Hi,

Static analysis on linux-next with Coverity had detected a potential
array out-of-bounds write issue in the following commit:

commit aa2e2785545aab21b6cb2e23f111ae0751cbcca7
Author: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Date:   Mon Oct 26 17:09:47 2020 +0000

    ASoC: qcom: sm8250: add sound card qrb5165-rb5 support

The analysis is as follows:

139 static int sm8250_snd_hw_free(struct snd_pcm_substream *substream)
140 {
141        struct snd_soc_pcm_runtime *rtd = substream->private_data;
142        struct sm8250_snd_data *data =
snd_soc_card_get_drvdata(rtd->card);
143        struct snd_soc_dai *cpu_dai = asoc_rtd_to_cpu(rtd, 0);
144        struct sdw_stream_runtime *sruntime =
data->sruntime[cpu_dai->id];
145

   1. Switch case value 105.

146        switch (cpu_dai->id) {

   2. equality_cond: Jumping to case 105.

147        case WSA_CODEC_DMA_RX_0:
148        case WSA_CODEC_DMA_RX_1:

   Out-of-bounds write (OVERRUN)
   3. Condition sruntime, taking true branch.
   4. Condition data->stream_prepared[cpu_dai->id], taking true branch.

149                if (sruntime && data->stream_prepared[cpu_dai->id]) {
150                        sdw_disable_stream(sruntime);
151                        sdw_deprepare_stream(sruntime);

   Out-of-bounds write (OVERRUN)
   5. overrun-local: Overrunning array data->stream_prepared of 16 bytes
at byte offset 105 using index cpu_dai->id (which evaluates to 105).

152                        data->stream_prepared[cpu_dai->id] = false;
153                }
154                break;
155        default:
156                break;
157        }
158
159        return 0;
160 }

So cpu_dia->id is 105 in this case statement, and yet
data->steam_prepared is an array of 16 elements, so this looks suspect.

Colin