[PDF] Fix shader fallback mapping when clipped

[PDF] Fix shader fallback mapping when clipped

The bounding box used for our pattern sizing is in device coordinates,
but SkPDFImageShader::Create() runs it through the inverse CTM to map
into shader space (with a notable effect of vertical flipping).

In order not to implode, the fallback shader path needs to compute its
adjustments in a consistent space.

BUG=skia:3398, chromium:454591, chromium:455052
R=halcanary@google.com

Committed: https://skia.googlesource.com/skia/+/7b4d4c7a4668f771641aaed88b83d22d5efe6873

[f(malita), 2015-02-05 23:16:27]

My head is spinning from all the nasty voodoo going on in
SkPDFImageShader::Create(), but I think I finally found the magic tweaks
required to make it grok our synthetic image shader.

All tests pass locally:
* GM:tiled_picture_shader
* https://code.google.com/p/chromium/issues/detail?id=454591
* https://code.google.com/p/chromium/issues/detail?id=455052

[f(malita), 2015-02-06 23:59:12]

PTAL

[f(malita), 2015-02-10 21:58:54]

fmalita@chromium.org changed reviewers:
+ reed@google.com, robertphillips@google.com

[f(malita), 2015-02-12 21:03:12]

The CQ bit was checked by fmalita@chromium.org

[commit-bot: I haz the power, 2015-02-12 21:03:48]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/901303003/1

[commit-bot: I haz the power, 2015-02-12 21:03:49]

Note for Reviewers:
The CQ is waiting for an approval.
If you believe that the CL is not ready yet, or if you would like to L-G-T-M
with comments then please uncheck the CQ checkbox.

Waiting for LGTM from valid reviewer(s) till 2015-02-13 03:03 UTC

[robertphillips, 2015-02-12 21:12:59]

I defer to Hal for the real l g t m

https://codereview.chromium.org/901303003/diff/1/src/pdf/SkPDFShader.cpp
File src/pdf/SkPDFShader.cpp (right):

https://codereview.chromium.org/901303003/diff/1/src/pdf/SkPDFShader.cpp#newc...
src/pdf/SkPDFShader.cpp:1225: SkRect shaderRect = SkRect::Make(bbox);
inverse_transform_bbox ?

[hal.canary, 2015-02-12 21:13:06]

On 2015/02/12 21:03:49, I haz the power (commit-bot) wrote:
> Note for Reviewers:
> The CQ is waiting for an approval.
> If you believe that the CL is not ready yet, or if you would like to L-G-T-M
> with comments then please uncheck the CQ checkbox.
> 
> Waiting for LGTM from valid reviewer(s) till 2015-02-13 03:03 UTC

I have verified that this patch renders tiled_picture_shader correctly. LGTM

[f(malita), 2015-02-12 21:13:56]

The CQ bit was unchecked by fmalita@chromium.org

[f(malita), 2015-02-12 21:24:34]

New patchsets have been uploaded after l-g-t-m from halcanary@google.com

[f(malita), 2015-02-12 21:28:33]

Thanks!

https://codereview.chromium.org/901303003/diff/1/src/pdf/SkPDFShader.cpp
File src/pdf/SkPDFShader.cpp (right):

https://codereview.chromium.org/901303003/diff/1/src/pdf/SkPDFShader.cpp#newc...
src/pdf/SkPDFShader.cpp:1225: SkRect shaderRect = SkRect::Make(bbox);
On 2015/02/12 21:12:59, robertphillips wrote:
> inverse_transform_bbox ?

Done.

[f(malita), 2015-02-12 21:28:40]

The CQ bit was checked by fmalita@chromium.org

[commit-bot: I haz the power, 2015-02-12 21:29:04]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/901303003/40001

[commit-bot: I haz the power, 2015-02-12 21:56:55]

Committed patchset #3 (id:40001) as
https://skia.googlesource.com/skia/+/7b4d4c7a4668f771641aaed88b83d22d5efe6873