Re: [Lynx-dev] SIGSEGV in scan_cookie_sublist

[Thorsten Glaser]

Thomas Dickey dixit:

>Is it something that you can reproduce without special data,
>e.g., stuff that you cannot share?

I cannot reliably reproduce it at all, it just happens after
a, usually very long (as in, days) time of browsing cookie-
happy sites (fanfiction sites, in this case, mostly).

>> This wasn’t in dev.14 though.
>
>The slice of code (the sprintf that you referred to) was not changed
>in dev.15 (src/LYCookie.c was last changed in June 2011).  If there's

Right, but that's only the place where it segfaulted, not
necessarily the place where the bug occurs (as I wrote, I
suspect a faulty but non-NULL entry in the list somewhere,
possibly a use-after-free, possibly a corruption).

>a bug introduced in dev.15, I'd look first at other files:

Hrm.

Of course I cannot reliably say it was introduced in dev.15,
just that it started to happen to me lately – maybe some site
redesign (fanfiction.net did one, AO3 maybe too) is the actual
cause.

Sorry for that.

I can do a debug build and save the core file next time, if
you want. (Would be useful if the signal handler can be dis-
abled, is there a compile-time option for that? Otherwise,
gdb cannot inspect the stack frames beyond the handler.)

bye,
//mirabilos
-- 
[DJBDNS Zone] TTL 86400 – <Natureshadow> kann man da auch 1d schreiben?
<mirabilos> nö, außerdem kann ein Deutscher oder ein Japaner mit 1d
    ja erstmal nix anfangen, oder könntest du 1日 im zone file lesen?
<Natureshadow> das heißt für mich: ein Regal, das u.U. schiefstehen könnte