i18n problem: citrus bug?

[Jukka Salmi <j+nbsd%2008.salmi.ch@localhost>]

Hello,

yesterday I received a [1]message which I can't read with Mutt (1.5.18
from pkgsrc) on a 5.99.01 i386 system: when trying to open the file Mutt
hangs, mremap(2)ping until being killed by the kernel, probably because
it runs out of memory.

$ ktrace -d -i /usr/pkg/bin/mutt -n -e 'unset header_cache' -f hang.message
$ kdump
 [...]
 29717      1 mutt     CALL  open(0xbfbf8628,0,0x80e5cb8)
 29717      1 mutt     NAMI  "/usr/share/i18n/esdb/esdb.dir.db"
 29717      1 mutt     RET   open 4
 29717      1 mutt     CALL  fcntl(4,2,1)
 29717      1 mutt     RET   fcntl 0
 29717      1 mutt     CALL  __fstat30(4,0xbfbf8588)
 29717      1 mutt     RET   __fstat30 0
 29717      1 mutt     CALL  mmap(0,0x3080,1,2,4,0,0,0)
 29717      1 mutt     RET   mmap -1149665280/0xbb798000
 29717      1 mutt     CALL  close(4)
 29717      1 mutt     RET   close 0
 29717      1 mutt     CALL  munmap(0xbb798000,0x3080)
 29717      1 mutt     RET   munmap 0
 29717      1 mutt     CALL  mmap(0,0x100000,3,0x14001002,0xffffffff,0,0,0)
 29717      1 mutt     RET   mmap -1152385024/0xbb500000
 29717      1 mutt     CALL  mmap(0,0x100000,3,0x14001002,0xffffffff,0,0,0)
 29717      1 mutt     RET   mmap -1153433600/0xbb400000
 29717      1 mutt     CALL  mmap(0,0x100000,3,0x14001002,0xffffffff,0,0,0)
 29717      1 mutt     RET   mmap -1154482176/0xbb300000
 29717      1 mutt     CALL  munmap(0xbb500000,0x100000)
 29717      1 mutt     RET   munmap 0
 29717      1 mutt     CALL  mremap(0xbb300000,0x100000,0,0x200000,0x14000000)
 29717      1 mutt     RET   mremap -1156579328/0xbb100000
 29717      1 mutt     CALL  mremap(0xbb100000,0x200000,0,0x300000,0x14000000)
 29717      1 mutt     RET   mremap -1156579328/0xbb100000
 29717      1 mutt     CALL  mremap(0xbb100000,0x300000,0,0x400000,0x14000000)
 29717      1 mutt     RET   mremap -1160773632/0xbad00000
 29717      1 mutt     CALL  mremap(0xbad00000,0x400000,0,0x500000,0x14000000)
 29717      1 mutt     RET   mremap -1160773632/0xbad00000
 [...]
 29717      1 mutt     CALL  
mremap(0x7b500000,0x32900000,0,0x32a00000,0x14000000)
 29717      1 mutt     RET   mremap 2068840448/0x7b500000
 29717      1 mutt     CALL  
mremap(0x7b500000,0x32a00000,0,0x32b00000,0x14000000)
 29717      1 mutt     RET   mremap 2068840448/0x7b500000
 29717      1 mutt     CALL  
mremap(0x7b500000,0x32b00000,0,0x32c00000,0x14000000)
 29717      1 mutt     RET   mremap 2068840448/0x7b500000
 29717      1 mutt     CALL  
mremap(0x7b500000,0x32c00000,0,0x32d00000,0x14000000)
 29717      1 mutt     RET   mremap 2068840448/0x7b500000
 29717      1 mutt     PSIG  SIGKILL SIG_DFL: code=[1], errno=0)

I can't reproduce this problem on a 4.0_STABLE i386 system using the
same version of Mutt, i.e. reading the mail works fine there.

Any hints?


Regards, Jukka

[1] http://salmi.ch/~jukka/mutt/hang.message

-- 
bashian roulette:
$ ((RANDOM%6)) || rm -rf ~