On Tue, Nov 22, 2011 at 2:11 AM, Tanmay Inamdar [off-list ref] wrote:
Fixing interrupt mapping of EMAC for canyonlands
Signed-off-by: Tanmay Inamdar <redacted>
As far as I can tell, your changes aren't really changing anything
just making it a bit clearer, correct? If so, do you mind if I change
the commit log to "clear up" instead of fix?
josh
quoted hunk
---
=A0arch/powerpc/boot/dts/canyonlands.dts | =A0 16 ++++++----------
=A01 files changed, 6 insertions(+), 10 deletions(-)
diff --git a/arch/powerpc/boot/dts/canyonlands.dts b/arch/powerpc/boot/dt=
s/canyonlands.dts
quoted hunk
index 3dc75de..c76bbcd 100644
--- a/arch/powerpc/boot/dts/canyonlands.dts
+++ b/arch/powerpc/boot/dts/canyonlands.dts
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0EMAC0: ethernet@ef600e00 {
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0device_typ=
e =3D "network";
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0compatible=
=3D "ibm,emac-460ex", "ibm,emac4sync";
- =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 interrupt-p=
arent =3D <&EMAC0>;
- =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 interrupts =
=3D <0x0 0x1>;
- =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 #interrupt-=
cells =3D <1>;
+ =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 interrupt-p=
arent =3D <&UIC2>;
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0#address-c=
ells =3D <0>;
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0#size-cell=
s =3D <0>;
- =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 interrupt-m=
ap =3D </*Status*/ 0x0 &UIC2 0x10 0x4
- =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0=
=A0 =A0 =A0 =A0 =A0 =A0/*Wake*/ =A0 0x1 &UIC2 0x14 0x4>;
+ =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 interrupts =
=3D </*Status*/0x10 0x4
+ =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0=
=A0 =A0 =A0 =A0 =A0 /*Wake*/0x14 0x4>;
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0reg =3D <0=
xef600e00 0x000000c4>;
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0local-mac-=
address =3D [000000000000]; /* Filled in by U-Boot */
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0mal-device=
=3D <&MAL0>;
quoted hunk
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0EMAC1: ethernet@ef600f00 {
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0device_typ=
e =3D "network";
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0compatible=
=3D "ibm,emac-460ex", "ibm,emac4sync";
- =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 interrupt-p=
arent =3D <&EMAC1>;
- =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 interrupts =
=3D <0x0 0x1>;
- =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 #interrupt-=
cells =3D <1>;
+ =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 interrupt-p=
arent =3D <&UIC2>;
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0#address-c=
ells =3D <0>;
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0#size-cell=
s =3D <0>;
- =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 interrupt-m=
ap =3D </*Status*/ 0x0 &UIC2 0x11 0x4
- =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0=
=A0 =A0 =A0 =A0 =A0 =A0/*Wake*/ =A0 0x1 &UIC2 0x15 0x4>;
+ =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 interrupts =
=3D </*Status*/0x11 0x4
+ =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0=
=A0 =A0 =A0 =A0 =A0 /*Wake*/0x15 0x4>;
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0reg =3D <0=
xef600f00 0x000000c4>;
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0local-mac-=
address =3D [000000000000]; /* Filled in by U-Boot */
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0mal-device=
=3D <&MAL0>;
--
1.6.1.rc3
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" i=
n
the body of a message to majordomo@vger.kernel.org
More majordomo info at =A0http://vger.kernel.org/majordomo-info.html
Please read the FAQ at =A0http://www.tux.org/lkml/
On Tue, Nov 22, 2011 at 5:00 PM, Josh Boyer [off-list ref] wrote:
On Tue, Nov 22, 2011 at 2:11 AM, Tanmay Inamdar [off-list ref] wrote:
quoted
Fixing interrupt mapping of EMAC for canyonlands
Signed-off-by: Tanmay Inamdar <redacted>
As far as I can tell, your changes aren't really changing anything
just making it a bit clearer, correct? If so, do you mind if I change
the commit log to "clear up" instead of fix?
Actually Rob Herring's commit (
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=commit;h=dc9372808412edbc653a675a526c2ee6c0c14a91)
breaks the interrupt mapping in EMAC driver.
I am trying to fix this issue by mapping interrupts in different way.
If you think "clear up" is fine, then please go ahead.
Thanks,
Tanmay
josh
quoted
---
arch/powerpc/boot/dts/canyonlands.dts | 16 ++++++----------
1 files changed, 6 insertions(+), 10 deletions(-)
diff --git a/arch/powerpc/boot/dts/canyonlands.dts
b/arch/powerpc/boot/dts/canyonlands.dts
quoted
index 3dc75de..c76bbcd 100644
--- a/arch/powerpc/boot/dts/canyonlands.dts
+++ b/arch/powerpc/boot/dts/canyonlands.dts
@@ -360,13 +360,11 @@
EMAC0: ethernet@ef600e00 {
device_type = "network";
compatible = "ibm,emac-460ex",
"ibm,emac4sync";
quoted
- interrupt-parent = <&EMAC0>;
- interrupts = <0x0 0x1>;
- #interrupt-cells = <1>;
+ interrupt-parent = <&UIC2>;
#address-cells = <0>;
#size-cells = <0>;
- interrupt-map = </*Status*/ 0x0 &UIC2
0x10 0x4
quoted
- /*Wake*/ 0x1 &UIC2
0x14 0x4>;
quoted
+ interrupts = </*Status*/0x10 0x4
+ /*Wake*/0x14 0x4>;
reg = <0xef600e00 0x000000c4>;
local-mac-address = [000000000000]; /*
Filled in by U-Boot */
quoted
mal-device = <&MAL0>;
@@ -390,13 +388,11 @@
EMAC1: ethernet@ef600f00 {
device_type = "network";
compatible = "ibm,emac-460ex",
"ibm,emac4sync";
quoted
- interrupt-parent = <&EMAC1>;
- interrupts = <0x0 0x1>;
- #interrupt-cells = <1>;
+ interrupt-parent = <&UIC2>;
#address-cells = <0>;
#size-cells = <0>;
- interrupt-map = </*Status*/ 0x0 &UIC2
0x11 0x4
quoted
- /*Wake*/ 0x1 &UIC2
0x15 0x4>;
quoted
+ interrupts = </*Status*/0x11 0x4
+ /*Wake*/0x15 0x4>;
reg = <0xef600f00 0x000000c4>;
local-mac-address = [000000000000]; /*
Filled in by U-Boot */
quoted
mal-device = <&MAL0>;
--
1.6.1.rc3
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel"
in
quoted
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
CONFIDENTIALITY NOTICE: This e-mail message, including any attachments,
is for the sole use of the intended recipient(s) and contains information
that is confidential and proprietary to AppliedMicro Corporation or its subsidiaries.
It is to be used solely for the purpose of furthering the parties' business relationship.
All unauthorized review, use, disclosure or distribution is prohibited.
If you are not the intended recipient, please contact the sender by reply e-mail
and destroy all copies of the original message.
On Tue, Nov 22, 2011 at 9:15 AM, Tanmay Inamdar [off-list ref] wrote:
On Tue, Nov 22, 2011 at 5:00 PM, Josh Boyer [off-list ref] wrote:
quoted
On Tue, Nov 22, 2011 at 2:11 AM, Tanmay Inamdar [off-list ref] wrote=
:
quoted
quoted
Fixing interrupt mapping of EMAC for canyonlands
Signed-off-by: Tanmay Inamdar <redacted>
As far as I can tell, your changes aren't really changing anything
just making it a bit clearer, correct? =A0If so, do you mind if I change
the commit log to "clear up" instead of fix?
Actually Rob Herring's commit
(http://git.kernel.org/?p=3Dlinux/kernel/git/torvalds/linux.git;a=3Dcommi=
t;h=3Ddc9372808412edbc653a675a526c2ee6c0c14a91)
breaks the interrupt mapping in EMAC driver.
I am trying to fix this issue by mapping interrupts in different way.
I see. That should have been in the commit log then. I'll add
something like that to it.
josh
On Tue, 2011-11-22 at 12:41 +0530, Tanmay Inamdar wrote:
Fixing interrupt mapping of EMAC for canyonlands
The previous stuff was odd .... but was it broken ?
It was done this way because the EMAC actually has more interrupts than
that which are routed to different UICs, and so doing a local map this
way allows to target multiple parents.
Cheers,
Ben.
quoted hunk
Signed-off-by: Tanmay Inamdar <redacted>
---
arch/powerpc/boot/dts/canyonlands.dts | 16 ++++++----------
1 files changed, 6 insertions(+), 10 deletions(-)
diff --git a/arch/powerpc/boot/dts/canyonlands.dts b/arch/powerpc/boot/dts/canyonlands.dts
index 3dc75de..c76bbcd 100644
--- a/arch/powerpc/boot/dts/canyonlands.dts
+++ b/arch/powerpc/boot/dts/canyonlands.dts
@@ -360,13 +360,11 @@
EMAC0: ethernet@ef600e00 {
device_type = "network";
compatible = "ibm,emac-460ex", "ibm,emac4sync";
- interrupt-parent = <&EMAC0>;
- interrupts = <0x0 0x1>;
- #interrupt-cells = <1>;
+ interrupt-parent = <&UIC2>;
#address-cells = <0>;
#size-cells = <0>;
- interrupt-map = </*Status*/ 0x0 &UIC2 0x10 0x4
- /*Wake*/ 0x1 &UIC2 0x14 0x4>;
+ interrupts = </*Status*/0x10 0x4
+ /*Wake*/0x14 0x4>;
reg = <0xef600e00 0x000000c4>;
local-mac-address = [000000000000]; /* Filled in by U-Boot */
mal-device = <&MAL0>;@@ -390,13 +388,11 @@
EMAC1: ethernet@ef600f00 {
device_type = "network";
compatible = "ibm,emac-460ex", "ibm,emac4sync";
- interrupt-parent = <&EMAC1>;
- interrupts = <0x0 0x1>;
- #interrupt-cells = <1>;
+ interrupt-parent = <&UIC2>;
#address-cells = <0>;
#size-cells = <0>;
- interrupt-map = </*Status*/ 0x0 &UIC2 0x11 0x4
- /*Wake*/ 0x1 &UIC2 0x15 0x4>;
+ interrupts = </*Status*/0x11 0x4
+ /*Wake*/0x15 0x4>;
reg = <0xef600f00 0x000000c4>;
local-mac-address = [000000000000]; /* Filled in by U-Boot */
mal-device = <&MAL0>;
On Tue, 2011-11-22 at 19:45 +0530, Tanmay Inamdar wrote:
On Tue, Nov 22, 2011 at 5:00 PM, Josh Boyer [off-list ref] wrote:
On Tue, Nov 22, 2011 at 2:11 AM, Tanmay Inamdar
[off-list ref] wrote:
> Fixing interrupt mapping of EMAC for canyonlands
>
> Signed-off-by: Tanmay Inamdar [off-list ref]
As far as I can tell, your changes aren't really changing
anything
just making it a bit clearer, correct? If so, do you mind if
I change
the commit log to "clear up" instead of fix?
Actually Rob Herring's commit
(http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=commit;h=dc9372808412edbc653a675a526c2ee6c0c14a91) breaks the interrupt mapping in EMAC driver.
I am trying to fix this issue by mapping interrupts in different way.
No. This commit needs to be reverted. It breaks existing practices.
Pointing to yourself as a parent in order to provide a map is an old
trick and it should be supported.
I'll ask Linus to revert.
Cheers,
Ben.
If you think "clear up" is fine, then please go ahead.
Thanks,
Tanmay
josh
> ---
> arch/powerpc/boot/dts/canyonlands.dts | 16
++++++----------
> 1 files changed, 6 insertions(+), 10 deletions(-)
>
> diff --git a/arch/powerpc/boot/dts/canyonlands.dts
b/arch/powerpc/boot/dts/canyonlands.dts
> index 3dc75de..c76bbcd 100644
> --- a/arch/powerpc/boot/dts/canyonlands.dts
> +++ b/arch/powerpc/boot/dts/canyonlands.dts
> @@ -360,13 +360,11 @@
> EMAC0: ethernet@ef600e00 {
> device_type = "network";
> compatible =
"ibm,emac-460ex", "ibm,emac4sync";
> - interrupt-parent = <&EMAC0>;
> - interrupts = <0x0 0x1>;
> - #interrupt-cells = <1>;
> + interrupt-parent = <&UIC2>;
> #address-cells = <0>;
> #size-cells = <0>;
> - interrupt-map = </*Status*/
0x0 &UIC2 0x10 0x4
> - /*Wake*/
0x1 &UIC2 0x14 0x4>;
> + interrupts = </*Status*/0x10
0x4
> + /*Wake*/0x14
0x4>;
> reg = <0xef600e00
0x000000c4>;
> local-mac-address =
[000000000000]; /* Filled in by U-Boot */
> mal-device = <&MAL0>;
> @@ -390,13 +388,11 @@
> EMAC1: ethernet@ef600f00 {
> device_type = "network";
> compatible =
"ibm,emac-460ex", "ibm,emac4sync";
> - interrupt-parent = <&EMAC1>;
> - interrupts = <0x0 0x1>;
> - #interrupt-cells = <1>;
> + interrupt-parent = <&UIC2>;
> #address-cells = <0>;
> #size-cells = <0>;
> - interrupt-map = </*Status*/
0x0 &UIC2 0x11 0x4
> - /*Wake*/
0x1 &UIC2 0x15 0x4>;
> + interrupts = </*Status*/0x11
0x4
> + /*Wake*/0x15
0x4>;
> reg = <0xef600f00
0x000000c4>;
> local-mac-address =
[000000000000]; /* Filled in by U-Boot */
> mal-device = <&MAL0>;
> --
> 1.6.1.rc3
>
> --
> To unsubscribe from this list: send the line "unsubscribe
linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at
http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
CONFIDENTIALITY NOTICE: This e-mail message, including any attachments,
is for the sole use of the intended recipient(s) and contains information
that is confidential and proprietary to AppliedMicro Corporation or its subsidiaries.
It is to be used solely for the purpose of furthering the parties' business relationship.
All unauthorized review, use, disclosure or distribution is prohibited.
If you are not the intended recipient, please contact the sender by reply e-mail
and destroy all copies of the original message.
On Wed, Nov 23, 2011 at 08:16:40AM +1100, Benjamin Herrenschmidt wrote:
On Tue, 2011-11-22 at 12:41 +0530, Tanmay Inamdar wrote:
quoted
Fixing interrupt mapping of EMAC for canyonlands
The previous stuff was odd .... but was it broken ?
It was done this way because the EMAC actually has more interrupts than
that which are routed to different UICs, and so doing a local map this
way allows to target multiple parents.
Well, in the canyonlands case, it appears that the interrupts went to
the same pic, so the simpler representation should be correct as
well. However, there certainly are boards where they go to multiple
pics, so we need this interrupt-map trick.
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
On Wed, 2011-11-23 at 12:15 +1100, David Gibson wrote:
On Wed, Nov 23, 2011 at 08:16:40AM +1100, Benjamin Herrenschmidt wrote:
quoted
On Tue, 2011-11-22 at 12:41 +0530, Tanmay Inamdar wrote:
quoted
Fixing interrupt mapping of EMAC for canyonlands
The previous stuff was odd .... but was it broken ?
It was done this way because the EMAC actually has more interrupts than
that which are routed to different UICs, and so doing a local map this
way allows to target multiple parents.
Well, in the canyonlands case, it appears that the interrupts went to
the same pic, so the simpler representation should be correct as
well. However, there certainly are boards where they go to multiple
pics, so we need this interrupt-map trick.
Doesn't emac has something like 4 more interrupts that we simply haven't
been bothered wiring up (because we don't use them ?)
Cheers,
Ben.