Re: [PATCH V3 XRT Alveo 13/18] fpga: xrt: devctl platform driver

From: Lizhi Hou
Date: Tue Mar 16 2021 - 19:56:36 EST

Next message: Florent Revest: "Re: [PATCH bpf-next 1/5] bpf: Add a ARG_PTR_TO_CONST_STR argument type"
Previous message: Guenter Roeck: "Re: [PATCH v4 RESEND net-next] net: socket: use BIT() for MSG_*"
Next in thread: Tom Rix: "Re: [PATCH V3 XRT Alveo 13/18] fpga: xrt: devctl platform driver"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On 03/04/2021 05:39 AM, Tom Rix wrote:

CAUTION: This message has originated from an External Source. Please use proper judgment and caution when opening attachments, clicking links, or responding to this email.

On 2/17/21 10:40 PM, Lizhi Hou wrote:

Add devctl driver. devctl is a type of hardware function which only has
few registers to read or write. They are discovered by walking firmware
metadata. A platform device node will be created for them.

Signed-off-by: Sonal Santan <sonal.santan@xxxxxxxxxx>
Signed-off-by: Max Zhen <max.zhen@xxxxxxxxxx>
Signed-off-by: Lizhi Hou <lizhih@xxxxxxxxxx>
---
drivers/fpga/xrt/include/xleaf/devctl.h | 43 +++++
drivers/fpga/xrt/lib/xleaf/devctl.c | 206 ++++++++++++++++++++++++
2 files changed, 249 insertions(+)
create mode 100644 drivers/fpga/xrt/include/xleaf/devctl.h
create mode 100644 drivers/fpga/xrt/lib/xleaf/devctl.c

diff --git a/drivers/fpga/xrt/include/xleaf/devctl.h b/drivers/fpga/xrt/include/xleaf/devctl.h
new file mode 100644
index 000000000000..96a40e066f83
--- /dev/null
+++ b/drivers/fpga/xrt/include/xleaf/devctl.h
@@ -0,0 +1,43 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Header file for XRT DEVCTL Leaf Driver
+ *
+ * Copyright (C) 2020-2021 Xilinx, Inc.
+ *
+ * Authors:
+ * Lizhi Hou <Lizhi.Hou@xxxxxxxxxx>
+ */
+
+#ifndef _XRT_DEVCTL_H_
+#define _XRT_DEVCTL_H_
+
+#include "xleaf.h"
+
+/*
+ * DEVCTL driver IOCTL calls.
+ */
+enum xrt_devctl_ioctl_cmd {
+ XRT_DEVCTL_READ = XRT_XLEAF_CUSTOM_BASE, /* See comments in xleaf.h */
+ XRT_DEVCTL_WRITE,
+};
+
+enum xrt_devctl_id {
+ XRT_DEVCTL_ROM_UUID,

Assumes 0, should make this explicit and initialize to 0

Sure.

+ XRT_DEVCTL_DDR_CALIB,
+ XRT_DEVCTL_GOLDEN_VER,
+ XRT_DEVCTL_MAX
+};
+
+struct xrt_devctl_ioctl_rw {
+ u32 xgir_id;
+ void *xgir_buf;
+ u32 xgir_len;
+ u32 xgir_offset;

similar to other patches, the xgir_ prefix is not needed

+};
+
+struct xrt_devctl_ioctl_intf_uuid {
+ u32 xgir_uuid_num;
+ uuid_t *xgir_uuids;
+};
+
+#endif /* _XRT_DEVCTL_H_ */
diff --git a/drivers/fpga/xrt/lib/xleaf/devctl.c b/drivers/fpga/xrt/lib/xleaf/devctl.c
new file mode 100644
index 000000000000..caf8c6569f0f
--- /dev/null
+++ b/drivers/fpga/xrt/lib/xleaf/devctl.c
@@ -0,0 +1,206 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Xilinx Alveo FPGA devctl Driver
+ *
+ * Copyright (C) 2020-2021 Xilinx, Inc.
+ *
+ * Authors:
+ * Lizhi Hou<Lizhi.Hou@xxxxxxxxxx>
+ */
+
+#include <linux/mod_devicetable.h>
+#include <linux/platform_device.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <linux/io.h>
+#include "metadata.h"
+#include "xleaf.h"
+#include "xleaf/devctl.h"
+
+#define XRT_DEVCTL "xrt_devctl"
+
+struct xrt_name_id {
+ char *ep_name;
+ int id;
+};
+
+static struct xrt_name_id name_id[XRT_DEVCTL_MAX] = {
+ { XRT_MD_NODE_BLP_ROM, XRT_DEVCTL_ROM_UUID },
+ { XRT_MD_NODE_GOLDEN_VER, XRT_DEVCTL_GOLDEN_VER },

DDR_CALIB is unused ?

Not sure if I understand the question correctly. ddr_calib will have more things need to handle other than just read/write.

+};
+
+struct xrt_devctl {
+ struct platform_device *pdev;
+ void __iomem *base_addrs[XRT_DEVCTL_MAX];
+ ulong sizes[XRT_DEVCTL_MAX];
+};

similar to other patches, why not use regmap ?

Will change to regmap.

+
+static int xrt_devctl_name2id(struct xrt_devctl *devctl, const char *name)
+{
+ int i;
+
+ for (i = 0; i < XRT_DEVCTL_MAX && name_id[i].ep_name; i++) {
+ if (!strncmp(name_id[i].ep_name, name, strlen(name_id[i].ep_name) + 1))
+ return name_id[i].id;
+ }
+
+ return -EINVAL;
+}
+
+static int
+xrt_devctl_leaf_ioctl(struct platform_device *pdev, u32 cmd, void *arg)
+{
+ struct xrt_devctl *devctl;
+ int ret = 0;
+
+ devctl = platform_get_drvdata(pdev);
+
+ switch (cmd) {
+ case XRT_XLEAF_EVENT:
+ /* Does not handle any event. */
+ break;
+ case XRT_DEVCTL_READ: {
+ struct xrt_devctl_ioctl_rw *rw_arg = arg;
+ u32 *p_src, *p_dst, i;
+
+ if (rw_arg->xgir_len & 0x3) {
+ xrt_err(pdev, "invalid len %d", rw_arg->xgir_len);
+ return -EINVAL;
+ }
+
+ if (rw_arg->xgir_id >= XRT_DEVCTL_MAX) {
+ xrt_err(pdev, "invalid id %d", rw_arg->xgir_id);
+ return -EINVAL;
+ }

needs a < 0 check ?

change xgir_id to u32.

+
+ p_src = devctl->base_addrs[rw_arg->xgir_id];
+ if (!p_src) {
+ xrt_err(pdev, "io not found, id %d",
+ rw_arg->xgir_id);
+ return -EINVAL;
+ }
+ if (rw_arg->xgir_offset + rw_arg->xgir_len >
+ devctl->sizes[rw_arg->xgir_id]) {
+ xrt_err(pdev, "invalid argument, off %d, len %d",
+ rw_arg->xgir_offset, rw_arg->xgir_len);
+ return -EINVAL;
+ }
+ p_dst = rw_arg->xgir_buf;
+ for (i = 0; i < rw_arg->xgir_len / sizeof(u32); i++) {
+ u32 val = ioread32(p_src + rw_arg->xgir_offset + i);
+
+ memcpy(p_dst + i, &val, sizeof(u32));
+ }
+ break;
+ }

The _WRITE msg is not handled Then why have it ?

Will remove write msg from this patch set and add it back in future patches.

Thanks,
Lizhi

Tom

+ default:
+ xrt_err(pdev, "unsupported cmd %d", cmd);
+ return -EINVAL;
+ }
+
+ return ret;
+}
+
+static int xrt_devctl_remove(struct platform_device *pdev)
+{
+ struct xrt_devctl *devctl;
+ int i;
+
+ devctl = platform_get_drvdata(pdev);
+
+ for (i = 0; i < XRT_DEVCTL_MAX; i++) {
+ if (devctl->base_addrs[i])
+ iounmap(devctl->base_addrs[i]);
+ }
+
+ platform_set_drvdata(pdev, NULL);
+ devm_kfree(&pdev->dev, devctl);
+
+ return 0;
+}
+
+static int xrt_devctl_probe(struct platform_device *pdev)
+{
+ struct xrt_devctl *devctl;
+ int i, id, ret = 0;
+ struct resource *res;
+
+ devctl = devm_kzalloc(&pdev->dev, sizeof(*devctl), GFP_KERNEL);
+ if (!devctl)
+ return -ENOMEM;
+
+ devctl->pdev = pdev;
+ platform_set_drvdata(pdev, devctl);
+
+ xrt_info(pdev, "probing...");
+ for (i = 0, res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+ res;
+ res = platform_get_resource(pdev, IORESOURCE_MEM, ++i)) {
+ id = xrt_devctl_name2id(devctl, res->name);
+ if (id < 0) {
+ xrt_err(pdev, "ep %s not found", res->name);
+ continue;
+ }
+ devctl->base_addrs[id] = ioremap(res->start, res->end - res->start + 1);
+ if (!devctl->base_addrs[id]) {
+ xrt_err(pdev, "map base failed %pR", res);
+ ret = -EIO;
+ goto failed;
+ }
+ devctl->sizes[id] = res->end - res->start + 1;
+ }
+
+failed:
+ if (ret)
+ xrt_devctl_remove(pdev);
+
+ return ret;
+}
+
+static struct xrt_subdev_endpoints xrt_devctl_endpoints[] = {
+ {
+ .xse_names = (struct xrt_subdev_ep_names[]) {
+ /* add name if ep is in same partition */
+ { .ep_name = XRT_MD_NODE_BLP_ROM },
+ { NULL },
+ },
+ .xse_min_ep = 1,
+ },
+ {
+ .xse_names = (struct xrt_subdev_ep_names[]) {
+ { .ep_name = XRT_MD_NODE_GOLDEN_VER },
+ { NULL },
+ },
+ .xse_min_ep = 1,
+ },
+ /* adding ep bundle generates devctl device instance */
+ { 0 },
+};
+
+static struct xrt_subdev_drvdata xrt_devctl_data = {
+ .xsd_dev_ops = {
+ .xsd_ioctl = xrt_devctl_leaf_ioctl,
+ },
+};
+
+static const struct platform_device_id xrt_devctl_table[] = {
+ { XRT_DEVCTL, (kernel_ulong_t)&xrt_devctl_data },
+ { },
+};
+
+static struct platform_driver xrt_devctl_driver = {
+ .driver = {
+ .name = XRT_DEVCTL,
+ },
+ .probe = xrt_devctl_probe,
+ .remove = xrt_devctl_remove,
+ .id_table = xrt_devctl_table,
+};
+
+void devctl_leaf_init_fini(bool init)
+{
+ if (init)
+ xleaf_register_driver(XRT_SUBDEV_DEVCTL, &xrt_devctl_driver, xrt_devctl_endpoints);
+ else
+ xleaf_unregister_driver(XRT_SUBDEV_DEVCTL);
+}

Next message: Florent Revest: "Re: [PATCH bpf-next 1/5] bpf: Add a ARG_PTR_TO_CONST_STR argument type"
Previous message: Guenter Roeck: "Re: [PATCH v4 RESEND net-next] net: socket: use BIT() for MSG_*"
Next in thread: Tom Rix: "Re: [PATCH V3 XRT Alveo 13/18] fpga: xrt: devctl platform driver"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]