Bug 704372

Summary: tolua++ segmentation fault while building Clanlib06
Product: [Fedora] Fedora Reporter: Karsten Hopp <karsten>
Component: luaAssignee: Tim Niemueller <tim>
Status: CLOSED RAWHIDE QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 15CC: michel, tim
Target Milestone: ---   
Target Release: ---   
Hardware: powerpc   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2011-10-11 18:15:20 EDT Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: ---

Description Karsten Hopp 2011-05-12 18:41:17 EDT
Description of problem:
the Clanlib06 spec file uses tolua++ to prepare some sources. tolua++ aborts with a segmentation fault, although it looks like it happens in liblua:

Core was generated by `tolua++ -o Sources/Lua/clanbindings.cpp Sources/Lua/clanbindings.pkg'.
Program terminated with signal 11, Segmentation fault.
#0  0x0ff88150 in luaV_execute (L=0x107c0008, nexeccalls=5) at lvm.c:389
389         const Instruction i = *pc++;
Missing separate debuginfos, use: debuginfo-install glibc-2.13.90-9.ppc
(gdb) bt
#0  0x0ff88150 in luaV_execute (L=0x107c0008, nexeccalls=5) at lvm.c:389
#1  0x0ff7c59c in luaD_call (L=0x107c0008, func=0x107c0290, nResults=<optimized out>) at ldo.c:377
#2  0x0ff7b5a4 in luaD_rawrunprotected (L=0x107c0008, f=0xff754e4 <f_call>, ud=0xff97a4b8) at ldo.c:116
#3  0x0ff7c7c0 in luaD_pcall (L=0x107c0008, func=<optimized out>, u=<optimized out>, old_top=64, ef=<optimized out>) at ldo.c:463
#4  0x0ff771c4 in lua_pcall (L=0x107c0008, nargs=<optimized out>, nresults=-1, errfunc=<optimized out>) at lapi.c:821
#5  0x0ff8b754 in luaB_pcall (L=0x107c0008) at lbaselib.c:377
#6  0x0ff7c0fc in luaD_precall (L=0x107c0008, func=<optimized out>, nresults=2) at ldo.c:319
#7  0x0ff88c74 in luaV_execute (L=0x107c0008, nexeccalls=1) at lvm.c:590
#8  0x0ff7c59c in luaD_call (L=0x107c0008, func=0x107c0270, nResults=<optimized out>) at ldo.c:377
#9  0x0ff7b5a4 in luaD_rawrunprotected (L=0x107c0008, f=0xff754e4 <f_call>, ud=0xff97a8d8) at ldo.c:116
#10 0x0ff7c7c0 in luaD_pcall (L=0x107c0008, func=<optimized out>, u=<optimized out>, old_top=32, ef=<optimized out>) at ldo.c:463
#11 0x0ff771c4 in lua_pcall (L=0x107c0008, nargs=<optimized out>, nresults=0, errfunc=<optimized out>) at lapi.c:821
#12 0x0ffd5f64 in tolua_dobuffer (L=0x107c0008, B=<optimized out>, size=<optimized out>, name=<optimized out>) at src/lib/tolua_map.c:699
#13 0x10001234 in tolua_tolua_open (tolua_S=0x107c0008) at src/bin/toluabind.c:8058
#14 0x10000a00 in main (argc=4, argv=0xff97abf4) at src/bin/tolua.c:152


Version-Release number of selected component (if applicable):
lua-5.1.4-8.fc15

Steps to Reproduce:
1. try to build ClanLib06 (on ppc, if that matters)
2.
3.
  

Additional info:
> rpm -q tolua++
tolua++-1.0.93-2.fc15
Comment 1 Karsten Hopp 2011-05-12 18:42:44 EDT
(gdb) p *L
$1 = {next = 0x1087a4f0, tt = 8 '\b', marked = 98 'b', status = 0 '\000', top = 0x107fba78, base = 0x107fb928, l_G = 0x107c0080, ci = 0x107eaa20, 
  savedpc = 0x11, stack_last = 0x107fc4a8, stack = 0x107fb788, end_ci = 0x107eac60, base_ci = 0x107ea978, stacksize = 216, size_ci = 32, nCcalls = 2, 
  baseCcalls = 0, hookmask = 0 '\000', allowhook = 1 '\001', basehookcount = 0, hookcount = 0, hook = 0, l_gt = {value = {gc = 0x107c0528, p = 0x107c0528, 
      n = 2.8877104096134679e-229, b = 276563240}, tt = 5}, env = {value = {gc = 0x107c2da0, p = 0x107c2da0, n = 2.90400198706782e-229, b = 276573600}, 
    tt = 5}, openupval = 0x0, gclist = 0x0, errorJmp = 0xff97a1e0, errfunc = 0}
Comment 2 Tim Niemueller 2011-05-13 16:58:11 EDT
Does the problem happen only on ppc or also on other architectures? I have no ppc machine and no intention to fix this bug if it is ppc specific. If it's broken only there I'll exclude the arch from building.

Maybe there is a PPC SIG which can help to debug and fix it? Otherwise you should try to report the problem to upstream and get it fixed there.
Comment 3 Karsten Hopp 2011-05-17 08:51:01 EDT
As this is package with seems to be used only for some games, an excludearch would probably be ok, even if I'd prefer a fix ;-)
Please make sure to follow the excludearch guidelines from 
http://fedoraproject.org/wiki/PackagingGuidelines#Architecture_Build_Failures so that we're able to lookup the reason for this change later on.
Comment 4 Tim Niemueller 2011-10-11 18:15:20 EDT
Added ExcludeArch in master branch. Not issuing an update just for this, but next clone and build should fail "properly".