3.1-rc3 fails to boot on my pandaboard

3.1-rc3 fails to boot on my pandaboard

[Luciano Coelho]

Hi,

There seems to be a problem with 3.1-rc3, which prevents it from booting
on my PandaBoard (ARM OMAP4).

3.1-rc2 works fine, so I have bisected the problem and it turns out to
be caused by this:

commit f3637a5f2e2eb391ff5757bc83fb5de8f9726464
Author: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Date:   Thu Jul 7 22:32:17 2011 +0200

    irq: Always set IRQF_ONESHOT if no primary handler is specified

After reverting this commit, -rc3 boots fine.  I'm not sure where
exactly the problem is.  Maybe it's something in the omap code which is
broken and this patch causes the bug to show up?

What happens is that the booting process hangs after printing out this:

[...] 
[    7.408477] EXT3-fs: barriers not enabled
[    7.415008] kjournald starting.  Commit interval 5 seconds
[    7.446228] EXT3-fs (mmcblk0p5): using internal journal
[    7.452087] EXT3-fs (mmcblk0p5): mounted filesystem with ordered data mode
[    7.459625] VFS: Mounted root (ext3 filesystem) on device 179:5.
[    7.466094] Freeing init memory: 312K
[    7.471710] Failed to execute /init.  Attempting defaults...
[   10.254333] EXT3-fs (mmcblk0p5): using internal journal

The full boot logs can be found here:

http://pastebin.pandaboard.org/index.php/view/88995306

Any ideas what may be going wrong?

-- 
Cheers,
Luca.

Re: 3.1-rc3 fails to boot on my pandaboard

[Sebastian Andrzej Siewior]

* Luciano Coelho | 2011-08-25 13:39:29 [+0300]:

>Hi,
Hi,

>Any ideas what may be going wrong?

I think my commit was identified as bogus and tglx is going to revert 
it. The problem is that I force ONESHOT mode for all threaded IRQs but
there are also others without the flag which is not allowed.

I was trying to check something but I don't get my board to boot. Could
you please try to boot with the patch reverted and paste me the
the output of 
 cat /proc/interrupts | grep 74 ?

I *think* that the flow handler is level (I can't find evidence of it
beeing edge, and omap_alloc_gc() is the place installing it).

So could you please gather additional debug info with this patch?

diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c
index 2e94258..eda25a0 100644
--- a/kernel/irq/manage.c
+++ b/kernel/irq/manage.c
@@ -1327,6 +1327,9 @@ int request_threaded_irq(unsigned int irq, irq_handler_t handler,
 	if (!irq_settings_can_request(desc))
 		return -EINVAL;
 
+	if (irq == 74)
+		printk(KERN_ERR "%s() h %pS th %pS f %lx dev %s flow %pS\n", __func__,
+				handler, thread_fn, irqflags, devname, desc->handle_irq);
 	if (!handler) {
 		if (!thread_fn)
 			return -EINVAL;


%pS should resolve the function names.

Sebastian

Re: 3.1-rc3 fails to boot on my pandaboard

[Luciano Coelho]

On Thu, 2011-08-25 at 13:04 +0200, Sebastian Andrzej Siewior wrote: 
> * Luciano Coelho | 2011-08-25 13:39:29 [+0300]:
> >Any ideas what may be going wrong?
> 
> I think my commit was identified as bogus and tglx is going to revert 
> it. The problem is that I force ONESHOT mode for all threaded IRQs but
> there are also others without the flag which is not allowed.
> 
> I was trying to check something but I don't get my board to boot. Could
> you please try to boot with the patch reverted and paste me the
> the output of 
>  cat /proc/interrupts | grep 74 ?

Hmmm... There doesn't seem to be an irq 74.  The only thing I get is
this:

root@xiru:~# cat /proc/interrupts | grep 74
IPI1:       2667       3743  Rescheduling interrupts

Which is probably not what you wanted. ;)


> I *think* that the flow handler is level (I can't find evidence of it
> beeing edge, and omap_alloc_gc() is the place installing it).
> 
> So could you please gather additional debug info with this patch?
> 
> diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c
> index 2e94258..eda25a0 100644
> --- a/kernel/irq/manage.c
> +++ b/kernel/irq/manage.c
> @@ -1327,6 +1327,9 @@ int request_threaded_irq(unsigned int irq, irq_handler_t handler,
>  	if (!irq_settings_can_request(desc))
>  		return -EINVAL;
>  
> +	if (irq == 74)
> +		printk(KERN_ERR "%s() h %pS th %pS f %lx dev %s flow %pS\n", __func__,
> +				handler, thread_fn, irqflags, devname, desc->handle_irq);
>  	if (!handler) {
>  		if (!thread_fn)
>  			return -EINVAL;
> 
> 
> %pS should resolve the function names.

As I mentioned above, there doesn't seem to be irq 74, so this code
doesn't hit.

These are the interrupts reported in /proc/interrupts on my board:

root@xiru:~# cat /proc/interrupts
           CPU0       CPU1       
39:          0          0       GIC  TWL6030-PIH
41:          0          0       GIC  l3-dbg-irq
42:          0          0       GIC  l3-app-irq
44:       5397          0       GIC  DMA
52:          0          0       GIC  gpmc
69:         16          0       GIC  gp timer
88:        270          0       GIC  omap_i2c
89:          0          0       GIC  omap_i2c
91:        183          0       GIC  mmc1
93:          0          0       GIC  omap_i2c
94:          0          0       GIC  omap_i2c
102:          0          0       GIC  serial idle
104:          0          0       GIC  serial idle
105:          0          0       GIC  serial idle
106:        290          0       GIC  serial idle, OMAP UART2
108:          1          0       GIC  ohci_hcd:usb3
109:        493          0       GIC  ehci_hcd:usb1
115:       6260          0       GIC  mmc0
124:          1          0       GIC  musb-hdrc
125:          0          0       GIC  musb-hdrc
213:          0          0      GPIO  wl1271
372:          0          0   twl6030  twl6030_usb
378:          0          0   twl6030  twl6030_usb
379:          0          0   twl6030  rtc0
IPI0:          0          0  Timer broadcast interrupts
IPI1:       2995       3606  Rescheduling interrupts
IPI2:          0          0  Function call interrupts
IPI3:         57         61  Single function call interrupts
IPI4:          0          0  CPU stop interrupts
LOC:       8879       5844  Local timer interrupts
Err:          0


-- 
Cheers,
Luca.

Re: 3.1-rc3 fails to boot on my pandaboard

[Sebastian Andrzej Siewior]

* Luciano Coelho | 2011-08-25 14:24:34 [+0300]:

>These are the interrupts reported in /proc/interrupts on my board:

Thank you, this is enough information.

Sebastian

Re: 3.1-rc3 fails to boot on my pandaboard

[Govindraj]

On Thu, Aug 25, 2011 at 4:09 PM, Luciano Coelho <coelho@ti.com> wrote:
> Hi,
>
> There seems to be a problem with 3.1-rc3, which prevents it from booting
> on my PandaBoard (ARM OMAP4).
>
> 3.1-rc2 works fine, so I have bisected the problem and it turns out to
> be caused by this:
>
> commit f3637a5f2e2eb391ff5757bc83fb5de8f9726464
> Author: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
> Date:   Thu Jul 7 22:32:17 2011 +0200
>
>    irq: Always set IRQF_ONESHOT if no primary handler is specified
>
> After reverting this commit, -rc3 boots fine.  I'm not sure where
> exactly the problem is.  Maybe it's something in the omap code which is
> broken and this patch causes the bug to show up?
>
> What happens is that the booting process hangs after printing out this:
>

fyi,

http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=69dd3d8e29e294caaf63eb5e8a72d250279f9e5f;hp=fcb8ce5cfe30ca9ca5c9a79cdfe26d1993e65e0c

--
Thanks,
Govindraj.R


> [...]
> [    7.408477] EXT3-fs: barriers not enabled
> [    7.415008] kjournald starting.  Commit interval 5 seconds
> [    7.446228] EXT3-fs (mmcblk0p5): using internal journal
> [    7.452087] EXT3-fs (mmcblk0p5): mounted filesystem with ordered data mode
> [    7.459625] VFS: Mounted root (ext3 filesystem) on device 179:5.
> [    7.466094] Freeing init memory: 312K
> [    7.471710] Failed to execute /init.  Attempting defaults...
> [   10.254333] EXT3-fs (mmcblk0p5): using internal journal
>
> The full boot logs can be found here:
>
> http://pastebin.pandaboard.org/index.php/view/88995306
>
> Any ideas what may be going wrong?
>
> --
> Cheers,
> Luca.
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-omap" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>

Re: 3.1-rc3 fails to boot on my pandaboard

[Luciano Coelho]

On Thu, 2011-08-25 at 17:05 +0530, Govindraj wrote: 
> On Thu, Aug 25, 2011 at 4:09 PM, Luciano Coelho <coelho@ti.com> wrote:
> > Hi,
> >
> > There seems to be a problem with 3.1-rc3, which prevents it from booting
> > on my PandaBoard (ARM OMAP4).
> >
> > 3.1-rc2 works fine, so I have bisected the problem and it turns out to
> > be caused by this:
> >
> > commit f3637a5f2e2eb391ff5757bc83fb5de8f9726464
> > Author: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
> > Date:   Thu Jul 7 22:32:17 2011 +0200
> >
> >    irq: Always set IRQF_ONESHOT if no primary handler is specified
> >
> > After reverting this commit, -rc3 boots fine.  I'm not sure where
> > exactly the problem is.  Maybe it's something in the omap code which is
> > broken and this patch causes the bug to show up?
> >
> > What happens is that the booting process hangs after printing out this:
> >
> 
> fyi,
> 
> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=69dd3d8e29e294caaf63eb5e8a72d250279f9e5f;hp=fcb8ce5cfe30ca9ca5c9a79cdfe26d1993e65e0c

Ah, thanks.  Now I see where the irq == 74 comes from. :) Still, I don't
see that on my system.

-- 
Cheers,
Luca.

Re: 3.1-rc3 fails to boot on my pandaboard

[Govindraj]

On Thu, Aug 25, 2011 at 5:09 PM, Luciano Coelho <coelho@ti.com> wrote:
> On Thu, 2011-08-25 at 17:05 +0530, Govindraj wrote:
>> On Thu, Aug 25, 2011 at 4:09 PM, Luciano Coelho <coelho@ti.com> wrote:
>> > Hi,
>> >
>> > There seems to be a problem with 3.1-rc3, which prevents it from booting
>> > on my PandaBoard (ARM OMAP4).
>> >
>> > 3.1-rc2 works fine, so I have bisected the problem and it turns out to
>> > be caused by this:
>> >
>> > commit f3637a5f2e2eb391ff5757bc83fb5de8f9726464
>> > Author: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
>> > Date:   Thu Jul 7 22:32:17 2011 +0200
>> >
>> >    irq: Always set IRQF_ONESHOT if no primary handler is specified
>> >
>> > After reverting this commit, -rc3 boots fine.  I'm not sure where
>> > exactly the problem is.  Maybe it's something in the omap code which is
>> > broken and this patch causes the bug to show up?
>> >
>> > What happens is that the booting process hangs after printing out this:
>> >
>>
>> fyi,
>>
>> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=69dd3d8e29e294caaf63eb5e8a72d250279f9e5f;hp=fcb8ce5cfe30ca9ca5c9a79cdfe26d1993e65e0c
>
> Ah, thanks.  Now I see where the irq == 74 comes from. :) Still, I don't
> see that on my system.
>

69dd3d8 Revert "irq: Always set IRQF_ONESHOT if no primary handler is specified"
f2b6071 drm/radeon: Extended DDC Probing for Toshiba L300D Radeon
Mobility X1100 HDMI-A Connector
c87019e HID: wiimote: Add status and return request handlers
2cb5e4b HID: wiimote: Add drm request
23a5a4a HID: wiimote: Register led class devices
26af174 HID: wiimote: Correctly call HID open/close callbacks
3989ef6 HID: wiimote: Simplify synchronization
8d3bb23 drm/ttm: ensure ttm for new node is bound before calling move_notify()
eac2095 drm/ttm: unbind ttm before destroying node in accel move cleanup
7c4c396 drm/ttm: fix ttm_bo_add_ttm(user) failure path
f5e4282 HID: usbhid: Add support for SiGma Micro chip
11f3a6b bridge: fix a possible net_device leak
b6bede3 xfs: fix tracing builds inside the source tree
832d80a Merge branch 'for-davem' of
git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless
fcb8ce5 Linux 3.1-rc3


--
Thanks,
Govindraj.R

RE: 3.1-rc3 fails to boot on my pandaboard

[Maupin, Chase]

> -----Original Message-----
> From: linux-omap-owner@vger.kernel.org [mailto:linux-omap-
> owner@vger.kernel.org] On Behalf Of Govindraj
> Sent: Thursday, August 25, 2011 6:46 AM
> To: Coelho, Luciano
> Cc: linux-kernel@vger.kernel.org; linux-omap@vger.kernel.org; Thomas
> Gleixner; Sebastian Andrzej Siewior
> Subject: Re: 3.1-rc3 fails to boot on my pandaboard
> 
> On Thu, Aug 25, 2011 at 5:09 PM, Luciano Coelho <coelho@ti.com> wrote:
> > On Thu, 2011-08-25 at 17:05 +0530, Govindraj wrote:
> >> On Thu, Aug 25, 2011 at 4:09 PM, Luciano Coelho <coelho@ti.com>
> wrote:
> >> > Hi,
> >> >
> >> > There seems to be a problem with 3.1-rc3, which prevents it from
> booting
> >> > on my PandaBoard (ARM OMAP4).
> >> >
> >> > 3.1-rc2 works fine, so I have bisected the problem and it turns
> out to
> >> > be caused by this:
> >> >
> >> > commit f3637a5f2e2eb391ff5757bc83fb5de8f9726464
> >> > Author: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
> >> > Date:   Thu Jul 7 22:32:17 2011 +0200
> >> >
> >> >    irq: Always set IRQF_ONESHOT if no primary handler is specified
> >> >
> >> > After reverting this commit, -rc3 boots fine.  I'm not sure where
> >> > exactly the problem is.  Maybe it's something in the omap code
> which is
> >> > broken and this patch causes the bug to show up?
> >> >
> >> > What happens is that the booting process hangs after printing out
> this:
> >> >
> >>
> >> fyi,
> >>
> >> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-
> 2.6.git;a=commitdiff;h=69dd3d8e29e294caaf63eb5e8a72d250279f9e5f;hp=fcb8
> ce5cfe30ca9ca5c9a79cdfe26d1993e65e0c
> >
> > Ah, thanks.  Now I see where the irq == 74 comes from. :) Still, I
> don't
> > see that on my system.
> >

FYI, I also saw this same issue on an AM37x-evm and beagleboard.  Reverting the above commit works for me to be able to see the boot messages from the file system after the kernel boots.

> 
> 69dd3d8 Revert "irq: Always set IRQF_ONESHOT if no primary handler is
> specified"
> f2b6071 drm/radeon: Extended DDC Probing for Toshiba L300D Radeon
> Mobility X1100 HDMI-A Connector
> c87019e HID: wiimote: Add status and return request handlers
> 2cb5e4b HID: wiimote: Add drm request
> 23a5a4a HID: wiimote: Register led class devices
> 26af174 HID: wiimote: Correctly call HID open/close callbacks
> 3989ef6 HID: wiimote: Simplify synchronization
> 8d3bb23 drm/ttm: ensure ttm for new node is bound before calling
> move_notify()
> eac2095 drm/ttm: unbind ttm before destroying node in accel move
> cleanup
> 7c4c396 drm/ttm: fix ttm_bo_add_ttm(user) failure path
> f5e4282 HID: usbhid: Add support for SiGma Micro chip
> 11f3a6b bridge: fix a possible net_device leak
> b6bede3 xfs: fix tracing builds inside the source tree
> 832d80a Merge branch 'for-davem' of
> git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless
> fcb8ce5 Linux 3.1-rc3
> 
> 
> --
> Thanks,
> Govindraj.R
> --
> To unsubscribe from this list: send the line "unsubscribe linux-omap"
> in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html