public inbox for intel-gfx@lists.freedesktop.org
 help / color / mirror / Atom feed
From: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
To: Zhi Wang <zhi.a.wang@intel.com>,
	intel-gfx@lists.freedesktop.org, tvrtko.ursulin@linux.intel.com,
	kevin.tian@intel.com, zhiyuan.lv@intel.com,
	chris@chris-wilson.co.uk
Subject: Re: [PATCH v6 4/9] drm/i915: Introduce host graphics memory partition for GVT-g
Date: Fri, 03 Jun 2016 12:17:55 +0300	[thread overview]
Message-ID: <1464945475.7292.21.camel@linux.intel.com> (raw)
In-Reply-To: <1464885380-7056-5-git-send-email-zhi.a.wang@intel.com>

On to, 2016-06-02 at 12:36 -0400, Zhi Wang wrote:
> From: Bing Niu <bing.niu@intel.com>
> 
> This patch introduces host graphics memory partition when GVT-g
> is enabled.
> 
> Under GVT-g, i915 host driver only owned limited graphics resources,
> others are managed by GVT-g resource allocator and kept for other vGPUs.
> 
> v6:
> 
> - Remove kernel parameters used to configure GGTT owned by host. (Chris)
> - Other coding style comments from Chris.
> - Add more comments for reviewer.
> 
> v3:
> 
> - Remove fence partition, will use i915 fence stealing in future.(Kevin)
> - Santinize GVT host gm kernel parameters. (Joonas)
> 
> v2:
> - Address all coding-style comments from Joonas previously.
> - Fix errors and warnning reported by checkpatch.pl. (Joonas)
> - Move the graphs into the header files. (Daniel)
> 
> Signed-off-by: Bing Niu <bing.niu@intel.com>
> Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
> ---
>  drivers/gpu/drm/i915/i915_vgpu.c | 23 +++++++++++++++++------
>  drivers/gpu/drm/i915/intel_gvt.h | 25 +++++++++++++++++++++++++
>  2 files changed, 42 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/i915_vgpu.c b/drivers/gpu/drm/i915/i915_vgpu.c
> index 5312816..0d891a3 100644
> --- a/drivers/gpu/drm/i915/i915_vgpu.c
> +++ b/drivers/gpu/drm/i915/i915_vgpu.c
> @@ -189,14 +189,25 @@ int intel_vgt_balloon(struct drm_i915_private *dev_priv)
>  	unsigned long unmappable_base, unmappable_size, unmappable_end;
>  	int ret;
>  
> -	if (!intel_vgpu_active(dev_priv))
> +	if (intel_gvt_active(dev_priv)) {
> +		/* Retrieve GGTT partition information from macros */
> +		mappable_base = 0;
> +		mappable_size = INTEL_GVT_HOST_LOW_GM_SIZE;
> +		unmappable_base = dev_priv->ggtt.mappable_end;
> +		unmappable_size = INTEL_GVT_HOST_HIGH_GM_SIZE;
> +	} else if (intel_vgpu_active(dev_priv)) {
> +		/* Retrieve GGTT partition information from PVINFO */
> +		mappable_base = I915_READ(
> +				vgtif_reg(avail_rs.mappable_gmadr.base));
> +		mappable_size = I915_READ(
> +				vgtif_reg(avail_rs.mappable_gmadr.size));
> +		unmappable_base = I915_READ(
> +				vgtif_reg(avail_rs.nonmappable_gmadr.base));
> +		unmappable_size = I915_READ(
> +				vgtif_reg(avail_rs.nonmappable_gmadr.size));
> +	} else
>  		return 0;
>  
> -	mappable_base = I915_READ(vgtif_reg(avail_rs.mappable_gmadr.base));
> -	mappable_size = I915_READ(vgtif_reg(avail_rs.mappable_gmadr.size));
> -	unmappable_base = I915_READ(vgtif_reg(avail_rs.nonmappable_gmadr.base));
> -	unmappable_size = I915_READ(vgtif_reg(avail_rs.nonmappable_gmadr.size));
> -
>  	mappable_end = mappable_base + mappable_size;
>  	unmappable_end = unmappable_base + unmappable_size;
>  
> diff --git a/drivers/gpu/drm/i915/intel_gvt.h b/drivers/gpu/drm/i915/intel_gvt.h
> index b9b361b..f0b9aac 100644
> --- a/drivers/gpu/drm/i915/intel_gvt.h
> +++ b/drivers/gpu/drm/i915/intel_gvt.h
> @@ -24,6 +24,31 @@
>  #ifndef _INTEL_GVT_H_
>  #define _INTEL_GVT_H_
>  
> +/*
> + * Under GVT-g, i915 host driver only owned limited graphics resources,
> + * others are managed by GVT-g resource allocator and kept for other vGPUs.
> + *
> + * For graphics memory space partition, a typical layout looks like:
> + *
> + * +-------+-----------------------+------+-----------------------+
> + * |* Host |   *GVT-g Resource     |* Host|   *GVT-g Resource     |
> + * | Owned |   Allocator Managed   | Owned|   Allocator Managed   |
> + * |       |                       |      |                       |
> + * +---------------+-------+----------------------+-------+-------+
> + * |       |       |       |       |      |       |       |       |
> + * | i915  | vm 1  | vm 2  | vm 3  | i915 | vm 1  | vm 2  | vm 3  |
> + * |       |       |       |       |      |       |       |       |
> + * +-------+-------+-------+--------------+-------+-------+-------+
> + * |           Aperture            |            Hidden            |
> + * +-------------------------------+------------------------------+
> + * |                       GGTT memory space                      |
> + * +--------------------------------------------------------------+
> + */
> +
> +/* GGTT memory space owned by host */
> +#define INTEL_GVT_HOST_LOW_GM_SIZE (96 * 1024 * 1024)
> +#define INTEL_GVT_HOST_HIGH_GM_SIZE (384 * 1024 * 1024)

Might be worth a comment where these numbers come from, with that;

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

> +
>  #ifdef CONFIG_DRM_I915_GVT
>  #include "gvt/gvt.h"
>  extern int intel_gvt_init(struct drm_i915_private *dev_priv);
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

  reply	other threads:[~2016-06-03  9:18 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-06-02 16:36 [PATCH v6 0/9] Introduce the implementation of GVT context Zhi Wang
2016-06-02 16:36 ` [PATCH v6 1/9] drm/i915: Factor out i915_pvinfo.h Zhi Wang
2016-06-03  8:45   ` Joonas Lahtinen
2016-06-02 16:36 ` [PATCH v6 2/9] drm/i915: Fold vGPU active check into inner functions Zhi Wang
2016-06-03  8:49   ` Joonas Lahtinen
2016-06-02 16:36 ` [PATCH v6 3/9] drm/i915: gvt: Introduce the basic architecture of GVT-g Zhi Wang
2016-06-03  9:14   ` Joonas Lahtinen
2016-06-02 16:36 ` [PATCH v6 4/9] drm/i915: Introduce host graphics memory partition for GVT-g Zhi Wang
2016-06-03  9:17   ` Joonas Lahtinen [this message]
2016-06-02 16:36 ` [PATCH v6 5/9] drm/i915: Make ring buffer size of a LRC context configurable Zhi Wang
2016-06-03  9:20   ` Joonas Lahtinen
2016-06-02 16:36 ` [PATCH v6 6/9] drm/i915: Make addressing mode bits in context descriptor configurable Zhi Wang
2016-06-03  9:25   ` Joonas Lahtinen
2016-06-02 16:36 ` [PATCH v6 7/9] drm/i915: Introduce execlist context status change notification Zhi Wang
2016-06-03  9:40   ` Joonas Lahtinen
2016-06-07 15:29     ` Wang, Zhi A
2016-06-08  7:49       ` Joonas Lahtinen
2016-06-02 16:36 ` [PATCH v6 8/9] drm/i915: Support LRC context single submission Zhi Wang
2016-06-03  9:47   ` Joonas Lahtinen
2016-06-03 11:25     ` Tian, Kevin
2016-06-07 14:13     ` Wang, Zhi A
2016-06-02 16:36 ` [PATCH v6 9/9] drm/i915: Introduce GVT context creation API Zhi Wang
2016-06-03  9:59   ` Joonas Lahtinen
2016-06-03  6:43 ` ✗ Ro.CI.BAT: warning for Introduce the implementation of GVT context (rev4) 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=1464945475.7292.21.camel@linux.intel.com \
    --to=joonas.lahtinen@linux.intel.com \
    --cc=chris@chris-wilson.co.uk \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=kevin.tian@intel.com \
    --cc=tvrtko.ursulin@linux.intel.com \
    --cc=zhi.a.wang@intel.com \
    --cc=zhiyuan.lv@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