On 03/22/2017 05:14 AM, Dan Carpenter wrote:
> Hello Eric Anholt,
>
> This is a semi-automatic email about new static checker warnings.
>
> The patch 5aa25d971ab4: "mmc: bcm2835: Add new driver for the sdhost
> controller." from Mar 8, 2017, leads to the following Smatch
> complaint:
>
> drivers/mmc/host/bcm2835.c:1203 bcm2835_request()
> error: we previously assumed 'host->mrq->data' could be null (see line 1186)
>
> drivers/mmc/host/bcm2835.c
> 1156 static void bcm2835_request(struct mmc_host *mmc, struct mmc_request *mrq)
> 1157 {
> 1158 struct bcm2835_host *host = mmc_priv(mmc);
> 1159 struct device *dev = &host->pdev->dev;
> 1160 u32 edm, fsm;
> 1161
> 1162 /* Reset the error statuses in case this is a retry */
> 1163 if (mrq->sbc)
> 1164 mrq->sbc->error = 0;
> 1165 if (mrq->cmd)
> 1166 mrq->cmd->error = 0;
> 1167 if (mrq->data)
> 1168 mrq->data->error = 0;
> 1169 if (mrq->stop)
> 1170 mrq->stop->error = 0;
> 1171
> 1172 if (mrq->data && !is_power_of_2(mrq->data->blksz)) {
> ^^^^^^^^^
> Tested here.
>
> 1173 dev_err(dev, "unsupported block size (%d bytes)\n",
> 1174 mrq->data->blksz);
> 1175 mrq->cmd->error = -EINVAL;
> 1176 mmc_request_done(mmc, mrq);
> 1177 return;
> 1178 }
> 1179
> 1180 if (host->use_dma && mrq->data && (mrq->data->blocks > PIO_THRESHOLD))
> ^^^^^^^^^
> And here.
>
> 1181 bcm2835_prepare_dma(host, mrq->data);
> 1182
> 1183 mutex_lock(&host->mutex);
> 1184
> 1185 WARN_ON(host->mrq);
> 1186 host->mrq = mrq;
>
> Now host->mrq->data is possibly NULL.
>
> 1187
> 1188 edm = readl(host->ioaddr + SDEDM);
> 1189 fsm = edm & SDEDM_FSM_MASK;
> 1190
> 1191 if ((fsm != SDEDM_FSM_IDENTMODE) &&
> 1192 (fsm != SDEDM_FSM_DATAMODE)) {
> 1193 dev_err(dev, "previous command (%d) not complete (EDM %08x)\n",
> 1194 readl(host->ioaddr + SDCMD) & SDCMD_CMD_MASK,
> 1195 edm);
> 1196 bcm2835_dumpregs(host);
> 1197 mrq->cmd->error = -EILSEQ;
> 1198 bcm2835_finish_request(host);
> 1199 mutex_unlock(&host->mutex);
> 1200 return;
> 1201 }
> 1202
> 1203 host->use_sbc = !!mrq->sbc && (host->mrq->data->flags & MMC_DATA_READ);
> ^^^^^^^^^^^^^^^^^
> Dereferenced without a test.
I think that it can be just added the condition checking for host->mrq->data.
Best Regards,
Jaehoon Chung
>
> 1204 if (host->use_sbc) {
> 1205 if (bcm2835_send_command(host, mrq->sbc)) {
>
> regards,
> dan carpenter
> --
> To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
>
--
To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
