| Message ID | 20150706165737.2337.95787.stgit@tlendack-t1.amdoffice.net |
|---|---|
| State | Accepted, archived |
| Delegated to: | David Miller |
| Headers | show |
On Mon, 6 Jul 2015 09:57:37 -0700 Tom Lendacky <thomas.lendacky@amd.com> wrote: > When running a kernel configured with CONFIG_DMA_API_DEBUG=y a warning > is issued: > DMA-API: device driver tries to sync DMA memory it has not allocated > > This warning is the result of mapping the full range of the Rx buffer > pages allocated and then performing a dma_sync_single_for_cpu against > a calculated DMA address. The proper thing to do is to use the > dma_sync_single_range_for_cpu with a base DMA address and an offset. > > Reported-by: Kim Phillips <kim.phillips@arm.com> > Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com> > --- That works, thanks! Tested-by: Kim Phillips <kim.phillips@arm.com> Kim -- IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you. ARM Limited, Registered office 110 Fulbourn Road, Cambridge CB1 9NJ, Registered in England & Wales, Company No: 2557590 ARM Holdings plc, Registered office 110 Fulbourn Road, Cambridge CB1 9NJ, Registered in England & Wales, Company No: 2548782 -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
From: Tom Lendacky <thomas.lendacky@amd.com> Date: Mon, 6 Jul 2015 11:57:37 -0500 > When running a kernel configured with CONFIG_DMA_API_DEBUG=y a warning > is issued: > DMA-API: device driver tries to sync DMA memory it has not allocated > > This warning is the result of mapping the full range of the Rx buffer > pages allocated and then performing a dma_sync_single_for_cpu against > a calculated DMA address. The proper thing to do is to use the > dma_sync_single_range_for_cpu with a base DMA address and an offset. > > Reported-by: Kim Phillips <kim.phillips@arm.com> > Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com> Applied, thanks Tom. -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/drivers/net/ethernet/amd/xgbe/xgbe-desc.c b/drivers/net/ethernet/amd/xgbe/xgbe-desc.c index 661cdaa..b3bc87f 100644 --- a/drivers/net/ethernet/amd/xgbe/xgbe-desc.c +++ b/drivers/net/ethernet/amd/xgbe/xgbe-desc.c @@ -303,7 +303,8 @@ static void xgbe_set_buffer_data(struct xgbe_buffer_data *bd, get_page(pa->pages); bd->pa = *pa; - bd->dma = pa->pages_dma + pa->pages_offset; + bd->dma_base = pa->pages_dma; + bd->dma_off = pa->pages_offset; bd->dma_len = len; pa->pages_offset += len; diff --git a/drivers/net/ethernet/amd/xgbe/xgbe-dev.c b/drivers/net/ethernet/amd/xgbe/xgbe-dev.c index 506e832..a4473d8 100644 --- a/drivers/net/ethernet/amd/xgbe/xgbe-dev.c +++ b/drivers/net/ethernet/amd/xgbe/xgbe-dev.c @@ -1110,6 +1110,7 @@ static void xgbe_rx_desc_reset(struct xgbe_prv_data *pdata, unsigned int rx_usecs = pdata->rx_usecs; unsigned int rx_frames = pdata->rx_frames; unsigned int inte; + dma_addr_t hdr_dma, buf_dma; if (!rx_usecs && !rx_frames) { /* No coalescing, interrupt for every descriptor */ @@ -1129,10 +1130,12 @@ static void xgbe_rx_desc_reset(struct xgbe_prv_data *pdata, * Set buffer 2 (hi) address to buffer dma address (hi) and * set control bits OWN and INTE */ - rdesc->desc0 = cpu_to_le32(lower_32_bits(rdata->rx.hdr.dma)); - rdesc->desc1 = cpu_to_le32(upper_32_bits(rdata->rx.hdr.dma)); - rdesc->desc2 = cpu_to_le32(lower_32_bits(rdata->rx.buf.dma)); - rdesc->desc3 = cpu_to_le32(upper_32_bits(rdata->rx.buf.dma)); + hdr_dma = rdata->rx.hdr.dma_base + rdata->rx.hdr.dma_off; + buf_dma = rdata->rx.buf.dma_base + rdata->rx.buf.dma_off; + rdesc->desc0 = cpu_to_le32(lower_32_bits(hdr_dma)); + rdesc->desc1 = cpu_to_le32(upper_32_bits(hdr_dma)); + rdesc->desc2 = cpu_to_le32(lower_32_bits(buf_dma)); + rdesc->desc3 = cpu_to_le32(upper_32_bits(buf_dma)); XGMAC_SET_BITS_LE(rdesc->desc3, RX_NORMAL_DESC3, INTE, inte); diff --git a/drivers/net/ethernet/amd/xgbe/xgbe-drv.c b/drivers/net/ethernet/amd/xgbe/xgbe-drv.c index 1e9c28d..aae9d5e 100644 --- a/drivers/net/ethernet/amd/xgbe/xgbe-drv.c +++ b/drivers/net/ethernet/amd/xgbe/xgbe-drv.c @@ -1765,8 +1765,9 @@ static struct sk_buff *xgbe_create_skb(struct xgbe_prv_data *pdata, /* Start with the header buffer which may contain just the header * or the header plus data */ - dma_sync_single_for_cpu(pdata->dev, rdata->rx.hdr.dma, - rdata->rx.hdr.dma_len, DMA_FROM_DEVICE); + dma_sync_single_range_for_cpu(pdata->dev, rdata->rx.hdr.dma_base, + rdata->rx.hdr.dma_off, + rdata->rx.hdr.dma_len, DMA_FROM_DEVICE); packet = page_address(rdata->rx.hdr.pa.pages) + rdata->rx.hdr.pa.pages_offset; @@ -1778,8 +1779,11 @@ static struct sk_buff *xgbe_create_skb(struct xgbe_prv_data *pdata, len -= copy_len; if (len) { /* Add the remaining data as a frag */ - dma_sync_single_for_cpu(pdata->dev, rdata->rx.buf.dma, - rdata->rx.buf.dma_len, DMA_FROM_DEVICE); + dma_sync_single_range_for_cpu(pdata->dev, + rdata->rx.buf.dma_base, + rdata->rx.buf.dma_off, + rdata->rx.buf.dma_len, + DMA_FROM_DEVICE); skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags, rdata->rx.buf.pa.pages, @@ -1945,8 +1949,9 @@ read_again: if (!skb) error = 1; } else if (rdesc_len) { - dma_sync_single_for_cpu(pdata->dev, - rdata->rx.buf.dma, + dma_sync_single_range_for_cpu(pdata->dev, + rdata->rx.buf.dma_base, + rdata->rx.buf.dma_off, rdata->rx.buf.dma_len, DMA_FROM_DEVICE); diff --git a/drivers/net/ethernet/amd/xgbe/xgbe.h b/drivers/net/ethernet/amd/xgbe/xgbe.h index 63d72a1..717ce21 100644 --- a/drivers/net/ethernet/amd/xgbe/xgbe.h +++ b/drivers/net/ethernet/amd/xgbe/xgbe.h @@ -337,7 +337,8 @@ struct xgbe_buffer_data { struct xgbe_page_alloc pa; struct xgbe_page_alloc pa_unmap; - dma_addr_t dma; + dma_addr_t dma_base; + unsigned long dma_off; unsigned int dma_len; };
When running a kernel configured with CONFIG_DMA_API_DEBUG=y a warning is issued: DMA-API: device driver tries to sync DMA memory it has not allocated This warning is the result of mapping the full range of the Rx buffer pages allocated and then performing a dma_sync_single_for_cpu against a calculated DMA address. The proper thing to do is to use the dma_sync_single_range_for_cpu with a base DMA address and an offset. Reported-by: Kim Phillips <kim.phillips@arm.com> Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com> --- drivers/net/ethernet/amd/xgbe/xgbe-desc.c | 3 ++- drivers/net/ethernet/amd/xgbe/xgbe-dev.c | 11 +++++++---- drivers/net/ethernet/amd/xgbe/xgbe-drv.c | 17 +++++++++++------ drivers/net/ethernet/amd/xgbe/xgbe.h | 3 ++- 4 files changed, 22 insertions(+), 12 deletions(-) -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html