Re: [PATCH 0/3] ARM: OMAP2+: omap_device: add dynamic pinctrl handling
On 07/17/2013 03:30 PM, Grygorii Strashko wrote: On 07/17/2013 02:57 PM, Roger Quadros wrote: Hi Grygorii, On 07/17/2013 02:41 PM, Grygorii Strashko wrote: Hi Tony, Kevin This patch series introduces dynamic pinctrl handling in OMAP device framework in the same way as it was before switching to DT. This allow OMAP devices driver's developers to simply add dynamic pinctrl handling for default, active, idle, sleep PIN states in their drivers by modifying DT definitions only - no modifications in drivers code are not needed. Overall I like the idea but can we make a provision for device drivers to override this default pin state handling? The OMAP EHCI driver is one such special case where the wakeup mechanism is tied to pinctrl states as it uses IO daisy chaining to implement wakeup. So depending on whether wakeup needs to be enabled or not I must be able to chose whether I put the pin in just sleep state or sleep with wakeup state. I think, in this case you can't use default behavior and need to define custom pins states like sleep_wakeup/sleep_no_wakeup and do not define pins state with name sleep', so Device core and OMAP device framework will not touch your pins. Yes, I think this should be fine. Thanks. cheers, -roger -- To unsubscribe from this list: send the line unsubscribe linux-omap in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[PATCH 0/3] ARM: OMAP2+: omap_device: add dynamic pinctrl handling
Hi Tony, Kevin This patch series introduces dynamic pinctrl handling in OMAP device framework in the same way as it was before switching to DT. This allow OMAP devices driver's developers to simply add dynamic pinctrl handling for default, active, idle, sleep PIN states in their drivers by modifying DT definitions only - no modifications in drivers code are not needed. Tested on OMAP4 SDP(Tablet2) board using omap4-sdp.dts - UART3/4 autoidle - suspend, by adding some fake sleep pin state - wake up (PRCM) IRQ generation by ading call to omap44xx_prm_reconfigure_io_chain() right after each call to pinctrl_pm_select_xx() API. UART3(console) reg dumps: - active 0x4A100140(RTS/CTS) - 0x0118 0x4A100140(TX/RX) - 0x0100 - idle 0x4A100140(RTS/CTS) - 0x0118 0x4A100144(TX/RX) - 0x4100 UART4 reg dumps: - active 0x4A10015C(TX/RX) - 0x0100 - idle 0x4A10015C(TX/RX) - 0x4100 Debugfs: # cat sys/kernel/debug/pinctrl/4a100040.pinmux/pinmux-pins [snip] pin 128 (4a100140.0): 4802.serial (GPIO UNCLAIMED) function pinmux_uart3_pins group pinmux_uart3_pins pin 129 (4a100142.0): 4802.serial (GPIO UNCLAIMED) function pinmux_uart3_pins group pinmux_uart3_pins pin 130 (4a100144.0): 4802.serial (GPIO UNCLAIMED) function pinmux_uart3_pins_active group pinmux_uart3_pins_active pin 131 (4a100146.0): 4802.serial (GPIO UNCLAIMED) function pinmux_uart3_pins group pinmux_uart3_pins [snip] pin 142 (4a10015c.0): 4806e000.serial (GPIO UNCLAIMED) function pinmux_uart4_pins_idle group pinmux_uart4_pins_idle pin 143 (4a10015e.0): 4806e000.serial (GPIO UNCLAIMED) function pinmux_uart4_pins group pinmux_uart4_pins Changes since RFC: - _od_resume_noirq() fixed - idle state was not selected on resume - description updated RFC: https://lkml.org/lkml/2013/6/21/309 Based on v3.11-rc1 plus patch series from Tony Lindgren [PATCH 0/4] improved support for runtime muxing for pinctrl http://www.spinics.net/lists/arm-kernel/msg258827.html: fd7937b drivers: Add pinctrl handling for dynamic pin states 479246b pinctrl: Add support for additional dynamic states c9f37e8 pinctrl: Allow pinctrl to have multiple active states 756f10b pinctrl: Remove duplicate code in pinctrl_pm_select_state functions ad81f05 Linux 3.11-rc1 Related discussions: - drivers: pinctrl sleep and idle states in the core http://lists.infradead.org/pipermail/linux-arm-kernel/2013-June/173514.html - [PATCH] drivers: pinctrl: add active state to core http://marc.info/?l=linux-kernelm=137094012703340w=2 - [PATCH] pinctrl: document the pinctrl PM states https://lkml.org/lkml/2013/6/11/501 - [3/3] i2c: nomadik: use pinctrl PM helpers https://patchwork.kernel.org/patch/2670291/ - mmc: omap_hsmmc: Remux pins to support SDIO interrupt and PM runtime https://patchwork.kernel.org/patch/2690191/ - [PATCH 00/11] drivers: Add Pinctrl PM support https://lkml.org/lkml/2013/5/31/210 Grygorii Strashko (3): pinctrl: rollback check for !dev-pins in pinctrl_pm_select*() APIs ARM: OMAP2+: omap_device: add pinctrl handling ARM: dts: omap4-sdp: add dynamic pin states for uart3/4 arch/arm/boot/dts/omap4-sdp.dts | 34 +++ arch/arm/mach-omap2/omap_device.c | 40 +++-- drivers/pinctrl/core.c| 12 +++ 3 files changed, 76 insertions(+), 10 deletions(-) Regards, -grygorii CC: Linus Walleij linus.wall...@linaro.org CC: Stephen Warren swar...@wwwdotorg.org CC: Roger Quadros rog...@ti.com -- 1.7.9.5 -- To unsubscribe from this list: send the line unsubscribe linux-omap in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 0/3] ARM: OMAP2+: omap_device: add dynamic pinctrl handling
Hi Grygorii, On 07/17/2013 02:41 PM, Grygorii Strashko wrote: Hi Tony, Kevin This patch series introduces dynamic pinctrl handling in OMAP device framework in the same way as it was before switching to DT. This allow OMAP devices driver's developers to simply add dynamic pinctrl handling for default, active, idle, sleep PIN states in their drivers by modifying DT definitions only - no modifications in drivers code are not needed. Overall I like the idea but can we make a provision for device drivers to override this default pin state handling? The OMAP EHCI driver is one such special case where the wakeup mechanism is tied to pinctrl states as it uses IO daisy chaining to implement wakeup. So depending on whether wakeup needs to be enabled or not I must be able to chose whether I put the pin in just sleep state or sleep with wakeup state. cheers, -roger -- To unsubscribe from this list: send the line unsubscribe linux-omap in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 0/3] ARM: OMAP2+: omap_device: add dynamic pinctrl handling
On 07/17/2013 02:57 PM, Roger Quadros wrote: Hi Grygorii, On 07/17/2013 02:41 PM, Grygorii Strashko wrote: Hi Tony, Kevin This patch series introduces dynamic pinctrl handling in OMAP device framework in the same way as it was before switching to DT. This allow OMAP devices driver's developers to simply add dynamic pinctrl handling for default, active, idle, sleep PIN states in their drivers by modifying DT definitions only - no modifications in drivers code are not needed. Overall I like the idea but can we make a provision for device drivers to override this default pin state handling? The OMAP EHCI driver is one such special case where the wakeup mechanism is tied to pinctrl states as it uses IO daisy chaining to implement wakeup. So depending on whether wakeup needs to be enabled or not I must be able to chose whether I put the pin in just sleep state or sleep with wakeup state. I think, in this case you can't use default behavior and need to define custom pins states like sleep_wakeup/sleep_no_wakeup and do not define pins state with name sleep', so Device core and OMAP device framework will not touch your pins. cheers, -roger Regards, -grygorii -- To unsubscribe from this list: send the line unsubscribe linux-omap in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
