feat: tiled grayscale rendering to drop the storeBwBuffer peak (#2106)

Tiled grayscale rendering to drop the storeBwBuffer peak largest
contiguous free block (the value that actually drives OOM on the C3)
from ~114 KB to ~82-90 KB.

This renders each grayscale plane band-by-band into a small (~8 KB)
scratch and
streams each band straight to controller RAM (community-sdk
writeGrayscalePlaneStrip), leaving the BW framebuffer intact. No save,
no
restore; controller RAM is re-synced for the next differential turn
directly
from the live framebuffer.

Three writers honor the active band target so per-band re-rendering
stays cheap
and correct:

- drawPixel (text) redirects writes to the band scratch and clips to it.
- renderCharImpl skips glyphs whose physical y-extent is outside the
band before
the bitmap decode (glyphIntersectsStrip), so the per-band re-render
doesn't
  pay N x glyph decode.
- DirectPixelWriter (images) writes the band scratch via getWriteTarget
instead
  of the framebuffer. Without this, image pixels wrote the live BW frame
directly and cleanup re-synced that corruption, leaving thin outlines
after
  navigating away from an image.

Controller specifics live in the SDK (X4 setRamArea windowing, X3 PTL);
the
reader checks supportsStripGrayscale() and is otherwise
controller-agnostic.

Measured on hardware (X4 and X3, text and images, visually correct):

- Grayscale scratch ~8 KB vs ~50 KB save; largest contiguous free block
held at
  full size during grayscale instead of dropping ~25-32 KB.
- X4 text page about +25 ms/page; X3 page time is dominated by its
intrinsic
  grayscale refresh, not tiling.

Depends on community-sdk #13 (the writeGrayscalePlaneStrip API). The
submodule
bump here points at that branch, so until #13 merges the submodule won't
resolve
from upstream and CI will fail there; keeping this a draft until then.
Will
rebase onto master and re-point the submodule to the merged SDK commit
once #13
lands.

Did you use AI tools to help write this code? partial

lib/Epub/Epub/converters/DirectPixelWriter.h

@@ -16,6 +16,12 @@ struct DirectPixelWriter {
   uint8_t* fb;
   GfxRenderer::RenderMode mode;
   uint16_t displayWidthBytes;  // Runtime framebuffer stride (X4: 100, X3: 99)
+  // Active write target: for tiled grayscale, fb is the band scratch, originY is
+  // the band's top physical row, and clipRows is the band height. Off-band
+  // pixels are dropped. With no strip active these collapse to the full frame
+  // (originY 0, clipRows panelHeight) so the clip doubles as a bounds guard.
+  int originY;
+  int clipRows;
 
   // Orientation is collapsed into a linear transform:
   //   phyX = phyXBase + x * phyXStepX + y * phyXStepY
@@ -28,7 +34,9 @@ struct DirectPixelWriter {
   int rowPhyXBase, rowPhyYBase;
 
   void init(GfxRenderer& renderer) {
-    fb = renderer.getFrameBuffer();
+    fb = renderer.getWriteTarget();
+    originY = renderer.getWriteOriginY();
+    clipRows = renderer.getWriteRows();
     mode = renderer.getRenderMode();
     displayWidthBytes = renderer.getDisplayWidthBytes();
 
@@ -120,7 +128,12 @@ struct DirectPixelWriter {
     const int phyX = rowPhyXBase + logicalX * phyXStepX;
     const int phyY = rowPhyYBase + logicalX * phyYStepX;
 
-    const uint16_t byteIndex = phyY * displayWidthBytes + (phyX >> 3);
+    // Band-local row. The unsigned compare drops both off-band pixels (strip
+    // mode) and any out-of-frame row (full-frame mode) in one branch.
+    const int sy = phyY - originY;
+    if (static_cast<unsigned>(sy) >= static_cast<unsigned>(clipRows)) return;
+
+    const uint16_t byteIndex = static_cast<uint16_t>(sy * displayWidthBytes + (phyX >> 3));
     const uint8_t bitMask = 1 << (7 - (phyX & 7));
 
     if (state) {