ARM: mvebu: network interface naming

PATCH New Thomas Petazzoni 5 participants 18 messages Jan 27, 2016 – Feb 25, 2016 Device tree bindings

lore

Share & tools

lore

Thomas Petazzoni

author

2016-01-27 15:08 UTC #0 lore raw

cover ARM: mvebu: network interface naming

Dear mvebu maintainers,

As was discussed back in November last year, the naming of network interfaces on Armada 38x is very confusing. Gregory Clement suggested to exceptionally violate the rule of "DT nodes should be ordered by register address" in order to reduce the confusion.

The first patch implements this proposal.

The second patch is a new iteration of the patch that adds comments to the Armada 385 DB AP Device Tree file. The contents of the patch is the same (add indications of which physical connector corresponds each port), only the commit log is adjusted to take into account the fact that thanks to the first patch, the naming of the network interfaces visible from Linux is now a lot more sane.

Thanks!

Thomas

Thomas Petazzoni (2):
ARM: mvebu: change order of ethernet DT nodes on Armada 38x
ARM: mvebu: add reference to ETH connectors for A385-AP

 arch/arm/boot/dts/armada-385-db-ap.dts |  3 +++
 arch/arm/boot/dts/armada-38x.dtsi      | 30 +++++++++++++++++++++---------
 2 files changed, 24 insertions(+), 9 deletions(-)

Signature · 1 line

2.6.4

Thomas Petazzoni

author

2016-01-27 15:08 UTC #1 lore raw

PATCH 1/2 ARM: mvebu: change order of ethernet DT nodes on Armada 38x

On Armada 38x, the available network interfaces are:

- port 0, at 0x70000
- port 1, at 0x30000
- port 2, at 0x34000

Due to the rule saying that DT nodes should be ordered by register addresses, the network interfaces are probed in this order:

- port 1, at 0x30000, which gets named eth0
- port 2, at 0x34000, which gets named eth1
- port 0, at 0x70000, which gets named eth2

(if all three ports are enabled at the board level)

Unfortunately, the network subsystem doesn't provide any way to rename network interfaces from the kernel (it can only be done from userspace). So, the default naming of the network interfaces is very confusing as it doesn't match the datasheet, nor the naming of the interfaces in the bootloader, nor the naming of the interfaces on labels printed on the board.

For example, on the Armada 388 GP, the board has two ports, labelled GE0 and GE1. One has to know that GE0 is eth1 and GE1 is eth0, which isn't really obvious.

In order to solve this, this patch proposes to exceptionaly violate the rule of "order DT nodes by register address", and put the 0x70000 node before the 0x30000 node, so that network interfaces get named in a more natural way.

Signed-off-by Thomas Petazzoni <[email protected]>

---
 arch/arm/boot/dts/armada-38x.dtsi | 30 +++++++++++++++++++++---------
 1 file changed, 21 insertions(+), 9 deletions(-)

arch/arm/boot/dts/armada-38x.dtsi +21 -9

index e8b7f67..b50784d 100644
@@ -429,6 +429,27 @@
 				reg = <0x22000 0x1000>;
 			};
 
+			/*
+			 * As a special exception to the "order by
+			 * register address" rule, the eth0 node is
+			 * placed here to ensure that it gets
+			 * registered as the first interface, since
+			 * the network subsystem doesn't allow naming
+			 * interfaces using DT aliases. Without this,
+			 * the ordering of interfaces is different
+			 * from the one used in U-Boot and the
+			 * labeling of interfaces on the boards, which
+			 * is very confusing for users.
+			 */
+			eth0: ethernet at 70000 {
+				compatible = "marvell,armada-370-neta";
+				reg = <0x70000 0x4000>;
+				interrupts-extended = <&mpic 8>;
+				clocks = <&gateclk 4>;
+				tx-csum-limit = <9800>;
+				status = "disabled";
+			};
+
 			eth1: ethernet at 30000 {
 				compatible = "marvell,armada-370-neta";
 				reg = <0x30000 0x4000>;

@@ -493,15 +514,6 @@
 				};
 			};
 
-			eth0: ethernet at 70000 {
-				compatible = "marvell,armada-370-neta";
-				reg = <0x70000 0x4000>;
-				interrupts-extended = <&mpic 8>;
-				clocks = <&gateclk 4>;
-				tx-csum-limit = <9800>;
-				status = "disabled";
-			};
-
 			mdio: mdio at 72004 {
 				#address-cells = <1>;
 				#size-cells = <0>;

Signature · 1 line

2.6.4

Russell King - ARM Linux

2016-01-27 19:31 UTC #2 lore raw

PATCH 1/2 ARM: mvebu: change order of ethernet DT nodes on Armada 38x

Quoted 22 lines · Thomas Petazzoni

On Armada 38x, the available network interfaces are:
- port 0, at 0x70000
- port 1, at 0x30000
- port 2, at 0x34000
Due to the rule saying that DT nodes should be ordered by register addresses, the network interfaces are probed in this order:
- port 1, at 0x30000, which gets named eth0
- port 2, at 0x34000, which gets named eth1
- port 0, at 0x70000, which gets named eth2
(if all three ports are enabled at the board level)
Unfortunately, the network subsystem doesn't provide any way to rename network interfaces from the kernel (it can only be done from userspace). So, the default naming of the network interfaces is very confusing as it doesn't match the datasheet, nor the naming of the interfaces in the bootloader, nor the naming of the interfaces on labels printed on the board.

Quoted 8 lines

For example, on the Armada 388 GP, the board has two ports, labelled GE0 and GE1. One has to know that GE0 is eth1 and GE1 is eth0, which isn't really obvious.
In order to solve this, this patch proposes to exceptionaly violate the rule of "order DT nodes by register address", and put the 0x70000 node before the 0x30000 node, so that network interfaces get named in a more natural way.

The danger is that this will completely mess up people's existing scripts/distro configuration which are now used to the current labelling of the network ports.

Not every distro provides a sane way to deal with this...

Signature · 3 lines

RMK's Patch system: http://www.arm.linux.org.uk/developer/patches/
FTTC broadband for 0.8mile line: currently at 9.6Mbps down 400kbps up
according to speedtest.net.

Willy Tarreau

2016-01-27 19:45 UTC #3 lore raw

PATCH 1/2 ARM: mvebu: change order of ethernet DT nodes on Armada 38x

On Wed, Jan 27, 2016 at 07:31:44PM +0000, Russell King - ARM Linux wrote:
(...)

Quoted 9 lines

Quoted 8 lines
For example, on the Armada 388 GP, the board has two ports, labelled GE0 and GE1. One has to know that GE0 is eth1 and GE1 is eth0, which isn't really obvious.
In order to solve this, this patch proposes to exceptionaly violate the rule of "order DT nodes by register address", and put the 0x70000 node before the 0x30000 node, so that network interfaces get named in a more natural way.

Quoted 5 lines

The danger is that this will completely mess up people's existing scripts/distro configuration which are now used to the current labelling of the network ports.
Not every distro provides a sane way to deal with this...

Well, may be that's one more reason for fixing it before boards start to ship and run distro kernels. If some boards are already in the wild running on a BSP kernel, people will complain that mainline reorders their network ports and will stick to the BSP kernel instead.

Just my two cents,
Willy

Russell King - ARM Linux

2016-01-29 11:48 UTC #4 lore raw

PATCH 1/2 ARM: mvebu: change order of ethernet DT nodes on Armada 38x

Quoted 15 lines · Willy Tarreau

On Wed, Jan 27, 2016 at 07:31:44PM +0000, Russell King - ARM Linux wrote:
(...)
Quoted 13 lines
Quoted 8 lines
For example, on the Armada 388 GP, the board has two ports, labelled GE0 and GE1. One has to know that GE0 is eth1 and GE1 is eth0, which isn't really obvious.
In order to solve this, this patch proposes to exceptionaly violate the rule of "order DT nodes by register address", and put the 0x70000 node before the 0x30000 node, so that network interfaces get named in a more natural way.
The danger is that this will completely mess up people's existing scripts/distro configuration which are now used to the current labelling of the network ports.

Quoted 6 lines

Quoted 1 line
Not every distro provides a sane way to deal with this...
Well, may be that's one more reason for fixing it before boards start to ship and run distro kernels. If some boards are already in the wild running on a BSP kernel, people will complain that mainline reorders their network ports and will stick to the BSP kernel instead.

I've checked, and we believe not many people are using mainline kernels on clearfog. I'll have to adjust my debian jessie setup to fix the resulting carnage though. On the plus side, it means that mainline matches Marvell's kernel for interface naming.

However, I will point out that changing the DT order may fix it for now, but if there's changes to the kernel initialisation which results in a different probe order (eg, because of a change to the way probing works, which is something that's been discussed a few times) it's likely that the ethX names will change again...

In other words, what I'm saying is that there's no guarantee that changing the DT order will always result in the desired network device naming.

Signature · 3 lines

RMK's Patch system: http://www.arm.linux.org.uk/developer/patches/
FTTC broadband for 0.8mile line: currently at 9.6Mbps down 400kbps up
according to speedtest.net.

Thomas Petazzoni

author

2016-02-03 14:56 UTC #5 lore raw

PATCH 1/2 ARM: mvebu: change order of ethernet DT nodes on Armada 38x

Hello Russell,

Quoted 16 lines · Russell King - ARM Linux

Quoted 4 lines
Well, may be that's one more reason for fixing it before boards start to ship and run distro kernels. If some boards are already in the wild running on a BSP kernel, people will complain that mainline reorders their network ports and will stick to the BSP kernel instead.
I've checked, and we believe not many people are using mainline kernels on clearfog. I'll have to adjust my debian jessie setup to fix the resulting carnage though. On the plus side, it means that mainline matches Marvell's kernel for interface naming.
However, I will point out that changing the DT order may fix it for now, but if there's changes to the kernel initialisation which results in a different probe order (eg, because of a change to the way probing works, which is something that's been discussed a few times) it's likely that the ethX names will change again...

Quoted 3 lines

In other words, what I'm saying is that there's no guarantee that changing the DT order will always result in the desired network device naming.

Yes, fully agreed, it is not a complete guarantee. But in "most cases", it will still lead to something that is a bit more logical for normal users.

In its vendor BSP, Marvell has gone all the way to changing the link order of certain PCIe drivers, to make sure that the mvneta driver gets probed first. And indeed, without this hack, if you plug an e1000e PCIe NIC, it gets probed first and therefore assigned as eth0, while all the built-in network interfaces get shifted at eth1, eth2, etc.

Really sad that the network subsystem doesn't follow the DT aliases to name network interfaces, but it has been rejected multiple times by Dave Miller as far as I know. In the absence of this, re-ordering the DT nodes is a cheap hack that makes things look a bit more logical in many situations.

Best regards,

Thomas

Signature · 3 lines

Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

Willy Tarreau

2016-02-24 18:41 UTC #6 lore raw

PATCH 1/2 ARM: mvebu: change order of ethernet DT nodes on Armada 38x

In reply to Russell King - ARM Linux

Hi Russell,

Quoted 23 lines · Russell King - ARM Linux

Quoted 15 lines · Willy Tarreau
On Wed, Jan 27, 2016 at 07:31:44PM +0000, Russell King - ARM Linux wrote:
(...)
Quoted 13 lines
Quoted 8 lines
For example, on the Armada 388 GP, the board has two ports, labelled GE0 and GE1. One has to know that GE0 is eth1 and GE1 is eth0, which isn't really obvious.
In order to solve this, this patch proposes to exceptionaly violate the rule of "order DT nodes by register address", and put the 0x70000 node before the 0x30000 node, so that network interfaces get named in a more natural way.
The danger is that this will completely mess up people's existing scripts/distro configuration which are now used to the current labelling of the network ports.
Quoted 6 lines
Quoted 1 line
Not every distro provides a sane way to deal with this...
Well, may be that's one more reason for fixing it before boards start to ship and run distro kernels. If some boards are already in the wild running on a BSP kernel, people will complain that mainline reorders their network ports and will stick to the BSP kernel instead.

Quoted 4 lines

I've checked, and we believe not many people are using mainline kernels on clearfog. I'll have to adjust my debian jessie setup to fix the resulting carnage though. On the plus side, it means that mainline matches Marvell's kernel for interface naming.

Well, now I'm one of these and I confirm that it's really painful to have a different ordering between the DTB provided in mainline and the DTB provided with the board. The worst thing is that using mainline, eth0 corresponds to the switch and appears up, so you don't immediately realize that it's not where your cable is connected :-/

Thus could we please get Thomas' patch to ensure that the board boots with similar interface naming with both the original and mainline kernel ?

Thanks in advance,
Willy

Imre Kaloz

2016-02-24 19:02 UTC #7 lore raw

PATCH 1/2 ARM: mvebu: change order of ethernet DT nodes on Armada 38x

Quoted 46 lines · Willy Tarreau

Hi Russell,
Quoted 28 lines · Russell King - ARM Linux
Quoted 1 line · Willy Tarreau
On Wed, Jan 27, 2016 at 07:31:44PM +0000, Russell King - ARM Linux
Quoted 2 lines · wrote:
(...)
Quoted 1 line
Quoted 1 line
For example, on the Armada 388 GP, the board has two ports,
labelled
Quoted 1 line
Quoted 1 line
Quoted 1 line
GE0 and GE1. One has to know that GE0 is eth1 and GE1 is eth0,
which
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
isn't really obvious.
In order to solve this, this patch proposes to exceptionaly
violate
Quoted 1 line
Quoted 1 line
Quoted 1 line
the rule of "order DT nodes by register address", and put the
0x70000
Quoted 1 line
Quoted 1 line
Quoted 1 line
node before the 0x30000 node, so that network interfaces get
named in
Quoted 9 lines
Quoted 7 lines
Quoted 1 line
a more natural way.
The danger is that this will completely mess up people's existing scripts/distro configuration which are now used to the current labelling of the network ports.
Not every distro provides a sane way to deal with this...
Well, may be that's one more reason for fixing it before boards start
to
Quoted 1 line
ship and run distro kernels. If some boards are already in the wild
running
Quoted 8 lines
Quoted 1 line
on a BSP kernel, people will complain that mainline reorders their
network
Quoted 1 line
ports and will stick to the BSP kernel instead.
I've checked, and we believe not many people are using mainline kernels on clearfog. I'll have to adjust my debian jessie setup to fix the resulting carnage though. On the plus side, it means that mainline matches Marvell's kernel for interface naming.
Well, now I'm one of these and I confirm that it's really painful to have a different ordering between the DTB provided in mainline and the DTB provided with the board. The worst thing is that using mainline, eth0 corresponds to the switch and appears up, so you don't immediately realize that it's not where your cable is connected :-/

Quoted 3 lines

Thus could we please get Thomas' patch to ensure that the board boots with similar interface naming with both the original and mainline kernel ?

Shouldn't this change be device specific then instead of messing up the dtsi?

Cheers,

Imre

Willy Tarreau

2016-02-24 19:07 UTC #8 lore raw

PATCH 1/2 ARM: mvebu: change order of ethernet DT nodes on Armada 38x

Quoted 11 lines · Imre Kaloz

Quoted 11 lines
Quoted 4 lines
I've checked, and we believe not many people are using mainline kernels on clearfog. I'll have to adjust my debian jessie setup to fix the resulting carnage though. On the plus side, it means that mainline matches Marvell's kernel for interface naming.
Well, now I'm one of these and I confirm that it's really painful to have a different ordering between the DTB provided in mainline and the DTB provided with the board. The worst thing is that using mainline, eth0 corresponds to the switch and appears up, so you don't immediately realize that it's not where your cable is connected :-/

Quoted 6 lines

Quoted 3 lines
Thus could we please get Thomas' patch to ensure that the board boots with similar interface naming with both the original and mainline kernel ?
Shouldn't this change be device specific then instead of messing up the dtsi?

According to what Thomas explained, it's related to the ordering of the NICs in the SoC's datasheet, so there are little chances that anyone would see them enumerated differently on their board.

Willy

Thomas Petazzoni

author

2016-02-25 10:32 UTC #9 lore raw

PATCH 1/2 ARM: mvebu: change order of ethernet DT nodes on Armada 38x

In reply to Imre Kaloz

Dear Imre Kaloz,

Quoted 2 lines · Imre Kaloz

Shouldn't this change be device specific then instead of messing up the dtsi?

It's the order in which the nodes are defined at the top-most level that counts. So if you change the order of the nodes in the .dts files, it won't change anything.

Best regards,

Thomas

Signature · 3 lines

Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

Russell King - ARM Linux

2016-02-24 22:33 UTC #10 lore raw

PATCH 1/2 ARM: mvebu: change order of ethernet DT nodes on Armada 38x

In reply to Willy Tarreau

Quoted 9 lines · Willy Tarreau

Hi Russell,
Well, now I'm one of these and I confirm that it's really painful to have a different ordering between the DTB provided in mainline and the DTB provided with the board. The worst thing is that using mainline, eth0 corresponds to the switch and appears up, so you don't immediately realize that it's not where your cable is connected :-/
Thus could we please get Thomas' patch to ensure that the board boots with similar interface naming with both the original and mainline kernel ?

I didn't say no to it, I merely asked a few pertinent questions and made some pertinent points.

Let me restate:

* Today, people who switch between mainline and vendor kernels experience some pain due to the NIC order changing.

* Mainline has had support for Armada 38x for 2 years now, which is long enough for it to have gained users. AFAICS, there haven't been any complaints about the different NIC ordering. Changing the NIC ordering is going to cause breakage to these users when they migrate across the change.

By making the change, we're effectively telling these mainline-only users "we don't care about your setups, we're going to break them" because that's exactly what we're going to do.

Of course, if no one complains about the change, you've got away with it.

I think the folk who want to make this change should be prepared for userspace breakage reports: Linus feels very strongly about zero userspace regressions, as we all should know.

What I'm saying, therefore, is make the change, but if you have one report that someone's userspace setup has broken as a result of the change, the change must be reverted.

Signature · 3 lines

RMK's Patch system: http://www.arm.linux.org.uk/developer/patches/
FTTC broadband for 0.8mile line: currently at 9.6Mbps down 400kbps up
according to speedtest.net.

Willy Tarreau

2016-02-24 22:56 UTC #11 lore raw

PATCH 1/2 ARM: mvebu: change order of ethernet DT nodes on Armada 38x

Quoted 8 lines · Russell King - ARM Linux

Quoted 7 lines · Willy Tarreau
Hi Russell,
Well, now I'm one of these and I confirm that it's really painful to have a different ordering between the DTB provided in mainline and the DTB provided with the board. The worst thing is that using mainline, eth0 corresponds to the switch and appears up, so you don't immediately realize that it's not where your cable is connected :-/

Quoted 5 lines

Quoted 2 lines
Thus could we please get Thomas' patch to ensure that the board boots with similar interface naming with both the original and mainline kernel ?
I didn't say no to it, I merely asked a few pertinent questions and made some pertinent points.

Yes I'm well aware, which is why I was bringing some feedback.

Quoted 7 lines

Let me restate:
* Today, people who switch between mainline and vendor kernels experience some pain due to the NIC order changing.
* Mainline has had support for Armada 38x for 2 years now, which is long enough for it to have gained users.

I didn't realize it had been *that* long, I thought early support was very limited, but I can indeed see that the first dtsi was already fairly complete (though lots of patches came up recently, including one to disable IP checksumming on jumbo frames that's only 3 months old).

Quoted 4 lines

AFAICS, there haven't been
any complaints about the different NIC ordering. Changing the NIC ordering is going to cause breakage to these users when they migrate across the change.

Yes, and the problem is that there are two expected breakages : going from LSP to "old mainline" and going from "old mainline" to "new mainline". Since boards ship with LSP by default, I think that not doing the change will lead to causing breakage at least once for each user going to mainline, while doing it will only cause breakage to users who have already experienced breakage.

Quoted 3 lines

By making the change, we're effectively telling these mainline-only users "we don't care about your setups, we're going to break them" because that's exactly what we're going to do.

Not exactly, we're telling "if you moved away from the original kernel, you have already experienced breakage so apparently you know how to fix it, and if it's the first time you move away from it you won't notice anything" :-)

Quoted 6 lines

Of course, if no one complains about the change, you've got away with it.
I think the folk who want to make this change should be prepared for userspace breakage reports: Linus feels very strongly about zero userspace regressions, as we all should know.

Oh yes and I've always agreed with him regarding this!

Quoted 3 lines

What I'm saying, therefore, is make the change, but if you have one report that someone's userspace setup has broken as a result of the change, the change must be reverted.

Makes sense. If any breakage is reported, it would be interested to know if it's on a multi-port board or single-port one.

Thanks,
Willy

Thomas Petazzoni

author

2016-02-25 10:36 UTC #12 lore raw

PATCH 1/2 ARM: mvebu: change order of ethernet DT nodes on Armada 38x

In reply to Russell King - ARM Linux

Hello,

Quoted 14 lines · Russell King - ARM Linux

I didn't say no to it, I merely asked a few pertinent questions and made some pertinent points.
Let me restate:
* Today, people who switch between mainline and vendor kernels experience some pain due to the NIC order changing.
* Mainline has had support for Armada 38x for 2 years now, which is long enough for it to have gained users. AFAICS, there haven't been any complaints about the different NIC ordering. Changing the NIC ordering is going to cause breakage to these users when they migrate across the change.

Quoted 6 lines

By making the change, we're effectively telling these mainline-only users "we don't care about your setups, we're going to break them" because that's exactly what we're going to do.
Of course, if no one complains about the change, you've got away with it.

I agree with you that the change isn't perfect, it's really a matter of trade-off. The perfect change would be to be able to help the network subsystem in its naming of network interfaces, but this has always been rejected. It would have indeed been better to add this DT node ordering work-around earlier, but we didn't do it, and we're in the present.

So again, yes the proposed change is not "great", but I believe it's relatively reasonable. If we indeed get feedback that it's breaking things for too many people, I guess we'll just revert. I don't mind if things remain as they are today, i.e with a weird ordering of the network interfaces. It's just annoying for new people using the platform, but it's not horrible either.

Best regards,

Thomas

Signature · 3 lines

Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

Imre Kaloz

2016-02-25 13:48 UTC #13 lore raw

PATCH 1/2 ARM: mvebu: change order of ethernet DT nodes on Armada 38x

Quoted 31 lines · Thomas Petazzoni

Hello,
Quoted 14 lines · Russell King - ARM Linux
I didn't say no to it, I merely asked a few pertinent questions and made some pertinent points.
Let me restate:
* Today, people who switch between mainline and vendor kernels experience some pain due to the NIC order changing.
* Mainline has had support for Armada 38x for 2 years now, which is long enough for it to have gained users. AFAICS, there haven't been any complaints about the different NIC ordering. Changing the NIC ordering is going to cause breakage to these users when they migrate across the change.
Quoted 6 lines
By making the change, we're effectively telling these mainline-only users "we don't care about your setups, we're going to break them" because that's exactly what we're going to do.
Of course, if no one complains about the change, you've got away with it.
I agree with you that the change isn't perfect, it's really a matter of trade-off. The perfect change would be to be able to help the network subsystem in its naming of network interfaces, but this has always been rejected. It would have indeed been better to add this DT node ordering work-around earlier, but we didn't do it, and we're in the present.

Quoted 6 lines

So again, yes the proposed change is not "great", but I believe it's relatively reasonable. If we indeed get feedback that it's breaking things for too many people, I guess we'll just revert. I don't mind if things remain as they are today, i.e with a weird ordering of the network interfaces. It's just annoying for new people using the platform, but it's not horrible either.

I can check during the weekend if the mainlined Linksys boards get broken because of this.

Imre

Thomas Petazzoni

author

2016-02-25 10:31 UTC #14 lore raw

PATCH 1/2 ARM: mvebu: change order of ethernet DT nodes on Armada 38x

In reply to Willy Tarreau

Willy,

Quoted 2 lines · Willy Tarreau

Thus could we please get Thomas' patch to ensure that the board boots with similar interface naming with both the original and mainline kernel ?

My patch has already been applied by the mvebu maintainers in the mvebu/dt branch:
http://git.infradead.org/linux-mvebu.git/commit/cb4f71c4298853db0c6751b1209e4535956f136c.

Best regards,

Thomas

Signature · 3 lines

Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

Willy Tarreau

2016-02-25 10:40 UTC #15 lore raw

PATCH 1/2 ARM: mvebu: change order of ethernet DT nodes on Armada 38x

Hi Thomas,

Quoted 3 lines · Thomas Petazzoni

My patch has already been applied by the mvebu maintainers in the mvebu/dt branch:
http://git.infradead.org/linux-mvebu.git/commit/cb4f71c4298853db0c6751b1209e4535956f136c.

Ah thanks, I didn't notice!

Best regards,
willy

Thomas Petazzoni

author

2016-01-27 15:08 UTC #16 lore raw

PATCH 2/2 ARM: mvebu: add reference to ETH connectors for A385-AP

In reply to Thomas Petazzoni

This commit adds some comments to the Armada 385 AP Device Tree description to indicate which Ethernet interface matches which physical connector on the board.

Signed-off-by Thomas Petazzoni <[email protected]>

Cc Maxime Ripard <[email protected]>

---
 arch/arm/boot/dts/armada-385-db-ap.dts | 3 +++
 1 file changed, 3 insertions(+)

arch/arm/boot/dts/armada-385-db-ap.dts +3 -0

index acd5b15..86fde0b 100644
@@ -134,18 +134,21 @@
 				};
 			};
 
+			/* CON3 */
 			ethernet at 30000 {
 				status = "okay";
 				phy = <&phy2>;
 				phy-mode = "sgmii";
 			};
 
+			/* CON2 */
 			ethernet at 34000 {
 				status = "okay";
 				phy = <&phy1>;
 				phy-mode = "sgmii";
 			};
 
+			/* CON4 */
 			ethernet at 70000 {
 				pinctrl-names = "default";

Signature · 1 line

2.6.4

Gregory CLEMENT

2016-02-08 14:48 UTC #17 lore raw

cover ARM: mvebu: network interface naming

In reply to Thomas Petazzoni

Hi Thomas,

Quoted 16 lines · Thomas Petazzoni

Dear mvebu maintainers,
As was discussed back in November last year, the naming of network interfaces on Armada 38x is very confusing. Gregory Clement suggested to exceptionally violate the rule of "DT nodes should be ordered by register address" in order to reduce the confusion.
The first patch implements this proposal.
The second patch is a new iteration of the patch that adds comments to the Armada 385 DB AP Device Tree file. The contents of the patch is the same (add indications of which physical connector corresponds each port), only the commit log is adjusted to take into account the fact that thanks to the first patch, the naming of the network interfaces visible from Linux is now a lot more sane.

Quoted 8 lines

Thanks!
Thomas
Thomas Petazzoni (2):
ARM: mvebu: change order of ethernet DT nodes on Armada 38x
ARM: mvebu: add reference to ETH connectors for A385-AP

Even if we know that it is not perfect, I applied both pacthes on mvebu/dt

Thanks,

Gregory

Quoted 7 lines

 arch/arm/boot/dts/armada-385-db-ap.dts |  3 +++
 arch/arm/boot/dts/armada-38x.dtsi      | 30 +++++++++++++++++++++---------
 2 files changed, 24 insertions(+), 9 deletions(-)

Signature · 1 line

2.6.4

Signature · 4 lines

Gregory Clement, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com