(double?)-clicking on room when start room has no exits causes seg. faults in A* routine

Bug #1263553 reported by Stephen Lyons
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Mudlet
Fix Released
Medium
Chris

Bug Description

I was getting fed up during debugging area and map code that I kept getting segmentation faults in the A* route finder if I clicked on rooms during usage of the 2D Map GUI. It turns out that this is caused if there are no (usable) exits from the start room for route finding - the attached commit/patch checks for at least one unlocked route from the start room and bails out before the route finder is started - returning a false result to indicate the failure to find a route.

Of course in practical usages rooms with no exits aren't much use in MUDs (death traps may not have a way out but unless you are an immortal that is the least of the problems for those type of rooms 8-( ), more likely is a room for which you have the ability to teleport to an unfixed and thus un-mappable destination. Alternatively the triggering situation may occur during off-line map creation / editing.

The patch - being from my current development source tree - may not match up on line number to the current Git version but as a single monolithic code block it should still apply with a bit of fuzziness to TMap.cpp .

Revision history for this message
Stephen Lyons (slysven) wrote :
Revision history for this message
Vadim Peretokin (vperetokin) wrote : Re: [Mudlet-makers] [Bug 1263553] [NEW] (double?)-clicking on room when start room has no exits causes seg. faults in A* routine
Revision history for this message
Chris (chrismudlet) wrote : Re: [Bug 1263553] [NEW] (double?)-clicking on room when start room has no exits causes seg. faults in A* routine

So this is actually the same bug I just fixed for the pathfinding between
non connected rooms. Any reason you didn't just commit this to the github
repo?

On Sun, Dec 22, 2013 at 5:37 PM, Stephen Lyons
<email address hidden>wrote:

> Public bug reported:
>
> I was getting fed up during debugging area and map code that I kept
> getting segmentation faults in the A* route finder if I clicked on rooms
> during usage of the 2D Map GUI. It turns out that this is caused if
> there are no (usable) exits from the start room for route finding - the
> attached commit/patch checks for at least one unlocked route from the
> start room and bails out before the route finder is started - returning
> a false result to indicate the failure to find a route.
>
> Of course in practical usages rooms with no exits aren't much use in
> MUDs (death traps may not have a way out but unless you are an immortal
> that is the least of the problems for those type of rooms 8-( ), more
> likely is a room for which you have the ability to teleport to an
> unfixed and thus un-mappable destination. Alternatively the triggering
> situation may occur during off-line map creation / editing.
>
> The patch - being from my current development source tree - may not
> match up on line number to the current Git version but as a single
> monolithic code block it should still apply with a bit of fuzziness to
> TMap.cpp .
>
> ** Affects: mudlet
> Importance: Medium
> Status: New
>
> ** Patch added: "commit-967b26d"
>
> https://bugs.launchpad.net/bugs/1263553/+attachment/3934501/+files/commit-967b26d
>
> --
> You received this bug notification because you are subscribed to Mudlet.
> https://bugs.launchpad.net/bugs/1263553
>
> Title:
> (double?)-clicking on room when start room has no exits causes seg.
> faults in A* routine
>
> Status in Mudlet the MUD client:
> New
>
> Bug description:
> I was getting fed up during debugging area and map code that I kept
> getting segmentation faults in the A* route finder if I clicked on
> rooms during usage of the 2D Map GUI. It turns out that this is
> caused if there are no (usable) exits from the start room for route
> finding - the attached commit/patch checks for at least one unlocked
> route from the start room and bails out before the route finder is
> started - returning a false result to indicate the failure to find a
> route.
>
> Of course in practical usages rooms with no exits aren't much use in
> MUDs (death traps may not have a way out but unless you are an
> immortal that is the least of the problems for those type of rooms 8-(
> ), more likely is a room for which you have the ability to teleport to
> an unfixed and thus un-mappable destination. Alternatively the
> triggering situation may occur during off-line map creation / editing.
>
> The patch - being from my current development source tree - may not
> match up on line number to the current Git version but as a single
> monolithic code block it should still apply with a bit of fuzziness to
> TMap.cpp .
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/mudlet/+bug/1263553/+subscriptions
>

Revision history for this message
Vadim Peretokin (vperetokin) wrote : Re: [Mudlet-makers] [Bug 1263553] [NEW] (double?)-clicking on room when start room has no exits causes seg. faults in A* routine

Addressing something else - which repo did this get fixed in,
pathfindingRevision or in another and merged into mudletDev? I didn't
notice such a thing and when reporting
https://bugs.launchpad.net/bugs/1263447 I tested with mudletDev.

Revision history for this message
Vadim Peretokin (vperetokin) wrote :

Ah OK.

Revision history for this message
Stephen Lyons (slysven) wrote :

@Chris:

I think we were working in parallel, I'd not been on-line during heavy
coding sessions over past few days and wasn't aware of
https://bugs.launchpad.net/bugs/1263447, I was just getting fed up with
segFaults that I couldn't back out of during debugging - what is it they
say about great minds thinking alike. 8-)

Do we want to mark one of the bug reports as a duplicate?

On 23/12/13 01:31, Vadim Peretokin wrote:
> Ah OK.
>
>
>
> _______________________________________________
> Mailing list: https://launchpad.net/~mudlet-makers
> Post to : <email address hidden>
> Unsubscribe : https://launchpad.net/~mudlet-makers
> More help : https://help.launchpad.net/ListHelp
>

Revision history for this message
Chris (chrismudlet) wrote :

Fixed in mudlet/development

Changed in mudlet:
assignee: nobody → Chris (chrismudlet)
milestone: none → 3.0
status: New → Fix Committed
Changed in mudlet:
status: Fix Committed → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.