assert, ABORT, and friends

[Vin Shelton]

On 6/21/06, Stephen J. Turnbull <stephen at xemacs.org> wrote:
> >>>>> "Jerry" == Jerry James <james at xemacs.org> writes:
>
>     Jerry> It does not preserve more data.  Our code base is riddled
>     Jerry> with code that assumes that ABORT() and failing assert()s
>     Jerry> do not return.  So in a double assertion failure scenario,
>     Jerry> what is likely to happen is that the second assert_failed()
>     Jerry> will return, and the code it returns to will immediately do
>     Jerry> something fatal based on the assumption that a return would
>     Jerry> not happen.  So XEmacs is dead anyway, without having done
>     Jerry> anything extra to preserve the user's data.
>
> I wonder if this scenario helps to account for the number of
> backtraces we see with no apparent connection to what the user said he
> was doing at the time?

That sounds like a good hypothesis and Jerry's work would be a great
boon if your hypothesis is true.

  - Vin