Currently, to select the "comedi_bond" driver, the "kcomedilib" support
has to be selected first.  It seems more natural to allow the
"comedi_bond" driver to be selected on its own and to automatically
select the "kcomedilib" module as a result of that.  Change the
dependency/select relationship between the two configuration options to
do that.

Also, make "kcomedilib" a module that can be selected independently,
perhaps for use by out-of-tree modules, although it is not terribly
useful for anything other than "comedi_bond" currently.

Also, improve the help text for the config options.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

The addition operation has higher precedence than ?: so this would write
to PCI230_DACOUT1 instead of devpriv->daqio + PCI230_DACOUT1.

Fixes: 4f9c63fe ('staging: comedi: amplc_pci230: refactor iobase addresses')
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use comedi_pci_detach() to handle the boilerplate part of the (*detach)
for this PCI driver.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Currently the dma buffers are freed during the detach of the driver using
the function s626_close_dma_b(). The buffers are also freed while the
interrupt handler is still attached. This could result in a race condition.

Refactor the code so that the buffers are freed after the interrupt handler
is released and the PCI device is disabled.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Factor the freeing of the dma buffers out of the (*detach).

Move the freeing of the buffers so that it occurs after the PCI device
has been disabled to avoid any race condition.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Factor the freeing of the dma buffers out of the (*detach).

Move the freeing of the buffers so that it occurs after the PCI device
has been disabled to avoid any race condition.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

This member is set at the end of the driver attach and is only used to
verify that icp_multi_reset() can be called in the (*detach). The only
requirement for icp_multi_reset() to work is that the dev->mmio is valid.

Remove the 'valid' member from the private data and use 'dev->mmio' to
see if the reset can be done.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

This member is set at the end of the driver attach and is only used to
verify that pci_dio_reset() can be called in the (*detach). The only
requirement for pci_dio_reset() to work is that the dev->iobase is valid.

Remove the 'valid' member from the private data and use 'dev->iobase' to
see if the reset can be done.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Introduce a generic (*detach) function for comedi PCI drivers to handle
the boilerplate code needed to detach a PCI driver.

This function works similar to comedi_legacy_detach() where it will:

  * free the dev->irq if it has been requested
  * iounmap the dev->mmio addres if it has been ioremap'ed

The helper then calls comedi_pci_disable() to release the regions and
disable the PCI device.

Use the new helper directly for the (*detach) in the following cases:

  * where comedi_pci_disable() is used directly for the (*detach)
  * where the detach function is just boilerplate

Use the new helper in the (*detach) of the simpler PCI drivers. Call
the helper after disabling interrupts (reset) and before any additional
cleanup (kfree) to avoid any race conditions with the interrupt handler.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

The comedi core does the sanity checking of the comedi_insn before
calling the (*insn_write). Remove the unnecessary checks.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

The comedi core validates that all the data values are <= s->maxdata
before calling the (*insn_write) function. The extra masking is not
needed. Remove it.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

The comedi core validates the insn->chanspec and data values before
calling the (*insn_write) functions. The 'chan' will always be valid
and the data values do not need to be masked.

Tidy up this function and remove the unnecessary code.

For aesthetics, rename the function.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

For aesthetics, rename this function and tidy it up a bit.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

For aesthetics, rename the function and tidy it up a bit. Also rename the
registers used by this function and add a macro to set the 'chan' bits in
the MSB data register.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.

For aesthetics, tidy up the (*insn_write) a bit.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

This code has been disabled since it's initial commit. It must not be
needed so just remove it.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

The unmunged data should be saved for readback not the munged data.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.

The das08_ao_initialize() function is only called when the subdevice
is first setup during the attach. It is used to initialize the channels
to a known state. Absorb that function to simplify the code a bit.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

The comedi core expects (*insn_write) functions to write insn->n values
to the hardware and return the number of values written.

Currently, this function only writes the first value. Fix it to work
like the core expects.

For aesthetics, rename the function and tidy it up a bit.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

The comedi core expects (*insn_write) functions to write insn->n values
to the hardware and return the number of values written.

Currently, this function only writes the first value. Fix it to work
like the core expects.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.

For aesthetics, rename the (*insn_write) function and tidy it up a bit.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.

Remove the unused private data and its allocation.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

The unmunged data should be saved for readback not the munged data.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

The comedi core expects (*insn_write) functions to write insn->n values
to the hardware and return the number of values written.

Currently, this function only writes the first value. The core also
validates all the data values so the extra masking with s->maxdata is
not needed. Fix this function to work like the core expects.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

The comedi core validates the insn->chanspec and data values before
calling the (*insn_write) functions. The 'chan' will always be valid.

Tidy up this function up a bit and remove the unnecessary code.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new comedi_subdevice 'readback' member and the core provided
(*insn_read) for the readback of the analog output subdevice channels.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Reviewed-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>