* ✗ LGCI.VerificationFailed: failure for Support Intel Xe GPU dirver Porting on RISC-V Architecture
2025-07-15 6:18 [PATCH v1] Support Intel Xe GPU dirver Porting on RISC-V Architecture zhangzhijie
@ 2025-07-15 20:47 ` Patchwork
2025-10-30 1:50 ` [PATCH v1] " Jeff Geerling
` (3 subsequent siblings)
4 siblings, 0 replies; 7+ messages in thread
From: Patchwork @ 2025-07-15 20:47 UTC (permalink / raw)
To: zhangzhijie; +Cc: intel-gfx
== Series Details ==
Series: Support Intel Xe GPU dirver Porting on RISC-V Architecture
URL : https://patchwork.freedesktop.org/series/151666/
State : failure
== Summary ==
Address 'zhangzhijie@bosc.ac.cn' is not on the allowlist, which prevents CI from being triggered for this patch.
If you want Intel GFX CI to accept this address, please contact the script maintainers at i915-ci-infra@lists.freedesktop.org.
Exception occurred during validation, bailing out!
^ permalink raw reply [flat|nested] 7+ messages in thread* Re: [PATCH v1] Support Intel Xe GPU dirver Porting on RISC-V Architecture
2025-07-15 6:18 [PATCH v1] Support Intel Xe GPU dirver Porting on RISC-V Architecture zhangzhijie
2025-07-15 20:47 ` ✗ LGCI.VerificationFailed: failure for " Patchwork
@ 2025-10-30 1:50 ` Jeff Geerling
2025-10-30 1:51 ` Jeff Geerling
` (2 subsequent siblings)
4 siblings, 0 replies; 7+ messages in thread
From: Jeff Geerling @ 2025-10-30 1:50 UTC (permalink / raw)
To: zhangzhijie
Cc: wangran, zhangjian, daniel, jani.nikula, rodrigo.vivi,
joonas.lahtinen, tursulin, airlied, intel-gfx, intel-xe,
dri-devel, linux-kernel
[-- Attachment #1: Type: text/plain, Size: 2582 bytes --]
+1 for this patch, as it would enable Xe on any non-x86 system.
I've successfully tested this change on Ampere and Broadcom (Raspberry Pi) systems.
We've been using the flag `CONFIG_VGA_CONSOLE` instead of `CONFIG_X86` but either should achieve the same goal. See: https://github.com/6by9/linux/commit/6bd4cfe79b5111986dd11a5c6e48d4a963fd7740
With some OSes a later Mesa version needs to be compiled to support newer Xe/Xe2 cards, however I've successfully tested a number of Intel dGPUs at this point. Sometimes setting force probe to '*' (or the particular GPU ID) is required, but otherwise stability is good.
For example, the A750: https://github.com/geerlingguy/raspberry-pi-pcie-devices/issues/510#issuecomment-3383284831
> On Jul 15, 2025, at 1:18 AM, zhangzhijie <zhangzhijie@bosc.ac.cn> wrote:
>
> 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 | 4 ++++
> 1 file changed, 4 insertions(+)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_vga.c b/drivers/gpu/drm/i915/display/intel_vga.c
> index 4b98833bfa8c..5e48e3282670 100644
> --- a/drivers/gpu/drm/i915/display/intel_vga.c
> +++ b/drivers/gpu/drm/i915/display/intel_vga.c
> @@ -33,6 +33,7 @@ void intel_vga_disable(struct drm_i915_private *dev_priv)
> if (intel_de_read(dev_priv, vga_reg) & VGA_DISP_DISABLE)
> return;
>
> +#if defined(CONFIG_X86) || defined(CONFIG_X86_64)
> /* WaEnableVGAAccessThroughIOPort:ctg,elk,ilk,snb,ivb,vlv,hsw */
> vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
> outb(0x01, VGA_SEQ_I);
> @@ -40,6 +41,7 @@ void intel_vga_disable(struct drm_i915_private *dev_priv)
> outb(sr1 | VGA_SR01_SCREEN_OFF, VGA_SEQ_D);
> vga_put(pdev, VGA_RSRC_LEGACY_IO);
> udelay(300);
> +#endif
>
> intel_de_write(dev_priv, vga_reg, VGA_DISP_DISABLE);
> intel_de_posting_read(dev_priv, vga_reg);
> @@ -80,6 +82,7 @@ void intel_vga_redisable(struct drm_i915_private *i915)
>
> void intel_vga_reset_io_mem(struct drm_i915_private *i915)
> {
> +#if defined(CONFIG_X86) || defined(CONFIG_X86_64)
> struct pci_dev *pdev = to_pci_dev(i915->drm.dev);
>
> /*
> @@ -95,6 +98,7 @@ void intel_vga_reset_io_mem(struct drm_i915_private *i915)
> vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
> outb(inb(VGA_MIS_R), VGA_MIS_W);
> vga_put(pdev, VGA_RSRC_LEGACY_IO);
> +#endif
> }
>
> int intel_vga_register(struct drm_i915_private *i915)
> --
> 2.34.1
>
>
[-- Attachment #2: Type: text/html, Size: 4315 bytes --]
^ permalink raw reply [flat|nested] 7+ messages in thread* Re: [PATCH v1] Support Intel Xe GPU dirver Porting on RISC-V Architecture
2025-07-15 6:18 [PATCH v1] Support Intel Xe GPU dirver Porting on RISC-V Architecture zhangzhijie
2025-07-15 20:47 ` ✗ LGCI.VerificationFailed: failure for " Patchwork
2025-10-30 1:50 ` [PATCH v1] " Jeff Geerling
@ 2025-10-30 1:51 ` Jeff Geerling
2025-10-30 1:55 ` Jeff Geerling
2025-10-30 13:34 ` Jani Nikula
4 siblings, 0 replies; 7+ messages in thread
From: Jeff Geerling @ 2025-10-30 1:51 UTC (permalink / raw)
To: zhangzhijie
Cc: wangran, zhangjian, daniel, jani.nikula, rodrigo.vivi,
joonas.lahtinen, tursulin, airlied, intel-gfx, intel-xe,
dri-devel, linux-kernel
[-- Attachment #1: Type: text/plain, Size: 2582 bytes --]
+1 for this patch, as it would enable Xe on any non-x86 system.
I've successfully tested this change on Ampere and Broadcom (Raspberry Pi) systems.
We've been using the flag `CONFIG_VGA_CONSOLE` instead of `CONFIG_X86` but either should achieve the same goal. See: https://github.com/6by9/linux/commit/6bd4cfe79b5111986dd11a5c6e48d4a963fd7740
With some OSes a later Mesa version needs to be compiled to support newer Xe/Xe2 cards, however I've successfully tested a number of Intel dGPUs at this point. Sometimes setting force probe to '*' (or the particular GPU ID) is required, but otherwise stability is good.
For example, the A750: https://github.com/geerlingguy/raspberry-pi-pcie-devices/issues/510#issuecomment-3383284831
> On Jul 15, 2025, at 1:18 AM, zhangzhijie <zhangzhijie@bosc.ac.cn> wrote:
>
> 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 | 4 ++++
> 1 file changed, 4 insertions(+)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_vga.c b/drivers/gpu/drm/i915/display/intel_vga.c
> index 4b98833bfa8c..5e48e3282670 100644
> --- a/drivers/gpu/drm/i915/display/intel_vga.c
> +++ b/drivers/gpu/drm/i915/display/intel_vga.c
> @@ -33,6 +33,7 @@ void intel_vga_disable(struct drm_i915_private *dev_priv)
> if (intel_de_read(dev_priv, vga_reg) & VGA_DISP_DISABLE)
> return;
>
> +#if defined(CONFIG_X86) || defined(CONFIG_X86_64)
> /* WaEnableVGAAccessThroughIOPort:ctg,elk,ilk,snb,ivb,vlv,hsw */
> vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
> outb(0x01, VGA_SEQ_I);
> @@ -40,6 +41,7 @@ void intel_vga_disable(struct drm_i915_private *dev_priv)
> outb(sr1 | VGA_SR01_SCREEN_OFF, VGA_SEQ_D);
> vga_put(pdev, VGA_RSRC_LEGACY_IO);
> udelay(300);
> +#endif
>
> intel_de_write(dev_priv, vga_reg, VGA_DISP_DISABLE);
> intel_de_posting_read(dev_priv, vga_reg);
> @@ -80,6 +82,7 @@ void intel_vga_redisable(struct drm_i915_private *i915)
>
> void intel_vga_reset_io_mem(struct drm_i915_private *i915)
> {
> +#if defined(CONFIG_X86) || defined(CONFIG_X86_64)
> struct pci_dev *pdev = to_pci_dev(i915->drm.dev);
>
> /*
> @@ -95,6 +98,7 @@ void intel_vga_reset_io_mem(struct drm_i915_private *i915)
> vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
> outb(inb(VGA_MIS_R), VGA_MIS_W);
> vga_put(pdev, VGA_RSRC_LEGACY_IO);
> +#endif
> }
>
> int intel_vga_register(struct drm_i915_private *i915)
> --
> 2.34.1
>
>
[-- Attachment #2: Type: text/html, Size: 4304 bytes --]
^ permalink raw reply [flat|nested] 7+ messages in thread* Re: [PATCH v1] Support Intel Xe GPU dirver Porting on RISC-V Architecture
2025-07-15 6:18 [PATCH v1] Support Intel Xe GPU dirver Porting on RISC-V Architecture zhangzhijie
` (2 preceding siblings ...)
2025-10-30 1:51 ` Jeff Geerling
@ 2025-10-30 1:55 ` Jeff Geerling
2025-10-30 3:28 ` ZhangZhiJie
2025-10-30 13:34 ` Jani Nikula
4 siblings, 1 reply; 7+ messages in thread
From: Jeff Geerling @ 2025-10-30 1:55 UTC (permalink / raw)
To: zhangzhijie
Cc: wangran, zhangjian, daniel, jani.nikula, rodrigo.vivi,
joonas.lahtinen, tursulin, airlied, intel-gfx, intel-xe,
dri-devel, linux-kernel
+1 for this patch, as it would enable Xe on any non-x86 system.
I've successfully tested this change on Ampere and Broadcom (Raspberry Pi) systems.
We've been using the flag `CONFIG_VGA_CONSOLE` instead of `CONFIG_X86` but either should achieve the same goal. See: https://github.com/6by9/linux/commit/6bd4cfe79b5111986dd11a5c6e48d4a963fd7740
With some OSes a later Mesa version needs to be compiled to support newer Xe/Xe2 cards, however I've successfully tested a number of Intel dGPUs at this point. Sometimes setting force probe to '*' (or the particular GPU ID) is required, but otherwise stability is good.
For example, the A750: https://github.com/geerlingguy/raspberry-pi-pcie-devices/issues/510#issuecomment-3383284831
> On Jul 15, 2025, at 1:18 AM, zhangzhijie <zhangzhijie@bosc.ac.cn> wrote:
>
> 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 | 4 ++++
> 1 file changed, 4 insertions(+)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_vga.c b/drivers/gpu/drm/i915/display/intel_vga.c
> index 4b98833bfa8c..5e48e3282670 100644
> --- a/drivers/gpu/drm/i915/display/intel_vga.c
> +++ b/drivers/gpu/drm/i915/display/intel_vga.c
> @@ -33,6 +33,7 @@ void intel_vga_disable(struct drm_i915_private *dev_priv)
> if (intel_de_read(dev_priv, vga_reg) & VGA_DISP_DISABLE)
> return;
>
> +#if defined(CONFIG_X86) || defined(CONFIG_X86_64)
> /* WaEnableVGAAccessThroughIOPort:ctg,elk,ilk,snb,ivb,vlv,hsw */
> vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
> outb(0x01, VGA_SEQ_I);
> @@ -40,6 +41,7 @@ void intel_vga_disable(struct drm_i915_private *dev_priv)
> outb(sr1 | VGA_SR01_SCREEN_OFF, VGA_SEQ_D);
> vga_put(pdev, VGA_RSRC_LEGACY_IO);
> udelay(300);
> +#endif
>
> intel_de_write(dev_priv, vga_reg, VGA_DISP_DISABLE);
> intel_de_posting_read(dev_priv, vga_reg);
> @@ -80,6 +82,7 @@ void intel_vga_redisable(struct drm_i915_private *i915)
>
> void intel_vga_reset_io_mem(struct drm_i915_private *i915)
> {
> +#if defined(CONFIG_X86) || defined(CONFIG_X86_64)
> struct pci_dev *pdev = to_pci_dev(i915->drm.dev);
>
> /*
> @@ -95,6 +98,7 @@ void intel_vga_reset_io_mem(struct drm_i915_private *i915)
> vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
> outb(inb(VGA_MIS_R), VGA_MIS_W);
> vga_put(pdev, VGA_RSRC_LEGACY_IO);
> +#endif
> }
>
> int intel_vga_register(struct drm_i915_private *i915)
> --
> 2.34.1
>
>
^ permalink raw reply [flat|nested] 7+ messages in thread* Re: [PATCH v1] Support Intel Xe GPU dirver Porting on RISC-V Architecture
2025-10-30 1:55 ` Jeff Geerling
@ 2025-10-30 3:28 ` ZhangZhiJie
0 siblings, 0 replies; 7+ messages in thread
From: ZhangZhiJie @ 2025-10-30 3:28 UTC (permalink / raw)
To: Jeff Geerling
Cc: wangran, zhangjian, daniel, jani.nikula, rodrigo.vivi,
joonas.lahtinen, tursulin, airlied, intel-gfx, intel-xe,
dri-devel, linux-kernel
On 2025/10/30 09:55, Jeff Geerling wrote:
> +1 for this patch, as it would enable Xe on any non-x86 system.
>
> I've successfully tested this change on Ampere and Broadcom (Raspberry Pi) systems.
>
> We've been using the flag `CONFIG_VGA_CONSOLE` instead of `CONFIG_X86` but either should achieve the same goal. See: https://github.com/6by9/linux/commit/6bd4cfe79b5111986dd11a5c6e48d4a963fd7740
>
> With some OSes a later Mesa version needs to be compiled to support newer Xe/Xe2 cards, however I've successfully tested a number of Intel dGPUs at this point. Sometimes setting force probe to '*' (or the particular GPU ID) is required, but otherwise stability is good.
>
> For example, the A750: https://github.com/geerlingguy/raspberry-pi-pcie-devices/issues/510#issuecomment-3383284831
>
>> On Jul 15, 2025, at 1:18 AM, zhangzhijie <zhangzhijie@bosc.ac.cn> wrote:
>>
>> 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 | 4 ++++
>> 1 file changed, 4 insertions(+)
>>
>> diff --git a/drivers/gpu/drm/i915/display/intel_vga.c b/drivers/gpu/drm/i915/display/intel_vga.c
>> index 4b98833bfa8c..5e48e3282670 100644
>> --- a/drivers/gpu/drm/i915/display/intel_vga.c
>> +++ b/drivers/gpu/drm/i915/display/intel_vga.c
>> @@ -33,6 +33,7 @@ void intel_vga_disable(struct drm_i915_private *dev_priv)
>> if (intel_de_read(dev_priv, vga_reg) & VGA_DISP_DISABLE)
>> return;
>>
>> +#if defined(CONFIG_X86) || defined(CONFIG_X86_64)
>> /* WaEnableVGAAccessThroughIOPort:ctg,elk,ilk,snb,ivb,vlv,hsw */
>> vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
>> outb(0x01, VGA_SEQ_I);
>> @@ -40,6 +41,7 @@ void intel_vga_disable(struct drm_i915_private *dev_priv)
>> outb(sr1 | VGA_SR01_SCREEN_OFF, VGA_SEQ_D);
>> vga_put(pdev, VGA_RSRC_LEGACY_IO);
>> udelay(300);
>> +#endif
>>
>> intel_de_write(dev_priv, vga_reg, VGA_DISP_DISABLE);
>> intel_de_posting_read(dev_priv, vga_reg);
>> @@ -80,6 +82,7 @@ void intel_vga_redisable(struct drm_i915_private *i915)
>>
>> void intel_vga_reset_io_mem(struct drm_i915_private *i915)
>> {
>> +#if defined(CONFIG_X86) || defined(CONFIG_X86_64)
>> struct pci_dev *pdev = to_pci_dev(i915->drm.dev);
>>
>> /*
>> @@ -95,6 +98,7 @@ void intel_vga_reset_io_mem(struct drm_i915_private *i915)
>> vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
>> outb(inb(VGA_MIS_R), VGA_MIS_W);
>> vga_put(pdev, VGA_RSRC_LEGACY_IO);
>> +#endif
>> }
>>
>> int intel_vga_register(struct drm_i915_private *i915)
>> --
>> 2.34.1
>>
>>
Congratulation! . So I think shoule be disable this ops or take other
ways to instead this code， Like None IO soc/ARCH
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH v1] Support Intel Xe GPU dirver Porting on RISC-V Architecture
2025-07-15 6:18 [PATCH v1] Support Intel Xe GPU dirver Porting on RISC-V Architecture zhangzhijie
` (3 preceding siblings ...)
2025-10-30 1:55 ` Jeff Geerling
@ 2025-10-30 13:34 ` Jani Nikula
4 siblings, 0 replies; 7+ messages in thread
From: Jani Nikula @ 2025-10-30 13:34 UTC (permalink / raw)
To: zhangzhijie, zhangzhijie, wangran, zhangjian, daniel,
rodrigo.vivi, joonas.lahtinen, tursulin, airlied, intel-gfx,
intel-xe, dri-devel, linux-kernel
On Tue, 15 Jul 2025, zhangzhijie <zhangzhijie@bosc.ac.cn> wrote:
> 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 | 4 ++++
> 1 file changed, 4 insertions(+)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_vga.c b/drivers/gpu/drm/i915/display/intel_vga.c
> index 4b98833bfa8c..5e48e3282670 100644
> --- a/drivers/gpu/drm/i915/display/intel_vga.c
> +++ b/drivers/gpu/drm/i915/display/intel_vga.c
> @@ -33,6 +33,7 @@ void intel_vga_disable(struct drm_i915_private *dev_priv)
> if (intel_de_read(dev_priv, vga_reg) & VGA_DISP_DISABLE)
> return;
>
> +#if defined(CONFIG_X86) || defined(CONFIG_X86_64)
> /* WaEnableVGAAccessThroughIOPort:ctg,elk,ilk,snb,ivb,vlv,hsw */
> vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
> outb(0x01, VGA_SEQ_I);
> @@ -40,6 +41,7 @@ void intel_vga_disable(struct drm_i915_private *dev_priv)
> outb(sr1 | VGA_SR01_SCREEN_OFF, VGA_SEQ_D);
> vga_put(pdev, VGA_RSRC_LEGACY_IO);
> udelay(300);
> +#endif
This should be abstracted to a separate function with the
#if/#else/#endif choosing a different implementation of the function
(empty stub in the else branch).
>
> intel_de_write(dev_priv, vga_reg, VGA_DISP_DISABLE);
> intel_de_posting_read(dev_priv, vga_reg);
> @@ -80,6 +82,7 @@ void intel_vga_redisable(struct drm_i915_private *i915)
>
> void intel_vga_reset_io_mem(struct drm_i915_private *i915)
> {
> +#if defined(CONFIG_X86) || defined(CONFIG_X86_64)
> struct pci_dev *pdev = to_pci_dev(i915->drm.dev);
>
> /*
> @@ -95,6 +98,7 @@ void intel_vga_reset_io_mem(struct drm_i915_private *i915)
> vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
> outb(inb(VGA_MIS_R), VGA_MIS_W);
> vga_put(pdev, VGA_RSRC_LEGACY_IO);
> +#endif
Ditto, but you have the function already.
> }
>
> int intel_vga_register(struct drm_i915_private *i915)
--
Jani Nikula, Intel
^ permalink raw reply [flat|nested] 7+ messages in thread