Re: [PATCH v5 net-next 7/8] sfc: add support for devlink port_function_hw_addr_get in ef100

On 2/7/23 15:01, Lucero Palau, Alejandro wrote:

On 2/2/23 12:09, Jiri Pirko wrote:

quoted

Thu, Feb 02, 2023 at 12:14:22PM CET, alejandro.lucero-palau@amd.com wrote:

quoted

From: Alejandro Lucero <redacted>

Using the builtin client handle id infrastructure, this patch adds

Don't talk about "this patch". Just tell the codebase what to do.

OK

quoted

quoted

support for obtaining the mac address linked to mports in ef100. This
implies to execute an MCDI command for getting the data from the
firmware for each devlink port.

Signed-off-by: Alejandro Lucero <redacted>
---
drivers/net/ethernet/sfc/ef100_nic.c   | 27 +++++++++++++
drivers/net/ethernet/sfc/ef100_nic.h   |  1 +
drivers/net/ethernet/sfc/ef100_rep.c   |  8 ++++
drivers/net/ethernet/sfc/ef100_rep.h   |  1 +
drivers/net/ethernet/sfc/efx_devlink.c | 53 ++++++++++++++++++++++++++
5 files changed, 90 insertions(+)

diff --git a/drivers/net/ethernet/sfc/ef100_nic.c b/drivers/net/ethernet/sfc/ef100_nic.c
index aa48c79a2149..becd21c2325d 100644
--- a/drivers/net/ethernet/sfc/ef100_nic.c
+++ b/drivers/net/ethernet/sfc/ef100_nic.c

@@ -1122,6 +1122,33 @@ static int ef100_probe_main(struct efx_nic *efx)

	return rc;
}

+/* MCDI commands are related to the same device issuing them. This function
+ * allows to do an MCDI command on behalf of another device, mainly PFs setting
+ * things for VFs.
+ */
+int efx_ef100_lookup_client_id(struct efx_nic *efx, efx_qword_t pciefn, u32 *id)
+{
+	MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_CLIENT_HANDLE_OUT_LEN);
+	MCDI_DECLARE_BUF(inbuf, MC_CMD_GET_CLIENT_HANDLE_IN_LEN);
+	u64 pciefn_flat = le64_to_cpu(pciefn.u64[0]);
+	size_t outlen;
+	int rc;
+
+	MCDI_SET_DWORD(inbuf, GET_CLIENT_HANDLE_IN_TYPE,
+		       MC_CMD_GET_CLIENT_HANDLE_IN_TYPE_FUNC);
+	MCDI_SET_QWORD(inbuf, GET_CLIENT_HANDLE_IN_FUNC,
+		       pciefn_flat);
+
+	rc = efx_mcdi_rpc(efx, MC_CMD_GET_CLIENT_HANDLE, inbuf, sizeof(inbuf),
+			  outbuf, sizeof(outbuf), &outlen);
+	if (rc)
+		return rc;
+	if (outlen < sizeof(outbuf))
+		return -EIO;
+	*id = MCDI_DWORD(outbuf, GET_CLIENT_HANDLE_OUT_HANDLE);
+	return 0;
+}
+
int ef100_probe_netdev_pf(struct efx_nic *efx)
{
	struct ef100_nic_data *nic_data = efx->nic_data;

diff --git a/drivers/net/ethernet/sfc/ef100_nic.h b/drivers/net/ethernet/sfc/ef100_nic.h
index e59044072333..f1ed481c1260 100644
--- a/drivers/net/ethernet/sfc/ef100_nic.h
+++ b/drivers/net/ethernet/sfc/ef100_nic.h

@@ -94,4 +94,5 @@ int ef100_filter_table_probe(struct efx_nic *efx);

int ef100_get_mac_address(struct efx_nic *efx, u8 *mac_address,
			  int client_handle, bool empty_ok);
+int efx_ef100_lookup_client_id(struct efx_nic *efx, efx_qword_t pciefn, u32 *id);
#endif	/* EFX_EF100_NIC_H */

diff --git a/drivers/net/ethernet/sfc/ef100_rep.c b/drivers/net/ethernet/sfc/ef100_rep.c
index 6b5bc5d6955d..0b3083ef0ead 100644
--- a/drivers/net/ethernet/sfc/ef100_rep.c
+++ b/drivers/net/ethernet/sfc/ef100_rep.c

@@ -361,6 +361,14 @@ bool ef100_mport_on_local_intf(struct efx_nic *efx,

		     mport_desc->interface_idx == nic_data->local_mae_intf;
}

+bool ef100_mport_is_vf(struct mae_mport_desc *mport_desc)
+{
+	bool pcie_func;
+
+	pcie_func = ef100_mport_is_pcie_vnic(mport_desc);
+	return pcie_func && (mport_desc->vf_idx != MAE_MPORT_DESC_VF_IDX_NULL);
+}
+
void efx_ef100_init_reps(struct efx_nic *efx)
{
	struct ef100_nic_data *nic_data = efx->nic_data;

diff --git a/drivers/net/ethernet/sfc/ef100_rep.h b/drivers/net/ethernet/sfc/ef100_rep.h
index ae6add4b0855..a042525a2240 100644
--- a/drivers/net/ethernet/sfc/ef100_rep.h
+++ b/drivers/net/ethernet/sfc/ef100_rep.h

@@ -76,4 +76,5 @@ void efx_ef100_fini_reps(struct efx_nic *efx);

struct mae_mport_desc;
bool ef100_mport_on_local_intf(struct efx_nic *efx,
			       struct mae_mport_desc *mport_desc);
+bool ef100_mport_is_vf(struct mae_mport_desc *mport_desc);
#endif /* EF100_REP_H */

diff --git a/drivers/net/ethernet/sfc/efx_devlink.c b/drivers/net/ethernet/sfc/efx_devlink.c
index afdb19f0c774..c44547b9894e 100644
--- a/drivers/net/ethernet/sfc/efx_devlink.c
+++ b/drivers/net/ethernet/sfc/efx_devlink.c

@@ -60,6 +60,56 @@ static int efx_devlink_add_port(struct efx_nic *efx,

	return devl_port_register(efx->devlink, &mport->dl_port, mport->mport_id);
}
+
+static int efx_devlink_port_addr_get(struct devlink_port *port, u8 *hw_addr,
+				     int *hw_addr_len,
+				     struct netlink_ext_ack *extack)
+{
+	struct efx_devlink *devlink = devlink_priv(port->devlink);
+	struct mae_mport_desc *mport_desc;
+	efx_qword_t pciefn;
+	u32 client_id;
+	int rc = 0;

Pointless init.

Right

quoted

quoted

+
+	mport_desc = container_of(port, struct mae_mport_desc, dl_port);
+
+	if (!ef100_mport_on_local_intf(devlink->efx, mport_desc)) {
+		rc = -EINVAL;
+		NL_SET_ERR_MSG_FMT(extack,
+				   "Port not on local interface (mport: %u)",
+				   mport_desc->mport_id);
+		goto out;
+	}
+
+	if (ef100_mport_is_vf(mport_desc))
+		EFX_POPULATE_QWORD_3(pciefn,
+				     PCIE_FUNCTION_PF, PCIE_FUNCTION_PF_NULL,
+				     PCIE_FUNCTION_VF, mport_desc->vf_idx,
+				     PCIE_FUNCTION_INTF, PCIE_INTERFACE_CALLER);
+	else
+		EFX_POPULATE_QWORD_3(pciefn,
+				     PCIE_FUNCTION_PF, mport_desc->pf_idx,
+				     PCIE_FUNCTION_VF, PCIE_FUNCTION_VF_NULL,
+				     PCIE_FUNCTION_INTF, PCIE_INTERFACE_CALLER);
+
+	rc = efx_ef100_lookup_client_id(devlink->efx, pciefn, &client_id);
+	if (rc) {
+		NL_SET_ERR_MSG_FMT(extack,
+				   "No internal client_ID for port (mport: %u)",
+				   mport_desc->mport_id);
+		goto out;
+	}
+
+	rc = ef100_get_mac_address(devlink->efx, hw_addr, client_id, true);
+	if (rc != 0)

why "if (rc)" is not enough here?

Right

quoted

quoted

+		NL_SET_ERR_MSG_FMT(extack,
+				   "No available MAC for port (mport: %u)",
+				   mport_desc->mport_id);

It is redundant to print mport_id which is exposed as devlink port id.
Please remove from the all the extack messages. No need to mention
"port" at all, as the user knows on which object he is performing the
command.

I think the idea was to have such a report to user space as input from a
client to our support team which could help.

But I guess you are right, and that info would also be reported without
specifically adding it here.

quoted

Also, perhaps it would sound better to say "No MAC available"?

That subtle change could imply a total different meaning ... at least in
my native language. But I will not fight this one :-) and change it in the
next version.

quoted

quoted

+out:
+	*hw_addr_len = ETH_ALEN;

Odd. I think you should not touch hw_addr_len in case of error.

It does not harm. Does it?

Otherwise, I need another goto in previous error checking what seems odd
for a single line.

Silly me. I do not need the gotos at all!

quoted

quoted

+	return rc;
+}
+
#endif

static int efx_devlink_info_nvram_partition(struct efx_nic *efx,

@@ -522,6 +572,9 @@ static int efx_devlink_info_get(struct devlink *devlink,

static const struct devlink_ops sfc_devlink_ops = {
	.info_get			= efx_devlink_info_get,
+#ifdef CONFIG_SFC_SRIOV
+	.port_function_hw_addr_get	= efx_devlink_port_addr_get,
+#endif
};

#ifdef CONFIG_SFC_SRIOV
-- 
2.17.1