| Age | Commit message (Collapse) | Author |
|---|
|
function. Can be useful for cards that only support 40-bit WEP or
where the card firmware lacks weak IVs avoidance. Prism/Symbol only.
In the future this will be expanded to support proposed WEP replacements.
Based on code from Jamison Adcock.
|
|
|
|
mickey and millert ok.
|
|
mickey and millert ok.
|
|
by millert@
|
|
evaluate the same thing for each loop iteration in wi_get_id().
[ previous commit contained unrelated changes and has been removed ]
|
|
since gcc doesn't seem to evaluate htole16() of a constant at
compile time after all. Noticed by deraadt@
|
|
make the table itself a #define so we can reuse it in wicontrol.
Also add a bunch of chip IDs from NetBSD/FreeBSD.
mickey@ OK
|
|
|
|
|
|
|
|
The reason for this is that it is possible in the future to have
wi(4) be a real AP using the AP tertiary firmware.
mickey@ OK
|
|
wi_init() if only IFF_PROMISC status has changed. There is no need
to change anything if we are in HostAP mode since wi_init() has
already turned off the hardware promisc bit when we entered HostAP
(HostAP already implies promisc behavior).
Noticed by imp@, OK by mickey@
|
|
|
|
|
|
somehow i didn't commit this. from denver airport.
|
|
|
|
doesn't do the wrong thing with quick card insert/ejects.
|
|
|
|
routines have been called. Avoids a panic when you insert & eject
a card very quickly on a slow machine.
|
|
sparc64 and macppc; tested by me and drahn@ on i386, macppc and sparc64.
|
|
Hopefully this will confuse people with crufty firmware a little less
than the old behavior. Supposedly it is possible to get hostap working
on prism firmware 0.7.6 but I'm not going to attempt that at this time.
|
|
|
|
the interface was up, even if it was the empty string. This tends
to confuse users who have not set the desired net name (ie: associate
with any AP). Now we only return the desired net name if it is not
empty and the interface is up. Otherwise we return the current
net name (ie: what we are associated with).
|
|
'struct arpcom foo' -> 'struct arpcom sc_arpcom'
ok itojun@
|
|
of comparing directly against WI_STAT_1042, WI_STAT_TUNNEL, etc.
This is needed by some Symbol cards in ad-hoc mode.
o change some ntohs(var) to htons(constant).
Adapted from NetBSD (mycroft).
|
|
The common.h include file has been incorporated into if_wi_ieee.h
similar to what is in FreeBSD.
|
|
From John Hay via Warner Losh
|
|
|
|
|
|
not actually set. Closed PR #2569.
|
|
or shared key).
|
|
early as possible. Also move the timeout_del() in wi_stop() to
before we disable wi(4) so it doesn't fire at the wrong time.
Joint effort between mickey@ and myself.
|
|
|
|
o Fix m->m_data alignment in wi_rxeof()
o Add a comment
|
|
EXCLUDE_UNENCRYPTED bit in the WI_RID_P2_ENCRYPTION value if we are
in shared key mode. Symbol cards, just to be different, supposedly
always want the EXCLUDE_UNENCRYPTED bit set (confirmed with the
Linux orinoco and spectrum24 drivers).
I have so far been unable to get my Symbol card to interop with
Lucent or Prism2 WEP. In BSS mode at least, this Symbol card does
not list WEP as a capability it supports. This is, to put it mildly,
rather annoying.
|
|
o Move the sc->has_wep flag into sc->wi_flags
o Only do the COR reset on Symbol cards just to play it safe.
o Remove unused sc->sc_enabled variable.
o Kill ibss_portmap array in favor of sc->wi_ibss_port, which is
stored in little endian.
mickey@ OK
|
|
o Map port type 4 to ibss regardless of firmware type. This gives
us a consistent way to set ibss mode.
|
|
|
|
only call it from wi_attach() and wi_pcmcia_activate() (ie: just once).
It would be nicer to have the COR reset be part of the bus-specific
code but we need to know whether or not we have a Lucent card since
old Lucent firmware revs get messed up on a COR soft reset.
Even with the COR reset we still need to avoid initializing Symbol
cards more than once. However, we *do* want to do a reset after
returning from suspend. Therefore, rename wi_gone to wi_flags and
store both the attach and init status in it. wi_reset() now checks
wi_flags to see if the card should be initialized in the Symbol case.
Info on initializing Symbol cards once from NetBSD.
|
|
Also add some more Symbol-specific RID values I noticed in the
Linux orinoco driver.
|
|
out of generic wi_attach() and into wi_pci_attach(). This means
we can't print the CIS info for embedded cards like the NCP130 but
since those don't have a "real" pcmcia card in them this is really
not a problem. Also check to see that there is actually a pcmcia
present and give and error if not. CIS magic number taken from the
Linux orinoco driver; there may be a better way to do this.
|
|
firmware version. Adapted from NetBSD and the Linux orinoco driver.
The Symbol firmware requires that we do a COR soft reset before
initializing the card in wi_reset(). Also, setting WI_RID_ROAMING_MODE
caused the Symbol card to go into a strange (non-working) mode with
firmware 2.x.x.
There are some other Symbol additions/oddities that will need to be
addressed in the future.
|
|
|
|
is available everywhere wi_init() is called. deraadt@ OK
|
|
"ifconfig nwid FOO" work correctly in Host-AP mode.
|
|
wrong index used for RC4SWAP during key setup
IV is stored in little endian format; works for millert@
|
|
|
|
Only Lucent uses the 'variant' form. Also, prism2 firmware 0.8.2 does
WEP w/o requiring promisc mode (from NetBSD). OK mickey@
|
|
formatiing and byte-swapping fixens as well
|
