- 07 Aug, 2015 40 commits
-
-
Sudip Mukherjee authored
Now since all cleanups are done and the code is ready to be merged lets move it out of staging into fbdev location. Signed-off-by: Sudip Mukherjee <sudip@vectorindia.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
H Hartley Sweeten authored
Change the MODULE_DESCRIPTION to something more useful than "Comedi low- level 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
Firmware loading was fixed by: Commit: ac584af5 "staging: comedi: me4000: fix firmware downloading" Change the driver status to "untested" and remove the comments about the driver being broken, 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
Format the multi-line comments in the kernel CodingStyle. 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
Fix checkpatch issue: "CHECK: usleep_range is preferred over udelay; see Documentation/timers/timers-howto.txt". `udelay()` is only used in the firmware upload process. Replace them with `usleep_range()` with a reasonable upper limit. 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 returns two's complement values for the analog input samples. These need to be converted to the unsigned binary format that the comedi core expects. Introduce a helper function to handle this. 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 irq handler does not need to manually stop conversions and disable interrupts when "end-of-acquisition", "error", or "overflow" events are detected. The comedi_handle_events() will call the subdevice (*cancel) when these are detected and stop the acquisition. 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
For aesthetics, add some white space to the analog output subdevice initialization. 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 is no real reason to reset the board when detaching. The comedi core will ensure that any commands are canceled before the detach. But the PLX interrupts should be 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>
-
H Hartley Sweeten authored
Currently me4000_reset() always enables the PLX interrupt. Move the enable of the interrupt into me4000_auto_attach() and only do the enable if we actually have and irq. Make sure the PLX interrupt is disabled in me4000_reset() before we try to request the irq. 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 a helper function to stop any ai conversions and reset the ai control register. This consolidates the common code in me4000_reset() and me4000_ai_cancel(). Use the new helper in the ai (*insn_read) to ensure that the ai control register is set to a known state after reading the samples. The ai control register will now always be '0' after the (*cancel) of a command or doing an (*insn_read). Knowing this the programming of the register for single acquisition mode in the (*insn_read) can be simplified. 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
Reset the analog input control register after ensuring that any active conversions have been stopped. This mimics what the ai subdevice (*cancel) does. 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 core ensures that the subdevice is not busy before it allows starting a new command. The subdevice (*cancel) is called when the subdevice is set to not busy. In this driver the me4000_ai_cancel() is the ai (*cancel) so the extra call in the ai (*do_cmd) is not necessary. 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 me4000_ai_cancel() already reset this 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>
-
H Hartley Sweeten authored
The STOP and IMMEDIATE_STOP bits in the ME4000_AI_CTRL_REG should be set, not cleared, to stop any running conversions. 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 function always returns 0 and the return value is never checked. Just return void. 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 function is only called by me4000_ai_do_cmd(). Absorb 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 function never fails and it's only called by me4000_ai_do_cmd(). Absorb it and remove the unnecessary failure check. 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 ai (*do_cmd_test) validates the trigger sources in Step 2b to ensure that they are compatible. Save the 'ai_ctrl_mode' that will be used in the private data so that ai_prepare(), which is called by the ai (*do_cmd), does not have to recheck the sources in order to figure it out. Also, tidy up the stop trigger checks so that the ME4000_AI_CTRL_HF_IRQ bit is set in the common code path. 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 ai (*do_cmd_test) calls me4000_ai_round_cmd_args() to calculate the timing values needed to the command. The the command test passes, the core will then call the ai (*do_cmd) which then has to call me4000_ai_round_cmd_args() again in order to get the same values to pass to ai_prepare() in order to program the timing. Add members to the private data to allow the (*do_cmd_test) to calculate and save to values needed by ai_prepare(). 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 function always succeeds. Change the return type to void and remove the unnecessary error check in me4000_ai_do_cmd(). Move the function call in me4000_ai_do_cmd_test() from before Step 1 to Step 3 where the arguments are validated. There is no reason to get the values if the previous steps fail. Rename the function so it has namespace associated with 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
Use the BIT() marco to define the bits of this register. For aesthetics, rename all the defines to remove the '_BIT' from the name. 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
Use the BIT() marco to define the bits of this register. For aesthetics, rename all the defines to remove the '_BIT' from the name. 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
Use the BIT() marco to define the bits of this register. For aesthetics, rename all the defines to remove the '_BIT' from the name. 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
Use the BIT() marco to define the bits of this register. For aesthetics, rename all the defines to remove the '_BIT' from the name. 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
Use the BIT() marco to define the bits of this register. For aesthetics, rename all the defines to remove the '_BIT' from the name. 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
Use the BIT() marco to define the bits of this register. For aesthetics, rename all the defines to remove the '_BIT' from the name. Also, use ME4000_AI_STATUS_REG instead of ME4000_AI_CTRL_REG when reading the register (they happen to be the same). 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
Use the helper function to check the range type instead of relying on the value. For aesthetics, rename the local variable used for the range. 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 chanspec channel, range, and aref are unsigned int values. Use the correct types when extracting them. 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 coemdi (*insn_read) functions are supposed to read insn->n values from the hardware. Make this function work like the core expects. Use the comedi_offset_munge() helper to munge the two's complement values to offset binary. 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
Use the comedi_timeout() helper to busy-wait for the analog input end-of- conversion instead of the udelay(). 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
Use the BIT() macro to define the bits of this 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>
-
H Hartley Sweeten authored
Rename this function so it has namespace associated with the driver. The last entry of the chanlist needs the ME4000_AI_LIST_LAST_ENTRY bit set to end the list. Fix the function and tidy if 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>
-
H Hartley Sweeten authored
The comedi_lrange table for the analog inputs is inverted compared to the values that need to be written to the ME4000_AI_CHANNEL_LIST_REG to select the range. Create a macro, ME4000_AI_LIST_RANGE(), to handle the inversion. Remove the old defines and 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>
-
H Hartley Sweeten authored
This define evaluates to 0 and is OR'ed with the 'entry' value that is written to the ME4000_AI_CHANNEL_LIST_REG when the channel aref is a single-ended type (AREF_GROUND or AREF_COMMON). OR'ing a zero value is pretty silly, just remove it. Remove the switch() in me4000_ai_insn_read() to simplify 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
In comedi drivers the local variable used for the dev->private pointer is normally named 'devpriv'. For aesthetics, rename the variables in this driver. Also, rename the struct to follow the norm. 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
Some of the boards supported by this driver can do analog input sample & hold on 8 of the channels. The 'ai_sh_nchan' member of the boardinfo is used to indicate which boards support this feature. To save a bit of space, convert this member to a bit-field, 'can_do_sh_ai'. Note, this feature is not currently supported. 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 always '4' for the boards that have an analog output FIFO. Covert it to a bit-field, 'has_ao_fifo', to save a bit of space. 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
For the boards that have analog output capability, there are always 4 analog output channels. Convert the 'ao_nchan' member of the boardinfo into a bit-field, 'has_ao', to save a bit of space and set the analog output subdevice 'n_chan' to 4 when supported. 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 only used as a flag indicating that the board supports differential analog inputs. Convert the member to a bit- field to save a bit of space. For aesthetics, rename the member to 'can_do_diff_ai'. 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>
-