drivers: bus: check cci device tree node status

The arndale-octa board was giving "imprecise external aborts" during boot-up with MCPM enabled. CCI enablement of the boot cluster was found to be the cause of these aborts (possibly because the secure f/w was not allowing it). Hence, disable CCI for the arndale-octa board.

Signed-off-by Abhilash Kesavan <[email protected]>

---
 arch/arm/boot/dts/exynos5420-arndale-octa.dts |    4 ++++
 arch/arm/boot/dts/exynos5420.dtsi             |    2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)

arch/arm/boot/dts/exynos5420-arndale-octa.dts +4 -0

index aa7a7d7..db2c1c4 100644
@@ -372,3 +372,7 @@
 &usbdrd_dwc3_1 {
 	dr_mode = "host";
 };
+
+&cci {
+	status = "disabled";
+};

arch/arm/boot/dts/exynos5420.dtsi +1 -1

index 517e50f..0a82ae7 100644
@@ -120,7 +120,7 @@
 		};
 	};
 
-	cci at 10d20000 {
+	cci: cci at 10d20000 {
 		compatible = "arm,cci-400";
 		#address-cells = <1>;
 		#size-cells = <1>;

Quoted 10 lines · Abhilash Kesavan

The arndale-octa board was giving "imprecise external aborts" during boot-up with MCPM enabled. CCI enablement of the boot cluster was found to be the cause of these aborts (possibly because the secure f/w was not allowing it). Hence, disable CCI for the arndale-octa board.

Signed-off-by Abhilash Kesavan <[email protected]>

---
 arch/arm/boot/dts/exynos5420-arndale-octa.dts |    4 ++++
 arch/arm/boot/dts/exynos5420.dtsi             |    2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)

I tested these 2 patches on Arndale Octa but there are no improvements. I still got imprecise aborts (some not fatal and sometimes killing init with full backtrace).

Tested on next-20141128. System booted from microSD (bootloader from Linaro Ubuntu Saucy server image):

==================================
U-Boot 2012.07 (Feb 15 2014 - 17:29:55) for ARNDALE OCTA

CPU: Exynos5420 Rev2.0 [Samsung SOC on SMP Platform Base on ARM CortexA15] APLL = 800MHz, KPLL = 600MHz
MPLL = 532MHz, BPLL = 800MHz

Board: ARNDALE OCTA
DRAM: 2 GiB
WARNING: Caches not enabled

TrustZone Enabled BSP
BL1 version:

Checking Boot Mode ... SDMMC
MMC: S5P_MSHC2: 0, S5P_MSHC0: 1
MMC Device 0: 29.3 GiB
MMC Device 1: 7.3 GiB
MMC Device 2: MMC Device 2 not found
there are pending interrupts 0x00000001

In:    serial
Out:   serial
Err:   serial
Net:   No ethernet found.

(Re)start USB...
USB: Register 1313 NbrPorts 3
USB EHCI 1.00
scanning bus for devices... The request port(2) is not configured The request port(2) is not configured
3 USB Device(s) found
scanning bus for storage devices... 0 Storage Device(s) found scanning usb for ethernet devices... 1 Ethernet Device(s) found
Hit any key to stop autoboot: 0
ARNDALE # uTTY
Unknown command 'uTTY' - try 'help'
ARNDALE # run nn
reading uImage

3674704 bytes read
reading uInitrd

2330480 bytes read
reading board.dtb

33680 bytes read
## Booting kernel from Legacy Image at 20007000 ...

   Image Name:   Linux-exynos5420-arndale-octa
   Image Type:   ARM Linux Kernel Image (uncompressed)
   Data Size:    3674640 Bytes = 3.5 MiB

Load Address: 41008000
Entry Point: 41008000
Verifying Checksum ... OK
## Loading init Ramdisk from Legacy Image at 22000000 ...

   Image Name:   initramfs
   Image Type:   ARM Linux RAMDisk Image (uncompressed)
   Data Size:    2330416 Bytes = 2.2 MiB

Load Address: 00000000
Entry Point: 00000000
Verifying Checksum ... OK
## Flattened Device Tree blob at 21f00000
Booting using the fdt blob at 0x21f00000 Loading Kernel Image ... OK
OK
Using Device Tree in place at 21f00000, end 21f0b38f

Starting kernel ...

[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 3.18.0-rc6-next-20141128-00007-g33b7bf7da133 (k.kozlowski at AMDC1943) (gcc version 4.7.3 (Ubuntu/Linaro 4.7.3-12ubuntu1) ) #24 SMP PREEMPT Fri Nov 28 16:13:43 CET 2014
[    0.000000] CPU: ARMv7 Processor [412fc0f3] revision 3 (ARMv7), cr=10c5387d

...

[   12.874514] VFS: Mounted root (ext4 filesystem) readonly on device 179:67.
[   12.887392] devtmpfs: mounted
[   12.889321] Freeing unused kernel memory: 324K (c0669000 - c06ba000)
[   12.951241] Unhandled fault: imprecise external abort (0x406) at 0x00000000
[   12.966312] Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000007
[   12.966312]
[   12.973980] CPU: 0 PID: 1 Comm: bash Tainted: G        W      3.18.0-rc6-next-20141128-00007-g33b7bf7da133 #24
[   12.983938] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   12.990014] [<c0014d44>] (unwind_backtrace) from [<c0011c88>] (show_stack+0x10/0x14)
[   12.997723] [<c0011c88>] (show_stack) from [<c048e82c>] (dump_stack+0x70/0xbc)
[   13.004911] [<c048e82c>] (dump_stack) from [<c048a8c0>] (panic+0x94/0x20c)
[   13.011753] [<c048a8c0>] (panic) from [<c0026500>] (do_exit+0x944/0x988)
[   13.018424] [<c0026500>] (do_exit) from [<c00265f4>] (do_group_exit+0x3c/0xbc)
[   13.025621] [<c00265f4>] (do_group_exit) from [<c0030d60>] (get_signal+0x218/0x8c4)
[   13.033247] [<c0030d60>] (get_signal) from [<c048a39c>] (do_signal+0x84/0x350)
[   13.040438] [<c048a39c>] (do_signal) from [<c00115ec>] (do_work_pending+0xbc/0xcc)
[   13.047979] [<c00115ec>] (do_work_pending) from [<c000f1f8>] (work_pending+0xc/0x20)
[   13.055693] CPU1: stopping
[   13.058385] CPU: 1 PID: 0 Comm: swapper/1 Tainted: G        W      3.18.0-rc6-next-20141128-00007-g33b7bf7da133 #24
[   13.068790] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   13.074858] [<c0014d44>] (unwind_backtrace) from [<c0011c88>] (show_stack+0x10/0x14)
[   13.082573] [<c0011c88>] (show_stack) from [<c048e82c>] (dump_stack+0x70/0xbc)
[   13.089762] [<c048e82c>] (dump_stack) from [<c0013e14>] (handle_IPI+0x158/0x18c)
[   13.097128] [<c0013e14>] (handle_IPI) from [<c00086f4>] (gic_handle_irq+0x60/0x68)
[   13.104669] [<c00086f4>] (gic_handle_irq) from [<c00127c4>] (__irq_svc+0x44/0x7c)
[   13.112112] Exception stack(0xee4c3fa0 to 0xee4c3fe8)
[   13.117135] 3fa0: c000fd7c ee4c3fe8 ee4c3ff0 00000000 c06c24e8 c04997e0 c06fa60a c06fa60a
[   13.125287] 3fc0: 00000001 412fc0f3 00000000 00000000 00000001 ee4c3fe8 c000fd7c c000fd80
[   13.133429] 3fe0: 60000013 ffffffff
[   13.136895] [<c00127c4>] (__irq_svc) from [<c000fd80>] (arch_cpu_idle+0x20/0x3c)
[   13.144272] [<c000fd80>] (arch_cpu_idle) from [<c0059d0c>] (cpu_startup_entry+0x11c/0x1a0)
[   13.152499] [<c0059d0c>] (cpu_startup_entry) from [<40008784>] (0x40008784)
[   13.159424] CPU2: stopping
[   13.162102] CPU: 2 PID: 0 Comm: swapper/2 Tainted: G        W      3.18.0-rc6-next-20141128-00007-g33b7bf7da133 #24
[   13.172508] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   13.178577] [<c0014d44>] (unwind_backtrace) from [<c0011c88>] (show_stack+0x10/0x14)
[   13.186291] [<c0011c88>] (show_stack) from [<c048e82c>] (dump_stack+0x70/0xbc)
[   13.193479] [<c048e82c>] (dump_stack) from [<c0013e14>] (handle_IPI+0x158/0x18c)
[   13.200845] [<c0013e14>] (handle_IPI) from [<c00086f4>] (gic_handle_irq+0x60/0x68)
[   13.208386] [<c00086f4>] (gic_handle_irq) from [<c00127c4>] (__irq_svc+0x44/0x7c)
[   13.215833] Exception stack(0xee4c5fa0 to 0xee4c5fe8)
[   13.220854] 5fa0: c000fd7c ee4c5fe8 ee4c5ff0 00000000 c06c24e8 c04997e0 c06fa60a c06fa60a
[   13.229006] 5fc0: 00000001 412fc0f3 00000000 00000000 00000001 ee4c5fe8 c000fd7c c000fd80
[   13.237149] 5fe0: 60000113 ffffffff
[   13.240612] [<c00127c4>] (__irq_svc) from [<c000fd80>] (arch_cpu_idle+0x20/0x3c)
[   13.247986] [<c000fd80>] (arch_cpu_idle) from [<c0059d0c>] (cpu_startup_entry+0x11c/0x1a0)
[   13.256218] [<c0059d0c>] (cpu_startup_entry) from [<40008784>] (0x40008784)
[   13.263144] CPU3: stopping
[   13.265823] CPU: 3 PID: 0 Comm: swapper/3 Tainted: G        W      3.18.0-rc6-next-20141128-00007-g33b7bf7da133 #24
[   13.276230] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   13.282300] [<c0014d44>] (unwind_backtrace) from [<c0011c88>] (show_stack+0x10/0x14)
[   13.290012] [<c0011c88>] (show_stack) from [<c048e82c>] (dump_stack+0x70/0xbc)
[   13.297200] [<c048e82c>] (dump_stack) from [<c0013e14>] (handle_IPI+0x158/0x18c)
[   13.304565] [<c0013e14>] (handle_IPI) from [<c00086f4>] (gic_handle_irq+0x60/0x68)
[   13.312108] [<c00086f4>] (gic_handle_irq) from [<c00127c4>] (__irq_svc+0x44/0x7c)
[   13.319553] Exception stack(0xee4c7fa0 to 0xee4c7fe8)
[   13.324576] 7fa0: c000fd7c ee4c7fe8 ee4c7ff0 00000000 c06c24e8 c04997e0 c06fa60a c06fa60a
[   13.332727] 7fc0: 00000001 412fc0f3 00000000 00000000 00000001 ee4c7fe8 c000fd7c c000fd80
[   13.340870] 7fe0: 60000113 ffffffff
[   13.344335] [<c00127c4>] (__irq_svc) from [<c000fd80>] (arch_cpu_idle+0x20/0x3c)
[   13.351710] [<c000fd80>] (arch_cpu_idle) from [<c0059d0c>] (cpu_startup_entry+0x11c/0x1a0)
[   13.359939] [<c0059d0c>] (cpu_startup_entry) from [<40008784>] (0x40008784)
[   13.366878] ---[ end Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000007
[   13.366878]

==================================

Best regards,
Krzysztof

Quoted 25 lines

arch/arm/boot/dts/exynos5420-arndale-octa.dts +4 -0

index aa7a7d7..db2c1c4 100644
@@ -372,3 +372,7 @@
 &usbdrd_dwc3_1 {
      dr_mode = "host";
 };
+
+&cci {
+     status = "disabled";
+};

arch/arm/boot/dts/exynos5420.dtsi +1 -1

index 517e50f..0a82ae7 100644
@@ -120,7 +120,7 @@
              };
      };
 
-     cci at 10d20000 {
+     cci: cci at 10d20000 {
              compatible = "arm,cci-400";
              #address-cells = <1>;
              #size-cells = <1>;

Hello Krzysztof,

Quoted 12 lines · Krzysztof Kozlowski

Quoted 10 lines · Abhilash Kesavan

The arndale-octa board was giving "imprecise external aborts" during boot-up with MCPM enabled. CCI enablement of the boot cluster was found to be the cause of these aborts (possibly because the secure f/w was not allowing it). Hence, disable CCI for the arndale-octa board.

Signed-off-by Abhilash Kesavan <[email protected]>

---
 arch/arm/boot/dts/exynos5420-arndale-octa.dts |    4 ++++
 arch/arm/boot/dts/exynos5420.dtsi             |    2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)

Quoted 3 lines

I tested these 2 patches on Arndale Octa but there are no improvements. I still got imprecise aborts (some not fatal and sometimes killing init with full backtrace).

Thanks for testing. Are you testing this with exynos_defconfig with no other changes ? Can you please confirm from the bootlog that MCPM and CCI are not being initialized.

Can you remove these 2 patches and on linux-next check if you are getting aborts even with 5420_MCPM disabled.

Quoted 16 lines · Abhilash Kesavan

Hello Krzysztof,

Quoted 12 lines · Krzysztof Kozlowski

Quoted 10 lines · Abhilash Kesavan

The arndale-octa board was giving "imprecise external aborts" during boot-up with MCPM enabled. CCI enablement of the boot cluster was found to be the cause of these aborts (possibly because the secure f/w was not allowing it). Hence, disable CCI for the arndale-octa board.

Signed-off-by Abhilash Kesavan <[email protected]>

---
 arch/arm/boot/dts/exynos5420-arndale-octa.dts |    4 ++++
 arch/arm/boot/dts/exynos5420.dtsi             |    2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)

Quoted 8 lines

Quoted 3 lines

I tested these 2 patches on Arndale Octa but there are no improvements. I still got imprecise aborts (some not fatal and sometimes killing init with full backtrace).

Thanks for testing. Are you testing this with exynos_defconfig with no other changes ? Can you please confirm from the bootlog that MCPM and CCI are not being initialized.

That was exynos_defconfig with disabled DRM and enabled some debug, next-20141128.

When I tried only exynos_defconfig (with disabled DRM) it worked fine... So the imprecise aborts were caused by one of following debug options:

DEBUG_SECTION_MISMATCH
DYNAMIC_DEBUG
DEBUG_ATOMIC_SLEEP
DEBUG_PREEMPT
PROVE_LOCKING
LOCKUP_DETECTOR
DEBUG_LOCK_ALLOC
PROVE_RCU
DEBUG_RT_MUTEXES
DEBUG_MUTEXES
DEBUG_SPINLOCK
DEBUG_LIST
DEBUG_PAGEALLOC
SPARSE_RCU_POINTER
DEBUG_FS
PM_DEBUG
PM_ADVANCED_DEBUG
GPIO_SYSFS

Quoted 2 lines

Can you remove these 2 patches and on linux-next check if you are getting aborts even with 5420_MCPM disabled.

I tried this already and imprecise aborts shown, however with my debugging options above.

Overall the patches seems to work properly (although the debugging issue needs to be resolved still), so:

On Arndale Octa (Exynos 5420):

Tested-by Krzysztof Kozlowski <[email protected]>

Quoted 53 lines · Krzysztof Kozlowski

Quoted 16 lines · Abhilash Kesavan

Hello Krzysztof,

Quoted 12 lines · Krzysztof Kozlowski

Quoted 10 lines · Abhilash Kesavan

The arndale-octa board was giving "imprecise external aborts" during boot-up with MCPM enabled. CCI enablement of the boot cluster was found to be the cause of these aborts (possibly because the secure f/w was not allowing it). Hence, disable CCI for the arndale-octa board.

Signed-off-by Abhilash Kesavan <[email protected]>

---
 arch/arm/boot/dts/exynos5420-arndale-octa.dts |    4 ++++
 arch/arm/boot/dts/exynos5420.dtsi             |    2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)

Quoted 8 lines

Quoted 3 lines

I tested these 2 patches on Arndale Octa but there are no improvements. I still got imprecise aborts (some not fatal and sometimes killing init with full backtrace).

Thanks for testing. Are you testing this with exynos_defconfig with no other changes ? Can you please confirm from the bootlog that MCPM and CCI are not being initialized.

That was exynos_defconfig with disabled DRM and enabled some debug, next-20141128.

When I tried only exynos_defconfig (with disabled DRM) it worked fine... So the imprecise aborts were caused by one of following debug options:

DEBUG_SECTION_MISMATCH
DYNAMIC_DEBUG
DEBUG_ATOMIC_SLEEP
DEBUG_PREEMPT
PROVE_LOCKING
LOCKUP_DETECTOR
DEBUG_LOCK_ALLOC
PROVE_RCU
DEBUG_RT_MUTEXES
DEBUG_MUTEXES
DEBUG_SPINLOCK
DEBUG_LIST
DEBUG_PAGEALLOC
SPARSE_RCU_POINTER
DEBUG_FS
PM_DEBUG
PM_ADVANCED_DEBUG
GPIO_SYSFS

Quoted 2 lines

Can you remove these 2 patches and on linux-next check if you are getting aborts even with 5420_MCPM disabled.

Quoted 8 lines

I tried this already and imprecise aborts shown, however with my debugging options above.

Overall the patches seems to work properly (although the debugging issue needs to be resolved still), so:

On Arndale Octa (Exynos 5420):

Tested-by Krzysztof Kozlowski <[email protected]>

Reading this message, it seems that this should *not* be given a tested-by, because it seems from what you've reported above, they don't work correctly.

If you have to turn debugging options off in order to get the kernel to apparently run correctly after applying some patches, it means those patches themselves are probably buggy, rather than the debug itself being buggy.

I'd suggest that you have some further work to do (a manual bisect of the config options you've disabled) to discover which is the cause of the problem.

It could be that the code introduces something like a use-after-free bug.

The arndale-octa board was giving "imprecise external aborts" during boot-up with MCPM enabled. CCI enablement of the boot cluster was found to be the cause of these aborts (possibly because the secure f/w was not allowing it). Hence, disable CCI for the arndale-octa board.

Signed-off-by Abhilash Kesavan <[email protected]>

Quoted 6 lines · Abhilash Kesavan

The arndale-octa board was giving "imprecise external aborts" during boot-up with MCPM enabled. CCI enablement of the boot cluster was found to be the cause of these aborts (possibly because the secure f/w was not allowing it). Hence, disable CCI for the arndale-octa board.

Signed-off-by Abhilash Kesavan <[email protected]>

Tested-by Kevin Hilman <[email protected]>

Tested on top of next-20141128 with exynos_defconfig on my Octa board and I'm not seeing the imprecise aborts anymore.

Thanks,

Kevin
--
To unsubscribe from this list: send the line "unsubscribe linux-samsung-soc" in the body of a message to majordomo at vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html

The arm-cci driver completes the probe sequence even if the cci node is marked as disabled. Add a check in the driver to honour the cci status in the device tree.

Signed-off-by Abhilash Kesavan <[email protected]>

---
 drivers/bus/arm-cci.c |    3 +++
 1 file changed, 3 insertions(+)

drivers/bus/arm-cci.c +3 -1

index 860da40..0ce5e2d 100644
@@ -1312,6 +1312,9 @@ static int cci_probe(void)
        if (!np)
                return -ENODEV;
 
+       if (!of_device_is_available(np))
+               return -ENODEV;
+
        cci_config = of_match_node(arm_cci_matches, np)->data;
        if (!cci_config)
                return -ENODEV;
--

1.7.9.5

Hi,

Quoted 5 lines · Abhilash Kesavan

The arm-cci driver completes the probe sequence even if the cci node is marked as disabled. Add a check in the driver to honour the cci status in the device tree.

Signed-off-by Abhilash Kesavan <[email protected]>

This patch helps disable CCI on the Arndale Octa board thus resolving some imprecise aborts seen on that board. Kindly review.

Regards,
Abhilash

Quoted 12 lines

---
  drivers/bus/arm-cci.c |    3 +++
  1 file changed, 3 insertions(+)

drivers/bus/arm-cci.c +0 -0

index 860da40..0ce5e2d 100644
@@ -1312,6 +1312,9 @@ static int cci_probe(void)
         if (!np)
                 return -ENODEV;
 

drivers/bus/arm-cci.c static int cci_probe(void) · line 1317

Quoted 3 lines

drivers/bus/arm-cci.c +3 -0

+       if (!of_device_is_available(np))
+               return -ENODEV;
+

Hi Abhilash,

Quoted 31 lines · Abhilash Kesavan

Hi,

Quoted 6 lines · Abhilash Kesavan

The arm-cci driver completes the probe sequence even if the cci node is marked as disabled. Add a check in the driver to honour the cci status in the device tree.

Signed-off-by Abhilash Kesavan <[email protected]>

This patch helps disable CCI on the Arndale Octa board thus resolving some imprecise aborts seen on that board. Kindly review.

Regards,
Abhilash

Quoted 13 lines

---
  drivers/bus/arm-cci.c |    3 +++
  1 file changed, 3 insertions(+)

drivers/bus/arm-cci.c +0 -0

index 860da40..0ce5e2d 100644
@@ -1312,6 +1312,9 @@ static int cci_probe(void)
         if (!np)
                 return -ENODEV;
 

Quoted 3 lines

drivers/bus/arm-cci.c static int cci_probe(void) · line 1317

Quoted 3 lines

drivers/bus/arm-cci.c +3 -0

+       if (!of_device_is_available(np))
+               return -ENODEV;
+

IIUC, by this change you are disabling the MCPM boot protocol here. Is there any alternative boot protocol that works on this platform to boot all 8 cores ? Sorry by quick grep couldn't find one, hence so I am asking.

Regards,
Sudeep

Hi Sudeep,

Quoted 34 lines · Sudeep Holla

Hi Abhilash,

Quoted 31 lines · Abhilash Kesavan

Hi,

Quoted 6 lines · Abhilash Kesavan

The arm-cci driver completes the probe sequence even if the cci node is marked as disabled. Add a check in the driver to honour the cci status in the device tree.

Signed-off-by Abhilash Kesavan <[email protected]>

This patch helps disable CCI on the Arndale Octa board thus resolving some imprecise aborts seen on that board. Kindly review.

Regards,
Abhilash

Quoted 13 lines

---
   drivers/bus/arm-cci.c |    3 +++
   1 file changed, 3 insertions(+)

drivers/bus/arm-cci.c +0 -0

index 860da40..0ce5e2d 100644
@@ -1312,6 +1312,9 @@ static int cci_probe(void)
          if (!np)
                  return -ENODEV;
 

Quoted 9 lines

Quoted 3 lines

drivers/bus/arm-cci.c static int cci_probe(void) · line 1317

Quoted 3 lines

drivers/bus/arm-cci.c +3 -0

+       if (!of_device_is_available(np))
+               return -ENODEV;
+

IIUC, by this change you are disabling the MCPM boot protocol here. Is there any alternative boot protocol that works on this platform to boot all 8 cores ? Sorry by quick grep couldn't find one, hence so I am asking.

Thanks for the reply.
On disabling MCPM, we will default to platsmp.c/firmware.c which boots 4 cores as per Kevin's comment here[1]. This was the original behavior before MCPM was enabled for all 5420 based SoCs.

Quoted 47 lines · Abhilash Kesavan

Hi Sudeep,

Quoted 34 lines · Sudeep Holla

Hi Abhilash,

Quoted 31 lines · Abhilash Kesavan

Hi,

Quoted 6 lines · Abhilash Kesavan

The arm-cci driver completes the probe sequence even if the cci node is marked as disabled. Add a check in the driver to honour the cci status in the device tree.

Signed-off-by Abhilash Kesavan <[email protected]>

This patch helps disable CCI on the Arndale Octa board thus resolving some imprecise aborts seen on that board. Kindly review.

Regards,
Abhilash

Quoted 13 lines

---
    drivers/bus/arm-cci.c |    3 +++
    1 file changed, 3 insertions(+)

drivers/bus/arm-cci.c +0 -0

index 860da40..0ce5e2d 100644
@@ -1312,6 +1312,9 @@ static int cci_probe(void)
           if (!np)
                   return -ENODEV;
 

Quoted 9 lines

Quoted 3 lines

drivers/bus/arm-cci.c static int cci_probe(void) · line 1317

Quoted 3 lines

drivers/bus/arm-cci.c +3 -0

+       if (!of_device_is_available(np))
+               return -ENODEV;
+

IIUC, by this change you are disabling the MCPM boot protocol here. Is there any alternative boot protocol that works on this platform to boot all 8 cores ? Sorry by quick grep couldn't find one, hence so I am asking.

Quoted 5 lines

Thanks for the reply.
On disabling MCPM, we will default to platsmp.c/firmware.c which boots 4 cores as per Kevin's comment here[1]. This was the original behavior before MCPM was enabled for all 5420 based SoCs.

Thanks for pointing that out. I assume the firmware can handle the alternate boot protocol and no more workarounds are needed especially when getting CPUIdle working in this mode.

Anyways the patch makes sense irrespective how it works on exynos, so you can add,

Acked-by Sudeep Holla <[email protected]>

Hi Abhilash,

Quoted 68 lines · Sudeep Holla

Quoted 58 lines · Abhilash Kesavan

Hi Sudeep,

Quoted 41 lines · Sudeep Holla

Hi Abhilash,

Quoted 37 lines · Abhilash Kesavan

Hi,

Quoted 8 lines · Abhilash Kesavan

The arm-cci driver completes the probe sequence even if the cci node is
marked as disabled. Add a check in the driver to honour the cci status in the device tree.

Signed-off-by Abhilash Kesavan <[email protected]>

This patch helps disable CCI on the Arndale Octa board thus resolving some imprecise aborts seen on that board. Kindly review.

Regards,
Abhilash

Quoted 14 lines

---
    drivers/bus/arm-cci.c |    3 +++
    1 file changed, 3 insertions(+)

drivers/bus/arm-cci.c +0 -0

index 860da40..0ce5e2d 100644
@@ -1312,6 +1312,9 @@ static int cci_probe(void)
           if (!np)
                   return -ENODEV;
 

Quoted 10 lines

Quoted 3 lines

drivers/bus/arm-cci.c static int cci_probe(void) · line 1317

Quoted 3 lines

drivers/bus/arm-cci.c +3 -0

+       if (!of_device_is_available(np))
+               return -ENODEV;
+

IIUC, by this change you are disabling the MCPM boot protocol here. Is there any alternative boot protocol that works on this platform to boot all 8 cores ? Sorry by quick grep couldn't find one, hence so I am asking.

Quoted 5 lines

Thanks for the reply.
On disabling MCPM, we will default to platsmp.c/firmware.c which boots 4 cores as per Kevin's comment here[1]. This was the original behavior before MCPM was enabled for all 5420 based SoCs.

Quoted 9 lines

Thanks for pointing that out. I assume the firmware can handle the alternate boot protocol and no more workarounds are needed especially when getting CPUIdle working in this mode.

Anyways the patch makes sense irrespective how it works on exynos, so you can add,

Acked-by Sudeep Holla <[email protected]>

What's the status of this patch. It was useful for me on vexpress for some testing. Please feel free to add

Tested-by Sudeep Holla <[email protected]>

if this is not yet queued.

Regards,
Sudeep

Hi Sudeep,

Quoted 1 line · Sudeep Holla

Hi Abhilash,

Quoted 7 lines

What's the status of this patch. It was useful for me on vexpress for some testing. Please feel free to add

Tested-by Sudeep Holla <[email protected]>

if this is not yet queued.

Thanks for the tested-by. This patch has not been merged yet; I am not quite sure who is supposed to pick this up.

Quoted 7 lines · Abhilash Kesavan

Hi Sudeep,

Quoted 2 lines · Sudeep Holla

Hi Abhilash,

[...]

Quoted 7 lines

Quoted 7 lines

What's the status of this patch. It was useful for me on vexpress for some
testing. Please feel free to add

Tested-by Sudeep Holla <[email protected]>

Quoted 6 lines

Quoted 1 line

if this is not yet queued.

Thanks for the tested-by. This patch has not been merged yet; I am not quite sure who is supposed to pick this up.

So far, most of the CCI patches are merged through arm-soc.

Regards,
Sudeep

Hi Arnd/Olof,

Quoted 24 lines · Sudeep Holla

Quoted 7 lines · Abhilash Kesavan

Hi Sudeep,

Quoted 2 lines · Sudeep Holla

Hi Abhilash,

[...]

Quoted 10 lines

Quoted 8 lines

What's the status of this patch. It was useful for me on vexpress for some
testing. Please feel free to add

Tested-by Sudeep Holla <[email protected]>

if this is not yet queued.

Quoted 5 lines

Quoted 3 lines

Thanks for the tested-by. This patch has not been merged yet; I am not quite sure who is supposed to pick this up.

So far, most of the CCI patches are merged through arm-soc.

Would you be OK picking this up as is or do you want me to re-send this with the RFT tag dropped ?

Quoted 27 lines · Abhilash Kesavan

Hi Arnd/Olof,

Quoted 24 lines · Sudeep Holla

Quoted 7 lines · Abhilash Kesavan

Hi Sudeep,

Quoted 2 lines · Sudeep Holla

Hi Abhilash,

[...]

Quoted 10 lines

Quoted 8 lines

What's the status of this patch. It was useful for me on vexpress for some
testing. Please feel free to add

Tested-by Sudeep Holla <[email protected]>

if this is not yet queued.

Quoted 8 lines

Quoted 5 lines

Quoted 3 lines

Thanks for the tested-by. This patch has not been merged yet; I am not quite sure who is supposed to pick this up.

So far, most of the CCI patches are merged through arm-soc.

Would you be OK picking this up as is or do you want me to re-send this with the RFT tag dropped ?

Please resend without the RFT, and collect the Tested-by tags you can add mine:

Tested-by Kevin Hilman <[email protected]>

Please send to arm at kernel.org where patches targeted for the arm-soc tree are collected.

Thanks,

Kevin

Hi,

Quoted 5 lines · Abhilash Kesavan

The arm-cci driver completes the probe sequence even if the cci node is marked as disabled. Add a check in the driver to honour the cci status in the device tree.

Signed-off-by Abhilash Kesavan <[email protected]>

This patch helps disable CCI on the Arndale Octa board thus resolving some imprecise aborts seen on that board. Kindly review.

Regards,
Abhilash

Quoted 21 lines

---
 drivers/bus/arm-cci.c |    3 +++
 1 file changed, 3 insertions(+)

drivers/bus/arm-cci.c +3 -1

index 860da40..0ce5e2d 100644
@@ -1312,6 +1312,9 @@ static int cci_probe(void)
        if (!np)
                return -ENODEV;
 
+       if (!of_device_is_available(np))
+               return -ENODEV;
+
        cci_config = of_match_node(arm_cci_matches, np)->data;
        if (!cci_config)
                return -ENODEV;
--

1.7.9.5