Fix #158269: Inconsistency between framebuffer default size and size

Due to an inconsistency in MTLFramebuffer `default_size` and its current
`size` could show an incorrect error message when reading from the
framebuffer using Python.

This has been fixed by making sure by disconnecting the default size in
Metal and base class. Sizes are set when needed. There is still
inconsistency when running in background.

Pull Request: https://projects.blender.org/blender/blender/pulls/158586

tests/python/bl_pyapi_gpu.py

@@ -21,6 +21,23 @@ class TestGpuInit(unittest.TestCase):
         gpu.types.GPUTexture(size=(1024, 1024), format="RGBA8")
 
 
+class TestGpuFrameBuffer(unittest.TestCase):
+    def test_read_color_bounds_check(self):
+        gpu.init()
+
+        # Setup a framebuffer.
+        tex = gpu.types.GPUTexture(size=(16, 16), format="RGBA8")
+        fb = gpu.types.GPUFrameBuffer(color_slots=[tex])
+
+        with fb.bind():
+            # Reading within bounds should succeed.
+            buf = fb.read_color(0, 0, 1, 1, 4, 0, "UBYTE")
+
+            # Reading outside bounds should raise ValueError.
+            with self.assertRaises(ValueError):
+                fb.read_color(-1, 0, 1, 1, 4, 0, "UBYTE")
+
+
 if __name__ == "__main__":
     import sys