Re: Rerolling Stats Bug

From: StormeRider (
Date: 11/27/98

At 10:23 PM 11/23/98 -0800, George wrote:
>On Tue, 24 Nov 1998, Silk wrote:
>>Having problems with some rerolling of stats. I can (usually) try to roll
>>a few times before it crashes, but sometimes it can crash right when it
>>goes to roll the first time. It seems to me that the problem is not in the
>>actual code involving the rolling of the stats, but something with that
>>particular CON_ value. I'm using a modified bpl11, but this has never
>>worked reliably since I first moved it from one server to another, and all
>>successive servers. (Oddly, the first was a SunOS box, and all successive
>>were Linux boxes of one flavor or another. Hmmm...)
>Nifty, memory corruption?

That's what I am afraid of. I know that there was a long thread on this
a while back, but short of using some replacements for malloc, etc.,
what's the easiest way to track this kind of stuff down?
>#0  0x804a8f2 in timeadd (a={tv_sec = 911885363, tv_usec = 855715}, b={
>      tv_sec = 1073958349, tv_usec = 136466629}) at comm.c:942
> 911885363 = 0x365a4433         6\nD3
>Don't suppose your re-rolling function can output that sequence?

Afraid I don't follow what you mean.
>and your b.tv_sec/b.tv_usec look like stack pointers.  What compiler do you

gcc version egcs-2.90.29 980515 (egcs-1.0.3 release) (slackware 3.5 with
circle 3.0bpl11... I havent had the time to wade through and upgrade...)


StormeRider                  ---                 --- Slackware 2.0.34

ICQ UIN: 2810688

The work I do on this project, any image creating while not at work,
any web designing done for Winds of Storm Web Design, and any
work done on any GPL'd project are not the property of my employer
Winstar iCi and I retain the sole rights to the source unless it is
stated publicly as an open source forum.

     | Ensure that you have read the CircleMUD Mailing List FAQ:  |
     | |

This archive was generated by hypermail 2b30 : 12/15/00 PST