Intel-XE Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Matthew Brost <matthew.brost@intel.com>
To: Maarten Lankhorst <dev@lankhorst.se>
Cc: <intel-xe@lists.freedesktop.org>
Subject: Re: [PATCH v7 03/12] drm/xe: Add xe_ggtt_node_addr() to avoid dereferencing xe_ggtt_node
Date: Wed, 15 Oct 2025 14:49:12 -0700	[thread overview]
Message-ID: <aPAW2H5JAyOFXl69@lstrano-desk.jf.intel.com> (raw)
In-Reply-To: <20251015074708.1654014-17-dev@lankhorst.se>

On Wed, Oct 15, 2025 at 09:47:12AM +0200, Maarten Lankhorst wrote:
> This function makes it possible to add an offset that is applied to
> all xe_ggtt_node's, and hides the internals from all its users.
> 
> Signed-off-by: Maarten Lankhorst <dev@lankhorst.se>
> ---
>  drivers/gpu/drm/xe/xe_bo.h   |  8 +++++---
>  drivers/gpu/drm/xe/xe_ggtt.c | 11 +++++++++++
>  drivers/gpu/drm/xe/xe_ggtt.h |  2 ++
>  3 files changed, 18 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/gpu/drm/xe/xe_bo.h b/drivers/gpu/drm/xe/xe_bo.h
> index 353d607d301da..dd67a6f40a16d 100644
> --- a/drivers/gpu/drm/xe/xe_bo.h
> +++ b/drivers/gpu/drm/xe/xe_bo.h
> @@ -9,6 +9,7 @@
>  #include <drm/ttm/ttm_tt.h>
>  
>  #include "xe_bo_types.h"
> +#include "xe_ggtt.h"
>  #include "xe_macros.h"
>  #include "xe_validation.h"
>  #include "xe_vm_types.h"
> @@ -251,13 +252,14 @@ static inline u32
>  __xe_bo_ggtt_addr(struct xe_bo *bo, u8 tile_id)
>  {
>  	struct xe_ggtt_node *ggtt_node = bo->ggtt_node[tile_id];
> +	u64 offset;
>  
>  	if (XE_WARN_ON(!ggtt_node))
>  		return 0;
>  
> -	XE_WARN_ON(ggtt_node->base.size > xe_bo_size(bo));

Is there any reason this warning was dropped? It's not immediately
obvious to me why it would be.

Matt

> -	XE_WARN_ON(ggtt_node->base.start + ggtt_node->base.size > (1ull << 32));
> -	return ggtt_node->base.start;
> +	offset = xe_ggtt_node_addr(ggtt_node);
> +	XE_WARN_ON(offset + xe_bo_size(bo) > (1ull << 32));
> +	return offset;
>  }
>  
>  static inline u32
> diff --git a/drivers/gpu/drm/xe/xe_ggtt.c b/drivers/gpu/drm/xe/xe_ggtt.c
> index 42e3caa3a6d9c..05515037e2d6a 100644
> --- a/drivers/gpu/drm/xe/xe_ggtt.c
> +++ b/drivers/gpu/drm/xe/xe_ggtt.c
> @@ -1079,3 +1079,14 @@ u64 xe_ggtt_read_pte(struct xe_ggtt *ggtt, u64 offset)
>  {
>  	return ioread64(ggtt->gsm + (offset / XE_PAGE_SIZE));
>  }
> +
> +/**
> + * xe_ggtt_node_addr - Get @node offset in GGTT.
> + * @node: &xe_ggtt_node
> + *
> + * Get the GGTT offset for allocated node.
> + */
> +u64 xe_ggtt_node_addr(const struct xe_ggtt_node *node)
> +{
> +	return node->base.start;
> +}
> diff --git a/drivers/gpu/drm/xe/xe_ggtt.h b/drivers/gpu/drm/xe/xe_ggtt.h
> index d54a2ae2b4aa8..4a8ef1b824156 100644
> --- a/drivers/gpu/drm/xe/xe_ggtt.h
> +++ b/drivers/gpu/drm/xe/xe_ggtt.h
> @@ -58,4 +58,6 @@ void xe_ggtt_might_lock(struct xe_ggtt *ggtt);
>  u64 xe_ggtt_encode_pte_flags(struct xe_ggtt *ggtt, struct xe_bo *bo, u16 pat_index);
>  u64 xe_ggtt_read_pte(struct xe_ggtt *ggtt, u64 offset);
>  
> +u64 xe_ggtt_node_addr(const struct xe_ggtt_node *node);
> +
>  #endif
> -- 
> 2.51.0
> 

  reply	other threads:[~2025-10-15 21:49 UTC|newest]

Thread overview: 34+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-10-15  7:47 [PATCH v7 00/12] drm/xe: Make all xe_ggtt structs private Maarten Lankhorst
2025-10-15  7:47 ` [PATCH v7 01/12] drm/xe: Start using ggtt->start in preparation of balloon removal Maarten Lankhorst
2025-10-15  7:47 ` [PATCH v7 02/12] drm/xe: Convert xe_fb_pin to use a callback for insertion into GGTT Maarten Lankhorst
2025-10-15  7:47 ` [PATCH v7 03/12] drm/xe: Add xe_ggtt_node_addr() to avoid dereferencing xe_ggtt_node Maarten Lankhorst
2025-10-15 21:49   ` Matthew Brost [this message]
2025-10-16  6:13     ` Maarten Lankhorst
2025-10-17 20:53       ` Matthew Brost
2025-10-15  7:47 ` [PATCH v7 04/12] drm/xe/display: Avoid " Maarten Lankhorst
2025-10-15 21:53   ` Matthew Brost
2025-10-15  7:47 ` [PATCH v7 05/12] drm/xe: Do not dereference ggtt_node in xe_bo.c Maarten Lankhorst
2025-10-15 21:58   ` Matthew Brost
2025-10-16  6:33     ` Maarten Lankhorst
2025-10-17 20:54       ` Matthew Brost
2025-10-15  7:47 ` [PATCH v7 06/12] drm/xe: Improve xe_gt_sriov_pf_config GGTT handling Maarten Lankhorst
2025-10-15 22:04   ` Matthew Brost
2025-10-16  8:50     ` Michal Wajdeczko
2025-10-16 13:46       ` Maarten Lankhorst
2025-10-16 13:58         ` Michal Wajdeczko
2025-10-15  7:47 ` [PATCH v7 07/12] drm/xe: Privatize xe_ggtt_node Maarten Lankhorst
2025-10-15 22:05   ` Matthew Brost
2025-10-15  7:47 ` [PATCH v7 08/12] drm/xe: Make xe_ggtt_node offset relative to starting offset Maarten Lankhorst
2025-10-15 22:38   ` Matthew Brost
2025-10-16  6:40     ` Maarten Lankhorst
2025-10-15  7:47 ` [PATCH v7 09/12] drm/xe: Rewrite GGTT VF initialisation Maarten Lankhorst
2025-10-16  0:30   ` Matthew Brost
2025-10-15  7:47 ` [PATCH v7 10/12] drm/xe: Move struct xe_ggtt to xe_ggtt.c Maarten Lankhorst
2025-10-15  7:47 ` [PATCH v7 11/12] drm/xe: Make xe_ggtt_node_insert return a node Maarten Lankhorst
2025-10-15 22:44   ` Matthew Brost
2025-10-15  7:47 ` [PATCH v7 12/12] drm/xe: Remove xe_ggtt_node_allocated Maarten Lankhorst
2025-10-15 22:41   ` Matthew Brost
2025-10-15  9:47 ` ✗ CI.checkpatch: warning for drm/xe: Make all xe_ggtt structs private. (rev2) Patchwork
2025-10-15  9:49 ` ✓ CI.KUnit: success " Patchwork
2025-10-15 11:03 ` ✓ Xe.CI.BAT: " Patchwork
2025-10-15 19:55 ` ✗ Xe.CI.Full: failure " Patchwork

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=aPAW2H5JAyOFXl69@lstrano-desk.jf.intel.com \
    --to=matthew.brost@intel.com \
    --cc=dev@lankhorst.se \
    --cc=intel-xe@lists.freedesktop.org \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox