Q: Software against hardware failures

[Ulrich Windl]

Hi!

I think Bertrand once stated that the Airbus software was proven to have no null pointer references. No recently I read that the Airbus A320 has a hardware problem, and it's recommended to install a software update against that (German: https://www.heise.de/news/Airbus-Update-nach-Zwischenfall-Airlines-reagieren-rasch-11097561.html).
That makes me wonder: If an external influence on the hardware can make the existing software unreliable, how can that be fixed in software?
Agreed, I lack knowledge of the details, but if the computer itself reacts incorrectly, how can software fix it?
However if the external influence only makes a sensor reporting unexpected values that could be handled, given that valid sensor data isn't essential for correct operation.

Any insights on that?
Kind regards,
Ulrich

[Alejandro Garcia]

I'm also confused,

They said that solar flares uncover the defect. For me that would  imply hardware failure.

But then they also say that the fix is to "downgrade" the software to a previous version that would imply a software error.

I guess we will have to wait for more information. 

Alejandro García F. (elviejo)

EOM – End Of Message. The whole message is in the subject don't need to open it.
NNTR – No Need To Respond. Help cut down on all those “cool” and “thanks” emails.
SINGLE SUBJECT. Send one email for one topic, this makes replies easy..
CLEAR CALL TO ACTION: Ask for some specific result very clearly.

--
You received this message because you are subscribed to the Google Groups "Eiffel Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to eiffel-users...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/eiffel-users/59a84a91-8afa-44ef-8fe1-0d0e39f027c0%40gmail.com.

[Alejandro Garcia]

Some more Information about the Airbus Incident:

https://www.bbc.com/future/article/20251201-how-cosmic-rays-grounded-thousands-of-aircraft

I guess a way to protect against this with software would be to have an invariant that says: the features should be the same value, if we haven't run any command yet.

--

Alejandro García F. (elviejo)

Too brief? Here's why! http://emailcharter.org

[Ulrich Windl]

Thanks for the additional information!
Some random thoughts:
Obviously the "fly by wire" made this problem possible only.
The other thing is that the Airbus has three independent flight control computers, but obviously no redundancy or validation for the control outputs.
And for the "bit flip" it seems they don't have ECC RAM (error correction), because that could detect and fix any single bit error, and would also detect any two-bit error.
There also exists "RAM mirroring" for some computer hardware, and older Intel CPUs supported "COU mirroring" also. I could imagine two memory boards at a 90 degree angle would make it extremely unlikely that the same bit will flip on both boards due to cosmic radiation.

And it seems the "solution" just reduces the likelihood: refresh (write) the critical variable more often. Classical assertions will fail here; maybe a new class of fuzzy assertions is needed:
"If variable A had been set to B at time X, the likelihood that A still contains B at time Y (Y >= X) is L"... ">=" allows an L for "instant corruption".

--
Kind regards,
Ulrich Windl

13.12.2025 06:05:33 Alejandro Garcia <agarc...@gmail.com>:

> Some more Information about the Airbus Incident:
>
> https://www.bbc.com/future/article/20251201-how-cosmic-rays-grounded-thousands-of-aircraft
>
> I guess a way to protect against this with software would be to have an invariant that says: the features should be the same value, if we haven't run any command yet.
>
>
>
> On Sun, Nov 30, 2025 at 9:19 PM Alejandro Garcia <agarc...@gmail.com> wrote:
>> I'm also confused,
>> They said that solar flares uncover the defect. For me that would  imply hardware failure.
>>
>> But then they also say that the fix is to "downgrade" the software to a previous version that would imply a software error.
>>
>> I guess we will have to wait for more information. 
>>
>>
>> Alejandro García F. (elviejo)
>>
>> EOM – End Of Message. The whole message is in the subject don't need to open it.
>> NNTR – No Need To Respond. Help cut down on all those “cool” and “thanks” emails.
>> SINGLE SUBJECT. Send one email for one topic, this makes replies easy..
>> CLEAR CALL TO ACTION: Ask for some specific result very clearly.
>>
>> On Sun, Nov 30, 2025, 04:33 Ulrich Windl <u202...@gmail.com> wrote:
>>> Hi!
>>>
>>> I think Bertrand once stated that the Airbus software was proven to have no null pointer references. No recently I read that the Airbus A320 has a hardware problem, and it's recommended to install a software update against that (German: https://www.heise.de/news/Airbus-Update-nach-Zwischenfall-Airlines-reagieren-rasch-11097561.html).
>>> That makes me wonder: If an external influence on the hardware can make the existing software unreliable, how can that be fixed in software?
>>> Agreed, I lack knowledge of the details, but if the computer itself reacts incorrectly, how can software fix it?
>>> However if the external influence only makes a sensor reporting unexpected values that could be handled, given that valid sensor data isn't essential for correct operation.
>>>
>>> Any insights on that?
>>> Kind regards,
>>> Ulrich
>>>
>>> --
>>> You received this message because you are subscribed to the Google Groups "Eiffel Users" group.

>>> To unsubscribe from this group and stop receiving emails from it, send an email to eiffel-users...@googlegroups.com[eiffel-users%2Bunsu...@googlegroups.com].

>>> To view this discussion visit https://groups.google.com/d/msgid/eiffel-users/59a84a91-8afa-44ef-8fe1-0d0e39f027c0%40gmail.com.
>
>
> --
> Alejandro García F. (elviejo)
>
> Too brief? Here's why! http://emailcharter.org
>
> EOM – End Of Message. The whole message is in the subject don't need to open it.
> NNTR – No Need To Respond. Help cut down on all those “cool” and “thanks” emails.
> SINGLE SUBJECT. Send one email for one topic, this makes replies easy..
> CLEAR CALL TO ACTION: Ask for some specific result very clearly.
>
> --
> You received this message because you are subscribed to the Google Groups "Eiffel Users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to eiffel-users...@googlegroups.com.

> To view this discussion visit https://groups.google.com/d/msgid/eiffel-users/CALFgaYDtDubT4vx2Ay1etuY%3DdPvgcNZrkp%2BJBohr_W1payC10A%40mail.gmail.com[https://groups.google.com/d/msgid/eiffel-users/CALFgaYDtDubT4vx2Ay1etuY%3DdPvgcNZrkp%2BJBohr_W1payC10A%40mail.gmail.com?utm_medium=email&utm_source=footer].