On 3/27/2020 7:56 PM, Pierre-Louis Bossart wrote:
On 3/27/20 3:47 PM, Amadeusz Sławiński wrote:
Functions soc_tplg_denum_create, soc_tplg_dmixer_create, soc_tplg_dbytes_create can fail, so their return values should be checked and error should be propagated.
Signed-off-by: Amadeusz Sławiński amadeuszx.slawinski@linux.intel.com
v2: Added this patch
sound/soc/soc-topology.c | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-)
diff --git a/sound/soc/soc-topology.c b/sound/soc/soc-topology.c index c3811dd66b68..860bced933d6 100644 --- a/sound/soc/soc-topology.c +++ b/sound/soc/soc-topology.c @@ -1124,6 +1124,7 @@ static int soc_tplg_kcontrol_elems_load(struct soc_tplg *tplg, struct snd_soc_tplg_hdr *hdr) { struct snd_soc_tplg_ctl_hdr *control_hdr; + int ret; int i; if (tplg->pass != SOC_TPLG_PASS_MIXER) { @@ -1152,25 +1153,30 @@ static int soc_tplg_kcontrol_elems_load(struct soc_tplg *tplg, case SND_SOC_TPLG_CTL_RANGE: case SND_SOC_TPLG_DAPM_CTL_VOLSW: case SND_SOC_TPLG_DAPM_CTL_PIN: - soc_tplg_dmixer_create(tplg, 1, - le32_to_cpu(hdr->payload_size)); + ret = soc_tplg_dmixer_create(tplg, 1, + le32_to_cpu(hdr->payload_size)); break; case SND_SOC_TPLG_CTL_ENUM: case SND_SOC_TPLG_CTL_ENUM_VALUE: case SND_SOC_TPLG_DAPM_CTL_ENUM_DOUBLE: case SND_SOC_TPLG_DAPM_CTL_ENUM_VIRT: case SND_SOC_TPLG_DAPM_CTL_ENUM_VALUE: - soc_tplg_denum_create(tplg, 1, - le32_to_cpu(hdr->payload_size)); + ret = soc_tplg_denum_create(tplg, 1, + le32_to_cpu(hdr->payload_size)); break; case SND_SOC_TPLG_CTL_BYTES: - soc_tplg_dbytes_create(tplg, 1, - le32_to_cpu(hdr->payload_size)); + ret = soc_tplg_dbytes_create(tplg, 1, + le32_to_cpu(hdr->payload_size)); break; default: soc_bind_err(tplg, control_hdr, i); return -EINVAL; } + if (ret < 0) { + dev_err(tplg->dev, "ASoC: invalid control\n"); + return ret; + }
Sounds good, but this happens in a loop, so would all the memory previously allocated by denum/dbytes/dmixer_create leak, or is it freed automatically somewhere else?
Well, now that error is propagated, snd_soc_tplg_component_remove() should be called by snd_soc_tplg_component_load() in case of errors while parsing. From quick look it seems like it should be able to free it up correctly by calling remove_enum/bytes/mixer.
Thanks, Amadeusz