* [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report
@ 2015-12-28 18:00 Grygorii Strashko
2015-12-28 18:00 ` Grygorii Strashko
` (2 more replies)
0 siblings, 3 replies; 16+ messages in thread
From: Grygorii Strashko @ 2015-12-28 18:00 UTC (permalink / raw)
To: bigeasy, linux-rt-users; +Cc: linux-kernel, Sekhar Nori, Grygorii Strashko
Hi All, Sebastian,
! Merry Christmas and Happy New year !
Sebastian, Thanks alot for this Christmas present.
I was able to test -RT 4.4-rc6-rt1 Kernel on ARM TI am437x-gpevm (UP)
1) UP build is broken, so there are two fixes.
2) I can boot using very minimal rootfs
3) Seems there are some problem with HIGHMEM :(. System crashes if I'm trying
to run hackbench (see log 1). I can run hackbench only if HIGHMEM is disabled.
I've also tried to disable CONFIG_CPU_SW_DOMAIN_PAN, CONFIG_ARM_PATCH_PHYS_VIRT,
and CONFIG_CPUSETS, before disabling HIGHMEM.
Sorry, I'll not able to investigate it further because of Christmas holidays here.
4) Could we revert commit 4fa5422 'ARM: OMAP2: Drop the concept of certain power
domains not being able to lose context.'? It's not required for 4.4-rt wakeup
and introduces build warnings.
Test1: cyclictest + hackbench
# ./cyclictest -D120s -m -n -Sp98 -q
T: 0 (14999) P:98 I:1000 C: 120000 Min: 3 Act: 4 Avg: 5 Max: 21
Test2: cyclictest + hackbench
T: 0 ( 151) P:98 I:1000 C: 120000 Min: 4 Act: 31 Avg: 32 Max: 61
----------- log 1 ----
/ # ./hackbench
Running in process mode with 10 groups using 40 file descriptors each (== 400 tasks)
Each sender will pass 100 messages of 100
[ 205.565705] Unable to handle kernel paging request at virtual address ffefe378
bytes
Unable to handle kernel paging request at virtual address ffefe378
pgd = ee8dc000
[ffefe378] *pgd=afffd861, *pte=00000000, *ppte=00000000
Internal error: Oops: 837 [#1] PREEMPT ARM
Modules linked in:
CPU: 0 PID: 111 Comm: hackbench Not tainted 4.4.0-rc6-rt1-00069-g17f26b6-dirty #15
Segmentation fault[ 205.574708] Hardware name: Generic AM43 (Flattened Device Tree)
task: ee8e8580 ti: ee92e000 task.ti: ee92e000
PC is at cpu_ca8_set_pte_ext+0x0/0x58
LR is at wp_page_copy+0x1ec/0x3c0
pc : [<c001c8e4>] lr : [<c00da3a0>] psr: 00000013
sp : ee92fd68 ip : ee92fd68 fp : ee92fdb4
r10: ffefe378 r9 : ee8a8540 r8 : b6ede000
r7 : 8942734f r6 : ee8af000 r5 : ef84657c r4 : ef9369c0
r3 : c001c8e4 r2 : 00000800 r1 : 8942734f r0 : ffefe378
Flags: nzcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user
Control: 10c53c7d Table: ae8dc059 DAC: 00000055
Process hackbench (pid: 111, stack limit = 0xee92e210)
Stack: (0xee92fd68 to 0xee930000)
fd60: ee92fd84 c07bf980 000000de c07bf980 ee8a8574 0008fef0
fd80: c0019c44 ef04e800 00000000 ef9369c0 ee8af000 8fef03cf b6ede000 ffefe378
fda0: ee8a8540 b6ede000 ee92fe1c ee92fdb8 c00dbc10 c00da1c0 ee8dedb8 8fef03cf
fdc0: ef9369c0 ef6f9000 0000004f c0776efc 000000de c07bf980 ee92fdfc ee92fde8
fde0: c00d994c c0572e44 effcd098 ef6f9000 ee92fe1c ee8a8574 ee8af000 ee8a8540
fe00: c0776efc 8fef03cf c07bf980 b6ede000 ee92fea4 ee92fe20 c00dd8d0 c00dbb68
fe20: ee8dedb8 ee8a8574 8fef03cf 00000040 ffefe000 b6e4f000 b6eb6000 ffefe13c
fe40: 000000de ffefe378 ffefe000 00000040 ee8dedb8 00000055 ee92ffb0 00000054
fe60: 0000005e b6e40000 ee8a8588 ee8e8580 b6ede1f8 ee8e8580 ee92fea4 ee92ffb0
fe80: ee8e8580 0000081f ee8a8540 b6ede1f8 ee8af000 ee8a8588 ee92fefc ee92fea8
fea0: c0573c48 c00dce84 00000000 00000054 00000000 c076aa0c 00000005 00000000
fec0: 00000000 00000800 00000000 00000055 ee92fefc c076a8ac 0000081f c05739d8
fee0: b6ede1f8 ee92ffb0 0000006e 0000006e ee92ffac ee92ff00 c00092a8 c05739e4
ff00: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
ff20: 00000000 c02aece0 80000113 ffffffff ee92ff7c 00000000 ee92ffac ee92ff48
ff40: c0573318 c0009274 b6f07068 beffffff 0000006f 00000000 beffffff b6f07068
ff60: 00000000 00000000 00000000 00000000 00000000 ee92ffac f7f0706b ee92ff98
ff80: c0050a80 c02aece0 80000113 b6e4fdce 00000030 ffffffff 10c53c7d 10c53c7d
ffa0: 00000000 ee92ffb0 c0573620 c0009274 00000000 00000004 0000006f b6ede1f8
ffc0: 00000000 be860c50 b6f07000 be860c70 b6f074c0 0000006e 0000006e b6ec0000
ffe0: 00000078 be860c50 b6e4fdaf b6e4fdce 00000030 ffffffff 00000000 00000000
Backtrace:
[<c00da1b4>] (wp_page_copy) from [<c00dbc10>] (do_wp_page+0xb4/0x574)
r10:b6ede000 r9:ee8a8540 r8:ffefe378 r7:b6ede000 r6:8fef03cf r5:ee8af000
r4:ef9369c0
[<c00dbb5c>] (do_wp_page) from [<c00dd8d0>] (handle_mm_fault+0xa58/0x1058)
r10:b6ede000 r9:c07bf980 r8:8fef03cf r7:c0776efc r6:ee8a8540 r5:ee8af000
r4:ee8a8574
[<c00dce78>] (handle_mm_fault) from [<c0573c48>] (do_page_fault+0x270/0x43c)
r10:ee8a8588 r9:ee8af000 r8:b6ede1f8 r7:ee8a8540 r6:0000081f r5:ee8e8580
r4:ee92ffb0
[<c05739d8>] (do_page_fault) from [<c00092a8>] (do_DataAbort+0x40/0xc0)
r10:0000006e r9:0000006e r8:ee92ffb0 r7:b6ede1f8 r6:c05739d8 r5:0000081f
r4:c076a8ac
[<c0009268>] (do_DataAbort) from [<c0573620>] (__dabt_usr+0x40/0x60)
Exception stack(0xee92ffb0 to 0xee92fff8)
ffa0: 00000000 00000004 0000006f b6ede1f8
ffc0: 00000000 be860c50 b6f07000 be860c70 b6f074c0 0000006e 0000006e b6ec0000
ffe0: 00000078 be860c50 b6e4fdaf b6e4fdce 00000030 ffffffff
r8:10c53c7d r7:10c53c7d r6:ffffffff r5:00000030 r4:b6e4fdce
Code: f57ff06f ee020f10 f57ff06f e12fff1e (e5801000)
---[ end trace 0000000000000002 ]---
Grygorii Strashko (2):
[4.4-rt] kernel/sched/core.c: UP: fix 'implicit declaration of
function 'update_migrate_disable'
[4.4-rt1] kernel/time/timer.c: UP: fix undefined reference to
`del_timer_sync'
kernel/sched/core.c | 18 +++++++++---------
kernel/time/timer.c | 2 +-
2 files changed, 10 insertions(+), 10 deletions(-)
--
2.6.4
^ permalink raw reply [flat|nested] 16+ messages in thread* [4.4-rc6-rt1 PATCH 1/2] kernel/sched/core.c: UP: fix 'implicit declaration of function 'update_migrate_disable'
2015-12-28 18:00 [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report Grygorii Strashko
@ 2015-12-28 18:00 ` Grygorii Strashko
2015-12-28 18:00 ` [4.4-rc6-rt1 PATCH 2/2] kernel/time/timer.c: UP: fix undefined reference to `del_timer_sync' Grygorii Strashko
2016-01-12 17:15 ` [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report Sebastian Andrzej Siewior
2 siblings, 0 replies; 16+ messages in thread
From: Grygorii Strashko @ 2015-12-28 18:00 UTC (permalink / raw)
To: bigeasy, linux-rt-users; +Cc: linux-kernel, Sekhar Nori, Grygorii Strashko
This patch fixes below build error for case SMP=n:
kernel/sched/core.c: In function '__schedule':
kernel/sched/core.c:3400:2: error: implicit declaration of function 'update_migrate_disable' [-Werror=implicit-function-declaration]
update_migrate_disable(prev);
^
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
---
kernel/sched/core.c | 18 +++++++++---------
1 file changed, 9 insertions(+), 9 deletions(-)
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 8d9f6a6..c10ac78 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -1092,6 +1092,15 @@ void check_preempt_curr(struct rq *rq, struct task_struct *p, int flags)
rq_clock_skip_update(rq, true);
}
+#if defined(CONFIG_PREEMPT_RT_FULL) && defined(CONFIG_SMP)
+#define MIGRATE_DISABLE_SET_AFFIN (1<<30) /* Can't make a negative */
+#define migrate_disabled_updated(p) ((p)->migrate_disable & MIGRATE_DISABLE_SET_AFFIN)
+#define migrate_disable_count(p) ((p)->migrate_disable & ~MIGRATE_DISABLE_SET_AFFIN)
+#else
+static inline void update_migrate_disable(struct task_struct *p) { }
+#define migrate_disabled_updated(p) 0
+#endif
+
#ifdef CONFIG_SMP
/*
* This is how migration works:
@@ -1209,15 +1218,6 @@ void set_cpus_allowed_common(struct task_struct *p, const struct cpumask *new_ma
p->nr_cpus_allowed = cpumask_weight(new_mask);
}
-#if defined(CONFIG_PREEMPT_RT_FULL) && defined(CONFIG_SMP)
-#define MIGRATE_DISABLE_SET_AFFIN (1<<30) /* Can't make a negative */
-#define migrate_disabled_updated(p) ((p)->migrate_disable & MIGRATE_DISABLE_SET_AFFIN)
-#define migrate_disable_count(p) ((p)->migrate_disable & ~MIGRATE_DISABLE_SET_AFFIN)
-#else
-static inline void update_migrate_disable(struct task_struct *p) { }
-#define migrate_disabled_updated(p) 0
-#endif
-
void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask)
{
struct rq *rq = task_rq(p);
--
2.6.4
^ permalink raw reply related [flat|nested] 16+ messages in thread* [4.4-rc6-rt1 PATCH 1/2] kernel/sched/core.c: UP: fix 'implicit declaration of function 'update_migrate_disable'
@ 2015-12-28 18:00 ` Grygorii Strashko
0 siblings, 0 replies; 16+ messages in thread
From: Grygorii Strashko @ 2015-12-28 18:00 UTC (permalink / raw)
To: bigeasy, linux-rt-users; +Cc: linux-kernel, Sekhar Nori, Grygorii Strashko
This patch fixes below build error for case SMP=n:
kernel/sched/core.c: In function '__schedule':
kernel/sched/core.c:3400:2: error: implicit declaration of function 'update_migrate_disable' [-Werror=implicit-function-declaration]
update_migrate_disable(prev);
^
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
---
kernel/sched/core.c | 18 +++++++++---------
1 file changed, 9 insertions(+), 9 deletions(-)
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 8d9f6a6..c10ac78 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -1092,6 +1092,15 @@ void check_preempt_curr(struct rq *rq, struct task_struct *p, int flags)
rq_clock_skip_update(rq, true);
}
+#if defined(CONFIG_PREEMPT_RT_FULL) && defined(CONFIG_SMP)
+#define MIGRATE_DISABLE_SET_AFFIN (1<<30) /* Can't make a negative */
+#define migrate_disabled_updated(p) ((p)->migrate_disable & MIGRATE_DISABLE_SET_AFFIN)
+#define migrate_disable_count(p) ((p)->migrate_disable & ~MIGRATE_DISABLE_SET_AFFIN)
+#else
+static inline void update_migrate_disable(struct task_struct *p) { }
+#define migrate_disabled_updated(p) 0
+#endif
+
#ifdef CONFIG_SMP
/*
* This is how migration works:
@@ -1209,15 +1218,6 @@ void set_cpus_allowed_common(struct task_struct *p, const struct cpumask *new_ma
p->nr_cpus_allowed = cpumask_weight(new_mask);
}
-#if defined(CONFIG_PREEMPT_RT_FULL) && defined(CONFIG_SMP)
-#define MIGRATE_DISABLE_SET_AFFIN (1<<30) /* Can't make a negative */
-#define migrate_disabled_updated(p) ((p)->migrate_disable & MIGRATE_DISABLE_SET_AFFIN)
-#define migrate_disable_count(p) ((p)->migrate_disable & ~MIGRATE_DISABLE_SET_AFFIN)
-#else
-static inline void update_migrate_disable(struct task_struct *p) { }
-#define migrate_disabled_updated(p) 0
-#endif
^ permalink raw reply related [flat|nested] 16+ messages in thread
* [4.4-rc6-rt1 PATCH 2/2] kernel/time/timer.c: UP: fix undefined reference to `del_timer_sync'
2015-12-28 18:00 [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report Grygorii Strashko
2015-12-28 18:00 ` Grygorii Strashko
@ 2015-12-28 18:00 ` Grygorii Strashko
2016-01-12 16:47 ` Sebastian Andrzej Siewior
2016-01-12 17:15 ` [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report Sebastian Andrzej Siewior
2 siblings, 1 reply; 16+ messages in thread
From: Grygorii Strashko @ 2015-12-28 18:00 UTC (permalink / raw)
To: bigeasy, linux-rt-users; +Cc: linux-kernel, Sekhar Nori, Grygorii Strashko
This patch fixes below build errors in case SMP=n:
drivers/built-in.o: In function `fbcon_del_cursor_timer':
/home/x0174654/kernel.org/linux-master/linux/drivers/video/console/fbcon.c:433: undefined reference to `del_timer_sync'
drivers/built-in.o: In function `sysrq_disconnect':
/home/x0174654/kernel.org/linux-master/linux/drivers/tty/sysrq.c:928: undefined reference to `del_timer_sync'
drivers/built-in.o: In function `kd_mksound':
/home/x0174654/kernel.org/linux-master/linux/drivers/tty/vt/keyboard.c:256: undefined reference to `del_timer_sync'
drivers/built-in.o: In function `do_blank_screen':
/home/x0174654/kernel.org/linux-master/linux/drivers/tty/vt/vt.c:3842: undefined reference to `del_timer_sync'
drivers/built-in.o: In function `univ8250_release_irq':
/home/x0174654/kernel.org/linux-master/linux/drivers/tty/serial/8250/8250_core.c:365: undefined reference to `del_timer_sync'
drivers/built-in.o:/home/x0174654/kernel.org/linux-master/linux/drivers/base/power/wakeup.c:114: more undefined references to `del_timer_sync' follow
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
---
kernel/time/timer.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/kernel/time/timer.c b/kernel/time/timer.c
index c3314fc..a460cbb 100644
--- a/kernel/time/timer.c
+++ b/kernel/time/timer.c
@@ -1118,7 +1118,7 @@ int try_to_del_timer_sync(struct timer_list *timer)
}
EXPORT_SYMBOL(try_to_del_timer_sync);
-#ifdef CONFIG_SMP
+#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL)
/**
* del_timer_sync - deactivate a timer and wait for the handler to finish.
* @timer: the timer to be deactivated
--
2.6.4
^ permalink raw reply related [flat|nested] 16+ messages in thread
* Re: [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report
2015-12-28 18:00 [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report Grygorii Strashko
2015-12-28 18:00 ` Grygorii Strashko
2015-12-28 18:00 ` [4.4-rc6-rt1 PATCH 2/2] kernel/time/timer.c: UP: fix undefined reference to `del_timer_sync' Grygorii Strashko
@ 2016-01-12 17:15 ` Sebastian Andrzej Siewior
2016-01-13 11:41 ` Grygorii Strashko
2 siblings, 1 reply; 16+ messages in thread
From: Sebastian Andrzej Siewior @ 2016-01-12 17:15 UTC (permalink / raw)
To: Grygorii Strashko; +Cc: linux-rt-users, linux-kernel, Sekhar Nori
* Grygorii Strashko | 2015-12-28 20:00:09 [+0200]:
>Hi All, Sebastian,
Hi Grygorii,
>3) Seems there are some problem with HIGHMEM :(. System crashes if I'm trying
> to run hackbench (see log 1). I can run hackbench only if HIGHMEM is disabled.
> I've also tried to disable CONFIG_CPU_SW_DOMAIN_PAN, CONFIG_ARM_PATCH_PHYS_VIRT,
> and CONFIG_CPUSETS, before disabling HIGHMEM.
> Sorry, I'll not able to investigate it further because of Christmas holidays here.
I see. It seems not to trigger anything on x86.
>4) Could we revert commit 4fa5422 'ARM: OMAP2: Drop the concept of certain power
> domains not being able to lose context.'? It's not required for 4.4-rt wakeup
> and introduces build warnings.
done.
Sebastian
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report
2016-01-12 17:15 ` [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report Sebastian Andrzej Siewior
@ 2016-01-13 11:41 ` Grygorii Strashko
2016-01-13 20:14 ` Grygorii Strashko
0 siblings, 1 reply; 16+ messages in thread
From: Grygorii Strashko @ 2016-01-13 11:41 UTC (permalink / raw)
To: Sebastian Andrzej Siewior; +Cc: linux-rt-users, linux-kernel, Sekhar Nori
On 01/12/2016 07:15 PM, Sebastian Andrzej Siewior wrote:
> * Grygorii Strashko | 2015-12-28 20:00:09 [+0200]:
>
>> Hi All, Sebastian,
> Hi Grygorii,
>
>> 3) Seems there are some problem with HIGHMEM :(. System crashes if I'm trying
>> to run hackbench (see log 1). I can run hackbench only if HIGHMEM is disabled.
>> I've also tried to disable CONFIG_CPU_SW_DOMAIN_PAN, CONFIG_ARM_PATCH_PHYS_VIRT,
>> and CONFIG_CPUSETS, before disabling HIGHMEM.
>> Sorry, I'll not able to investigate it further because of Christmas holidays here.
>
> I see. It seems not to trigger anything on x86.
Yah. I saw other reports - it seems this issue is not triggered on ARM SMP also
(as per report from Tim Sander). Will try to dig deeper.
>
>> 4) Could we revert commit 4fa5422 'ARM: OMAP2: Drop the concept of certain power
>> domains not being able to lose context.'? It's not required for 4.4-rt wakeup
>> and introduces build warnings.
>
> done.
Thanks.
--
regards,
-grygorii
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report
2016-01-13 11:41 ` Grygorii Strashko
@ 2016-01-13 20:14 ` Grygorii Strashko
2016-01-14 20:11 ` Grygorii Strashko
0 siblings, 1 reply; 16+ messages in thread
From: Grygorii Strashko @ 2016-01-13 20:14 UTC (permalink / raw)
To: Sebastian Andrzej Siewior; +Cc: linux-rt-users, linux-kernel, Sekhar Nori
Hi Sebastian,
On 01/13/2016 01:41 PM, Grygorii Strashko wrote:
> On 01/12/2016 07:15 PM, Sebastian Andrzej Siewior wrote:
>> * Grygorii Strashko | 2015-12-28 20:00:09 [+0200]:
>>
>>> Hi All, Sebastian,
>> Hi Grygorii,
>>
>>> 3) Seems there are some problem with HIGHMEM :(. System crashes if I'm trying
>>> to run hackbench (see log 1). I can run hackbench only if HIGHMEM is disabled.
>>> I've also tried to disable CONFIG_CPU_SW_DOMAIN_PAN, CONFIG_ARM_PATCH_PHYS_VIRT,
>>> and CONFIG_CPUSETS, before disabling HIGHMEM.
>>> Sorry, I'll not able to investigate it further because of Christmas holidays here.
>>
>> I see. It seems not to trigger anything on x86.
>
> Yah. I saw other reports - it seems this issue is not triggered on ARM SMP also
> (as per report from Tim Sander). Will try to dig deeper.
>
I did more tests, but still don't know what is wrong.
1) vanilla 4.4 - no issue
2) 4.4-rt: CONFIG_PREEMPT_RT_BASE=n and CONFIG_PREEMPT_RT_FULL=n -- no issue
3) 4.4-rt: CONFIG_DEBUG_HIGHMEM=y -- issue is present log1
4) 4.4-rt: this is interesting ;)
I can run hackbench few times without issues If I run hackbench with RT priority (log2)
But I got crash immediately once I run it using non-RT priorities.
======================== log1 =============================
/ #
/ # ./hackbench
Running in process mode with 10 groups using 40 file descriptors each (== 400 tasks)
Each sender will pass 100 messages of 100 bytes
[ 104.105133] ------------[ cut here ]------------
[ 104.105163] kernel BUG at arch/arm/mm/highmem.c:90!
[ 104.105163] Internal error: Oops - BUG: 0 [#1] PREEMPT SMP ARM
[ 104.105163] Modules linked in:
[ 104.105194] CPU: 0 PID: 168 Comm: hackbench Not tainted 4.4.0-rt2-debug-00261-g16b2b9c #51
[ 104.105194] Hardware name: Generic AM43 (Flattened Device Tree)
[ 104.105194] task: edeb9a00 ti: edeec000 task.ti: edeec000
[ 104.105224] PC is at kmap_atomic+0x174/0x178
[ 104.105224] LR is at 0x7ff
[ 104.105224] pc : [<c00200d8>] lr : [<000007ff>] psr: a0000113
[ 104.105224] sp : edeedd80 ip : c0004000 fp : b6fce000
[ 104.105224] r10: 000be875 r9 : 00000001 r8 : ffefe000
[ 104.105224] r7 : 00080000 r6 : ef6f9000 r5 : 0000004f r4 : effc4074
[ 104.105224] r3 : efffd000 r2 : bec6524f r1 : c084e414 r0 : 000000fe
[ 104.105255] Flags: NzCv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user
[ 104.105255] Control: 10c5387d Table: adef0059 DAC: 00000055
[ 104.105255] Process hackbench (pid: 168, stack limit = 0xedeec218)
[ 104.105255] Stack: (0xedeedd80 to 0xedeee000)
[ 104.105255] dd80: effc4074 effc558c effc558c edeae780 b6fce000 c00208f0 000be875 c0020900
[ 104.105285] dda0: effc4074 eded0000 effc558c c01585a4 edecb738 c003658c edeb9a00 eded0000
[ 104.105285] ddc0: edeedddc c00667c4 2e0cb000 effc4074 eded0000 be8753cf b6fce000 edecb738
[ 104.105285] dde0: edeae780 edecb738 b6fce000 c0159fa8 edef2db8 be8753cf effc4074 edeae7b4
[ 104.105285] de00: c015b668 edeae7b4 be8753cf c09af5c4 c120c640 edecb738 b6fce000 c06b06fc
[ 104.105316] de20: eded0000 edeae780 edeae7b4 be8753cf c09af5c4 c120c640 edecb738 c015bdc8
[ 104.105316] de40: edef2db8 edeae7b4 be8753cf c0a2e768 00000055 edef2db8 00000040 000001ce
[ 104.105316] de60: edecb000 00000001 00000000 00000000 0000006d edeae808 c09c0f28 60000113
[ 104.105316] de80: c082d628 0000012f b6fce1f8 c008c334 c0a2d88a 00000000 c082d628 edeedfb0
[ 104.105316] dea0: edeb9a00 0000081f edeae780 b6fce1f8 eded0000 edeae808 00000055 c06b18ac
[ 104.105346] dec0: 00000001 00000001 000003b4 00000079 00000001 00000000 00000000 00000800
[ 104.105346] dee0: 00000000 edeb9a00 edeb9f00 c09b1aa4 0000081f c06b1570 b6fce1f8 edeedfb0
[ 104.105346] df00: 0000006d 0000006d b6fb0000 c000935c 00000000 00000001 c11fc8fc edeb9f00
[ 104.105346] df20: 00000000 c008dcb4 60000113 c008dfe4 c0383ac0 80000113 c0383ac0 80000113
[ 104.105377] df40: c0383ac0 80000113 ffffffff c06b0ea8 b6ff7068 beffffff 000000a8 60000113
[ 104.105377] df60: beffffff b6ff7068 00000000 00000000 00000000 00000000 00000000 edeedfac
[ 104.105377] df80: f7ff706b edeedfa0 c0065d1c c0383ac0 b6f3fdce 00000030 b6f3fdce 00000030
[ 104.105377] dfa0: ffffffff 10c5387d 10c5387d c06b11a0 00000000 00000004 000000a8 b6fce1f8
[ 104.105377] dfc0: 00000000 bef9cc50 b6ff7000 bef9cc70 b6ff74c0 0000006d 0000006d b6fb0000
[ 104.105407] dfe0: 00000078 bef9cc50 b6f3fdaf b6f3fdce 00000030 ffffffff 84e7c1f5 00b90c8b
[ 104.105407] [<c00200d8>] (kmap_atomic) from [<c0020900>] (v6_copy_user_highpage_nonaliasing+0x10/0x3c)
[ 104.105438] [<c0020900>] (v6_copy_user_highpage_nonaliasing) from [<c01585a4>] (wp_page_copy+0x90/0x42c)
[ 104.105438] [<c01585a4>] (wp_page_copy) from [<c0159fa8>] (do_wp_page+0xb0/0x58c)
[ 104.105468] [<c0159fa8>] (do_wp_page) from [<c015bdc8>] (handle_mm_fault+0xba4/0x131c)
[ 104.105468] [<c015bdc8>] (handle_mm_fault) from [<c06b18ac>] (do_page_fault+0x33c/0x464)
[ 104.105468] [<c06b18ac>] (do_page_fault) from [<c000935c>] (do_DataAbort+0x38/0xb8)
[ 104.105499] [<c000935c>] (do_DataAbort) from [<c06b11a0>] (__dabt_usr+0x40/0x60)
[ 104.105499] Exception stack(0xedeedfb0 to 0xedeedff8)
[ 104.105499] dfa0: 00000000 00000004 000000a8 b6fce1f8
[ 104.105499] dfc0: 00000000 bef9cc50 b6ff7000 bef9cc70 b6ff74c0 0000006d 0000006d b6fb0000
[ 104.105499] dfe0: 00000078 bef9cc50 b6f3fdaf b6f3fdce 00000030 ffffffff
[ 104.105529] Code: f57ff06f e1a00008 e8bd87f0 e7f001f2 (e7f001f2)
[ 104.459960] ---[ end trace 0000000000000002 ]---
[ 104.459960] Kernel panic - not syncing: Fatal exception
[ 104.469879] ---[ end Kernel panic - not syncing: Fatal exception
[ 104.470123] ------------[ cut here ]------------
[ 104.470123] kernel BUG at arch/arm/mm/highmem.c:90!
[ 104.470123] Internal error: Oops - BUG: 0 [#2] PREEMPT SMP ARM
[ 104.470153] Modules linked in:
[ 104.470153] CPU: 0 PID: 169 Comm: hackbench Tainted: G D 4.4.0-rt2-debug-00261-g16b2b9c #51
[ 104.470153] Hardware name: Generic AM43 (Flattened Device Tree)
[ 104.470153] task: edeba700 ti: edeee000 task.ti: edeee000
[ 104.470184] PC is at kmap_atomic+0x174/0x178
[ 104.470184] LR is at 0x7ff
[ 104.470184] pc : [<c00200d8>] lr : [<000007ff>] psr: a0000113
[ 104.470184] sp : edeefdc0 ip : c0004000 fp : b6ff7000
[ 104.470184] r10: 000007dc r9 : 00000001 r8 : ffefe000
[ 104.470184] r7 : 00080000 r6 : ef6f9000 r5 : 0000004f r4 : effccd80
[ 104.470184] r3 : efffd000 r2 : bec6524f r1 : c084e414 r0 : 000000fe
[ 104.470184] Flags: NzCv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user
[ 104.470214] Control: 10c5387d Table: adef4059 DAC: 00000055
[ 104.470214] Process hackbench (pid: 169, stack limit = 0xedeee218)
[ 104.470214] Stack: (0xedeefdc0 to 0xedef0000)
[ 104.470214] fdc0: eded09f8 edeaea00 c09c501c 000001f7 c09af5c4 c120c640 000007dc c015b3e8
[ 104.470214] fde0: 00000000 00000000 c015b294 c0a2e768 00000015 edef6db8 00000000 000001f7
[ 104.470245] fe00: 000007dc c06b1600 00000000 00000000 00000000 edeaea88 00000015 00000001
[ 104.470245] fe20: edeaea88 c06b1600 edeaea00 b6ff7068 00000000 edeaea88 00000015 edeeff50
[ 104.470245] fe40: edeba700 00000817 edeaea00 b6ff7068 eded09f8 edeaea88 00000015 c06b18ac
[ 104.470245] fe60: 46645842 17d0a331 40676928 65e14918 0121ab18 0802f85c 2d40a304 00000800
[ 104.470245] fe80: 89a57429 57055c13 bd1cad2d c09b1a24 00000817 c06b1570 b6ff7068 edeeff50
[ 104.470275] fea0: 00000000 00000000 edeeffac c000935c 3934894a c2228357 b030396c a0149d40
[ 104.470275] fec0: 28cf2586 88402298 4227321a 0098d964 6100c3ba 45034574 da227e42 92a62662
[ 104.470275] fee0: b103ebd0 51162a81 1b14562f 8ab07f9d b159415c 00000000 00000001 ef6ae150
[ 104.470275] ff00: 00000000 00000000 00000000 00000000 00000001 c09c0f38 00000000 00000000
[ 104.470306] ff20: 00000002 c0a2e768 60000113 c008dfe4 00000000 00000000 c0383ac0 80000113
[ 104.470306] ff40: ffffffff edeeff84 00000000 c06b0e9c b6ff7068 beffffff 000000a9 60000113
[ 104.470306] ff60: beffffff b6ff7068 00000000 00000000 00000000 00000000 00000000 edeeffac
[ 104.470306] ff80: f7ff706b edeeffa0 c0065d1c c0383ac0 80000113 ffffffff 00000000 c0065d0c
[ 104.470306] ffa0: 00000000 00000000 00000000 c0010338 00000000 00000000 00000000 00000000
[ 104.470336] ffc0: b6ff7068 bef9cc50 b6ff7000 00000078 b6ff74c0 0000006d 0000006d b6fb0000
[ 104.470336] ffe0: 00000078 bef9cc48 b6f3fdaf b6ee9276 00000030 01200011 77168c61 1e1444b0
[ 104.470336] [<c00200d8>] (kmap_atomic) from [<c015b3e8>] (handle_mm_fault+0x1c4/0x131c)
[ 104.470367] [<c015b3e8>] (handle_mm_fault) from [<c06b18ac>] (do_page_fault+0x33c/0x464)
[ 104.470367] [<c06b18ac>] (do_page_fault) from [<c000935c>] (do_DataAbort+0x38/0xb8)
[ 104.470367] [<c000935c>] (do_DataAbort) from [<c06b0e9c>] (__dabt_svc+0x3c/0x60)
[ 104.470367] Exception stack(0xedeeff50 to 0xedeeff98)
[ 104.470397] ff40: b6ff7068 beffffff 000000a9 60000113
[ 104.470397] ff60: beffffff b6ff7068 00000000 00000000 00000000 00000000 00000000 edeeffac
[ 104.470397] ff80: f7ff706b edeeffa0 c0065d1c c0383ac0 80000113 ffffffff
[ 104.470428] [<c06b0e9c>] (__dabt_svc) from [<c0383ac0>] (__put_user_4+0xc/0x18)
[ 104.470428] Code: f57ff06f e1a00008 e8bd87f0 e7f001f2 (e7f001f2)
======================== log2 =============================
/ # chrt -f 99 ./hackbench
Running in process mode with 10 groups using 40 [ 78.947326] sched: RT throttling activated
file descriptors each (== 400 tasks)
Each sende[ 78.948150] NOHZ: local_softirq_pending 02
r will pass 100 messages of 100 bytes
Time: 10.[ 78.956817] NOHZ: local_softirq_pending 02
600
[ 78.966949] NOHZ: local_softirq_pending 02
[ 78.976593] NOHZ: local_softirq_pending 02
[ 78.986572] NOHZ: local_softirq_pending 02
[ 78.996582] NOHZ: local_softirq_pending 02
[ 79.006561] NOHZ: local_softirq_pending 02
[ 79.956848] NOHZ: local_softirq_pending 02
[ 79.966735] NOHZ: local_softirq_pending 02
[ 79.976562] NOHZ: local_softirq_pending 02
/ #
/ # chrt -f 98 ./hackbench
Running in process mode with 10 groups using 40 file descriptors each (== 400 tasks)
Each sender will pass 100 messages of 100 bytes
Time: 11.468
/ # ./hackbench
Running in process mode with 10 groups using 40 file descriptors each (== 400 tasks)
Each sender will pass 100 messages of 100 bytes
[ 158.488372] Unable to handle kernel paging request at virtual address ffefe2a0
[ 158.522277] pgd = ed718000
[ 158.525115] [ffefe2a0] *pgd=afffd861, *pte=00000000, *ppte=00000000
[ 158.543548] Internal error: Oops: 837 [#1] PREEMPT SMP ARM
[ 158.543548] Modules linked in:
[ 158.543548] CPU: 0 PID: 904 Comm: hackbench Not tainted 4.4.0-rt2-00261-g16b2b9c #52
[ 158.543579] Hardware name: Generic AM43 (Flattened Device Tree)
[ 158.543579] task: edbd9b80 ti: eefb6000 task.ti: eefb6000
[ 158.543609] PC is at cpu_ca8_set_pte_ext+0x0/0x58
[ 158.543609] LR is at filemap_map_pages+0x3bc/0x414
[ 158.543609] pc : [<c00222a8>] lr : [<c0125c88>] psr: 40000113
[ 158.543609] sp : eefb7de0 ip : 00000000 fp : ee634d5c
[ 158.543640] r10: edd60000 r9 : c0a324ee r8 : 00000000
[ 158.543640] r7 : ee634bf8 r6 : eefb7e7c r5 : ef937a34 r4 : ee5bbf14
[ 158.543640] r3 : c00222a8 r2 : 00000800 r1 : 8ff6518f r0 : ffefe2a0
[ 158.543640] Flags: nZcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment none
[ 158.543640] Control: 10c5387d Table: ad718059 DAC: 00000051
[ 158.543640] Process hackbench (pid: 904, stack limit = 0xeefb6218)
[ 158.543670] Stack: (0xeefb7de0 to 0xeefb8000)
[ 158.543670] 7de0: 00000000 00000000 c015b178 c015b078 ee634d60 c015b178 b6ea0000 ed9804d0
[ 158.543670] 7e00: 000000d8 00000000 ee634cac ee5bbf14 c0977128 00000057 00000080 b6ea1000
[ 158.543670] 7e20: 000000a1 ed9804d0 edbb0500 b6ea1000 000d4000 b6f25000 00000050 ffefe284
[ 158.543701] 7e40: 000000a1 c015b178 00000000 00000000 c015a934 c0097314 00000001 00000054
[ 158.543701] 7e60: ed71adb8 00000040 00000284 edbb0534 ffefe000 00000000 00000008 00000054
[ 158.543701] 7e80: 0000004f b6ea0000 edbb0500 b6ea1e70 0000005e ffefe280 00000054 eefb7fb0
[ 158.543701] 7ea0: edbd9b80 80000007 edbb0500 b6ea1e70 ed9804d0 edbb0578 00000054 c068ece0
[ 158.543731] 7ec0: 00000000 edbda170 00000000 00000000 00000000 00000000 00000008 00000000
[ 158.543731] 7ee0: 00000000 c082e334 00000000 c097c8f4 00000007 c068e9b8 b6ea1e70 eefb7fb0
[ 158.543731] 7f00: 00000386 00000386 b6f2f000 c000943c 00000001 c09bc35c 00000000 00000000
[ 158.543731] 7f20: 00000002 c0a331d4 60000113 c0097314 00000000 00000000 c03761c0 80000113
[ 158.543762] 7f40: c03761c0 80000113 ffffffff c068e138 b6f76068 beffffff 00000388 00000000
[ 158.543762] 7f60: 00000051 b6f76068 00000000 00000000 00000000 00000000 00000000 eefb7fac
[ 158.543762] 7f80: f7f7606b eefb7fa0 00000000 beefac50 b6f76000 c000f99c b6ea1e70 00000030
[ 158.543762] 7fa0: ffffffff 10c5387d 10c5387d c068e814 00000000 0000000c 00000388 b6f37bfd
[ 158.543792] 7fc0: 00000000 beefac50 b6f76000 beefac70 b6f764c0 00000386 00000386 b6f2f000
[ 158.543792] 7fe0: 00000152 beefac50 b6ebede3 b6ea1e70 00000030 ffffffff afffd861 afffdc61
[ 158.543823] [<c00222a8>] (cpu_ca8_set_pte_ext) from [<c015b178>] (handle_mm_fault+0x8ac/0x12f8)
[ 158.543823] [<c015b178>] (handle_mm_fault) from [<c068ece0>] (do_page_fault+0x328/0x45c)
[ 158.543853] [<c068ece0>] (do_page_fault) from [<c000943c>] (do_PrefetchAbort+0x34/0x98)
[ 158.543853] [<c000943c>] (do_PrefetchAbort) from [<c068e814>] (ret_from_exception+0x0/0x2c)
[ 158.543853] Exception stack(0xeefb7fb0 to 0xeefb7ff8)
[ 158.543853] 7fa0: 00000000 0000000c 00000388 b6f37bfd
[ 158.543884] 7fc0: 00000000 beefac50 b6f76000 beefac70 b6f764c0 00000386 00000386 b6f2f000
[ 158.543884] 7fe0: 00000152 beefac50 b6ebede3 b6ea1e70 00000030 ffffffff
[ 158.543884] Code: f57ff06f ee020f10 f57ff06f e12fff1e (e5801000)
[ 158.987609] Unable to handle kernel paging request at virtual address ffefe5c8
[ 158.995178] pgd = ed80c000
[ 159.015991] [ffefe5c8] *pgd=afffd861, *pte=00000000, *ppte=00000000
[ 159.032867] Internal error: Oops: 837 [#2] PREEMPT SMP ARM
[ 159.032867] Modules linked in:
[ 159.032867] CPU: 0 PID: 1012 Comm: hackbench Tainted: G D 4.4.0-rt2-00261-g16b2b9c #52
[ 159.032867] Hardware name: Generic AM43 (Flattened Device Tree)
[ 159.032897] task: ed536e00 ti: ed808000 task.ti: ed808000
[ 159.032897] PC is at cpu_ca8_set_pte_ext+0x0/0x58
[ 159.032928] LR is at filemap_map_pages+0x3bc/0x414
[ 159.032928] pc : [<c00222a8>] lr : [<c0125c88>] psr: 40000113
[ 159.032928] sp : ed809de0 ip : 00000000 fp : ee6128dc
[ 159.032928] r10: ed271000 r9 : c0a324ee r8 : 00000000
[ 159.032958] r7 : ee612778 r6 : ed809e7c r5 : ef937350 r4 : ee5bb688
[ 159.032958] r3 : c00222a8 r2 : 00000800 r1 : 8ff3418f r0 : ffefe5c8
[ 159.032958] Flags: nZcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment none
[ 159.032958] Control: 10c5387d Table: ad80c059 DAC: 00000051
[ 159.032958] Process hackbench (pid: 1012, stack limit = 0xed808218)
[ 159.032958] Stack: (0xed809de0 to 0xed80a000)
[ 159.032989] 9de0: 00000000 00000000 c015b178 c015b078 ee6128e0 c015b178 b6f70000 ed0c1790
[ 159.032989] 9e00: 00000018 00000000 ee61282c ee5bb688 c0977128 00000016 00000040 b6f70000
[ 159.032989] 9e20: 00000170 ed0c1790 ed80a000 b6f70000 00017000 b6f73000 00000014 ffefe5c0
[ 159.033020] 9e40: 00000170 c015b178 00000000 00000000 c015a934 c0097314 00000001 00000054
[ 159.033020] 9e60: ed80edb8 00000040 000005c0 ed80a034 ffefe000 00000000 00000008 00000054
[ 159.033020] 9e80: 00000014 b6f70000 ed80a000 b6f70378 00000016 ffefe5c0 00000054 ed809fb0
[ 159.033020] 9ea0: ed536e00 00000017 ed80a000 b6f70378 ed0c1790 ed80a078 00000054 c068ece0
[ 159.033020] 9ec0: 00000000 00000000 00000000 c0a331d4 60000013 00000000 00000000 00000000
[ 159.033050] 9ee0: 00000000 c005093c 00000000 c097c6f4 00000017 c068e9b8 b6f70378 ed809fb0
[ 159.033050] 9f00: 00000000 b6f7adc8 10a8b550 c0009388 ed809f68 c003dc08 ed536e00 00000001
[ 159.033050] 9f20: ed809f3c c006e5a0 ed7405e8 00000001 ed536e00 00000001 000221ac c00509d8
[ 159.033050] 9f40: ffffffff beefaa34 00000051 00000051 c000fa64 beefaa34 00000014 00000000
[ 159.033081] 9f60: 00000000 c0050e50 00000000 14000000 b6e76aa0 00000002 00000000 0001121d
[ 159.033081] 9f80: 14000000 b6e76aa0 00000002 00000000 80000013 beefab64 b6f6d254 80000030
[ 159.033081] 9fa0: ffffffff 10c5387d 10c5387d c068e508 000105f2 b6f70378 0000000b 0000006c
[ 159.033081] 9fc0: b6f7b960 000105f1 b6f7759c b6f7b960 00000005 00000000 b6f7adc8 10a8b550
[ 159.033111] 9fe0: beefaa90 beefaa58 b6f678cd b6f6d254 80000030 ffffffff 00000000 00000000
[ 159.033142] [<c00222a8>] (cpu_ca8_set_pte_ext) from [<c015b178>] (handle_mm_fault+0x8ac/0x12f8)
[ 159.033142] [<c015b178>] (handle_mm_fault) from [<c068ece0>] (do_page_fault+0x328/0x45c)
[ 159.033172] [<c068ece0>] (do_page_fault) from [<c0009388>] (do_DataAbort+0x34/0xb4)
[ 159.033172] [<c0009388>] (do_DataAbort) from [<c068e508>] (__dabt_usr+0x48/0x60)
[ 159.033172] Exception stack(0xed809fb0 to 0xed809ff8)
[ 159.033172] 9fa0: 000105f2 b6f70378 0000000b 0000006c
[ 159.033172] 9fc0: b6f7b960 000105f1 b6f7759c b6f7b960 00000005 00000000 b6f7adc8 10a8b550
[ 159.033203] 9fe0: beefaa90 beefaa58 b6f678cd b6f6d254 80000030 ffffffff
[ 159.033203] Code: f57ff06f ee020f10 f57ff06f e12fff1e (e5801000)
[ 159.607635] Unable to handle kernel paging request at virtual address ffefe350
[ 159.615203] pgd = eefe0000
[ 159.636474] [ffefe350] *pgd=afffd861, *pte=00000000, *ppte=00000000
[ 159.653350] Internal error: Oops: 837 [#3] PREEMPT SMP ARM
[ 159.653381] Modules linked in:
[ 159.653381] CPU: 0 PID: 1208 Comm: hackbench Tainted: G D 4.4.0-rt2-00261-g16b2b9c #52
[ 159.653381] Hardware name: Generic AM43 (Flattened Device Tree)
[ 159.653381] task: edc35280 ti: eef5c000 task.ti: eef5c000
[ 159.653411] PC is at cpu_ca8_set_pte_ext+0x0/0x58
[ 159.653442] LR is at filemap_map_pages+0x3bc/0x414
[ 159.653442] pc : [<c00222a8>] lr : [<c0125c88>] psr: 40000113
[ 159.653442] sp : eef5dde0 ip : 00000000 fp : ee634d5c
[ 159.653442] r10: edd60000 r9 : c0a324ee r8 : 00000000
[ 159.653442] r7 : ee634bf8 r6 : eef5de7c r5 : efd97424 r4 : ee618024
[ 159.653472] r3 : c00222a8 r2 : 00000800 r1 : af10118f r0 : ffefe350
[ 159.653472] Flags: nZcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment none
[ 159.653472] Control: 10c5387d Table: aefe0059 DAC: 00000051
[ 159.653472] Process hackbench (pid: 1208, stack limit = 0xeef5c218)
[ 159.653472] Stack: (0xeef5dde0 to 0xeef5e000)
[ 159.653503] dde0: 00000000 00000000 c015b178 c015b078 ee634d60 c015b178 b6ed0000 ed311b00
[ 159.653503] de00: 000000d8 00000000 ee634cac ee618024 c0977128 00000083 000000c0 b6ed7000
[ 159.653503] de20: 000000d7 ed311b00 ed304a00 b6ed7000 000d4000 b6f25000 00000086 ffefe35c
[ 159.653503] de40: 000000d7 c015b178 00000000 00000000 c015a934 c0097314 00000001 00000054
[ 159.653533] de60: eefe2db8 00000040 0000035c ed304a34 ffefe000 00000000 00000000 00000054
[ 159.653533] de80: 0000007f b6ed0000 ed304a00 b6ed7470 0000008e ffefe340 00000054 eef5dfb0
[ 159.653533] dea0: edc35280 80000007 ed304a00 b6ed7470 ed311b00 ed304a78 00000054 c068ece0
[ 159.653533] dec0: 00000000 00000000 00000000 00000001 00000000 00000000 00000000 00000000
[ 159.653564] dee0: 00000000 00000000 00000000 c097c8f4 00000007 c068e9b8 b6ed7470 eef5dfb0
[ 159.653564] df00: 00024610 000221ac 000221ac c000943c eef5df68 beefabe7 00000001 00000001
[ 159.653564] df20: 00000000 00000000 eef5df1c 00000000 edd60300 00000000 00000000 00000000
[ 159.653564] df40: 00000000 edd60300 00000001 ee6682e0 00000002 edd60308 eef5c000 00000000
[ 159.653564] df60: 000221ac c0187d6c 00000000 00000000 edd60300 edd60300 beefabe7 00000001
[ 159.653594] df80: c000fa64 c018850c 00000000 00000000 80000013 0000002a b6ed7470 80000030
[ 159.653594] dfa0: ffffffff 10c5387d 10c5387d c068e814 beefabe8 00000001 ffffffff 00000001
[ 159.653594] dfc0: 0000002a 00000000 0001131d beefac68 000113c1 00024610 000221ac 000221ac
[ 159.653594] dfe0: b6ed7471 beefabe0 000112fd b6ed7470 80000030 ffffffff 00000000 00000000
[ 159.653625] [<c00222a8>] (cpu_ca8_set_pte_ext) from [<c015b178>] (handle_mm_fault+0x8ac/0x12f8)
[ 159.653656] [<c015b178>] (handle_mm_fault) from [<c068ece0>] (do_page_fault+0x328/0x45c)
[ 159.653656] [<c068ece0>] (do_page_fault) from [<c000943c>] (do_PrefetchAbort+0x34/0x98)
[ 159.653717] [<c000943c>] (do_PrefetchAbort) from [<c068e814>] (ret_from_exception+0x0/0x2c)
[ 159.653717] Exception stack(0xeef5dfb0 to 0xeef5dff8)
[ 159.653717] dfa0: beefabe8 00000001 ffffffff 00000001
[ 159.653717] dfc0: 0000002a 00000000 0001131d beefac68 000113c1 00024610 000221ac 000221ac
[ 159.653717] dfe0: b6ed7471 beefabe0 000112fd b6ed7470 80000030 ffffffff
[ 159.653747] Code: f57ff06f ee020f10 f57ff06f e12fff1e (e5801000)
[ 160.039001] Unable to handle kernel paging request at virtual address ffefdb20
[ 160.046752] pgd = edab8000
[ 160.049560] [ffefdb20] *pgd=afffd861, *pte=00000000, *ppte=00000000
[ 160.056152] Internal error: Oops: 837 [#4] PREEMPT SMP ARM
[ 160.056152] Modules linked in:
[ 160.056152] CPU: 0 PID: 902 Comm: hackbench Tainted: G D 4.4.0-rt2-00261-g16b2b9c #52
[ 160.056182] Hardware name: Generic AM43 (Flattened Device Tree)
[ 160.056182] task: edbd8dc0 ti: eefb2000 task.ti: eefb2000
[ 160.056182] PC is at copy_page+0x2c/0x80
[ 160.056182] LR is at 0x19ef1cf1
[ 160.056213] pc : [<c0373e0c>] lr : [<19ef1cf1>] psr: 20000113
[ 160.056213] sp : eefb3da0 ip : abfb4583 fp : 000000fa
[ 160.056213] r10: 000be6a9 r9 : edbb0280 r8 : c0021684
[ 160.056213] r7 : beefa000 r6 : effc00c0 r5 : ffefd000 r4 : bfdf5a74
[ 160.056213] r3 : 3f3779db r2 : 00000009 r1 : ffefeb30 r0 : ffefdb20
[ 160.056213] Flags: nzCv IRQs on FIQs on Mode SVC_32 ISA ARM Segment none
[ 160.056213] Control: 10c5387d Table: adab8059 DAC: 00000051
[ 160.056243] Process hackbench (pid: 902, stack limit = 0xeefb2218)
[ 160.056243] Stack: (0xeefb3da0 to 0xeefb4000)
[ 160.056243] 3da0: ffefe000 c00216ac effbffc4 edbb8108 effc00c0 c0157d88 edbd8dc0 edbb8108
[ 160.056243] 3dc0: eefb3ddc c006e5a0 effbffc4 edbb8108 be6a93cf effbffc4 edbb8108 be6a93cf
[ 160.056274] 3de0: beefa000 ed0913e8 edbb0280 ed0913e8 000000fa c0159664 edabafb8 be6a93cf
[ 160.056274] 3e00: effbffc4 edbb02b4 c015ace0 beefa000 be6a93cf edbb02b4 c09bf68c ed0913e8
[ 160.056274] 3e20: 000000fa c068d88c edbb8108 edbb0280 beefa000 be6a93cf edbb02b4 c09bf68c
[ 160.056274] 3e40: ed0913e8 c015b400 edabafb8 edbb02b4 be6a93cf c0097314 00000001 00000055
[ 160.056304] 3e60: edabafb8 00000040 ed091000 00000001 eefb3f2c c007e94c 00000001 00000001
[ 160.056304] 3e80: edbb02f8 c068ea48 edbb0280 beefac48 00000386 edbb02f8 00000055 eefb3fb0
[ 160.056304] 3ea0: edbd8dc0 0000081f edbb0280 beefac48 edbb8108 edbb02f8 00000055 c068ece0
[ 160.056304] 3ec0: edbd8dc0 c007c3a8 60000093 edbd8dc0 ef6b2e80 00000000 00000008 00000800
[ 160.056335] 3ee0: 00000000 c0a33008 ed3c4a40 c097c774 0000081f c068e9b8 beefac48 eefb3fb0
[ 160.056335] 3f00: 00000386 00000386 b6f2f000 c0009388 ed031000 00000000 b6f2f000 c068d51c
[ 160.056335] 3f20: ed3c44c0 beefac50 ed3c44c0 c003c4b0 ed031000 00000000 00000000 eefb3f58
[ 160.056335] 3f40: c06e9440 c01892d4 c0977ba0 00000000 ee677080 c0527538 ee949410 00000007
[ 160.056365] 3f60: ed02e600 00000000 00000051 ed02e480 00000018 b6f76068 beefac50 b6f76000
[ 160.056365] 3f80: 00000078 c000fa64 eefb2000 00000000 b6f2f000 c003cbb0 b6ea1ee6 60000030
[ 160.056365] 3fa0: ffffffff 10c5387d 10c5387d c068e508 000004ff 00000000 00000000 0008e7c4
[ 160.056365] 3fc0: 000004ff beefac50 b6f76000 beefac70 b6f764c0 00000386 00000386 b6f2f000
[ 160.056396] 3fe0: 00000078 beefac50 b6ebee77 b6ea1ee6 60000030 ffffffff afffd861 afffdc61
[ 160.056396] Code: e8a05018 e8b15018 e8a05018 e8b15018 (e8a05018)
[ 160.317626] ---[ end trace 0000000000000002 ]---
[ 160.322509] ------------[ cut here ]------------
[ 160.322509] Kernel BUG at c068bc0c [verbose debug info unavailable]
[ 160.322509] Internal error: Oops - BUG: 0 [#5] PREEMPT SMP ARM
[ 160.322540] Modules linked in:
[ 160.322540] CPU: 0 PID: 904 Comm: hackbench Tainted: G D 4.4.0-rt2-00261-g16b2b9c #52
[ 160.322540] Hardware name: Generic AM43 (Flattened Device Tree)
[ 160.322540] task: edbd9b80 ti: eefb6000 task.ti: eefb6000
[ 160.322570] PC is at rt_spin_lock_slowlock+0x248/0x2f8
[ 160.322570] LR is at 0xedbb055c
[ 160.322570] pc : [<c068bc0c>] lr : [<edbb055c>] psr: 60000113
[ 160.322570] sp : eefb7a78 ip : edbd9b80 fp : 00011000
[ 160.322570] r10: c0976454 r9 : 00012000 r8 : 00000001
[ 160.322570] r7 : edbd9b80 r6 : eefb7a88 r5 : 00000000 r4 : edbb0534
[ 160.322570] r3 : 00000001 r2 : edbd9b80 r1 : edbd9b80 r0 : 00000000
[ 160.322601] Flags: nZCv IRQs on FIQs on Mode SVC_32 ISA ARM Segment none
[ 160.322601] Control: 10c5387d Table: ad718059 DAC: 00000051
[ 160.322601] Process hackbench (pid: 904, stack limit = 0xeefb6218)
[ 160.322601] Stack: (0xeefb7a78 to 0xeefb8000)
[ 160.322601] 7a60: c0a2f2c0 c0a30108
[ 160.322601] 7a80: 00000100 00000000 eefb7a88 00000000 c0968ad8 eefb7a94 c0820b1b 00000000
[ 160.322631] 7aa0: 00000000 c0968ad8 00000001 00000141 00000000 edbb0534 c0158ef0 eefb7b88
[ 160.322631] 7ac0: 00000000 eefb7b88 00012000 c0976454 00011000 c068d89c 00000339 c0020bfc
[ 160.322631] 7ae0: 00000001 ffffffff ffefd040 ffefd040 ffffffff c0158ef0 eefb7af8 ed9805d8
[ 160.322631] 7b00: ed718004 00000000 edbb0534 ffefd040 edbb0500 edbb06c8 00012000 00000000
[ 160.322662] 7b20: ed718000 00011fff c0136d70 00000000 00000000 00000000 00000000 ed9805d8
[ 160.322662] 7b40: ffffffff eefb7b88 00000000 00000000 00000001 edbb0578 c00222aa c015a164
[ 160.322662] 7b60: 00000000 edbb0500 ed9805d8 edbb0500 eefb7bb0 60000113 edbb0500 c01608ec
[ 160.322662] 7b80: 60000113 c0097314 edbb0500 00000001 00000000 00000000 ffffffff 00000000
[ 160.322692] 7ba0: 00000000 00000000 00000400 ed032000 c0041644 c097befc 60000113 edbb0500
[ 160.322692] 7bc0: edbb0500 00000000 c097befc 60000113 edbb0500 00000000 c097befc c003a7a8
[ 160.322692] 7be0: edbd9b80 edbda0bc c097befc c0041670 60000113 c0123600 00000001 c0a38244
[ 160.322692] 7c00: eefb7d90 c097befc 60000113 0000000b 00000000 00000001 c00222aa c0014724
[ 160.322723] 7c20: eefb6218 0000000b c00222ac eefb6000 bf000000 00000008 660a2314 66663735
[ 160.322723] 7c40: 20663630 32306565 30316630 37356620 36306666 31652066 66666632 28206531
[ 160.322723] 7c60: 30383565 30303031 c0002029 eefb7c8c ee634d5c ffefe2a0 eefb7d90 00000837
[ 160.322723] 7c80: edbb0500 ffefe2a0 c0a324ee edd60000 ee634d5c c0023704 edbb0500 c068ede8
[ 160.322723] 7ca0: edbd9b80 c00947cc c068d584 edbd9b80 00000001 edbda940 c068a110 c0094930
[ 160.322753] 7cc0: ef6b2e80 00000000 eefb7d14 c097c6f4 00000837 c068e9b8 ffefe2a0 eefb7d90
[ 160.322753] 7ce0: c0a324ee edd60000 ee634d5c c0009388 edbd9b80 ef6b2e90 edbd9b80 ee8b1b80
[ 160.322753] 7d00: 00000000 edbb0500 00000000 c0a33008 eefb7d64 c068a110 c09690f4 00000000
[ 160.322753] 7d20: 00000001 edbda198 e681b990 00000024 00000001 edbda198 edbd9b80 00000004
[ 160.322784] 7d40: 00000006 eefb6000 edd60000 c00947cc c068e1f0 edbd9b80 00000001 eefb7dac
[ 160.322784] 7d60: 00000001 c0094930 c017913c 20000113 ffffffff c068e1f0 c00222a8 40000113
[ 160.322784] 7d80: ffffffff eefb7dc4 00000000 c068e12c ffefe2a0 8ff6518f 00000800 c00222a8
[ 160.322784] 7da0: ee5bbf14 ef937a34 eefb7e7c ee634bf8 00000000 c0a324ee edd60000 ee634d5c
[ 160.322814] 7dc0: 00000000 eefb7de0 c0125c88 c00222a8 40000113 ffffffff 00000051 c0125c88
[ 160.322814] 7de0: 00000000 00000000 c015b178 c015b078 ee634d60 c015b178 b6ea0000 ed9804d0
[ 160.322814] 7e00: 000000d8 00000000 ee634cac ee5bbf14 c0977128 00000057 00000080 b6ea1000
[ 160.322814] 7e20: 000000a1 ed9804d0 edbb0500 b6ea1000 000d4000 b6f25000 00000050 ffefe284
[ 160.322845] 7e40: 000000a1 c015b178 00000000 00000000 c015a934 c0097314 00000001 00000054
[ 160.322845] 7e60: ed71adb8 00000040 00000284 edbb0534 ffefe000 00000000 00000008 00000054
[ 160.322845] 7e80: 0000004f b6ea0000 edbb0500 b6ea1e70 0000005e ffefe280 00000054 eefb7fb0
[ 160.322845] 7ea0: edbd9b80 80000007 edbb0500 b6ea1e70 ed9804d0 edbb0578 00000054 c068ece0
[ 160.322875] 7ec0: 00000000 edbda170 00000000 00000000 00000000 00000000 00000008 00000000
[ 160.322875] 7ee0: 00000000 c082e334 00000000 c097c8f4 00000007 c068e9b8 b6ea1e70 eefb7fb0
[ 160.322875] 7f00: 00000386 00000386 b6f2f000 c000943c 00000001 c09bc35c 00000000 00000000
[ 160.322875] 7f20: 00000002 c0a331d4 60000113 c0097314 00000000 00000000 c03761c0 80000113
[ 160.322906] 7f40: c03761c0 80000113 ffffffff c068e138 b6f76068 beffffff 00000388 00000000
[ 160.322906] 7f60: 00000051 b6f76068 00000000 00000000 00000000 00000000 00000000 eefb7fac
[ 160.322906] 7f80: f7f7606b eefb7fa0 00000000 beefac50 b6f76000 c000f99c b6ea1e70 00000030
[ 160.322906] 7fa0: ffffffff 10c5387d 10c5387d c068e814 00000000 0000000c 00000388 b6f37bfd
[ 160.322906] 7fc0: 00000000 beefac50 b6f76000 beefac70 b6f764c0 00000386 00000386 b6f2f000
[ 160.322937] 7fe0: 00000152 beefac50 b6ebede3 b6ea1e70 00000030 ffffffff afffd861 afffdc61
[ 160.322937] [<c068bc0c>] (rt_spin_lock_slowlock) from [<c068d89c>] (rt_spin_lock+0x80/0x84)
[ 160.322967] [<c068d89c>] (rt_spin_lock) from [<c0158ef0>] (unmap_single_vma+0x16c/0x56c)
[ 160.322967] [<c0158ef0>] (unmap_single_vma) from [<c015a164>] (unmap_vmas+0x54/0x68)
[ 160.322998] [<c015a164>] (unmap_vmas) from [<c01608ec>] (exit_mmap+0xd8/0x1dc)
[ 160.322998] [<c01608ec>] (exit_mmap) from [<c003a7a8>] (mmput+0x44/0xec)
[ 160.323028] [<c003a7a8>] (mmput) from [<c0041670>] (do_exit+0x2c0/0xb98)
[ 160.323028] [<c0041670>] (do_exit) from [<c0014724>] (die+0x430/0x480)
[ 160.323059] [<c0014724>] (die) from [<c0023704>] (__do_kernel_fault.part.0+0x54/0x1e4)
[ 160.323089] [<c0023704>] (__do_kernel_fault.part.0) from [<c068ede8>] (do_page_fault+0x430/0x45c)
[ 160.323089] [<c068ede8>] (do_page_fault) from [<c0009388>] (do_DataAbort+0x34/0xb4)
[ 160.323089] [<c0009388>] (do_DataAbort) from [<c068e12c>] (__dabt_svc+0x4c/0x80)
[ 160.323089] Exception stack(0xeefb7d90 to 0xeefb7dd8)
[ 160.323120] 7d80: ffefe2a0 8ff6518f 00000800 c00222a8
[ 160.323120] 7da0: ee5bbf14 ef937a34 eefb7e7c ee634bf8 00000000 c0a324ee edd60000 ee634d5c
[ 160.323120] 7dc0: 00000000 eefb7de0 c0125c88 c00222a8 40000113 ffffffff
[ 160.323120] [<c068e12c>] (__dabt_svc) from [<c00222a8>] (cpu_ca8_set_pte_ext+0x0/0x58)
[ 160.323150] [<c00222a8>] (cpu_ca8_set_pte_ext) from [<c015b178>] (handle_mm_fault+0x8ac/0x12f8)
[ 160.323150] [<c015b178>] (handle_mm_fault) from [<c068ece0>] (do_page_fault+0x328/0x45c)
[ 160.323150] [<c068ece0>] (do_page_fault) from [<c000943c>] (do_PrefetchAbort+0x34/0x98)
[ 160.323181] [<c000943c>] (do_PrefetchAbort) from [<c068e814>] (ret_from_exception+0x0/0x2c)
[ 160.323181] Exception stack(0xeefb7fb0 to 0xeefb7ff8)
[ 160.323181] 7fa0: 00000000 0000000c 00000388 b6f37bfd
[ 160.323181] 7fc0: 00000000 beefac50 b6f76000 beefac70 b6f764c0 00000386 00000386 b6f2f000
[ 160.323181] 7fe0: 00000152 beefac50 b6ebede3 b6ea1e70 00000030 ffffffff
[ 160.323211] Code: e1540002 1a000001 e1530006 1a000010 (e7f001f2)
[ 161.006896] ---[ end trace 0000000000000003 ]---
[ 161.006896] Fixing recursive fault but reboot is needed!
[ 161.007080] ---[ end trace 0000000000000004 ]---
--
regards,
-grygorii
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report
2016-01-13 20:14 ` Grygorii Strashko
@ 2016-01-14 20:11 ` Grygorii Strashko
0 siblings, 0 replies; 16+ messages in thread
From: Grygorii Strashko @ 2016-01-14 20:11 UTC (permalink / raw)
To: Sebastian Andrzej Siewior
Cc: linux-rt-users, linux-kernel, Sekhar Nori, linux-arm
Hi Sebastian,
On 01/13/2016 10:14 PM, Grygorii Strashko wrote:
> On 01/13/2016 01:41 PM, Grygorii Strashko wrote:
>> On 01/12/2016 07:15 PM, Sebastian Andrzej Siewior wrote:
>>> * Grygorii Strashko | 2015-12-28 20:00:09 [+0200]:
>>>
>>>
>>>> 3) Seems there are some problem with HIGHMEM :(. System crashes if I'm trying
>>>> to run hackbench (see log 1). I can run hackbench only if HIGHMEM is disabled.
>>>> I've also tried to disable CONFIG_CPU_SW_DOMAIN_PAN, CONFIG_ARM_PATCH_PHYS_VIRT,
>>>> and CONFIG_CPUSETS, before disabling HIGHMEM.
>>>> Sorry, I'll not able to investigate it further because of Christmas holidays here.
>>>
>>> I see. It seems not to trigger anything on x86.
>>
>> Yah. I saw other reports - it seems this issue is not triggered on ARM SMP also
>> (as per report from Tim Sander). Will try to dig deeper.
>>
>
> I did more tests, but still don't know what is wrong.
> 1) vanilla 4.4 - no issue
> 2) 4.4-rt: CONFIG_PREEMPT_RT_BASE=n and CONFIG_PREEMPT_RT_FULL=n -- no issue
> 3) 4.4-rt: CONFIG_DEBUG_HIGHMEM=y -- issue is present log1
> 4) 4.4-rt: this is interesting ;)
> I can run hackbench few times without issues If I run hackbench with RT priority (log2)
> But I got crash immediately once I run it using non-RT priorities.
Seems I've found root cause of the issue - FIX_KMAP_BEGIN offset need to be taken into
account in switch_kmaps() and below diff fixes this issue for me.
What I'm not fully understand is why I'm the only one who hit this issue :(
New functionality was added by commit:
commit a5f4c561b3b19a9bc43a81da6382b0098ebbc1fb
ARM: 8415/1: early fixmap support for earlycon
and it is enabled by default.
Luckiness :(
diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c
index deabc36..b9b4f9c 100644
--- a/arch/arm/mm/highmem.c
+++ b/arch/arm/mm/highmem.c
@@ -167,7 +167,7 @@ void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p)
* Clear @prev's kmap_atomic mappings
*/
for (i = 0; i < prev_p->kmap_idx; i++) {
- int idx = i + KM_TYPE_NR * smp_processor_id();
+ int idx = FIX_KMAP_BEGIN + i + KM_TYPE_NR * smp_processor_id();
set_fixmap_pte(idx, __pte(0));
}
@@ -175,7 +175,7 @@ void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p)
* Restore @next_p's kmap_atomic mappings
*/
for (i = 0; i < next_p->kmap_idx; i++) {
- int idx = i + KM_TYPE_NR * smp_processor_id();
+ int idx = FIX_KMAP_BEGIN + i + KM_TYPE_NR * smp_processor_id();
if (!pte_none(next_p->kmap_pte[i]))
set_fixmap_pte(idx, next_p->kmap_pte[i]);
--
regards,
-grygorii
^ permalink raw reply related [flat|nested] 16+ messages in thread* [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report
@ 2016-01-14 20:11 ` Grygorii Strashko
0 siblings, 0 replies; 16+ messages in thread
From: Grygorii Strashko @ 2016-01-14 20:11 UTC (permalink / raw)
To: linux-arm-kernel
Hi Sebastian,
On 01/13/2016 10:14 PM, Grygorii Strashko wrote:
> On 01/13/2016 01:41 PM, Grygorii Strashko wrote:
>> On 01/12/2016 07:15 PM, Sebastian Andrzej Siewior wrote:
>>> * Grygorii Strashko | 2015-12-28 20:00:09 [+0200]:
>>>
>>>
>>>> 3) Seems there are some problem with HIGHMEM :(. System crashes if I'm trying
>>>> to run hackbench (see log 1). I can run hackbench only if HIGHMEM is disabled.
>>>> I've also tried to disable CONFIG_CPU_SW_DOMAIN_PAN, CONFIG_ARM_PATCH_PHYS_VIRT,
>>>> and CONFIG_CPUSETS, before disabling HIGHMEM.
>>>> Sorry, I'll not able to investigate it further because of Christmas holidays here.
>>>
>>> I see. It seems not to trigger anything on x86.
>>
>> Yah. I saw other reports - it seems this issue is not triggered on ARM SMP also
>> (as per report from Tim Sander). Will try to dig deeper.
>>
>
> I did more tests, but still don't know what is wrong.
> 1) vanilla 4.4 - no issue
> 2) 4.4-rt: CONFIG_PREEMPT_RT_BASE=n and CONFIG_PREEMPT_RT_FULL=n -- no issue
> 3) 4.4-rt: CONFIG_DEBUG_HIGHMEM=y -- issue is present log1
> 4) 4.4-rt: this is interesting ;)
> I can run hackbench few times without issues If I run hackbench with RT priority (log2)
> But I got crash immediately once I run it using non-RT priorities.
Seems I've found root cause of the issue - FIX_KMAP_BEGIN offset need to be taken into
account in switch_kmaps() and below diff fixes this issue for me.
What I'm not fully understand is why I'm the only one who hit this issue :(
New functionality was added by commit:
commit a5f4c561b3b19a9bc43a81da6382b0098ebbc1fb
ARM: 8415/1: early fixmap support for earlycon
and it is enabled by default.
Luckiness :(
diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c
index deabc36..b9b4f9c 100644
--- a/arch/arm/mm/highmem.c
+++ b/arch/arm/mm/highmem.c
@@ -167,7 +167,7 @@ void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p)
* Clear @prev's kmap_atomic mappings
*/
for (i = 0; i < prev_p->kmap_idx; i++) {
- int idx = i + KM_TYPE_NR * smp_processor_id();
+ int idx = FIX_KMAP_BEGIN + i + KM_TYPE_NR * smp_processor_id();
set_fixmap_pte(idx, __pte(0));
}
@@ -175,7 +175,7 @@ void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p)
* Restore @next_p's kmap_atomic mappings
*/
for (i = 0; i < next_p->kmap_idx; i++) {
- int idx = i + KM_TYPE_NR * smp_processor_id();
+ int idx = FIX_KMAP_BEGIN + i + KM_TYPE_NR * smp_processor_id();
if (!pte_none(next_p->kmap_pte[i]))
set_fixmap_pte(idx, next_p->kmap_pte[i]);
--
regards,
-grygorii
^ permalink raw reply related [flat|nested] 16+ messages in thread* Re: [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report
2016-01-14 20:11 ` Grygorii Strashko
@ 2016-01-14 21:30 ` Russell King - ARM Linux
-1 siblings, 0 replies; 16+ messages in thread
From: Russell King - ARM Linux @ 2016-01-14 21:30 UTC (permalink / raw)
To: Grygorii Strashko
Cc: Sebastian Andrzej Siewior, Sekhar Nori, linux-rt-users,
linux-arm, linux-kernel
On Thu, Jan 14, 2016 at 10:11:09PM +0200, Grygorii Strashko wrote:
> diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c
> index deabc36..b9b4f9c 100644
> --- a/arch/arm/mm/highmem.c
> +++ b/arch/arm/mm/highmem.c
> @@ -167,7 +167,7 @@ void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p)
> * Clear @prev's kmap_atomic mappings
> */
> for (i = 0; i < prev_p->kmap_idx; i++) {
> - int idx = i + KM_TYPE_NR * smp_processor_id();
> + int idx = FIX_KMAP_BEGIN + i + KM_TYPE_NR * smp_processor_id();
>
> set_fixmap_pte(idx, __pte(0));
> }
> @@ -175,7 +175,7 @@ void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p)
> * Restore @next_p's kmap_atomic mappings
> */
> for (i = 0; i < next_p->kmap_idx; i++) {
> - int idx = i + KM_TYPE_NR * smp_processor_id();
> + int idx = FIX_KMAP_BEGIN + i + KM_TYPE_NR * smp_processor_id();
>
> if (!pte_none(next_p->kmap_pte[i]))
> set_fixmap_pte(idx, next_p->kmap_pte[i]);
This looks like it introduces the 4th and 5th copies of the same
calcuation, so can I suggest that we do this to prevent this kind
of error?
diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c
index d02f8187b1cc..61f0d5941116 100644
--- a/arch/arm/mm/highmem.c
+++ b/arch/arm/mm/highmem.c
@@ -34,6 +34,11 @@ static inline pte_t get_fixmap_pte(unsigned long vaddr)
return *ptep;
}
+static unsigned int fixmap_idx(int type)
+{
+ return FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id();
+}
+
void *kmap(struct page *page)
{
might_sleep();
@@ -79,7 +84,7 @@ void *kmap_atomic(struct page *page)
type = kmap_atomic_idx_push();
- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id();
+ idx = fixmap_idx(type);
vaddr = __fix_to_virt(idx);
#ifdef CONFIG_DEBUG_HIGHMEM
/*
@@ -106,7 +111,7 @@ void __kunmap_atomic(void *kvaddr)
if (kvaddr >= (void *)FIXADDR_START) {
type = kmap_atomic_idx();
- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id();
+ idx = fixmap_idx(type);
if (cache_is_vivt())
__cpuc_flush_dcache_area((void *)vaddr, PAGE_SIZE);
@@ -138,7 +143,7 @@ void *kmap_atomic_pfn(unsigned long pfn)
return page_address(page);
type = kmap_atomic_idx_push();
- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id();
+ idx = fixmap_idx(type);
vaddr = __fix_to_virt(idx);
#ifdef CONFIG_DEBUG_HIGHMEM
BUG_ON(!pte_none(get_fixmap_pte(vaddr)));
--
RMK's Patch system: http://www.arm.linux.org.uk/developer/patches/
FTTC broadband for 0.8mile line: currently at 9.6Mbps down 400kbps up
according to speedtest.net.
^ permalink raw reply related [flat|nested] 16+ messages in thread* [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report
@ 2016-01-14 21:30 ` Russell King - ARM Linux
0 siblings, 0 replies; 16+ messages in thread
From: Russell King - ARM Linux @ 2016-01-14 21:30 UTC (permalink / raw)
To: linux-arm-kernel
On Thu, Jan 14, 2016 at 10:11:09PM +0200, Grygorii Strashko wrote:
> diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c
> index deabc36..b9b4f9c 100644
> --- a/arch/arm/mm/highmem.c
> +++ b/arch/arm/mm/highmem.c
> @@ -167,7 +167,7 @@ void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p)
> * Clear @prev's kmap_atomic mappings
> */
> for (i = 0; i < prev_p->kmap_idx; i++) {
> - int idx = i + KM_TYPE_NR * smp_processor_id();
> + int idx = FIX_KMAP_BEGIN + i + KM_TYPE_NR * smp_processor_id();
>
> set_fixmap_pte(idx, __pte(0));
> }
> @@ -175,7 +175,7 @@ void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p)
> * Restore @next_p's kmap_atomic mappings
> */
> for (i = 0; i < next_p->kmap_idx; i++) {
> - int idx = i + KM_TYPE_NR * smp_processor_id();
> + int idx = FIX_KMAP_BEGIN + i + KM_TYPE_NR * smp_processor_id();
>
> if (!pte_none(next_p->kmap_pte[i]))
> set_fixmap_pte(idx, next_p->kmap_pte[i]);
This looks like it introduces the 4th and 5th copies of the same
calcuation, so can I suggest that we do this to prevent this kind
of error?
diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c
index d02f8187b1cc..61f0d5941116 100644
--- a/arch/arm/mm/highmem.c
+++ b/arch/arm/mm/highmem.c
@@ -34,6 +34,11 @@ static inline pte_t get_fixmap_pte(unsigned long vaddr)
return *ptep;
}
+static unsigned int fixmap_idx(int type)
+{
+ return FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id();
+}
+
void *kmap(struct page *page)
{
might_sleep();
@@ -79,7 +84,7 @@ void *kmap_atomic(struct page *page)
type = kmap_atomic_idx_push();
- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id();
+ idx = fixmap_idx(type);
vaddr = __fix_to_virt(idx);
#ifdef CONFIG_DEBUG_HIGHMEM
/*
@@ -106,7 +111,7 @@ void __kunmap_atomic(void *kvaddr)
if (kvaddr >= (void *)FIXADDR_START) {
type = kmap_atomic_idx();
- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id();
+ idx = fixmap_idx(type);
if (cache_is_vivt())
__cpuc_flush_dcache_area((void *)vaddr, PAGE_SIZE);
@@ -138,7 +143,7 @@ void *kmap_atomic_pfn(unsigned long pfn)
return page_address(page);
type = kmap_atomic_idx_push();
- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id();
+ idx = fixmap_idx(type);
vaddr = __fix_to_virt(idx);
#ifdef CONFIG_DEBUG_HIGHMEM
BUG_ON(!pte_none(get_fixmap_pte(vaddr)));
--
RMK's Patch system: http://www.arm.linux.org.uk/developer/patches/
FTTC broadband for 0.8mile line: currently at 9.6Mbps down 400kbps up
according to speedtest.net.
^ permalink raw reply related [flat|nested] 16+ messages in thread* Re: [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report
2016-01-14 21:30 ` Russell King - ARM Linux
@ 2016-01-15 11:35 ` Grygorii Strashko
-1 siblings, 0 replies; 16+ messages in thread
From: Grygorii Strashko @ 2016-01-15 11:35 UTC (permalink / raw)
To: Russell King - ARM Linux
Cc: Sebastian Andrzej Siewior, Sekhar Nori, linux-rt-users,
linux-arm, linux-kernel
Hi Russell,
On 01/14/2016 11:30 PM, Russell King - ARM Linux wrote:
> On Thu, Jan 14, 2016 at 10:11:09PM +0200, Grygorii Strashko wrote:
>> diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c
>> index deabc36..b9b4f9c 100644
>> --- a/arch/arm/mm/highmem.c
>> +++ b/arch/arm/mm/highmem.c
>> @@ -167,7 +167,7 @@ void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p)
>> * Clear @prev's kmap_atomic mappings
>> */
>> for (i = 0; i < prev_p->kmap_idx; i++) {
>> - int idx = i + KM_TYPE_NR * smp_processor_id();
>> + int idx = FIX_KMAP_BEGIN + i + KM_TYPE_NR * smp_processor_id();
>>
>> set_fixmap_pte(idx, __pte(0));
>> }
>> @@ -175,7 +175,7 @@ void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p)
>> * Restore @next_p's kmap_atomic mappings
>> */
>> for (i = 0; i < next_p->kmap_idx; i++) {
>> - int idx = i + KM_TYPE_NR * smp_processor_id();
>> + int idx = FIX_KMAP_BEGIN + i + KM_TYPE_NR * smp_processor_id();
>>
>> if (!pte_none(next_p->kmap_pte[i]))
>> set_fixmap_pte(idx, next_p->kmap_pte[i]);
>
> This looks like it introduces the 4th and 5th copies of the same
> calcuation, so can I suggest that we do this to prevent this kind
> of error?
>
> diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c
> index d02f8187b1cc..61f0d5941116 100644
> --- a/arch/arm/mm/highmem.c
> +++ b/arch/arm/mm/highmem.c
> @@ -34,6 +34,11 @@ static inline pte_t get_fixmap_pte(unsigned long vaddr)
> return *ptep;
> }
>
> +static unsigned int fixmap_idx(int type)
> +{
> + return FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id();
> +}
> +
This looks very reasonable - I'll updated and re-send.
Would you agree if I'll add your Signed-off-by: in final patch?
Thanks for review.
[...]
--
regards,
-grygorii
^ permalink raw reply [flat|nested] 16+ messages in thread* [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report
@ 2016-01-15 11:35 ` Grygorii Strashko
0 siblings, 0 replies; 16+ messages in thread
From: Grygorii Strashko @ 2016-01-15 11:35 UTC (permalink / raw)
To: linux-arm-kernel
Hi Russell,
On 01/14/2016 11:30 PM, Russell King - ARM Linux wrote:
> On Thu, Jan 14, 2016 at 10:11:09PM +0200, Grygorii Strashko wrote:
>> diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c
>> index deabc36..b9b4f9c 100644
>> --- a/arch/arm/mm/highmem.c
>> +++ b/arch/arm/mm/highmem.c
>> @@ -167,7 +167,7 @@ void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p)
>> * Clear @prev's kmap_atomic mappings
>> */
>> for (i = 0; i < prev_p->kmap_idx; i++) {
>> - int idx = i + KM_TYPE_NR * smp_processor_id();
>> + int idx = FIX_KMAP_BEGIN + i + KM_TYPE_NR * smp_processor_id();
>>
>> set_fixmap_pte(idx, __pte(0));
>> }
>> @@ -175,7 +175,7 @@ void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p)
>> * Restore @next_p's kmap_atomic mappings
>> */
>> for (i = 0; i < next_p->kmap_idx; i++) {
>> - int idx = i + KM_TYPE_NR * smp_processor_id();
>> + int idx = FIX_KMAP_BEGIN + i + KM_TYPE_NR * smp_processor_id();
>>
>> if (!pte_none(next_p->kmap_pte[i]))
>> set_fixmap_pte(idx, next_p->kmap_pte[i]);
>
> This looks like it introduces the 4th and 5th copies of the same
> calcuation, so can I suggest that we do this to prevent this kind
> of error?
>
> diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c
> index d02f8187b1cc..61f0d5941116 100644
> --- a/arch/arm/mm/highmem.c
> +++ b/arch/arm/mm/highmem.c
> @@ -34,6 +34,11 @@ static inline pte_t get_fixmap_pte(unsigned long vaddr)
> return *ptep;
> }
>
> +static unsigned int fixmap_idx(int type)
> +{
> + return FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id();
> +}
> +
This looks very reasonable - I'll updated and re-send.
Would you agree if I'll add your Signed-off-by: in final patch?
Thanks for review.
[...]
--
regards,
-grygorii
^ permalink raw reply [flat|nested] 16+ messages in thread* Re: [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report
2016-01-15 11:35 ` Grygorii Strashko
@ 2016-01-15 11:36 ` Russell King - ARM Linux
-1 siblings, 0 replies; 16+ messages in thread
From: Russell King - ARM Linux @ 2016-01-15 11:36 UTC (permalink / raw)
To: Grygorii Strashko
Cc: Sebastian Andrzej Siewior, Sekhar Nori, linux-rt-users,
linux-arm, linux-kernel
On Fri, Jan 15, 2016 at 01:35:33PM +0200, Grygorii Strashko wrote:
> This looks very reasonable - I'll updated and re-send.
> Would you agree if I'll add your Signed-off-by: in final patch?
Yes, that's fine. Please use:
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Thanks.
--
RMK's Patch system: http://www.arm.linux.org.uk/developer/patches/
FTTC broadband for 0.8mile line: currently at 9.6Mbps down 400kbps up
according to speedtest.net.
^ permalink raw reply [flat|nested] 16+ messages in thread
end of thread, other threads:[~2016-01-15 11:36 UTC | newest]
Thread overview: 16+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2015-12-28 18:00 [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report Grygorii Strashko
2015-12-28 18:00 ` [4.4-rc6-rt1 PATCH 1/2] kernel/sched/core.c: UP: fix 'implicit declaration of function 'update_migrate_disable' Grygorii Strashko
2015-12-28 18:00 ` Grygorii Strashko
2015-12-28 18:00 ` [4.4-rc6-rt1 PATCH 2/2] kernel/time/timer.c: UP: fix undefined reference to `del_timer_sync' Grygorii Strashko
2016-01-12 16:47 ` Sebastian Andrzej Siewior
2016-01-12 17:15 ` [4.4-rc6-rt1 PATCH 0/2] ARM: am437x: boot test report Sebastian Andrzej Siewior
2016-01-13 11:41 ` Grygorii Strashko
2016-01-13 20:14 ` Grygorii Strashko
2016-01-14 20:11 ` Grygorii Strashko
2016-01-14 20:11 ` Grygorii Strashko
2016-01-14 21:30 ` Russell King - ARM Linux
2016-01-14 21:30 ` Russell King - ARM Linux
2016-01-15 11:35 ` Grygorii Strashko
2016-01-15 11:35 ` Grygorii Strashko
2016-01-15 11:36 ` Russell King - ARM Linux
2016-01-15 11:36 ` Russell King - ARM Linux
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.