All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Christian König" <ckoenig.leichtzumerken@gmail.com>
To: chong li <chongli2@amd.com>, amd-gfx@lists.freedesktop.org
Cc: HaiJun.Chang@amd.com, Emily.Deng@amd.com,
	Cursor <cursoragent@cursor.com>
Subject: Re: [PATCH 1/2] drm/amdgpu: read FB through BAR0 when aperture is unavailable
Date: Mon, 6 Jul 2026 13:36:14 +0200	[thread overview]
Message-ID: <48d65314-02fd-401e-b2a6-1b09c6f8da5f@gmail.com> (raw)
In-Reply-To: <20260701055355.20478-1-chongli2@amd.com>

On 7/1/26 07:53, chong li wrote:
> Allow early VRAM reads to fall back to a temporary BAR0 mapping
> when the normal aperture mapping has not been established yet.

Clear NAK.

You are just trying to fix a symptom of incorrect initialization order here.

Regards,
Christian.

> 
> Signed-off-by: chong li <chongli2@amd.com>
> Co-authored-by: Cursor <cursoragent@cursor.com>
> ---
>  drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 43 +++++++++++++++++++++-
>  1 file changed, 42 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
> index 5442a1fc1c37..610d82b79de3 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
> @@ -743,6 +743,44 @@ void amdgpu_device_mm_access(struct amdgpu_device *adev, loff_t pos,
>  	drm_dev_exit(idx);
>  }
>  
> +static int amdgpu_device_read_fb_via_bar0(struct amdgpu_device *adev,
> +					  u64 offset, void *buf, size_t size)
> +{
> +	resource_size_t bar_start, bar_size, map_base;
> +	void __iomem *vram;
> +	size_t map_offset, map_size;
> +	unsigned long flags;
> +	u64 end;
> +
> +	if (!buf || !size)
> +		return -EINVAL;
> +
> +	flags = pci_resource_flags(adev->pdev, 0);
> +	if ((flags & IORESOURCE_UNSET) || !(flags & IORESOURCE_MEM))
> +		return -EINVAL;
> +
> +	bar_size = pci_resource_len(adev->pdev, 0);
> +	if (!bar_size)
> +		return -ENODEV;
> +
> +	if (check_add_overflow(offset, size, &end) || end > bar_size)
> +		return -EINVAL;
> +
> +	bar_start = pci_resource_start(adev->pdev, 0);
> +	map_offset = offset_in_page(offset);
> +	map_base = bar_start + (offset & PAGE_MASK);
> +	map_size = PAGE_ALIGN(map_offset + size);
> +
> +	vram = ioremap_wc(map_base, map_size);
> +	if (!vram)
> +		return -ENOMEM;
> +
> +	memcpy_fromio(buf, (u8 __iomem *)vram + map_offset, size);
> +	iounmap(vram);
> +
> +	return 0;
> +}
> +
>  /**
>   * amdgpu_device_aper_access - access vram by vram aperture
>   *
> @@ -762,8 +800,11 @@ size_t amdgpu_device_aper_access(struct amdgpu_device *adev, loff_t pos,
>  	size_t count = 0;
>  	uint64_t last;
>  
> -	if (!adev->mman.aper_base_kaddr)
> +	if (!adev->mman.aper_base_kaddr) {
> +		if (!write && !amdgpu_device_read_fb_via_bar0(adev, pos, buf, size))
> +			return size;
>  		return 0;
> +	}
>  
>  	last = min(pos + size, adev->gmc.visible_vram_size);
>  	if (last > pos) {


      parent reply	other threads:[~2026-07-06 11:36 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-07-01  5:53 [PATCH 1/2] drm/amdgpu: read FB through BAR0 when aperture is unavailable chong li
2026-07-01  5:53 ` [PATCH 2/2] drm/amdgpu: improve the amdgpu device init progress in sriov mode chong li
2026-07-06 11:39   ` Christian König
2026-07-06 11:36 ` Christian König [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=48d65314-02fd-401e-b2a6-1b09c6f8da5f@gmail.com \
    --to=ckoenig.leichtzumerken@gmail.com \
    --cc=Emily.Deng@amd.com \
    --cc=HaiJun.Chang@amd.com \
    --cc=amd-gfx@lists.freedesktop.org \
    --cc=chongli2@amd.com \
    --cc=cursoragent@cursor.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.