- 29 Oct, 2014 40 commits
-
-
H Hartley Sweeten authored
Factor the urb handling and resubmit out of the completion handler and tidy it up. This allows a common exit path to be used in the completion handler to stop the async command and handle the events. Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com> Reviewed-by: Ian Abbott <abbotti@mev.co.uk> Reviewed-by: Bernd Porr <mail@berndporr.me.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
H Hartley Sweeten authored
Factor the urb handling and resubmit out of the analog input urb completion handler and tidy it up. This allows a common exit path to be used in the completion handler to stop the async command and handle the events. Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com> Reviewed-by: Ian Abbott <abbotti@mev.co.uk> Reviewed-by: Bernd Porr <mail@berndporr.me.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
H Hartley Sweeten authored
Factor the urb handling and resubmit out of the analog output urb completion handler and tidy it up. This allows a common exit path to be used in the completion handler to stop the async command and handle the events. Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com> Reviewed-by: Ian Abbott <abbotti@mev.co.uk> Reviewed-by: Bernd Porr <mail@berndporr.me.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
H Hartley Sweeten authored
This header is no longer included by any of the addi-data drivers. 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>
-
H Hartley Sweeten authored
Remove the need to include addi_common.h by introducing a new private data definition. Only include the members that are actually used by the 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>
-
H Hartley Sweeten authored
Remove the need to include addi_common.h by introducing a new private data definition. Only include the members that are actually used by the 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>
-
H Hartley Sweeten authored
Remove the need to include addi_common.h by introducing a new private data definition. Only include the members that are actually used by the 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>
-
H Hartley Sweeten authored
There are two boards supported by this driver. One with 16-bit analog inputs and the other 12-bit analog inputs. Add a bit field flag, 'ai_is_16bit', to the boardinfo and remove the 'i_AiMaxdata' member. Use the new flag to initialize tne analog input subdevice maxdata correctly. For aesthetics, adjust the whitespace in the boardinfo declaration. 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>
-
H Hartley Sweeten authored
There are two boards supported by this driver. One with eight 14-bit analog outputs and the other without analog outputs. Add a bit field flag, 'has_ao', to the boardinfo and remove the 'i_NbrAoChannel' and 'i_AoMaxdata' members. Use the new flag to determine if the analog output subdevice needs to be initialized and open code the n_chan and maxdata. 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>
-
H Hartley Sweeten authored
All boardinfo entries have this member set as '16'. This is also the Analog Input 'n_chan'. For aesthetics, initialize the Analog Input 'len_chanlist' with s->n_chan and remove the boardinfo member. 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>
-
H Hartley Sweeten authored
All boardinfo entries have this member set as '16'. This will initialize the Analog Input 'n_chan' and cause the 'i_NbrAiChannelDiff' member to never get used. Remove both members and initialize the Analog Input 'n_chan' with 16 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>
-
H Hartley Sweeten authored
All boardinfo entries have this member set as '4'. Remove the member and initialize the Digial Input 'n_chan' with 4 directly. Also, remove the initialization of 'len_chanlist'. This member is only used with subdevices that support async commands. 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>
-
H Hartley Sweeten authored
All boardinfo entries have this member set as '4'. Remove the member and initialize the Digial Output 'n_chan' with 4 directly. Also, remove the initialization of 'len_chanlist'. This member is only used with subdevices that support async commands. 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>
-
H Hartley Sweeten authored
The maxdata for a Digital Output subdevice is always '1'. Remove this incorrect boardinfo. 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>
-
H Hartley Sweeten authored
Remove all the unused members from the addi_private definition. 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>
-
H Hartley Sweeten authored
The addi_board definition is not only used by the addi_apci_3120 driver. Introduce a private definition in that driver and remove the global definition from addi_common.h. 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>
-
H Hartley Sweeten authored
The apci3120_ai_insn_config() function is broken in this driver. It does not follow the comedi API and will fail to work correctly. For now just remove the need for the boardinfo access to allow additional cleanup. The 'i_NbrAiChannel' is actually the subdevice 'n_chan' use that instead and remove the boardinfo 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>
-
H Hartley Sweeten authored
This source file is no longer included by any of the addi-data drivers. 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>
-
H Hartley Sweeten authored
The hardware has an eeprom but the boardinfo properly defines the subdevice support so reading the eeprom and setting the 's_EeParameters' use is not necessary. Since the boardinfo only has one entry its not really necessary. Also, providing a subdevice the read the eeprom is not really interesting. Remove the eeprom support code along with the boardinfo and initialize the subdevices 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>
-
H Hartley Sweeten authored
The hardware does not have an eeprom so the 's_EeParameters' use is not necessary. Remove it. Since the boardinfo only has one entry its not really necessary. Remove it and initialize the subdevices 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>
-
H Hartley Sweeten authored
This member of the boardinfo is not used by any of the drivers. 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>
-
H Hartley Sweeten authored
As indicated in the boardinfo, the hardware has an eeprom. Simplify the board attach code and remove the boardinfo. 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>
-
H Hartley Sweeten authored
As indicated in the boardinfo, the hardware does not have an eeprom. Remove the subdevice and support code for 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>
-
H Hartley Sweeten authored
This member of the boardinfo is used as a flag to determine what PCI BAR to use for the main 'iobase' used by the driver. Remove the boardinfo member and refactor the (*auto_attach) of the drivers to use the correct PCI BAR. 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>
-
H Hartley Sweeten authored
Rename this CamelCase member of the boardinfo. The comedi_device 'board_name' is set to this member during the (*auto_attach) of the addi-data that still use this boardinfo drivers. For aesthetics, use the dev->board_name instead of the boardinfo in the drivers. 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>
-
H Hartley Sweeten authored
Only allocate space for the subdevicess used by the driver. Remove all the COMEDI_SUBD_UNUSED subdevices. 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>
-
H Hartley Sweeten authored
The subdevices that don't initialize any of the callbacks don't really exist. Remove the unnecessary initialzation and just set the type to COMEDI_SUBD_UNUSED. 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>
-
H Hartley Sweeten authored
Only allocate space for the subdevicess used by the driver. Remove all the COMEDI_SUBD_UNUSED subdevices. 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>
-
H Hartley Sweeten authored
The subdevices that don't initialize any of the callbacks don't really exist. Remove the unnecessary initialzation and just set the type to COMEDI_SUBD_UNUSED. 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>
-
H Hartley Sweeten authored
None of the drivers that still include this header use the subdevice callbacks. Remove all of them to reduce the bloat. 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>
-
H Hartley Sweeten authored
This driver only has one boardinfo entry so the subdevice callbacks are always the same. Remvoe the callbacks from the boardinfo and use them directly when initializing the subdevices. Remove all the subdevice callback initialization that would be set to NULL. 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>
-
H Hartley Sweeten authored
This driver only has one boardinfo entry so the subdevice callbacks are always the same. Remvoe the callbacks from the boardinfo and use them directly when initializing the subdevices. Remove all the subdevice callback initialization that would be set to NULL. 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>
-
H Hartley Sweeten authored
The addi_apci_035 and addi_apci_1500 are the only drivers left that include this source file. Copy the i_ADDIDATA_InsnReadEeprom() function from that file to those drivers and remove the addi_common.c file. 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>
-
H Hartley Sweeten authored
The addi_apci_035 and addi_apci_1500 are the only drivers left that use this function in addi_common.c. The function simply calls the 'interrupt' function that is in the boardinfo of the driver. Both drivers use the same 'interrupt' function for all boardnfo entries. Remove the i_ADDI_Reset() function as well as the 'interrupt' boardinfo and use the 'interrupt' function directly when doing the request_irq(). In addition, the addi_apci_3120 driver has a private v_ADDI_Interrupt() function that is doing the same thing. Remove that one as well. Fix the return type of 'interrupt' functions in the drivers and add the return vaules. 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>
-
H Hartley Sweeten authored
The addi_apci_035 and addi_apci_1500 are the only drivers left that use this function in addi_common.c. The addi_common code adds a lot of bloat to the drivers. Copy the addi_auto_attach() code to the (*auto_attach) functions of the drivers and remove the function. This will help with removing the bloat. 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>
-
H Hartley Sweeten authored
The addi_apci_035 and addi_apci_1500 are the only drivers left that use this function in addi_common.c. The function simply calls the 'reset' function that is in the boardinfo of the driver. Both drivers use the same 'reset' function for all boardnfo entries. Remove the i_ADDI_Reset() function as well as the 'reset' boardinfo and just call the 'reset' function directly. The i_ADDI_Reset() is called by addi_auto_attach() in addi_common.c after a sucessful attach. Modify the (*auto_attach) in the drivers to call the 'reset' function directly and remove it from addi_auto_attach(). 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>
-
H Hartley Sweeten authored
The addi_apci_035 and addi_apci_1500 are the only drivers left that use this function in addi_common.c. Remove the function an add a private comedi_driver (*detach) function to those drivers. 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>
-
H Hartley Sweeten authored
This driver has some serious bitrot. In addition, it's not in the Kconfig or Makefile so it can't even be built. Just remove it. It would be easier to rewrite the driver than fix 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>
-
H Hartley Sweeten authored
Introduce COMEDI_CB_ERROR_MASK and COMEDI_CB_CANCEL_MASK to clarify the async->events that indicate errors and cancel an async command. Use the new defines to tidy up and clarify the 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>
-
H Hartley Sweeten authored
The COMEDI_CB_* defines are the comedi_async "events" that the drivers set to let the core detect the state of running async commands. These "events" are only relevant to the kernel modules and should not be exposed to userspace in the comedi.h user API header. Move the defines to comedidev.h to avoid exposing them. For aesthetics, convert the defines to bit shifts to better indicate that they are bitmask values. Cleanup the documentation. 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>
-