off-by: John W. Linville -
When you have an AP on channel 13, it will currently often enough
be listed in scan results even when the regulatory domain restricts
to channels 1-11. This is due to channel overlap. To avoid getting
very strange failures, don't show such APs in the scan results. The
failure mode will now go from "I can see the AP but not associate"
to "I can't see the AP although I know it's there" which is easier
to debug.This problem was first really noticed by Jes Sorensen.
Signed-off-by: Johannes Berg
Cc: Jes Sorensen
Signed-off-by: John W. Linville -
Use the new ieee80211_get_channel() function instead of open-coding it.
Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville -
Add ieee80211_get_channel() which gets you a channel struct for a
specific wiphy if that channel is present in that wiphy.Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville -
1. Host command sending functions moved from iwl4965-base.c to iwl-hcmd.c
in iwlcore module
2. enqueue_hcmd function currently stays in iwl4965-base.c. It is invoked
through the new 'utils' field in priv's ops.Signed-off-by: Tomas Winkler
Signed-off-by: Assaf Krauss
Signed-off-by: John W. Linville -
iwl4965_get_channel_info was moved to iwlcore module
4965 needs to be stripped offSigned-off-by: Assaf Krauss
Signed-off-by: Tomas Winkler
Signed-off-by: John W. Linville -
This patch removes 4965 host commands from iwl-3945-commands.h
Signed-off-by: Tomas Winkler
Signed-off-by: Reinette Chatre
Signed-off-by: John W. Linville -
This patch enables RX TKIP decryption in HW.
Signed-off-by: Emmanuel Grumbach
Signed-off-by: Tomas Winkler
Signed-off-by: John W. Linville -
This patch add support for TKIP encryption (TX) in HW.
Signed-off-by: Emmanuel Grumbach
Signed-off-by: Tomas Winkler
Signed-off-by: John W. Linville -
This patch makes mac80211 able to send a phase1 key for TKIP
decryption.
This is needed for drivers that don't do the rekeying by themselves
(i.e. iwlwifi). Upon IV16 wrap around, the packet is decrypted in SW,
if decryption is ok, mac80211 calls to update_tkip_key with a new
phase 1 RX key.Signed-off-by: Emmanuel Grumbach
Signed-off-by: Tomas Winkler
Signed-off-by: John W. Linville -
This patch makes mac80211 able to compute a TKIP key from an skb.
The requested key can be a phase 1 or a phase 2 key.
This is useful for drivers who need to provide tkip key to their
HW to enable HW encryption.Signed-off-by: Emmanuel Grumbach
Signed-off-by: Tomas Winkler
Signed-off-by: John W. Linville
