* IMX8M Nano compatibility
@ 2021-07-28 13:47 Kristiansen, Morten (INT)
2021-07-29 2:42 ` Chen, Hongzhan
2021-07-29 7:39 ` Jan Kiszka
0 siblings, 2 replies; 5+ messages in thread
From: Kristiansen, Morten (INT) @ 2021-07-28 13:47 UTC (permalink / raw)
To: xenomai
Hi Everybody.
I have a boot problem on IMX8M Nano.
NXP provide a ported kernel for their evaluation board based on Linux 5.4.70. (https://source.codeaurora.org/external/imx/linux-imx, branch imx_5.4.70_2.3.0)
I managed to modify apply the ipipe-core-5.4.72-arm64-0.patch to apply to kernel. It required a few modifications for irq-imx-gpcv2.c.
The following files files was removed during the NXP port:
drivers/gpu/ipu-v3/ipu-prv.h
drivers/gpu/ipu-v3/ipu-common.c
For reference, I’ve attached my revised patch.
The end result is the kernel mostly boots fine, but about 1 out of 5 it fails with the message below. Has anybody had any luck running Xenomai on the IMX8M series?
Any help is highly appreciated. 😊
/Morten Kristiansen
[ 12.672926] irq 103: nobody cared (try booting with the "irqpoll" option)
[ 12.680938] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.4.70-2.3.2+g8c73bc625c4d #1
[ 12.689246] Hardware name: NXP i.MX8MNano EVK board (DT)
[ 12.695120] I-pipe domain: Linux
[ 12.698889] Call trace:
[ 12.702267] dump_backtrace+0x0/0x140
[ 12.706694] show_stack+0x14/0x20
[ 12.711023] dump_stack+0xe4/0x12c
[ 12.715277] __report_bad_irq+0x48/0xd8
[ 12.719873] note_interrupt+0x2c4/0x37c
[ 12.724448] handle_irq_event_percpu+0x80/0x88
[ 12.729604] handle_irq_event+0x44/0xd0
[ 12.734249] handle_level_irq+0x90/0xe8
[ 12.739058] generic_handle_irq+0x24/0x38
[ 12.743762] __handle_domain_irq+0x60/0xb8
[ 12.748789] __ipipe_do_IRQ+0x28/0x30
[ 12.753419] __ipipe_do_sync_stage+0x198/0x1c8
[ 12.758628] ipipe_unstall_root+0x40/0x50
[ 12.763370] __do_softirq+0xa4/0x248
[ 12.767896] irq_exit+0xb8/0xd8
[ 12.771730] __handle_domain_irq+0x64/0xb8
[ 12.776598] __ipipe_do_IRQ+0x28/0x30
[ 12.781055] __ipipe_do_sync_stage+0x198/0x1c8
[ 12.786295] ipipe_unstall_root+0x40/0x50
[ 12.791263] cpuidle_enter_state+0xa4/0x330
[ 12.796213] cpuidle_enter+0x34/0x48
[ 12.800673] call_cpuidle+0x18/0x38
[ 12.804939] do_idle+0x1f0/0x288
[ 12.808940] cpu_startup_entry+0x20/0x50
[ 12.813665] rest_init+0xd4/0xe0
[ 12.817842] arch_call_rest_init+0xc/0x14
[ 12.822659] start_kernel+0x410/0x444
[ 12.826855] handlers:
[ 12.829882] [<00000000ab81b488>] irq_default_primary_handler threaded [<000000002543e636>] _tcpci_irq
[ 12.840309] Disabling IRQ #103
[ 24.092075] rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
[ 24.098209] rcu: 0-...0: (5 ticks this GP) idle=b8e/0/0x3 softirq=80/80 fqs=2621
[ 24.105801] (detected by 2, t=5252 jiffies, g=-959, q=16)
[ 24.111304] Task dump for CPU 0:
[ 24.114554] swapper/0 R running task 0 0 0 0x0000002a
[ 24.121611] Call trace:
[ 24.124099] __switch_to+0x108/0x1c0
[ 24.127695] 0x18
Teledyne Confidential; Commercially Sensitive Business Data
-------------- next part --------------
A non-text attachment was scrubbed...
Name: imx-ipipe-core-5.4.72-arm64-0.patch
Type: application/octet-stream
Size: 515901 bytes
Desc: imx-ipipe-core-5.4.72-arm64-0.patch
URL: <http://xenomai.org/pipermail/xenomai/attachments/20210728/4fde70e5/attachment.obj>
^ permalink raw reply [flat|nested] 5+ messages in thread
* RE: IMX8M Nano compatibility
2021-07-28 13:47 IMX8M Nano compatibility Kristiansen, Morten (INT)
@ 2021-07-29 2:42 ` Chen, Hongzhan
2021-07-29 7:39 ` Jan Kiszka
1 sibling, 0 replies; 5+ messages in thread
From: Chen, Hongzhan @ 2021-07-29 2:42 UTC (permalink / raw)
To: Kristiansen, Morten (INT), xenomai
>-----Original Message-----
>From: Xenomai <xenomai-bounces@xenomai.org> On Behalf Of Kristiansen, Morten (INT) via Xenomai
>Sent: Wednesday, July 28, 2021 9:47 PM
>To: xenomai@xenomai.org
>Subject: IMX8M Nano compatibility
>
>Hi Everybody.
>
>I have a boot problem on IMX8M Nano.
>
>NXP provide a ported kernel for their evaluation board based on Linux 5.4.70. (https://source.codeaurora.org/external/imx/linux-imx, branch imx_5.4.70_2.3.0)
>
>I managed to modify apply the ipipe-core-5.4.72-arm64-0.patch to apply to kernel. It required a few modifications for irq-imx-gpcv2.c.
>
>The following files files was removed during the NXP port:
>drivers/gpu/ipu-v3/ipu-prv.h
>drivers/gpu/ipu-v3/ipu-common.c
>
>For reference, I’ve attached my revised patch.
>
>The end result is the kernel mostly boots fine, but about 1 out of 5 it fails with the message below. Has anybody had any luck running Xenomai on the IMX8M series?
>
>Any help is highly appreciated. ??
>
>/Morten Kristiansen
>
>
>
>[ 12.672926] irq 103: nobody cared (try booting with the "irqpoll" option)
>[ 12.680938] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.4.70-2.3.2+g8c73bc625c4d #1
>[ 12.689246] Hardware name: NXP i.MX8MNano EVK board (DT)
>[ 12.695120] I-pipe domain: Linux
>[ 12.698889] Call trace:
>[ 12.702267] dump_backtrace+0x0/0x140
>[ 12.706694] show_stack+0x14/0x20
>[ 12.711023] dump_stack+0xe4/0x12c
>[ 12.715277] __report_bad_irq+0x48/0xd8
>[ 12.719873] note_interrupt+0x2c4/0x37c
>[ 12.724448] handle_irq_event_percpu+0x80/0x88
>[ 12.729604] handle_irq_event+0x44/0xd0
>[ 12.734249] handle_level_irq+0x90/0xe8
>[ 12.739058] generic_handle_irq+0x24/0x38
>[ 12.743762] __handle_domain_irq+0x60/0xb8
>[ 12.748789] __ipipe_do_IRQ+0x28/0x30
>[ 12.753419] __ipipe_do_sync_stage+0x198/0x1c8
>[ 12.758628] ipipe_unstall_root+0x40/0x50
>[ 12.763370] __do_softirq+0xa4/0x248
>[ 12.767896] irq_exit+0xb8/0xd8
>[ 12.771730] __handle_domain_irq+0x64/0xb8
>[ 12.776598] __ipipe_do_IRQ+0x28/0x30
>[ 12.781055] __ipipe_do_sync_stage+0x198/0x1c8
>[ 12.786295] ipipe_unstall_root+0x40/0x50
>[ 12.791263] cpuidle_enter_state+0xa4/0x330
>[ 12.796213] cpuidle_enter+0x34/0x48
>[ 12.800673] call_cpuidle+0x18/0x38
>[ 12.804939] do_idle+0x1f0/0x288
>[ 12.808940] cpu_startup_entry+0x20/0x50
>[ 12.813665] rest_init+0xd4/0xe0
>[ 12.817842] arch_call_rest_init+0xc/0x14
>[ 12.822659] start_kernel+0x410/0x444
>[ 12.826855] handlers:
>[ 12.829882] [<00000000ab81b488>] irq_default_primary_handler threaded [<000000002543e636>] _tcpci_irq
>[ 12.840309] Disabling IRQ #103
>[ 24.092075] rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
>[ 24.098209] rcu: 0-...0: (5 ticks this GP) idle=b8e/0/0x3 softirq=80/80 fqs=2621
>[ 24.105801] (detected by 2, t=5252 jiffies, g=-959, q=16)
>[ 24.111304] Task dump for CPU 0:
>[ 24.114554] swapper/0 R running task 0 0 0 0x0000002a
>[ 24.121611] Call trace:
>[ 24.124099] __switch_to+0x108/0x1c0
>[ 24.127695] 0x18
>
>
>Teledyne Confidential; Commercially Sensitive Business Data
>-------------- next part --------------
>A non-text attachment was scrubbed...
>Name: imx-ipipe-core-5.4.72-arm64-0.patch
>Type: application/octet-stream
>Size: 515901 bytes
>Desc: imx-ipipe-core-5.4.72-arm64-0.patch
>URL: <http://xenomai.org/pipermail/xenomai/attachments/20210728/4fde70e5/attachment.obj>
>-static void imx_gpcv2_irq_unmask(struct irq_data *d)
>+static void __imx_gpcv2_irq_unmask(struct irq_data *d)
> {
> struct gpcv2_irqchip_data *cd = d->chip_data;
>+ unsigned long flags;
> void __iomem *reg;
> u32 val;
>
>- raw_spin_lock(&cd->rlock);
> reg = gpcv2_idx_to_reg(cd, d->hwirq / 32);
> val = readl_relaxed(reg);
> val &= ~BIT(d->hwirq % 32);
> writel_relaxed(val, reg);
>- raw_spin_unlock(&cd->rlock);
>
>+ raw_spin_lock_irqsave(&cd->rlock, flags);
>+ __imx_gpcv2_irq_unmask(d);
Why it need recursive call here?
Regards
Hongzhan Chen
>+ raw_spin_unlock_irqrestore(&cd->rlock, flags);
> irq_chip_unmask_parent(d);
> }
^ permalink raw reply [flat|nested] 5+ messages in thread* Re: IMX8M Nano compatibility
2021-07-28 13:47 IMX8M Nano compatibility Kristiansen, Morten (INT)
2021-07-29 2:42 ` Chen, Hongzhan
@ 2021-07-29 7:39 ` Jan Kiszka
2021-07-29 8:48 ` Peng Fan
1 sibling, 1 reply; 5+ messages in thread
From: Jan Kiszka @ 2021-07-29 7:39 UTC (permalink / raw)
To: Kristiansen, Morten (INT), xenomai
On 28.07.21 15:47, Kristiansen, Morten (INT) via Xenomai wrote:
> Hi Everybody.
>
> I have a boot problem on IMX8M Nano.
>
> NXP provide a ported kernel for their evaluation board based on Linux 5.4.70. (https://source.codeaurora.org/external/imx/linux-imx, branch imx_5.4.70_2.3.0)
>
You generally don't want to use this, you want an upstream version.
> I managed to modify apply the ipipe-core-5.4.72-arm64-0.patch to apply to kernel. It required a few modifications for irq-imx-gpcv2.c.
>
See https://xenomai.org/pipermail/xenomai/2021-July/045793.html.
Jan
> The following files files was removed during the NXP port:
> drivers/gpu/ipu-v3/ipu-prv.h
> drivers/gpu/ipu-v3/ipu-common.c
>
> For reference, I’ve attached my revised patch.
>
> The end result is the kernel mostly boots fine, but about 1 out of 5 it fails with the message below. Has anybody had any luck running Xenomai on the IMX8M series?
>
> Any help is highly appreciated. 😊
>
> /Morten Kristiansen
>
>
>
> [ 12.672926] irq 103: nobody cared (try booting with the "irqpoll" option)
> [ 12.680938] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.4.70-2.3.2+g8c73bc625c4d #1
> [ 12.689246] Hardware name: NXP i.MX8MNano EVK board (DT)
> [ 12.695120] I-pipe domain: Linux
> [ 12.698889] Call trace:
> [ 12.702267] dump_backtrace+0x0/0x140
> [ 12.706694] show_stack+0x14/0x20
> [ 12.711023] dump_stack+0xe4/0x12c
> [ 12.715277] __report_bad_irq+0x48/0xd8
> [ 12.719873] note_interrupt+0x2c4/0x37c
> [ 12.724448] handle_irq_event_percpu+0x80/0x88
> [ 12.729604] handle_irq_event+0x44/0xd0
> [ 12.734249] handle_level_irq+0x90/0xe8
> [ 12.739058] generic_handle_irq+0x24/0x38
> [ 12.743762] __handle_domain_irq+0x60/0xb8
> [ 12.748789] __ipipe_do_IRQ+0x28/0x30
> [ 12.753419] __ipipe_do_sync_stage+0x198/0x1c8
> [ 12.758628] ipipe_unstall_root+0x40/0x50
> [ 12.763370] __do_softirq+0xa4/0x248
> [ 12.767896] irq_exit+0xb8/0xd8
> [ 12.771730] __handle_domain_irq+0x64/0xb8
> [ 12.776598] __ipipe_do_IRQ+0x28/0x30
> [ 12.781055] __ipipe_do_sync_stage+0x198/0x1c8
> [ 12.786295] ipipe_unstall_root+0x40/0x50
> [ 12.791263] cpuidle_enter_state+0xa4/0x330
> [ 12.796213] cpuidle_enter+0x34/0x48
> [ 12.800673] call_cpuidle+0x18/0x38
> [ 12.804939] do_idle+0x1f0/0x288
> [ 12.808940] cpu_startup_entry+0x20/0x50
> [ 12.813665] rest_init+0xd4/0xe0
> [ 12.817842] arch_call_rest_init+0xc/0x14
> [ 12.822659] start_kernel+0x410/0x444
> [ 12.826855] handlers:
> [ 12.829882] [<00000000ab81b488>] irq_default_primary_handler threaded [<000000002543e636>] _tcpci_irq
> [ 12.840309] Disabling IRQ #103
> [ 24.092075] rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
> [ 24.098209] rcu: 0-...0: (5 ticks this GP) idle=b8e/0/0x3 softirq=80/80 fqs=2621
> [ 24.105801] (detected by 2, t=5252 jiffies, g=-959, q=16)
> [ 24.111304] Task dump for CPU 0:
> [ 24.114554] swapper/0 R running task 0 0 0 0x0000002a
> [ 24.121611] Call trace:
> [ 24.124099] __switch_to+0x108/0x1c0
> [ 24.127695] 0x18
>
>
> Teledyne Confidential; Commercially Sensitive Business Data
> -------------- next part --------------
> A non-text attachment was scrubbed...
> Name: imx-ipipe-core-5.4.72-arm64-0.patch
> Type: application/octet-stream
> Size: 515901 bytes
> Desc: imx-ipipe-core-5.4.72-arm64-0.patch
> URL: <http://xenomai.org/pipermail/xenomai/attachments/20210728/4fde70e5/attachment.obj>
>
--
Siemens AG, T RDA IOT
Corporate Competence Center Embedded Linux
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: IMX8M Nano compatibility
2021-07-29 7:39 ` Jan Kiszka
@ 2021-07-29 8:48 ` Peng Fan
2021-07-29 9:12 ` Jan Kiszka
0 siblings, 1 reply; 5+ messages in thread
From: Peng Fan @ 2021-07-29 8:48 UTC (permalink / raw)
To: Jan Kiszka; +Cc: Kristiansen, Morten (INT), xenomai
Jan Kiszka via Xenomai <xenomai@xenomai.org> 于2021年7月29日周四 下午3:39写道：
> On 28.07.21 15:47, Kristiansen, Morten (INT) via Xenomai wrote:
> > Hi Everybody.
> >
> > I have a boot problem on IMX8M Nano.
> >
> > NXP provide a ported kernel for their evaluation board based on Linux
> 5.4.70. (https://source.codeaurora.org/external/imx/linux-imx, branch
> imx_5.4.70_2.3.0)
> >
>
> You generally don't want to use this, you want an upstream version.
>
Jan,
Do you see an issue with that version?
THanks,
Peng.
>
> > I managed to modify apply the ipipe-core-5.4.72-arm64-0.patch to apply
> to kernel. It required a few modifications for irq-imx-gpcv2.c.
> >
>
> See https://xenomai.org/pipermail/xenomai/2021-July/045793.html.
>
> Jan
>
> > The following files files was removed during the NXP port:
> > drivers/gpu/ipu-v3/ipu-prv.h
> > drivers/gpu/ipu-v3/ipu-common.c
> >
> > For reference, I’ve attached my revised patch.
> >
> > The end result is the kernel mostly boots fine, but about 1 out of 5 it
> fails with the message below. Has anybody had any luck running Xenomai on
> the IMX8M series?
> >
> > Any help is highly appreciated. 😊
> >
> > /Morten Kristiansen
> >
> >
> >
> > [ 12.672926] irq 103: nobody cared (try booting with the "irqpoll"
> option)
> > [ 12.680938] CPU: 0 PID: 0 Comm: swapper/0 Not tainted
> 5.4.70-2.3.2+g8c73bc625c4d #1
> > [ 12.689246] Hardware name: NXP i.MX8MNano EVK board (DT)
> > [ 12.695120] I-pipe domain: Linux
> > [ 12.698889] Call trace:
> > [ 12.702267] dump_backtrace+0x0/0x140
> > [ 12.706694] show_stack+0x14/0x20
> > [ 12.711023] dump_stack+0xe4/0x12c
> > [ 12.715277] __report_bad_irq+0x48/0xd8
> > [ 12.719873] note_interrupt+0x2c4/0x37c
> > [ 12.724448] handle_irq_event_percpu+0x80/0x88
> > [ 12.729604] handle_irq_event+0x44/0xd0
> > [ 12.734249] handle_level_irq+0x90/0xe8
> > [ 12.739058] generic_handle_irq+0x24/0x38
> > [ 12.743762] __handle_domain_irq+0x60/0xb8
> > [ 12.748789] __ipipe_do_IRQ+0x28/0x30
> > [ 12.753419] __ipipe_do_sync_stage+0x198/0x1c8
> > [ 12.758628] ipipe_unstall_root+0x40/0x50
> > [ 12.763370] __do_softirq+0xa4/0x248
> > [ 12.767896] irq_exit+0xb8/0xd8
> > [ 12.771730] __handle_domain_irq+0x64/0xb8
> > [ 12.776598] __ipipe_do_IRQ+0x28/0x30
> > [ 12.781055] __ipipe_do_sync_stage+0x198/0x1c8
> > [ 12.786295] ipipe_unstall_root+0x40/0x50
> > [ 12.791263] cpuidle_enter_state+0xa4/0x330
> > [ 12.796213] cpuidle_enter+0x34/0x48
> > [ 12.800673] call_cpuidle+0x18/0x38
> > [ 12.804939] do_idle+0x1f0/0x288
> > [ 12.808940] cpu_startup_entry+0x20/0x50
> > [ 12.813665] rest_init+0xd4/0xe0
> > [ 12.817842] arch_call_rest_init+0xc/0x14
> > [ 12.822659] start_kernel+0x410/0x444
> > [ 12.826855] handlers:
> > [ 12.829882] [<00000000ab81b488>] irq_default_primary_handler threaded
> [<000000002543e636>] _tcpci_irq
> > [ 12.840309] Disabling IRQ #103
> > [ 24.092075] rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
> > [ 24.098209] rcu: 0-...0: (5 ticks this GP) idle=b8e/0/0x3
> softirq=80/80 fqs=2621
> > [ 24.105801] (detected by 2, t=5252 jiffies, g=-959, q=16)
> > [ 24.111304] Task dump for CPU 0:
> > [ 24.114554] swapper/0 R running task 0 0 0
> 0x0000002a
> > [ 24.121611] Call trace:
> > [ 24.124099] __switch_to+0x108/0x1c0
> > [ 24.127695] 0x18
> >
> >
> > Teledyne Confidential; Commercially Sensitive Business Data
> > -------------- next part --------------
> > A non-text attachment was scrubbed...
> > Name: imx-ipipe-core-5.4.72-arm64-0.patch
> > Type: application/octet-stream
> > Size: 515901 bytes
> > Desc: imx-ipipe-core-5.4.72-arm64-0.patch
> > URL: <
> http://xenomai.org/pipermail/xenomai/attachments/20210728/4fde70e5/attachment.obj
> >
> >
>
> --
> Siemens AG, T RDA IOT
> Corporate Competence Center Embedded Linux
>
>
--
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: IMX8M Nano compatibility
2021-07-29 8:48 ` Peng Fan
@ 2021-07-29 9:12 ` Jan Kiszka
0 siblings, 0 replies; 5+ messages in thread
From: Jan Kiszka @ 2021-07-29 9:12 UTC (permalink / raw)
To: Peng Fan; +Cc: Kristiansen, Morten (INT), xenomai
On 29.07.21 10:48, Peng Fan wrote:
> Jan Kiszka via Xenomai <xenomai@xenomai.org
> <mailto:xenomai@xenomai.org>> 于2021年7月29日周四 下午3:39写道：
>
> On 28.07.21 15:47, Kristiansen, Morten (INT) via Xenomai wrote:
> > Hi Everybody.
> >
> > I have a boot problem on IMX8M Nano.
> >
> > NXP provide a ported kernel for their evaluation board based on
> Linux 5.4.70. (https://source.codeaurora.org/external/imx/linux-imx
> <https://source.codeaurora.org/external/imx/linux-imx>,
> branch imx_5.4.70_2.3.0)
> >
>
> You generally don't want to use this, you want an upstream version.
>
>
> Jan,
>
> Do you see an issue with that version?
It's not upstream, thus not able to provide official stable interfaces.
And it didn't pass the extra upstream QA yet. And my not live long
enough to be used in products. The common vendor kernel problem.
Plus, when you combine things with another downstream patch series -
that's what Xenomai is responsible for -, things do not become simpler.
Sorry, just burned again, not by NXP, but by the same pattern.
Jan
--
Siemens AG, T RDA IOT
Corporate Competence Center Embedded Linux
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2021-07-29 9:12 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-07-28 13:47 IMX8M Nano compatibility Kristiansen, Morten (INT)
2021-07-29 2:42 ` Chen, Hongzhan
2021-07-29 7:39 ` Jan Kiszka
2021-07-29 8:48 ` Peng Fan
2021-07-29 9:12 ` Jan Kiszka
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.