diff mbox

arm/mach-exynos: suspend: Don't try to initialize suspend/suspend on old DT

Message ID 1431438307-25410-1-git-send-email-julien.grall@citrix.com (mailing list archive)
State New, archived
Headers show

Commit Message

Julien Grall May 12, 2015, 1:45 p.m. UTC
Since commit 8b283c0 "ARM: exynos4/5: convert pmu wakeup to stacked domains",
a suspend/resume is not supported on old DT.

Although, rather than printing a warning and continue to boot, the
kernel will segfault just after:

------------[ cut here ]------------

WARNING: CPU: 1 PID: 1 at arch/arm/mach-exynos/suspend.c:726 exynos_pm_init+0x4c/0xc8()
Modules linked in:
CPU: 1 PID: 1 Comm: swapper/0 Not tainted 4.1.0-rc3 #1
Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[<c02181c4>] (unwind_backtrace) from [<c0213b2c>] (show_stack+0x10/0x14)
[<c0213b2c>] (show_stack) from [<c0949890>] (dump_stack+0x70/0x8c)
[<c0949890>] (dump_stack) from [<c024f0b0>] (warn_slowpath_common+0x74/0xac)
[<c024f0b0>] (warn_slowpath_common) from [<c024f104>] (warn_slowpath_null+0x1c/0x24)
[<c024f104>] (warn_slowpath_null) from [<c0cf1d28>] (exynos_pm_init+0x4c/0xc8)
[<c0cf1d28>] (exynos_pm_init) from [<c0ceaae8>] (init_machine_late+0x1c/0x28)
[<c0ceaae8>] (init_machine_late) from [<c020aa64>] (do_one_initcall+0x80/0x1d0)
[<c020aa64>] (do_one_initcall) from [<c0ce8d4c>] (kernel_init_freeable+0x10c/0x1d8)
[<c0ce8d4c>] (kernel_init_freeable) from [<c0944a2c>] (kernel_init+0x8/0xe4)
[<c0944a2c>] (kernel_init) from [<c0210e60>] (ret_from_fork+0x14/0x34)
---[ end trace 335bd937d409f3c7 ]---
Outdated DT detected, suspend/resume will NOT work
Unable to handle kernel NULL pointer dereference at virtual address 00000608
pgd = c0204000
[00000608] *pgd=00000000
Internal error: Oops: 5 [#1] SMP ARM
Modules linked in:
CPU: 1 PID: 1 Comm: swapper/0 Tainted: G        W       4.1.0-rc3 #1
Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
task: db06c000 ti: db05a000 task.ti: db05a000
PC is at exynos_pm_init+0x6c/0xc8
LR is at exynos_pm_init+0x54/0xc8
pc : [<c0cf1d48>]    lr : [<c0cf1d30>]    psr: 60000113
sp : db05bee8  ip : 00000000  fp : 00000000
r10: 00000116  r9 : c0dab2d4  r8 : d8d5f440
r7 : c0db7ad8  r6 : c0db7ad8  r5 : 00000000  r4 : c0ceaacc
r3 : c0eb2aec  r2 : c0951e40  r1 : 00000000  r0 : c0eb2acc
Flags: nZCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment kernel
Control: 10c5387d  Table: 6020406a  DAC: 00000015
Process swapper/0 (pid: 1, stack limit = 0xdb05a220)
Stack: (0xdb05bee8 to 0xdb05c000)
bee0:                   c0db7ad8 c0d8fe34 c0cf17c8 c0ceaae8 00000000 c020aa64
bf00: 00000033 c09580b8 db04fd00 c0ed79a4 c0eb1000 c0ce8588 c0ca2bc4 c0353fcc
bf20: 00000000 c0df358c 60000113 00000000 dbfffba4 00000000 c0ca2bc4 c026654c
bf40: c0b80134 c0ca1a64 00000007 00000007 c0df3554 c0d6c2f4 00000007 c0d6c2d4
bf60: c0eb1000 c0ce8588 c0dab2d4 00000116 00000000 c0ce8d4c 00000007 00000007
bf80: c0ce8588 c0944a24 00000000 c0944a24 00000000 00000000 00000000 00000000
bfa0: 00000000 c0944a2c 00000000 c0210e60 00000000 00000000 00000000 00000000
bfc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
bfe0: 00000000 00000000 00000000 00000000 00000013 00000000 00000000 00000000
[<c0cf1d48>] (exynos_pm_init) from [<c0ceaae8>] (init_machine_late+0x1c/0x28)
[<c0ceaae8>] (init_machine_late) from [<c020aa64>] (do_one_initcall+0x80/0x1d0)
[<c020aa64>] (do_one_initcall) from [<c0ce8d4c>] (kernel_init_freeable+0x10c/0x1d8)
[<c0ce8d4c>] (kernel_init_freeable) from [<c0944a2c>] (kernel_init+0x8/0xe4)
[<c0944a2c>] (kernel_init) from [<c0210e60>] (ret_from_fork+0x14/0x34)
Code: e59f005c e59220c0 e5901000 e5832000 (e591e608)
---[ end trace 335bd937d409f3c8 ]---

This is happening because pmu_base_addr is only initialized when the
PMU is an interrupt controller. It's not the case on old DT.

Signed-off-by: Julien Grall <julien.grall@citrix.com>
---
 arch/arm/mach-exynos/suspend.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

Comments

kgene@kernel.org May 12, 2015, 5:18 p.m. UTC | #1
On 05/12/15 22:45, Julien Grall wrote:
> Since commit 8b283c0 "ARM: exynos4/5: convert pmu wakeup to stacked domains",
> a suspend/resume is not supported on old DT.
> 
> Although, rather than printing a warning and continue to boot, the
> kernel will segfault just after:
> 
> ------------[ cut here ]------------
> 
> WARNING: CPU: 1 PID: 1 at arch/arm/mach-exynos/suspend.c:726 exynos_pm_init+0x4c/0xc8()
> Modules linked in:
> CPU: 1 PID: 1 Comm: swapper/0 Not tainted 4.1.0-rc3 #1
> Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
> [<c02181c4>] (unwind_backtrace) from [<c0213b2c>] (show_stack+0x10/0x14)
> [<c0213b2c>] (show_stack) from [<c0949890>] (dump_stack+0x70/0x8c)
> [<c0949890>] (dump_stack) from [<c024f0b0>] (warn_slowpath_common+0x74/0xac)
> [<c024f0b0>] (warn_slowpath_common) from [<c024f104>] (warn_slowpath_null+0x1c/0x24)
> [<c024f104>] (warn_slowpath_null) from [<c0cf1d28>] (exynos_pm_init+0x4c/0xc8)
> [<c0cf1d28>] (exynos_pm_init) from [<c0ceaae8>] (init_machine_late+0x1c/0x28)
> [<c0ceaae8>] (init_machine_late) from [<c020aa64>] (do_one_initcall+0x80/0x1d0)
> [<c020aa64>] (do_one_initcall) from [<c0ce8d4c>] (kernel_init_freeable+0x10c/0x1d8)
> [<c0ce8d4c>] (kernel_init_freeable) from [<c0944a2c>] (kernel_init+0x8/0xe4)
> [<c0944a2c>] (kernel_init) from [<c0210e60>] (ret_from_fork+0x14/0x34)
> ---[ end trace 335bd937d409f3c7 ]---
> Outdated DT detected, suspend/resume will NOT work
> Unable to handle kernel NULL pointer dereference at virtual address 00000608
> pgd = c0204000
> [00000608] *pgd=00000000
> Internal error: Oops: 5 [#1] SMP ARM
> Modules linked in:
> CPU: 1 PID: 1 Comm: swapper/0 Tainted: G        W       4.1.0-rc3 #1
> Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
> task: db06c000 ti: db05a000 task.ti: db05a000
> PC is at exynos_pm_init+0x6c/0xc8
> LR is at exynos_pm_init+0x54/0xc8
> pc : [<c0cf1d48>]    lr : [<c0cf1d30>]    psr: 60000113
> sp : db05bee8  ip : 00000000  fp : 00000000
> r10: 00000116  r9 : c0dab2d4  r8 : d8d5f440
> r7 : c0db7ad8  r6 : c0db7ad8  r5 : 00000000  r4 : c0ceaacc
> r3 : c0eb2aec  r2 : c0951e40  r1 : 00000000  r0 : c0eb2acc
> Flags: nZCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment kernel
> Control: 10c5387d  Table: 6020406a  DAC: 00000015
> Process swapper/0 (pid: 1, stack limit = 0xdb05a220)
> Stack: (0xdb05bee8 to 0xdb05c000)
> bee0:                   c0db7ad8 c0d8fe34 c0cf17c8 c0ceaae8 00000000 c020aa64
> bf00: 00000033 c09580b8 db04fd00 c0ed79a4 c0eb1000 c0ce8588 c0ca2bc4 c0353fcc
> bf20: 00000000 c0df358c 60000113 00000000 dbfffba4 00000000 c0ca2bc4 c026654c
> bf40: c0b80134 c0ca1a64 00000007 00000007 c0df3554 c0d6c2f4 00000007 c0d6c2d4
> bf60: c0eb1000 c0ce8588 c0dab2d4 00000116 00000000 c0ce8d4c 00000007 00000007
> bf80: c0ce8588 c0944a24 00000000 c0944a24 00000000 00000000 00000000 00000000
> bfa0: 00000000 c0944a2c 00000000 c0210e60 00000000 00000000 00000000 00000000
> bfc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
> bfe0: 00000000 00000000 00000000 00000000 00000013 00000000 00000000 00000000
> [<c0cf1d48>] (exynos_pm_init) from [<c0ceaae8>] (init_machine_late+0x1c/0x28)
> [<c0ceaae8>] (init_machine_late) from [<c020aa64>] (do_one_initcall+0x80/0x1d0)
> [<c020aa64>] (do_one_initcall) from [<c0ce8d4c>] (kernel_init_freeable+0x10c/0x1d8)
> [<c0ce8d4c>] (kernel_init_freeable) from [<c0944a2c>] (kernel_init+0x8/0xe4)
> [<c0944a2c>] (kernel_init) from [<c0210e60>] (ret_from_fork+0x14/0x34)
> Code: e59f005c e59220c0 e5901000 e5832000 (e591e608)
> ---[ end trace 335bd937d409f3c8 ]---
> 
> This is happening because pmu_base_addr is only initialized when the
> PMU is an interrupt controller. It's not the case on old DT.
> 
> Signed-off-by: Julien Grall <julien.grall@citrix.com>
> ---
>  arch/arm/mach-exynos/suspend.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/arm/mach-exynos/suspend.c b/arch/arm/mach-exynos/suspend.c
> index 3e6aea7..b6f3ddc 100644
> --- a/arch/arm/mach-exynos/suspend.c
> +++ b/arch/arm/mach-exynos/suspend.c
> @@ -723,8 +723,10 @@ void __init exynos_pm_init(void)
>  		return;
>  	}
>  
> -	if (WARN_ON(!of_find_property(np, "interrupt-controller", NULL)))
> +	if (WARN_ON(!of_find_property(np, "interrupt-controller", NULL))) {
>  		pr_warn("Outdated DT detected, suspend/resume will NOT work\n");
> +		return;
> +	}
>  
>  	pm_data = (const struct exynos_pm_data *) match->data;
>  
I'm fine on this, applied.

Thanks,
Kukjin
--
To unsubscribe from this list: send the line "unsubscribe linux-samsung-soc" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Ian Campbell May 14, 2015, 8:50 a.m. UTC | #2
On Wed, 2015-05-13 at 02:18 +0900, Kukjin Kim wrote:
> On 05/12/15 22:45, Julien Grall wrote:
> > Since commit 8b283c0 "ARM: exynos4/5: convert pmu wakeup to stacked domains",
> > a suspend/resume is not supported on old DT.
[...]
> I'm fine on this, applied.

Thanks. Do you also plan to send it to Linus (or to whomeever you feed
things through) for 4.1-rc$NEXT?

Cheers,
Ian.

--
To unsubscribe from this list: send the line "unsubscribe linux-samsung-soc" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
kgene@kernel.org May 16, 2015, 3:45 a.m. UTC | #3
On 05/14/15 17:50, Ian Campbell wrote:
> On Wed, 2015-05-13 at 02:18 +0900, Kukjin Kim wrote:
>> On 05/12/15 22:45, Julien Grall wrote:
>>> Since commit 8b283c0 "ARM: exynos4/5: convert pmu wakeup to stacked domains",
>>> a suspend/resume is not supported on old DT.
> [...]
>> I'm fine on this, applied.
> 
> Thanks. Do you also plan to send it to Linus (or to whomeever you feed
> things through) for 4.1-rc$NEXT?
> 
Already it has been sent to Linus via arm-soc. I think, it should be
fine after v4.1-rc4?

- Kukjin
--
To unsubscribe from this list: send the line "unsubscribe linux-samsung-soc" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Ian Campbell May 18, 2015, 9:33 a.m. UTC | #4
On Sat, 2015-05-16 at 12:45 +0900, Kukjin Kim wrote:
> On 05/14/15 17:50, Ian Campbell wrote:
> > On Wed, 2015-05-13 at 02:18 +0900, Kukjin Kim wrote:
> >> On 05/12/15 22:45, Julien Grall wrote:
> >>> Since commit 8b283c0 "ARM: exynos4/5: convert pmu wakeup to stacked domains",
> >>> a suspend/resume is not supported on old DT.
> > [...]
> >> I'm fine on this, applied.
> > 
> > Thanks. Do you also plan to send it to Linus (or to whomeever you feed
> > things through) for 4.1-rc$NEXT?
> > 
> Already it has been sent to Linus via arm-soc. I think, it should be
> fine after v4.1-rc4?

Thanks, I can see that our latest linux-latest test flight (still in
progress) has indeed picked this up and has avoided this particular
issue.

Cheers,
Ian.

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

Patch

diff --git a/arch/arm/mach-exynos/suspend.c b/arch/arm/mach-exynos/suspend.c
index 3e6aea7..b6f3ddc 100644
--- a/arch/arm/mach-exynos/suspend.c
+++ b/arch/arm/mach-exynos/suspend.c
@@ -723,8 +723,10 @@  void __init exynos_pm_init(void)
 		return;
 	}
 
-	if (WARN_ON(!of_find_property(np, "interrupt-controller", NULL)))
+	if (WARN_ON(!of_find_property(np, "interrupt-controller", NULL))) {
 		pr_warn("Outdated DT detected, suspend/resume will NOT work\n");
+		return;
+	}
 
 	pm_data = (const struct exynos_pm_data *) match->data;