From: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
To: Radhey Shyam Pandey <radhey.shyam.pandey@amd.com>
Cc: Thinh Nguyen <Thinh.Nguyen@synopsys.com>,
"gregkh@linuxfoundation.org" <gregkh@linuxfoundation.org>,
"michal.simek@amd.com" <michal.simek@amd.com>,
"linux-usb@vger.kernel.org" <linux-usb@vger.kernel.org>,
"linux-arm-kernel@lists.infradead.org"
<linux-arm-kernel@lists.infradead.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
"git@amd.com" <git@amd.com>
Subject: Re: [PATCH] usb: dwc3: xilinx: set coherency mode for AMD versal adaptive platform
Date: Thu, 5 Jun 2025 00:32:44 +0000 [thread overview]
Message-ID: <20250605003240.f66u3jf4r4ft2qtr@synopsys.com> (raw)
In-Reply-To: <1748976818-710088-1-git-send-email-radhey.shyam.pandey@amd.com>
On Wed, Jun 04, 2025, Radhey Shyam Pandey wrote:
> If device is coherent or if DMA (direct memory access) is translated by
> an IOMMU then program USB2.0 IP to route transactions through the CCI for
> coherency even if the target of transaction is in low power domain (LPD).
>
> Signed-off-by: Radhey Shyam Pandey <radhey.shyam.pandey@amd.com>
> ---
> drivers/usb/dwc3/dwc3-xilinx.c | 34 ++++++++++++++++++++++------------
> 1 file changed, 22 insertions(+), 12 deletions(-)
>
> diff --git a/drivers/usb/dwc3/dwc3-xilinx.c b/drivers/usb/dwc3/dwc3-xilinx.c
> index 4ca7f6240d07..00e613c4d8c0 100644
> --- a/drivers/usb/dwc3/dwc3-xilinx.c
> +++ b/drivers/usb/dwc3/dwc3-xilinx.c
> @@ -32,6 +32,9 @@
> #define XLNX_USB_TRAFFIC_ROUTE_CONFIG 0x005C
> #define XLNX_USB_TRAFFIC_ROUTE_FPD 0x1
>
> +/* USB 2.0 IP Register */
> +#define XLNX_USB2_TRAFFIC_ROUTE_CONFIG 0x0044
> +
> #define XLNX_USB_FPD_PIPE_CLK 0x7c
> #define PIPE_CLK_DESELECT 1
> #define PIPE_CLK_SELECT 0
> @@ -66,6 +69,23 @@ static void dwc3_xlnx_mask_phy_rst(struct dwc3_xlnx *priv_data, bool mask)
> writel(reg, priv_data->regs + XLNX_USB_PHY_RST_EN);
> }
>
> +static void dwc3_xlnx_set_coherency(struct dwc3_xlnx *priv_data, u32 coherency_offset)
> +{
> + struct device *dev = priv_data->dev;
> + u32 reg;
> +
> + /*
> + * This routes the USB DMA traffic to go through FPD path instead
> + * of reaching DDR directly. This traffic routing is needed to
> + * make SMMU and CCI work with USB DMA.
> + */
> + if (of_dma_is_coherent(dev->of_node) || device_iommu_mapped(dev)) {
> + reg = readl(priv_data->regs + coherency_offset);
> + reg |= XLNX_USB_TRAFFIC_ROUTE_FPD;
> + writel(reg, priv_data->regs + coherency_offset);
> + }
> +}
> +
> static int dwc3_xlnx_init_versal(struct dwc3_xlnx *priv_data)
> {
> struct device *dev = priv_data->dev;
> @@ -92,6 +112,7 @@ static int dwc3_xlnx_init_versal(struct dwc3_xlnx *priv_data)
> }
>
> dwc3_xlnx_mask_phy_rst(priv_data, true);
> + dwc3_xlnx_set_coherency(priv_data, XLNX_USB2_TRAFFIC_ROUTE_CONFIG);
>
> return 0;
> }
> @@ -102,7 +123,6 @@ static int dwc3_xlnx_init_zynqmp(struct dwc3_xlnx *priv_data)
> struct reset_control *crst, *hibrst, *apbrst;
> struct gpio_desc *reset_gpio;
> int ret = 0;
> - u32 reg;
>
> priv_data->usb3_phy = devm_phy_optional_get(dev, "usb3-phy");
> if (IS_ERR(priv_data->usb3_phy)) {
> @@ -219,17 +239,7 @@ static int dwc3_xlnx_init_zynqmp(struct dwc3_xlnx *priv_data)
> usleep_range(5000, 10000);
> }
>
> - /*
> - * This routes the USB DMA traffic to go through FPD path instead
> - * of reaching DDR directly. This traffic routing is needed to
> - * make SMMU and CCI work with USB DMA.
> - */
> - if (of_dma_is_coherent(dev->of_node) || device_iommu_mapped(dev)) {
> - reg = readl(priv_data->regs + XLNX_USB_TRAFFIC_ROUTE_CONFIG);
> - reg |= XLNX_USB_TRAFFIC_ROUTE_FPD;
> - writel(reg, priv_data->regs + XLNX_USB_TRAFFIC_ROUTE_CONFIG);
> - }
> -
> + dwc3_xlnx_set_coherency(priv_data, XLNX_USB_TRAFFIC_ROUTE_CONFIG);
> err:
> return ret;
> }
>
> base-commit: a0bea9e39035edc56a994630e6048c8a191a99d8
> --
> 2.34.1
>
Acked-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
BR,
Thinh
prev parent reply other threads:[~2025-06-05 0:32 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-06-03 18:53 Radhey Shyam Pandey
2025-06-05 0:32 ` Thinh Nguyen [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20250605003240.f66u3jf4r4ft2qtr@synopsys.com \
--to=thinh.nguyen@synopsys.com \
--cc=git@amd.com \
--cc=gregkh@linuxfoundation.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-usb@vger.kernel.org \
--cc=michal.simek@amd.com \
--cc=radhey.shyam.pandey@amd.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®