Subject: Re: [issue] asic3 irq - msg#00008

List: handhelds.ipaq.hx4700.linux

Date: Prev Next Index Thread: Prev Next Index

On Sat, Dec 09, 2006 at 03:21:10PM +0300, Anton Vorontsov wrote:
> Hi all,
>
> I've caught ASIC3 IRQ issue.
>
> It's easily reproducible for me (using 2.6.19-hh4):
> 1. Put hx4700 into cradle
> 2. attach USB and AC
> 3. Pull AC adapter's supply from cradle (not from the wall)
>
> Then I'm getting unstoppable messages:
> <6>hx4700 power_change: ac_in=0
> <6>hx4700 power_change: usb_in=1
> <6>hx4700 power_change: ac_in=0
> <6>hx4700 power_change: usb_in=1
> <6>hx4700 power_change: ac_in=0
> <6>hx4700 power_change: usb_in=1
> <6>hx4700 power_change: ac_in=0
>
> I've made patch (attached) to debug it a little and now getting:
>
> <4>ac attach
> <6>hx4700 power_change: ac_in=0
> <6>hx4700 power_change: usb_in=1
> <4>falling
> <4>ac attach
> <6>hx4700 power_change: ac_in=0
> <6>hx4700 power_change: usb_in=1
> <4>falling
> <4>ac attach
> <6>hx4700 power_change: ac_in=0
> <6>hx4700 power_change: usb_in=1
> <4>falling
> <4>ac attach
> <6>hx4700 power_change: ac_in=0
> <6>hx4700 power_change: usb_in=1
> <4>falling
> <4>ac attach
> <6>hx4700 power_change: ac_in=0
> <6>hx4700 power_change: usb_in=1
> <4>falling
>
> Thus, it's ac irq "going slightly mad" (c)
>
> That not happen for me if USB not attached, or if AC pulled from the
> wall. Thus I'm suspecting some ASIC3 timing issue: AC pulled from cradle
> asserts irq fast, and pulling AC from the wall asserts irq a bit "slowly"
> (current drops slowly because step-down transformer in AC adapter).
>
> Also, it seems not related to ASIC3 irq changes made lately, I've
> reverted them and still got issue. Though, I've not tested 2.6.{18,17}
> yet. I've tried to implement it through bothedge but then I'm not
> getting any irqs. :-/

Actually, when I set up BOTHEDGE irq I'm getting:

"interrupt processing overrun" from asic3_base. :-/

>
> Can somebody confirm that issue? Or any ideas?
>
>
> Thanks,
>
> -- Anton (irc: bd2)

Was this page helpful?

Thread at a glance:

Previous Message by Date: click to view message preview

Re: "Erase failed" ... how to fix?

On Tue, 2006-12-05 at 15:32 -0500, Aric D. Blumer wrote: > Stephen Strowes wrote: > > status=00C000C0 > > This status means that the block is in erase suspend, but the bootldr does not > use erase suspend. So perhaps a hard reset would fix it? (Press reset button > with stylus while holding down Calendar and the launch button.) Hmm, I'd tried this before posting; a hard reset didn't solve the problem. However, I left the battery out overnight, and I no longer get the error. I have now reflashed my root filesystem to get me back to a clean system, and the ipaq appears to have booted flawlessly. Hmm. Cheers, -S.

Next Message by Date: click to view message preview

[issue] asic3 irq

Hi all, I've caught ASIC3 IRQ issue. It's easily reproducible for me (using 2.6.19-hh4): 1. Put hx4700 into cradle 2. attach USB and AC 3. Pull AC adapter's supply from cradle (not from the wall) Then I'm getting unstoppable messages: <6>hx4700 power_change: ac_in=0 <6>hx4700 power_change: usb_in=1 <6>hx4700 power_change: ac_in=0 <6>hx4700 power_change: usb_in=1 <6>hx4700 power_change: ac_in=0 <6>hx4700 power_change: usb_in=1 <6>hx4700 power_change: ac_in=0 I've made patch (attached) to debug it a little and now getting: <4>ac attach <6>hx4700 power_change: ac_in=0 <6>hx4700 power_change: usb_in=1 <4>falling <4>ac attach <6>hx4700 power_change: ac_in=0 <6>hx4700 power_change: usb_in=1 <4>falling <4>ac attach <6>hx4700 power_change: ac_in=0 <6>hx4700 power_change: usb_in=1 <4>falling <4>ac attach <6>hx4700 power_change: ac_in=0 <6>hx4700 power_change: usb_in=1 <4>falling <4>ac attach <6>hx4700 power_change: ac_in=0 <6>hx4700 power_change: usb_in=1 <4>falling Thus, it's ac irq "going slightly mad" (c) That not happen for me if USB not attached, or if AC pulled from the wall. Thus I'm suspecting some ASIC3 timing issue: AC pulled from cradle asserts irq fast, and pulling AC from the wall asserts irq a bit "slowly" (current drops slowly because step-down transformer in AC adapter). Also, it seems not related to ASIC3 irq changes made lately, I've reverted them and still got issue. Though, I've not tested 2.6.{18,17} yet. I've tried to implement it through bothedge but then I'm not getting any irqs. :-/ Can somebody confirm that issue? Or any ideas? Thanks, -- Anton (irc: bd2) debug.patch Description: Text document bothedge.patch Description: Text document _______________________________________________ Kernel-discuss mailing list Kernel-discuss-CN5wO63fgwogsBAKwltoeQ@xxxxxxxxxxxxxxxx https://handhelds.org/mailman/listinfo/kernel-discuss

Previous Message by Thread: click to view message preview

Next Message by Thread: click to view message preview

Re: [Hx4700-port] Re: [issue] asic3 irq

On 12/9/06, Anton Vorontsov <cbou-JGs/UdohzUI@xxxxxxxxxxxxxxxx> wrote: On Sat, Dec 09, 2006 at 03:21:10PM +0300, Anton Vorontsov wrote: > Hi all, > > I've caught ASIC3 IRQ issue. > > It's easily reproducible for me (using 2.6.19-hh4): > 1. Put hx4700 into cradle > 2. attach USB and AC > 3. Pull AC adapter's supply from cradle (not from the wall) > > Then I'm getting unstoppable messages: > <6>hx4700 power_change: ac_in=0 > <6>hx4700 power_change: usb_in=1 > <6>hx4700 power_change: ac_in=0 > <6>hx4700 power_change: usb_in=1 > <6>hx4700 power_change: ac_in=0 > <6>hx4700 power_change: usb_in=1 > <6>hx4700 power_change: ac_in=0 > > I've made patch (attached) to debug it a little and now getting: > > <4>ac attach > <6>hx4700 power_change: ac_in=0 > <6>hx4700 power_change: usb_in=1 > <4>falling > <4>ac attach > <6>hx4700 power_change: ac_in=0 > <6>hx4700 power_change: usb_in=1 > <4>falling > <4>ac attach > <6>hx4700 power_change: ac_in=0 > <6>hx4700 power_change: usb_in=1 > <4>falling > <4>ac attach > <6>hx4700 power_change: ac_in=0 > <6>hx4700 power_change: usb_in=1 > <4>falling > <4>ac attach > <6>hx4700 power_change: ac_in=0 > <6>hx4700 power_change: usb_in=1 > <4>falling > > Thus, it's ac irq "going slightly mad" (c) > > That not happen for me if USB not attached, or if AC pulled from the > wall. Thus I'm suspecting some ASIC3 timing issue: AC pulled from cradle > asserts irq fast, and pulling AC from the wall asserts irq a bit "slowly" > (current drops slowly because step-down transformer in AC adapter). > > Also, it seems not related to ASIC3 irq changes made lately, I've > reverted them and still got issue. Though, I've not tested 2.6.{18,17} > yet. I've tried to implement it through bothedge but then I'm not > getting any irqs. :-/ Actually, when I set up BOTHEDGE irq I'm getting: "interrupt processing overrun" from asic3_base. :-/ This sounds like a level interrupt being treated like an edge interrupt, or an edge interrupt not getting ACK'd. Matt

Loading Comments...