Intel-XE Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Balasubramani Vivekanandan <balasubramani.vivekanandan@intel.com>
To: Lucas De Marchi <lucas.demarchi@intel.com>,
	<intel-xe@lists.freedesktop.org>
Cc: Matt Roper <matthew.d.roper@intel.com>
Subject: Re: [Intel-xe] [PATCH v2 10/15] drm/xe/xe2: Track VA bits independently of max page table level
Date: Tue, 22 Aug 2023 11:24:22 +0530	[thread overview]
Message-ID: <ZORNjghEvtzgjc4P@bvivekan-mobl> (raw)
In-Reply-To: <20230818220824.700519-11-lucas.demarchi@intel.com>

On 18.08.2023 15:08, Lucas De Marchi wrote:
> From: Matt Roper <matthew.d.roper@intel.com>
> 
> Starting with Xe2, a 5-level page table is always used, regardless of
> the actual virtual address range supported by the platform.  The two
> values need to be tracked separately in the device descriptor since Xe2
> platforms only have a 48 bit virtual address range.
> 
> Bspec: 59505, 65637, 70817
> Cc: Balasubramani Vivekanandan <balasubramani.vivekanandan@intel.com>
> Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
> Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
> ---
>  drivers/gpu/drm/xe/xe_device_types.h | 2 ++
>  drivers/gpu/drm/xe/xe_pci.c          | 6 ++++++
>  drivers/gpu/drm/xe/xe_pci_types.h    | 1 +
>  drivers/gpu/drm/xe/xe_query.c        | 3 +--
>  drivers/gpu/drm/xe/xe_vm.c           | 2 +-
>  5 files changed, 11 insertions(+), 3 deletions(-)

Reviewed-by: Balasubramani Vivekanandan <balasubramani.vivekanandan@intel.com>

Regards,
Bala

> 
> diff --git a/drivers/gpu/drm/xe/xe_device_types.h b/drivers/gpu/drm/xe/xe_device_types.h
> index 303447c093c5..18292aad87e9 100644
> --- a/drivers/gpu/drm/xe/xe_device_types.h
> +++ b/drivers/gpu/drm/xe/xe_device_types.h
> @@ -216,6 +216,8 @@ struct xe_device {
>  		u8 gt_count;
>  		/** @vm_max_level: Max VM level */
>  		u8 vm_max_level;
> +		/** @va_bits: Maximum bits of a virtual address */
> +		u8 va_bits;
>  
>  		/** @is_dgfx: is discrete device */
>  		u8 is_dgfx:1;
> diff --git a/drivers/gpu/drm/xe/xe_pci.c b/drivers/gpu/drm/xe/xe_pci.c
> index fdd2eba1b91f..70d5f25c0d20 100644
> --- a/drivers/gpu/drm/xe/xe_pci.c
> +++ b/drivers/gpu/drm/xe/xe_pci.c
> @@ -106,6 +106,7 @@ static const struct xe_graphics_desc graphics_xelp = {
>  	.hw_engine_mask = BIT(XE_HW_ENGINE_RCS0) | BIT(XE_HW_ENGINE_BCS0),
>  
>  	.dma_mask_size = 39,
> +	.va_bits = 48,
>  	.vm_max_level = 3,
>  };
>  
> @@ -117,6 +118,7 @@ static const struct xe_graphics_desc graphics_xelpp = {
>  	.hw_engine_mask = BIT(XE_HW_ENGINE_RCS0) | BIT(XE_HW_ENGINE_BCS0),
>  
>  	.dma_mask_size = 39,
> +	.va_bits = 48,
>  	.vm_max_level = 3,
>  };
>  
> @@ -124,6 +126,7 @@ static const struct xe_graphics_desc graphics_xelpp = {
>  	.has_range_tlb_invalidation = true, \
>  	.has_flat_ccs = true, \
>  	.dma_mask_size = 46, \
> +	.va_bits = 48, \
>  	.vm_max_level = 3
>  
>  static const struct xe_graphics_desc graphics_xehpg = {
> @@ -157,6 +160,7 @@ static const struct xe_graphics_desc graphics_xehpc = {
>  	XE_HP_FEATURES,
>  	.dma_mask_size = 52,
>  	.max_remote_tiles = 1,
> +	.va_bits = 57,
>  	.vm_max_level = 4,
>  	.vram_flags = XE_VRAM_FLAGS_NEED64K,
>  
> @@ -182,6 +186,7 @@ static const struct xe_graphics_desc graphics_xelpg = {
>  	.has_flat_ccs = 0 /* FIXME: implementation missing */, \
>  	.has_range_tlb_invalidation = 1, \
>  	.supports_usm = 0 /* FIXME: implementation missing */, \
> +	.va_bits = 48, \
>  	.vm_max_level = 4, \
>  	.hw_engine_mask = \
>  		BIT(XE_HW_ENGINE_RCS0) | \
> @@ -587,6 +592,7 @@ static int xe_info_init(struct xe_device *xe,
>  
>  	xe->info.dma_mask_size = graphics_desc->dma_mask_size;
>  	xe->info.vram_flags = graphics_desc->vram_flags;
> +	xe->info.va_bits = graphics_desc->va_bits;
>  	xe->info.vm_max_level = graphics_desc->vm_max_level;
>  	xe->info.supports_usm = graphics_desc->supports_usm;
>  	xe->info.has_asid = graphics_desc->has_asid;
> diff --git a/drivers/gpu/drm/xe/xe_pci_types.h b/drivers/gpu/drm/xe/xe_pci_types.h
> index ba31b933eb8e..df6ddbc52d7f 100644
> --- a/drivers/gpu/drm/xe/xe_pci_types.h
> +++ b/drivers/gpu/drm/xe/xe_pci_types.h
> @@ -14,6 +14,7 @@ struct xe_graphics_desc {
>  	u8 rel;
>  
>  	u8 dma_mask_size;	/* available DMA address bits */
> +	u8 va_bits;
>  	u8 vm_max_level;
>  	u8 vram_flags;
>  
> diff --git a/drivers/gpu/drm/xe/xe_query.c b/drivers/gpu/drm/xe/xe_query.c
> index 7ea235c71385..1db77a7c9039 100644
> --- a/drivers/gpu/drm/xe/xe_query.c
> +++ b/drivers/gpu/drm/xe/xe_query.c
> @@ -197,8 +197,7 @@ static int query_config(struct xe_device *xe, struct drm_xe_device_query *query)
>  			XE_QUERY_CONFIG_FLAGS_HAS_VRAM;
>  	config->info[XE_QUERY_CONFIG_MIN_ALIGNEMENT] =
>  		xe->info.vram_flags & XE_VRAM_FLAGS_NEED64K ? SZ_64K : SZ_4K;
> -	config->info[XE_QUERY_CONFIG_VA_BITS] = 12 +
> -		(9 * (xe->info.vm_max_level + 1));
> +	config->info[XE_QUERY_CONFIG_VA_BITS] = xe->info.va_bits;
>  	config->info[XE_QUERY_CONFIG_GT_COUNT] = xe->info.gt_count;
>  	config->info[XE_QUERY_CONFIG_MEM_REGION_COUNT] =
>  		hweight_long(xe->info.mem_region_mask);
> diff --git a/drivers/gpu/drm/xe/xe_vm.c b/drivers/gpu/drm/xe/xe_vm.c
> index 2e99f865d7ec..389ac5ba8ddf 100644
> --- a/drivers/gpu/drm/xe/xe_vm.c
> +++ b/drivers/gpu/drm/xe/xe_vm.c
> @@ -1207,7 +1207,7 @@ struct xe_vm *xe_vm_create(struct xe_device *xe, u32 flags)
>  	kref_init(&vm->refcount);
>  	dma_resv_init(&vm->resv);
>  
> -	vm->size = 1ull << xe_pt_shift(xe->info.vm_max_level + 1);
> +	vm->size = 1ull << xe->info.va_bits;
>  
>  	vm->flags = flags;
>  
> -- 
> 2.40.1
> 

  reply	other threads:[~2023-08-22  5:54 UTC|newest]

Thread overview: 33+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-08-18 22:08 [Intel-xe] [PATCH v2 00/15] Add Lunar Lake support Lucas De Marchi
2023-08-18 22:08 ` [Intel-xe] [PATCH v2 01/15] drm/xe/xe2: Update render/compute context image sizes Lucas De Marchi
2023-08-18 22:08 ` [Intel-xe] [PATCH v2 02/15] drm/xe/xe2: Add GT topology readout Lucas De Marchi
2023-08-18 22:08 ` [Intel-xe] [PATCH v2 03/15] drm/xe/xe2: Add MCR register steering for primary GT Lucas De Marchi
2023-08-21 14:32   ` Balasubramani Vivekanandan
2023-08-22 16:55     ` Matt Atwood
2023-08-18 22:08 ` [Intel-xe] [PATCH v2 04/15] drm/xe/xe2: Add MCR register steering for media GT Lucas De Marchi
2023-08-18 22:08 ` [Intel-xe] [PATCH v2 05/15] drm/xe/xe2: Update context image layouts Lucas De Marchi
2023-08-18 22:08 ` [Intel-xe] [PATCH v2 06/15] drm/xe/xe2: Handle fused-off CCS engines Lucas De Marchi
2023-08-18 22:08 ` [Intel-xe] [PATCH v2 07/15] drm/xe/xe2: AuxCCS is no longer used Lucas De Marchi
2023-08-18 22:08 ` [Intel-xe] [PATCH v2 08/15] drm/xe/xe2: Define Xe2_LPG IP features Lucas De Marchi
2023-08-21 14:58   ` Balasubramani Vivekanandan
2023-08-21 16:03     ` Lucas De Marchi
2023-08-22  5:38       ` Balasubramani Vivekanandan
2023-08-18 22:08 ` [Intel-xe] [PATCH v2 09/15] drm/xe/xe2: Define Xe2_LPM " Lucas De Marchi
2023-08-21 14:23   ` Matt Roper
2023-08-21 18:46     ` Lucas De Marchi
2023-08-22  5:41   ` Balasubramani Vivekanandan
2023-08-18 22:08 ` [Intel-xe] [PATCH v2 10/15] drm/xe/xe2: Track VA bits independently of max page table level Lucas De Marchi
2023-08-22  5:54   ` Balasubramani Vivekanandan [this message]
2023-08-18 22:08 ` [Intel-xe] [PATCH v2 11/15] drm/xe/xe2: Add MOCS table Lucas De Marchi
2023-08-18 22:08 ` [Intel-xe] [PATCH v2 12/15] drm/xe/xe2: Program GuC's MOCS on Xe2 and beyond Lucas De Marchi
2023-08-22  6:27   ` Balasubramani Vivekanandan
2023-08-18 22:08 ` [Intel-xe] [PATCH v2 13/15] drm/xe/lnl: Add LNL platform definition Lucas De Marchi
2023-08-18 22:08 ` [Intel-xe] [PATCH v2 14/15] drm/xe/lnl: Add GuC firmware definition Lucas De Marchi
2023-08-18 22:08 ` [Intel-xe] [PATCH v2 15/15] drm/xe/lnl: Hook up MOCS table Lucas De Marchi
2023-08-18 22:11 ` [Intel-xe] ✓ CI.Patch_applied: success for Add Lunar Lake support (rev2) Patchwork
2023-08-18 22:11 ` [Intel-xe] ✗ CI.checkpatch: warning " Patchwork
2023-08-18 22:12 ` [Intel-xe] ✓ CI.KUnit: success " Patchwork
2023-08-18 22:16 ` [Intel-xe] ✓ CI.Build: " Patchwork
2023-08-18 22:17 ` [Intel-xe] ✓ CI.Hooks: " Patchwork
2023-08-18 22:17 ` [Intel-xe] ✗ CI.checksparse: warning " Patchwork
2023-08-18 22:44 ` [Intel-xe] ✓ CI.BAT: success " 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=ZORNjghEvtzgjc4P@bvivekan-mobl \
    --to=balasubramani.vivekanandan@intel.com \
    --cc=intel-xe@lists.freedesktop.org \
    --cc=lucas.demarchi@intel.com \
    --cc=matthew.d.roper@intel.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox