registers are registers that overlap with others on many controllers, but
which may actually be distinct on some controllers. Right now, the two
shadows are:
- wd_status (usually overlaps wd_command)
- wd_features (usually overlaps wd_error)
Add a new helper function, wdc_init_shadow_regs(), used to initialize
the shadow register handles on controllers where they do actually overlap.
Partially from Jordan Rhody @ Wasabi Systems, Inc.
Some ATAPI device never get out of busy if touched too fast after a reset.
Delay value from atapi_wdc.c; fix problem reported by Nicolas Joly on
current-users.
real save data pointers when we get the message (or rather, at disconnect
time following the message).
Factor out code to do this, and to deal with xs->resid, in siop_common.c.
What we do here seems to trigger interrupts on some pcmcia adapters, which
cause the kernel to hang.
Add some WDCDEBUG_PRINT((), DEBUG_PROBE).
Avoid touching registers of nonexistent drives, once we know which drive is
or is not here.
This makes the "IBM PCMCIA Portable CD-ROM Drive" (external CD drive with
PCMCIA adapter) work.
Some controllers/drives (e.g. SataLink 3114 with WD Raptor) require
it. Should fix kern/23808 by Chris Gilbert, patch suplied by Chris Gilbert
on tech-kern, extended to all places enabling interrupts by me.
log.
----------------------------
update copyright notice for 2004
----------------------------
add new statistics
Obtained from: madwifi
----------------------------
update radiotap support to reflect recent changes:
o move tx taps from ath_start to ath_tx_start so lots more
state is available to tap
o add tx flags
o add tx rate
o add tx power (constant for the moment)
o add tx antenna state
----------------------------
do proper subclassing of node free+copy; the previous hack falls apart when
the 802.11 layer does useful work
Obtained from: madwifi
----------------------------
transmit beacon frames directly instead of defering them to a swi; there
was too much delay
Obtained from: madwifi
----------------------------
update copyright notice for 2004
----------------------------
radiotap updates:
o force little-endian byte order for header
o pad header to 32-bit boundary to guard against applications that assume
packet data alignment
FreeBSD commit log:
----------------------------
use correct malloc type to allocate struct ieee80211_node's
Noticed by: phk
----------------------------
do proper subclassing of node free+copy; the previous hack falls apart when
the 802.11 layer does useful work
Obtained from: madwifi
----------------------------
transmit beacon frames directly instead of defering them to a swi; there
was too much delay
Obtained from: madwifi
----------------------------
update copyright notice for 2004
----------------------------
check more quickly (and directly) if an interrupt is pending; this reduces
work done in ath_intr when the irq is shared
Obtained from: madwifi
----------------------------
cleanup descriptor allocation if attach fails
Obtained from: madwifi
----------------------------
remove use IEEE80211_C_RCVMGT
----------------------------
radiotap updates:
o force little-endian byte order for header
o pad header to 32-bit boundary to guard against applications that assume
packet data alignment
----------------------------
Don't announce MAC addresses twice.
(ieee80211_ifattach() calls ether_ifattach().)
----------------------------
Make this compile on amd64.
"I'll cope" by: sam
----------------------------
When draining the tx queue reclaim any node references held in packets.
This fixes a problem when operating as an AP where clients would get
stuck in the node table because the reference count never went to zero.
----------------------------
When ath_hal_stoptxdma returns an error dma is still likely stopped
so don't just stop trying to send a beacon frame or we'll be more likely
to lose sync. This only seems to happen on some older chips.
----------------------------
use ath_reset instead of ath_init when recovering from a watchdog timeout:
resetting the hardware is sufficient, no need to reset the 802.11 fsm
----------------------------
make hw.ath.debug a tunable
----------------------------
make hw.ath.outdoor and hw.ath.countrycode tunables
----------------------------
split debugging messages up into classes;
ah_debug is now treated as a bit vector
----------------------------
update radiotap support to reflect recent changes:
o move tx taps from ath_start to ath_tx_start so lots more
state is available to tap
o add tx flags
o add tx rate
o add tx power (constant for the moment)
o add tx antenna state
----------------------------
o eliminate widespread on-stack mbuf use for bpf by introducing
a new bpf_mtap2 routine that does the right thing for an mbuf
and a variable-length chunk of data that should be prepended.
o while we're sweeping the drivers, use u_int32_t uniformly when
when prepending the address family (several places were assuming
sizeof(int) was 4)
o return M_ASSERTVALID to BPF_MTAP* now that all stack-allocated
mbufs have been eliminated; this may better be moved to the bpf
routines
on x86 boxes. There must be a better way to deal with this that works
well for everyone. In the meantime, back out to the version that works
for more people.
in tsleep()/wakeup() to appease gcc. Otherwise, the ch_flags value may
be cached in a register in atabus_thread(), and when it sets the WDCF_TH_RUN
bit after tsleep() it may loose loose the changes made by an interrupt handler
or another thread.
Problem analysed by Jukka Andberg on tech-kern.
THRE interrupt occurs between the LSR read and IIR read, we won't see the
LSR_TXRDY bit when testing it in the variable `lsr' and we don't interrupted
again (as the corresponding bit in the IIR is cleared by reading, except
for some broken device).
Tested by Matthias Scheler and me, reviewed by Allen Briggs.
Closes PR#25010.
it if we think it's probably necessary. Then do it again in the
tx interrupt handler, if we again think it's necessary. This
reduces the number of commands we issue the chip. Prior to this
change, the i82550 (running without extended feature set, so like
a '559) would sometimes fail in fxp_scb_wait() prior to issuing
another CU_RESUME, resulting in
fxp0: WARNING: SCB timed out!
messages on the console, often followed by device timeouts.
(tightly scoped) reason for recording the node address by recording
the assigned number. Dink pci/if_bge.c to match, since ic/ath.c was
used as the archetype.
commit. The rate adaptation code expects them. Usually wi gets
lucky, and an Rx/Alloc/Info event triggers the interrupt handler,
but I had not intended for wi to count on it.
Without Tx/TxExc interrupts enabled, wi will sometimes exhaust all
its rssdescs and cease transmitting. Usually it sets IFF_OACTIVE
in that situation.
(c) TNF
line from 4-clause UCB to 3-clause UCB license; in other words,
remove UCB's ad clause from the license TNF grants.
There is no point in TNF demanding that UCB's ad clause be followed
when even UCB doesn't demand it any longer.
Ok'd by board@ and agc@.
Figure out whether the shared memory region is word-accessible in the same
place we figure out its size, and store this in the softc, rather than using a
series of comparisons later.
I have pulled that code into the function ieee80211_rssadapt_choose
so that I can re-use it in ath(4), atw(4), and in other drivers.
In rssadapt(9), I have also created a struct ieee80211_rssadapt_expavgctl
that contains parameters for rate adaptation. When IEEE80211_RSSADAPT_DEBUG
is enabled, I will using sysctl to expose an ieee80211_rssadapt_expavgctl
for each wireless device.
Also in rssadapt(9), I have introduced an interpolate() macro which
makes the exponential-averaging code more compact.
