Skip to contentSkip to conversation

OpenPandora: make wl1251 connected to mmc3 sdio port of OpenPandora work again

PATCH Superseded H. Nikolaus Schaller 4 participants 24 messages Oct 19, 2019 – Oct 31, 2019

Subsystems (11)

v2 of 3 Superseded Latest is v3 →

lore
Share & tools
lore
#0 lore raw

v2 cover OpenPandora: make wl1251 connected to mmc3 sdio port of OpenPandora work again

* added acked-by for wl1251 patches - Kalle Valo <[email protected]>
* really removed old pdata-quirks code (not through #if 0)
* splited out a partial revert of
efdfeb079cc3b ("regulator: fixed: Convert to use GPIO descriptor only")
because that was introduced after v4.19 and stops the removal of the pdata-quirks patch from cleanly applying to v4.9, v4.14, v4.19
- reported by Sasha Levin <[email protected]>
* added a new patch to remove old omap hsmmc since pdata quirks were last user - suggested by Tony Lindgren <[email protected]>

PATCH V1 2019-10-18 22:25:39:

Here we have a set of scattered patches to make the OpenPandora WiFi work again.

v4.7 did break the pdata-quirks which made the mmc3 interface fail completely, because some code now assumes device tree based instantiation.

Fixes 81eef6ca9201 ("mmc: omap_hsmmc: Use dma_request_chan() for requesting DMA channel")

v4.11 did break the sdio qirks for wl1251 which made the driver no longer load, although the device was found as an sdio client.

Fixes 884f38607897 ("mmc: core: move some sdio IDs out of quirks file")

To solve these issues:
* we convert mmc3 and wl1251 initialization from pdata-quirks to device tree
* we make the wl1251 driver read properties from device tree
* we fix the mmc core vendor ids and quirks
* we fix the wl1251 (and wl1271) driver to use only vendor ids from header file instead of (potentially conflicting) local definitions

H. Nikolaus Schaller (11):
Documentation: dt: wireless: update wl1251 for sdio
net: wireless: ti: wl1251 add device tree support
DTS: ARM: pandora-common: define wl1251 as child node of mmc3
mmc: host: omap_hsmmc: add code for special init of wl1251 to get rid

    of pandora_wl1251_init_card

omap: pdata-quirks: revert pandora specific gpiod additions
omap: pdata-quirks: remove openpandora quirks for mmc3 and wl1251
omap: remove old hsmmc.[ch] and in Makefile
mmc: sdio: fix wl1251 vendor id
mmc: core: fix wl1251 sdio quirks
net: wireless: ti: wl1251 use new SDIO_VENDOR_ID_TI_WL1251 definition
net: wireless: ti: remove local VENDOR_ID and DEVICE_ID definitions

 .../bindings/net/wireless/ti,wl1251.txt       |  26 +++
 arch/arm/boot/dts/omap3-pandora-common.dtsi   |  37 +++-
 arch/arm/mach-omap2/Makefile                  |   1 -
 arch/arm/mach-omap2/hsmmc.c                   | 171 ------------------
 arch/arm/mach-omap2/hsmmc.h                   |  32 ----
 arch/arm/mach-omap2/pdata-quirks.c            | 105 -----------
 drivers/mmc/core/quirks.h                     |   7 +
 drivers/mmc/host/omap_hsmmc.c                 |  21 +++
 drivers/net/wireless/ti/wl1251/sdio.c         |  23 ++-
 drivers/net/wireless/ti/wlcore/sdio.c         |   8 -
 include/linux/mmc/sdio_ids.h                  |   2 +
 11 files changed, 105 insertions(+), 328 deletions(-)
 delete mode 100644 arch/arm/mach-omap2/hsmmc.c
 delete mode 100644 arch/arm/mach-omap2/hsmmc.h
Signature · 1 line
2.19.1
#1 lore raw

PATCH v2 1/11 Documentation: dt: wireless: update wl1251 for sdio

The standard method for sdio devices connected to an sdio interface is to define them as a child node like we can see with wlcore.

---
 .../bindings/net/wireless/ti,wl1251.txt       | 26 +++++++++++++++++++
 1 file changed, 26 insertions(+)
Documentation/devicetree/bindings/net/wireless/ti,wl1251.txt +26 -0
index bb2fcde6f7ff..88612ff29f2d 100644
@@ -35,3 +35,29 @@ Examples:
 		ti,power-gpio = <&gpio3 23 GPIO_ACTIVE_HIGH>; /* 87 */
 	};
 };
+
+&mmc3 {
+	vmmc-supply = <&wlan_en>;
+
+	bus-width = <4>;
+	non-removable;
+	ti,non-removable;
+	cap-power-off-card;
+
+	pinctrl-names = "default";
+	pinctrl-0 = <&mmc3_pins>;
+
+	#address-cells = <1>;
+	#size-cells = <0>;
+
+	wlan: wl1251@1 {
+		compatible = "ti,wl1251";
+
+		reg = <1>;
+
+		interrupt-parent = <&gpio1>;
+		interrupts = <21 IRQ_TYPE_LEVEL_HIGH>;	/* GPIO_21 */
+
+		ti,wl1251-has-eeprom;
+	};
+};
Signature · 1 line
2.19.1
#4 lore raw

PATCH v2 2/11 net: wireless: ti: wl1251 add device tree support

In reply to H. Nikolaus Schaller

We will have the wl1251 defined as a child node of the mmc interface and can read setup for gpios, interrupts and the ti,use-eeprom property from there instead of pdata to be provided by pdata-quirks.

Fixes 81eef6ca9201 ("mmc: omap_hsmmc: Use dma_request_chan() for requesting DMA channel")
---
 drivers/net/wireless/ti/wl1251/sdio.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)
drivers/net/wireless/ti/wl1251/sdio.c +13 -0
index 677f1146ccf0..c54a273713ed 100644
@@ -16,6 +16,9 @@
 #include <linux/irq.h>
 #include <linux/pm_runtime.h>
 #include <linux/gpio.h>
+#include <linux/of.h>
+#include <linux/of_gpio.h>
+#include <linux/of_irq.h>
 
 #include "wl1251.h"
 
@@ -217,6 +220,7 @@ static int wl1251_sdio_probe(struct sdio_func *func,
 	struct ieee80211_hw *hw;
 	struct wl1251_sdio *wl_sdio;
 	const struct wl1251_platform_data *wl1251_board_data;
+	struct device_node *np = func->dev.of_node;
 
 	hw = wl1251_alloc_hw();
 	if (IS_ERR(hw))
@@ -248,6 +252,15 @@ static int wl1251_sdio_probe(struct sdio_func *func,
 		wl->power_gpio = wl1251_board_data->power_gpio;
 		wl->irq = wl1251_board_data->irq;
 		wl->use_eeprom = wl1251_board_data->use_eeprom;
+	} else if (np) {
+		wl->use_eeprom =of_property_read_bool(np, "ti,wl1251-has-eeprom");
+		wl->power_gpio = of_get_named_gpio(np, "ti,power-gpio", 0);
+		wl->irq = of_irq_get(np, 0);
+
+		if (wl->power_gpio == -EPROBE_DEFER || wl->irq == -EPROBE_DEFER) {
+			ret = -EPROBE_DEFER;
+			goto disable;
+		}
 	}
 
 	if (gpio_is_valid(wl->power_gpio)) {
Signature · 1 line
2.19.1
#5 lore raw

PATCH v2 3/11 DTS: ARM: pandora-common: define wl1251 as child node of mmc3

In reply to H. Nikolaus Schaller

Since v4.7 the dma initialization requires that there is a device tree property for "rx" and "tx" channels which is not provided by the pdata-quirks initialization.

By conversion of the mmc3 setup to device tree this will finally allows to remove the OpenPandora wlan specific omap3 data-quirks.

Fixes 81eef6ca9201 ("mmc: omap_hsmmc: Use dma_request_chan() for requesting DMA channel")
---
 arch/arm/boot/dts/omap3-pandora-common.dtsi | 37 +++++++++++++++++++--
 1 file changed, 35 insertions(+), 2 deletions(-)
arch/arm/boot/dts/omap3-pandora-common.dtsi +35 -2
index ec5891718ae6..c595b3eb314d 100644
@@ -226,6 +226,18 @@
 		gpio = <&gpio6 4 GPIO_ACTIVE_HIGH>;	/* GPIO_164 */
 	};
 
+	/* wl1251 wifi+bt module */
+	wlan_en: fixed-regulator-wg7210_en {
+		compatible = "regulator-fixed";
+		regulator-name = "vwlan";
+		regulator-min-microvolt = <1800000>;
+		regulator-max-microvolt = <1800000>;
+		startup-delay-us = <50000>;
+		regulator-always-on;
+		enable-active-high;
+		gpio = <&gpio1 23 GPIO_ACTIVE_HIGH>;
+	};
+
 	/* wg7210 (wifi+bt module) 32k clock buffer */
 	wg7210_32k: fixed-regulator-wg7210_32k {
 		compatible = "regulator-fixed";
@@ -522,9 +534,30 @@
 	/*wp-gpios = <&gpio4 31 GPIO_ACTIVE_HIGH>;*/	/* GPIO_127 */
 };
 
-/* mmc3 is probed using pdata-quirks to pass wl1251 card data */
 &mmc3 {
-	status = "disabled";
+	vmmc-supply = <&wlan_en>;
+
+	bus-width = <4>;
+	non-removable;
+	ti,non-removable;
+	cap-power-off-card;
+
+	pinctrl-names = "default";
+	pinctrl-0 = <&mmc3_pins>;
+
+	#address-cells = <1>;
+	#size-cells = <0>;
+
+	wlan: wl1251@1 {
+		compatible = "ti,wl1251";
+
+		reg = <1>;
+
+		interrupt-parent = <&gpio1>;
+		interrupts = <21 IRQ_TYPE_LEVEL_HIGH>;	/* GPIO_21 */
+
+		ti,wl1251-has-eeprom;
+	};
 };
 
 /* bluetooth*/
Signature · 1 line
2.19.1
#9 lore raw

PATCH v2 4/11 mmc: host: omap_hsmmc: add code for special init of wl1251 to get rid of pandora_wl1251_init_card

In reply to H. Nikolaus Schaller

Pandora_wl1251_init_card was used to do special pdata based setup of the sdio mmc interface. This does no longer work with v4.7 and later. A fix requires a device tree based mmc3 setup.

Therefore we move the special setup to omap_hsmmc.c instead of calling some pdata supplied init_card function.

The new code checks for a DT child node compatible to wl1251 so it will not affect other MMC3 use cases.

Fixes 81eef6ca9201 ("mmc: omap_hsmmc: Use dma_request_chan() for requesting DMA channel")
---
 drivers/mmc/host/omap_hsmmc.c | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)
drivers/mmc/host/omap_hsmmc.c +21 -0
index 952fa4063ff8..03ba80bcf319 100644
@@ -1512,6 +1512,27 @@ static void omap_hsmmc_init_card(struct mmc_host *mmc, struct mmc_card *card)
 
 	if (mmc_pdata(host)->init_card)
 		mmc_pdata(host)->init_card(card);
+	else if (card->type == MMC_TYPE_SDIO || card->type == MMC_TYPE_SD_COMBO) {
+		struct device_node *np = mmc_dev(mmc)->of_node;
+
+		np = of_get_compatible_child(np, "ti,wl1251");
+		if (np) {
+			/*
+			 * We have TI wl1251 attached to MMC3. Pass this information to
+			 * SDIO core because it can't be probed by normal methods.
+			 */
+
+			dev_info(host->dev, "found wl1251\n");
+			card->quirks |= MMC_QUIRK_NONSTD_SDIO;
+			card->cccr.wide_bus = 1;
+			card->cis.vendor = 0x104c;
+			card->cis.device = 0x9066;
+			card->cis.blksize = 512;
+			card->cis.max_dtr = 24000000;
+			card->ocr = 0x80;
+			of_node_put(np);
+		}
+	}
 }
 
 static void omap_hsmmc_enable_sdio_irq(struct mmc_host *mmc, int enable)
Signature · 1 line
2.19.1
#13 lore raw

PATCH v2 5/11 omap: pdata-quirks: revert pandora specific gpiod additions

In reply to H. Nikolaus Schaller

introduced by commit

efdfeb079cc3b ("regulator: fixed: Convert to use GPIO descriptor only")

We must remove this from mainline first, so that the following patch to remove the openpandora quirks for mmc3 and wl1251 cleanly applies to stable v4.9, v4.14, v4.19 where the above mentioned patch is not yet present.

---
 arch/arm/mach-omap2/pdata-quirks.c | 19 ++++---------------
 1 file changed, 4 insertions(+), 15 deletions(-)
arch/arm/mach-omap2/pdata-quirks.c +4 -15
index d942a3357090..89734ef9ab1e 100644
@@ -7,7 +7,6 @@
 #include <linux/clk.h>
 #include <linux/davinci_emac.h>
 #include <linux/gpio.h>
-#include <linux/gpio/machine.h>
 #include <linux/init.h>
 #include <linux/kernel.h>
 #include <linux/of_platform.h>
@@ -327,7 +326,9 @@ static struct regulator_init_data pandora_vmmc3 = {
 static struct fixed_voltage_config pandora_vwlan = {
 	.supply_name		= "vwlan",
 	.microvolts		= 1800000, /* 1.8V */
+	.gpio			= PANDORA_WIFI_NRESET_GPIO,
 	.startup_delay		= 50000, /* 50ms */
+	.enable_high		= 1,
 	.init_data		= &pandora_vmmc3,
 };
 
@@ -339,19 +340,6 @@ static struct platform_device pandora_vwlan_device = {
 	},
 };
 
-static struct gpiod_lookup_table pandora_vwlan_gpiod_table = {
-	.dev_id = "reg-fixed-voltage.1",
-	.table = {
-		/*
-		 * As this is a low GPIO number it should be at the first
-		 * GPIO bank.
-		 */
-		GPIO_LOOKUP("gpio-0-31", PANDORA_WIFI_NRESET_GPIO,
-			    NULL, GPIO_ACTIVE_HIGH),
-		{ },
-	},
-};
-
 static void pandora_wl1251_init_card(struct mmc_card *card)
 {
 	/*
@@ -373,6 +361,8 @@ static struct omap2_hsmmc_info pandora_mmc3[] = {
 	{
 		.mmc		= 3,
 		.caps		= MMC_CAP_4_BIT_DATA | MMC_CAP_POWER_OFF_CARD,
+		.gpio_cd	= -EINVAL,
+		.gpio_wp	= -EINVAL,
 		.init_card	= pandora_wl1251_init_card,
 	},
 	{}	/* Terminator */
@@ -411,7 +401,6 @@ static void __init pandora_wl1251_init(void)
 static void __init omap3_pandora_legacy_init(void)
 {
 	platform_device_register(&pandora_backlight);
-	gpiod_add_lookup_table(&pandora_vwlan_gpiod_table);
 	platform_device_register(&pandora_vwlan_device);
 	omap_hsmmc_init(pandora_mmc3);
 	omap_hsmmc_late_init(pandora_mmc3);
Signature · 1 line
2.19.1
#14 lore raw

PATCH v2 6/11 omap: pdata-quirks: remove openpandora quirks for mmc3 and wl1251

In reply to H. Nikolaus Schaller

With a wl1251 child node of mmc3 in the device tree decoded in omap_hsmmc.c to handle special wl1251 initialization, we do no longer need to instantiate the mmc3 through pdata quirks.

We also can remove the wlan regulator and reset/interrupt definitions and do them through device tree.

Fixes 81eef6ca9201 ("mmc: omap_hsmmc: Use dma_request_chan() for requesting DMA channel")
---
 arch/arm/mach-omap2/pdata-quirks.c | 93 ------------------------------
 1 file changed, 93 deletions(-)
arch/arm/mach-omap2/pdata-quirks.c +0 -93
index 89734ef9ab1e..ecc1ef632951 100644
@@ -303,108 +303,15 @@ static void __init omap3_logicpd_torpedo_init(void)
 }
 
 /* omap3pandora legacy devices */
-#define PANDORA_WIFI_IRQ_GPIO		21
-#define PANDORA_WIFI_NRESET_GPIO	23
 
 static struct platform_device pandora_backlight = {
 	.name	= "pandora-backlight",
 	.id	= -1,
 };
 
-static struct regulator_consumer_supply pandora_vmmc3_supply[] = {
-	REGULATOR_SUPPLY("vmmc", "omap_hsmmc.2"),
-};
-
-static struct regulator_init_data pandora_vmmc3 = {
-	.constraints = {
-		.valid_ops_mask		= REGULATOR_CHANGE_STATUS,
-	},
-	.num_consumer_supplies	= ARRAY_SIZE(pandora_vmmc3_supply),
-	.consumer_supplies	= pandora_vmmc3_supply,
-};
-
-static struct fixed_voltage_config pandora_vwlan = {
-	.supply_name		= "vwlan",
-	.microvolts		= 1800000, /* 1.8V */
-	.gpio			= PANDORA_WIFI_NRESET_GPIO,
-	.startup_delay		= 50000, /* 50ms */
-	.enable_high		= 1,
-	.init_data		= &pandora_vmmc3,
-};
-
-static struct platform_device pandora_vwlan_device = {
-	.name		= "reg-fixed-voltage",
-	.id		= 1,
-	.dev = {
-		.platform_data = &pandora_vwlan,
-	},
-};
-
-static void pandora_wl1251_init_card(struct mmc_card *card)
-{
-	/*
-	 * We have TI wl1251 attached to MMC3. Pass this information to
-	 * SDIO core because it can't be probed by normal methods.
-	 */
-	if (card->type == MMC_TYPE_SDIO || card->type == MMC_TYPE_SD_COMBO) {
-		card->quirks |= MMC_QUIRK_NONSTD_SDIO;
-		card->cccr.wide_bus = 1;
-		card->cis.vendor = 0x104c;
-		card->cis.device = 0x9066;
-		card->cis.blksize = 512;
-		card->cis.max_dtr = 24000000;
-		card->ocr = 0x80;
-	}
-}
-
-static struct omap2_hsmmc_info pandora_mmc3[] = {
-	{
-		.mmc		= 3,
-		.caps		= MMC_CAP_4_BIT_DATA | MMC_CAP_POWER_OFF_CARD,
-		.gpio_cd	= -EINVAL,
-		.gpio_wp	= -EINVAL,
-		.init_card	= pandora_wl1251_init_card,
-	},
-	{}	/* Terminator */
-};
-
-static void __init pandora_wl1251_init(void)
-{
-	struct wl1251_platform_data pandora_wl1251_pdata;
-	int ret;
-
-	memset(&pandora_wl1251_pdata, 0, sizeof(pandora_wl1251_pdata));
-
-	pandora_wl1251_pdata.power_gpio = -1;
-
-	ret = gpio_request_one(PANDORA_WIFI_IRQ_GPIO, GPIOF_IN, "wl1251 irq");
-	if (ret < 0)
-		goto fail;
-
-	pandora_wl1251_pdata.irq = gpio_to_irq(PANDORA_WIFI_IRQ_GPIO);
-	if (pandora_wl1251_pdata.irq < 0)
-		goto fail_irq;
-
-	pandora_wl1251_pdata.use_eeprom = true;
-	ret = wl1251_set_platform_data(&pandora_wl1251_pdata);
-	if (ret < 0)
-		goto fail_irq;
-
-	return;
-
-fail_irq:
-	gpio_free(PANDORA_WIFI_IRQ_GPIO);
-fail:
-	pr_err("wl1251 board initialisation failed\n");
-}
-
 static void __init omap3_pandora_legacy_init(void)
 {
 	platform_device_register(&pandora_backlight);
-	platform_device_register(&pandora_vwlan_device);
-	omap_hsmmc_init(pandora_mmc3);
-	omap_hsmmc_late_init(pandora_mmc3);
-	pandora_wl1251_init();
 }
 #endif /* CONFIG_ARCH_OMAP3 */
 
Signature · 1 line
2.19.1
#15 lore raw

PATCH v2 7/11 omap: remove old hsmmc.[ch] and in Makefile

In reply to H. Nikolaus Schaller

There is a new one in drivers/mmc/host/omap_hsmmc.c configured by CONFIG_MMC_OMAP_HS and the last user was the pdata-quirks for pandora.

---
 arch/arm/mach-omap2/Makefile       |   1 -
 arch/arm/mach-omap2/hsmmc.c        | 171 -----------------------------
 arch/arm/mach-omap2/hsmmc.h        |  32 ------
 arch/arm/mach-omap2/pdata-quirks.c |   1 -
 4 files changed, 205 deletions(-)
 delete mode 100644 arch/arm/mach-omap2/hsmmc.c
 delete mode 100644 arch/arm/mach-omap2/hsmmc.h
arch/arm/mach-omap2/Makefile +0 -1
index 8f208197988f..1b3062f32899 100644
@@ -216,7 +216,6 @@ obj-$(CONFIG_MACH_NOKIA_N8X0)		+= board-n8x0.o
 
 # Platform specific device init code
 
-omap-hsmmc-$(CONFIG_MMC_OMAP_HS)	:= hsmmc.o
 obj-y					+= $(omap-hsmmc-m) $(omap-hsmmc-y)
 
 obj-y					+= omap_phy_internal.o
arch/arm/mach-omap2/hsmmc.c deleted +0 -171

171 lines not on this page · raw message

arch/arm/mach-omap2/hsmmc.h deleted +0 -32
deleted file mode 100644
index 76c5ed2afa72..000000000000
@@ -1,32 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0-only */
-/*
- * MMC definitions for OMAP2
- */
-
-struct mmc_card;
-
-struct omap2_hsmmc_info {
-	u8	mmc;		/* controller 1/2/3 */
-	u32	caps;		/* 4/8 wires and any additional host
-				 * capabilities OR'd (ref. linux/mmc/host.h) */
-	struct platform_device *pdev;	/* mmc controller instance */
-	/* init some special card */
-	void (*init_card)(struct mmc_card *card);
-};
-
-#if IS_ENABLED(CONFIG_MMC_OMAP_HS)
-
-void omap_hsmmc_init(struct omap2_hsmmc_info *);
-void omap_hsmmc_late_init(struct omap2_hsmmc_info *);
-
-#else
-
-static inline void omap_hsmmc_init(struct omap2_hsmmc_info *info)
-{
-}
-
-static inline void omap_hsmmc_late_init(struct omap2_hsmmc_info *info)
-{
-}
-
-#endif
arch/arm/mach-omap2/pdata-quirks.c +0 -1
index ecc1ef632951..4ce05d408a67 100644
@@ -32,7 +32,6 @@
 #include "omap_device.h"
 #include "omap-secure.h"
 #include "soc.h"
-#include "hsmmc.h"
 
 static struct omap_hsmmc_platform_data __maybe_unused mmc_pdata[2];
 
Signature · 1 line
2.19.1
#20 lore raw

PATCH v2 8/11 mmc: sdio: fix wl1251 vendor id

In reply to H. Nikolaus Schaller

v4.11-rc1 did introduce a patch series that rearranged the sdio quirks into a header file. Unfortunately this did forget to handle SDIO_VENDOR_ID_TI differently between wl1251 and wl1271 with the result that although the wl1251 was found on the sdio bus, the firmware did not load any more and there was no interface registration.

This patch defines separate constants to be used by sdio quirks and drivers.

Fixes 884f38607897 ("mmc: core: move some sdio IDs out of quirks file")
---
 include/linux/mmc/sdio_ids.h | 2 ++
 1 file changed, 2 insertions(+)
include/linux/mmc/sdio_ids.h +2 -0
index d1a5d5df02f5..08b25c02b5a1 100644
@@ -71,6 +71,8 @@
 
 #define SDIO_VENDOR_ID_TI			0x0097
 #define SDIO_DEVICE_ID_TI_WL1271		0x4076
+#define SDIO_VENDOR_ID_TI_WL1251		0x104c
+#define SDIO_DEVICE_ID_TI_WL1251		0x9066
 
 #define SDIO_VENDOR_ID_STE			0x0020
 #define SDIO_DEVICE_ID_STE_CW1200		0x2280
Signature · 1 line
2.19.1
#21 lore raw

PATCH v2 9/11 mmc: core: fix wl1251 sdio quirks

In reply to H. Nikolaus Schaller

wl1251 and wl1271 have different vendor id and device id. So we need to handle both with sdio quirks.

Fixes 884f38607897 ("mmc: core: move some sdio IDs out of quirks file")
---
 drivers/mmc/core/quirks.h | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/mmc/core/quirks.h +7 -0
index 2d2d9ea8be4f..3dba15bccce2 100644
@@ -119,7 +119,14 @@ static const struct mmc_fixup mmc_ext_csd_fixups[] = {
 	END_FIXUP
 };
 
+
 static const struct mmc_fixup sdio_fixup_methods[] = {
+	SDIO_FIXUP(SDIO_VENDOR_ID_TI_WL1251, SDIO_DEVICE_ID_TI_WL1251,
+		   add_quirk, MMC_QUIRK_NONSTD_FUNC_IF),
+
+	SDIO_FIXUP(SDIO_VENDOR_ID_TI_WL1251, SDIO_DEVICE_ID_TI_WL1251,
+		   add_quirk, MMC_QUIRK_DISABLE_CD),
+
 	SDIO_FIXUP(SDIO_VENDOR_ID_TI, SDIO_DEVICE_ID_TI_WL1271,
 		   add_quirk, MMC_QUIRK_NONSTD_FUNC_IF),
 
Signature · 1 line
2.19.1
#22 lore raw

PATCH v2 10/11 net: wireless: ti: wl1251 use new SDIO_VENDOR_ID_TI_WL1251 definition

In reply to H. Nikolaus Schaller

SDIO_VENDOR_ID_TI_WL1251 is now defined in mmc/sdio_ids.h separately from SDIO_VENDOR_ID_TI for wl1271.

Fixes 884f38607897 ("mmc: core: move some sdio IDs out of quirks file")
---
 drivers/net/wireless/ti/wl1251/sdio.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/wireless/ti/wl1251/sdio.c +1 -1
index c54a273713ed..42b55f3a50df 100644
@@ -52,7 +52,7 @@ static void wl1251_sdio_interrupt(struct sdio_func *func)
 }
 
 static const struct sdio_device_id wl1251_devices[] = {
-	{ SDIO_DEVICE(SDIO_VENDOR_ID_TI, SDIO_DEVICE_ID_TI_WL1251) },
+	{ SDIO_DEVICE(SDIO_VENDOR_ID_TI_WL1251, SDIO_DEVICE_ID_TI_WL1251) },
 	{}
 };
 MODULE_DEVICE_TABLE(sdio, wl1251_devices);
Signature · 1 line
2.19.1
#23 lore raw

PATCH v2 11/11 net: wireless: ti: remove local VENDOR_ID and DEVICE_ID definitions

In reply to H. Nikolaus Schaller

They are already included from mmc/sdio_ids.h and do not need a local definition.

Fixes 884f38607897 ("mmc: core: move some sdio IDs out of quirks file")
---
 drivers/net/wireless/ti/wl1251/sdio.c | 8 --------
 drivers/net/wireless/ti/wlcore/sdio.c | 8 --------
 2 files changed, 16 deletions(-)
drivers/net/wireless/ti/wl1251/sdio.c +0 -8
index 42b55f3a50df..3c4d5e38c66c 100644
@@ -22,14 +22,6 @@
 
 #include "wl1251.h"
 
-#ifndef SDIO_VENDOR_ID_TI
-#define SDIO_VENDOR_ID_TI		0x104c
-#endif
-
-#ifndef SDIO_DEVICE_ID_TI_WL1251
-#define SDIO_DEVICE_ID_TI_WL1251	0x9066
-#endif
-
 struct wl1251_sdio {
 	struct sdio_func *func;
 	u32 elp_val;
drivers/net/wireless/ti/wlcore/sdio.c +0 -8
index 7afaf35f2453..9fd8cf2d270c 100644
@@ -26,14 +26,6 @@
 #include "wl12xx_80211.h"
 #include "io.h"
 
-#ifndef SDIO_VENDOR_ID_TI
-#define SDIO_VENDOR_ID_TI		0x0097
-#endif
-
-#ifndef SDIO_DEVICE_ID_TI_WL1271
-#define SDIO_DEVICE_ID_TI_WL1271	0x4076
-#endif
-
 static bool dump = false;
 
 struct wl12xx_sdio_glue {
Signature · 1 line
2.19.1
maintainer
#16 lore raw

Re: [PATCH v2 07/11] omap: remove old hsmmc.[ch] and in Makefile

In reply to H. Nikolaus Schaller

arch/arm/mach-omap2/Makefile obj-$(CONFIG_MACH_NOKIA_N8X0) += board-n8x0.o · line 219

Quoted 8 lines · H. Nikolaus Schaller
arch/arm/mach-omap2/Makefile +0 -1
@@ -216,7 +216,6 @@ obj-$(CONFIG_MACH_NOKIA_N8X0)		+= board-n8x0.o
 
 # Platform specific device init code
 
-omap-hsmmc-$(CONFIG_MMC_OMAP_HS)	:= hsmmc.o
 obj-y					+= $(omap-hsmmc-m) $(omap-hsmmc-y)

The related obj-y line can go now too, right?

And looks like common.h also has struct omap2_hsmmc_info so maybe check by grepping for hsmmc_info to see it's gone.

Regards,

Tony

#17 lore raw

Re: [PATCH v2 07/11] omap: remove old hsmmc.[ch] and in Makefile

Quoted 8 lines

Am 21.10.2019 um 16:30 schrieb Tony Lindgren <[email protected]>:

Quoted 4 lines · H. Nikolaus Schaller
arch/arm/mach-omap2/Makefile +0 -0
@@ -216,7 +216,6 @@ obj-$(CONFIG_MACH_NOKIA_N8X0)		+= board-n8x0.o
 
Quoted 6 lines

arch/arm/mach-omap2/Makefile obj-$(CONFIG_MACH_NOKIA_N8X0) += board-n8x0.o · line 219

Quoted 4 lines
arch/arm/mach-omap2/Makefile +0 -1
 # Platform specific device init code
 
-omap-hsmmc-$(CONFIG_MMC_OMAP_HS)	:= hsmmc.o
 obj-y					+= $(omap-hsmmc-m) $(omap-hsmmc-y)

The related obj-y line can go now too, right?

Yes, I think so. It is a construction that I have never seen before :) Therefore I did not recognize that it is related.

Quoted 2 lines

And looks like common.h also has struct omap2_hsmmc_info so maybe check by grepping for hsmmc_info to see it's gone.

Yes, it is just a forward-declaration of the struct name with no user anywhere.

Scheduled for v3.

BTW: should this series go through your tree since it is an omap machine?

BR and thanks,
Nikolaus

maintainer
#18 lore raw

Re: [PATCH v2 07/11] omap: remove old hsmmc.[ch] and in Makefile

Quoted 21 lines · H. Nikolaus Schaller
Quoted 8 lines

Am 21.10.2019 um 16:30 schrieb Tony Lindgren <[email protected]>:

Quoted 4 lines · H. Nikolaus Schaller
arch/arm/mach-omap2/Makefile +0 -0
@@ -216,7 +216,6 @@ obj-$(CONFIG_MACH_NOKIA_N8X0)		+= board-n8x0.o
 
Quoted 6 lines

arch/arm/mach-omap2/Makefile obj-$(CONFIG_MACH_NOKIA_N8X0) += board-n8x0.o · line 219

Quoted 4 lines
arch/arm/mach-omap2/Makefile +0 -1
 # Platform specific device init code
 
-omap-hsmmc-$(CONFIG_MMC_OMAP_HS)	:= hsmmc.o
 obj-y					+= $(omap-hsmmc-m) $(omap-hsmmc-y)

The related obj-y line can go now too, right?

Yes, I think so. It is a construction that I have never seen before :) Therefore I did not recognize that it is related.

Quoted 2 lines

And looks like common.h also has struct omap2_hsmmc_info so maybe check by grepping for hsmmc_info to see it's gone.

Quoted 7 lines

Yes, it is just a forward-declaration of the struct name with no user anywhere.

Scheduled for v3.

BTW: should this series go through your tree since it is an omap machine?

Or MMC tree as that's where the code change really are.

Regards,

Tony

#19 lore raw

Re: [PATCH v2 07/11] omap: remove old hsmmc.[ch] and in Makefile

- Trimmed cc-list (could be a good idea for next submission as well)

Quoted 26 lines · Tony Lindgren
Quoted 24 lines · H. Nikolaus Schaller
Quoted 8 lines

Am 21.10.2019 um 16:30 schrieb Tony Lindgren <[email protected]>:

Quoted 4 lines · H. Nikolaus Schaller
arch/arm/mach-omap2/Makefile +0 -0
@@ -216,7 +216,6 @@ obj-$(CONFIG_MACH_NOKIA_N8X0)          += board-n8x0.o
 
Quoted 6 lines

arch/arm/mach-omap2/Makefile obj-$(CONFIG_MACH_NOKIA_N8X0) += board-n8x0.o · line 219

Quoted 4 lines
arch/arm/mach-omap2/Makefile +0 -1
 # Platform specific device init code
 
-omap-hsmmc-$(CONFIG_MMC_OMAP_HS)  := hsmmc.o
 obj-y                                      += $(omap-hsmmc-m) $(omap-hsmmc-y)

The related obj-y line can go now too, right?

Yes, I think so. It is a construction that I have never seen before :) Therefore I did not recognize that it is related.

Quoted 2 lines

And looks like common.h also has struct omap2_hsmmc_info so maybe check by grepping for hsmmc_info to see it's gone.

Yes, it is just a forward-declaration of the struct name with no user anywhere.

Quoted 6 lines
Quoted 4 lines

Scheduled for v3.

BTW: should this series go through your tree since it is an omap machine?

Or MMC tree as that's where the code change really are.

I am okay with that. I will have a look at the series and provide some comments.

Kind regards
Uffe

maintainer
#2 lore raw

Re: [PATCH v2 01/11] Documentation: dt: wireless: update wl1251 for sdio

In reply to H. Nikolaus Schaller
Quoted 19 lines · H. Nikolaus Schaller

The standard method for sdio devices connected to an sdio interface is to define them as a child node like we can see with wlcore.

---
 .../bindings/net/wireless/ti,wl1251.txt       | 26 +++++++++++++++++++
 1 file changed, 26 insertions(+)
Documentation/devicetree/bindings/net/wireless/ti,wl1251.txt +1 -0
index bb2fcde6f7ff..88612ff29f2d 100644
@@ -35,3 +35,29 @@ Examples:
 		ti,power-gpio = <&gpio3 23 GPIO_ACTIVE_HIGH>; /* 87 */
 	};
 };
+

Documentation/devicetree/bindings/net/wireless/ti,wl1251.txt Examples: · line 48

Quoted 10 lines
Documentation/devicetree/bindings/net/wireless/ti,wl1251.txt +10 -0
+&mmc3 {
+	vmmc-supply = <&wlan_en>;
+
+	bus-width = <4>;
+	non-removable;
+	ti,non-removable;
+	cap-power-off-card;
+
+	pinctrl-names = "default";
+	pinctrl-0 = <&mmc3_pins>;

None of the above are really relevant to this binding.

Documentation/devicetree/bindings/net/wireless/ti,wl1251.txt Examples: · line 53

Quoted 5 lines
Documentation/devicetree/bindings/net/wireless/ti,wl1251.txt +5 -0
+
+	#address-cells = <1>;
+	#size-cells = <0>;
+
+	wlan: wl1251@1 {

wifi@1

Quoted 18 lines
Documentation/devicetree/bindings/net/wireless/ti,wl1251.txt +10 -0
+		compatible = "ti,wl1251";
+
+		reg = <1>;
+
+		interrupt-parent = <&gpio1>;
+		interrupts = <21 IRQ_TYPE_LEVEL_HIGH>;	/* GPIO_21 */
+
+		ti,wl1251-has-eeprom;
+	};
+};
Signature · 7 lines
2.19.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
#3 lore raw

Re: [PATCH v2 01/11] Documentation: dt: wireless: update wl1251 for sdio

Quoted 27 lines

Am 25.10.2019 um 23:13 schrieb Rob Herring <[email protected]>:

Quoted 23 lines · H. Nikolaus Schaller

The standard method for sdio devices connected to an sdio interface is to define them as a child node like we can see with wlcore.

---
.../bindings/net/wireless/ti,wl1251.txt | 26 +++++++++++++++++++

1 file changed, 26 insertions(+)
Documentation/devicetree/bindings/net/wireless/ti,wl1251.txt +5 -0
index bb2fcde6f7ff..88612ff29f2d 100644
@@ -35,3 +35,29 @@ Examples:
 		ti,power-gpio = <&gpio3 23 GPIO_ACTIVE_HIGH>; /* 87 */
 	};
 };
+
+&mmc3 {
+	vmmc-supply = <&wlan_en>;
+
+	bus-width = <4>;
Quoted 8 lines

Documentation/devicetree/bindings/net/wireless/ti,wl1251.txt Examples: · line 48

Quoted 6 lines
Documentation/devicetree/bindings/net/wireless/ti,wl1251.txt +6 -0
+	non-removable;
+	ti,non-removable;
+	cap-power-off-card;
+
+	pinctrl-names = "default";
+	pinctrl-0 = <&mmc3_pins>;

None of the above are really relevant to this binding.

Ok, but how and where do we document that they are needed to make both ends of the interface work together?

Quoted 8 lines

Documentation/devicetree/bindings/net/wireless/ti,wl1251.txt Examples: · line 53

Quoted 6 lines
Documentation/devicetree/bindings/net/wireless/ti,wl1251.txt +5 -0
+
+	#address-cells = <1>;
+	#size-cells = <0>;
+
+	wlan: wl1251@1 {

wifi@1

Ok.

Quoted 14 lines
Quoted 14 lines
Documentation/devicetree/bindings/net/wireless/ti,wl1251.txt +10 -0
+		compatible = "ti,wl1251";
+
+		reg = <1>;
+
+		interrupt-parent = <&gpio1>;
+		interrupts = <21 IRQ_TYPE_LEVEL_HIGH>;	/* GPIO_21 */
+
+		ti,wl1251-has-eeprom;
+	};
+};
Signature · 1 line
2.19.1

BR and thanks,
Nikolaus

#10 lore raw

Re: [PATCH v2 04/11] mmc: host: omap_hsmmc: add code for special init of wl1251 to get rid of pandora_wl1251_init_card

In reply to H. Nikolaus Schaller
Quoted 35 lines · H. Nikolaus Schaller

Pandora_wl1251_init_card was used to do special pdata based setup of the sdio mmc interface. This does no longer work with v4.7 and later. A fix requires a device tree based mmc3 setup.

Therefore we move the special setup to omap_hsmmc.c instead of calling some pdata supplied init_card function.

The new code checks for a DT child node compatible to wl1251 so it will not affect other MMC3 use cases.

Fixes 81eef6ca9201 ("mmc: omap_hsmmc: Use dma_request_chan() for requesting DMA channel")
---
 drivers/mmc/host/omap_hsmmc.c | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)
drivers/mmc/host/omap_hsmmc.c +8 -0
index 952fa4063ff8..03ba80bcf319 100644
@@ -1512,6 +1512,27 @@ static void omap_hsmmc_init_card(struct mmc_host *mmc, struct mmc_card *card)
 
        if (mmc_pdata(host)->init_card)
                mmc_pdata(host)->init_card(card);
+       else if (card->type == MMC_TYPE_SDIO || card->type == MMC_TYPE_SD_COMBO) {
+               struct device_node *np = mmc_dev(mmc)->of_node;
+
+               np = of_get_compatible_child(np, "ti,wl1251");
+               if (np) {
+                       /*
+                        * We have TI wl1251 attached to MMC3. Pass this information to
+                        * SDIO core because it can't be probed by normal methods.

drivers/mmc/host/omap_hsmmc.c static void omap_hsmmc_init_card(struct mmc_host *mmc, struct mmc_card *card) · line 1532

Quoted 10 lines
drivers/mmc/host/omap_hsmmc.c +10 -0
+                        */
+
+                       dev_info(host->dev, "found wl1251\n");
+                       card->quirks |= MMC_QUIRK_NONSTD_SDIO;
+                       card->cccr.wide_bus = 1;
+                       card->cis.vendor = 0x104c;
+                       card->cis.device = 0x9066;
+                       card->cis.blksize = 512;
+                       card->cis.max_dtr = 24000000;
+                       card->ocr = 0x80;

These things should really be figured out by the mmc core during SDIO card initialization itself, not via the host ops ->init_card() callback. That is just poor hack, which in the long run should go away.

Moreover, I think we should add a subnode to the host node in the DT, to describe the embedded SDIO card, rather than parsing the subnode for the SDIO func - as that seems wrong to me.

To add a subnode for the SDIO card, we already have a binding that I think we should extend. Please have a look at Documentation/devicetree/bindings/mmc/mmc-card.txt.

If you want an example of how to implement this for your case, do a git grep "broken-hpi" in the driver/mmc/core/, I think it will tell you more of what I have in mind.

Quoted 9 lines
drivers/mmc/host/omap_hsmmc.c +3 -1
+                       of_node_put(np);
+               }
+       }
 }
 
 static void omap_hsmmc_enable_sdio_irq(struct mmc_host *mmc, int enable)
--

2.19.1

Kind regards
Uffe

#6 lore raw

Re: [PATCH v2 03/11] DTS: ARM: pandora-common: define wl1251 as child node of mmc3

In reply to H. Nikolaus Schaller
Quoted 22 lines · H. Nikolaus Schaller

Since v4.7 the dma initialization requires that there is a device tree property for "rx" and "tx" channels which is not provided by the pdata-quirks initialization.

By conversion of the mmc3 setup to device tree this will finally allows to remove the OpenPandora wlan specific omap3 data-quirks.

Fixes 81eef6ca9201 ("mmc: omap_hsmmc: Use dma_request_chan() for requesting DMA channel")
---
 arch/arm/boot/dts/omap3-pandora-common.dtsi | 37 +++++++++++++++++++--
 1 file changed, 35 insertions(+), 2 deletions(-)
arch/arm/boot/dts/omap3-pandora-common.dtsi +0 -0
index ec5891718ae6..c595b3eb314d 100644

arch/arm/boot/dts/omap3-pandora-common.dtsi line 234

Quoted 9 lines
arch/arm/boot/dts/omap3-pandora-common.dtsi +6 -0
@@ -226,6 +226,18 @@
                gpio = <&gpio6 4 GPIO_ACTIVE_HIGH>;     /* GPIO_164 */
        };
 
+       /* wl1251 wifi+bt module */
+       wlan_en: fixed-regulator-wg7210_en {
+               compatible = "regulator-fixed";
+               regulator-name = "vwlan";
+               regulator-min-microvolt = <1800000>;
+               regulator-max-microvolt = <1800000>;

I doubt these are correct.

I guess this should be in the range of 2.7V-3.6V.

arch/arm/boot/dts/omap3-pandora-common.dtsi line 236

Quoted 2 lines
arch/arm/boot/dts/omap3-pandora-common.dtsi +2 -0
+               startup-delay-us = <50000>;
+               regulator-always-on;

Always on?

Quoted 41 lines
arch/arm/boot/dts/omap3-pandora-common.dtsi +27 -3
+               enable-active-high;
+               gpio = <&gpio1 23 GPIO_ACTIVE_HIGH>;
+       };
+
        /* wg7210 (wifi+bt module) 32k clock buffer */
        wg7210_32k: fixed-regulator-wg7210_32k {
                compatible = "regulator-fixed";
@@ -522,9 +534,30 @@
        /*wp-gpios = <&gpio4 31 GPIO_ACTIVE_HIGH>;*/    /* GPIO_127 */
 };
 
-/* mmc3 is probed using pdata-quirks to pass wl1251 card data */
 &mmc3 {
-       status = "disabled";
+       vmmc-supply = <&wlan_en>;
+
+       bus-width = <4>;
+       non-removable;
+       ti,non-removable;
+       cap-power-off-card;
+
+       pinctrl-names = "default";
+       pinctrl-0 = <&mmc3_pins>;
+
+       #address-cells = <1>;
+       #size-cells = <0>;
+
+       wlan: wl1251@1 {
+               compatible = "ti,wl1251";
+
+               reg = <1>;
+
+               interrupt-parent = <&gpio1>;
+               interrupts = <21 IRQ_TYPE_LEVEL_HIGH>;  /* GPIO_21 */
+
+               ti,wl1251-has-eeprom;
+       };
 };
 
 /* bluetooth*/
--
Quoted 2 lines

2.19.1

Kind regards
Uffe

#11 lore raw

Re: [PATCH v2 04/11] mmc: host: omap_hsmmc: add code for special init of wl1251 to get rid of pandora_wl1251_init_card

In reply to Ulf Hansson

Hi Ulf,

Quoted 15 lines

Am 30.10.2019 um 16:51 schrieb Ulf Hansson <[email protected]>:

Quoted 13 lines
drivers/mmc/host/omap_hsmmc.c +13 -0
+
+               np = of_get_compatible_child(np, "ti,wl1251");
+               if (np) {
+                       /*
+                        * We have TI wl1251 attached to MMC3. Pass this information to
+                        * SDIO core because it can't be probed by normal methods.
+                        */
+
+                       dev_info(host->dev, "found wl1251\n");
+                       card->quirks |= MMC_QUIRK_NONSTD_SDIO;
+                       card->cccr.wide_bus = 1;
+                       card->cis.vendor = 0x104c;
+                       card->cis.device = 0x9066;
Quoted 8 lines

drivers/mmc/host/omap_hsmmc.c static void omap_hsmmc_init_card(struct mmc_host *mmc, struct mmc_card *card) · line 1532

Quoted 3 lines
drivers/mmc/host/omap_hsmmc.c +3 -0
+                       card->cis.blksize = 512;
+                       card->cis.max_dtr = 24000000;
+                       card->ocr = 0x80;

These things should really be figured out by the mmc core during SDIO card initialization itself, not via the host ops ->init_card() callback. That is just poor hack, which in the long run should go away.

Yes, I agree.

But I am just the poor guy who is trying to fix really broken code with as low effort as possible.

I don't even have a significant clue what this code is exactly doing and what the magic values mean. They were setup by pandora_wl1251_init_card() in the same way so that I have just moved the code here and make it called in (almost) the same situation.

Quoted 3 lines

Moreover, I think we should add a subnode to the host node in the DT, to describe the embedded SDIO card, rather than parsing the subnode for the SDIO func - as that seems wrong to me.

You mean a second subnode?

The wl1251 is the child node of the mmc node and describes the SDIO card. We just check if it is a wl1251 or e.g. wl1837 or something else or even no child.

Quoted 7 lines

To add a subnode for the SDIO card, we already have a binding that I think we should extend. Please have a look at Documentation/devicetree/bindings/mmc/mmc-card.txt.

If you want an example of how to implement this for your case, do a git grep "broken-hpi" in the driver/mmc/core/, I think it will tell you more of what I have in mind.

So while I agree that it should be improved in the long run, we should IMHO fix the hack first (broken since v4.9!), even if it remains a hack for now. Improving this part seems to be quite independent and focussed on the mmc subsystem, while the other patches involve other subsystems.

Maybe should we make a REVISIT note in the code? Or add something to the commit message about the idea how it should be done right?

Quoted 6 lines
Quoted 6 lines
drivers/mmc/host/omap_hsmmc.c +3 -0
+                       of_node_put(np);
+               }
+       }

}

Quoted 7 lines
Quoted 4 lines

static void omap_hsmmc_enable_sdio_irq(struct mmc_host *mmc, int enable) --
2.19.1

Kind regards
Uffe

BR and thanks,
Nikolaus

#7 lore raw

Re: [PATCH v2 03/11] DTS: ARM: pandora-common: define wl1251 as child node of mmc3

In reply to Ulf Hansson
Quoted 38 lines

Am 30.10.2019 um 17:44 schrieb Ulf Hansson <[email protected]>:

Quoted 22 lines · H. Nikolaus Schaller

Since v4.7 the dma initialization requires that there is a device tree property for "rx" and "tx" channels which is not provided by the pdata-quirks initialization.

By conversion of the mmc3 setup to device tree this will finally allows to remove the OpenPandora wlan specific omap3 data-quirks.

Fixes 81eef6ca9201 ("mmc: omap_hsmmc: Use dma_request_chan() for requesting DMA channel")

---
arch/arm/boot/dts/omap3-pandora-common.dtsi | 37 +++++++++++++++++++--

1 file changed, 35 insertions(+), 2 deletions(-)
arch/arm/boot/dts/omap3-pandora-common.dtsi +0 -0
index ec5891718ae6..c595b3eb314d 100644

arch/arm/boot/dts/omap3-pandora-common.dtsi line 234

Quoted 9 lines
arch/arm/boot/dts/omap3-pandora-common.dtsi +6 -0
@@ -226,6 +226,18 @@
               gpio = <&gpio6 4 GPIO_ACTIVE_HIGH>;     /* GPIO_164 */
       };
 
+       /* wl1251 wifi+bt module */
+       wlan_en: fixed-regulator-wg7210_en {
+               compatible = "regulator-fixed";
+               regulator-name = "vwlan";
+               regulator-min-microvolt = <1800000>;
+               regulator-max-microvolt = <1800000>;

I doubt these are correct.

Quoted 1 line

I guess this should be in the range of 2.7V-3.6V.

Well, it is a gpio which enables some LDO inside the wifi chip. We do not really know the voltage it produces and it does not matter. The gpio voltage is 1.8V.

Basically we use a fixed-regulator to "translate" a regulator into a control gpio because the mmc interface wants to see a vmmc-supply.

Quoted 5 lines

arch/arm/boot/dts/omap3-pandora-common.dtsi line 236

Quoted 3 lines
arch/arm/boot/dts/omap3-pandora-common.dtsi +2 -0
+               startup-delay-us = <50000>;
+               regulator-always-on;

Always on?

Oops. Yes, that is something to check!

Quoted 40 lines
Quoted 40 lines
arch/arm/boot/dts/omap3-pandora-common.dtsi +27 -2
+               enable-active-high;
+               gpio = <&gpio1 23 GPIO_ACTIVE_HIGH>;
+       };
+
       /* wg7210 (wifi+bt module) 32k clock buffer */
       wg7210_32k: fixed-regulator-wg7210_32k {
               compatible = "regulator-fixed";
@@ -522,9 +534,30 @@
       /*wp-gpios = <&gpio4 31 GPIO_ACTIVE_HIGH>;*/    /* GPIO_127 */
 };
 
-/* mmc3 is probed using pdata-quirks to pass wl1251 card data */
 &mmc3 {
-       status = "disabled";
+       vmmc-supply = <&wlan_en>;
+
+       bus-width = <4>;
+       non-removable;
+       ti,non-removable;
+       cap-power-off-card;
+
+       pinctrl-names = "default";
+       pinctrl-0 = <&mmc3_pins>;
+
+       #address-cells = <1>;
+       #size-cells = <0>;
+
+       wlan: wl1251@1 {
+               compatible = "ti,wl1251";
+
+               reg = <1>;
+
+               interrupt-parent = <&gpio1>;
+               interrupts = <21 IRQ_TYPE_LEVEL_HIGH>;  /* GPIO_21 */
+
+               ti,wl1251-has-eeprom;
+       };
 };
 
Quoted 4 lines
Quoted 4 lines
arch/arm/boot/dts/omap3-pandora-common.dtsi +0 -1
 /* bluetooth*/
--

2.19.1

BR and thanks,
Nikolaus

#12 lore raw

Re: [PATCH v2 04/11] mmc: host: omap_hsmmc: add code for special init of wl1251 to get rid of pandora_wl1251_init_card

In reply to H. Nikolaus Schaller
Quoted 27 lines · H. Nikolaus Schaller

Hi Ulf,

Quoted 15 lines

Am 30.10.2019 um 16:51 schrieb Ulf Hansson <[email protected]>:

Quoted 13 lines
drivers/mmc/host/omap_hsmmc.c +13 -0
+
+               np = of_get_compatible_child(np, "ti,wl1251");
+               if (np) {
+                       /*
+                        * We have TI wl1251 attached to MMC3. Pass this information to
+                        * SDIO core because it can't be probed by normal methods.
+                        */
+
+                       dev_info(host->dev, "found wl1251\n");
+                       card->quirks |= MMC_QUIRK_NONSTD_SDIO;
+                       card->cccr.wide_bus = 1;
+                       card->cis.vendor = 0x104c;
+                       card->cis.device = 0x9066;
Quoted 8 lines

drivers/mmc/host/omap_hsmmc.c static void omap_hsmmc_init_card(struct mmc_host *mmc, struct mmc_card *card) · line 1532

Quoted 3 lines
drivers/mmc/host/omap_hsmmc.c +3 -0
+                       card->cis.blksize = 512;
+                       card->cis.max_dtr = 24000000;
+                       card->ocr = 0x80;

These things should really be figured out by the mmc core during SDIO card initialization itself, not via the host ops ->init_card() callback. That is just poor hack, which in the long run should go away.

Quoted 4 lines

Yes, I agree.

But I am just the poor guy who is trying to fix really broken code with as low effort as possible.

I see. Thanks for looking at this mess!

In general, as long as we improve the code, I am happy to move forward.

However, my main concern at this point, is to make sure we get the DT bindings and the DTS files updated correctly. We don't want to come back to this again.

Quoted 5 lines

I don't even have a significant clue what this code is exactly doing and what the magic values mean. They were setup by pandora_wl1251_init_card() in the same way so that I have just moved the code here and make it called in (almost) the same situation.

Okay!

Quoted 10 lines
Quoted 4 lines

Moreover, I think we should add a subnode to the host node in the DT, to describe the embedded SDIO card, rather than parsing the subnode for the SDIO func - as that seems wrong to me.

You mean a second subnode?

The wl1251 is the child node of the mmc node and describes the SDIO card. We just check if it is a wl1251 or e.g. wl1837 or something else or even no child.

The reason why I brought this up, was because there are sometimes cases where an SDIO card is shared between more than one SDIO func. WiFi+Bluetooth for example, but if I am correct, that is not the case for wl1251?

That said, I am happy to continue with your approach.

Quoted 5 lines
Quoted 5 lines

To add a subnode for the SDIO card, we already have a binding that I think we should extend. Please have a look at Documentation/devicetree/bindings/mmc/mmc-card.txt.

Quoted 8 lines
Quoted 3 lines

If you want an example of how to implement this for your case, do a git grep "broken-hpi" in the driver/mmc/core/, I think it will tell you more of what I have in mind.

So while I agree that it should be improved in the long run, we should IMHO fix the hack first (broken since v4.9!), even if it remains a hack for now. Improving this part seems to be quite independent and focussed on the mmc subsystem, while the other patches involve other subsystems.

I agree.

Quoted 3 lines

Maybe should we make a REVISIT note in the code? Or add something to the commit message about the idea how it should be done right?

Just add a note that we should move this DT parsing of the subnode to the mmc core, but that we are leaving that as a future improvement. That's good enough. Then I can have a look as a second step, and when I get some time, to move this to the mmc core.

However, there is one thing I would like you to add to the series. That is:

In the struct omap_hsmmc_platform_data, there is an ->init_card() callback. Beyond the changes of this series, there is no longer any users of that, unless I am mistaken. Going forward, let's make sure it doesn't get used again, so can you please remove it!?

[...]

Kind regards
Uffe

#8 lore raw

Re: [PATCH v2 03/11] DTS: ARM: pandora-common: define wl1251 as child node of mmc3

In reply to H. Nikolaus Schaller
Quoted 41 lines · H. Nikolaus Schaller
Quoted 39 lines

Am 30.10.2019 um 17:44 schrieb Ulf Hansson <[email protected]>:

Quoted 22 lines · H. Nikolaus Schaller

Since v4.7 the dma initialization requires that there is a device tree property for "rx" and "tx" channels which is not provided by the pdata-quirks initialization.

By conversion of the mmc3 setup to device tree this will finally allows to remove the OpenPandora wlan specific omap3 data-quirks.

Fixes 81eef6ca9201 ("mmc: omap_hsmmc: Use dma_request_chan() for requesting DMA channel")

---
arch/arm/boot/dts/omap3-pandora-common.dtsi | 37 +++++++++++++++++++--

1 file changed, 35 insertions(+), 2 deletions(-)
arch/arm/boot/dts/omap3-pandora-common.dtsi +0 -0
index ec5891718ae6..c595b3eb314d 100644

arch/arm/boot/dts/omap3-pandora-common.dtsi line 234

Quoted 9 lines
arch/arm/boot/dts/omap3-pandora-common.dtsi +6 -0
@@ -226,6 +226,18 @@
               gpio = <&gpio6 4 GPIO_ACTIVE_HIGH>;     /* GPIO_164 */
       };
 
+       /* wl1251 wifi+bt module */
+       wlan_en: fixed-regulator-wg7210_en {
+               compatible = "regulator-fixed";
+               regulator-name = "vwlan";
+               regulator-min-microvolt = <1800000>;
+               regulator-max-microvolt = <1800000>;

I doubt these are correct.

Quoted 1 line

I guess this should be in the range of 2.7V-3.6V.

Quoted 7 lines

Well, it is a gpio which enables some LDO inside the wifi chip. We do not really know the voltage it produces and it does not matter. The gpio voltage is 1.8V.

Basically we use a fixed-regulator to "translate" a regulator into a control gpio because the mmc interface wants to see a vmmc-supply.

The vmmc supply represent the core power to the SDIO card (or SD/(e)MMC). Depending on what voltage range the vmmc supply supports, the so called OCR mask is created by the mmc core. The mask is then used to let the core negotiate the voltage level with the SDIO card, during the card initialization. This is not to confuse with the I/O voltage level, which is a different regulator.

Anyway, according to the TI WiLink series specifications, it looks like vmmc should be a regulator supporting 3-3.3V (in many schematics it's called VBAT).

Furthermore I decided to dig into various DTS files that specifies the vmmc regulator, of course for mmc nodes having a subnode specifying an SDIO card for a TI WiLink. In most cases a 1.8V fixed GPIO regulator is used. This looks wrong to me. The fixed GPIO regulator isn't really the one that should model vmmc.

The proper solution, would rather be to use separate regulator for vmmc and instead use a so called mmc-pwrseq node to manage the GPIO.

To conclude from my side, as we have lots of DTS that are wrong, I don't really care if we add another one in the way you suggest above. But feel free to look into the mmc-pwrseq option.

Quoted 8 lines
Quoted 6 lines

arch/arm/boot/dts/omap3-pandora-common.dtsi line 236

Quoted 4 lines
arch/arm/boot/dts/omap3-pandora-common.dtsi +2 -0
+               startup-delay-us = <50000>;
+               regulator-always-on;

Always on?

Oops. Yes, that is something to check!

As it's a GPIO regulator, for sure it's not always on.

Quoted 41 lines
Quoted 41 lines
Quoted 41 lines
arch/arm/boot/dts/omap3-pandora-common.dtsi +27 -2
+               enable-active-high;
+               gpio = <&gpio1 23 GPIO_ACTIVE_HIGH>;
+       };
+
       /* wg7210 (wifi+bt module) 32k clock buffer */
       wg7210_32k: fixed-regulator-wg7210_32k {
               compatible = "regulator-fixed";
@@ -522,9 +534,30 @@
       /*wp-gpios = <&gpio4 31 GPIO_ACTIVE_HIGH>;*/    /* GPIO_127 */
 };
 
-/* mmc3 is probed using pdata-quirks to pass wl1251 card data */
 &mmc3 {
-       status = "disabled";
+       vmmc-supply = <&wlan_en>;
+
+       bus-width = <4>;
+       non-removable;
+       ti,non-removable;
+       cap-power-off-card;
+
+       pinctrl-names = "default";
+       pinctrl-0 = <&mmc3_pins>;
+
+       #address-cells = <1>;
+       #size-cells = <0>;
+
+       wlan: wl1251@1 {
+               compatible = "ti,wl1251";
+
+               reg = <1>;
+
+               interrupt-parent = <&gpio1>;
+               interrupts = <21 IRQ_TYPE_LEVEL_HIGH>;  /* GPIO_21 */
+
+               ti,wl1251-has-eeprom;
+       };
 };
 
Quoted 8 lines
Quoted 4 lines
Quoted 4 lines
arch/arm/boot/dts/omap3-pandora-common.dtsi +0 -1
 /* bluetooth*/
--

2.19.1

BR and thanks,
Nikolaus

Kind regards
Uffe