DDE on real machines

Alexander Boettcher alexander.boettcher at ...1...
Thu Jun 27 09:47:01 CEST 2013


Sorry, for sending twice. The first message was send a bit to early ...

On 26.06.2013 18:12, Daniel Waddington wrote:
> Hi Alex,
> 
> I just tried your branch (head) with foc_x86_64 and nova_x86_64 and it
> also does not work on the machines in question (Dell T3600 and T3500).
> I don't think the issue is connected with LWIP or DDE since we have had
> "disappearing interrupts" with non-DDE network drivers without LWIP.  

That's odd. And you are saying that this happens for nova as foc.

Are the "disappearing interrupts" are maybe shared between several
devices ? If yes, you could try with a Genode version before release
12.11 in which we introduced the support for handling shared interrupts
in general. It's just a guess.

> Have you other machines you can test on?

I tested it today on another machine, based on a Intel DX58SO
motherboard and a Core i7 CPU. Could not reproduce it here.

Alex.

> 
> Daniel
> 
> On Wed, 2013-06-26 at 09:45 +0200, Alexander Boettcher wrote:
>> Hello,
>>
>> can you please give the last commit on branch [0] a try?
>>
>> The run/lwip script uses on x86 platforms with ACPI support the wrong
>> irq_session - that one of core, not the one of the acpi driver. If in
>> the ACPI table of your machine the IRQs are rewritten to other IRQs
>> (which is likely the case), core would use the wrong one.
>>
>> Does this commit solve your problem ?
>> (On our x86 test machine the run/lwip test passes for foc/nova 32/64
>> without and with the commit above.)
>>
>> Thanks for reporting and testing,
>>
>> Alex.
>>
>> [0] https://github.com/alex-ab/genode/tree/issue_lwip_acpi
>>
>> On 25.06.2013 21:49, Daniel Waddington wrote:
>>> I can confirm that the problem has crept in since release 13.02.
>>>
>>> Daniel
>>>
>>> On Tue, 2013-06-25 at 10:31 -0700, Daniel Waddington wrote:
>>>> Hi,
>>>> Can someone check DDE is still working on a real PC target?  I am
>>>> using fiasco.oc and 64bit but I think the problem is across kernels.
>>>> Basically my issue is that DHCP is not working and it appears that the
>>>> rx interrupt is not firing (according to foc jdb).  I've tried the
>>>> latest genode with run/lwip.
>>>>
>>>> Daniel 
>>>> ------------------------------------------------------------------------------
>>>> This SF.net email is sponsored by Windows:
>>>>
>>>> Build for Windows Store.
>>>>
>>>> http://p.sf.net/sfu/windows-dev2dev
>>>> _______________________________________________
>>>> Genode-main mailing list
>>>> Genode-main at lists.sourceforge.net
>>>> https://lists.sourceforge.net/lists/listinfo/genode-main
>>>
>>>
>>>
>>> ------------------------------------------------------------------------------
>>> This SF.net email is sponsored by Windows:
>>>
>>> Build for Windows Store.
>>>
>>> http://p.sf.net/sfu/windows-dev2dev
>>> _______________________________________________
>>> Genode-main mailing list
>>> Genode-main at lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/genode-main
>>>
>>
>>
>> ------------------------------------------------------------------------------
>> This SF.net email is sponsored by Windows:
>>
>> Build for Windows Store.
>>
>> http://p.sf.net/sfu/windows-dev2dev
>> _______________________________________________
>> Genode-main mailing list
>> Genode-main at lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/genode-main
> 
> 
> 
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Windows:
> 
> Build for Windows Store.
> 
> http://p.sf.net/sfu/windows-dev2dev
> _______________________________________________
> Genode-main mailing list
> Genode-main at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/genode-main
> 





More information about the users mailing list