| Message ID | cover.1514475654.git.vilhelm.gray@gmail.com |
|---|---|
| Headers | show |
| Series | Change ISA_BUS_API dependency to selection | expand |
On 12/28/2017 08:01 AM, William Breathitt Gray wrote: > The ISA_BUS_API Kconfig option enables the compilation of the ISA bus > driver. The ISA bus driver does not perform any hardware interaction, > and is instead just a thin layer of software abstraction to eliminate > boilerplate code common to ISA-style device drivers. Since ISA_BUS_API > has no dependencies and does not jeopardize the integrity of the system > when enabled, drivers should select it when the ISA bus driver > functionality is needed. > > Originally, when the ISA_BUS_API Kconfig option was introduced, it > served a dual-purpose of masking drivers ISA-style devices not commonly > found for desktop systems, such as the PC/104 device drivers. This > secondary semantic was inappropriate for the ISA_BUS_API option, and > proper masking of these device drivers is now accomplished via dedicated > Kconfig options such as CONFIG_PC104. > > Linus, please pickup this entire patchset through your GPIO subsystem > tree; a recursive dependency error is present if these patches are > cherry-picked (see https://lkml.org/lkml/2017/12/26/235), so they should > be merged together in the same tree. > > Maciej, this patchset resolves the recursive dependency issue you > encountered, so now you should be able to submit your Winbond GPIO > driver with the ISA_BUS_API selection as desired. > > William Breathitt Gray (3): > iio: Change ISA_BUS_API dependency to selection > watchdog: Change ISA_BUS_API dependency to selection > gpio: Change ISA_BUS_API dependency to selection > > drivers/gpio/Kconfig | 14 +++++++++----- > drivers/iio/adc/Kconfig | 3 ++- > drivers/iio/counter/Kconfig | 3 ++- > drivers/iio/dac/Kconfig | 3 ++- > drivers/watchdog/Kconfig | 3 ++- > 5 files changed, 17 insertions(+), 9 deletions(-) > But why keep "config ISA_BUS" ? Its only purpose is to select ISA_BUS_API. Guenter -- To unsubscribe from this list: send the line "unsubscribe linux-gpio" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 28.12.2017 17:01, William Breathitt Gray wrote: (..) > > Linus, please pickup this entire patchset through your GPIO subsystem > tree; a recursive dependency error is present if these patches are > cherry-picked (see https://lkml.org/lkml/2017/12/26/235), so they should > be merged together in the same tree. > > Maciej, this patchset resolves the recursive dependency issue you > encountered, so now you should be able to submit your Winbond GPIO > driver with the ISA_BUS_API selection as desired. Thanks for these patches, will submit the driver as soon as I integrate Andy's comments. Maciej -- To unsubscribe from this list: send the line "unsubscribe linux-gpio" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Thu, Dec 28, 2017 at 09:03:22AM -0800, Guenter Roeck wrote: >On 12/28/2017 08:01 AM, William Breathitt Gray wrote: >> The ISA_BUS_API Kconfig option enables the compilation of the ISA bus >> driver. The ISA bus driver does not perform any hardware interaction, >> and is instead just a thin layer of software abstraction to eliminate >> boilerplate code common to ISA-style device drivers. Since ISA_BUS_API >> has no dependencies and does not jeopardize the integrity of the system >> when enabled, drivers should select it when the ISA bus driver >> functionality is needed. >> >> Originally, when the ISA_BUS_API Kconfig option was introduced, it >> served a dual-purpose of masking drivers ISA-style devices not commonly >> found for desktop systems, such as the PC/104 device drivers. This >> secondary semantic was inappropriate for the ISA_BUS_API option, and >> proper masking of these device drivers is now accomplished via dedicated >> Kconfig options such as CONFIG_PC104. >> >> Linus, please pickup this entire patchset through your GPIO subsystem >> tree; a recursive dependency error is present if these patches are >> cherry-picked (see https://lkml.org/lkml/2017/12/26/235), so they should >> be merged together in the same tree. >> >> Maciej, this patchset resolves the recursive dependency issue you >> encountered, so now you should be able to submit your Winbond GPIO >> driver with the ISA_BUS_API selection as desired. >> >> William Breathitt Gray (3): >> iio: Change ISA_BUS_API dependency to selection >> watchdog: Change ISA_BUS_API dependency to selection >> gpio: Change ISA_BUS_API dependency to selection >> >> drivers/gpio/Kconfig | 14 +++++++++----- >> drivers/iio/adc/Kconfig | 3 ++- >> drivers/iio/counter/Kconfig | 3 ++- >> drivers/iio/dac/Kconfig | 3 ++- >> drivers/watchdog/Kconfig | 3 ++- >> 5 files changed, 17 insertions(+), 9 deletions(-) >> >But why keep "config ISA_BUS" ? Its only purpose is to select ISA_BUS_API. I intend to have CONFIG_ISA_BUS now serve purely as a mask for true ISA device drivers (e.g. CIO_DAC within this patchset). The legacy CONFIG_ISA has a X86_32 dependency which prevents its use for X86_64 builds -- and unfortunately existing drivers depending on CONFIG_ISA expect the X86_32 arch, thus making the task of decoupling that arch dependency difficult. I'm hoping CONFIG_ISA_BUS can therefore supersede the legacy CONFIG_ISA because an ISA bus can physically exist on a non-X86_32 system. However, there is no longer a need to select ISA_BUS_API, so that should be removed, as well as the help text updated to match the changes. I'll make these updates and submit a version 2 of this patchset some time tomorrow then. William Breathitt Gray > >Guenter -- To unsubscribe from this list: send the line "unsubscribe linux-gpio" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html