GitLab

This subdevice function is not implemented in the driver. 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>

For aesthetic reasons, add some whitespace to the subdevice init.
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>

Convert the comedi_subdevice access from pointer match to array
access.
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>

Add some local variables to make this function a bit clearer and
remove the unnecessary comments.
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>

Add some local variables to make this function a bit cleaner.

The comedi core expects this function to return the number of
data parameters used. Change the return from '1' to insn->n to
make this more aparent.
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>

There are a number of uint16_t casts used in the #define's of the
constant bit field values as well as the calls to DEBIreplace().
These cause a number of sparse warnings of the type:

warning: cast truncates bits from constant value (ffff1cff becomes 1cff)

Remove all of the casts and change the types of the parameters to
DEBIreplace from uin16_t to unsigned int. This fixes all the warnings.

Mask the addr that is or'ed with DEBI_CMD_RDWORD then written to the
P_DEBICMD register as well as the val written to the P_DEBIAD register
with 0xffff. The addr and val are only 16-bits but the registers are
32-bits.
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 base_address variable in the private data is the ioremap'ed
PCI bar 0 resource.

For aesthetic reasons, and to shorten some of the lines, rename
this variable to 'mmio'.
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 macro relies on a local variable having a specific name its
also just a wrapper around a readl() call. Remove the macro and
just call readl() directly.
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 WR7146 macro relies on a local variable having a specific name.
This macro is a wrapper around a writel() call. Remove the macro and
just call writel() directly.

The SETVEC macro uses the WR7146 macro so remove it as well and just
do the writel() directly.
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 macro relies on a local variable having a specific name. Replace
it with a new helper function, s626_mc_test().
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 macro relies on a local variable having a specific name. Replace
it with a new helper function, s626_mc_disable().
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

This macro relies on a local variable having a specific name. Replace
it with a new helper function, s626_mc_enable().
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Currently the unused s626_ai_rinsn() function is commented out with
/* ... */ on every line of the function. Remove those by using an

This block will be removed once it is determined that it really
is not needed.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Reformat the whitespace in the comedi_lrange table and use the
BIP_RANGE() macro instead of generic RANGE() macro.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Factor the code that handles the counter interrupt out of s626_irq_handler().
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Factor the code that handles the end of scan interrupt out of
s626_irq_handler().

Remove the printk noise when cfc_write_to_buffer() fails. The user
can't do anything about it anyway.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

The comedi_async pointer is used a number of times in this function.
For aesthetic reasons, use a local variable for the pointer.

Also, since was already have a pointer to the "cmd" use that in the
for loop instead of s->async->cmd (or even async->cmd).
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

The comedi_subdevice in this function is actually the dev->read_subdev
that was initialized during the attach of the board. Use that instead
of accessing the dev->subdevices array directly.

Also, get the comedi_cmd used in s626_irq_handler() when the function
is first entered.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Factor the code that checks and handles the dio interrupts out of
s626_irq_handler().

This allows reducing the number of indents in the code and makes it
easier to follow.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

This driver has three dio subdevices, each with 16 channels. These
subdevices use hardware registers that are spaced 0x10 bytes apart
for each "group" of 16 channels.

Create new macros to get the correct addresses for each register
based on the dio "group". Use the s->private (void *) to hold the
"group" number to use in the subdevice functions.

This allows removing struct dio_private and all its uses. It also
removes the ugly casts used to get the register address.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Some PCI drivers use the "spare" `iobase` member of `struct
comedi_device` as a flag to indicate that the call to
`comedi_pci_enable()` was successful.  This is no longer necessary now
that `comedi_pci_enable()` and `comedi_pci_disable()` use the
`ioenabled` member of `struct comedi_device` themselves to keep track of
what needs to be done.

Remove the unnecessary assignments to the `iobase` member in the
relevant drivers.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Change the `attached` member of `struct comedi_device` to a 1-bit
bit-field of type `bool`.  Change assigned values to `true` and `false`
and replace or remove comparison operations with simple boolean tests.

We'll put some extra bit-fields in the gap later to save space.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Make comedi_pci_enable() use the same parameter type as
comedi_pci_disable(). This also allows comedi_pci_enable
to automatically determine the resource name passed to
pci_request_regions().

Make sure the errno value returned is passed on instead of
assuming an errno. Also, remove any kernel noise that is
generated when the call fails.

The National Instruments drivers that use the mite module
currently enable the PCI device in the mite module. For
those drivers move the call to comedi_pci_enable into the
driver.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Currently all the comedi PCI drivers need to do some checking in
their (*detach) before calling comedi_pci_disable() in order to
make sure the PCI device has actually be enabled.

Change the parameter passed to comedi_pci_disable() from a struct
pci_dev pointer to a comedi_device pointer and have comedi_pci_disable()
handle all the checking.

For most comedi PCI drivers this also allows removing the local
variable holding the pointer to the pci_dev. For some of the drivers
comedi_pci_disable can now be used directly as the (*detach) function.

The National Instruments drivers that use the mite module currently
enable/disable the PCI device in the mite module. For those drivers
move the call to comedi_pci_disable into the driver and make sure
dev->iobase is set to a non-zero value.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

The comedi_pci_auto_config() function is used to allow the PCI driver
(*probe) function to automatically call the comedi driver (*auto_attach).
This allows the comedi driver to be part of the PnP process when the
PCI device is detected.

Currently the comedi_pci_auto_config() always passes a 'context' of '0'
to comedi_auto_config(). This makes the 'context' a bit useless.

Modify comedi_pci_auto_config() to allow the comedi pci drivers to pass
a 'context' from the PCI driver.

Make all the comedi pci drivers pass the pci_device_id 'driver_data' as
the 'context'. Since none of the comedi pci drivers currently set the
'driver_data' the 'context' will still be '0'.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Separate the comedi_pci_* functions out of drivers.c into a new
source file, comedi_pci.c. This allows conditionally building
support for comedi PCI drivers into the comedi core. Fix the
Kconfig and Makefile appropriately.

Group all the comedi_pci_* prototypes and related defines into one
place in comedidev.h. Protect these prototypes with an #ifdef and
provide some dummy functions so that the mixed ISA/PCI comedi
drivers will still build correctly.

Remove the #include <linux/pci.h> from comedidev.h and drivers.c. This
include is only needed by the comedi PCI driver support code and the
PCI drivers. The include should occur in those files.

Also, remove the #include <linux/pci.h> from a couple non-PCI drivers
since it's not needed.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

(Almost) all comedi pci drivers have some wrapper for their
pci_driver.remove function which simply calls comedi_pci_auto_unconfig
which has the same function prototype as the wrapper.

-> we can remove these wrappers and call comedi_pci_auto_unconfig
directly. This removes a lot some boilerplate code and saves some bytes.
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

CONFIG_HOTPLUG is going away as an option so __devexit is no
longer needed.
Signed-off-by: Bill Pemberton <wfp5p@virginia.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

CONFIG_HOTPLUG is going away as an option so __devinit is no longer
needed.
Signed-off-by: Bill Pemberton <wfp5p@virginia.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

CONFIG_HOTPLUG is going away as an option so __devexit_p is no longer
needed.
Signed-off-by: Bill Pemberton <wfp5p@virginia.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the new helpers in the step 3 tests of s626_ai_cmdtest().
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Change comedi drivers for PCI boards to use the new `auto_attach()`
method instead of the `attach_pci()` method.  I plan to remove the
`attach_pci()` and `attach_usb()` methods from `struct comedi_driver`
once nothing is using them.

Tag the functions with `__devinit` where they are not already so tagged,
as they are only called during PCI probe.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

The comedi core now sets the 'hw_dev' pointer in the function
comedi_auto_config_helper() before calling the auto attach
function in the driver.

Remove the now unnecessary call to comedi_set_hw_dev() in the
drivers that use the auto-config attach mechanism.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

This inline function has a very generic name and it's only a
wrapper around a simple kzalloc(). Since the inline function
does not save any lines-of-code, instead of renaming it just
remove it and do the kzalloc() directly.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Missed one... This macro relies on a local variable having
a specific name. Remove its use by replacing it with a local
variable where used.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Use the cfc_check_trigger_src() helper for Step 1 in all the
driver cmdtest functions.

Use the cfc_check_trigger_is_unique() helper for Step 2 in all
the driver cmdtest functions. Note that single source triggers
do not need to be checked, they are already unique if they pass
Step 1.

For aesthetic reasons, change the comments in the cmdtest
functions for steps 1 and 2 so that they are all the same.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

`s626_enc_insn_config()` is the `insn_config()` handler for a counter
subdevice.  The `data[0]` value is supposed to be one of the
`INSN_CONFIG_...` constants defined in "comedi.h" indicating the type of
configuration instruction, but this function seems to be using it as a
variable value to preload the counter with.  Various values of `data[0]`
are going to cause `check_insn_config_length()` in the comedi core
("comedi_fops.c") to return an error, and this function won't be called
in those cases.  Most other values will log a warning to the kernel log.

It's not entirely clear what constant should be checked for in
`data[0]`, so add a "FIXME" comment for now.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

`s626_enc_insn_config()` is incorrectly dereferencing `insn->data` which
is a pointer to user memory.  It should be dereferencing the separate
`data` parameter that points to a copy of the data in kernel memory.
Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Reviewed-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Cleanup the comments to follow the coding style of the kernel.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

The private data is kzalloc'ed. There is no need to set any
of the initial data to '0'.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>