On Wed, Jun 10, 2015 at 3:44 PM, Daniel Drake <[email protected]> wrote: > On Wed, Jun 10, 2015 at 2:53 AM, Carlo Caione <[email protected]> wrote: >> +static int meson_mmc_map_dma(struct meson_mmc_host *host, >> + struct mmc_data *data, >> + unsigned int flags) >> +{ u32 i, dma_len; >> + struct scatterlist *sg; >> + >> + dma_len = dma_map_sg(mmc_dev(host->mmc), data->sg, data->sg_len, >> + ((data->flags & MMC_DATA_READ) ? >> + DMA_FROM_DEVICE : DMA_TO_DEVICE)); >> + if (dma_len == 0) { >> + dev_err(mmc_dev(host->mmc), "dma_map_sg failed\n"); >> + return -ENOMEM; >> + } >> + >> + for_each_sg(data->sg, sg, data->sg_len, i) { >> + if (sg->offset & 3 || sg->length & 3) { >> + dev_err(mmc_dev(host->mmc), >> + "unaligned scatterlist: os %x length %d\n", >> + sg->offset, sg->length); >> + return -EINVAL; >> + } >> + } > > If the offset/length check fails here, do you leak the DMA mapping? > If so, can you do the offset/length check first to avoid that?
Good idea. Fix in v3. > There should only ever be 1 entry in the scatterlist, right? In that > case you can do a simple check there, and then instead of looping, > just check the first entry in the scatterlist. Right. I'm going to wait a couple of days for more comments before submitting v3. Thanks, -- Carlo Caione | +39.340.80.30.096 | Endless -- To unsubscribe from this list: send the line "unsubscribe linux-mmc" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
