[PATCH 0/4] ipmi: bt-bmc: Add configurable LPC host interface

[PATCH 0/4] ipmi: bt-bmc: Add configurable LPC host interface

[Yu-Che Hsieh via B4 Relay]

The Aspeed BT BMC driver currently programs a fixed LPC IO address and

SerIRQ value for the host-facing BT interface. That matches the original

single-interface setup, but newer systems may need the host interface

parameters to be described by firmware.

The Aspeed KCS BMC and VUART bindings already use aspeed,lpc-io-reg and

aspeed,lpc-interrupts for this purpose. Reuse the same properties for

the BT BMC binding and teach the driver to consume them while preserving

the existing default LPC IO address and level-low SerIRQ configuration

when the properties are absent.

The first two patches are small preparation patches. The register

definitions are converted to bitfield helpers so BT_CR0 fields can be

programmed by name, and the open state is moved from a global variable

to the device instance so multiple BT devices are not blocked by a

single shared open count.

---
Yu-Che Hsieh (4):
      ipmi: bt-bmc: Use bitfield helpers for register definitions
      ipmi: bt-bmc: Track open state per device
      dt-bindings: ipmi: Add optional LPC properties to ASPEED BT devices
      ipmi: bt-bmc: Read LPC address and SerIRQ from device tree

 .../bindings/ipmi/aspeed,ast2400-ibt-bmc.yaml      |  21 ++++
 drivers/char/ipmi/bt-bmc.c                         | 118 ++++++++++++++-------
 2 files changed, 99 insertions(+), 40 deletions(-)
---
base-commit: 493181e2f2f1bdfd4f09a988008653ae73b30688
change-id: 20260625-aspeed-bt-bmc-multichannel-b44a10ff407c

Best regards,
-- 
Yu-Che Hsieh <yc_hsieh@aspeedtech.com>

[PATCH 1/4] ipmi: bt-bmc: Use bitfield helpers for register definitions

[Yu-Che Hsieh via B4 Relay]

From: Yu-Che Hsieh <yc_hsieh@aspeedtech.com>

Use BIT(), GENMASK(), and FIELD_PREP() for the BT register definitions

and register field programming.

This makes the register layout easier to read and prepares the driver

for later changes that need to program the BT_CR0 fields from device

configuration.

Signed-off-by: Yu-Che Hsieh <yc_hsieh@aspeedtech.com>
---
 drivers/char/ipmi/bt-bmc.c | 72 ++++++++++++++++++++++++----------------------
 1 file changed, 37 insertions(+), 35 deletions(-)

diff --git a/drivers/char/ipmi/bt-bmc.c b/drivers/char/ipmi/bt-bmc.c
index a179d4797011..f3c67272502f 100644
--- a/drivers/char/ipmi/bt-bmc.c
+++ b/drivers/char/ipmi/bt-bmc.c
@@ -15,6 +15,7 @@
 #include <linux/poll.h>
 #include <linux/sched.h>
 #include <linux/timer.h>
+#include <linux/bitfield.h>
 
 /*
  * This is a BMC device used to communicate to the host
@@ -24,33 +25,34 @@
 #define BT_IO_BASE	0xe4
 #define BT_IRQ		10
 
-#define BT_CR0		0x0
-#define   BT_CR0_IO_BASE		16
-#define   BT_CR0_IRQ			12
-#define   BT_CR0_EN_CLR_SLV_RDP		0x8
-#define   BT_CR0_EN_CLR_SLV_WRP		0x4
-#define   BT_CR0_ENABLE_IBT		0x1
-#define BT_CR1		0x4
-#define   BT_CR1_IRQ_H2B	0x01
-#define   BT_CR1_IRQ_HBUSY	0x40
-#define BT_CR2		0x8
-#define   BT_CR2_IRQ_H2B	0x01
-#define   BT_CR2_IRQ_HBUSY	0x40
-#define BT_CR3		0xc
-#define BT_CTRL		0x10
-#define   BT_CTRL_B_BUSY		0x80
-#define   BT_CTRL_H_BUSY		0x40
-#define   BT_CTRL_OEM0			0x20
-#define   BT_CTRL_SMS_ATN		0x10
-#define   BT_CTRL_B2H_ATN		0x08
-#define   BT_CTRL_H2B_ATN		0x04
-#define   BT_CTRL_CLR_RD_PTR		0x02
-#define   BT_CTRL_CLR_WR_PTR		0x01
-#define BT_BMC2HOST	0x14
-#define BT_INTMASK	0x18
-#define   BT_INTMASK_B2H_IRQEN		0x01
-#define   BT_INTMASK_B2H_IRQ		0x02
-#define   BT_INTMASK_BMC_HWRST		0x80
+#define BT_CR0			0x0
+#define   BT_CR0_IO_BASE	GENMASK(31, 16)
+#define   BT_CR0_SIRQ		GENMASK(15, 12)
+#define   BT_CR0_SIRQ_TYPE	GENMASK(11, 10)
+#define   BT_CR0_EN_CLR_SLV_RDP	BIT(3)
+#define   BT_CR0_EN_CLR_SLV_WRP	BIT(2)
+#define   BT_CR0_ENABLE_IBT	BIT(0)
+#define BT_CR1			0x4
+#define   BT_CR1_IRQ_EN_HBUSY	BIT(6)
+#define   BT_CR1_IRQ_EN_H2B	BIT(0)
+#define BT_CR2			0x8
+#define   BT_CR2_IRQ_STS_HBUSY	BIT(6)
+#define   BT_CR2_IRQ_STS_H2B	BIT(0)
+#define BT_CR3			0xc
+#define BT_CTRL			0x10
+#define   BT_CTRL_B_BUSY	BIT(7)
+#define   BT_CTRL_H_BUSY	BIT(6)
+#define   BT_CTRL_OEM0		BIT(5)
+#define   BT_CTRL_SMS_ATN	BIT(4)
+#define   BT_CTRL_B2H_ATN	BIT(3)
+#define   BT_CTRL_H2B_ATN	BIT(2)
+#define   BT_CTRL_CLR_RD_PTR	BIT(1)
+#define   BT_CTRL_CLR_WR_PTR	BIT(0)
+#define BT_BMC2HOST		0x14
+#define BT_INTMASK		0x18
+#define   BT_INTMASK_BMC_HWRST	BIT(7)
+#define   BT_INTMASK_B2H_IRQ	BIT(1)
+#define   BT_INTMASK_B2H_IRQEN	BIT(0)
 
 #define BT_BMC_BUFFER_SIZE 256
 
@@ -361,7 +363,7 @@ static irqreturn_t bt_bmc_irq(int irq, void *arg)
 
 	reg = readl(bt_bmc->base + BT_CR2);
 
-	reg &= BT_CR2_IRQ_H2B | BT_CR2_IRQ_HBUSY;
+	reg &= BT_CR2_IRQ_STS_H2B | BT_CR2_IRQ_STS_HBUSY;
 	if (!reg)
 		return IRQ_NONE;
 
@@ -398,7 +400,7 @@ static int bt_bmc_config_irq(struct bt_bmc *bt_bmc,
 	 * message to the BT buffer
 	 */
 	reg = readl(bt_bmc->base + BT_CR1);
-	reg |= BT_CR1_IRQ_H2B | BT_CR1_IRQ_HBUSY;
+	reg |= BT_CR1_IRQ_EN_H2B | BT_CR1_IRQ_EN_HBUSY;
 	writel(reg, bt_bmc->base + BT_CR1);
 
 	return 0;
@@ -447,12 +449,12 @@ static int bt_bmc_probe(struct platform_device *pdev)
 		add_timer(&bt_bmc->poll_timer);
 	}
 
-	writel((BT_IO_BASE << BT_CR0_IO_BASE) |
-		     (BT_IRQ << BT_CR0_IRQ) |
-		     BT_CR0_EN_CLR_SLV_RDP |
-		     BT_CR0_EN_CLR_SLV_WRP |
-		     BT_CR0_ENABLE_IBT,
-		bt_bmc->base + BT_CR0);
+	writel(FIELD_PREP(BT_CR0_IO_BASE, BT_IO_BASE) |
+	       FIELD_PREP(BT_CR0_SIRQ, BT_IRQ) |
+	       BT_CR0_EN_CLR_SLV_RDP |
+	       BT_CR0_EN_CLR_SLV_WRP |
+	       BT_CR0_ENABLE_IBT,
+	       bt_bmc->base + BT_CR0);
 
 	clr_b_busy(bt_bmc);
 

-- 
2.34.1

[PATCH 2/4] ipmi: bt-bmc: Track open state per device

[Yu-Che Hsieh via B4 Relay]

From: Yu-Che Hsieh <yc_hsieh@aspeedtech.com>

The driver uses a global open count to allow only one userspace client

to open the BT device at a time. This works for a single device, but

also prevents independent BT device instances from being opened

concurrently.

Move the open count into struct bt_bmc so each device instance tracks

its own open state. This preserves the single-open behavior per device

while allowing multiple BT devices to operate independently.

Signed-off-by: Yu-Che Hsieh <yc_hsieh@aspeedtech.com>
---
 drivers/char/ipmi/bt-bmc.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

diff --git a/drivers/char/ipmi/bt-bmc.c b/drivers/char/ipmi/bt-bmc.c
index f3c67272502f..486ecc0b6815 100644
--- a/drivers/char/ipmi/bt-bmc.c
+++ b/drivers/char/ipmi/bt-bmc.c
@@ -64,10 +64,9 @@ struct bt_bmc {
 	wait_queue_head_t	queue;
 	struct timer_list	poll_timer;
 	struct mutex		mutex;
+	atomic_t		open_count;
 };
 
-static atomic_t open_count = ATOMIC_INIT(0);
-
 static u8 bt_inb(struct bt_bmc *bt_bmc, int reg)
 {
 	return readb(bt_bmc->base + reg);
@@ -152,12 +151,12 @@ static int bt_bmc_open(struct inode *inode, struct file *file)
 {
 	struct bt_bmc *bt_bmc = file_bt_bmc(file);
 
-	if (atomic_inc_return(&open_count) == 1) {
+	if (atomic_inc_return(&bt_bmc->open_count) == 1) {
 		clr_b_busy(bt_bmc);
 		return 0;
 	}
 
-	atomic_dec(&open_count);
+	atomic_dec(&bt_bmc->open_count);
 	return -EBUSY;
 }
 
@@ -313,7 +312,7 @@ static int bt_bmc_release(struct inode *inode, struct file *file)
 {
 	struct bt_bmc *bt_bmc = file_bt_bmc(file);
 
-	atomic_dec(&open_count);
+	atomic_dec(&bt_bmc->open_count);
 	set_b_busy(bt_bmc);
 	return 0;
 }
@@ -425,6 +424,8 @@ static int bt_bmc_probe(struct platform_device *pdev)
 	if (IS_ERR(bt_bmc->base))
 		return PTR_ERR(bt_bmc->base);
 
+	atomic_set(&bt_bmc->open_count, 0);
+
 	mutex_init(&bt_bmc->mutex);
 	init_waitqueue_head(&bt_bmc->queue);
 

-- 
2.34.1

[PATCH 3/4] dt-bindings: ipmi: Add optional LPC properties to ASPEED BT devices

[Yu-Che Hsieh via B4 Relay]

From: Yu-Che Hsieh <yc_hsieh@aspeedtech.com>

Allocating IO and IRQ resources to LPC devices is in-theory an operation

for the host, however ASPEED systems describe these resources through

BMC-internal configuration, as already supported by the ASPEED KCS BMC

binding.

Add aspeed,lpc-io-reg and aspeed,lpc-interrupts to the ASPEED BT BMC

binding so firmware can describe the host LPC IO address and SerIRQ

configuration using the same properties as KCS devices.

Signed-off-by: Yu-Che Hsieh <yc_hsieh@aspeedtech.com>
---
 .../bindings/ipmi/aspeed,ast2400-ibt-bmc.yaml       | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)

diff --git a/Documentation/devicetree/bindings/ipmi/aspeed,ast2400-ibt-bmc.yaml b/Documentation/devicetree/bindings/ipmi/aspeed,ast2400-ibt-bmc.yaml
index c4f7cdbbe16b..1803c6bbae93 100644
--- a/Documentation/devicetree/bindings/ipmi/aspeed,ast2400-ibt-bmc.yaml
+++ b/Documentation/devicetree/bindings/ipmi/aspeed,ast2400-ibt-bmc.yaml
@@ -25,6 +25,24 @@ properties:
   interrupts:
     maxItems: 1
 
+  aspeed,lpc-io-reg:
+    $ref: /schemas/types.yaml#/definitions/uint32-array
+    maxItems: 1
+    description: |
+      The host CPU LPC IO address for the BT device.
+
+  aspeed,lpc-interrupts:
+    $ref: /schemas/types.yaml#/definitions/uint32-array
+    minItems: 2
+    maxItems: 2
+    description: |
+      A 2-cell property expressing the LPC SerIRQ number and the interrupt
+      level/sense encoding (specified in the standard fashion).
+
+      Note that the generated interrupt is issued from the BMC to the host, and
+      thus the target interrupt controller is not captured by the BMC's
+      devicetree.
+
 required:
   - compatible
   - reg
@@ -35,10 +53,13 @@ additionalProperties: false
 examples:
   - |
     #include <dt-bindings/clock/aspeed-clock.h>
+    #include <dt-bindings/interrupt-controller/irq.h>
 
     bt@1e789140 {
         compatible = "aspeed,ast2400-ibt-bmc";
         reg = <0x1e789140 0x18>;
         interrupts = <8>;
         clocks = <&syscon ASPEED_CLK_GATE_LCLK>;
+        aspeed,lpc-io-reg = <0xe4>;
+        aspeed,lpc-interrupts = <10 IRQ_TYPE_LEVEL_LOW>;
     };

-- 
2.34.1

[PATCH 4/4] ipmi: bt-bmc: Read LPC address and SerIRQ from device tree

[Yu-Che Hsieh via B4 Relay]

From: Yu-Che Hsieh <yc_hsieh@aspeedtech.com>

The BT interface currently programs a fixed host LPC IO address and

SerIRQ number. This works for the existing single-channel setup, but

does not allow the host interface parameters to be described by firmware.

Read the LPC IO address from aspeed,lpc-io-reg and the SerIRQ number

and interrupt type from aspeed,lpc-interrupts. Keep the existing IO

address, SerIRQ number, and level-low interrupt type as defaults when

the properties are not present.

Signed-off-by: Yu-Che Hsieh <yc_hsieh@aspeedtech.com>
---
 drivers/char/ipmi/bt-bmc.c | 39 +++++++++++++++++++++++++++++++++++++--
 1 file changed, 37 insertions(+), 2 deletions(-)

diff --git a/drivers/char/ipmi/bt-bmc.c b/drivers/char/ipmi/bt-bmc.c
index 486ecc0b6815..6e1f941e63db 100644
--- a/drivers/char/ipmi/bt-bmc.c
+++ b/drivers/char/ipmi/bt-bmc.c
@@ -65,6 +65,12 @@ struct bt_bmc {
 	struct timer_list	poll_timer;
 	struct mutex		mutex;
 	atomic_t		open_count;
+	u32			io_addr;
+
+	struct {
+		u32 id;
+		u32 type;
+	} sirq;
 };
 
 static u8 bt_inb(struct bt_bmc *bt_bmc, int reg)
@@ -429,6 +435,33 @@ static int bt_bmc_probe(struct platform_device *pdev)
 	mutex_init(&bt_bmc->mutex);
 	init_waitqueue_head(&bt_bmc->queue);
 
+	rc = of_property_read_u32(dev->of_node, "aspeed,lpc-io-reg",
+				  &bt_bmc->io_addr);
+	if (rc) {
+		bt_bmc->io_addr = BT_IO_BASE;
+	} else if (bt_bmc->io_addr > FIELD_MAX(BT_CR0_IO_BASE)) {
+		dev_err(dev, "invalid LPC IO address\n");
+		return -EINVAL;
+	}
+
+	rc = of_property_read_u32_array(dev->of_node, "aspeed,lpc-interrupts",
+					(u32 *)&bt_bmc->sirq, 2);
+	if (rc) {
+		bt_bmc->sirq.id = BT_IRQ;
+		bt_bmc->sirq.type = IRQ_TYPE_LEVEL_LOW;
+	} else {
+		if (bt_bmc->sirq.id > FIELD_MAX(BT_CR0_SIRQ)) {
+			dev_err(dev, "invalid SerIRQ number\n");
+			return -EINVAL;
+		}
+
+		if (bt_bmc->sirq.type != IRQ_TYPE_LEVEL_HIGH &&
+		    bt_bmc->sirq.type != IRQ_TYPE_LEVEL_LOW) {
+			dev_err(dev, "invalid SerIRQ type\n");
+			return -EINVAL;
+		}
+	}
+
 	bt_bmc->miscdev.minor	= MISC_DYNAMIC_MINOR;
 	bt_bmc->miscdev.name	= DEVICE_NAME;
 	bt_bmc->miscdev.fops	= &bt_bmc_fops;
@@ -450,8 +483,10 @@ static int bt_bmc_probe(struct platform_device *pdev)
 		add_timer(&bt_bmc->poll_timer);
 	}
 
-	writel(FIELD_PREP(BT_CR0_IO_BASE, BT_IO_BASE) |
-	       FIELD_PREP(BT_CR0_SIRQ, BT_IRQ) |
+	writel(FIELD_PREP(BT_CR0_IO_BASE, bt_bmc->io_addr) |
+	       FIELD_PREP(BT_CR0_SIRQ, bt_bmc->sirq.id) |
+	       FIELD_PREP(BT_CR0_SIRQ_TYPE,
+			  bt_bmc->sirq.type == IRQ_TYPE_LEVEL_LOW ? 0 : 1) |
 	       BT_CR0_EN_CLR_SLV_RDP |
 	       BT_CR0_EN_CLR_SLV_WRP |
 	       BT_CR0_ENABLE_IBT,

-- 
2.34.1

Re: [PATCH 3/4] dt-bindings: ipmi: Add optional LPC properties to ASPEED BT devices

[Conor Dooley]

[-- Attachment #1: Type: text/plain, Size: 2563 bytes --]

On Mon, Jun 29, 2026 at 02:49:00PM +0800, Yu-Che Hsieh via B4 Relay wrote:
> From: Yu-Che Hsieh <yc_hsieh@aspeedtech.com>
> 
> Allocating IO and IRQ resources to LPC devices is in-theory an operation
> 
> for the host, however ASPEED systems describe these resources through
> 
> BMC-internal configuration, as already supported by the ASPEED KCS BMC
> 
> binding.
> 
> Add aspeed,lpc-io-reg and aspeed,lpc-interrupts to the ASPEED BT BMC
> 
> binding so firmware can describe the host LPC IO address and SerIRQ
> 
> configuration using the same properties as KCS devices.
> 
> Signed-off-by: Yu-Che Hsieh <yc_hsieh@aspeedtech.com>
> ---
>  .../bindings/ipmi/aspeed,ast2400-ibt-bmc.yaml       | 21 +++++++++++++++++++++
>  1 file changed, 21 insertions(+)
> 
> diff --git a/Documentation/devicetree/bindings/ipmi/aspeed,ast2400-ibt-bmc.yaml b/Documentation/devicetree/bindings/ipmi/aspeed,ast2400-ibt-bmc.yaml
> index c4f7cdbbe16b..1803c6bbae93 100644
> --- a/Documentation/devicetree/bindings/ipmi/aspeed,ast2400-ibt-bmc.yaml
> +++ b/Documentation/devicetree/bindings/ipmi/aspeed,ast2400-ibt-bmc.yaml
> @@ -25,6 +25,24 @@ properties:
>    interrupts:
>      maxItems: 1
>  
> +  aspeed,lpc-io-reg:
> +    $ref: /schemas/types.yaml#/definitions/uint32-array
> +    maxItems: 1
> +    description: |
> +      The host CPU LPC IO address for the BT device.
> +
> +  aspeed,lpc-interrupts:
> +    $ref: /schemas/types.yaml#/definitions/uint32-array
> +    minItems: 2
> +    maxItems: 2
> +    description: |
> +      A 2-cell property expressing the LPC SerIRQ number and the interrupt
> +      level/sense encoding (specified in the standard fashion).
> +
> +      Note that the generated interrupt is issued from the BMC to the host, and
> +      thus the target interrupt controller is not captured by the BMC's
> +      devicetree.

Why can these two properties not just be an additional reg and
interrupts entry?


Cheers,
Conor.

> +
>  required:
>    - compatible
>    - reg
> @@ -35,10 +53,13 @@ additionalProperties: false
>  examples:
>    - |
>      #include <dt-bindings/clock/aspeed-clock.h>
> +    #include <dt-bindings/interrupt-controller/irq.h>
>  
>      bt@1e789140 {
>          compatible = "aspeed,ast2400-ibt-bmc";
>          reg = <0x1e789140 0x18>;
>          interrupts = <8>;
>          clocks = <&syscon ASPEED_CLK_GATE_LCLK>;
> +        aspeed,lpc-io-reg = <0xe4>;
> +        aspeed,lpc-interrupts = <10 IRQ_TYPE_LEVEL_LOW>;
>      };
> 
> -- 
> 2.34.1
> 
> 

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 228 bytes --]

Re: [PATCH 3/4] dt-bindings: ipmi: Add optional LPC properties to ASPEED BT devices

[YC Hsieh]

Hi Conor,

Thanks for the review.

> Why can these two properties not just be an additional reg and
> interrupts entry?

The existing reg and interrupts properties describe BMC-side resources.

For this device, reg is the BMC-visible MMIO register range, and
interrupts is the interrupt from the BT controller to the BMC CPU. The
LPC IO address and SerIRQ described here are host-facing resources: the
LPC IO decode address is in the host LPC IO space, and the SerIRQ is
generated from the BMC toward the host.

There is no host interrupt controller represented in the BMC devicetree
to target with an interrupts entry, and adding another reg entry would
mix the BMC MMIO address space with the host LPC IO address space.

This follows the existing ASPEED KCS BMC bindings, which use
aspeed,lpc-io-reg and aspeed,lpc-interrupts for the same host-facing LPC
configuration.

Regards,
Yu-Che
************* Email Confidentiality Notice ********************
免責聲明:
本信件(或其附件)可能包含機密資訊，並受法律保護。如 台端非指定之收件者，請以電子郵件通知本電子郵件之發送者, 並請立即刪除本電子郵件及其附件和銷毀所有複印件。謝謝您的合作!
信驊科技以誠信正直原則永續經營企業，並已委由第三方公正單位勤業眾信及信驊科技獨立董事來管理匿名舉報系統，如各個利害關係人等有發現任何違法及違反公司從業道德、違反法令法規及專業準則、亦或霸凌及違反性別平等之情事，請直接透過以下可選擇匿名之舉報系統舉報，再次感謝您協助信驊持續邁向永續經營。信驊科技舉報網站連結：https://secure.conductwatch.com/aspeedtech/

DISCLAIMER:
This message (and any attachments) may contain legally privileged and/or other confidential information. If you have received it in error, please notify the sender by reply e-mail and immediately delete the e-mail and any attachments without copying or disclosing the contents. Thank you.
ASPEED Technology is committed to sustainable business practices based on integrity and honesty principles. In order to ensure that all information can be openly and transparently communicated, a third-party independent organization, Deloitte and ASPEED Technology's independent directors, have been entrusted to manage the anonymous reporting system. If any stakeholders discover any illegal activities, violations of the company's professional ethics, infringements of laws and regulations, or incidents of bullying and gender inequality, please directly report through the anonymous reporting system provided below. We thank you for your assistance in helping ASPEED Technology continue its journey towards sustainable operations: https://secure.conductwatch.com/aspeedtech/

Re: [PATCH 3/4] dt-bindings: ipmi: Add optional LPC properties to ASPEED BT devices

[Krzysztof Kozlowski]

On 29/06/2026 08:49, Yu-Che Hsieh via B4 Relay wrote:
> From: Yu-Che Hsieh <yc_hsieh@aspeedtech.com>
> 
> Allocating IO and IRQ resources to LPC devices is in-theory an operation
> 
> for the host, however ASPEED systems describe these resources through
> 
> BMC-internal configuration, as already supported by the ASPEED KCS BMC

What

is

with

this

line breaks?


> 
> binding.
> 
> Add aspeed,lpc-io-reg and aspeed,lpc-interrupts to the ASPEED BT BMC
> 
> binding so firmware can describe the host LPC IO address and SerIRQ
> 
> configuration using the same properties as KCS devices.
> 
> Signed-off-by: Yu-Che Hsieh <yc_hsieh@aspeedtech.com>
> ---
>  .../bindings/ipmi/aspeed,ast2400-ibt-bmc.yaml       | 21 +++++++++++++++++++++
>  1 file changed, 21 insertions(+)
> 
> diff --git a/Documentation/devicetree/bindings/ipmi/aspeed,ast2400-ibt-bmc.yaml b/Documentation/devicetree/bindings/ipmi/aspeed,ast2400-ibt-bmc.yaml
> index c4f7cdbbe16b..1803c6bbae93 100644
> --- a/Documentation/devicetree/bindings/ipmi/aspeed,ast2400-ibt-bmc.yaml
> +++ b/Documentation/devicetree/bindings/ipmi/aspeed,ast2400-ibt-bmc.yaml
> @@ -25,6 +25,24 @@ properties:
>    interrupts:
>      maxItems: 1
>  
> +  aspeed,lpc-io-reg:
> +    $ref: /schemas/types.yaml#/definitions/uint32-array
> +    maxItems: 1
> +    description: |
> +      The host CPU LPC IO address for the BT device.

No, you do not get second reg property.

> +
> +  aspeed,lpc-interrupts:
> +    $ref: /schemas/types.yaml#/definitions/uint32-array
> +    minItems: 2
> +    maxItems: 2
> +    description: |
> +      A 2-cell property expressing the LPC SerIRQ number and the interrupt
> +      level/sense encoding (specified in the standard fashion).
> +
> +      Note that the generated interrupt is issued from the BMC to the host, and
> +      thus the target interrupt controller is not captured by the BMC's
> +      devicetree.

No, you do not get second interrupts property.

> 


Best regards,
Krzysztof

Re: [PATCH 3/4] dt-bindings: ipmi: Add optional LPC properties to ASPEED BT devices

[Conor Dooley]

[-- Attachment #1: Type: text/plain, Size: 2057 bytes --]

On Tue, Jun 30, 2026 at 02:24:52AM +0000, YC Hsieh wrote:

> ************* Email Confidentiality Notice ********************
> 免責聲明:
> 本信件(或其附件)可能包含機密資訊，並受法律保護。如 台端非指定之收件者，請以電子郵件通知本電子郵件之發送者, 並請立即刪除本電子郵件及其附件和銷毀所有複印件。謝謝您的合作!
> 信驊科技以誠信正直原則永續經營企業，並已委由第三方公正單位勤業眾信及信驊科技獨立董事來管理匿名舉報系統，如各個利害關係人等有發現任何違法及違反公司從業道德、違反法令法規及專業準則、亦或霸凌及違反性別平等之情事，請直接透過以下可選擇匿名之舉報系統舉報，再次感謝您協助信驊持續邁向永續經營。信驊科技舉報網站連結：https://secure.conductwatch.com/aspeedtech/
> 
> DISCLAIMER:
> This message (and any attachments) may contain legally privileged and/or other confidential information. If you have received it in error, please notify the sender by reply e-mail and immediately delete the e-mail and any attachments without copying or disclosing the contents. Thank you.
> ASPEED Technology is committed to sustainable business practices based on integrity and honesty principles. In order to ensure that all information can be openly and transparently communicated, a third-party independent organization, Deloitte and ASPEED Technology's independent directors, have been entrusted to manage the anonymous reporting system. If any stakeholders discover any illegal activities, violations of the company's professional ethics, infringements of laws and regulations, or incidents of bullying and gender inequality, please directly report through the anonymous reporting system provided below. We thank you for your assistance in helping ASPEED Technology continue its journey towards sustainable operations: https://secure.conductwatch.com/aspeedtech/

Sort this out, it's not compatible with mailing lists.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 228 bytes --]

Re: [PATCH 3/4] dt-bindings: ipmi: Add optional LPC properties to ASPEED BT devices

[Rob Herring]

On Tue, Jun 30, 2026 at 08:11:34AM +0200, Krzysztof Kozlowski wrote:
> On 29/06/2026 08:49, Yu-Che Hsieh via B4 Relay wrote:
> > From: Yu-Che Hsieh <yc_hsieh@aspeedtech.com>
> > 
> > Allocating IO and IRQ resources to LPC devices is in-theory an operation
> > 
> > for the host, however ASPEED systems describe these resources through
> > 
> > BMC-internal configuration, as already supported by the ASPEED KCS BMC
> 
> What
> 
> is
> 
> with
> 
> this
> 
> line breaks?

I've seen Codex do this... It amazes me how hard it is to get it to 
write properly formatted commit messages and then not forget how to 
write them.

Rob