Auto merge of #1772 - christolliday:fix_gpu_location_panic, r=glennw

Fix 'RenderBackend' panicked at 'handle not requested or allocated!'

I started writing up an issue for this, but since it's not easily reproduced I did some debugging and found the issue, so thought I'd just make a PR.

After upgrading webrender I started seeing this crash in an example app using a custom webrender based UI library. I could reproduce it somewhat regularly but unfortunately not through a reliable set of steps, so I don't have a simple test case.

```
thread 'RenderBackend' panicked at 'handle not requested or allocated!', /checkout/src/libcore/option.rs:839:4
note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.
stack backtrace:
   0: std::sys::imp::backtrace::tracing::imp::unwind_backtrace
             at /checkout/src/libstd/sys/unix/backtrace/tracing/gcc_s.rs:49
   1: std::sys_common::backtrace::_print
             at /checkout/src/libstd/sys_common/backtrace.rs:71
   2: std::panicking::default_hook::{{closure}}
             at /checkout/src/libstd/sys_common/backtrace.rs:60
             at /checkout/src/libstd/panicking.rs:381
   3: std::panicking::default_hook
             at /checkout/src/libstd/panicking.rs:397
   4: std::panicking::rust_panic_with_hook
             at /checkout/src/libstd/panicking.rs:577
   5: std::panicking::begin_panic
             at /checkout/src/libstd/panicking.rs:538
   6: std::panicking::begin_panic_fmt
             at /checkout/src/libstd/panicking.rs:522
   7: rust_begin_unwind
             at /checkout/src/libstd/panicking.rs:498
   8: core::panicking::panic_fmt
             at /checkout/src/libcore/panicking.rs:71
   9: core::option::expect_failed
             at /checkout/src/libcore/option.rs:839
  10: <core::option::Option<T>>::expect
             at /checkout/src/libcore/option.rs:302
  11: webrender::gpu_cache::GpuCache::get_address
             at /media/chris/linux_data/dev/rust/gui/webrender/webrender/src/gpu_cache.rs:592
  12: webrender::tiling::<impl webrender::render_task::AlphaRenderItem>::add_to_batch
             at /media/chris/linux_data/dev/rust/gui/webrender/webrender/src/tiling.rs:364
  13: webrender::tiling::AlphaBatcher::build
             at /media/chris/linux_data/dev/rust/gui/webrender/webrender/src/tiling.rs:745
  14: <webrender::tiling::ColorRenderTarget as webrender::tiling::RenderTarget>::build
             at /media/chris/linux_data/dev/rust/gui/webrender/webrender/src/tiling.rs:1091
  15: <webrender::tiling::RenderTargetList<T>>::build
             at /media/chris/linux_data/dev/rust/gui/webrender/webrender/src/tiling.rs:983
  16: webrender::tiling::RenderPass::build
             at /media/chris/linux_data/dev/rust/gui/webrender/webrender/src/tiling.rs:1450
  17: webrender::frame_builder::FrameBuilder::build
             at /media/chris/linux_data/dev/rust/gui/webrender/webrender/src/frame_builder.rs:2358
  18: webrender::frame::Frame::build_frame::{{closure}}
             at /media/chris/linux_data/dev/rust/gui/webrender/webrender/src/frame.rs:1253
  19: <core::option::Option<T>>::map
             at /checkout/src/libcore/option.rs:398
  20: webrender::frame::Frame::build_frame
             at /media/chris/linux_data/dev/rust/gui/webrender/webrender/src/frame.rs:1252
  21: webrender::frame::Frame::build
             at /media/chris/linux_data/dev/rust/gui/webrender/webrender/src/frame.rs:1229
  22: webrender::render_backend::Document::render
             at /media/chris/linux_data/dev/rust/gui/webrender/webrender/src/render_backend.rs:106
  23: webrender::render_backend::RenderBackend::process_document
             at /media/chris/linux_data/dev/rust/gui/webrender/webrender/src/render_backend.rs:400
  24: webrender::render_backend::RenderBackend::run
             at /media/chris/linux_data/dev/rust/gui/webrender/webrender/src/render_backend.rs:468
  25: webrender::renderer::Renderer::new::{{closure}}
             at /media/chris/linux_data/dev/rust/gui/webrender/webrender/src/renderer.rs:1713
```

The problem is that there's no gpu cache location set for an `AlphaRenderItem::Primitive` inside of `AlphaRenderItem::add_to_batch`, because during the previous call to `PrimitiveStore::prepare_prim_for_render` which is supposed fill the gpu cache, the early return in this block can prevent the call to `gpu_cache.request`:

        let clip_task = if prim_clips.is_masking() {
            // Take into account the actual clip info of the primitive, and
            // mutate the current bounds accordingly.
            let mask_rect = match prim_clips.bounds.outer {
                Some(ref outer) => match prim_screen_rect.intersection(&outer.device_rect) {
                    Some(rect) => rect,
                    None => return None,
                },
                _ => prim_screen_rect,
            };

The change just makes sure the gpu location is set before adding the item. The `is_some` method lets the location be checked without exposing the `CacheLocation` type.

<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/webrender/1772)
<!-- Reviewable:end -->

Author: bors-servo

webrender/src/prim_store.rs

@@ -1148,7 +1148,10 @@ impl PrimitiveStore {
             let mask_rect = match prim_clips.bounds.outer {
                 Some(ref outer) => match prim_screen_rect.intersection(&outer.device_rect) {
                     Some(rect) => rect,
-                    None => return None,
+                    None => {
+                        metadata.screen_rect = None;
+                        return None;
+                    }
                 },
                 _ => prim_screen_rect,
             };