Fix un-antialiased rendering on Qualcomm

On Qualcomm we artificially disable support for pixel local storage because framebuffer fetch is slow. However, we weren't specifying a default sample count for this scenario, which caused us to use 1x "MSAA". Use 4x for now. If in the future we want 8x for lower resolution displays, we will have to get smarter about this.

Diffs=
09d9cd5cb Fix un-antialiased rendering on Qualcomm (#7270)

Co-authored-by: Chris Dalton <99840794+csmartdalton@users.noreply.github.com>

.rive_head

diff --git a/.rive_head b/.rive_head
index 23dbc07..063c254 100644
--- a/.rive_head
+++ b/.rive_head
@@ -1 +1 @@
-b9585c9ec382a5231a82d76b608a19f0ecc5a818
+09d9cd5cb661dc81833749d856856f0d4929a322

renderer/pls_render_context.cpp

diff --git a/renderer/pls_render_context.cpp b/renderer/pls_render_context.cpp
index 9f17286..ae183a1 100644
--- a/renderer/pls_render_context.cpp
+++ b/renderer/pls_render_context.cpp
@@ -219,10 +219,15 @@
     assert(frameDescriptor.renderTargetWidth > 0);
     assert(frameDescriptor.renderTargetHeight > 0);
     m_frameDescriptor = frameDescriptor;
-    if (m_frameDescriptor.msaaSampleCount > 0 || !platformFeatures().supportsPixelLocalStorage)
+    if (!platformFeatures().supportsPixelLocalStorage)
+    {
+        // Use 4x MSAA if we don't have pixel local storage and MSAA wasn't specified.
+        m_frameDescriptor.msaaSampleCount =
+            m_frameDescriptor.msaaSampleCount > 0 ? m_frameDescriptor.msaaSampleCount : 4;
+    }
+    if (m_frameDescriptor.msaaSampleCount > 0)
     {
         m_frameInterlockMode = pls::InterlockMode::depthStencil;
-        m_frameDescriptor.msaaSampleCount = std::max(m_frameDescriptor.msaaSampleCount, 1);
     }
     else if (m_frameDescriptor.disableRasterOrdering || !platformFeatures().supportsRasterOrdering)
     {