If a card stops generating interrupts for any reason, it can take awhile to
unload the driver while waiting for the commands to timeout. Now if there is a
problem, immediately set a bit that the card is failed so that no new commands
will be submitted.
Also, do not free and commands in submit since all commands are freed in get
results now. This prevents list/memory corruption when commands time out.
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Signed-off-by: Russ Meyerriecks <rmeyerriecks@digium.com>
When spans are unassigned, dahdi_span_ops.shutdown was called, but the RUNNING
flag was never cleared. Now make sure all calls to the shutdown span ops
callback are the same.
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Signed-off-by: Russ Meyerriecks <rmeyerriecks@digium.com>
If you have a multiple span system configured for t1 mode, and you try to change
only the first span from t1 to e1 via sysfs, you could get an error in the
kernel log about trying to create duplicate channel.
The problem is that the check for whether there was enough room for the all the
channels of the switched span was wrong.
Reported-by: James Brown <jbrown@digium.com>
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Signed-off-by: Russ Meyerriecks <rmeyerriecks@digium.com>
This brings the wcte12xp driver in line with the wct4xxp driver. It also
eliminates the chance that the local side will errouneously go into loopup when
switched from E1 to T1 mode via the spantype sysfs attribute.
Internal-Issue-ID: DAHDI-1038
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Signed-off-by: Russ Meyerriecks <rmeyerriecks@digium.com>
This adds definitions for bool and pr_fmt which are needed when compiling
against vanilla 2.6.18.
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Signed-off-by: Russ Meyerriecks <rmeyerriecks@digium.com>
In 2.6.2 I introduced an error in (41639330a "wctdm24xxp: Eliminate chance for
channel to be stuck in RED alarm.") which would get an FXO stuck in detecting an
ONHOOK event from the remote side. An FXO port could go into the
BATTERY_DEBOUNCING_LOST_ALARM state and then back to the BATTERY_PRESENT state
without sending the dahdi_hooksig up to Asterisk. Therefore, Asterisk would
never detect the state changes needed for a remote side disconnect.
This change adds two more states, BATTERY_DEBOUNCING_PRESENT_FROM_LOST_ALARM and
BATTERY_DEBOUNCING_LOST_FROM_PRESENT_ALARM so that the state machine does not
miss any needed transitions when going back to BATTERY_PRESENT or BATTERY_LOST
regardless of why the debounce was started.
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Signed-off-by: Russ Meyerriecks <rmeyerriecks@digium.com>
The channels now have embedded 'struct devices' that contain a device_private
pointer in which to store the device private data. When a device is
unregistered in the system the device privata data is freed by the device core,
even though the chan_release function doesn't free the memory
associated with the device, So when dev_set_drvdata() is called,
it's writing into freed memory.
We also need zero out the embedded struct device before registration,
since we do not have any guarantee that it points to valid memory (or, if a
channel was unregisterd with sysfs, and then reregistered without being cleared
out). Otherwise the core could try to use the previously freed private
data again.
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Signed-off-by: Russ Meyerriecks <rmeyerriecks@digium.com>
If a user switches linemode with sysfs, the VPM would never switch the
companding mode to alaw, and the result would be extreme static on all channels
of the span when the VPM is enabeld.
Now if the deflaw of the span has changed when an echocan is created, the
companding mode of the VPM channel is changed.
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Soon there will be more than one driver in the source tree that will want to use
these files. Compiling it as a library speeds the build since it won't have to
be built for each driver that wants to link it in.
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Signed-off-by: Russ Meyerriecks <rmeyerriecks@digium.com>
Some of the card callbacks needed maintstat set when processing loopup codes
from the remote side. This change is a modification of commit: (6c02c3c
"dahdi-base: Minor maint mode error")
Without this change cetain cards would not properly process the loopup /
loopdown codes (I know I should have more specific information here but the
details escape me at the moment).
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
This was an accidental commit that slipped in as part of (a682401 "dahdi: Expose
dahdi devices in sysfs.")
Signed-off-by: Russ Meyerriecks <rmeyerriecks@digium.com>
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Adds support for setting a value of 3 to the module parameter
caller_id_style: passthrough - always pass and don't try to emulate
DTMFs.
Signed-off-by: Tzafrir Cohen <tzafrir.cohen@xorcom.com>
An unset conditional compile flag was triggering the unused variable compile
warning. Added the condition around the variable define.
Signed-off-by: Russ Meyerriecks <rmeyerriecks@digium.com>
Fixes up the kbuild to work with compiling OSLEC from the kernel source. See
HOWTO here: http://forums.digium.com/viewtopic.php?t=67164
Internal-Issue-ID: DAHLIN-317
Reported-By: Vladimir Mikhelson
Signed-off-by: Russ Meyerriecks <rmeyerriecks@digium.com>
Quote: "It's a change. People hate change"
Make the version string say something like 2.6.2 instead of v2.6.2.
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Otherwise, it's possible for a link to remain in use if a process quits without
unlinking the conferences. Now when all the channels are removed from the
conference, any links are also cleaned up.
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
This is mostly a revert of commit r9463. If you need to use DAHDI_CONFLINK
ioctl, make sure to define CONFIG_DAHDI_CONFLINK in
include/dahdi/dahdi_config.h. Apparently there were some users of CONFLINK out
there still.
It's a compile time option now since most users won't need to run the test for
conflinks in the hot-path that is the process_masterspan function.
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Tested-by: Ted Gerold <ted@twg.org>
Trades the memory of a file_operations structure to eliminate some tests in the
timer operations.
Internal-Issue-ID: ABE-2904
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
On SMP systems there isn't a need for all dahdi timers to
synchronize on dahdi_timer_lock. Instead give each timer it's own
lock to synchronize with process_timers and use dahdi_timer_lock to
protect the global list of locks.
Also, the dahdi_timers list now only contains timers that are set to
fire. This eliminates a test for each timer in the hot path of the
process_masterspan context.
Reduces system load on many-core systems which are heavy users of
DAHDI timers.
Internal-Issue-ID: ABE-2904
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
The Big Kernel Lock removal is no longer experimental so we can eliminate the
call to lock_kernel when calling unlock_ioctl.
Internal-Issue-ID: ABE-2904
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Also, if there is no other version information use the directory name. Downloads
from gitweb will include the sha information in the build and otherwise a user
could locate the source directory via the embedded version information. I
believe this is better than an empty string.
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
The __dev* directives and functions were removed in 3.8, as they
are no-ops. We still have use of them for older versions, thus
we should define them (as noops) if they don't exist.
Signed-off-by: Tzafrir Cohen <tzafrir.cohen@xorcom.com>
There was a code patch where it was possible to get stuck in RED ALARM on a
channel when debouncing the battery states. The state transitions would look
like this:
BATTERY_PRESENT -> BATTERY_DEBOUNCING_LOST -> BATTERY_DEBOUNCING_LOST_ALARM --
(send alarm up to asterisk) --> BATTERY_LOST -> BATTERY_DEBOUNCING_PRESENT ->
BATTERY_DEBOUNCING_PRESENT_ALARM -> BATTERY_DEBOUNCING_LOST -> BATTERY_PRESENT
In the above sequence there was never any transition from
BATTERY_DEBOUNCING_PRESENT_ALARM to BATTERY_PRESENT so the alarm to Asterisk was
never cleared and the channel stayed stuck.
Now when you loose battery when in the BATTERY_DEBOUNCING_PRESENT_ALARM go all
the way back to the BATTERY_LOST state instead of the BATTERY_DEBOUNCING_LOST
state so that all the events are properly sent up.
This fixes a regression introduced in 2.6.0 with commit (r10169 "wctdm24xxp: Use
interval for debouncing FXO battery." 874b76bd22).
Internal-Issue-ID: DAHDI-1019
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
The logic to check for battery lost and battery present were using different
time bases. One was using jiffies and the other was using framecount. Since
framecount is always in milliseconds, let's use that to stay consistent.
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
hfc_decode_st_state() will be called from interrupt context when the debug flag
is set to 32. Therefore, must use GFP_ATOMIC when allocating memory.
Only affects the wcb4xxp driver when called with particular debug flags set.
Internal-Issue-ID: DAHLIN-314
Reported-by: Gerald Schnabel
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
This is necessary to prevent a crash when opening files by the new
device files, which do not have a valid file_operations structure.
This fixes a problem does not exist in any releases of DAHDI.
Since none of the existing tools or applications open files from the
new location of /dev/dahdi/channels/<span>/<offset>, this wasn't
seen until just recently.
Reported-by: Ted Gerold <ted@twg.org>
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Without digging into the specifics, it looks like Red Hat Linux 5.9
removed the hex_asc definition that was previously used to determine
if the bool definition was backported.
We can simply use the RHEL_RELEASE_CODE now since we do not support any
releases before the 5 series now.
Reported-By: Vladimir Mikhelson
Internal-Issue-ID: DAHLIN-312
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Acked-By: Tzafrir Cohen <tzafrir.cohen@xorcom.com>
It's wrong (and explicit oops with kernel >= 2.6.19) to set both
bus and class for the same device. But setting the class is not needed
in order to create a device. So just remove our dynamic devices from the
dahdi_class. It will only contain the fixed-named devices.
This fixes regression from cb4e4d0068.
Signed-off-by: Tzafrir Cohen <tzafrir.cohen@xorcom.com>
The driver iterates through all the spans on a given device during assignment,
checking for unassigned spans, but it was erroneously testing the span on which
assigned was called.
This just removes some unexpected behavior and provides a slight performance
increase on load and does not impact the functionality of the driver as far as
I'm aware.
Reported-by: Doug Bailey <dbailey@digium.com>
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Eliminates the following warning from sparse:
drivers/dahdi/dahdi-sysfs-chan.c:50:29: error: dubious one-bit signed bitfield
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
struct dahdi_count is not directly exposed to user space, so we can use the
native u32 type instead of __u32 to clarify that.
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
This allows timingslips to be reset along with the other counters and clarifies
the intended use.
This came up when Doug Bailey asked why he couldn't use dahdi_maint to clear
timing slips in addition to the other counters.
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
* Kernel (and SysFS) objects for channels.
* A bus of their own.
* Dynamically allocate a minor (no conflict with /dev/dahdi/timer et. al.).
* Device files located under /dev/dahdi/channels instead of
/dev/dahdi/spans [e.g: /dev/dahdi/channels/2/4)
- A link back should is doable with a simple udev rule.
* Bus name is dahdi_channels instead of dahdi_chans
[e.g: /sys/bus/dahdi_channels]
* Driver name is dahdi instead of dahdi_chans
[e.g: /sys/bus/dahdi_channels/drivers/dahdi]
* Attributes are all strings (or lists of strings), including 'sig' and
'sigcap'.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)
iQEcBAABCAAGBQJQyL3jAAoJEMZ1bgEN3R/0YHEH/RFKSftR0GWNo8TkRbDnojjK
bU9YwnLucAKe2AgwRKcKLStmEHYK/e2Jar1TVuDBlgvE3hdYRzddjbB045Ff4un2
0483Yx3JMQFYQep3W5DNpT1R7YU9rn9iurObZXnXryWe0Zxw57sKC7ixCl3tEv0m
x58cl9UDMhknf+LtY8b/QNMneIu0MbzBgxiow6yj4g8x9Vv1obnPFbfui871Yvlt
2H4LdVchZCfmYYTcT93AxfV6G/1RuF9wWGbSUN+VigcXuvhUWERsjSVlq0c8GAWO
pocXlxxwN1W8gYC/yhNKKLZHVTBi8YLvNAjUCCr/N7JxX9YnMljCcBjuiljDVlg=
=sZHz
-----END PGP SIGNATURE-----
Merge tag 'review-sysfs-chan' of http://git.tzafrir.org.il/git/dahdi-linux into for-trunk
SysFS Channel representation changes
* Kernel (and SysFS) objects for channels.
* A bus of their own.
* Dynamically allocate a minor (no conflict with /dev/dahdi/timer et. al.).
* Device files located under /dev/dahdi/channels instead of
/dev/dahdi/spans [e.g: /dev/dahdi/channels/2/4)
- A link back should is doable with a simple udev rule.
* Bus name is dahdi_channels instead of dahdi_chans
[e.g: /sys/bus/dahdi_channels]
* Driver name is dahdi instead of dahdi_chans
[e.g: /sys/bus/dahdi_channels/drivers/dahdi]
* Attributes are all strings (or lists of strings), including 'sig' and
'sigcap'.
* Channels have their own dynamically allocated major number
* No arbitrary limit of minor < 250 (no collision with /dev/timer, etc)
* We still have arbitrary limit of channo < DAHDI_MAX_CHANNELS
(should be raised?)
* FIXME: need to check to FIXME's
Signed-off-by: Oron Peled <oron.peled@xorcom.com>
* Added minimal infrastructure:
- A 'chan_device' member to struct dahdi_chan
- An empty 'device_attribute' array
- A 'bus_type' with its methods
- A 'device_driver' with its methods
- Initialization/Cleanup code
Signed-off-by: Oron Peled <oron.peled@xorcom.com>
This resolves an issue with a Glenayre GL3000 Paging Terminal with v8.000
software. The Glenayre would change both A and B bits to signal on/off hook
states, but there were a couple of milliseconds between when those bits changed.
This resulted in DAHDI generating an on-hook event to Asterisk before generating
the off-hook event and therefore Asterisk terminated the call prematurely.
Looking the A and B bits before this patch:
Asterisk (AB) Glenayre (AB)
00 00 Both sides on-hook
11 00 Asterisk goes off hook to sieze line.
11 01 Glenayre starts going-offhook. On-hook event
sent to Asterisk from drivers since bits
changed but ABIT was still 0.
11 11 Glenayre finishes going off-hook. Off-hook
event sent to Asterisk since ABIT changed
from 0 to 1.
00 11 Asterisk, processes on-hook event and goes
on-hook itself to release line.
00 00 Glenayre releases line. Call fails.
After this patch:
Asterisk (AB) Glenayre (AB)
00 00 Both sides on-hook
11 00 Asterisk initiates call
11 01 Glenayre starts handling call. No event is
sent to Asterisk since only ABIT is checked.
11 11 Glanayre finishes going off-hook. Call
proceeds normally.
Internal-Issue-ID: DAHDI-1009
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Signed-off-by: Russ Meyerriecks <rmeyerriecks@digium.com>
dahdi_dynamic now always calls the flush function in softirq context so packet
flushing no longer needs to be pushed off to process context since interrupts
are enabled.
Signed-off-by: Shaun Ruffell <sruffell@digium.com>
Signed-off-by: Russ Meyerriecks <rmeyerriecks@digium.com>
