For some client driver, it want to reduce latency between excuting previous packet command and next packet command, so append jump command to the end of previous packet and the jump destination address is the start address of next packet command buffer. Before next packet exist, the previous packet has no information of where to jump to, so append nop command first. When next packet exist, change nop command to jump command. For mediatek drm driver, it never has next packet, so appending nop command is redundant. Because cmdq_pkt_finalize() would append nop command, so change calling cmdq_pkt_finalize() to cmdq_pkt_eoc() to prevent append redundant nop command. Signed-off-by: Chun-Kuang Hu <chunkuang.hu@xxxxxxxxxx> --- drivers/gpu/drm/mediatek/mtk_crtc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/mediatek/mtk_crtc.c b/drivers/gpu/drm/mediatek/mtk_crtc.c index 6f34f573e127..a1976c3ae001 100644 --- a/drivers/gpu/drm/mediatek/mtk_crtc.c +++ b/drivers/gpu/drm/mediatek/mtk_crtc.c @@ -607,7 +607,7 @@ static void mtk_crtc_update_config(struct mtk_crtc *mtk_crtc, bool needs_vblank) cmdq_pkt_clear_event(cmdq_handle, mtk_crtc->cmdq_event); cmdq_pkt_wfe(cmdq_handle, mtk_crtc->cmdq_event, false); mtk_crtc_ddp_config(crtc, cmdq_handle); - cmdq_pkt_finalize(cmdq_handle); + cmdq_pkt_eoc(cmdq_handle); dma_sync_single_for_device(mtk_crtc->cmdq_client.chan->mbox->dev, cmdq_handle->pa_base, cmdq_handle->cmd_buf_size, -- 2.34.1
