ARM: Kconfig: Do not select ARCH_VEXPRESS on ARCH_MULTI_V7
PATCH 1/1 ARM: Kconfig: Do not select ARCH_VEXPRESS on ARCH_MULTI_V7
Booting a kernel on a mx53qsb board (built with imx_v6_v7_defconfig) leads to:
Unable to handle kernel NULL pointer dereference at virtual address 00000008 pgd = 80004000 [00000008] *pgd=00000000 Internal error: Oops: 805 [#1] SMP ARM Modules linked in: CPU: 0 Not tainted (3.7.0-rc5-next-20121115+ #715) PC is at vexpress_sysreg_init_leds+0x30/0xb8 LR is at do_one_initcall+0x3c/0x190 pc : [<8065e170>] lr : [<800086a0>] psr: 60000153
sp : df855ef0 ip : df855f18 fp : df855f14
r10: 00000000 r9 : 8067eb5c r8 : 8065e140
r7 : 80711cac r6 : 806c2cc0 r5 : 00000000 r4 : 80673850 r3 : 00000000 r2 : 00000000 r1 : df854000 r0 : 8065e140
...
Since commit c5a0d4977 (ARM: imx: enable multi-platform build), imx_v6_v7_defconfig selects CONFIG_ARCH_MULTI_V7=y, which in turn select ARCH_VEXPRESS.
Do not select ARCH_VEXPRESS on ARCH_MULTI_V7.
--- arch/arm/Kconfig | 1 - 1 file changed, 1 deletion(-)
arch/arm/Kconfig +0 -1
@@ -1000,7 +1000,6 @@ config ARCH_MULTI_V7 bool "ARMv7 based platforms (Cortex-A, PJ4, Krait)" default y select ARCH_MULTI_V6_V7 - select ARCH_VEXPRESS select CPU_V7 config ARCH_MULTI_V6_V7
Signature · 1 line
1.7.9.5
PATCH 1/1 ARM: Kconfig: Do not select ARCH_VEXPRESS on ARCH_MULTI_V7
Quoted 11 lines · Fabio Estevam
Booting a kernel on a mx53qsb board (built with
imx_v6_v7_defconfig) leads to:Unable to handle kernel NULL pointer dereference at virtual address 00000008 pgd = 80004000 [00000008] *pgd=00000000 Internal error: Oops: 805 [#1] SMP ARM Modules linked in: CPU: 0 Not tainted (3.7.0-rc5-next-20121115+ #715) PC is at vexpress_sysreg_init_leds+0x30/0xb8 LR is at do_one_initcall+0x3c/0x190 pc : [<8065e170>] lr : [<800086a0>] psr: 60000153
Quoted 5 lines
sp :
df855ef0ip :df855f18fp :df855f14
r10: 00000000 r9 :8067eb5cr8 :8065e140
r7 :80711cacr6 :806c2cc0r5 : 00000000 r4 : 80673850 r3 : 00000000 r2 : 00000000 r1 :df854000r0 :8065e140
...
It's a bug in vexpress-sysreg driver. 3 hours ago I've posted patch removing this code...
https://patchwork.kernel.org/patch/1764801/
Apologies about causing the issue in the first place.
Pawel
PATCH 1/1 ARM: Kconfig: Do not select ARCH_VEXPRESS on ARCH_MULTI_V7
Quoted 6 lines · Pawel Moll
It's a bug in vexpress-sysreg driver. 3 hours ago I've posted patch removing this code...
https://patchwork.kernel.org/patch/1764801/
Apologies about causing the issue in the first place.
But still, the patch that Fabio raises is interesting: why would ARCH_VEXPRESS be unconditionally enabled when building an ARM multiplatform kernel for v6/v7 ? I may want to build a kernel with several platforms, but not necessarily vexpress, no?
Thanks!
Thomas
Signature · 4 lines
Thomas Petazzoni, Free Electrons Kernel, drivers, real-time and embedded Linux development, consulting, training and support. http://free-electrons.com
PATCH 1/1 ARM: Kconfig: Do not select ARCH_VEXPRESS on ARCH_MULTI_V7
Quoted 5 lines · Thomas Petazzoni
Quoted 3 lines · Pawel Moll
It's a bug in vexpress-sysreg driver. 3 hours ago I've posted patch removing this code...
Quoted 8 lines
Quoted 3 lines
https://patchwork.kernel.org/patch/1764801/
Apologies about causing the issue in the first place.
But still, the patch that Fabio raises is interesting: why would
ARCH_VEXPRESSbe unconditionally enabled when building an ARM multiplatform kernel for v6/v7 ? I may want to build a kernel with several platforms, but not necessarily vexpress, no?
I'm not arguing this. As far as I remember Rob wanted to make vexpress a default for the multiplatform kernel, but you're right saying that forcing it with the SELECT was probably a step too far.
Pawe?
PATCH 1/1 ARM: Kconfig: Do not select ARCH_VEXPRESS on ARCH_MULTI_V7
Quoted 3 lines · Pawel Moll
I'm not arguing this. As far as I remember Rob wanted to make vexpress a default for the multiplatform kernel, but you're right saying that forcing it with the SELECT was probably a step too far.
IIRC, the problem is that we need at least one platform to be enabled, otherwise it doesn't build, but I might be wrong.
Best regards,
Thomas
Signature · 4 lines
Thomas Petazzoni, Free Electrons Kernel, drivers, real-time and embedded Linux development, consulting, training and support. http://free-electrons.com