powerpc/83xx: don't probe broken PCI on mpc837x_mds boards

PATCH Superseded Anton Vorontsov 2 participants 8 messages Oct 2, 2008 – Oct 7, 2008

Subsystems (2)

PowerPC Linux for PowerPC embedded PPC8XX and PPC83XX

lore

Share & tools

lore

Anton Vorontsov

author

2008-10-02 15:18 UTC #0 lore raw

PATCH 1/1 powerpc/83xx: don't probe broken PCI on mpc837x_mds boards

In the standalone setup the board's CPLD disables the PCI internal arbiter, thus any access to the PCI bus will hang the board.

When there is no PCI arbiter on the bus the u-boot marks the PCI bus as broken and we should not try to probe it.

Signed-off-by Anton Vorontsov <[email protected]>

---
 arch/powerpc/platforms/83xx/mpc837x_mds.c |    8 +++++++-
 1 files changed, 7 insertions(+), 1 deletions(-)

arch/powerpc/platforms/83xx/mpc837x_mds.c +7 -1

index be62de2..8bb13c8 100644
@@ -85,8 +85,14 @@ static void __init mpc837x_mds_setup_arch(void)
 		ppc_md.progress("mpc837x_mds_setup_arch()", 0);
 
 #ifdef CONFIG_PCI
-	for_each_compatible_node(np, "pci", "fsl,mpc8349-pci")
+	for_each_compatible_node(np, "pci", "fsl,mpc8349-pci") {
+		if (!of_device_is_available(np)) {
+			pr_warning("%s: disabled by the firmware.\n",
+				   np->full_name);
+			continue;
+		}
 		mpc83xx_add_bridge(np);
+	}
 #endif
 	mpc837xmds_usb_cfg();
 }

Signature · 1 line

1.5.6.3

Kumar Gala

2008-10-02 19:48 UTC #1 lore raw

Quoted 7 lines · Anton Vorontsov

In the standalone setup the board's CPLD disables the PCI internal arbiter, thus any access to the PCI bus will hang the board.
When there is no PCI arbiter on the bus the u-boot marks the PCI bus as broken and we should not try to probe it.
Signed-off-by Anton Vorontsov <[email protected]>

Can you add to the commit message something about using the 'status field in the device tree to determine if the pci controller is available'

- k

Anton Vorontsov

author

2008-10-03 16:08 UTC #2 lore raw

PATCH v2 1/1 powerpc/83xx: don't probe broken PCI on mpc837x_mds boards

In the standalone setup the board's CPLD disables the PCI internal arbiter, thus any access to the PCI bus will hang the board.

When there is no PCI arbiter on the bus the u-boot adds status = "broken (no arbiter)" property into the PCI controller's node, and so marks the PCI controller as unavailable.

Signed-off-by Anton Vorontsov <[email protected]>

---

Quoted 3 lines · Kumar Gala

Can you add to the commit message something about using the 'status field in the device tree to determine if the pci controller is available'

Something like this?

Thanks,

 arch/powerpc/platforms/83xx/mpc837x_mds.c |    8 +++++++-
 1 files changed, 7 insertions(+), 1 deletions(-)

arch/powerpc/platforms/83xx/mpc837x_mds.c +7 -1

index be62de2..8bb13c8 100644
@@ -85,8 +85,14 @@ static void __init mpc837x_mds_setup_arch(void)
 		ppc_md.progress("mpc837x_mds_setup_arch()", 0);
 
 #ifdef CONFIG_PCI
-	for_each_compatible_node(np, "pci", "fsl,mpc8349-pci")
+	for_each_compatible_node(np, "pci", "fsl,mpc8349-pci") {
+		if (!of_device_is_available(np)) {
+			pr_warning("%s: disabled by the firmware.\n",
+				   np->full_name);
+			continue;
+		}
 		mpc83xx_add_bridge(np);
+	}
 #endif
 	mpc837xmds_usb_cfg();
 }

Signature · 1 line

1.5.6.3

Kumar Gala

2008-10-03 16:14 UTC #3 lore raw

Quoted 10 lines · Anton Vorontsov

In the standalone setup the board's CPLD disables the PCI internal arbiter, thus any access to the PCI bus will hang the board.
When there is no PCI arbiter on the bus the u-boot adds status = "broken (no arbiter)" property into the PCI controller's node, and so marks the PCI controller as unavailable.
Signed-off-by Anton Vorontsov <[email protected]>
---

Quoted 8 lines

Quoted 3 lines · Kumar Gala
Can you add to the commit message something about using the 'status field in the device tree to determine if the pci controller is available'
Something like this?
Thanks,

yes, but should we just have "status = disabled" since that is the effect?

- k

Anton Vorontsov

author

2008-10-03 16:27 UTC #4 lore raw

Quoted 18 lines · Kumar Gala

Quoted 15 lines · Anton Vorontsov
In the standalone setup the board's CPLD disables the PCI internal arbiter, thus any access to the PCI bus will hang the board.
When there is no PCI arbiter on the bus the u-boot adds status = "broken (no arbiter)" property into the PCI controller's node, and so marks the PCI controller as unavailable.
Signed-off-by Anton Vorontsov <[email protected]>
---
Quoted 3 lines · Kumar Gala
Can you add to the commit message something about using the 'status field in the device tree to determine if the pci controller is available'

Quoted 6 lines

Quoted 3 lines
Something like this?
Thanks,
yes, but should we just have "status = disabled" since that is the effect?

I don't know, should we? For the unavailable/disabled case the status can be anything but not 'ok' or 'okay' (the only status values for the available devices). So if we can encode the reason, why not do this?

Signature · 3 lines

Anton Vorontsov
email: [email protected]
irc://irc.freenode.net/bd2

Kumar Gala

2008-10-03 17:52 UTC #5 lore raw

Quoted 21 lines · Anton Vorontsov

Quoted 20 lines · Kumar Gala
Quoted 17 lines · Anton Vorontsov
In the standalone setup the board's CPLD disables the PCI internal arbiter, thus any access to the PCI bus will hang the board.
When there is no PCI arbiter on the bus the u-boot adds status = "broken (no arbiter)" property into the PCI controller's node, and so marks the PCI controller as unavailable.
Signed-off-by Anton Vorontsov <[email protected]>
---
Quoted 3 lines · Kumar Gala
Can you add to the commit message something about using the 'status field in the device tree to determine if the pci controller is available'
Something like this?

Quoted 8 lines

Quoted 4 lines
Quoted 1 line
Thanks,
yes, but should we just have "status = disabled" since that is the effect?
I don't know, should we? For the unavailable/disabled case the status can be anything but not 'ok' or 'okay' (the only status values for the available devices). So if we can encode the reason, why not do this?

that works for me, just add the fact to the commit msg that the "valid status's are 'ok' and 'okay' and everything else is treated as "not available or disabled"

- k

Anton Vorontsov

author

2008-10-06 17:08 UTC #6 lore raw

PATCH v3 1/1 powerpc/83xx: don't probe broken PCI on mpc837x_mds boards

In the standalone setup the board's CPLD disables the PCI internal arbiter, thus any access to the PCI bus will hang the board.

The common way to disable particular devices in the device tree is to put the "status" property with any value other than "ok" or "okay" into the device node we want to disable.

So, when there is no PCI arbiter on the bus the u-boot adds status = "broken (no arbiter)" property into the PCI controller's node, and so marks the PCI controller as unavailable.

Signed-off-by Anton Vorontsov <[email protected]>

---

On Fri, Oct 03, 2008 at 12:51:41PM -0500, Kumar Gala wrote:
[...]

Quoted 10 lines

Quoted 6 lines
Quoted 2 lines
yes, but should we just have "status = disabled" since that is the effect?
I don't know, should we? For the unavailable/disabled case the status can be anything but not 'ok' or 'okay' (the only status values for the available devices). So if we can encode the reason, why not do this?
that works for me, just add the fact to the commit msg that the "valid status's are 'ok' and 'okay' and everything else is treated as "not available or disabled"

Done.

 arch/powerpc/platforms/83xx/mpc837x_mds.c |    8 +++++++-
 1 files changed, 7 insertions(+), 1 deletions(-)

arch/powerpc/platforms/83xx/mpc837x_mds.c +7 -1

index be62de2..8bb13c8 100644
@@ -85,8 +85,14 @@ static void __init mpc837x_mds_setup_arch(void)
 		ppc_md.progress("mpc837x_mds_setup_arch()", 0);
 
 #ifdef CONFIG_PCI
-	for_each_compatible_node(np, "pci", "fsl,mpc8349-pci")
+	for_each_compatible_node(np, "pci", "fsl,mpc8349-pci") {
+		if (!of_device_is_available(np)) {
+			pr_warning("%s: disabled by the firmware.\n",
+				   np->full_name);
+			continue;
+		}
 		mpc83xx_add_bridge(np);
+	}
 #endif
 	mpc837xmds_usb_cfg();
 }

Signature · 1 line

1.5.6.3

Kumar Gala

Applied

2008-10-07 12:32 UTC #7 lore raw

Quoted 7 lines · Anton Vorontsov

In the standalone setup the board's CPLD disables the PCI internal arbiter, thus any access to the PCI bus will hang the board.
The common way to disable particular devices in the device tree is to put the "status" property with any value other than "ok" or "okay" into the device node we want to disable.

Quoted 6 lines

So, when there is no PCI arbiter on the bus the u-boot adds status = "broken (no arbiter)" property into the PCI controller's node, and so marks the PCI controller as unavailable.
Signed-off-by Anton Vorontsov <[email protected]>
---

applied

- k