[Intel-gfx] [PATCH v2 2/3] drm/i915: Document our internal limit on object size

[Chris Wilson]

On Tue, Oct 18, 2016 at 10:27:58AM +0100, Tvrtko Ursulin wrote:
> 
> On 17/10/2016 09:00, Chris Wilson wrote:
> >In many places, we try to count pages using a 32 bit integer. That
> >implies if we are asked to create an object larger than 43bits, we will
> >subtly crash much later. Catch this on the boundary, and add a warning
> >to remind ourselves later on our exabyte systems.
> >
> >Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
> >---
> >  drivers/gpu/drm/i915/i915_drv.h |  2 +-
> >  drivers/gpu/drm/i915/i915_gem.c | 17 +++++++++++++++--
> >  2 files changed, 16 insertions(+), 3 deletions(-)
> >
> >diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
> >index 092c5a0a44f0..a2b5fc72fdd9 100644
> >--- a/drivers/gpu/drm/i915/i915_drv.h
> >+++ b/drivers/gpu/drm/i915/i915_drv.h
> >@@ -3105,7 +3105,7 @@ void i915_gem_object_free(struct drm_i915_gem_object *obj);
> >  void i915_gem_object_init(struct drm_i915_gem_object *obj,
> >  			 const struct drm_i915_gem_object_ops *ops);
> >  struct drm_i915_gem_object *i915_gem_object_create(struct drm_device *dev,
> >-						  size_t size);
> >+						   u64 size);
> >  struct drm_i915_gem_object *i915_gem_object_create_from_data(
> >  		struct drm_device *dev, const void *data, size_t size);
> >  void i915_gem_close_object(struct drm_gem_object *gem, struct drm_file *file);
> >diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
> >index 838dc159a2d1..181bda2db587 100644
> >--- a/drivers/gpu/drm/i915/i915_gem.c
> >+++ b/drivers/gpu/drm/i915/i915_gem.c
> >@@ -4131,14 +4131,27 @@ static const struct drm_i915_gem_object_ops i915_gem_object_ops = {
> >  	.put_pages = i915_gem_object_put_pages_gtt,
> >  };
> >-struct drm_i915_gem_object *i915_gem_object_create(struct drm_device *dev,
> >-						  size_t size)
> >+#define overflows_type(x, T) \
> >+	(sizeof(x) > sizeof(T) && (x) >> (sizeof(T) * BITS_PER_BYTE))
> >+
> 
> Looks like it wouldn't detect storing unsigned int in a signed int
> but I guess we don't care that much as long as this is local use
> only. Just slightly relevant because of the int page_count situation
> we mention below.

Hmm. Yeah, definitely worth improving. Quick googling shows that you are
the first to notice! :-p

I was thinking of trying gcc's __builtin_add_overflowp(x, 0, T) or
something like that.

But I also wonder if we can use signed T *var vs unsigned T *var in any
way to generalise the number of positive bits in a type.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre