svn commit: r281075 - in stable/10: sys/conf sys/contrib/dev/acpica sys/contrib/dev/acpica/common sys/contrib/dev/acpica/compiler sys/contrib/dev/acpica/components/debugger sys/contrib/dev/acpica/c...

[Slawa Olhovchenkov]

On Thu, Apr 09, 2015 at 05:36:48PM -0400, Jung-uk Kim wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA256
> 
> On 04/09/2015 17:11, Slawa Olhovchenkov wrote:
> > On Thu, Apr 09, 2015 at 11:56:39PM +0300, Slawa Olhovchenkov
> > wrote:
> > 
> >> On Thu, Apr 09, 2015 at 03:32:08PM -0400, Jung-uk Kim wrote:
> >> 
> >>>>>> `acpidump -d` work `acpidump -t` don't work
> >>>>>> 
> >>>>>>> From a quick look, it seems that your RSDT contains an
> >>>>>>> entry with an address size which is neither 4 nor 8,
> >>>>>>> and that leads to the assertion failure.  I am unsure
> >>>>>>> whether there is any upstream fix for it.
> >>>>>> 
> >>>>>> I am not sure how long acpidump -t don't work.
> >>>> 
> >>>> addr_size must be 4 or 8 here.  Otherwise, something is
> >>>> corrupting memory.
> >>> 
> >>> Sorry, I misread your message.  Please install
> >>> sysutils/acpica-tools from ports, then do
> >>> /usr/local/bin/acpidump and send me the output.
> >> 
> >> Host crashed
> > 
> > Captured inforamtion you can take here: http://zxy.spb.ru/acpi/
> 
> The last entry in your XSDT is actually empty.  Although it is not
> clearly defined in the specification, I believe it should be skipped.
>  I just committed a workaround on head (r281331) and it will be merged
> to stable branches after 3 days.

I am try some timr later.
I never expect that 'Fatal trap 12: page fault while in user mode' may
trigger kernel crash.