[PATCH 0/3] net: tsec: Fix issues of DM driver
From: Hou Zhiqiang Zhiqiang.Hou@nxp.com
This patch set is to fix some issues of eTSEC DM driver.
Hou Zhiqiang (3): doc: dt-bindings: tsec: Correct the Ethernet port compatible string net: tsec: Access eTSEC registers using virtual address net: tsec: Access TBI PHY through the corresponding MII
doc/device-tree-bindings/net/fsl-tsec-phy.txt | 4 ++-- drivers/net/tsec.c | 22 +++++++++++++++---- 2 files changed, 20 insertions(+), 6 deletions(-)
From: Hou Zhiqiang Zhiqiang.Hou@nxp.com
Change the compatible string to "fsl,etsec2" for the Ethernet ports, which is used in the current driver's match table.
Fixes: 69a00875e3db ("doc: dt-bindings: Describe Freescale TSEC ethernet controller") Signed-off-by: Hou Zhiqiang Zhiqiang.Hou@nxp.com --- doc/device-tree-bindings/net/fsl-tsec-phy.txt | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/doc/device-tree-bindings/net/fsl-tsec-phy.txt b/doc/device-tree-bindings/net/fsl-tsec-phy.txt index 59989e3b09..8e8574bc97 100644 --- a/doc/device-tree-bindings/net/fsl-tsec-phy.txt +++ b/doc/device-tree-bindings/net/fsl-tsec-phy.txt @@ -2,7 +2,7 @@
Properties:
- - compatible : Should be "fsl,tsec" + - compatible : Should be "fsl,etsec2" - reg : Offset and length of the register set for the device - phy-handle : See ethernet.txt file in the same directory. - phy-connection-type : See ethernet.txt file in the same directory. This @@ -12,7 +12,7 @@ Properties:
Example: ethernet@24000 { - compatible = "fsl,tsec"; + compatible = "fsl,etsec2"; reg = <0x24000 0x1000>; phy-handle = <&phy0>; phy-connection-type = "sgmii";
On Sun, 3 May 2020 at 09:28, Zhiqiang Hou Zhiqiang.Hou@nxp.com wrote:
From: Hou Zhiqiang Zhiqiang.Hou@nxp.com
Change the compatible string to "fsl,etsec2" for the Ethernet ports, which is used in the current driver's match table.
Fixes: 69a00875e3db ("doc: dt-bindings: Describe Freescale TSEC ethernet controller") Signed-off-by: Hou Zhiqiang Zhiqiang.Hou@nxp.com
Acked-by: Vladimir Oltean vladimir.oltean@nxp.com
Although I am unclear how the G&D MPC8308 boards (arch/powerpc/dts/gdsys/mpc8308.dtsi) have ever probed their TSEC interfaces. Probably not based on DM, since the fsl,tsec compatible string has never been supported by the mainline tsec driver.
doc/device-tree-bindings/net/fsl-tsec-phy.txt | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/doc/device-tree-bindings/net/fsl-tsec-phy.txt b/doc/device-tree-bindings/net/fsl-tsec-phy.txt index 59989e3b09..8e8574bc97 100644 --- a/doc/device-tree-bindings/net/fsl-tsec-phy.txt +++ b/doc/device-tree-bindings/net/fsl-tsec-phy.txt @@ -2,7 +2,7 @@
Properties:
- compatible : Should be "fsl,tsec"
- compatible : Should be "fsl,etsec2"
- reg : Offset and length of the register set for the device
- phy-handle : See ethernet.txt file in the same directory.
- phy-connection-type : See ethernet.txt file in the same directory. This
@@ -12,7 +12,7 @@ Properties:
Example: ethernet@24000 {
compatible = "fsl,tsec";
compatible = "fsl,etsec2"; reg = <0x24000 0x1000>; phy-handle = <&phy0>; phy-connection-type = "sgmii";-- 2.17.1
Hi Vladimir,
Thanks a lot for your ack!
-----Original Message----- From: Vladimir Oltean olteanv@gmail.com Sent: 2020年5月3日 19:13 To: Z.q. Hou zhiqiang.hou@nxp.com Cc: u-boot u-boot@lists.denx.de; Joe Hershberger joe.hershberger@ni.com; Bin Meng bmeng.cn@gmail.com; Priyanka Jain priyanka.jain@nxp.com; mario.six@gdsys.cc Subject: Re: [PATCH 1/3] doc: dt-bindings: tsec: Correct the Ethernet port compatible string
On Sun, 3 May 2020 at 09:28, Zhiqiang Hou Zhiqiang.Hou@nxp.com wrote:
From: Hou Zhiqiang Zhiqiang.Hou@nxp.com
Change the compatible string to "fsl,etsec2" for the Ethernet ports, which is used in the current driver's match table.
Fixes: 69a00875e3db ("doc: dt-bindings: Describe Freescale TSEC ethernet controller") Signed-off-by: Hou Zhiqiang Zhiqiang.Hou@nxp.com
Acked-by: Vladimir Oltean vladimir.oltean@nxp.com
Although I am unclear how the G&D MPC8308 boards (arch/powerpc/dts/gdsys/mpc8308.dtsi) have ever probed their TSEC interfaces. Probably not based on DM, since the fsl,tsec compatible string has never been supported by the mainline tsec driver.
I agree with you, they seems using legacy driver. I'm working on the support of compatible "gianfar", the MPC8308 boards can shift to the DM driver then.
Thanks, Zhiqiang
doc/device-tree-bindings/net/fsl-tsec-phy.txt | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/doc/device-tree-bindings/net/fsl-tsec-phy.txt b/doc/device-tree-bindings/net/fsl-tsec-phy.txt index 59989e3b09..8e8574bc97 100644 --- a/doc/device-tree-bindings/net/fsl-tsec-phy.txt +++ b/doc/device-tree-bindings/net/fsl-tsec-phy.txt @@ -2,7 +2,7 @@
Properties:
- compatible : Should be "fsl,tsec"
- compatible : Should be "fsl,etsec2"
- reg : Offset and length of the register set for the device
- phy-handle : See ethernet.txt file in the same directory.
- phy-connection-type : See ethernet.txt file in the same
directory. This @@ -12,7 +12,7 @@ Properties:
Example: ethernet@24000 {
compatible = "fsl,tsec";
compatible = "fsl,etsec2"; reg = <0x24000 0x1000>; phy-handle = <&phy0>; phy-connection-type = "sgmii";-- 2.17.1
From: Hou Zhiqiang Zhiqiang.Hou@nxp.com
The current code accesses eTSEC registers using physical address directly, it's not correct, though no problem on current platforms. It won't work on platforms, which does not support 1:1 virtual-physical address map.
Signed-off-by: Hou Zhiqiang Zhiqiang.Hou@nxp.com --- drivers/net/tsec.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-)
diff --git a/drivers/net/tsec.c b/drivers/net/tsec.c index f85cdcb97e..ce41aec5cf 100644 --- a/drivers/net/tsec.c +++ b/drivers/net/tsec.c @@ -798,7 +798,7 @@ int tsec_probe(struct udevice *dev) int ret;
pdata->iobase = (phys_addr_t)dev_read_addr(dev); - priv->regs = (struct tsec *)pdata->iobase; + priv->regs = dev_remap_addr(dev);
if (dev_read_phandle_with_args(dev, "phy-handle", NULL, 0, 0, &phandle_args)) { @@ -817,8 +817,11 @@ int tsec_probe(struct udevice *dev) }
reg = ofnode_get_addr_index(parent, 0); - priv->phyregs_sgmii = (struct tsec_mii_mng *) - (reg + TSEC_MDIO_REGS_OFFSET); + if (reg == FDT_ADDR_T_NONE) + return -ENOENT; + + priv->phyregs_sgmii = map_physmem(reg + TSEC_MDIO_REGS_OFFSET, 0, + MAP_NOCACHE);
ret = dev_read_phandle_with_args(dev, "tbi-handle", NULL, 0, 0, &phandle_args);
Hi Zhiqiang,
On Sun, 3 May 2020 at 09:28, Zhiqiang Hou Zhiqiang.Hou@nxp.com wrote:
From: Hou Zhiqiang Zhiqiang.Hou@nxp.com
The current code accesses eTSEC registers using physical address directly, it's not correct, though no problem on current platforms. It won't work on platforms, which does not support 1:1 virtual-physical address map.
Signed-off-by: Hou Zhiqiang Zhiqiang.Hou@nxp.com
Reviewed-by: Vladimir Oltean vladimir.oltean@nxp.com Tested-by: Vladimir Oltean vladimir.oltean@nxp.com
(on LS1021A)
drivers/net/tsec.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-)
diff --git a/drivers/net/tsec.c b/drivers/net/tsec.c index f85cdcb97e..ce41aec5cf 100644 --- a/drivers/net/tsec.c +++ b/drivers/net/tsec.c @@ -798,7 +798,7 @@ int tsec_probe(struct udevice *dev) int ret;
pdata->iobase = (phys_addr_t)dev_read_addr(dev);
priv->regs = (struct tsec *)pdata->iobase;
priv->regs = dev_remap_addr(dev); if (dev_read_phandle_with_args(dev, "phy-handle", NULL, 0, 0, &phandle_args)) {@@ -817,8 +817,11 @@ int tsec_probe(struct udevice *dev) }
reg = ofnode_get_addr_index(parent, 0);
priv->phyregs_sgmii = (struct tsec_mii_mng *)(reg + TSEC_MDIO_REGS_OFFSET);
if (reg == FDT_ADDR_T_NONE)return -ENOENT;priv->phyregs_sgmii = map_physmem(reg + TSEC_MDIO_REGS_OFFSET, 0,MAP_NOCACHE); ret = dev_read_phandle_with_args(dev, "tbi-handle", NULL, 0, 0, &phandle_args);-- 2.17.1
Hi Vladimir,
Thanks a lot for your review and test!
Thanks, Zhiqiang
-----Original Message----- From: Vladimir Oltean olteanv@gmail.com Sent: 2020年5月3日 19:36 To: Z.q. Hou zhiqiang.hou@nxp.com Cc: u-boot u-boot@lists.denx.de; Joe Hershberger joe.hershberger@ni.com; Bin Meng bmeng.cn@gmail.com; Priyanka Jain priyanka.jain@nxp.com Subject: Re: [PATCH 2/3] net: tsec: Access eTSEC registers using virtual address
Hi Zhiqiang,
On Sun, 3 May 2020 at 09:28, Zhiqiang Hou Zhiqiang.Hou@nxp.com wrote:
From: Hou Zhiqiang Zhiqiang.Hou@nxp.com
The current code accesses eTSEC registers using physical address directly, it's not correct, though no problem on current platforms. It won't work on platforms, which does not support 1:1 virtual-physical address map.
Signed-off-by: Hou Zhiqiang Zhiqiang.Hou@nxp.com
Reviewed-by: Vladimir Oltean vladimir.oltean@nxp.com Tested-by: Vladimir Oltean vladimir.oltean@nxp.com
(on LS1021A)
drivers/net/tsec.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-)
diff --git a/drivers/net/tsec.c b/drivers/net/tsec.c index f85cdcb97e..ce41aec5cf 100644 --- a/drivers/net/tsec.c +++ b/drivers/net/tsec.c @@ -798,7 +798,7 @@ int tsec_probe(struct udevice *dev) int ret;
pdata->iobase = (phys_addr_t)dev_read_addr(dev);
priv->regs = (struct tsec *)pdata->iobase;
priv->regs = dev_remap_addr(dev); if (dev_read_phandle_with_args(dev, "phy-handle", NULL, 0, 0, &phandle_args)) { @@-817,8
+817,11 @@ int tsec_probe(struct udevice *dev) }
reg = ofnode_get_addr_index(parent, 0);
priv->phyregs_sgmii = (struct tsec_mii_mng *)(reg + TSEC_MDIO_REGS_OFFSET);
if (reg == FDT_ADDR_T_NONE)return -ENOENT;priv->phyregs_sgmii = map_physmem(reg +TSEC_MDIO_REGS_OFFSET, 0,
MAP_NOCACHE); ret = dev_read_phandle_with_args(dev, "tbi-handle", NULL, 0,0,
&phandle_args);-- 2.17.1
From: Hou Zhiqiang Zhiqiang.Hou@nxp.com
When an eTSEC is configured to use TBI, configuration of the TBI is done through the MIIM registers for that eTSEC. For example, if a TBI interface is required on eTSEC2, then the MIIM registers starting at offset 0x2_5520 are used to configure it.
Fixes: 9a1d6af55ecd ("net: tsec: Add driver model ethernet support") Signed-off-by: Hou Zhiqiang Zhiqiang.Hou@nxp.com --- drivers/net/tsec.c | 17 ++++++++++++++--- 1 file changed, 14 insertions(+), 3 deletions(-)
diff --git a/drivers/net/tsec.c b/drivers/net/tsec.c index ce41aec5cf..31056f3117 100644 --- a/drivers/net/tsec.c +++ b/drivers/net/tsec.c @@ -789,6 +789,7 @@ int tsec_probe(struct udevice *dev) { struct eth_pdata *pdata = dev_get_platdata(dev); struct tsec_private *priv = dev_get_priv(dev); + struct tsec_mii_mng __iomem *ext_phyregs_mii; struct ofnode_phandle_args phandle_args; u32 tbiaddr = CONFIG_SYS_TBIPA_VALUE; struct fsl_pq_mdio_info mdio_info; @@ -820,8 +821,8 @@ int tsec_probe(struct udevice *dev) if (reg == FDT_ADDR_T_NONE) return -ENOENT;
- priv->phyregs_sgmii = map_physmem(reg + TSEC_MDIO_REGS_OFFSET, 0, - MAP_NOCACHE); + ext_phyregs_mii = map_physmem(reg + TSEC_MDIO_REGS_OFFSET, 0, + MAP_NOCACHE);
ret = dev_read_phandle_with_args(dev, "tbi-handle", NULL, 0, 0, &phandle_args); @@ -830,6 +831,16 @@ int tsec_probe(struct udevice *dev)
priv->tbiaddr = tbiaddr;
+ parent = ofnode_get_parent(phandle_args.node); + if (!ofnode_valid(parent)) { + printf("No parent node for TBI PHY?\n"); + return -ENOENT; + } + + reg = ofnode_get_addr_index(parent, 0); + priv->phyregs_sgmii = map_physmem(reg + TSEC_MDIO_REGS_OFFSET, 0, + MAP_NOCACHE); + phy_mode = dev_read_prop(dev, "phy-connection-type", NULL); if (phy_mode) pdata->phy_interface = phy_get_interface_by_name(phy_mode); @@ -844,7 +855,7 @@ int tsec_probe(struct udevice *dev) if (priv->interface == PHY_INTERFACE_MODE_SGMII) priv->flags |= TSEC_SGMII;
- mdio_info.regs = priv->phyregs_sgmii; + mdio_info.regs = ext_phyregs_mii; mdio_info.name = (char *)dev->name; ret = fsl_pq_mdio_init(NULL, &mdio_info); if (ret)
On Sun, 3 May 2020 at 09:28, Zhiqiang Hou Zhiqiang.Hou@nxp.com wrote:
From: Hou Zhiqiang Zhiqiang.Hou@nxp.com
When an eTSEC is configured to use TBI, configuration of the TBI is done through the MIIM registers for that eTSEC. For example, if a TBI interface is required on eTSEC2, then the MIIM registers starting at offset 0x2_5520 are used to configure it.
Fixes: 9a1d6af55ecd ("net: tsec: Add driver model ethernet support") Signed-off-by: Hou Zhiqiang Zhiqiang.Hou@nxp.com
drivers/net/tsec.c | 17 ++++++++++++++--- 1 file changed, 14 insertions(+), 3 deletions(-)
diff --git a/drivers/net/tsec.c b/drivers/net/tsec.c index ce41aec5cf..31056f3117 100644 --- a/drivers/net/tsec.c +++ b/drivers/net/tsec.c @@ -789,6 +789,7 @@ int tsec_probe(struct udevice *dev) { struct eth_pdata *pdata = dev_get_platdata(dev); struct tsec_private *priv = dev_get_priv(dev);
struct tsec_mii_mng __iomem *ext_phyregs_mii; struct ofnode_phandle_args phandle_args; u32 tbiaddr = CONFIG_SYS_TBIPA_VALUE; struct fsl_pq_mdio_info mdio_info;@@ -820,8 +821,8 @@ int tsec_probe(struct udevice *dev) if (reg == FDT_ADDR_T_NONE) return -ENOENT;
priv->phyregs_sgmii = map_physmem(reg + TSEC_MDIO_REGS_OFFSET, 0,MAP_NOCACHE);
ext_phyregs_mii = map_physmem(reg + TSEC_MDIO_REGS_OFFSET, 0,MAP_NOCACHE); ret = dev_read_phandle_with_args(dev, "tbi-handle", NULL, 0, 0, &phandle_args);@@ -830,6 +831,16 @@ int tsec_probe(struct udevice *dev)
priv->tbiaddr = tbiaddr;
parent = ofnode_get_parent(phandle_args.node);if (!ofnode_valid(parent)) {printf("No parent node for TBI PHY?\n");return -ENOENT;}reg = ofnode_get_addr_index(parent, 0);
Missing a check
if (reg == FDT_ADDR_T_NONE) return -ENOENT;
here?
priv->phyregs_sgmii = map_physmem(reg + TSEC_MDIO_REGS_OFFSET, 0,MAP_NOCACHE);phy_mode = dev_read_prop(dev, "phy-connection-type", NULL); if (phy_mode) pdata->phy_interface = phy_get_interface_by_name(phy_mode);@@ -844,7 +855,7 @@ int tsec_probe(struct udevice *dev) if (priv->interface == PHY_INTERFACE_MODE_SGMII) priv->flags |= TSEC_SGMII;
mdio_info.regs = priv->phyregs_sgmii;
mdio_info.regs = ext_phyregs_mii; mdio_info.name = (char *)dev->name; ret = fsl_pq_mdio_init(NULL, &mdio_info); if (ret)-- 2.17.1
With that:
Reviewed-by: Vladimir Oltean vladimir.oltean@nxp.com Tested-by: Vladimir Oltean vladimir.oltean@nxp.com
(on LS1021A)
Thanks! -Vladimir
Hi Vladimir,
Thanks a lot for your review and test!
-----Original Message----- From: Vladimir Oltean olteanv@gmail.com Sent: 2020年5月3日 19:35 To: Z.q. Hou zhiqiang.hou@nxp.com Cc: u-boot u-boot@lists.denx.de; Joe Hershberger joe.hershberger@ni.com; Bin Meng bmeng.cn@gmail.com; Priyanka Jain priyanka.jain@nxp.com Subject: Re: [PATCH 3/3] net: tsec: Access TBI PHY through the corresponding MII
On Sun, 3 May 2020 at 09:28, Zhiqiang Hou Zhiqiang.Hou@nxp.com wrote:
From: Hou Zhiqiang Zhiqiang.Hou@nxp.com
When an eTSEC is configured to use TBI, configuration of the TBI is done through the MIIM registers for that eTSEC. For example, if a TBI interface is required on eTSEC2, then the MIIM registers starting at offset 0x2_5520 are used to configure it.
Fixes: 9a1d6af55ecd ("net: tsec: Add driver model ethernet support") Signed-off-by: Hou Zhiqiang Zhiqiang.Hou@nxp.com
drivers/net/tsec.c | 17 ++++++++++++++--- 1 file changed, 14 insertions(+), 3 deletions(-)
diff --git a/drivers/net/tsec.c b/drivers/net/tsec.c index ce41aec5cf..31056f3117 100644 --- a/drivers/net/tsec.c +++ b/drivers/net/tsec.c @@ -789,6 +789,7 @@ int tsec_probe(struct udevice *dev) { struct eth_pdata *pdata = dev_get_platdata(dev); struct tsec_private *priv = dev_get_priv(dev);
struct tsec_mii_mng __iomem *ext_phyregs_mii; struct ofnode_phandle_args phandle_args; u32 tbiaddr = CONFIG_SYS_TBIPA_VALUE; struct fsl_pq_mdio_info mdio_info; @@ -820,8 +821,8 @@ inttsec_probe(struct udevice *dev) if (reg == FDT_ADDR_T_NONE) return -ENOENT;
priv->phyregs_sgmii = map_physmem(reg +TSEC_MDIO_REGS_OFFSET, 0,
MAP_NOCACHE);
ext_phyregs_mii = map_physmem(reg +TSEC_MDIO_REGS_OFFSET, 0,
MAP_NOCACHE); ret = dev_read_phandle_with_args(dev, "tbi-handle", NULL, 0,0,
&phandle_args); @@-830,6
+831,16 @@ int tsec_probe(struct udevice *dev)
priv->tbiaddr = tbiaddr;
parent = ofnode_get_parent(phandle_args.node);if (!ofnode_valid(parent)) {printf("No parent node for TBI PHY?\n");return -ENOENT;}reg = ofnode_get_addr_index(parent, 0);Missing a check
if (reg == FDT_ADDR_T_NONE) return -ENOENT;here?
Yes, will fix in v2.
Thanks, Zhiqiang
priv->phyregs_sgmii = map_physmem(reg +TSEC_MDIO_REGS_OFFSET, 0,
MAP_NOCACHE);phy_mode = dev_read_prop(dev, "phy-connection-type", NULL); if (phy_mode) pdata->phy_interface =phy_get_interface_by_name(phy_mode); @@ -844,7 +855,7 @@ int tsec_probe(struct udevice *dev) if (priv->interface == PHY_INTERFACE_MODE_SGMII) priv->flags |= TSEC_SGMII;
mdio_info.regs = priv->phyregs_sgmii;
mdio_info.regs = ext_phyregs_mii; mdio_info.name = (char *)dev->name; ret = fsl_pq_mdio_init(NULL, &mdio_info); if (ret)-- 2.17.1
With that:
Reviewed-by: Vladimir Oltean vladimir.oltean@nxp.com Tested-by: Vladimir Oltean vladimir.oltean@nxp.com
(on LS1021A)
Thanks! -Vladimir
Hi Zhiqiang,
On Sun, 3 May 2020 at 16:49, Z.q. Hou zhiqiang.hou@nxp.com wrote:
Hi Vladimir,
Thanks a lot for your review and test!
-----Original Message----- From: Vladimir Oltean olteanv@gmail.com Sent: 2020年5月3日 19:35 To: Z.q. Hou zhiqiang.hou@nxp.com Cc: u-boot u-boot@lists.denx.de; Joe Hershberger joe.hershberger@ni.com; Bin Meng bmeng.cn@gmail.com; Priyanka Jain priyanka.jain@nxp.com Subject: Re: [PATCH 3/3] net: tsec: Access TBI PHY through the corresponding MII
On Sun, 3 May 2020 at 09:28, Zhiqiang Hou Zhiqiang.Hou@nxp.com wrote:
From: Hou Zhiqiang Zhiqiang.Hou@nxp.com
When an eTSEC is configured to use TBI, configuration of the TBI is done through the MIIM registers for that eTSEC. For example, if a TBI interface is required on eTSEC2, then the MIIM registers starting at offset 0x2_5520 are used to configure it.
Fixes: 9a1d6af55ecd ("net: tsec: Add driver model ethernet support") Signed-off-by: Hou Zhiqiang Zhiqiang.Hou@nxp.com
drivers/net/tsec.c | 17 ++++++++++++++--- 1 file changed, 14 insertions(+), 3 deletions(-)
diff --git a/drivers/net/tsec.c b/drivers/net/tsec.c index ce41aec5cf..31056f3117 100644 --- a/drivers/net/tsec.c +++ b/drivers/net/tsec.c @@ -789,6 +789,7 @@ int tsec_probe(struct udevice *dev) { struct eth_pdata *pdata = dev_get_platdata(dev); struct tsec_private *priv = dev_get_priv(dev);
struct tsec_mii_mng __iomem *ext_phyregs_mii; struct ofnode_phandle_args phandle_args; u32 tbiaddr = CONFIG_SYS_TBIPA_VALUE; struct fsl_pq_mdio_info mdio_info; @@ -820,8 +821,8 @@ inttsec_probe(struct udevice *dev) if (reg == FDT_ADDR_T_NONE) return -ENOENT;
priv->phyregs_sgmii = map_physmem(reg +TSEC_MDIO_REGS_OFFSET, 0,
MAP_NOCACHE);
ext_phyregs_mii = map_physmem(reg +TSEC_MDIO_REGS_OFFSET, 0,
MAP_NOCACHE); ret = dev_read_phandle_with_args(dev, "tbi-handle", NULL, 0,0,
&phandle_args); @@-830,6
+831,16 @@ int tsec_probe(struct udevice *dev)
priv->tbiaddr = tbiaddr;
parent = ofnode_get_parent(phandle_args.node);if (!ofnode_valid(parent)) {printf("No parent node for TBI PHY?\n");return -ENOENT;}reg = ofnode_get_addr_index(parent, 0);Missing a check
if (reg == FDT_ADDR_T_NONE) return -ENOENT;here?
Yes, will fix in v2.
Thanks, Zhiqiang
Actually since the TBI PHY is optional (not present on RGMII interfaces and such), I don't believe you should return an error, but just skip the phyregs_sgmii initialization.
priv->phyregs_sgmii = map_physmem(reg +TSEC_MDIO_REGS_OFFSET, 0,
MAP_NOCACHE);phy_mode = dev_read_prop(dev, "phy-connection-type", NULL); if (phy_mode) pdata->phy_interface =phy_get_interface_by_name(phy_mode); @@ -844,7 +855,7 @@ int tsec_probe(struct udevice *dev) if (priv->interface == PHY_INTERFACE_MODE_SGMII) priv->flags |= TSEC_SGMII;
mdio_info.regs = priv->phyregs_sgmii;
mdio_info.regs = ext_phyregs_mii; mdio_info.name = (char *)dev->name; ret = fsl_pq_mdio_init(NULL, &mdio_info); if (ret)-- 2.17.1
With that:
Reviewed-by: Vladimir Oltean vladimir.oltean@nxp.com Tested-by: Vladimir Oltean vladimir.oltean@nxp.com
(on LS1021A)
Thanks! -Vladimir
Thanks, -Vladimir
Hi Vladimir,
Thanks a lot for your comments!
-----Original Message----- From: Vladimir Oltean olteanv@gmail.com Sent: 2020年5月3日 22:07 To: Z.q. Hou zhiqiang.hou@nxp.com Cc: u-boot u-boot@lists.denx.de; Joe Hershberger joe.hershberger@ni.com; Bin Meng bmeng.cn@gmail.com; Priyanka Jain priyanka.jain@nxp.com Subject: Re: [PATCH 3/3] net: tsec: Access TBI PHY through the corresponding MII
Hi Zhiqiang,
On Sun, 3 May 2020 at 16:49, Z.q. Hou zhiqiang.hou@nxp.com wrote:
Hi Vladimir,
Thanks a lot for your review and test!
-----Original Message----- From: Vladimir Oltean olteanv@gmail.com Sent: 2020年5月3日 19:35 To: Z.q. Hou zhiqiang.hou@nxp.com Cc: u-boot u-boot@lists.denx.de; Joe Hershberger joe.hershberger@ni.com; Bin Meng bmeng.cn@gmail.com;
Priyanka
Jain priyanka.jain@nxp.com Subject: Re: [PATCH 3/3] net: tsec: Access TBI PHY through the corresponding MII
On Sun, 3 May 2020 at 09:28, Zhiqiang Hou Zhiqiang.Hou@nxp.com wrote:
From: Hou Zhiqiang Zhiqiang.Hou@nxp.com
When an eTSEC is configured to use TBI, configuration of the TBI is done through the MIIM registers for that eTSEC. For example, if a TBI interface is required on eTSEC2, then the MIIM registers starting at offset 0x2_5520 are used to configure it.
Fixes: 9a1d6af55ecd ("net: tsec: Add driver model ethernet support") Signed-off-by: Hou Zhiqiang Zhiqiang.Hou@nxp.com
drivers/net/tsec.c | 17 ++++++++++++++--- 1 file changed, 14 insertions(+), 3 deletions(-)
diff --git a/drivers/net/tsec.c b/drivers/net/tsec.c index ce41aec5cf..31056f3117 100644 --- a/drivers/net/tsec.c +++ b/drivers/net/tsec.c @@ -789,6 +789,7 @@ int tsec_probe(struct udevice *dev) { struct eth_pdata *pdata = dev_get_platdata(dev); struct tsec_private *priv = dev_get_priv(dev);
struct tsec_mii_mng __iomem *ext_phyregs_mii; struct ofnode_phandle_args phandle_args; u32 tbiaddr = CONFIG_SYS_TBIPA_VALUE; struct fsl_pq_mdio_info mdio_info; @@ -820,8 +821,8 @@int
tsec_probe(struct udevice *dev) if (reg == FDT_ADDR_T_NONE) return -ENOENT;
priv->phyregs_sgmii = map_physmem(reg +TSEC_MDIO_REGS_OFFSET, 0,
MAP_NOCACHE);
ext_phyregs_mii = map_physmem(reg +TSEC_MDIO_REGS_OFFSET, 0,
MAP_NOCACHE); ret = dev_read_phandle_with_args(dev, "tbi-handle", NULL,0,
0,
&phandle_args);@@
-830,6
+831,16 @@ int tsec_probe(struct udevice *dev)
priv->tbiaddr = tbiaddr;
parent = ofnode_get_parent(phandle_args.node);if (!ofnode_valid(parent)) {printf("No parent node for TBI PHY?\n");return -ENOENT;}reg = ofnode_get_addr_index(parent, 0);Missing a check
if (reg == FDT_ADDR_T_NONE) return -ENOENT;here?
Yes, will fix in v2.
Thanks, Zhiqiang
Actually since the TBI PHY is optional (not present on RGMII interfaces and such), I don't believe you should return an error, but just skip the phyregs_sgmii initialization.
Correct, I just realized it.
Thanks, Zhiqiang
priv->phyregs_sgmii = map_physmem(reg +TSEC_MDIO_REGS_OFFSET, 0,
MAP_NOCACHE);phy_mode = dev_read_prop(dev, "phy-connection-type",NULL);
if (phy_mode) pdata->phy_interface =phy_get_interface_by_name(phy_mode); @@ -844,7 +855,7 @@ int tsec_probe(struct udevice *dev) if (priv->interface == PHY_INTERFACE_MODE_SGMII) priv->flags |= TSEC_SGMII;
mdio_info.regs = priv->phyregs_sgmii;
mdio_info.regs = ext_phyregs_mii; mdio_info.name = (char *)dev->name; ret = fsl_pq_mdio_init(NULL, &mdio_info); if (ret)-- 2.17.1
With that:
Reviewed-by: Vladimir Oltean vladimir.oltean@nxp.com Tested-by: Vladimir Oltean vladimir.oltean@nxp.com
(on LS1021A)
Thanks! -Vladimir
Thanks, -Vladimir
participants (3)
-
Vladimir Oltean -
Z.q. Hou -
Zhiqiang Hou