Safety harnesses (was: Is a successful call to write(2) atomic?)

[Ronald F. Guilmette]

In message <a17cf47d-5aef-1623-46da-51c83054f083 at gmail.com>, 
Graham Perrin <grahamperrin at gmail.com> wrote:

>The surprise, to me, was losing fifty-something seconds' worth of data 
>in a kernel panic situation...

In an ideal universe, the kernel never panics.

Mine however still does from time to time.

I have been informed that this is attributable to me continuing to use an
otherwise perfectly servicable AMD graphics card that the X maintainer
has elected to no longer support.

I have suggested that there might possibly be some more graceful failure
mode than a kernel panic in response to some unhandled condition, but to
no avail.

(Hardware upgrade has been pednding on my end for more than a year.  I have
all of the necessary parts... new motherboard, new CPU/APU and new memory...
however I just never seem to find the time.)


Regards,
rfg