Menu
▾
▴
1 Attachments
#300 fgfs-construct segfaults (Linux x86_64)
Originally created by: kri... (code.google.com)@gmail.com
Trying to build scenery using terragear leads to a segfault in the last step (fgfs-construct).
The scenery I try to get to work consists of v0_landmass, CORINE- and OpenStreetMap-data. Building with "--lat=49.92 --lon=9.15 --ydist=0.02 --xdist=0.02" works, trying "--lat=49.92 --lon=9.15 --ydist=0.5 --xdist=0.5" fails.
Versions used:
gpc: v2.32
newmat: "newmat11.tar.gz"
simgear-cs: git HEAD, commit [rcc1e359a04a57e1815db5f19b4c016eb9271931d]
terragear-cs: git HEAD, commit [r0fdbcaff635a761c8f92c9600b470c0a02ce3e4b]
I'm using Gentoo-Linux unstable, gcc is "gcc-Version 4.5.2 (Gentoo 4.5.2 p1.0, pie-0.4.5)"
Segfault is reproducible, valgrind output is attached. Scenery can be uploaded on request, just might be *a bit* large (it's most of south-germany with detailed roads from OSM)
View and moderate all "codetickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: aeitsch...@yahoo.de
(No comment was entered for this change.)
Labels: Scenery
View and moderate all "codetickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: kri... (code.google.com)@gmail.com
After a short discussion in FlightGear's IRC-channel yesterday, I started fgfs-construct with "--lat=50 --lon=9 --ydist=1 --xdist=1". The result is the same: segfault with similar valgrind-output
View and moderate all "codetickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: gijsrooy
(No comment was entered for this change.)
Labels: TerraGear
View and moderate all "codetickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: goo...@mgras.net
Segfaults in GPS have been a permanent companion to using TerraGear's "fgfs-construct".
Status: Stalled
View and moderate all "codetickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: chrschm... (code.google.com)@gmail.com
Have you set the recommended GPC_EPSILON value in gpc.h as recommended in the TG Readme file for GPC?
View and moderate all "codetickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: clement.... (code.google.com)@gmail.com
I have create a script to install it automatically.
See here : http://www.flightgear.org/forums/viewtopic.php?f=20&t=14849
View and moderate all "codetickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: psadrozi... (code.google.com)@gmail.com
I'm working on replacing GPC with an integer based clipping library (clipper). It is actively maintained by Angus Johnson. It is used by default in the genapt850 parser. Work is ongoing to get it working with fgfs-construct.
View and moderate all "codetickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: chrschm... (code.google.com)@gmail.com
This issue should be revisited. A lot of changes have been done meanwhile. The original reporter should either try TG again (built with normal SG, not SG-CS) and report back here. Else this bug should be closed.
View and moderate all "codetickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: kri... (code.google.com)@gmail.com
I'll take a look at this again when I can find the time. Currently my university-exams are keeping me busy enough :/
View and moderate all "codetickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: bre... (code.google.com)@gmail.com
(No comment was entered for this change.)
Status: Testing
View and moderate all "codetickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: chrschm... (code.google.com)@gmail.com
(No comment was entered for this change.)
Status: Stalled
View and moderate all "codetickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: chrschm... (code.google.com)@gmail.com
(No comment was entered for this change.)
Status: NeedInfo
View and moderate all "codetickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: chrschm... (code.google.com)@gmail.com
It is highly unlikely that this problem still exists. TG has seen *a lot* of changes/fixes and construct is almost a rewrite right now. Anyway, If your exams are already behind you, please test again (note that the program is now called "tg-construct" :)
Status: Stalled
View and moderate all "codetickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: bre... (code.google.com)@gmail.com
See Christian's comment #13. Issue will be closed soon, unless anyone is able to reproduce the issue with the latest version.
Status: Testing
View and moderate all "codetickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: kri... (code.google.com)@gmail.com
Yeah, please just close it. I'm currently writing my master-thesis and so don't have time to reproduce this.
I fully understand that you cannot work on issues no one can reproduce or verify and I don't want to stop you from doing other useful things beeing held back by this.
View and moderate all "codetickets Discussion" comments posted by this user
Mark all as spam, and block user from posting to "Tickets"
Originally posted by: chrschm... (code.google.com)@gmail.com
(No comment was entered for this change.)
Status: Expired
Diff:
tg-construct is currently working and OP requested this ticket be closed.