[PATCH 00/15] staging: comedi: 8255: tidy up the (*io) operations
Greg KH
gregkh at linuxfoundation.org
Sun Aug 10 04:15:51 UTC 2014
On Mon, Aug 04, 2014 at 02:18:43PM +0100, Ian Abbott wrote:
> On 2014-08-01 22:18, H Hartley Sweeten wrote:
> >Now that the comedi_device has a 'mmio' and a 'iobase' member, this module
> >can detect if the (*io) operations need to use readb/writeb for memory mapped
> >I/O or inb/outb for port I/O. This can be use to cleanup a lot of the code
> >duplication in the drivers that use the 8255 module.
> >
> >1) Modify a couple comedi drivers that use port I/O for the 8255 device so
> > that the 'iobase' is used for the base address of the 8255 device.
> >
> >2) Add a comedi_device parameter to the (*io) callback. Tidy up the private
> > (*io) callbacks to use this parameter.
> >
> >3) Modify the default (*io) callback to use the comedi_device 'iobase'. Remove
> > the, now redundant, dev->iobase from the 'iobase' that is passed to
> > subdev_8255_init() for all callers that use the default (*io) callback.
> >
> >4) Modify the default (*io) callback to use the comedi_device 'mmio'. Remove
> > the private (*io) callbacks that use readb/writeb to access the hardware.
> >
> >5) Modify the amplc_dio200_common module to use the 8255 module and remove the
> > code duplication.
> >
> >H Hartley Sweeten (15):
> > staging: comedi: cb_pcidas: use dev->iobase for PCI bar 3
> > staging: comedi: cb_pcimdas: refactor iobase addresses
> > staging: comedi: cb_pcidda: refactor iobase addresses
> > staging: comedi: amplc_pci230: refactor iobase addresses
> > staging: comedi: 8255: add a comedi_device param to the (*io) callback
> > staging: comedi: ni_mio_common: tidy up ni_8255_callback()
> > staging: comedi: ni_labpc: tidy up labpc_8255_mmio()
> > staging: comedi: ni_labpc: tidy up daqboard2000_8255_cb()
> > staging: comedi: cb_pcidas64: tidy up dio_callback()
> > staging: comedi: cb_pcidas64: tidy up dio_callback_4020()
> > staging: comedi: 8255_pci: tidy up pci_8255_mmio()
> > staging: comedi: 8255: refactor how the (*io) function works
> > staging: comedi: 8255: make default (*io) function handle memory mapped io
> > staging: comedi: 8255: provide common defines for registers
> > staging: comedi: amplc_dio200_common: use 8255 module
> >
> > drivers/staging/comedi/Kconfig | 1 +
> > drivers/staging/comedi/drivers/8255.c | 100 +++++++------
> > drivers/staging/comedi/drivers/8255.h | 24 ++-
> > drivers/staging/comedi/drivers/8255_pci.c | 26 +---
> > drivers/staging/comedi/drivers/adv_pci_dio.c | 4 +-
> > drivers/staging/comedi/drivers/aio_aio12_8.c | 3 +-
> > .../staging/comedi/drivers/amplc_dio200_common.c | 127 ++--------------
> > .../staging/comedi/drivers/amplc_pc236_common.c | 2 +-
> > drivers/staging/comedi/drivers/amplc_pci230.c | 162 ++++++++++-----------
> > drivers/staging/comedi/drivers/cb_pcidas.c | 10 +-
> > drivers/staging/comedi/drivers/cb_pcidas64.c | 28 +---
> > drivers/staging/comedi/drivers/cb_pcidda.c | 27 ++--
> > drivers/staging/comedi/drivers/cb_pcimdas.c | 16 +-
> > drivers/staging/comedi/drivers/cb_pcimdda.c | 3 +-
> > drivers/staging/comedi/drivers/daqboard2000.c | 13 +-
> > drivers/staging/comedi/drivers/das08.c | 3 +-
> > drivers/staging/comedi/drivers/das16.c | 3 +-
> > drivers/staging/comedi/drivers/das16m1.c | 2 +-
> > drivers/staging/comedi/drivers/ni_atmio16d.c | 2 +-
> > drivers/staging/comedi/drivers/ni_daq_dio24.c | 2 +-
> > drivers/staging/comedi/drivers/ni_labpc.c | 20 +--
> > drivers/staging/comedi/drivers/ni_mio_common.c | 12 +-
> > drivers/staging/comedi/drivers/pcl724.c | 10 +-
> > drivers/staging/comedi/drivers/pcm3724.c | 27 +---
> > 24 files changed, 233 insertions(+), 394 deletions(-)
> >
>
> Patch 08 has an incorrect title. Patch 13 has an implicit assumption about
> the type of I/O to use that I don't like. I'd rather drop patch 15 as it
> stops me supporting additional features of the Amplicon DIO PCIe boards.
>
> Reviewed-by: Ian Abbott <abbotti at mev.co.uk>
> for patches 01 to 07, 09 to 12, and 14, and for patch 08 once the patch
> title is corrected.
Ok, that's a mess, I'll just wait for a resend of this series with
things fixed up so I don't have to pick-and-choose here.
thanks,
greg k-h
More information about the devel
mailing list