diff mbox

[v2,1/6] make RAM device optional

Message ID 1423639301-27304-2-git-send-email-acourbot@nvidia.com
State Not Applicable, archived
Headers show

Commit Message

Alexandre Courbot Feb. 11, 2015, 7:21 a.m. UTC 
Having a RAM device does not make sense for chips like GK20A which have
no dedicated video memory. The dummy RAM device that we used so far
works as a temporary band-aid, but in the long-term it is desirable for
the driver to be able to work without any kind of VRAM.

This patch adds a few conditionals in places where a RAM device was
assumed to be present and allows some more objects to be allocated from
the TT domain, allowing Nouveau to handle GPUs for which
pfb->ram == NULL.

Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
---
 drm/nouveau/nouveau_display.c         |  8 +++++++-
 drm/nouveau/nouveau_ttm.c             |  3 +++
 drm/nouveau/nv84_fence.c              | 14 +++++++++++---
 drm/nouveau/nvkm/engine/device/base.c |  9 ++++++---
 drm/nouveau/nvkm/subdev/clk/base.c    |  2 +-
 drm/nouveau/nvkm/subdev/fb/base.c     | 26 ++++++++++++++++++--------
 drm/nouveau/nvkm/subdev/ltc/gf100.c   | 15 +++++++++++----
 7 files changed, 57 insertions(+), 20 deletions(-)

Comments

Ilia Mirkin Feb. 11, 2015, 7:40 a.m. UTC | #1 
On Wed, Feb 11, 2015 at 2:21 AM, Alexandre Courbot <gnurou@gmail.com> wrote:
> @@ -150,7 +151,10 @@ gf100_ltc_init_tag_ram(struct nvkm_fb *pfb, struct nvkm_ltc_priv *priv)
>         int ret;
>
>         /* tags for 1/4 of VRAM should be enough (8192/4 per GiB of VRAM) */
> -       priv->num_tags = (pfb->ram->size >> 17) / 4;
> +       if (pfb->ram)
> +               priv->num_tags = (pfb->ram->size >> 17) / 4;
> +       else
> +               priv->num_tags = (1 << 17);
>         if (priv->num_tags > (1 << 17))
>                 priv->num_tags = 1 << 17; /* we have 17 bits in PTE */
>         priv->num_tags = (priv->num_tags + 63) & ~63; /* round up to 64 */
> @@ -170,8 +174,11 @@ gf100_ltc_init_tag_ram(struct nvkm_fb *pfb, struct nvkm_ltc_priv *priv)
>         tag_size += tag_align;
>         tag_size  = (tag_size + 0xfff) >> 12; /* round up */
>
> -       ret = nvkm_mm_tail(&pfb->vram, 1, 1, tag_size, tag_size, 1,
> -                          &priv->tag_ram);
> +       if (pfb->ram)
> +               ret = nvkm_mm_tail(&pfb->vram, 1, 1, tag_size, tag_size, 1,
> +                                  &priv->tag_ram);
> +       else
> +               ret = -1;
>         if (ret) {
>                 priv->num_tags = 0;
>         } else {

Seems like you could just avoid a lot of this and just start the
function out with

if (!pfb->ram) {
    priv->num_tags = 0; (is that even necessary? the object is
probably kzalloc'd)
    ret = nouveau_mm_init(&priv->tags, 0, 0, 1);
    return ret;
}

That's essentially what happens, no? Perhaps you could go a step
further and teach the base code about uninitialized tags mm, but that
may not be worth it.

  -ilia
--
To unsubscribe from this list: send the line "unsubscribe linux-tegra" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Alexandre Courbot Feb. 17, 2015, 7:22 a.m. UTC | #2 
On Wed, Feb 11, 2015 at 4:40 PM, Ilia Mirkin <imirkin@alum.mit.edu> wrote:
>
> On Wed, Feb 11, 2015 at 2:21 AM, Alexandre Courbot <gnurou@gmail.com> wrote:
> > @@ -150,7 +151,10 @@ gf100_ltc_init_tag_ram(struct nvkm_fb *pfb, struct nvkm_ltc_priv *priv)
> >         int ret;
> >
> >         /* tags for 1/4 of VRAM should be enough (8192/4 per GiB of VRAM) */
> > -       priv->num_tags = (pfb->ram->size >> 17) / 4;
> > +       if (pfb->ram)
> > +               priv->num_tags = (pfb->ram->size >> 17) / 4;
> > +       else
> > +               priv->num_tags = (1 << 17);
> >         if (priv->num_tags > (1 << 17))
> >                 priv->num_tags = 1 << 17; /* we have 17 bits in PTE */
> >         priv->num_tags = (priv->num_tags + 63) & ~63; /* round up to 64 */
> > @@ -170,8 +174,11 @@ gf100_ltc_init_tag_ram(struct nvkm_fb *pfb, struct nvkm_ltc_priv *priv)
> >         tag_size += tag_align;
> >         tag_size  = (tag_size + 0xfff) >> 12; /* round up */
> >
> > -       ret = nvkm_mm_tail(&pfb->vram, 1, 1, tag_size, tag_size, 1,
> > -                          &priv->tag_ram);
> > +       if (pfb->ram)
> > +               ret = nvkm_mm_tail(&pfb->vram, 1, 1, tag_size, tag_size, 1,
> > +                                  &priv->tag_ram);
> > +       else
> > +               ret = -1;
> >         if (ret) {
> >                 priv->num_tags = 0;
> >         } else {
>
> Seems like you could just avoid a lot of this and just start the
> function out with
>
> if (!pfb->ram) {
>     priv->num_tags = 0; (is that even necessary? the object is
> probably kzalloc'd)
>     ret = nouveau_mm_init(&priv->tags, 0, 0, 1);
>     return ret;
> }
>
> That's essentially what happens, no? Perhaps you could go a step
> further and teach the base code about uninitialized tags mm, but that
> may not be worth it.

You're absolutely correct - had I tried to understand the code instead
of workaround my way around it, I would have figured that out.

Actually the intention should be to keep some space for tags even when
VRAM is not present, so my code was not even correct in that respect.
We can keep it like this for the time being though, until I figure out
how this works for GK20A.
--
To unsubscribe from this list: send the line "unsubscribe linux-tegra" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
diff mbox

Patch

diff --git a/drm/nouveau/nouveau_display.c b/drm/nouveau/nouveau_display.c
index 860b0e2..68ee0af 100644
--- a/drm/nouveau/nouveau_display.c
+++ b/drm/nouveau/nouveau_display.c
@@ -869,13 +869,19 @@  nouveau_display_dumb_create(struct drm_file *file_priv, struct drm_device *dev,
 			    struct drm_mode_create_dumb *args)
 {
 	struct nouveau_bo *bo;
+	uint32_t domain;
 	int ret;
 
 	args->pitch = roundup(args->width * (args->bpp / 8), 256);
 	args->size = args->pitch * args->height;
 	args->size = roundup(args->size, PAGE_SIZE);
 
-	ret = nouveau_gem_new(dev, args->size, 0, NOUVEAU_GEM_DOMAIN_VRAM, 0, 0, &bo);
+	if (nvxx_fb(&nouveau_drm(dev)->device)->ram)
+		domain = NOUVEAU_GEM_DOMAIN_VRAM;
+	else
+		domain = NOUVEAU_GEM_DOMAIN_GART;
+
+	ret = nouveau_gem_new(dev, args->size, 0, domain, 0, 0, &bo);
 	if (ret)
 		return ret;
 
diff --git a/drm/nouveau/nouveau_ttm.c b/drm/nouveau/nouveau_ttm.c
index 273e501..a3c2e9b 100644
--- a/drm/nouveau/nouveau_ttm.c
+++ b/drm/nouveau/nouveau_ttm.c
@@ -85,6 +85,9 @@  nouveau_vram_manager_new(struct ttm_mem_type_manager *man,
 	if (nvbo->tile_flags & NOUVEAU_GEM_TILE_NONCONTIG)
 		size_nc = 1 << nvbo->page_shift;
 
+	if (!pfb->ram)
+		return -ENOMEM;
+
 	ret = pfb->ram->get(pfb, mem->num_pages << PAGE_SHIFT,
 			   mem->page_alignment << PAGE_SHIFT, size_nc,
 			   (nvbo->tile_flags >> 8) & 0x3ff, &node);
diff --git a/drm/nouveau/nv84_fence.c b/drm/nouveau/nv84_fence.c
index bf429ca..b981f85 100644
--- a/drm/nouveau/nv84_fence.c
+++ b/drm/nouveau/nv84_fence.c
@@ -215,6 +215,7 @@  nv84_fence_create(struct nouveau_drm *drm)
 {
 	struct nvkm_fifo *pfifo = nvxx_fifo(&drm->device);
 	struct nv84_fence_priv *priv;
+	u32 domain;
 	int ret;
 
 	priv = drm->fence = kzalloc(sizeof(*priv), GFP_KERNEL);
@@ -231,10 +232,17 @@  nv84_fence_create(struct nouveau_drm *drm)
 	priv->base.context_base = fence_context_alloc(priv->base.contexts);
 	priv->base.uevent = true;
 
-	ret = nouveau_bo_new(drm->dev, 16 * priv->base.contexts, 0,
-			     TTM_PL_FLAG_VRAM, 0, 0, NULL, NULL, &priv->bo);
+	domain = nvxx_fb(&drm->device)->ram ?
+			 TTM_PL_FLAG_VRAM :
+			 /*
+			  * fences created in TT must be coherent or we will
+			  * wait on old CPU cache values!
+			  */
+			 TTM_PL_FLAG_TT | TTM_PL_FLAG_UNCACHED;
+	ret = nouveau_bo_new(drm->dev, 16 * priv->base.contexts, 0, domain, 0,
+			     0, NULL, NULL, &priv->bo);
 	if (ret == 0) {
-		ret = nouveau_bo_pin(priv->bo, TTM_PL_FLAG_VRAM, false);
+		ret = nouveau_bo_pin(priv->bo, domain, false);
 		if (ret == 0) {
 			ret = nouveau_bo_map(priv->bo);
 			if (ret)
diff --git a/drm/nouveau/nvkm/engine/device/base.c b/drm/nouveau/nvkm/engine/device/base.c
index 6efa8f3..48f8537 100644
--- a/drm/nouveau/nvkm/engine/device/base.c
+++ b/drm/nouveau/nvkm/engine/device/base.c
@@ -139,9 +139,12 @@  nvkm_devobj_info(struct nvkm_object *object, void *data, u32 size)
 
 	args->v0.chipset  = device->chipset;
 	args->v0.revision = device->chiprev;
-	if (pfb)  args->v0.ram_size = args->v0.ram_user = pfb->ram->size;
-	else      args->v0.ram_size = args->v0.ram_user = 0;
-	if (imem) args->v0.ram_user = args->v0.ram_user - imem->reserved;
+	if (pfb  && pfb->ram)
+		args->v0.ram_size = args->v0.ram_user = pfb->ram->size;
+	else
+		args->v0.ram_size = args->v0.ram_user = 0;
+	if (imem)
+		args->v0.ram_user = args->v0.ram_user - imem->reserved;
 	return 0;
 }
 
diff --git a/drm/nouveau/nvkm/subdev/clk/base.c b/drm/nouveau/nvkm/subdev/clk/base.c
index b24a9cc..39a83d8 100644
--- a/drm/nouveau/nvkm/subdev/clk/base.c
+++ b/drm/nouveau/nvkm/subdev/clk/base.c
@@ -184,7 +184,7 @@  nvkm_pstate_prog(struct nvkm_clk *clk, int pstatei)
 	nv_debug(clk, "setting performance state %d\n", pstatei);
 	clk->pstate = pstatei;
 
-	if (pfb->ram->calc) {
+	if (pfb->ram && pfb->ram->calc) {
 		int khz = pstate->base.domain[nv_clk_src_mem];
 		do {
 			ret = pfb->ram->calc(pfb, khz);
diff --git a/drm/nouveau/nvkm/subdev/fb/base.c b/drm/nouveau/nvkm/subdev/fb/base.c
index 16589fa..61fde43 100644
--- a/drm/nouveau/nvkm/subdev/fb/base.c
+++ b/drm/nouveau/nvkm/subdev/fb/base.c
@@ -55,9 +55,11 @@  _nvkm_fb_fini(struct nvkm_object *object, bool suspend)
 	struct nvkm_fb *pfb = (void *)object;
 	int ret;
 
-	ret = nv_ofuncs(pfb->ram)->fini(nv_object(pfb->ram), suspend);
-	if (ret && suspend)
-		return ret;
+	if (pfb->ram) {
+		ret = nv_ofuncs(pfb->ram)->fini(nv_object(pfb->ram), suspend);
+		if (ret && suspend)
+			return ret;
+	}
 
 	return nvkm_subdev_fini(&pfb->base, suspend);
 }
@@ -72,9 +74,11 @@  _nvkm_fb_init(struct nvkm_object *object)
 	if (ret)
 		return ret;
 
-	ret = nv_ofuncs(pfb->ram)->init(nv_object(pfb->ram));
-	if (ret)
-		return ret;
+	if (pfb->ram) {
+		ret = nv_ofuncs(pfb->ram)->init(nv_object(pfb->ram));
+		if (ret)
+			return ret;
+	}
 
 	for (i = 0; i < pfb->tile.regions; i++)
 		pfb->tile.prog(pfb, i, &pfb->tile.region[i]);
@@ -91,9 +95,12 @@  _nvkm_fb_dtor(struct nvkm_object *object)
 	for (i = 0; i < pfb->tile.regions; i++)
 		pfb->tile.fini(pfb, i, &pfb->tile.region[i]);
 	nvkm_mm_fini(&pfb->tags);
-	nvkm_mm_fini(&pfb->vram);
 
-	nvkm_object_ref(NULL, (struct nvkm_object **)&pfb->ram);
+	if (pfb->ram) {
+		nvkm_mm_fini(&pfb->vram);
+		nvkm_object_ref(NULL, (struct nvkm_object **)&pfb->ram);
+	}
+
 	nvkm_subdev_destroy(&pfb->base);
 }
 
@@ -127,6 +134,9 @@  nvkm_fb_create_(struct nvkm_object *parent, struct nvkm_object *engine,
 
 	pfb->memtype_valid = impl->memtype;
 
+	if (!impl->ram)
+		return 0;
+
 	ret = nvkm_object_ctor(nv_object(pfb), NULL, impl->ram, NULL, 0, &ram);
 	if (ret) {
 		nv_fatal(pfb, "error detecting memory configuration!!\n");
diff --git a/drm/nouveau/nvkm/subdev/ltc/gf100.c b/drm/nouveau/nvkm/subdev/ltc/gf100.c
index 8e7cc62..7ae8e91 100644
--- a/drm/nouveau/nvkm/subdev/ltc/gf100.c
+++ b/drm/nouveau/nvkm/subdev/ltc/gf100.c
@@ -136,7 +136,8 @@  gf100_ltc_dtor(struct nvkm_object *object)
 	struct nvkm_ltc_priv *priv = (void *)object;
 
 	nvkm_mm_fini(&priv->tags);
-	nvkm_mm_free(&pfb->vram, &priv->tag_ram);
+	if (pfb->ram)
+		nvkm_mm_free(&pfb->vram, &priv->tag_ram);
 
 	nvkm_ltc_destroy(priv);
 }
@@ -150,7 +151,10 @@  gf100_ltc_init_tag_ram(struct nvkm_fb *pfb, struct nvkm_ltc_priv *priv)
 	int ret;
 
 	/* tags for 1/4 of VRAM should be enough (8192/4 per GiB of VRAM) */
-	priv->num_tags = (pfb->ram->size >> 17) / 4;
+	if (pfb->ram)
+		priv->num_tags = (pfb->ram->size >> 17) / 4;
+	else
+		priv->num_tags = (1 << 17);
 	if (priv->num_tags > (1 << 17))
 		priv->num_tags = 1 << 17; /* we have 17 bits in PTE */
 	priv->num_tags = (priv->num_tags + 63) & ~63; /* round up to 64 */
@@ -170,8 +174,11 @@  gf100_ltc_init_tag_ram(struct nvkm_fb *pfb, struct nvkm_ltc_priv *priv)
 	tag_size += tag_align;
 	tag_size  = (tag_size + 0xfff) >> 12; /* round up */
 
-	ret = nvkm_mm_tail(&pfb->vram, 1, 1, tag_size, tag_size, 1,
-			   &priv->tag_ram);
+	if (pfb->ram)
+		ret = nvkm_mm_tail(&pfb->vram, 1, 1, tag_size, tag_size, 1,
+				   &priv->tag_ram);
+	else
+		ret = -1;
 	if (ret) {
 		priv->num_tags = 0;
 	} else {