| Message ID | E1gWjE3-0006mG-2S@rmk-PC.armlinux.org.uk (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | mmc: omap_hsmmc: fix DMA API warning | expand |
Hi, * Russell King <rmk+kernel@armlinux.org.uk> [181211 14:41]: > @@ -1939,6 +1938,17 @@ static int omap_hsmmc_probe(struct platform_device *pdev) > goto err_irq; > } > > + /* > + * Limit the maximum segment size to the lower of the request size > + * and the DMA engine device segment size limits. In reality, with > + * 32-bit transfers, the DMA engine can do longer segments than this > + * but there is no way to represent that in the DMA model - if we > + * increase this figure here, we get warnings from the DMA API debug. > + */ > + mmc->max_seg_size = min(mmc->max_req_size, > + min(dma_get_max_seg_size(host->rx_chan->device->dev), > + dma_get_max_seg_size(host->tx_chan->device->dev))); > + Looks like using min3() here would be handy? Regards, Tony
On Tue, 11 Dec 2018 at 15:41, Russell King <rmk+kernel@armlinux.org.uk> wrote: > > While booting with rootfs on MMC, the following warning is encountered > on OMAP4430: > > omap-dma-engine 4a056000.dma-controller: DMA-API: mapping sg segment longer than device claims to support [len=69632] [max=65536] > > This is because the DMA engine has a default maximum segment size of 64K > but HSMMC sets: > > mmc->max_blk_size = 512; /* Block Length at max can be 1024 */ > mmc->max_blk_count = 0xFFFF; /* No. of Blocks is 16 bits */ > mmc->max_req_size = mmc->max_blk_size * mmc->max_blk_count; > mmc->max_seg_size = mmc->max_req_size; > > which ends up telling the block layer that we support a maximum segment > size of 65535*512, which exceeds the advertised DMA engine capabilities. > > Fix this by clamping the maximum segment size to the lower of the > maximum request size and of the DMA engine device used for either DMA > channel. > > Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk> I have applied this for fixes, amending the change by modifying the min-min check to min3 according to the suggestion from Tony. I assumed adding a stable tag makes sense so I did, however, please tell if you want me to drop that. Thanks and kind regards Uffe > --- > drivers/mmc/host/omap_hsmmc.c | 12 +++++++++++- > 1 file changed, 11 insertions(+), 1 deletion(-) > > diff --git a/drivers/mmc/host/omap_hsmmc.c b/drivers/mmc/host/omap_hsmmc.c > index 467d889a1638..55100974fb95 100644 > --- a/drivers/mmc/host/omap_hsmmc.c > +++ b/drivers/mmc/host/omap_hsmmc.c > @@ -1909,7 +1909,6 @@ static int omap_hsmmc_probe(struct platform_device *pdev) > mmc->max_blk_size = 512; /* Block Length at max can be 1024 */ > mmc->max_blk_count = 0xFFFF; /* No. of Blocks is 16 bits */ > mmc->max_req_size = mmc->max_blk_size * mmc->max_blk_count; > - mmc->max_seg_size = mmc->max_req_size; > > mmc->caps |= MMC_CAP_MMC_HIGHSPEED | MMC_CAP_SD_HIGHSPEED | > MMC_CAP_WAIT_WHILE_BUSY | MMC_CAP_ERASE | MMC_CAP_CMD23; > @@ -1939,6 +1938,17 @@ static int omap_hsmmc_probe(struct platform_device *pdev) > goto err_irq; > } > > + /* > + * Limit the maximum segment size to the lower of the request size > + * and the DMA engine device segment size limits. In reality, with > + * 32-bit transfers, the DMA engine can do longer segments than this > + * but there is no way to represent that in the DMA model - if we > + * increase this figure here, we get warnings from the DMA API debug. > + */ > + mmc->max_seg_size = min(mmc->max_req_size, > + min(dma_get_max_seg_size(host->rx_chan->device->dev), > + dma_get_max_seg_size(host->tx_chan->device->dev))); > + > /* Request IRQ for MMC operations */ > ret = devm_request_irq(&pdev->dev, host->irq, omap_hsmmc_irq, 0, > mmc_hostname(mmc), host); > -- > 2.7.4 >
diff --git a/drivers/mmc/host/omap_hsmmc.c b/drivers/mmc/host/omap_hsmmc.c index 467d889a1638..55100974fb95 100644 --- a/drivers/mmc/host/omap_hsmmc.c +++ b/drivers/mmc/host/omap_hsmmc.c @@ -1909,7 +1909,6 @@ static int omap_hsmmc_probe(struct platform_device *pdev) mmc->max_blk_size = 512; /* Block Length at max can be 1024 */ mmc->max_blk_count = 0xFFFF; /* No. of Blocks is 16 bits */ mmc->max_req_size = mmc->max_blk_size * mmc->max_blk_count; - mmc->max_seg_size = mmc->max_req_size; mmc->caps |= MMC_CAP_MMC_HIGHSPEED | MMC_CAP_SD_HIGHSPEED | MMC_CAP_WAIT_WHILE_BUSY | MMC_CAP_ERASE | MMC_CAP_CMD23; @@ -1939,6 +1938,17 @@ static int omap_hsmmc_probe(struct platform_device *pdev) goto err_irq; } + /* + * Limit the maximum segment size to the lower of the request size + * and the DMA engine device segment size limits. In reality, with + * 32-bit transfers, the DMA engine can do longer segments than this + * but there is no way to represent that in the DMA model - if we + * increase this figure here, we get warnings from the DMA API debug. + */ + mmc->max_seg_size = min(mmc->max_req_size, + min(dma_get_max_seg_size(host->rx_chan->device->dev), + dma_get_max_seg_size(host->tx_chan->device->dev))); + /* Request IRQ for MMC operations */ ret = devm_request_irq(&pdev->dev, host->irq, omap_hsmmc_irq, 0, mmc_hostname(mmc), host);
While booting with rootfs on MMC, the following warning is encountered on OMAP4430: omap-dma-engine 4a056000.dma-controller: DMA-API: mapping sg segment longer than device claims to support [len=69632] [max=65536] This is because the DMA engine has a default maximum segment size of 64K but HSMMC sets: mmc->max_blk_size = 512; /* Block Length at max can be 1024 */ mmc->max_blk_count = 0xFFFF; /* No. of Blocks is 16 bits */ mmc->max_req_size = mmc->max_blk_size * mmc->max_blk_count; mmc->max_seg_size = mmc->max_req_size; which ends up telling the block layer that we support a maximum segment size of 65535*512, which exceeds the advertised DMA engine capabilities. Fix this by clamping the maximum segment size to the lower of the maximum request size and of the DMA engine device used for either DMA channel. Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk> --- drivers/mmc/host/omap_hsmmc.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-)