Skip to contentSkip to conversation

TQMx86: TQMx110EB and TQMxE40x MFD/GPIO support

PATCH Applied Matthias Schiffer 5 participants 17 messages Jul 16, 2021 – Aug 16, 2021

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

v3 cover TQMx86: TQMx110EB and TQMxE40x MFD/GPIO support

v3:
- Removed Fixes tag from patch 3 again

v2:
- A number of new patches (more hardware support and a few fixes)
- Patches 1-3 have gained Fixes tags
- Patch 2 depends on 1, so maybe we can push the GPIO patch through the MFD tree to keep them together?
- The change in patch 7 was somewhat controversial. I've added a warning, but it is the last patch of the series, so it doesn't affect the rest of the series if it is rejected.

Matthias Schiffer (7):
gpio: tqmx86: really make IRQ optional
mfd: tqmx86: clear GPIO IRQ resource when no IRQ is set
mfd: tqmx86: remove incorrect TQMx90UC board ID
mfd: tqmx86: fix typo in "platform"
mfd: tqmx86: add support for TQMx110EB and TQMxE40x
mfd: tqmx86: add support for TQ-Systems DMI IDs
mfd: tqmx86: assume 24MHz LPC clock for unknown boards

 drivers/gpio/gpio-tqmx86.c |  6 ++---
 drivers/mfd/tqmx86.c       | 48 ++++++++++++++++++++++++++++++--------
 2 files changed, 41 insertions(+), 13 deletions(-)
Signature · 1 line
2.17.1
#1 lore raw

PATCH v3 1/7 gpio: tqmx86: really make IRQ optional

The tqmx86 MFD driver was passing IRQ 0 for "no IRQ" in the past. This causes warnings with newer kernels.

Prepare the gpio-tqmx86 driver for the fixed MFD driver by handling a missing IRQ properly.

Fixes b868db94a6a7 ("gpio: tqmx86: Add GPIO from for this IO controller")

---

v2: add Fixes line
v3: no changes

 drivers/gpio/gpio-tqmx86.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/gpio/gpio-tqmx86.c +3 -3
index 5022e0ad0fae..0f5d17f343f1 100644
@@ -238,8 +238,8 @@ static int tqmx86_gpio_probe(struct platform_device *pdev)
 	struct resource *res;
 	int ret, irq;
 
-	irq = platform_get_irq(pdev, 0);
-	if (irq < 0)
+	irq = platform_get_irq_optional(pdev, 0);
+	if (irq < 0 && irq != -ENXIO)
 		return irq;
 
 	res = platform_get_resource(pdev, IORESOURCE_IO, 0);
@@ -278,7 +278,7 @@ static int tqmx86_gpio_probe(struct platform_device *pdev)
 
 	pm_runtime_enable(&pdev->dev);
 
-	if (irq) {
+	if (irq > 0) {
 		struct irq_chip *irq_chip = &gpio->irq_chip;
 		u8 irq_status;
 
Signature · 1 line
2.17.1
Acked-by on v3 1/7
#2 lore raw

Re: [PATCH v3 1/7] gpio: tqmx86: really make IRQ optional

Quoted 6 lines · Matthias Schiffer

The tqmx86 MFD driver was passing IRQ 0 for "no IRQ" in the past. This causes warnings with newer kernels.

Prepare the gpio-tqmx86 driver for the fixed MFD driver by handling a missing IRQ properly.

Quoted 7 lines
Fixes b868db94a6a7 ("gpio: tqmx86: Add GPIO from for this IO controller")

---

v2: add Fixes line
v3: no changes

I suppose Lee will merge this into MFD with the rest of the patches? I don't see anything stopping Bart from just merging this one patch into the GPIO tree though. Only runtime dependencies.

Yours,
Linus Walleij

#3 lore raw

Re: [PATCH v3 1/7] gpio: tqmx86: really make IRQ optional

Quoted 18 lines · Linus Walleij
Quoted 6 lines · Matthias Schiffer

The tqmx86 MFD driver was passing IRQ 0 for "no IRQ" in the past. This causes warnings with newer kernels.

Prepare the gpio-tqmx86 driver for the fixed MFD driver by handling a missing IRQ properly.

Quoted 7 lines
Fixes b868db94a6a7 ("gpio: tqmx86: Add GPIO from for this IO controller")

---

v2: add Fixes line
v3: no changes

Quoted 8 lines

I suppose Lee will merge this into MFD with the rest of the patches? I don't see anything stopping Bart from just merging this one patch into the GPIO tree though. Only runtime dependencies.

Yours,
Linus Walleij

I applied this patch for fixes and will shortly send a PR with some other fixes to Linus.

Bart

#4 lore raw

PATCH v3 2/7 mfd: tqmx86: clear GPIO IRQ resource when no IRQ is set

In reply to Matthias Schiffer

The driver was registering IRQ 0 when no IRQ was set. This leads to warnings with newer kernels.

Clear the resource flags, so no resource is registered at all in this case.

Fixes 2f17dd34ffed ("mfd: tqmx86: IO controller with I2C, Wachdog and GPIO")

---

v2: add Fixes line
v3: no changes

 drivers/mfd/tqmx86.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/mfd/tqmx86.c +2 -0
index ddddf08b6a4c..732013f40e4e 100644
@@ -209,6 +209,8 @@ static int tqmx86_probe(struct platform_device *pdev)
 
 		/* Assumes the IRQ resource is first. */
 		tqmx_gpio_resources[0].start = gpio_irq;
+	} else {
+		tqmx_gpio_resources[0].flags = 0;
 	}
 
 	ocores_platfom_data.clock_khz = tqmx86_board_id_to_clk_rate(board_id);
Signature · 1 line
2.17.1
#5 lore raw

Re: [PATCH v3 2/7] mfd: tqmx86: clear GPIO IRQ resource when no IRQ is set

Quoted 22 lines · Matthias Schiffer

The driver was registering IRQ 0 when no IRQ was set. This leads to warnings with newer kernels.

Clear the resource flags, so no resource is registered at all in this case.

Fixes 2f17dd34ffed ("mfd: tqmx86: IO controller with I2C, Wachdog and GPIO")

---

v2: add Fixes line
v3: no changes

 drivers/mfd/tqmx86.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/mfd/tqmx86.c +0 -0
index ddddf08b6a4c..732013f40e4e 100644

drivers/mfd/tqmx86.c static int tqmx86_probe(struct platform_device *pdev) · line 213

Quoted 5 lines
drivers/mfd/tqmx86.c +2 -0
@@ -209,6 +209,8 @@ static int tqmx86_probe(struct platform_device *pdev)
 
 		/* Assumes the IRQ resource is first. */
 		tqmx_gpio_resources[0].start = gpio_irq;
+	} else {
+		tqmx_gpio_resources[0].flags = 0;

Strange - why is this !0 in the first place?

Quoted 3 lines
drivers/mfd/tqmx86.c +0 -0
 	}
 
 	ocores_platfom_data.clock_khz = tqmx86_board_id_to_clk_rate(board_id);
Signature · 4 lines
Lee Jones [李琼斯]
Senior Technical Lead - Developer Services
Linaro.org │ Open source software for Arm SoCs
Follow Linaro: Facebook | Twitter | Blog
#6 lore raw

Re: [PATCH v3 2/7] mfd: tqmx86: clear GPIO IRQ resource when no IRQ is set

Quoted 25 lines · Lee Jones
Quoted 23 lines · Matthias Schiffer

The driver was registering IRQ 0 when no IRQ was set. This leads to warnings with newer kernels.

Clear the resource flags, so no resource is registered at all in this case.

Fixes 2f17dd34ffed ("mfd: tqmx86: IO controller with I2C, Wachdog and GPIO")

---

v2: add Fixes line
v3: no changes

 drivers/mfd/tqmx86.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/mfd/tqmx86.c +0 -0
index ddddf08b6a4c..732013f40e4e 100644
@@ -209,6 +209,8 @@ static int tqmx86_probe(struct platform_device *pdev)
 
Quoted 6 lines

drivers/mfd/tqmx86.c static int tqmx86_probe(struct platform_device *pdev) · line 213

Quoted 4 lines
drivers/mfd/tqmx86.c +2 -0
 		/* Assumes the IRQ resource is first. */
 		tqmx_gpio_resources[0].start = gpio_irq;
+	} else {
+		tqmx_gpio_resources[0].flags = 0;

Strange - why is this !0 in the first place?

I don't see anything strange here. DEFINE_RES_IRQ() sets flags to IORESOURCE_IRQ. We reset it to 0 when there is no IRQ to clear that resource entry.

An alternative would be to start with an empty entry and only fill in the fields when an IRQ is used, but that seems more cumbersome than the current code to me.

Quoted 6 lines
Quoted 4 lines
drivers/mfd/tqmx86.c +0 -0
 	}
 
 	ocores_platfom_data.clock_khz = tqmx86_board_id_to_clk_rate(board_id);
#7 lore raw

Re: [PATCH v3 2/7] mfd: tqmx86: clear GPIO IRQ resource when no IRQ is set

Quoted 2 lines · Matthias Schiffer
Quoted 1 line · Lee Jones

On Fri, 16 Jul 2021, Matthias Schiffer wrote:

...

Quoted 9 lines
Quoted 1 line

Strange - why is this !0 in the first place?

I don't see anything strange here. DEFINE_RES_IRQ() sets flags to IORESOURCE_IRQ. We reset it to 0 when there is no IRQ to clear that resource entry.

An alternative would be to start with an empty entry and only fill in the fields when an IRQ is used, but that seems more cumbersome than the current code to me.

Another alternative is to start using the IRQ DISABLED resource flag, but I'm afraid that OF code is not ready for that.
https://elixir.bootlin.com/linux/latest/source/include/linux/ioport.h#L331

Signature · 2 lines
With Best Regards,
Andy Shevchenko
#8 lore raw

Re: [PATCH v3 2/7] mfd: tqmx86: clear GPIO IRQ resource when no IRQ is set

Quoted 13 lines · Andy Shevchenko
Quoted 2 lines · Matthias Schiffer
Quoted 1 line · Lee Jones

On Fri, 16 Jul 2021, Matthias Schiffer wrote:

...

Quoted 6 lines
Quoted 1 line

Strange - why is this !0 in the first place?

I don't see anything strange here. DEFINE_RES_IRQ() sets flags to IORESOURCE_IRQ. We reset it to 0 when there is no IRQ to clear that resource entry.

Quoted 8 lines
Quoted 3 lines

An alternative would be to start with an empty entry and only fill in the fields when an IRQ is used, but that seems more cumbersome than the current code to me.

Another alternative is to start using the IRQ DISABLED resource flag, but I'm afraid that OF code is not ready for that.
https://elixir.bootlin.com/linux/latest/source/include/linux/ioport.h#L331

As this patch is a fairly simple bugfix, I'd prefer to get it (or a similar fix) applied without having to wait for improvements of the core code - also for the sake of stable backports.

#9 lore raw

PATCH v3 3/7 mfd: tqmx86: remove incorrect TQMx90UC board ID

In reply to Matthias Schiffer

No TQMx90UC exists at the moment, and it is undecided whether ID 10 will be used eventually (and if it is, how that SoM will be named).

---

v2: new patch
v3: remove Fixes line

 drivers/mfd/tqmx86.c | 4 ----
 1 file changed, 4 deletions(-)
drivers/mfd/tqmx86.c +0 -4
index 732013f40e4e..9b65dbedc1bb 100644
@@ -35,7 +35,6 @@
 #define TQMX86_REG_BOARD_ID_E39x	7
 #define TQMX86_REG_BOARD_ID_70EB	8
 #define TQMX86_REG_BOARD_ID_80UC	9
-#define TQMX86_REG_BOARD_ID_90UC	10
 #define TQMX86_REG_BOARD_REV	0x21
 #define TQMX86_REG_IO_EXT_INT	0x26
 #define TQMX86_REG_IO_EXT_INT_NONE		0
@@ -128,8 +127,6 @@ static const char *tqmx86_board_id_to_name(u8 board_id)
 		return "TQMx70EB";
 	case TQMX86_REG_BOARD_ID_80UC:
 		return "TQMx80UC";
-	case TQMX86_REG_BOARD_ID_90UC:
-		return "TQMx90UC";
 	default:
 		return "Unknown";
 	}
@@ -142,7 +139,6 @@ static int tqmx86_board_id_to_clk_rate(u8 board_id)
 	case TQMX86_REG_BOARD_ID_60EB:
 	case TQMX86_REG_BOARD_ID_70EB:
 	case TQMX86_REG_BOARD_ID_80UC:
-	case TQMX86_REG_BOARD_ID_90UC:
 		return 24000;
 	case TQMX86_REG_BOARD_ID_E39M:
 	case TQMX86_REG_BOARD_ID_E39C:
Signature · 1 line
2.17.1
#10 lore raw

Re: [PATCH v3 3/7] mfd: tqmx86: remove incorrect TQMx90UC board ID

Quoted 5 lines · Matthias Schiffer

No TQMx90UC exists at the moment, and it is undecided whether ID 10 will be used eventually (and if it is, how that SoM will be named).

Quoted 7 lines

---

v2: new patch
v3: remove Fixes line

 drivers/mfd/tqmx86.c | 4 ----
 1 file changed, 4 deletions(-)

For my own reference (apply this as-is to your sign-off block):

Signature · 4 lines
Lee Jones [李琼斯]
Senior Technical Lead - Developer Services
Linaro.org │ Open source software for Arm SoCs
Follow Linaro: Facebook | Twitter | Blog
#11 lore raw

PATCH v3 4/7 mfd: tqmx86: fix typo in "platform"

In reply to Matthias Schiffer

Rename variable from "ocores_platfom_data" to "ocores_platform_data".

---

v2: new patch
v3: new patch

 drivers/mfd/tqmx86.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/mfd/tqmx86.c +4 -4
index 9b65dbedc1bb..ff1bdb742e3f 100644
@@ -76,7 +76,7 @@ static struct i2c_board_info tqmx86_i2c_devices[] = {
 	},
 };
 
-static struct ocores_i2c_platform_data ocores_platfom_data = {
+static struct ocores_i2c_platform_data ocores_platform_data = {
 	.num_devices = ARRAY_SIZE(tqmx86_i2c_devices),
 	.devices = tqmx86_i2c_devices,
 };
@@ -84,8 +84,8 @@ static struct ocores_i2c_platform_data ocores_platfom_data = {
 static const struct mfd_cell tqmx86_i2c_soft_dev[] = {
 	{
 		.name = "ocores-i2c",
-		.platform_data = &ocores_platfom_data,
-		.pdata_size = sizeof(ocores_platfom_data),
+		.platform_data = &ocores_platform_data,
+		.pdata_size = sizeof(ocores_platform_data),
 		.resources = tqmx_i2c_soft_resources,
 		.num_resources = ARRAY_SIZE(tqmx_i2c_soft_resources),
 	},
@@ -209,7 +209,7 @@ static int tqmx86_probe(struct platform_device *pdev)
 		tqmx_gpio_resources[0].flags = 0;
 	}
 
-	ocores_platfom_data.clock_khz = tqmx86_board_id_to_clk_rate(board_id);
+	ocores_platform_data.clock_khz = tqmx86_board_id_to_clk_rate(board_id);
 
 	if (i2c_det == TQMX86_REG_I2C_DETECT_SOFT) {
 		err = devm_mfd_add_devices(dev, PLATFORM_DEVID_NONE,
Signature · 1 line
2.17.1
#12 lore raw

Re: [PATCH v3 4/7] mfd: tqmx86: fix typo in "platform"

Quoted 8 lines · Matthias Schiffer

Rename variable from "ocores_platfom_data" to "ocores_platform_data".

---

v2: new patch
v3: new patch

Heh?

Quoted 2 lines
 drivers/mfd/tqmx86.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

For my own reference (apply this as-is to your sign-off block):

Signature · 4 lines
Lee Jones [李琼斯]
Senior Technical Lead - Developer Services
Linaro.org │ Open source software for Arm SoCs
Follow Linaro: Facebook | Twitter | Blog
#13 lore raw

PATCH v3 5/7 mfd: tqmx86: add support for TQMx110EB and TQMxE40x

In reply to Matthias Schiffer

Add the board IDs for the TQMx110EB and the TQMxE40x family. All use a 24MHz LPC clock.

---

v2:
- add more new modules
- explicitly list each module for LPC clock

v3: no changes

 drivers/mfd/tqmx86.c | 20 ++++++++++++++++++++
 1 file changed, 20 insertions(+)
drivers/mfd/tqmx86.c +20 -0
index ff1bdb742e3f..9eb05b3ef573 100644
@@ -35,6 +35,11 @@
 #define TQMX86_REG_BOARD_ID_E39x	7
 #define TQMX86_REG_BOARD_ID_70EB	8
 #define TQMX86_REG_BOARD_ID_80UC	9
+#define TQMX86_REG_BOARD_ID_110EB	11
+#define TQMX86_REG_BOARD_ID_E40M	12
+#define TQMX86_REG_BOARD_ID_E40S	13
+#define TQMX86_REG_BOARD_ID_E40C1	14
+#define TQMX86_REG_BOARD_ID_E40C2	15
 #define TQMX86_REG_BOARD_REV	0x21
 #define TQMX86_REG_IO_EXT_INT	0x26
 #define TQMX86_REG_IO_EXT_INT_NONE		0
@@ -127,6 +132,16 @@ static const char *tqmx86_board_id_to_name(u8 board_id)
 		return "TQMx70EB";
 	case TQMX86_REG_BOARD_ID_80UC:
 		return "TQMx80UC";
+	case TQMX86_REG_BOARD_ID_110EB:
+		return "TQMx110EB";
+	case TQMX86_REG_BOARD_ID_E40M:
+		return "TQMxE40M";
+	case TQMX86_REG_BOARD_ID_E40S:
+		return "TQMxE40S";
+	case TQMX86_REG_BOARD_ID_E40C1:
+		return "TQMxE40C1";
+	case TQMX86_REG_BOARD_ID_E40C2:
+		return "TQMxE40C2";
 	default:
 		return "Unknown";
 	}
@@ -139,6 +154,11 @@ static int tqmx86_board_id_to_clk_rate(u8 board_id)
 	case TQMX86_REG_BOARD_ID_60EB:
 	case TQMX86_REG_BOARD_ID_70EB:
 	case TQMX86_REG_BOARD_ID_80UC:
+	case TQMX86_REG_BOARD_ID_110EB:
+	case TQMX86_REG_BOARD_ID_E40M:
+	case TQMX86_REG_BOARD_ID_E40S:
+	case TQMX86_REG_BOARD_ID_E40C1:
+	case TQMX86_REG_BOARD_ID_E40C2:
 		return 24000;
 	case TQMX86_REG_BOARD_ID_E39M:
 	case TQMX86_REG_BOARD_ID_E39C:
Signature · 1 line
2.17.1
#14 lore raw

PATCH v3 6/7 mfd: tqmx86: add support for TQ-Systems DMI IDs

In reply to Matthias Schiffer

Newer TQMx86 modules use TQ-Systems instead of TQ-Group as their vendor ID.

---

v2: new patch
v3: no changes

 drivers/mfd/tqmx86.c | 8 ++++++++
 1 file changed, 8 insertions(+)
drivers/mfd/tqmx86.c +8 -0
index 9eb05b3ef573..58f35c8b5a45 100644
@@ -271,6 +271,14 @@ static const struct dmi_system_id tqmx86_dmi_table[] __initconst = {
 		},
 		.callback = tqmx86_create_platform_device,
 	},
+	{
+		.ident = "TQMX86",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "TQ-Systems"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "TQMx"),
+		},
+		.callback = tqmx86_create_platform_device,
+	},
 	{}
 };
 MODULE_DEVICE_TABLE(dmi, tqmx86_dmi_table);
Signature · 1 line
2.17.1
#15 lore raw

PATCH v3 7/7 mfd: tqmx86: assume 24MHz LPC clock for unknown boards

In reply to Matthias Schiffer

All future TQMx86 modules should use a 24MHz LPC clock. Warn about unknown boards, but assume this is the case.

---

v2:
- Separated from TQMxE40M support patch
- Add warning for unknown boards

v3: no changes

 drivers/mfd/tqmx86.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)
drivers/mfd/tqmx86.c +5 -3
index 58f35c8b5a45..7ae906ff8e35 100644
@@ -147,7 +147,7 @@ static const char *tqmx86_board_id_to_name(u8 board_id)
 	}
 }
 
-static int tqmx86_board_id_to_clk_rate(u8 board_id)
+static int tqmx86_board_id_to_clk_rate(struct device *dev, u8 board_id)
 {
 	switch (board_id) {
 	case TQMX86_REG_BOARD_ID_50UC:
@@ -168,7 +168,9 @@ static int tqmx86_board_id_to_clk_rate(u8 board_id)
 	case TQMX86_REG_BOARD_ID_E38C:
 		return 33000;
 	default:
-		return 0;
+		dev_warn(dev, "unknown board %d, assuming 24MHz LPC clock\n",
+			 board_id);
+		return 24000;
 	}
 }
 
@@ -229,7 +231,7 @@ static int tqmx86_probe(struct platform_device *pdev)
 		tqmx_gpio_resources[0].flags = 0;
 	}
 
-	ocores_platform_data.clock_khz = tqmx86_board_id_to_clk_rate(board_id);
+	ocores_platform_data.clock_khz = tqmx86_board_id_to_clk_rate(dev, board_id);
 
 	if (i2c_det == TQMX86_REG_I2C_DETECT_SOFT) {
 		err = devm_mfd_add_devices(dev, PLATFORM_DEVID_NONE,
Signature · 1 line
2.17.1
Applied
#16 lore raw
In reply to Matthias Schiffer
Quoted 15 lines · Matthias Schiffer

v3:
- Removed Fixes tag from patch 3 again

v2:
- A number of new patches (more hardware support and a few fixes)
- Patches 1-3 have gained Fixes tags
- Patch 2 depends on 1, so maybe we can push the GPIO patch through the MFD tree to keep them together?
- The change in patch 7 was somewhat controversial. I've added a warning, but it is the last patch of the series, so it doesn't affect the rest of the series if it is rejected.

Matthias Schiffer (7):
gpio: tqmx86: really make IRQ optional

Quoted 6 lines

mfd: tqmx86: clear GPIO IRQ resource when no IRQ is set
mfd: tqmx86: remove incorrect TQMx90UC board ID
mfd: tqmx86: fix typo in "platform"
mfd: tqmx86: add support for TQMx110EB and TQMxE40x
mfd: tqmx86: add support for TQ-Systems DMI IDs
mfd: tqmx86: assume 24MHz LPC clock for unknown boards

Applied, thanks.

Signature · 4 lines
Lee Jones [李琼斯]
Senior Technical Lead - Developer Services
Linaro.org │ Open source software for Arm SoCs
Follow Linaro: Facebook | Twitter | Blog