Properly recompute SVG font size under zoom

Properly recompute SVG font size under zoom

This patch fixes a regression from [1] where an optimization in
RenderSVGInlineText::computeNewScaledFontForStyle would skip updating
the font size on lowdpi devices while zooming. The computed font size
calculation depends on zoom and the font scaling factor depends on the
device's scaling factor. Previously, we would skip updating the font
when the scaling factor was 1 which would fail to account for zoom.
This patch updates the optimization to only be used when both the
scaling factor and zoom are 1.

[1] http://src.chromium.org/viewvc/blink?revision=153683&view=revision

BUG=405468
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=180995

[pdr., 2014-08-27 03:32:47]

pdr@chromium.org changed reviewers:
+ fmalita@google.com, fs@opera.com

[f(malita), 2014-08-27 12:44:32]

fmalita@chromium.org changed reviewers:
+ fmalita@chromium.org

[f(malita), 2014-08-27 12:44:32]

lgtm

https://codereview.chromium.org/508823002/diff/1/Source/core/rendering/svg/Re...
File Source/core/rendering/svg/RenderSVGInlineText.cpp (left):

https://codereview.chromium.org/508823002/diff/1/Source/core/rendering/svg/Re...
Source/core/rendering/svg/RenderSVGInlineText.cpp:223: if (scalingFactor == 1 ||
!scalingFactor) {
Not related to this CL, but do you know why we reset 0.0 -> 1.0? It seems odd
that at the low end of our range, the font snaps back to visible.

[pdr., 2014-08-27 18:09:51]

https://codereview.chromium.org/508823002/diff/1/Source/core/rendering/svg/Re...
File Source/core/rendering/svg/RenderSVGInlineText.cpp (left):

https://codereview.chromium.org/508823002/diff/1/Source/core/rendering/svg/Re...
Source/core/rendering/svg/RenderSVGInlineText.cpp:223: if (scalingFactor == 1 ||
!scalingFactor) {
On 2014/08/27 12:44:31, Florin Malita wrote:
> Not related to this CL, but do you know why we reset 0.0 -> 1.0? It seems odd
> that at the low end of our range, the font snaps back to visible.

I agree, this is weird.

I tried converting this to an assert:
    ASSERT(scalingFactor);
    if (style->effectiveZoom() == 1 && scalingFactor == 1)

Unfortunately it looks like scalingFactor will sometimes actually be zero. This
is probably a bug since scalingFactor is effectively the full ctm. This code has
some gross bugs and does not work properly with the compositing update :/

[pdr., 2014-08-27 18:09:55]

The CQ bit was checked by pdr@chromium.org

[commit-bot: I haz the power, 2014-08-27 18:10:50]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/pdr@chromium.org/508823002/1

[commit-bot: I haz the power, 2014-08-27 18:49:28]

Committed patchset #1 (id:1) as 180995