On Fri, Oct 18, 2013 at 8:20 AM, NeilBrown <neilb@xxxxxxx> wrote: > On Sat, 5 Oct 2013 13:17:09 +0200 Andreas Fenkart <afenkart@xxxxxxxxx> wrote: > >> The am335x can't detect pending cirq in PM runtime suspend. >> This patch reconfigures dat1 as a GPIO before going to suspend. >> SDIO interrupts are detected with the GPIO, while in runtime >> suspend, standard detection of the module block otherwise. >> >> Signed-off-by: Andreas Fenkart <afenkart@xxxxxxxxx> >> >> diff --git a/Documentation/devicetree/bindings/mmc/ti-omap-hsmmc.txt b/Documentation/devicetree/bindings/mmc/ti-omap-hsmmc.txt >> index 1136e6b..146f3ad 100644 >> --- a/Documentation/devicetree/bindings/mmc/ti-omap-hsmmc.txt >> +++ b/Documentation/devicetree/bindings/mmc/ti-omap-hsmmc.txt >> @@ -21,8 +21,11 @@ ti,non-removable: non-removable slot (like eMMC) >> ti,needs-special-reset: Requires a special softreset sequence >> ti,needs-special-hs-handling: HSMMC IP needs special setting for handling High Speed >> ti,quirk-swakup-missing: SOC missing the swakeup line, will not detect >> -SDIO irq while in suspend. Fallback to polling. Affected chips are >> -am335x, >> +SDIO irq while in suspend. The workaround is to reconfigure the dat1 line as a >> +GPIO upon suspend. Beyond this option and the GPIO config, you also need to set >> +named pinctrl states "default", "active" and "idle ", see example below. The >> +MMC driver will then then toggle between default and idle during the runtime >> +Affected chips are am335x, >> >> ------ >> | PRCM | >> @@ -49,3 +52,24 @@ Example: >> vmmc-supply = <&vmmc>; /* phandle to regulator node */ >> ti,non-removable; >> }; >> + >> +[am335x with with gpio for sdio irq] >> + >> + mmc1_cirq_pin: pinmux_cirq_pin { >> + pinctrl-single,pins = < >> + 0x0f8 0x3f /* MMC0_DAT1 as GPIO2_28 */ >> + >; >> + }; >> + >> + mmc1: mmc@48060000 { >> + ti,non-removable; >> + bus-width = <4>; >> + vmmc-supply = <&ldo2_reg>; >> + vmmc_aux-supply = <&vmmc>; >> + ti,quirk-swakeup-missing; >> + pinctrl-names = "default", "active", "idle"; >> + pinctrl-0 = <&mmc1_pins>; >> + pinctrl-1 = <&mmc1_pins>; >> + pinctrl-2 = <&mmc1_cirq_pin>; >> + ti,cirq-gpio = <&gpio3 28 0>; >> + }; > > > hi, > I've been trying to get SD irq to work on my OMAP3 DM3730. > I seems to need the magic to catch interrupts while FCLK is off, as > the only way I can get it to work at the moment is to keep FCLK on. > > I discovered your patch and tried it out, but it doesn't seem to work for me. > > I have a Libertas WIFI chip attached to the second mmc (which is sometimes > called mmc1, and sometimes mmc2 - very confusing!). Hi Neil, I have a DM3730 board with the same setup, Libertas (Marvell SD8686) wifi + bt chip attached to mmc2. I was going to try to add support for this to the DTS but it would be great if I can use your as a reference. Would you be so kind to share your DTS? Thanks a lot and best regards, Javier > I added: > mmc2_cirq_pin: pinmux_cirq_pin { > pinctrl-single,pins = < > 0x012e (PIN_INPUT_PULLUP | MUX_MODE4) /* MMC2_DAT1 as GPIO5_5 */ > >; > }; > > and > > &mmc2 { > ti,quirk-swakeup-missing; > pinctrl-names = "default", "active", "idle"; > pinctrl-0 = <&mmc2_pins>; > pinctrl-1 = <&mmc2_pins>; > pinctrl-2 = <&mmc2_cirq_pin>; > ti,cirq-gpio = <&gpio5 5 0>; /* GPIO133 = 128+5 */ > }; > > to my dts file but it doesn't make any apparent difference. > > Any idea what I might be doing wrong? > (the base kernel I am applying patches to is from 8th Oct, > commit 0e7a3ed04f0cd4311096d691888f88569310ee6c) > > BTW, > - with the default polling, I get about 1Mb/sec with iperf > - with sd-irq enabled and FCLK kept on the whole time, I get 4Mb/sec > - with sd-irq enable, FCLK kept on, and the 5ms polling also > running, I get over 5Mb/sec! > Still much less than the 40Mb/sec that I would like to get... > > Thanks, > NeilBrown -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html