Hi Dan,
On 04/28/2017 01:28 PM, Dan Carpenter wrote:
Hello Arnaud Pouliquen,
The patch 76c2145ded6b: "ASoC: sti: Add CPU DAI driver for playback" from Jun 22, 2015, leads to the following static checker warning:
sound/soc/sti/uniperif_player.c:1077 uni_player_init() warn: unused return: ret = PTR_ERR()sound/soc/sti/uniperif_player.c 1047 int uni_player_init(struct platform_device *pdev, 1048 struct uniperif *player) 1049 { 1050 int ret = 0; 1051 1052 player->dev = &pdev->dev; 1053 player->state = UNIPERIF_STATE_STOPPED; 1054 player->dai_ops = &uni_player_dai_ops; 1055 1056 /* Get PCM_CLK_SEL & PCMP_VALID_SEL from audio-glue-ctrl SoC reg */ 1057 ret = uni_player_parse_dt_audio_glue(pdev, player); 1058 1059 if (ret < 0) { 1060 dev_err(player->dev, "Failed to parse DeviceTree\n"); 1061 return ret; 1062 } 1063 1064 /* Underflow recovery is only supported on later ip revisions */ 1065 if (player->ver >= SND_ST_UNIPERIF_VERSION_UNI_PLR_TOP_1_0) 1066 player->underflow_enabled = 1; 1067 1068 if (UNIPERIF_TYPE_IS_TDM(player)) 1069 player->hw = &uni_tdm_hw; 1070 else 1071 player->hw = &uni_player_pcm_hw; 1072 1073 /* Get uniperif resource */ 1074 player->clk = of_clk_get(pdev->dev.of_node, 0); 1075 if (IS_ERR(player->clk)) { 1076 dev_err(player->dev, "Failed to get clock\n"); 1077 ret = PTR_ERR(player->clk); ^^^^^^^^^^^^^^^^^^^^^^^^^^ This should probably be "return PTR_ERR(player->clk);? Can we run without a clk?
Yes you are right, thanks for highlighting the issue. the clock is mandatory.
Don't hesitate to propose the patch, i will ack it. Or tell me if you prefer that i fix it on my side...
Regards Arnaud