Merged revisions 6699,6706,6714,6768,6771,6785,6812-6818,6821,6838,6928-6929,6941,6945-6946,6952,6981-6982,7003-7004,7008,7023,7027,7094-7097,7110,7117-7118,7125,7140,7147,7155,7194,7284,7293,7319,7437-7438,7445,7480,7486-7487,7512-7524,7527-7528,7534-7536 via svnmerge from

https://origsvn.digium.com/svn/dahdi/linux/trunk

........
  r6699 | kpfleming | 2009-06-23 15:32:01 -0500 (Tue, 23 Jun 2009) | 3 lines
  
  Use the same mutex lock for channel allocation and license checking in dahdi_echocan_hpec, so that channel allocation won't happen while the license is being checked (or rechecked)
........
  r6706 | sruffell | 2009-06-23 18:21:25 -0500 (Tue, 23 Jun 2009) | 1 line
  
  README: Minor corrections to the README.
........
  r6714 | twilson | 2009-06-24 15:23:07 -0500 (Wed, 24 Jun 2009) | 1 line
  
  Remove unused torisa code from header
........
  r6768 | sruffell | 2009-06-26 12:07:39 -0500 (Fri, 26 Jun 2009) | 1 line
  
  wct4xxp: Unmap the same size DMA window that was mapped.
........
  r6771 | sruffell | 2009-06-26 12:22:45 -0500 (Fri, 26 Jun 2009) | 6 lines
  
  dahdi_dynamic: Release the dlock before calling accross modules.
  
  Resolves a hard lock due to a recursive spinlock grab at startup.
  
  Reported by: mapacheco
  (closes issue #15210)
........
  r6785 | sruffell | 2009-06-28 23:47:26 -0500 (Sun, 28 Jun 2009) | 5 lines
  
  echocan: Properly keep the reference counts for the echocan modules.
  
  (closes issue #13504)
  (closes issue #15327)
  Reported by: sruffell, tzafrir
........
  r6812 | tzafrir | 2009-07-02 14:25:54 -0500 (Thu, 02 Jul 2009) | 5 lines
  
  wcb4xxp: Don't assume we have 4 spans (ports).
  
  First part of extra hfcmulti drivers: the number of ports is still
  hardwired to 4, but just in a single place.
........
  r6813 | tzafrir | 2009-07-02 14:34:05 -0500 (Thu, 02 Jul 2009) | 4 lines
  
  wcb4xxp: Extra bit macros that a needed for kernels < 2.6.24
  
  Macros borrowed from xpp/ .
........
  r6814 | tzafrir | 2009-07-02 14:42:25 -0500 (Thu, 02 Jul 2009) | 2 lines
  
  wcb4xxp: Adjust debug filter code to number of ports.
........
  r6815 | tzafrir | 2009-07-02 14:52:14 -0500 (Thu, 02 Jul 2009) | 6 lines
  
  wcb4xxp: support for other HFC-xS cards (info, not code)
  
  This commit includes skeleton for the support of other HFC-[248]S-based
  cards.  It still does not include all the different cases for different
  cards.
........
  r6816 | tzafrir | 2009-07-02 14:59:44 -0500 (Thu, 02 Jul 2009) | 2 lines
  
  The B410P differs from other HFC-xS cards with respect to the EC unit
........
  r6817 | tzafrir | 2009-07-02 15:07:36 -0500 (Thu, 02 Jul 2009) | 7 lines
  
  More B410P differences: Clock and NT/TE.
  
  * The B410P reads the NT/TE switches the other way around from other
    cards.
  * Its clock is also 1/2 of that of other cards, which causes wierd PCM 
    on an unmodified driver.
........
  r6818 | tzafrir | 2009-07-02 15:14:39 -0500 (Thu, 02 Jul 2009) | 4 lines
  
  wcb4xxp: Fix PCM handling for various cards.
  
  HFC-8S cards behave quite differently than HFC-4S cards here.
........
  r6821 | tzafrir | 2009-07-02 15:18:59 -0500 (Thu, 02 Jul 2009) | 6 lines
  
  wcb4xxp: Fix LED handling in OpenVox cards (maybe also others)
  
  This commit adds extra functions to handle LEDs in the non-B410P cards.
  Only tested on OpenVox cards. OpenVox cards are known to have slightly
  different LEDs so this is likely to be slightly broken for others.
........
  r6838 | sruffell | 2009-07-13 09:33:39 -0500 (Mon, 13 Jul 2009) | 1 line
  
  wctc4xxp: Remove deprecated DMA_xxBIT_MASK usage.
........
  r6928 | sruffell | 2009-08-04 11:22:23 -0500 (Tue, 04 Aug 2009) | 1 line
  
  wcte12xp: Remove unused 'schluffen' declaration in wcte12xp.h
........
  r6929 | sruffell | 2009-08-04 11:22:26 -0500 (Tue, 04 Aug 2009) | 4 lines
  
  wctdm24xxp: Eliminate 'variety' and 'type' members from 'struct wctdm'.
  
  struct wctdm can hold a pointer to struct wctdm_desc directly, and eliminate
  the need to copy members of wctdm_desc into wctdm.
........
  r6941 | dbailey | 2009-08-05 09:40:45 -0500 (Wed, 05 Aug 2009) | 14 lines
  
  Change proslic linefeed register setting
  
  Insure that proslic linefeed register is not transitioned from Active to
  On-Hook Transmission while the channel is off-hook.
  Replaced magic numbers assigned to linefeed associated variables with more
  descriptive constants. 
  
  (issue #15352)
  Reported by: alecdavis
  Patches:
        wctdm_prevent_ohttimer_click.diff3.txt uploaded by dbailey (license 819)
  Tested by: alecdavis, dbailey, vmikhelson
........
  r6945 | sruffell | 2009-08-05 14:39:10 -0500 (Wed, 05 Aug 2009) | 6 lines
  
  wctdm24xxp, wctdm: Formatting changes.
  
  Update the lines affected by revision 6941. I'm taking every opportunity
  to move DAHDI closer to the kernel coding conventions short of just
  reformatting for the sake of reformatting. The majority of these changes
  are to bring the line lengths under 80 chars.
........
  r6946 | sruffell | 2009-08-05 14:39:13 -0500 (Wed, 05 Aug 2009) | 1 line
  
  dahdi_dummy: Remove some trailing whitespace.
........
  r6952 | sruffell | 2009-08-11 13:47:21 -0500 (Tue, 11 Aug 2009) | 3 lines
  
  wctdm24xxp: Remove unused members related to hardware DTMF detection.
  
  Not used anymore, so they are gone.
........
  r6981 | sruffell | 2009-08-13 09:42:05 -0500 (Thu, 13 Aug 2009) | 1 line
  
  wctc4xxp: Remove flag member that is not used.
........
  r6982 | sruffell | 2009-08-13 09:42:08 -0500 (Thu, 13 Aug 2009) | 1 line
  
  wctdm24xxp: Use the ARRAY_SIZE macro where appropriate.
........
  r7003 | sruffell | 2009-08-13 19:46:26 -0500 (Thu, 13 Aug 2009) | 9 lines
  
  wctdm24xxp, wctdm: Detect if our hookstate has been set back to the initial state.
  
  Check if our hookstate has been set back to the initial state, typically the
  result of a chanconfig, and if so, if we're an FXO port, forget our current
  battery state.  This allows the driver to determine and report again what the
  hook state of the port is.
  
  (related to issue #14577)
  (closes issue #15429)
........
  r7004 | sruffell | 2009-08-13 19:46:28 -0500 (Thu, 13 Aug 2009) | 1 line
  
  dahdi-base: Add comment to explain why rxhooksig is reset on span start.
........
  r7008 | sruffell | 2009-08-14 10:47:39 -0500 (Fri, 14 Aug 2009) | 11 lines
  
  wcfxo: Reset the DAA on module initialization.
  
  The X100p and clones will sometimes work and sometimes not depending on wether
  the DAA powers up in running state- this seems to be related to the power
  supply. This problem is caused by the driver not reseting the DAA and may be
  the source of a great many intermittent problems with this card.
  
  (closes issue #14232)
  Reported by: tallen8840
  Patch by: tallen8840
  Tested by: explidous, Flavio
........
  r7023 | sruffell | 2009-08-17 09:07:06 -0500 (Mon, 17 Aug 2009) | 4 lines
  
  README: Update known issues section.
  
  Remove note about echocanceler reference counts, since that is fixed, and add
  a note about issue with KB1 when configured with more than 128 taps.
........
  r7027 | seanbright | 2009-08-17 14:31:54 -0500 (Mon, 17 Aug 2009) | 7 lines
  
  Silence spurious warnings when trying to remove Zaptel directories during install.
  
  (closes issue #15479)
  Reported by: pprindeville
  Patches:
        dahdi-linux-rm.patch uploaded by pprindeville (license 347)
........
  r7094 | sruffell | 2009-09-07 16:40:19 -0500 (Mon, 07 Sep 2009) | 4 lines
  
  wctdm24xxp: Remove a few more unneeded 'volatile' keywords.
  
  The writechunk and readchunk parameters are never accessed by hardware at
  the same time that the software is accessing them anymore.
........
  r7095 | sruffell | 2009-09-07 16:40:22 -0500 (Mon, 07 Sep 2009) | 5 lines
  
  dahdi-base: Reduce the stack usage of dahdi_common_ioctl.
  
  Split the DAHDI_GETGAINS and DAHDI_SETGAINS ioctls into their own functions
  and dynamically allocate the 'struct dahdi_gains' structure to reduce the
  pressure on the stack.
........
  r7096 | sruffell | 2009-09-07 16:40:25 -0500 (Mon, 07 Sep 2009) | 1 line
  
  dahdi-base: Prevent compilation if both EMPULSE and EMFLASH are defined.
........
  r7097 | sruffell | 2009-09-07 16:40:29 -0500 (Mon, 07 Sep 2009) | 4 lines
  
  dahdi-base: Fix flag check in dahdi_rbs_sethook.
  
  As long as any of the flags were set, this check would have always
  passed.
........
  r7110 | rmeyerriecks | 2009-09-14 15:30:24 -0500 (Mon, 14 Sep 2009) | 1 line
  
  Fixed issue where the clear channel flags were not being set at the appropriate time causing a channel re-configure to mis-set the last channel in each span
........
  r7117 | dbailey | 2009-09-14 15:51:56 -0500 (Mon, 14 Sep 2009) | 17 lines
  
  Race condition in handling writes to proslic LINEFEED register (64)
  
  The wctdm24xxp driver has a problem where a VMWI IOCTL call followed immediately
  by a ONHOOKTRANSFER IOCTL call will cause the ONHOOK transfer request to be
  dropped. This occurs if the write to the proslic's LINEFEED register for the
  VMWI ICTL call is not completed when the ONHOOK transfer request IOCTL is
  processed.
  
  I also cleaned out some magic numbers used in setting the linefeed register.
  
  (closes issue #15875)
  Reported by: dbailey
  Patches:
        15875-wctdm24xxp.diff uploaded by dbailey (license 819)
  Tested by: dbailey
........
  r7118 | seanbright | 2009-09-14 16:10:38 -0500 (Mon, 14 Sep 2009) | 1 line
  
  Change zap -> dahdi.
........
  r7125 | rmeyerriecks | 2009-09-15 09:59:06 -0500 (Tue, 15 Sep 2009) | 2 lines
  
  dahdi-base: Minor syntax change to meet style guidelines
........
  r7140 | dbailey | 2009-09-15 15:50:45 -0500 (Tue, 15 Sep 2009) | 12 lines
  
  Change WCTDM SPI clock off state polarity and read timing
  
  Change the off state of the SPI clock to high and provide more time for data to
  settle out on SPI reads.
  
  (closes issue #15261)
  Reported by: alecdavis
  Patches:
        wctdm_spi_clocking.diff2.txt uploaded by alecdavis (license 585)
  Tested by: alecdavis, dbailey
........
  r7147 | sruffell | 2009-09-16 13:19:00 -0500 (Wed, 16 Sep 2009) | 4 lines
  
  wct4xxp: Check the alarm state if we're debouncing a red alarm.
  
  This fixes a problem where if you set the alarmdebounce module parameter on
  gen2+ cards, you never detect when you go into red alarm.
........
  r7155 | sruffell | 2009-09-21 10:24:36 -0500 (Mon, 21 Sep 2009) | 8 lines
  
  dahdi-base: dahdi_ioctl_[get|set]gains should return the res value.
  
  In function dahdi_ioctl_getgains() and dahdi_ioctl_setgains() return value
  assigned to res variable, but these function always return 0 which is an
  error.
  
  (closes issue #15916.)
  Patch by: ys
........
  r7194 | dbailey | 2009-09-22 09:03:53 -0500 (Tue, 22 Sep 2009) | 12 lines
  
  wctdm: Add missing break
  
  A break was missing that caused DAHDI_ONHOOKTRANSFER ioctl call to fall into 
  DAHDI_SETPOLARITY ioctl call. 
  
  (issue #14261)
  Reported by: alecdavis
  Patches:
        wctdm_fix_ONHOOKTRANSFER.diff.txt uploaded by alecdavis (license 585)
  Tested by: alecdavis
........
  r7284 | mattf | 2009-09-30 11:34:11 -0500 (Wed, 30 Sep 2009) | 1 line
  
  Update echocan API so it only uses channel offset in free routine
........
  r7293 | tzafrir | 2009-09-30 13:09:42 -0500 (Wed, 30 Sep 2009) | 2 lines
  
  move the dev->bus_id fix from xpp to kernel.h: needed elsewhere
........
  r7319 | sruffell | 2009-10-02 16:09:01 -0500 (Fri, 02 Oct 2009) | 1 line
  
  wcte12xp: The timer is called every 200ms, not every 100ms. Fix comment.
........
  r7437 | sruffell | 2009-10-29 13:26:16 -0500 (Thu, 29 Oct 2009) | 10 lines
  
  dahdi-base: Do not allow jumps in system time to lock up the system w/core_timer
  
  Since dahdi coretimer uses the number of milliseconds that has actually passed
  to determine how many times to call dahdi_receive, it is possible that if the
  system time shifts after dahdi is started, that the system can appear to lock
  up while the core timer attempts to catch up.  This change prevents soft lock
  ups under these conditions.  This is brings the dahdi_dummy changes in r6933
  into dahdi-base.
  
  (related to issue #15647)
........
  r7438 | sruffell | 2009-10-29 13:26:17 -0500 (Thu, 29 Oct 2009) | 1 line
  
  wcte12xp, wctdm24xxp: VPMADT032 firmware update to 1.20.
........
  r7445 | mspiceland | 2009-10-29 16:37:45 -0500 (Thu, 29 Oct 2009) | 3 lines
  
  Debounce alarms by default for wct4xxp per AT&T 54016.  Also, the
  various alarm conditions can be debounced separately.
........
  r7480 | sruffell | 2009-11-04 14:43:05 -0600 (Wed, 04 Nov 2009) | 4 lines
  
  voicebus: Increase the NLP converged threshold to 18.
  
  Brings in the change from r7065 that was on the team/sruffell/dahdi-linux-vpm119
  branch.
........
  r7486 | mspiceland | 2009-11-04 17:25:32 -0600 (Wed, 04 Nov 2009) | 9 lines
  
  Adding alarm debounce to single span driver (wcte12xp).
  
  Debounce yellow alarm also.
  
  In wcte12xp, change check alarm frequency to 100ms for better debounce
  granularity.
  
  Fix lines over 80 cols from last alarm debounce commit.
........
  r7487 | mspiceland | 2009-11-04 17:28:21 -0600 (Wed, 04 Nov 2009) | 2 lines
  
  Remove commented out code block that was unintentionally left in.
........
  r7512 | sruffell | 2009-11-06 18:35:38 -0600 (Fri, 06 Nov 2009) | 1 line
  
  wcte12xp: Remove unused flag member and make const the t1_descs.
........
  r7513 | sruffell | 2009-11-06 18:35:38 -0600 (Fri, 06 Nov 2009) | 4 lines
  
  voicebus: Remove the VB_PRINTK macro.
  
  Unnecessarily duplicates the dev_xxx macros.  Also removes the need for the
  board_name member from struct voicebus.
........
  r7514 | sruffell | 2009-11-06 18:35:39 -0600 (Fri, 06 Nov 2009) | 4 lines
  
  voicebus: Remove sdi member from 'struct voicebus'
  
  This is only used during startup so we don't need to carry it around in the
  structure at all times.
........
  r7515 | sruffell | 2009-11-06 18:35:40 -0600 (Fri, 06 Nov 2009) | 1 line
  
  voicebus: Set the DMA_BIT_MASK
........
  r7516 | sruffell | 2009-11-06 18:35:40 -0600 (Fri, 06 Nov 2009) | 1 line
  
  voicebus: Use DAHDI_IRQ_SHARED instead of defining our own.
........
  r7517 | sruffell | 2009-11-06 18:35:40 -0600 (Fri, 06 Nov 2009) | 24 lines
  
  voicebus: Send 'idle' buffers when the transmit descriptor underruns.
  
  Previously, when the host system fails to service the interrupt in a timely
  fashion, the transmit descriptor ring for the voicebus card would "go empty"
  since the interface wouldn't have another descriptor to read in.  The driver
  only knows that it went empty, not how far behind it actually was.  Therefore,
  the driver could just increase the latency by a millisecond and keep going
  waiting for another bump.
  
  Additionally, when the transmit descriptor actually goes empty, there are some
  cases where an in process SPI transaction to one of the modules is interrupted,
  which may result in corrupted module register writes on rare occassions.
  
  This now makes it possible for the voicebus drivers to coexist with some devices
  that periodically lock interrupts for longer than 25ms.  Before this patch, the
  latency would constantly increase until either the modules received a corrupted
  frame.
  
  This patch preconfigures all the receive descriptors to send an "idle" packet
  that will be transmitted to the onboard modules when the host doesn't service
  the interrupt within (latency - 2)ms.  There are now two kinds of underruns,
  softunderuns where the driver can detect that these idlebuffers have made it to
  the TX FIFO, and the normal hard underrun where the part signals a transmit
  descriptor unavailable interrupt.  DAHDI-278.
........
  r7518 | sruffell | 2009-11-06 18:35:41 -0600 (Fri, 06 Nov 2009) | 7 lines
  
  voicebus: Add function to lock the latency.
  
  Now that increases in the latency produce less undefined behavior on the SPI
  busses, provide an interface for client drivers to inform the voicebus library
  to not increase the latency if underruns are detected. This can speed up loads
  of the driver since latency bumps do not trigger a restart of the driver
  initialization. DAHDI-278.
........
  r7519 | sruffell | 2009-11-06 18:35:42 -0600 (Fri, 06 Nov 2009) | 4 lines
  
  wcte12xp: Lock latency when loading
  
  No longer need to restart board initialization if the latency would have
  increased during initialization. DAHDI-278.
........
  r7520 | sruffell | 2009-11-06 18:35:42 -0600 (Fri, 06 Nov 2009) | 4 lines
  
  wctdm24xxp: Lock latency when loading
  
  We no longer need to retry board initialization if the latency would have
  increased during the initialization. DAHDI-278
........
  r7521 | sruffell | 2009-11-06 18:35:43 -0600 (Fri, 06 Nov 2009) | 4 lines
  
  voicebus, wctdm24xxp, wcte12xp: Move a print out of the interrupt handler.
  
  This can be handled just as well in process context and printing to a serial
  console from the interrupt handler has the potential to cause long latencies.
........
  r7522 | sruffell | 2009-11-06 18:35:44 -0600 (Fri, 06 Nov 2009) | 5 lines
  
  voicebus: Add optional sysfs entry for reading a boards current latency.
  
  This is off by default since it hasn't been tested on a full range of kernels,
  but can be useful for quickly seeing differences for latencies on different
  cards installed in the system.
........
  r7523 | sruffell | 2009-11-06 18:35:44 -0600 (Fri, 06 Nov 2009) | 1 line
  
  voicebus: Remove 'assert' macros and use BUG_ON/WARN_ON directly.
........
  r7524 | sruffell | 2009-11-06 18:35:45 -0600 (Fri, 06 Nov 2009) | 1 line
  
  voicebus: Be just a little more graceful if we cannot grab our interrupt line.
........
  r7527 | sruffell | 2009-11-06 18:58:03 -0600 (Fri, 06 Nov 2009) | 1 line
  
  wcte12xp: Fix up some continued strings.
........
  r7528 | sruffell | 2009-11-06 18:58:03 -0600 (Fri, 06 Nov 2009) | 1 line
  
  wct4xxp: Fix up some continued strings.
........
  r7534 | sruffell | 2009-11-09 12:02:40 -0600 (Mon, 09 Nov 2009) | 1 line
  
  wct4xxp: Only print the new debounce messages when debug is set.
........
  r7535 | sruffell | 2009-11-09 12:02:41 -0600 (Mon, 09 Nov 2009) | 1 line
  
  wcte12xp: Only print the new debounce messages when debug is set.
........
  r7536 | sruffell | 2009-11-09 12:11:06 -0600 (Mon, 09 Nov 2009) | 4 lines
  
  wcte12xp: use the dev_xxx macro for the debounce messages.
  
  We want to know which device is reporting the debounce when there are more than
  one card in the system.
........


git-svn-id: http://svn.asterisk.org/svn/dahdi/linux/branches/2.2@7539 a0bf4364-ded3-4de4-8d8a-66a801d63aff

2 files changed, 10 insertions, 13 deletions

diff --git a/include/dahdi/kernel.h b/include/dahdi/kernel.h
index 0731056..6058779 100644
--- a/include/dahdi/kernel.h
+++ b/include/dahdi/kernel.h
@@ -85,6 +85,12 @@
 #endif
 #endif
 
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,26)
+#define dev_name(dev)		(dev)->bus_id
+#define dev_set_name(dev, format, ...) \
+	snprintf((dev)->bus_id, BUS_ID_SIZE, format, ## __VA_ARGS__);
+#endif
+
 /*! Default chunk size for conferences and such -- static right now, might make
    variable sometime.  8 samples = 1 ms = most frequent service interval possible
    for a USB device */
@@ -1166,4 +1172,8 @@ static inline short dahdi_txtone_nextsample(struct dahdi_chan *ss)
 #define kzalloc(a, b) kcalloc(1, a, b)
 #endif
 
+#ifndef DMA_BIT_MASK
+#define DMA_BIT_MASK(n)	(((n) == 64) ? ~0ULL : ((1ULL<<(n))-1))
+#endif
+
 #endif /* _DAHDI_KERNEL_H */
diff --git a/include/dahdi/user.h b/include/dahdi/user.h
index 5c7121a..1d0570a 100644
--- a/include/dahdi/user.h
+++ b/include/dahdi/user.h
@@ -1027,19 +1027,6 @@ struct dahdi_vmwi_info {
 
 #define DAHDI_ECHOCANCEL_FAX_MODE	_IOW(DAHDI_CODE, 102, int)
 
-struct torisa_debug {
-	unsigned int txerrors;
-	unsigned int irqcount;
-	unsigned int taskletsched;
-	unsigned int taskletrun;
-	unsigned int taskletexec;
-	int span1flags;
-	int span2flags;
-};
-
-/* Special torisa ioctl */
-#define TORISA_GETDEBUG			_IOW(DAHDI_CODE, 60, struct torisa_debug)
-
 /* Get current status IOCTL */
 /* Defines for Radio Status (dahdi_radio_stat.radstat) bits */