Re: [RFC PATCH V6 1/7] drm/xe: Link VRAM object with gpu buddy

[Matthew Brost <matthew.brost@intel.com>]

On Thu, Apr 02, 2026 at 03:10:02AM -0600, Upadhyay, Tejas wrote:
> 
> 
> > -----Original Message-----
> > From: Brost, Matthew <matthew.brost@intel.com>
> > Sent: 02 April 2026 05:27
> > To: Upadhyay, Tejas <tejas.upadhyay@intel.com>
> > Cc: intel-xe@lists.freedesktop.org; Auld, Matthew
> > <matthew.auld@intel.com>; thomas.hellstrom@linux.intel.com; Ghimiray,
> > Himal Prasad <himal.prasad.ghimiray@intel.com>
> > Subject: Re: [RFC PATCH V6 1/7] drm/xe: Link VRAM object with gpu buddy
> > 
> > On Fri, Mar 27, 2026 at 05:18:14PM +0530, Tejas Upadhyay wrote:
> > > Setup to link TTM buffer object inside gpu buddy. This functionality
> > > is critical for supporting the memory page offline feature on CRI,
> > > where identified faulty pages must be traced back to their originating
> > > buffer for safe removal.
> > >
> > 
> > I just checked the SVM code is still setting 'block->private'.
> > 
> > I thought we had patch for that but I don't see it in this series.
> 
> Yeah, you are right, it was there when I sent earlier revision, I somehow missed adding that patch in this series. I will add it back in next revision. Sorry for confusion here.
> 

Also free to send that as indepent patch as we should be able to merge
that one whenever.

Another comment.

> Tejas
> > 
> > Matt
> > 
> > > Signed-off-by: Tejas Upadhyay <tejas.upadhyay@intel.com>
> > > ---
> > >  drivers/gpu/drm/xe/xe_ttm_vram_mgr.c | 3 +++
> > >  1 file changed, 3 insertions(+)
> > >
> > > diff --git a/drivers/gpu/drm/xe/xe_ttm_vram_mgr.c
> > > b/drivers/gpu/drm/xe/xe_ttm_vram_mgr.c
> > > index 5fd0d5506a7e..c627dbf94552 100644
> > > --- a/drivers/gpu/drm/xe/xe_ttm_vram_mgr.c
> > > +++ b/drivers/gpu/drm/xe/xe_ttm_vram_mgr.c
> > > @@ -54,6 +54,7 @@ static int xe_ttm_vram_mgr_new(struct
> > ttm_resource_manager *man,
> > >  	struct xe_ttm_vram_mgr *mgr = to_xe_ttm_vram_mgr(man);
> > >  	struct xe_ttm_vram_mgr_resource *vres;
> > >  	struct gpu_buddy *mm = &mgr->mm;
> > > +	struct gpu_buddy_block *block;
> > >  	u64 size, min_page_size;
> > >  	unsigned long lpfn;
> > >  	int err;
> > > @@ -138,6 +139,8 @@ static int xe_ttm_vram_mgr_new(struct
> > ttm_resource_manager *man,
> > >  	}
> > >
> > >  	mgr->visible_avail -= vres->used_visible_size;
> > > +	list_for_each_entry(block, &vres->blocks, link)
> > > +		block->private = tbo;

Should we clear block->private in xe_ttm_vram_mgr_del() to avoid blocks
having stale references to BOs? Or are there guarantees elsewhere in the
code that a free block cannot reference block->private as a BO if it
encounters an error while in a free state? For safety and clarity, even
if such guarantees exist, it may still make sense to clear
block->private in xe_ttm_vram_mgr_del().

Matt

> > >  	mutex_unlock(&mgr->lock);
> > >
> > >  	if (!(vres->base.placement & TTM_PL_FLAG_CONTIGUOUS) &&
> > > --
> > > 2.52.0
> > >