[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Gcl-devel] New GMP configuration avoidance and generalcommunication
From: |
Camm Maguire |
Subject: |
Re: [Gcl-devel] New GMP configuration avoidance and generalcommunication notes |
Date: |
07 Sep 2004 12:28:26 -0400 |
User-agent: |
Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 |
Greetings, and thanks as always Mike for your great work here!
"Mike Thomas" <address@hidden> writes:
> Hi Camm.
>
> | > 2. Automating the setting of C stack size and related items.
> | This has been
> | > working well for almost a month now with the default stack size, but I'm
> | > stuck on how to pass the configurable stack size (CSSIZE) into the
> | > configure.in test for the C Stack address (CSTACK_ADDRESS) - the final
> | > connection.
> |
> | Wondering whether configure can read an rusage limit component on Mingw...
>
> No, and as far as I can tell, the only way to determine the start of the C
> stack is by checking the location of a local variable at runtime. I was
> hoping to do that at configuration time to speed up NULL_OR_ON_CSTACK but
> I'm starting to think that is not necessarily a major optimisation, so maybe
> I'll just leave it as it is currently and just retain the configuration of
> the actual size of the C stack.
I don't think the performance issue is of much significance, but I do
think that getting robustly correct bounds on the C stack is important
on correctly handling a lot of stack overflow conditions. I have a
patch set, not yet committed, which restores the older sigaltstack
functionality enabling safe traps of C stack overruns, which trigger a
trappable segfault at least on Linux. Would be great to extend this
across the board.
>
> | > I am also curious as to the state of your 2.7.0 planning.
> | >
> |
> | Have obviously been a bit delayed in putting together this message --
> | will try to soon.
>
> Don't hurry - I think we may all have other things to do presently, and I
> was merely indulging my curiousity.
>
Indeed. The clouds appear to be parting soon ....
Take care,
> Cheers
>
> Mike Thomas.
>
>
>
>
> _______________________________________________
> Gcl-devel mailing list
> address@hidden
> http://lists.gnu.org/mailman/listinfo/gcl-devel
>
>
>
--
Camm Maguire address@hidden
==========================================================================
"The earth is but one country, and mankind its citizens." -- Baha'u'llah
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- Re: [Gcl-devel] New GMP configuration avoidance and generalcommunication notes,
Camm Maguire <=