[PATCH] aspeed: Add Supermicro X11 SPI machine type

[PATCH] aspeed: Add Supermicro X11 SPI machine type

[Guenter Roeck]

supermicrox11-bmc is configured with ast2400-a1 SoC. This does not match
the Supermicro documentation for X11 BMCs, and it does not match the
devicetree file in the Linux kernel.

As it turns out, some Supermicro X11 motherboards use AST2400 SoCs,
while others use AST2500.

Introduce new machine type supermicrox11-spi-bmc with AST2500 SoC
to match the devicetree description in the Linux kernel. Hardware
configuration details for this machine type are guesswork and taken
from defaults as well as from the Linux kernel devicetree file.

The new machine type was tested with aspeed-bmc-supermicro-x11spi.dts
from the Linux kernel and with Linux versions 6.0.3 and 6.1-rc2.
Linux booted successfully from initrd and from both SPI interfaces.
Ethernet interfaces were confirmed to be operational.

Signed-off-by: Guenter Roeck <linux@roeck-us.net>
---
 hw/arm/aspeed.c | 33 +++++++++++++++++++++++++++++++++
 1 file changed, 33 insertions(+)

diff --git a/hw/arm/aspeed.c b/hw/arm/aspeed.c
index c282ead38f..56d007de5a 100644
--- a/hw/arm/aspeed.c
+++ b/hw/arm/aspeed.c
@@ -71,6 +71,16 @@ struct AspeedMachineState {
         SCU_HW_STRAP_VGA_SIZE_SET(VGA_16M_DRAM) |                       \
         SCU_AST2400_HW_STRAP_BOOT_MODE(AST2400_SPI_BOOT))
 
+/* TODO: Find the actual hardware value */
+#define SUPERMICROX11_SPI_BMC_HW_STRAP1 (                               \
+        AST2500_HW_STRAP1_DEFAULTS |                                    \
+        SCU_AST2500_HW_STRAP_SPI_AUTOFETCH_ENABLE |                     \
+        SCU_AST2500_HW_STRAP_GPIO_STRAP_ENABLE |                        \
+        SCU_AST2500_HW_STRAP_UART_DEBUG |                               \
+        SCU_AST2500_HW_STRAP_DDR4_ENABLE |                              \
+        SCU_HW_STRAP_SPI_WIDTH |                                        \
+        SCU_HW_STRAP_SPI_MODE(SCU_HW_STRAP_SPI_M_S_EN))
+
 /* AST2500 evb hardware value: 0xF100C2E6 */
 #define AST2500_EVB_HW_STRAP1 ((                                        \
         AST2500_HW_STRAP1_DEFAULTS |                                    \
@@ -1172,6 +1182,25 @@ static void aspeed_machine_supermicrox11_bmc_class_init(ObjectClass *oc,
     mc->default_ram_size = 256 * MiB;
 }
 
+static void aspeed_machine_supermicrox11_spi_bmc_class_init(ObjectClass *oc,
+                                                            void *data)
+{
+    MachineClass *mc = MACHINE_CLASS(oc);
+    AspeedMachineClass *amc = ASPEED_MACHINE_CLASS(oc);
+
+    mc->desc       = "Supermicro X11 SPI BMC (ARM1176)";
+    amc->soc_name  = "ast2500-a1";
+    amc->hw_strap1 = SUPERMICROX11_SPI_BMC_HW_STRAP1;
+    amc->fmc_model = "mx25l25635e";
+    amc->spi_model = "mx25l25635e";
+    amc->num_cs    = 1;
+    amc->macs_mask = ASPEED_MAC0_ON | ASPEED_MAC1_ON;
+    amc->i2c_init  = palmetto_bmc_i2c_init;
+    mc->default_ram_size = 512 * MiB;
+    mc->default_cpus = mc->min_cpus = mc->max_cpus =
+        aspeed_soc_num_cpus(amc->soc_name);
+}
+
 static void aspeed_machine_ast2500_evb_class_init(ObjectClass *oc, void *data)
 {
     MachineClass *mc = MACHINE_CLASS(oc);
@@ -1546,6 +1575,10 @@ static const TypeInfo aspeed_machine_types[] = {
         .name          = MACHINE_TYPE_NAME("supermicrox11-bmc"),
         .parent        = TYPE_ASPEED_MACHINE,
         .class_init    = aspeed_machine_supermicrox11_bmc_class_init,
+    }, {
+        .name          = MACHINE_TYPE_NAME("supermicrox11-spi-bmc"),
+        .parent        = TYPE_ASPEED_MACHINE,
+        .class_init    = aspeed_machine_supermicrox11_spi_bmc_class_init,
     }, {
         .name          = MACHINE_TYPE_NAME("ast2500-evb"),
         .parent        = TYPE_ASPEED_MACHINE,
-- 
2.36.2

Re: [PATCH] aspeed: Add Supermicro X11 SPI machine type

[Philippe Mathieu-Daudé]

On 25/10/22 18:51, Guenter Roeck wrote:
> supermicrox11-bmc is configured with ast2400-a1 SoC. This does not match
> the Supermicro documentation for X11 BMCs, and it does not match the
> devicetree file in the Linux kernel.
> 
> As it turns out, some Supermicro X11 motherboards use AST2400 SoCs,
> while others use AST2500.
> 
> Introduce new machine type supermicrox11-spi-bmc with AST2500 SoC
> to match the devicetree description in the Linux kernel. Hardware
> configuration details for this machine type are guesswork and taken
> from defaults as well as from the Linux kernel devicetree file.
> 
> The new machine type was tested with aspeed-bmc-supermicro-x11spi.dts
> from the Linux kernel and with Linux versions 6.0.3 and 6.1-rc2.
> Linux booted successfully from initrd and from both SPI interfaces.
> Ethernet interfaces were confirmed to be operational.
> 
> Signed-off-by: Guenter Roeck <linux@roeck-us.net>
> ---
>   hw/arm/aspeed.c | 33 +++++++++++++++++++++++++++++++++
>   1 file changed, 33 insertions(+)

Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>

Re: [PATCH] aspeed: Add Supermicro X11 SPI machine type

[Cédric Le Goater]

On 10/25/22 18:51, Guenter Roeck wrote:
> supermicrox11-bmc is configured with ast2400-a1 SoC. This does not match
> the Supermicro documentation for X11 BMCs, and it does not match the
> devicetree file in the Linux kernel.
> 
> As it turns out, some Supermicro X11 motherboards use AST2400 SoCs,
> while others use AST2500.
> 
> Introduce new machine type supermicrox11-spi-bmc with AST2500 SoC
> to match the devicetree description in the Linux kernel. Hardware
> configuration details for this machine type are guesswork and taken
> from defaults as well as from the Linux kernel devicetree file.
> 
> The new machine type was tested with aspeed-bmc-supermicro-x11spi.dts
> from the Linux kernel and with Linux versions 6.0.3 and 6.1-rc2.
> Linux booted successfully from initrd and from both SPI interfaces.
> Ethernet interfaces were confirmed to be operational.
> 
> Signed-off-by: Guenter Roeck <linux@roeck-us.net>

Hello Guenter

Would it be possible to contribute a functional test for this
machine ?

Since this machine contributes little to the Aspeed models,
its value lies in the firmware it can run to exercise the
models. Without functional tests, I plan to schedule the
removal in the QEMU 10.2 cycle.


Thanks,

C.



> ---
>   hw/arm/aspeed.c | 33 +++++++++++++++++++++++++++++++++
>   1 file changed, 33 insertions(+)
> 
> diff --git a/hw/arm/aspeed.c b/hw/arm/aspeed.c
> index c282ead38f..56d007de5a 100644
> --- a/hw/arm/aspeed.c
> +++ b/hw/arm/aspeed.c
> @@ -71,6 +71,16 @@ struct AspeedMachineState {
>           SCU_HW_STRAP_VGA_SIZE_SET(VGA_16M_DRAM) |                       \
>           SCU_AST2400_HW_STRAP_BOOT_MODE(AST2400_SPI_BOOT))
>   
> +/* TODO: Find the actual hardware value */
> +#define SUPERMICROX11_SPI_BMC_HW_STRAP1 (                               \
> +        AST2500_HW_STRAP1_DEFAULTS |                                    \
> +        SCU_AST2500_HW_STRAP_SPI_AUTOFETCH_ENABLE |                     \
> +        SCU_AST2500_HW_STRAP_GPIO_STRAP_ENABLE |                        \
> +        SCU_AST2500_HW_STRAP_UART_DEBUG |                               \
> +        SCU_AST2500_HW_STRAP_DDR4_ENABLE |                              \
> +        SCU_HW_STRAP_SPI_WIDTH |                                        \
> +        SCU_HW_STRAP_SPI_MODE(SCU_HW_STRAP_SPI_M_S_EN))
> +
>   /* AST2500 evb hardware value: 0xF100C2E6 */
>   #define AST2500_EVB_HW_STRAP1 ((                                        \
>           AST2500_HW_STRAP1_DEFAULTS |                                    \
> @@ -1172,6 +1182,25 @@ static void aspeed_machine_supermicrox11_bmc_class_init(ObjectClass *oc,
>       mc->default_ram_size = 256 * MiB;
>   }
>   
> +static void aspeed_machine_supermicrox11_spi_bmc_class_init(ObjectClass *oc,
> +                                                            void *data)
> +{
> +    MachineClass *mc = MACHINE_CLASS(oc);
> +    AspeedMachineClass *amc = ASPEED_MACHINE_CLASS(oc);
> +
> +    mc->desc       = "Supermicro X11 SPI BMC (ARM1176)";
> +    amc->soc_name  = "ast2500-a1";
> +    amc->hw_strap1 = SUPERMICROX11_SPI_BMC_HW_STRAP1;
> +    amc->fmc_model = "mx25l25635e";
> +    amc->spi_model = "mx25l25635e";
> +    amc->num_cs    = 1;
> +    amc->macs_mask = ASPEED_MAC0_ON | ASPEED_MAC1_ON;
> +    amc->i2c_init  = palmetto_bmc_i2c_init;
> +    mc->default_ram_size = 512 * MiB;
> +    mc->default_cpus = mc->min_cpus = mc->max_cpus =
> +        aspeed_soc_num_cpus(amc->soc_name);
> +}
> +
>   static void aspeed_machine_ast2500_evb_class_init(ObjectClass *oc, void *data)
>   {
>       MachineClass *mc = MACHINE_CLASS(oc);
> @@ -1546,6 +1575,10 @@ static const TypeInfo aspeed_machine_types[] = {
>           .name          = MACHINE_TYPE_NAME("supermicrox11-bmc"),
>           .parent        = TYPE_ASPEED_MACHINE,
>           .class_init    = aspeed_machine_supermicrox11_bmc_class_init,
> +    }, {
> +        .name          = MACHINE_TYPE_NAME("supermicrox11-spi-bmc"),
> +        .parent        = TYPE_ASPEED_MACHINE,
> +        .class_init    = aspeed_machine_supermicrox11_spi_bmc_class_init,
>       }, {
>           .name          = MACHINE_TYPE_NAME("ast2500-evb"),
>           .parent        = TYPE_ASPEED_MACHINE,

Re: [PATCH] aspeed: Add Supermicro X11 SPI machine type

[Guenter Roeck]

On 9/29/25 22:55, Cédric Le Goater wrote:
> On 10/25/22 18:51, Guenter Roeck wrote:
>> supermicrox11-bmc is configured with ast2400-a1 SoC. This does not match
>> the Supermicro documentation for X11 BMCs, and it does not match the
>> devicetree file in the Linux kernel.
>>
>> As it turns out, some Supermicro X11 motherboards use AST2400 SoCs,
>> while others use AST2500.
>>
>> Introduce new machine type supermicrox11-spi-bmc with AST2500 SoC
>> to match the devicetree description in the Linux kernel. Hardware
>> configuration details for this machine type are guesswork and taken
>> from defaults as well as from the Linux kernel devicetree file.
>>
>> The new machine type was tested with aspeed-bmc-supermicro-x11spi.dts
>> from the Linux kernel and with Linux versions 6.0.3 and 6.1-rc2.
>> Linux booted successfully from initrd and from both SPI interfaces.
>> Ethernet interfaces were confirmed to be operational.
>>
>> Signed-off-by: Guenter Roeck <linux@roeck-us.net>
> 
> Hello Guenter
> 
> Would it be possible to contribute a functional test for this
> machine ?
> 
> Since this machine contributes little to the Aspeed models,
> its value lies in the firmware it can run to exercise the
> models. Without functional tests, I plan to schedule the
> removal in the QEMU 10.2 cycle.
> 

I use it to test the Linux kernel. Nor worth enough to spend more time on.
Just drop it; I'll drop the respective test runs.

Guenter

Re: [PATCH] aspeed: Add Supermicro X11 SPI machine type

[Cédric Le Goater]

On 9/30/25 12:08, Guenter Roeck wrote:
> On 9/29/25 22:55, Cédric Le Goater wrote:
>> On 10/25/22 18:51, Guenter Roeck wrote:
>>> supermicrox11-bmc is configured with ast2400-a1 SoC. This does not match
>>> the Supermicro documentation for X11 BMCs, and it does not match the
>>> devicetree file in the Linux kernel.
>>>
>>> As it turns out, some Supermicro X11 motherboards use AST2400 SoCs,
>>> while others use AST2500.
>>>
>>> Introduce new machine type supermicrox11-spi-bmc with AST2500 SoC
>>> to match the devicetree description in the Linux kernel. Hardware
>>> configuration details for this machine type are guesswork and taken
>>> from defaults as well as from the Linux kernel devicetree file.
>>>
>>> The new machine type was tested with aspeed-bmc-supermicro-x11spi.dts
>>> from the Linux kernel and with Linux versions 6.0.3 and 6.1-rc2.
>>> Linux booted successfully from initrd and from both SPI interfaces.
>>> Ethernet interfaces were confirmed to be operational.
>>>
>>> Signed-off-by: Guenter Roeck <linux@roeck-us.net>
>>
>> Hello Guenter
>>
>> Would it be possible to contribute a functional test for this
>> machine ?
>>
>> Since this machine contributes little to the Aspeed models,
>> its value lies in the firmware it can run to exercise the
>> models. Without functional tests, I plan to schedule the
>> removal in the QEMU 10.2 cycle.
>>
> 
> I use it to test the Linux kernel. Nor worth enough to spend more time on.
> Just drop it; I'll drop the respective test runs.
Can the FW be downloaded ?

C.

Re: [PATCH] aspeed: Add Supermicro X11 SPI machine type

[Guenter Roeck]

On 9/30/25 04:13, Cédric Le Goater wrote:
> On 9/30/25 12:08, Guenter Roeck wrote:
>> On 9/29/25 22:55, Cédric Le Goater wrote:
>>> On 10/25/22 18:51, Guenter Roeck wrote:
>>>> supermicrox11-bmc is configured with ast2400-a1 SoC. This does not match
>>>> the Supermicro documentation for X11 BMCs, and it does not match the
>>>> devicetree file in the Linux kernel.
>>>>
>>>> As it turns out, some Supermicro X11 motherboards use AST2400 SoCs,
>>>> while others use AST2500.
>>>>
>>>> Introduce new machine type supermicrox11-spi-bmc with AST2500 SoC
>>>> to match the devicetree description in the Linux kernel. Hardware
>>>> configuration details for this machine type are guesswork and taken
>>>> from defaults as well as from the Linux kernel devicetree file.
>>>>
>>>> The new machine type was tested with aspeed-bmc-supermicro-x11spi.dts
>>>> from the Linux kernel and with Linux versions 6.0.3 and 6.1-rc2.
>>>> Linux booted successfully from initrd and from both SPI interfaces.
>>>> Ethernet interfaces were confirmed to be operational.
>>>>
>>>> Signed-off-by: Guenter Roeck <linux@roeck-us.net>
>>>
>>> Hello Guenter
>>>
>>> Would it be possible to contribute a functional test for this
>>> machine ?
>>>
>>> Since this machine contributes little to the Aspeed models,
>>> its value lies in the firmware it can run to exercise the
>>> models. Without functional tests, I plan to schedule the
>>> removal in the QEMU 10.2 cycle.
>>>
>>
>> I use it to test the Linux kernel. Nor worth enough to spend more time on.
>> Just drop it; I'll drop the respective test runs.
> Can the FW be downloaded ?
> 

I have no idea. I only test the Linux kernel (more specifically, the Linux kernel
drivers). I already have to work around issues with the board that because the
Linux kernel no longer supports the Macronix flash on that board as emulated
by qemu. My patch to let me specify the flash model for aspeed boards in qemu
was rejected. So I already have to carry a downstream qemu change to support it.
Obviously no one else uses/tests it, or they would have reported that problem.
So you might as well drop it.

The flash problem is also seen with ast2600-evb, making me wonder if that board
could/should be dropped as well.

Guenter

Re: [PATCH] aspeed: Add Supermicro X11 SPI machine type

[Cédric Le Goater]

On 9/30/25 16:35, Guenter Roeck wrote:
> On 9/30/25 04:13, Cédric Le Goater wrote:
>> On 9/30/25 12:08, Guenter Roeck wrote:
>>> On 9/29/25 22:55, Cédric Le Goater wrote:
>>>> On 10/25/22 18:51, Guenter Roeck wrote:
>>>>> supermicrox11-bmc is configured with ast2400-a1 SoC. This does not match
>>>>> the Supermicro documentation for X11 BMCs, and it does not match the
>>>>> devicetree file in the Linux kernel.
>>>>>
>>>>> As it turns out, some Supermicro X11 motherboards use AST2400 SoCs,
>>>>> while others use AST2500.
>>>>>
>>>>> Introduce new machine type supermicrox11-spi-bmc with AST2500 SoC
>>>>> to match the devicetree description in the Linux kernel. Hardware
>>>>> configuration details for this machine type are guesswork and taken
>>>>> from defaults as well as from the Linux kernel devicetree file.
>>>>>
>>>>> The new machine type was tested with aspeed-bmc-supermicro-x11spi.dts
>>>>> from the Linux kernel and with Linux versions 6.0.3 and 6.1-rc2.
>>>>> Linux booted successfully from initrd and from both SPI interfaces.
>>>>> Ethernet interfaces were confirmed to be operational.
>>>>>
>>>>> Signed-off-by: Guenter Roeck <linux@roeck-us.net>
>>>>
>>>> Hello Guenter
>>>>
>>>> Would it be possible to contribute a functional test for this
>>>> machine ?
>>>>
>>>> Since this machine contributes little to the Aspeed models,
>>>> its value lies in the firmware it can run to exercise the
>>>> models. Without functional tests, I plan to schedule the
>>>> removal in the QEMU 10.2 cycle.
>>>>
>>>
>>> I use it to test the Linux kernel. Nor worth enough to spend more time on.
>>> Just drop it; I'll drop the respective test runs.
>> Can the FW be downloaded ?
>>
> 
> I have no idea. I only test the Linux kernel (more specifically, the Linux kernel
> drivers). 

OK. you boot with -kernel and a custom rootfs I guess. That's OK
for a functional test too.

> I already have to work around issues with the board that because the
> Linux kernel no longer supports the Macronix flash on that board as emulated
> by qemu. 

Ah. I am the kernel maintainer of the spi driver. So interested
to know more. Is that related to the fact that there are mx25l25635e
and mx25l25635f flash devices with the same device ID but different
SFDP tables ? One has 4B support, the other not.

> My patch to let me specify the flash model for aspeed boards in qemu
> was rejected. 

The aspeed machines have these options :

   fmc-model=<string>     - Change the FMC Flash model
   spi-model=<string>     - Change the SPI Flash model

Isn't that enough ?

> So I already have to carry a downstream qemu change to support it.

I don't see any change under :

    https://github.com/groeck/qemu/commits/v10.0.2-local/

May be you are mistaking with the Nuvoton boards ?

Upstream lacks your UHCI proposal. Unfortunately, I don't know any
reviewers who could help us merge.

> Obviously no one else uses/tests it, or they would have reported that problem.
> So you might as well drop it.

Since you are using it for kernel, I probably won't.
  
> The flash problem is also seen with ast2600-evb, making me wonder if that board
> could/should be dropped as well.
the ast2600 EVB has numerous firmware images. See

   https://github.com/AspeedTech-BMC/openbmc/releases/

It's heavily used and none reported an issue.

buildroot has a defconfig for it too.

Thanks,

C.

Re: [PATCH] aspeed: Add Supermicro X11 SPI machine type

[Guenter Roeck]

On 9/30/25 07:54, Cédric Le Goater wrote:
> On 9/30/25 16:35, Guenter Roeck wrote:
>> On 9/30/25 04:13, Cédric Le Goater wrote:
>>> On 9/30/25 12:08, Guenter Roeck wrote:
>>>> On 9/29/25 22:55, Cédric Le Goater wrote:
>>>>> On 10/25/22 18:51, Guenter Roeck wrote:
>>>>>> supermicrox11-bmc is configured with ast2400-a1 SoC. This does not match
>>>>>> the Supermicro documentation for X11 BMCs, and it does not match the
>>>>>> devicetree file in the Linux kernel.
>>>>>>
>>>>>> As it turns out, some Supermicro X11 motherboards use AST2400 SoCs,
>>>>>> while others use AST2500.
>>>>>>
>>>>>> Introduce new machine type supermicrox11-spi-bmc with AST2500 SoC
>>>>>> to match the devicetree description in the Linux kernel. Hardware
>>>>>> configuration details for this machine type are guesswork and taken
>>>>>> from defaults as well as from the Linux kernel devicetree file.
>>>>>>
>>>>>> The new machine type was tested with aspeed-bmc-supermicro-x11spi.dts
>>>>>> from the Linux kernel and with Linux versions 6.0.3 and 6.1-rc2.
>>>>>> Linux booted successfully from initrd and from both SPI interfaces.
>>>>>> Ethernet interfaces were confirmed to be operational.
>>>>>>
>>>>>> Signed-off-by: Guenter Roeck <linux@roeck-us.net>
>>>>>
>>>>> Hello Guenter
>>>>>
>>>>> Would it be possible to contribute a functional test for this
>>>>> machine ?
>>>>>
>>>>> Since this machine contributes little to the Aspeed models,
>>>>> its value lies in the firmware it can run to exercise the
>>>>> models. Without functional tests, I plan to schedule the
>>>>> removal in the QEMU 10.2 cycle.
>>>>>
>>>>
>>>> I use it to test the Linux kernel. Nor worth enough to spend more time on.
>>>> Just drop it; I'll drop the respective test runs.
>>> Can the FW be downloaded ?
>>>
>>
>> I have no idea. I only test the Linux kernel (more specifically, the Linux kernel
>> drivers). 
> 
> OK. you boot with -kernel and a custom rootfs I guess. That's OK
> for a functional test too.
> 
>> I already have to work around issues with the board that because the
>> Linux kernel no longer supports the Macronix flash on that board as emulated
>> by qemu. 
> 
> Ah. I am the kernel maintainer of the spi driver. So interested
> to know more. Is that related to the fact that there are mx25l25635e
> and mx25l25635f flash devices with the same device ID but different
> SFDP tables ? One has 4B support, the other not.
> 
>> My patch to let me specify the flash model for aspeed boards in qemu
>> was rejected. 
> q
> The aspeed machines have these options :
> 
>    fmc-model=<string>     - Change the FMC Flash model
>    spi-model=<string>     - Change the SPI Flash model
> 
> Isn't that enough ?
> 
>> So I already have to carry a downstream qemu change to support it.
> 
> I don't see any change under :
> 
>     https://github.com/groeck/qemu/commits/v10.0.2-local/
> 
> May be you are mistaking with the Nuvoton boards ?
> 

Yes. Sorry, I am confused. That was for quanta-gsj and kudo-bmc where I had
the same problem with Macronix flashes.

> Upstream lacks your UHCI proposal. Unfortunately, I don't know any
> reviewers who could help us merge.
> 

I'll just carry it downstream until I am getting tired of maintaining it.

Thanks,
Guenter