Ian Abbott [Mon, 5 Sep 2011 12:29:20 +0000 (13:29 +0100)]
Fix warnings about including deprecated <linux/config.h>
Change Comedi's <linux/config.h> compatibility header so it only
includes the real <linux/config.h> for kernel versions below 2.6.15.
Previously it was included for kernel versions below 2.6.19 (when it was
removed from the kernel sources) but wasn't needed for 2.6.15 onwards.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Mon, 5 Sep 2011 12:17:20 +0000 (13:17 +0100)]
Deal with backported bool type for RHEL5's 2.6.18 kernel
RHEL5's 2.6.18 kernel includes a change backported from 2.6.19 that
declares a generic 'bool' type in <linux/types.h> and declares the
'false' and 'true' enum values in <linux/stddef.h>. Comedi's
compatibility header for <linux/stddef.h> declared the 'false' and
'true' enum values for any kernel version below 2.6.19. This resulted
in comedi failing to build due to redeclaration errors.
Add an autoconf test to see if the generic 'bool' type is defined in
<linux/types.h> and if so define the macro
CONFIG_COMEDI_HAVE_GENERIC_BOOL_TYPE in <config.h>. Conditionally
declare the 'bool' type in the <linux/types.h> compatibility header and
change the condition for declaring the 'false' and 'true' enum values in
the <linux/stddef.h> compatibility header so they are only declared if
the CONFIG_COMEDI_HAVE_GENERIC_BOOL_TYPE macro is undefined.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Mon, 18 Jul 2011 11:16:18 +0000 (12:16 +0100)]
ni_pcidio: Forgot to include irq flags variable for spinlock.
I messed up the previous commit. Need to include a variable in
setup_mite_dma() to hold the IRQ flags when grabbing the spinlock.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Mon, 18 Jul 2011 11:08:38 +0000 (12:08 +0100)]
ni_pcidio: Grab MITE channel spinlock while preparing and arming DMA.
When setting up the DMA for 'read' streaming acquisition command, grab
the MITE channel spinlock before preparing and arming the DMA.
Change inspired by ni_ai_setup_MITE_dma() in ni_mio_common.c.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Mon, 18 Jul 2011 10:30:08 +0000 (11:30 +0100)]
ni_pcidio: Mark buffer for writing when setting up DMA.
When setting up the DMA for 'read' streaming acquisition on the DIO
subdevice, mark the whole buffer as writable before starting the DMA.
This prevents a spurious detection of a DMA overwrite of good data
during the first interrupt.
Problem reported by Nico Nell. Fix suggested by Frank Mori Hess.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Mon, 27 Jun 2011 10:05:39 +0000 (11:05 +0100)]
Fix infoleak to userspace.
Following Vasiliy Kulikov <segoon _at_ openwall _dot_ org>'s patch for
the staging sources:
driver_name and board_name are pointers to strings, not buffers of size
COMEDI_NAMELEN. Copying COMEDI_NAMELEN bytes of a string containing
less than COMEDI_NAMELEN-1 bytes would leak some unrelated bytes.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Nico Nell [Wed, 15 Jun 2011 19:46:19 +0000 (20:46 +0100)]
Advantech PCI-1739U support
Hello,
I've added a bit of simple configuration to adv_pci_dio.c in order to
make the PCI-1739U work with the adv_pci_dio driver. I have tested
inputs only so far but they seem to work. A git style patch is
attached.
Thanks,
--
Nicholas Nell
Professional Research Assistant
University of Colorado
nicholas.nell@colorado.edu
303-492-5661
From
c2a97fef832aa70b7fb2eff60de6412829b314ba Mon Sep 17 00:00:00 2001
From: Gene Ray <user@SLICE-testing.Colorado.EDU>
Date: Thu, 2 Jun 2011 12:58:37 -0600
Subject: [PATCH] Added functionality for the Advantech PCI-1739U DIO board
Ian Abbott [Thu, 2 Jun 2011 13:47:01 +0000 (14:47 +0100)]
configure: Recognize kernel 3.x
Support Linus's alpha-maleness kernel version numbering change.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Mon, 16 May 2011 15:06:32 +0000 (16:06 +0100)]
addi-data: Get rid of redundant ps_BoardInfo
The ps_BoardInfo pointer in the device private data is redundant as we
can just use the this_board macro to access the same data, as is done
elsewhere in the code. Get rid of the pointer and change the code to
use the this_board macro instead.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Mon, 16 May 2011 14:57:24 +0000 (15:57 +0100)]
addi_data: Don't overwrite read-only data
The drivers for ADDI-DATA cards can override some static parameters for
the board type using information read from EEPROM. Unfortunately, they
currently write the parameters from the EEPROM back to the shared,
read-only board data! The problem has been masked during compilation by
type-casting away the const-ness of the data.
This patch changes the code to use an area in the private data for the
board instance to hold the parameters read from EEPROM (after
initializing the parameters from the static board data). It also
changes the type-casts to the read-only data to preserve the const
qualifier.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Mon, 16 May 2011 09:53:21 +0000 (10:53 +0100)]
AUTHORS: Added Ivan Russkih (Иван Русских).
Added for patch to adv_pci_dio to support counter subdevice on PCI-1751.
Ian Abbott [Mon, 16 May 2011 09:31:42 +0000 (10:31 +0100)]
adv_pci_dio: Add 8254 counter subdevice to PCI-1751
The Advantech PCI-1751 has a 8254 counter chip providing 3 counter
channels. Add a counter subdevice for this card.
Patch from Иван Русских (vanekrus at gmail dot com).
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Fri, 13 May 2011 13:55:19 +0000 (14:55 +0100)]
comedi: Removal of <linux/smp_lock.h>
The <linux/smp_lock.h> header (used for the Big Kernel Lock) is removed
in kernel 2.6.39. It was included by:
comedi/comedi_compat32.c
comedi/drivers/usbdux.c
comedi/drivers/usbduxfast.c
The usbdux and usbduxfast drivers shouldn't need to include the header
as they don't grab the BKL, so remove the #include line for these
drivers.
The comedi_compat32 code only uses it when calling the ioctl file
operation, and that can only happen for kernels up to 2.6.35 (and even
then it wouldn't be called if the unlocked_ioctl file operation is
available!). Ifdef out the #include line. Also include
<linux/version.h> explicitly to avoid compiler errors for 2.6.39 kernel.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Tue, 10 May 2011 13:26:31 +0000 (14:26 +0100)]
configure.ac: Change reference to CVS to Git.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Tue, 10 May 2011 13:18:16 +0000 (14:18 +0100)]
INSTALL: Changed reference to cvs to git.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Tue, 10 May 2011 13:14:14 +0000 (14:14 +0100)]
.gitignore: Ignore the file 'missing'.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Tue, 10 May 2011 13:13:05 +0000 (14:13 +0100)]
Remove references to .cvsignore from .gitignore files.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Tue, 10 May 2011 13:08:53 +0000 (14:08 +0100)]
Rename the .cvsignore files to .gitignore
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Tue, 10 May 2011 12:51:35 +0000 (13:51 +0100)]
README: Updated mailing list details.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Tue, 10 May 2011 11:47:01 +0000 (12:47 +0100)]
README: Removed reference to non-existant comedi FTP site.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Tue, 10 May 2011 11:43:50 +0000 (12:43 +0100)]
README: Changed CVS references to Git.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Tue, 10 May 2011 11:33:37 +0000 (12:33 +0100)]
README.CVS: Renamed to README.Git and edited slightly
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Thu, 3 Mar 2011 15:47:44 +0000 (15:47 +0000)]
Add #include <linux/version.h> to include/asm/uaccess.h
The include/asm/uaccess.h compatibility header uses LINUX_VERSION_CODE
and KERNEL_VERSION, but doesn't #include <linux/version.h> to define
those macros. This caused a build failure for at least one person.
Perhaps we should just remove the file though, as it's only needed for
really ancient kernels (before 2.2.0) that we no longer support.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Ian Abbott [Wed, 19 Jan 2011 10:54:39 +0000 (10:54 +0000)]
ni_labpc: Use shared IRQ for PCMCIA card.
Ian Abbott [Tue, 18 Jan 2011 16:26:27 +0000 (16:26 +0000)]
Most drivers ignore the base channel in insn->chanspec for INSN_BITS
instructions. If the subdevice has no more than 32 channels, and the
base channel is non-zero, modify the instruction chanspec and data before
passing it to the driver with base channel set to 0, and modify the
returned data.
Ian Abbott [Thu, 13 Jan 2011 12:07:33 +0000 (12:07 +0000)]
pcmmio: Avoid compiler warning about uninitialized irq[1] which is out of
bounds of irq[] anyway (as MAX_ASICS = 1).
Ian Abbott [Mon, 22 Nov 2010 10:28:01 +0000 (10:28 +0000)]
Added Brice DUBOST for cb_pcidas contributions (edge and inverted level
external triggering for start argument of AI command on PCI-DAS1602 series).
Ian Abbott [Mon, 22 Nov 2010 10:23:56 +0000 (10:23 +0000)]
cb_pcidas: Applied patch from Brice DUBOST to support CR_EDGE and CR_INVERT
start_arg flags for AI commands on PCI_DAS1602 series.
Ian Abbott [Mon, 15 Nov 2010 10:25:43 +0000 (10:25 +0000)]
Fix memory leak spotted by Nicolas Kaiser in the staging sources.
Ian Abbott [Tue, 9 Nov 2010 12:28:29 +0000 (12:28 +0000)]
ni_mio_cs: I broke support for 2.6.35 when adding support for 2.6.36 due
to using a wrong #ifdef in mio_cs_config(). Sorry! This fixes it.
Ian Abbott [Mon, 8 Nov 2010 15:15:05 +0000 (15:15 +0000)]
More ugly PCMCIA changes for 2.6.37 kernel.
Ian Abbott [Fri, 8 Oct 2010 14:32:24 +0000 (14:32 +0000)]
PCMCIA support broke again with 2.6.36(-rc) kernel, so add more uqly hacks.
Ian Abbott [Fri, 8 Oct 2010 14:30:02 +0000 (14:30 +0000)]
serial2002: old 'ioctl' file operation removed in 2.6.36.
Ian Abbott [Fri, 8 Oct 2010 11:10:02 +0000 (11:10 +0000)]
comedi_compat32: old 'ioctl' file operation removed in 2.6.36.
Ian Abbott [Fri, 8 Oct 2010 09:59:20 +0000 (09:59 +0000)]
I messed up the #ifdefs on the previous commit. Sorry!
Ian Abbott [Thu, 30 Sep 2010 16:31:27 +0000 (16:31 +0000)]
quatech_daqp_cs: Replace eos semaphore with a completion. Build tested only.
Note: the wait_for_completion_interruptible() function needs 2.6.11 kernel
or above, but I'm pretty sure this driver (and other PCMCIA drivers) is
currently broken for those kernels anyway, so I'm not bothering to implement
the function in compatibility headers just yet.
Ian Abbott [Thu, 30 Sep 2010 15:59:41 +0000 (15:59 +0000)]
PCMCIA support broke again with 2.6.35 kernel, so add yet more ugly hacks.
Ian Abbott [Mon, 26 Jul 2010 11:47:08 +0000 (11:47 +0000)]
dt9812, usbdux, usbduxfast: Use mutex instead of semaphore.
Ian Abbott [Mon, 26 Jul 2010 11:16:30 +0000 (11:16 +0000)]
Add #include <linux/semaphore.h>.
TODO: Migrate to use mutex instead, but note that it currently initializes
the semaphore in the 'locked' state.
Ian Abbott [Mon, 26 Jul 2010 11:14:07 +0000 (11:14 +0000)]
Compatibility header for <linux/semaphore.h>. Just includes the proper
<linux/semaphore.h> if CONFIG_COMEDI_HAVE_LINUX_SEMAPHORE_H is defined,
otherwise it includes <asm/semaphore.h> instead.
Ian Abbott [Mon, 26 Jul 2010 11:12:19 +0000 (11:12 +0000)]
Add check for existance of <linux/semaphore.h> header and define the
CONFIG_COMEDI_HAVE_LINUX_SEMAPHORE_H macro in config.h if present. Note
<linux/semaphore.h> was added in kernel version 2.6.26, replacing
<asm/semaphore.h>. We need a compatibility header for it.
Ian Abbott [Wed, 23 Jun 2010 15:12:40 +0000 (15:12 +0000)]
Add a dkms.conf file for building comedi with DKMS
(Dynamic Kernel Module Support).
Ian Abbott [Tue, 22 Jun 2010 16:12:24 +0000 (16:12 +0000)]
PCMCIA support broke with the 2.6.33 kernel, so add support back in.
I don't know if it works, but at least it should compile!
The driver source code changes are very inelegant.
Note that the IRQ attributes have been changed from IRQ_TYPE_EXCLUSIVE to
IRQ_TYPE_DYNAMIC_SHARING.
Ian Abbott [Fri, 18 Jun 2010 10:00:26 +0000 (10:00 +0000)]
adl_pci9111: bug fix for AI commands when scan_begin_src is TRIG_FOLLOW.
The dev_private->scan_delay should have been left set to 0 in this case,
but it ended up being set to (unsigned int)-1.
Ian Abbott [Mon, 14 Jun 2010 10:42:02 +0000 (10:42 +0000)]
Change example Driver: comment block slightly to avoid it being matched by
updated 'dump_doc' script.
Ian Abbott [Mon, 14 Jun 2010 10:34:38 +0000 (10:34 +0000)]
Allow "*" (optionally surrounded by a space on each side) to precede
"Driver:" and strip this from all the lines in the comment block.
Ian Abbott [Wed, 26 May 2010 11:17:33 +0000 (11:17 +0000)]
Check for RTAI include directory "${RTAI_DIR}/include/rtai" before checking
for "${RTAI_DIR}/include". This is because Debian uses --with-rtaidir=/usr
but puts the RTAI include files in /usr/include/rtai.
Ian Abbott [Wed, 19 May 2010 16:49:49 +0000 (16:49 +0000)]
Fix previous commit. For COMEDI_BUFINFO, if command not active, report back
the last buffer position instead of random junk.
Ian Abbott [Wed, 19 May 2010 14:40:46 +0000 (14:40 +0000)]
adv_pci_dio: Forgot to update the 'Updated:' line when the driver
description was changed in Revision 1.26.
Ian Abbott [Wed, 19 May 2010 14:31:19 +0000 (14:31 +0000)]
adv_pci_dio: Protect access to counter subdevice with a spin lock.
Ian Abbott [Tue, 18 May 2010 16:28:07 +0000 (16:28 +0000)]
serial2002: Fix memory leak on 'detach' (off-by-one error in subdevice loop).
Ian Abbott [Mon, 17 May 2010 12:18:35 +0000 (12:18 +0000)]
For COMEDI_BUFINFO ioctl, check for locked or busy subdevice.
Ian Abbott [Mon, 17 May 2010 11:26:07 +0000 (11:26 +0000)]
For COMEDI_BUFINFO ioctl, if subdevice has no async capability, set
bytes_read and bytes_written to 0 in returned structure.
Ian Abbott [Fri, 14 May 2010 15:29:11 +0000 (15:29 +0000)]
For write(), don't bother writing data to buffer if command already
finished, as the data won't be used anyway. This allows user code to
see the command has finished before it has refilled the buffer completely.
Ian Abbott [Tue, 4 May 2010 13:58:49 +0000 (13:58 +0000)]
Support Advantech PCI-1735, including 8254-based counter subdevice (the
counter subdevice might not be working though).
Frank Mori Hess [Mon, 12 Apr 2010 18:07:09 +0000 (18:07 +0000)]
Some changes to make new driver pass kernel's anal formatting checks
in checkpatch.pl.
Frank Mori Hess [Mon, 12 Apr 2010 17:37:51 +0000 (17:37 +0000)]
Added Adlink PCI-7230 driver from David Fernandez.
Ian Abbott [Wed, 24 Mar 2010 17:04:59 +0000 (17:04 +0000)]
addi_common: Just use the driver name in the call to comedi_request_irq()
instead of a static char buffer that gets overwritten.
Ian Abbott [Wed, 24 Mar 2010 16:34:46 +0000 (16:34 +0000)]
amplc_dio200: Protect access to counter subdevices with a spin lock.
Ian Abbott [Wed, 24 Mar 2010 12:54:35 +0000 (12:54 +0000)]
serial2002: Slight change to previous commit. If failed to allocate
maxdata_list or range_table_list for a subdevice during the 'open' method,
free these for all subdevices as they may be left over from a previous
open.
Ian Abbott [Wed, 24 Mar 2010 12:45:23 +0000 (12:45 +0000)]
serial2002: In serial_2002_open(), handle failure to allocate maxdata_list
or range_table_list for a subdevice.
Ian Abbott [Wed, 24 Mar 2010 12:19:30 +0000 (12:19 +0000)]
serial2002: Reduce stack usage in serial_2002_open() by allocating
dig_in_config, dig_out_config, chan_in_config and chan_out_config arrays
using kcalloc().
Ian Abbott [Wed, 24 Mar 2010 11:53:09 +0000 (11:53 +0000)]
The 'open' method in struct comedi_device_struct now returns 0 on success
and a negative errno on failure. The 'open' method must clean up after
itself on failure as the 'close' method won't be called. A failure causes
the 'open' file operation for the comedi device to fail.
The dt9812 and serial2002 drivers now make use of the ability to return
an error.
Ian Abbott [Thu, 18 Mar 2010 15:39:58 +0000 (15:39 +0000)]
Give the different addi_apci_* drivers different comedi driver names.
Ian Abbott [Mon, 15 Mar 2010 17:02:46 +0000 (17:02 +0000)]
Prototype of request_firmware_nowait() changed in 2.6.33.
Use new prototype and add compatibility code for earlier kernels.
Frank Mori Hess [Mon, 22 Feb 2010 14:23:25 +0000 (14:23 +0000)]
Increased timeout for digital output dma loading, due to report of it
being too short for some hardware.
Ian Abbott [Wed, 20 Jan 2010 12:33:02 +0000 (12:33 +0000)]
pcl818: No need to check chanlist_len for AI command test.
Ian Abbott [Wed, 20 Jan 2010 12:17:37 +0000 (12:17 +0000)]
pcl816: No need to check chanlist_len for AI command test.
Ian Abbott [Wed, 20 Jan 2010 12:04:26 +0000 (12:04 +0000)]
pcl812: Don't need to set cur_chan twice at end of scan.
Ian Abbott [Wed, 20 Jan 2010 11:57:34 +0000 (11:57 +0000)]
pcl812: No need to check chanlist_len for AI command test.
Ian Abbott [Wed, 20 Jan 2010 11:52:49 +0000 (11:52 +0000)]
pcl812: Untested bug fix for multi-channel AI commands. Update MUX on
AI interrupt.
Ian Abbott [Mon, 18 Jan 2010 19:38:39 +0000 (19:38 +0000)]
adv_pci1710: Don't need to check chanlist_len in pci171x_ai_cmdtest().
Ian Abbott [Mon, 18 Jan 2010 19:23:46 +0000 (19:23 +0000)]
adv_pci1710: Bug fixes: For AI commands, keep track of current channel when
moving block of data from FIFO. Setting up of internal channel list should
fill in the whole list, not just the first repeated segment.
Ian Abbott [Mon, 18 Jan 2010 17:14:07 +0000 (17:14 +0000)]
pcl816: Check the channel list in step 5 of AI command test.
Ian Abbott [Mon, 18 Jan 2010 16:46:11 +0000 (16:46 +0000)]
pcl818: bug fixes: For AI commands, the scan counter was being updated
after every sample, except for DMA mode where is was being updated after
each repeated segment of the channel list. Keep track of current sample
and update the scan counter at the end of each scan. Also, AI commands
with multiple channels were being terminated with an error incorrectly in
non-DMA mode. This was because the current position within a repeated
segment of the channel list was not being updated, causing it to think
there was a "channel dropout" when comparing the assumed current channel
with the channel number read from the FIFO.
Ian Abbott [Mon, 18 Jan 2010 16:25:33 +0000 (16:25 +0000)]
pcl816: bug fix: AI scan counter was being updated after every channel in
non-DMA mode, or after every repeated segment of the channel list in DMA
mode. Keep track of the current channel for each sample and update the
scan counter at the end of the scan.
Ian Abbott [Mon, 18 Jan 2010 16:20:08 +0000 (16:20 +0000)]
pcl812: bug fix: scan counter was being incremented after every sample.
Keep track of current channel and update scan counter at end of scan.
Ian Abbott [Mon, 4 Jan 2010 11:39:28 +0000 (11:39 +0000)]
pcmmio: reorder check for second irq in pcmmio_attach() to avoid reading
past end of array.
(Staging patch from Dan Carpenter.)
Ian Abbott [Mon, 4 Jan 2010 11:22:14 +0000 (11:22 +0000)]
comedi_bond: fix off by one in doDevConfig().
(Staging patch from Dan Carpenter.)
Bernd Porr [Thu, 26 Nov 2009 20:18:59 +0000 (20:18 +0000)]
removed "deprecated" from the define COMEDI_CB_BLOCK and added
a comment so that other know that this flag has to be set whenever
fresh data has arrived from an A/D or has been written to a D/A.
Bernd Porr [Thu, 26 Nov 2009 20:10:14 +0000 (20:10 +0000)]
Jan-Matthias Braun spotted the problem that once the ring buffer
of comedi has run empty the next read command would block forever
even after new data has arrived. Thanks for Jan-Matthias and Ian
who fixed the bug.
Ian Abbott [Thu, 26 Nov 2009 14:59:46 +0000 (14:59 +0000)]
jr3-pci: Only ioremap register space for actual number of channels on the
card. Also, check result of ioremap() for failure.
Thanks to Anders Blomdell for input and Sami Hussein for testing.
Ian Abbott [Thu, 19 Nov 2009 18:37:30 +0000 (18:37 +0000)]
adv_pci1710: Also initialize divisor1 and divisor2 in pci171x_ai_cmdtest().
Ian Abbott [Thu, 19 Nov 2009 11:43:52 +0000 (11:43 +0000)]
pcl818: In ai_cmd_test(), pcl818_ai_cmd_mode(), and pcl818_ao_mode13(),
initialize divisor1 and divisor2 values passed by reference to
i8153_cascade_ns_to_timer().
Ian Abbott [Thu, 19 Nov 2009 11:39:10 +0000 (11:39 +0000)]
pcl816: In pcl816_ai_cmdtest(), initialize divisor1 and divisor2 values
passed by reference to i8253_cascase_ns_to_timer().
Ian Abbott [Thu, 19 Nov 2009 11:35:13 +0000 (11:35 +0000)]
cb_das16_cs: In das16cs_ai_cmdtest(), initialize div1 and div2 values
passed by reference to i8253_cascade_ns_to_timer() (two instances).
Ian Abbott [Thu, 19 Nov 2009 11:32:37 +0000 (11:32 +0000)]
adv_pci1710: In pci171x_ai_docmd_and_mode(), initialize divisor1 and
divisor2 values passed by reference to i8253_cascase_ns_to_timer().
Ian Abbott [Thu, 19 Nov 2009 11:30:05 +0000 (11:30 +0000)]
adl_pci9118: In pci9118_ai_cmdtest(), initialize divisor1 and divisor2
values, passed by reference to i8253_cascade_ns_to_timer().
Ian Abbott [Thu, 19 Nov 2009 11:21:51 +0000 (11:21 +0000)]
pcl816: In pcl816_ai_cmdtest(), correct step 1 test for valid convert_src
(operator precedence error).
Ian Abbott [Thu, 19 Nov 2009 11:10:08 +0000 (11:10 +0000)]
pcl711: In pcl711_ai_cmd(), fix uninitialized variables timer1 and timer2.
Values are passed by reference to i8253_cascade_ns_to_timer().
Ian Abbott [Thu, 19 Nov 2009 11:05:09 +0000 (11:05 +0000)]
serial2002: In serial_2002_open(), reduce size of config_t by 4 bytes.
This reduces stack usage by 512 bytes.
Ian Abbott [Thu, 19 Nov 2009 10:59:13 +0000 (10:59 +0000)]
me4000: In xilinx_download(), correct test for /INIT pin being set during
initialization. Parentheses were missing.
Bernd Porr [Tue, 17 Nov 2009 23:18:49 +0000 (23:18 +0000)]
I've set all timeout values to 1000ms. Some bulk transfers used 30ms
as a timeout which should be OK but it's just a bit short. Nothing
detrimental.
Bernd Porr [Tue, 17 Nov 2009 23:09:47 +0000 (23:09 +0000)]
Fixed an issue with jiffies vs miliseconds for the timeout value
of bulk transfers. From 2.6.12 the timing information is miliseconds
and not jiffies. However in the driver there was a mix of constants
referring to jiffies and miliseconds. I've fixed it and introduced
the constant BULK_TIMEOUT in miliseconds for all transfers.
Ian Abbott [Mon, 19 Oct 2009 09:34:54 +0000 (09:34 +0000)]
[Following Linus's linux-2.6 tree...]
commit
d43c36dc removed sched.h from interrupt.h. This broke some of
the comedi drivers. This patch fixes this.
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Ian Abbott [Mon, 19 Oct 2009 09:33:01 +0000 (09:33 +0000)]
[Following Linus's linux-2.6 tree...]
Now that sched.h was removed from poll.h, serial2002.c needs
to include it otherwise it does not build properly.
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Frank Mori Hess [Mon, 21 Sep 2009 20:09:52 +0000 (20:09 +0000)]
Added device id for pxi-6225 to ni_pci_table.
Frank Mori Hess [Mon, 17 Aug 2009 16:07:53 +0000 (16:07 +0000)]
Added device id for pxi-6225.
Ian Abbott [Wed, 12 Aug 2009 16:04:43 +0000 (16:04 +0000)]
When reading DO subdevice with inverted outputs invert the values read back
from the ports to match the inversion of values written.
Ian Abbott [Wed, 12 Aug 2009 15:56:14 +0000 (15:56 +0000)]
Fix insn_bits bitshift calculation for subdevice with non-zero base_port.
Thanks to Jérôme Carretero (cJ-comedi at zougloub dot eu) for spotting the
bug.