qsvenc: take only the allocated dimensions from the frames context

Other parameters, like the display size, should still be taken from the
codec context.

libavcodec/qsvenc.c

@@ -355,6 +355,9 @@ static int rc_supported(QSVEncContext *q)
 
 static int init_video_param(AVCodecContext *avctx, QSVEncContext *q)
 {
+    enum AVPixelFormat sw_format = avctx->pix_fmt == AV_PIX_FMT_QSV ?
+                                   avctx->sw_pix_fmt : avctx->pix_fmt;
+    const AVPixFmtDescriptor *desc;
     float quant;
     int ret;
 
@@ -380,15 +383,6 @@ static int init_video_param(AVCodecContext *avctx, QSVEncContext *q)
     q->param.mfx.EncodedOrder       = 0;
     q->param.mfx.BufferSizeInKB     = 0;
 
-    if (avctx->hw_frames_ctx) {
-        AVHWFramesContext *frames_ctx = (AVHWFramesContext*)avctx->hw_frames_ctx->data;
-        AVQSVFramesContext *frames_hwctx = frames_ctx->hwctx;
-        q->param.mfx.FrameInfo = frames_hwctx->surfaces[0].Info;
-    } else {
-        enum AVPixelFormat sw_format = avctx->pix_fmt == AV_PIX_FMT_QSV ?
-                                       avctx->sw_pix_fmt : avctx->pix_fmt;
-        const AVPixFmtDescriptor *desc;
-
     desc = av_pix_fmt_desc_get(sw_format);
     if (!desc)
         return AVERROR_BUG;
@@ -408,6 +402,12 @@ static int init_video_param(AVCodecContext *avctx, QSVEncContext *q)
     q->param.mfx.FrameInfo.BitDepthLuma   = desc->comp[0].depth;
     q->param.mfx.FrameInfo.BitDepthChroma = desc->comp[0].depth;
     q->param.mfx.FrameInfo.Shift          = desc->comp[0].depth > 8;
+
+    if (avctx->hw_frames_ctx) {
+        AVHWFramesContext *frames_ctx = (AVHWFramesContext*)avctx->hw_frames_ctx->data;
+        AVQSVFramesContext *frames_hwctx = frames_ctx->hwctx;
+        q->param.mfx.FrameInfo.Width  = frames_hwctx->surfaces[0].Info.Width;
+        q->param.mfx.FrameInfo.Height = frames_hwctx->surfaces[0].Info.Height;
     }
 
     if (avctx->framerate.den > 0 && avctx->framerate.num > 0) {