Fix in-place conversion of certain QImage formats

In the copying/transcription of the generic converter to the generic in-place converter, a test for Format_RGB32 destination format was mistakenly replaced with test of source format. The result was that a suboptimal pixel store function was selected, and the resulting image data could end up with non-0xff in the unused alpha field. Task-number: QTBUG-132051 Change-Id: If3ebf5fdd7ab6e377c8ad479ea38ce665f922b7c Reviewed-by: Allan Sandfeld Jensen <[email protected]> (cherry picked from commit 0ace5ba) Reviewed-by: Qt Cherry-pick Bot <[email protected]> (cherry picked from commit 4d00967)
qt · Dec 14, 2024 · 933128d · 933128d
1 parent bde5851
commit 933128d
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/src/gui/image/qimage_conversions.cpp b/src/gui/image/qimage_conversions.cpp
@@ -385,7 +385,7 @@ bool convert_generic_inplace(QImageData *data, QImage::Format dst_format, Qt::Im
             !destLayout->hasAlphaChannel && destLayout->storeFromRGB32) {
         // Avoid unnecessary premultiply and unpremultiply when converting from unpremultiplied src format.
         fetch = qPixelLayouts[qt_toPremultipliedFormat(data->format)].fetchToARGB32PM;
-        if (data->format == QImage::Format_RGB32)
+        if (dst_format == QImage::Format_RGB32)
             store = storeRGB32FromARGB32;
         else
             store = destLayout->storeFromRGB32;