[Intel-gfx] [PATCH 4/4] drm/i915: Support for pread/pwrite from/to non shmem backed objects

Tvrtko Ursulin tvrtko.ursulin at linux.intel.com
Wed Jul 22 08:46:00 PDT 2015


Hi,

On 07/22/2015 02:51 PM, ankitprasad.r.sharma at intel.com wrote:
> From: Ankitprasad Sharma <ankitprasad.r.sharma at intel.com>
>
> This patch adds support for extending the pread/pwrite functionality
> for objects not backed by shmem. The access will be made through
> gtt interface.
> This will cover prime objects as well as stolen memory backed objects
> but for userptr objects it is still forbidden.
>
> v2: Drop locks around slow_user_access, prefault the pages before
> access (Chris)
>
> v3: Rebased to the latest drm-intel-nightly (Ankit)
>
> v4: Moved page base & offset calculations outside the copy loop,
> corrected data types for size and offset variables, corrected if-else
> braces format (Tvrtko/kerneldocs)
>
> Testcase: igt/gem_stolen
>
> Signed-off-by: Ankitprasad Sharma <ankitprasad.r.sharma at intel.com>
> ---
>   drivers/gpu/drm/i915/i915_gem.c | 138 +++++++++++++++++++++++++++++++++++-----
>   1 file changed, 121 insertions(+), 17 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
> index 9e7e182..4321178 100644
> --- a/drivers/gpu/drm/i915/i915_gem.c
> +++ b/drivers/gpu/drm/i915/i915_gem.c
> @@ -631,6 +631,102 @@ shmem_pread_slow(struct page *page, int shmem_page_offset, int page_length,
>   	return ret ? - EFAULT : 0;
>   }
>
> +static inline uint64_t
> +slow_user_access(struct io_mapping *mapping,
> +		 uint64_t page_base, int page_offset,
> +		 char __user *user_data,
> +		 int length, bool pwrite)
> +{
> +	void __iomem *vaddr_inatomic;
> +	void *vaddr;
> +	uint64_t unwritten;
> +
> +	vaddr_inatomic = io_mapping_map_wc(mapping, page_base);
> +	/* We can use the cpu mem copy function because this is X86. */
> +	vaddr = (void __force *)vaddr_inatomic + page_offset;
> +	if (pwrite)
> +		unwritten = __copy_from_user(vaddr, user_data, length);
> +	else
> +		unwritten = __copy_to_user(user_data, vaddr, length);
> +
> +	io_mapping_unmap(vaddr_inatomic);
> +	return unwritten;
> +}
> +
> +static int
> +i915_gem_gtt_pread_pwrite(struct drm_device *dev,
> +			  struct drm_i915_gem_object *obj, uint64_t size,
> +			  uint64_t data_offset, uint64_t data_ptr, bool pwrite)
> +{
> +	struct drm_i915_private *dev_priv = dev->dev_private;
> +	char __user *user_data;
> +	uint64_t remain;
> +	uint64_t offset, page_base;
> +	int page_offset, page_length, ret = 0;
> +
> +	ret = i915_gem_obj_ggtt_pin(obj, 0, PIN_MAPPABLE);
> +	if (ret)
> +		goto out;
> +
> +	ret = i915_gem_object_set_to_gtt_domain(obj, pwrite);
> +	if (ret)
> +		goto out_unpin;
> +
> +	ret = i915_gem_object_put_fence(obj);
> +	if (ret)
> +		goto out_unpin;
> +
> +	user_data = to_user_ptr(data_ptr);
> +	remain = size;
> +	offset = i915_gem_obj_ggtt_offset(obj) + data_offset;
> +
> +	if (pwrite)
> +		intel_fb_obj_invalidate(obj, ORIGIN_GTT);
> +
> +	mutex_unlock(&dev->struct_mutex);
> +	if (!pwrite && likely(!i915.prefault_disable))
> +		ret = fault_in_multipages_writeable(user_data, remain);
> +
> +	/*
> +	 * page_offset = offset within page
> +	 * page_base = page offset within aperture
> +	 */
> +	page_offset = offset_in_page(offset);
> +	page_base = offset & PAGE_MASK;
> +
> +	while (remain > 0) {
> +		/* page_length = bytes to copy for this page */
> +		page_length = remain;
> +		if ((page_offset + remain) > PAGE_SIZE)
> +			page_length = PAGE_SIZE - page_offset;
> +
> +		/* This is a slow read/write as it tries to read from
> +		 * and write to user memory which may result into page
> +		 * faults
> +		 */
> +		ret = slow_user_access(dev_priv->gtt.mappable, page_base,
> +				       page_offset, user_data,
> +				       page_length, pwrite);
> +
> +		if (ret) {
> +			ret = -EFAULT;
> +			break;
> +		}
> +
> +		remain -= page_length;
> +		user_data += page_length;
> +		page_base += page_length;
> +		page_offset = 0;
> +	}
> +
> +	mutex_lock(&dev->struct_mutex);

My objection here was that we are re-acquiring the mutex in 
non-interruptible mode, while the caller had it interruptible. It am not 
100% what the conclusion back then was?

But also, why it is even necessary to drop the mutex here?

Regards,

Tvrtko


More information about the Intel-gfx mailing list