[v2] i915: Support Intel GPU porting on any non-x86 system.

[PATCH v2] i915: Support Intel GPU porting on any non-x86 system.

Posted by zhangzhijie 1 week ago

inb/outb speccial wire not support on other ARCH.
Should detect whether arch platform support or not.

Signed-off-by: zhangzhijie <zhangzhijie@bosc.ac.cn>
---
 drivers/gpu/drm/i915/display/intel_vga.c | 27 ++++++++++++++++++------
 1 file changed, 20 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/i915/display/intel_vga.c b/drivers/gpu/drm/i915/display/intel_vga.c
index 6e125564db34..d5d6c4ba6434 100644
--- a/drivers/gpu/drm/i915/display/intel_vga.c
+++ b/drivers/gpu/drm/i915/display/intel_vga.c
@@ -41,6 +41,15 @@ static bool has_vga_pipe_sel(struct intel_display *display)
 	return DISPLAY_VER(display) < 7;
 }
 
+static bool intel_arch_support_vga_pm(struct intel_display *display)
+{
+#if defined(CONFIG_X86) || defined(CONFIG_X86_64)
+	return true;
+#else
+	return false;
+#endif
+}
+
 /* Disable the VGA plane that we never use */
 void intel_vga_disable(struct intel_display *display)
 {
@@ -66,11 +75,13 @@ void intel_vga_disable(struct intel_display *display)
 
 	/* WaEnableVGAAccessThroughIOPort:ctg,elk,ilk,snb,ivb,vlv,hsw */
 	vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
-	outb(0x01, VGA_SEQ_I);
-	sr1 = inb(VGA_SEQ_D);
-	outb(sr1 | VGA_SR01_SCREEN_OFF, VGA_SEQ_D);
-	vga_put(pdev, VGA_RSRC_LEGACY_IO);
-	udelay(300);
+	if (likely(intel_arch_support_vga_pm(display))) {
+		outb(0x01, VGA_SEQ_I);
+		sr1 = inb(VGA_SEQ_D);
+		outb(sr1 | VGA_SR01_SCREEN_OFF, VGA_SEQ_D);
+		vga_put(pdev, VGA_RSRC_LEGACY_IO);
+		udelay(300);
+	}
 
 	intel_de_write(display, vga_reg, VGA_DISP_DISABLE);
 	intel_de_posting_read(display, vga_reg);
@@ -91,8 +102,10 @@ void intel_vga_reset_io_mem(struct intel_display *display)
 	 * and error messages.
 	 */
 	vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
-	outb(inb(VGA_MIS_R), VGA_MIS_W);
-	vga_put(pdev, VGA_RSRC_LEGACY_IO);
+	if (likely(intel_arch_support_vga_pm(display))) {
+		outb(inb(VGA_MIS_R), VGA_MIS_W);
+		vga_put(pdev, VGA_RSRC_LEGACY_IO);
+	}
 }
 
 int intel_vga_register(struct intel_display *display)
-- 
2.34.1

Re: [PATCH v2] i915: Support Intel GPU porting on any non-x86 system.

Posted by Simon Richter 6 days, 11 hours ago

Hi,

On 11/24/25 3:56 PM, zhangzhijie wrote:

 > inb/outb speccial wire not support on other ARCH.
 > Should detect whether arch platform support or not.

These platforms still have inb/outb definitions though that do 
something. https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/1824 
has a few cases where this goes wrong.

There are also a few Intel chipsets that also allow MMIO accesses to VGA 
registers. Whether there is an intersection of chipsets that allow it, 
and chipsets where that is useful is another question though.

 >   	vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);

This function really should check if inb/outb will address pdev after 
the function returns, and return an error if not, then the accesses here 
and in the power well code could be conditional on this, and it would be 
generic.

For this, the vgaarb code would need to check that the bridges on the 
way up have accepted the VGA forwarding bit (i.e. read back after 
writing it), and there probably should be a mechanism for the bridge 
between PCI(e) and the platform bus, so systems with multiple PCI(e) 
domains can also be handled.

E.g. on POWER, there are global variables "isa_io_base" and 
"pci_io_base", and it is assumed that these point to a place that 
generates IO accesses from MMIO, but this assumption is not actually 
verified, and I can pretty much guarantee that it's wrong if the PCIe 
bus in question is not even attached to the first socket.

Just checking that upstream bridges accept the VGA bit while vgaarb 
switches decoding over would solve all of my problems though, because 
the root bridge doesn't, so that would be a quick way of inhibiting 
these accesses. It might work for you too, just try setting the bit 
manually with setpci and check if it sticks.

    Simon

Re: [PATCH v2] i915: Support Intel GPU porting on any non-x86 system.

Posted by Jani Nikula 1 week ago

On Mon, 24 Nov 2025, zhangzhijie <zhangzhijie@bosc.ac.cn> wrote:
> inb/outb speccial wire not support on other ARCH.
> Should detect whether arch platform support or not.

Seems to me it inb/outb depend on CONFIG_HAS_IOPORT. Which arch are you
talking about specifically?

> Signed-off-by: zhangzhijie <zhangzhijie@bosc.ac.cn>
> ---
>  drivers/gpu/drm/i915/display/intel_vga.c | 27 ++++++++++++++++++------
>  1 file changed, 20 insertions(+), 7 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_vga.c b/drivers/gpu/drm/i915/display/intel_vga.c
> index 6e125564db34..d5d6c4ba6434 100644
> --- a/drivers/gpu/drm/i915/display/intel_vga.c
> +++ b/drivers/gpu/drm/i915/display/intel_vga.c
> @@ -41,6 +41,15 @@ static bool has_vga_pipe_sel(struct intel_display *display)
>  	return DISPLAY_VER(display) < 7;
>  }
>  
> +static bool intel_arch_support_vga_pm(struct intel_display *display)
> +{
> +#if defined(CONFIG_X86) || defined(CONFIG_X86_64)
> +	return true;
> +#else
> +	return false;
> +#endif
> +}
> +
>  /* Disable the VGA plane that we never use */
>  void intel_vga_disable(struct intel_display *display)
>  {
> @@ -66,11 +75,13 @@ void intel_vga_disable(struct intel_display *display)
>  
>  	/* WaEnableVGAAccessThroughIOPort:ctg,elk,ilk,snb,ivb,vlv,hsw */
>  	vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
> -	outb(0x01, VGA_SEQ_I);
> -	sr1 = inb(VGA_SEQ_D);
> -	outb(sr1 | VGA_SR01_SCREEN_OFF, VGA_SEQ_D);
> -	vga_put(pdev, VGA_RSRC_LEGACY_IO);
> -	udelay(300);
> +	if (likely(intel_arch_support_vga_pm(display))) {

As a rule of thumb, please do not use likely/unlikely, at all.

Here, intel_arch_support_vga_pm() is evaluated at build time, and the
likely has no functional value, only dubious documentation value.

> +		outb(0x01, VGA_SEQ_I);
> +		sr1 = inb(VGA_SEQ_D);
> +		outb(sr1 | VGA_SR01_SCREEN_OFF, VGA_SEQ_D);
> +		vga_put(pdev, VGA_RSRC_LEGACY_IO);

Both or neither get/put need to be inside the branch. Now you only get,
never put, on non-x86.

BR,
Jani.

> +		udelay(300);
> +	}
>  
>  	intel_de_write(display, vga_reg, VGA_DISP_DISABLE);
>  	intel_de_posting_read(display, vga_reg);
> @@ -91,8 +102,10 @@ void intel_vga_reset_io_mem(struct intel_display *display)
>  	 * and error messages.
>  	 */
>  	vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
> -	outb(inb(VGA_MIS_R), VGA_MIS_W);
> -	vga_put(pdev, VGA_RSRC_LEGACY_IO);
> +	if (likely(intel_arch_support_vga_pm(display))) {
> +		outb(inb(VGA_MIS_R), VGA_MIS_W);
> +		vga_put(pdev, VGA_RSRC_LEGACY_IO);
> +	}
>  }
>  
>  int intel_vga_register(struct intel_display *display)

-- 
Jani Nikula, Intel

Re: [PATCH v2] i915: Support Intel GPU porting on any non-x86 system.

Posted by ZhangZhiJie 6 days, 14 hours ago


On 2025/11/24 20:24, Jani Nikula wrote:
> On Mon, 24 Nov 2025, zhangzhijie <zhangzhijie@bosc.ac.cn> wrote:
>> inb/outb speccial wire not support on other ARCH.
>> Should detect whether arch platform support or not.
> 
> Seems to me it inb/outb depend on CONFIG_HAS_IOPORT. Which arch are you
> talking about specifically?

riscv , has CONFIG_HAS_IOPORT.  is using for serial, GPIO., etc.

Thanks.
BR.
ZhiJie
> 
>> Signed-off-by: zhangzhijie <zhangzhijie@bosc.ac.cn>
>> ---
>>   drivers/gpu/drm/i915/display/intel_vga.c | 27 ++++++++++++++++++------
>>   1 file changed, 20 insertions(+), 7 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/i915/display/intel_vga.c b/drivers/gpu/drm/i915/display/intel_vga.c
>> index 6e125564db34..d5d6c4ba6434 100644
>> --- a/drivers/gpu/drm/i915/display/intel_vga.c
>> +++ b/drivers/gpu/drm/i915/display/intel_vga.c
>> @@ -41,6 +41,15 @@ static bool has_vga_pipe_sel(struct intel_display *display)
>>   	return DISPLAY_VER(display) < 7;
>>   }
>>   
>> +static bool intel_arch_support_vga_pm(struct intel_display *display)
>> +{
>> +#if defined(CONFIG_X86) || defined(CONFIG_X86_64)
>> +	return true;
>> +#else
>> +	return false;
>> +#endif
>> +}
>> +
>>   /* Disable the VGA plane that we never use */
>>   void intel_vga_disable(struct intel_display *display)
>>   {
>> @@ -66,11 +75,13 @@ void intel_vga_disable(struct intel_display *display)
>>   
>>   	/* WaEnableVGAAccessThroughIOPort:ctg,elk,ilk,snb,ivb,vlv,hsw */
>>   	vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
>> -	outb(0x01, VGA_SEQ_I);
>> -	sr1 = inb(VGA_SEQ_D);
>> -	outb(sr1 | VGA_SR01_SCREEN_OFF, VGA_SEQ_D);
>> -	vga_put(pdev, VGA_RSRC_LEGACY_IO);
>> -	udelay(300);
>> +	if (likely(intel_arch_support_vga_pm(display))) {
> 
> As a rule of thumb, please do not use likely/unlikely, at all.
> 
> Here, intel_arch_support_vga_pm() is evaluated at build time, and the
> likely has no functional value, only dubious documentation value.
> 
>> +		outb(0x01, VGA_SEQ_I);
>> +		sr1 = inb(VGA_SEQ_D);
>> +		outb(sr1 | VGA_SR01_SCREEN_OFF, VGA_SEQ_D);
>> +		vga_put(pdev, VGA_RSRC_LEGACY_IO);
> 
> Both or neither get/put need to be inside the branch. Now you only get,
> never put, on non-x86.
> 
> BR,
> Jani.
> 
>> +		udelay(300);
>> +	}
>>   
>>   	intel_de_write(display, vga_reg, VGA_DISP_DISABLE);
>>   	intel_de_posting_read(display, vga_reg);
>> @@ -91,8 +102,10 @@ void intel_vga_reset_io_mem(struct intel_display *display)
>>   	 * and error messages.
>>   	 */
>>   	vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
>> -	outb(inb(VGA_MIS_R), VGA_MIS_W);
>> -	vga_put(pdev, VGA_RSRC_LEGACY_IO);
>> +	if (likely(intel_arch_support_vga_pm(display))) {
>> +		outb(inb(VGA_MIS_R), VGA_MIS_W);
>> +		vga_put(pdev, VGA_RSRC_LEGACY_IO);
>> +	}
>>   }
>>   
>>   int intel_vga_register(struct intel_display *display)
>

Re: [PATCH v2] i915: Support Intel GPU porting on any non-x86 system.

Posted by Jani Nikula 6 days, 14 hours ago

On Tue, 25 Nov 2025, ZhangZhiJie <zhangzhijie@bosc.ac.cn> wrote:
> On 2025/11/24 20:24, Jani Nikula wrote:
>> On Mon, 24 Nov 2025, zhangzhijie <zhangzhijie@bosc.ac.cn> wrote:
>>> inb/outb speccial wire not support on other ARCH.
>>> Should detect whether arch platform support or not.
>> 
>> Seems to me it inb/outb depend on CONFIG_HAS_IOPORT. Which arch are you
>> talking about specifically?
>
> riscv , has CONFIG_HAS_IOPORT.  is using for serial, GPIO., etc.

What is the actual failure mode you see?

Does it build?

BR,
Jani.



>
> Thanks.
> BR.
> ZhiJie
>> 
>>> Signed-off-by: zhangzhijie <zhangzhijie@bosc.ac.cn>
>>> ---
>>>   drivers/gpu/drm/i915/display/intel_vga.c | 27 ++++++++++++++++++------
>>>   1 file changed, 20 insertions(+), 7 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/i915/display/intel_vga.c b/drivers/gpu/drm/i915/display/intel_vga.c
>>> index 6e125564db34..d5d6c4ba6434 100644
>>> --- a/drivers/gpu/drm/i915/display/intel_vga.c
>>> +++ b/drivers/gpu/drm/i915/display/intel_vga.c
>>> @@ -41,6 +41,15 @@ static bool has_vga_pipe_sel(struct intel_display *display)
>>>   	return DISPLAY_VER(display) < 7;
>>>   }
>>>   
>>> +static bool intel_arch_support_vga_pm(struct intel_display *display)
>>> +{
>>> +#if defined(CONFIG_X86) || defined(CONFIG_X86_64)
>>> +	return true;
>>> +#else
>>> +	return false;
>>> +#endif
>>> +}
>>> +
>>>   /* Disable the VGA plane that we never use */
>>>   void intel_vga_disable(struct intel_display *display)
>>>   {
>>> @@ -66,11 +75,13 @@ void intel_vga_disable(struct intel_display *display)
>>>   
>>>   	/* WaEnableVGAAccessThroughIOPort:ctg,elk,ilk,snb,ivb,vlv,hsw */
>>>   	vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
>>> -	outb(0x01, VGA_SEQ_I);
>>> -	sr1 = inb(VGA_SEQ_D);
>>> -	outb(sr1 | VGA_SR01_SCREEN_OFF, VGA_SEQ_D);
>>> -	vga_put(pdev, VGA_RSRC_LEGACY_IO);
>>> -	udelay(300);
>>> +	if (likely(intel_arch_support_vga_pm(display))) {
>> 
>> As a rule of thumb, please do not use likely/unlikely, at all.
>> 
>> Here, intel_arch_support_vga_pm() is evaluated at build time, and the
>> likely has no functional value, only dubious documentation value.
>> 
>>> +		outb(0x01, VGA_SEQ_I);
>>> +		sr1 = inb(VGA_SEQ_D);
>>> +		outb(sr1 | VGA_SR01_SCREEN_OFF, VGA_SEQ_D);
>>> +		vga_put(pdev, VGA_RSRC_LEGACY_IO);
>> 
>> Both or neither get/put need to be inside the branch. Now you only get,
>> never put, on non-x86.
>> 
>> BR,
>> Jani.
>> 
>>> +		udelay(300);
>>> +	}
>>>   
>>>   	intel_de_write(display, vga_reg, VGA_DISP_DISABLE);
>>>   	intel_de_posting_read(display, vga_reg);
>>> @@ -91,8 +102,10 @@ void intel_vga_reset_io_mem(struct intel_display *display)
>>>   	 * and error messages.
>>>   	 */
>>>   	vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
>>> -	outb(inb(VGA_MIS_R), VGA_MIS_W);
>>> -	vga_put(pdev, VGA_RSRC_LEGACY_IO);
>>> +	if (likely(intel_arch_support_vga_pm(display))) {
>>> +		outb(inb(VGA_MIS_R), VGA_MIS_W);
>>> +		vga_put(pdev, VGA_RSRC_LEGACY_IO);
>>> +	}
>>>   }
>>>   
>>>   int intel_vga_register(struct intel_display *display)
>> 
>

-- 
Jani Nikula, Intel

Re: [PATCH v2] i915: Support Intel GPU porting on any non-x86 system.

Posted by ZhangZhiJie 6 days, 14 hours ago


On 2025/11/25 18:13, Jani Nikula wrote:
> On Tue, 25 Nov 2025, ZhangZhiJie <zhangzhijie@bosc.ac.cn> wrote:
>> On 2025/11/24 20:24, Jani Nikula wrote:
>>> On Mon, 24 Nov 2025, zhangzhijie <zhangzhijie@bosc.ac.cn> wrote:
>>>> inb/outb speccial wire not support on other ARCH.
>>>> Should detect whether arch platform support or not.
>>>
>>> Seems to me it inb/outb depend on CONFIG_HAS_IOPORT. Which arch are you
>>> talking about specifically?
>>
>> riscv , has CONFIG_HAS_IOPORT.  is using for serial, GPIO., etc.
> 
> What is the actual failure mode you see?
> 
> Does it build?
> 
> BR,
> Jani.
> 
Yes, i compiled Xe driver, and it's occured load access fault when not 
disable  VGA_SEQ_*.

can found this link 
https://github.com/geerlingguy/raspberry-pi-pcie-devices/issues/510#issuecomment-3383284831

Thanks.
BR.
ZhiJie
> 
> 
>>
>> Thanks.
>> BR.
>> ZhiJie
>>>
>>>> Signed-off-by: zhangzhijie <zhangzhijie@bosc.ac.cn>
>>>> ---
>>>>    drivers/gpu/drm/i915/display/intel_vga.c | 27 ++++++++++++++++++------
>>>>    1 file changed, 20 insertions(+), 7 deletions(-)
>>>>
>>>> diff --git a/drivers/gpu/drm/i915/display/intel_vga.c b/drivers/gpu/drm/i915/display/intel_vga.c
>>>> index 6e125564db34..d5d6c4ba6434 100644
>>>> --- a/drivers/gpu/drm/i915/display/intel_vga.c
>>>> +++ b/drivers/gpu/drm/i915/display/intel_vga.c
>>>> @@ -41,6 +41,15 @@ static bool has_vga_pipe_sel(struct intel_display *display)
>>>>    	return DISPLAY_VER(display) < 7;
>>>>    }
>>>>    
>>>> +static bool intel_arch_support_vga_pm(struct intel_display *display)
>>>> +{
>>>> +#if defined(CONFIG_X86) || defined(CONFIG_X86_64)
>>>> +	return true;
>>>> +#else
>>>> +	return false;
>>>> +#endif
>>>> +}
>>>> +
>>>>    /* Disable the VGA plane that we never use */
>>>>    void intel_vga_disable(struct intel_display *display)
>>>>    {
>>>> @@ -66,11 +75,13 @@ void intel_vga_disable(struct intel_display *display)
>>>>    
>>>>    	/* WaEnableVGAAccessThroughIOPort:ctg,elk,ilk,snb,ivb,vlv,hsw */
>>>>    	vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
>>>> -	outb(0x01, VGA_SEQ_I);
>>>> -	sr1 = inb(VGA_SEQ_D);
>>>> -	outb(sr1 | VGA_SR01_SCREEN_OFF, VGA_SEQ_D);
>>>> -	vga_put(pdev, VGA_RSRC_LEGACY_IO);
>>>> -	udelay(300);
>>>> +	if (likely(intel_arch_support_vga_pm(display))) {
>>>
>>> As a rule of thumb, please do not use likely/unlikely, at all.
>>>
>>> Here, intel_arch_support_vga_pm() is evaluated at build time, and the
>>> likely has no functional value, only dubious documentation value.
>>>
>>>> +		outb(0x01, VGA_SEQ_I);
>>>> +		sr1 = inb(VGA_SEQ_D);
>>>> +		outb(sr1 | VGA_SR01_SCREEN_OFF, VGA_SEQ_D);
>>>> +		vga_put(pdev, VGA_RSRC_LEGACY_IO);
>>>
>>> Both or neither get/put need to be inside the branch. Now you only get,
>>> never put, on non-x86.
>>>
>>> BR,
>>> Jani.
>>>
>>>> +		udelay(300);
>>>> +	}
>>>>    
>>>>    	intel_de_write(display, vga_reg, VGA_DISP_DISABLE);
>>>>    	intel_de_posting_read(display, vga_reg);
>>>> @@ -91,8 +102,10 @@ void intel_vga_reset_io_mem(struct intel_display *display)
>>>>    	 * and error messages.
>>>>    	 */
>>>>    	vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
>>>> -	outb(inb(VGA_MIS_R), VGA_MIS_W);
>>>> -	vga_put(pdev, VGA_RSRC_LEGACY_IO);
>>>> +	if (likely(intel_arch_support_vga_pm(display))) {
>>>> +		outb(inb(VGA_MIS_R), VGA_MIS_W);
>>>> +		vga_put(pdev, VGA_RSRC_LEGACY_IO);
>>>> +	}
>>>>    }
>>>>    
>>>>    int intel_vga_register(struct intel_display *display)
>>>
>>
>

Re: [PATCH v2] i915: Support Intel GPU porting on any non-x86 system.

Posted by Lucas De Marchi 6 days, 2 hours ago

On Tue, Nov 25, 2025 at 06:20:00PM +0800, ZhangZhiJie wrote:
>
>
>On 2025/11/25 18:13, Jani Nikula wrote:
>>On Tue, 25 Nov 2025, ZhangZhiJie <zhangzhijie@bosc.ac.cn> wrote:
>>>On 2025/11/24 20:24, Jani Nikula wrote:
>>>>On Mon, 24 Nov 2025, zhangzhijie <zhangzhijie@bosc.ac.cn> wrote:
>>>>>inb/outb speccial wire not support on other ARCH.
>>>>>Should detect whether arch platform support or not.
>>>>
>>>>Seems to me it inb/outb depend on CONFIG_HAS_IOPORT. Which arch are you
>>>>talking about specifically?
>>>
>>>riscv , has CONFIG_HAS_IOPORT.  is using for serial, GPIO., etc.
>>
>>What is the actual failure mode you see?
>>
>>Does it build?
>>
>>BR,
>>Jani.
>>
>Yes, i compiled Xe driver, and it's occured load access fault when not 
>disable  VGA_SEQ_*.
>
>can found this link https://github.com/geerlingguy/raspberry-pi-pcie-devices/issues/510#issuecomment-3383284831

and this is what I tested with rpi:
https://lore.kernel.org/intel-xe/20251119-ioport-v1-1-ec43f1e12c49@intel.com/

not sure why you are changing the intel_vga_disable() function: out of
reset that bit is disabled and the function does nothing:

	tmp = intel_de_read(display, vga_reg);
	if (tmp & VGA_DISP_DISABLE)
		return;

If there's no VGA, no bios, then there isn't anything enabling that and
we don't need to disable it.

I have a patch that moves the vga access to a separate function,
intel_vga_set_screen_off(), but that's mostly for clarity, not to fix
anything. If later we want to add an ifdef then we'd probably have to
implement the alternative.

Also note that not having CONFIG_HAS_IOPORT doesn't mean inb()/outb()
are not implemented. See arch/{arm,powerpc,sparc}/include/asm/io.h
for a few variants.

Lucas De Marchi

Re: [PATCH v2] i915: Support Intel GPU porting on any non-x86 system.

Posted by ZhangZhiJie 5 days, 23 hours ago


On 2025/11/26 05:54, Lucas De Marchi wrote:
> On Tue, Nov 25, 2025 at 06:20:00PM +0800, ZhangZhiJie wrote:
>>
>>
>> On 2025/11/25 18:13, Jani Nikula wrote:
>>> On Tue, 25 Nov 2025, ZhangZhiJie <zhangzhijie@bosc.ac.cn> wrote:
>>>> On 2025/11/24 20:24, Jani Nikula wrote:
>>>>> On Mon, 24 Nov 2025, zhangzhijie <zhangzhijie@bosc.ac.cn> wrote:
>>>>>> inb/outb speccial wire not support on other ARCH.
>>>>>> Should detect whether arch platform support or not.
>>>>>
>>>>> Seems to me it inb/outb depend on CONFIG_HAS_IOPORT. Which arch are 
>>>>> you
>>>>> talking about specifically?
>>>>
>>>> riscv , has CONFIG_HAS_IOPORT.  is using for serial, GPIO., etc.
>>>
>>> What is the actual failure mode you see?
>>>
>>> Does it build?
>>>
>>> BR,
>>> Jani.
>>>
>> Yes, i compiled Xe driver, and it's occured load access fault when not 
>> disable  VGA_SEQ_*.
>>
>> can found this link https://github.com/geerlingguy/raspberry-pi-pcie- 
>> devices/issues/510#issuecomment-3383284831
> 
> and this is what I tested with rpi:
> https://lore.kernel.org/intel-xe/20251119-ioport-v1-1- 
> ec43f1e12c49@intel.com/
> 
Yep. agree like this patch.

In my mind. if machine not implement VGA_SEQ_* ioport.
should avoid access ioports about VGA_SEQ_* .

Can you give  log for bootting on non-x86 arch?

> not sure why you are changing the intel_vga_disable() function: out of
> reset that bit is disabled and the function does nothing:
> 
>      tmp = intel_de_read(display, vga_reg);
>      if (tmp & VGA_DISP_DISABLE)
>          return;
> 
> If there's no VGA, no bios, then there isn't anything enabling that and
> we don't need to disable it.
> 
> I have a patch that moves the vga access to a separate function,
> intel_vga_set_screen_off(), but that's mostly for clarity, not to fix
> anything. If later we want to add an ifdef then we'd probably have to
> implement the alternative.
> 
> Also note that not having CONFIG_HAS_IOPORT doesn't mean inb()/outb()
> are not implemented. See arch/{arm,powerpc,sparc}/include/asm/io.h
> for a few variants.
> 
> Lucas De Marchi

BR.
ZhiJie