segfault while compiling tcl

Sam Tygier samtygier at
Mon Oct 31 08:47:09 PST 2005

Ken Moffat wrote:
> On Mon, 31 Oct 2005, Sam Tygier wrote:
>> hello
>> this is my first try at a lfs system. the computer is a via epia 
>> MII6000. i am using the live cd 6.1-3.
>  Some via cpus in the past didn't have an instruction gcc  used for 
> i686.  But if you can bootstrap the first compiler and build glibc, I 
> doubt that is the problem.

sounds possible, i know this machine can't run the 686 kernels in ubuntu. i think it does not support all the extra instructions.

is there a way to force it to only use 386 (or c3 compatable) instructions?
>> everything goes fine up till compiling tcl. it segfaults with the 
>> following message
>> gcc -pipe -c -O2  -DTCL_DBGX= -Wall -Wno-implicit-int 
>> -fno-strict-aliasing -fPIC -I./../generic -I. -DPEEK_XCLOSEIM=1 
> [...]
>> -DHAVE_SYS_IOCTL_H=1         -DTCL_SHLIB_EXT=\".so\" 
>> ./../unix/tclLoadDl.c
>> rm -f libtclstub8.4.a
>> ar cr libtclstub8.4.a tclStubLib.o  ; ranlib libtclstub8.4.a
>> /bin/sh: line 1: 23253 Segmentation fault      ranlib libtclstub8.4.a
>> make: *** [libtclstub8.4.a] Error 139
>> it always fails in the same place with the same error
>  Segmentation faults ('signal 11') are usually a sign that something is 
> not well - bad memory, insufficient processor cooling, perhaps bad power 
> supply.  But, gcc and glibc are usually a much harder test of the 
> machine, and failing in the exact same place is not especially common.
>  I recommend running memtest86+ for several hours.  Via epias tend to 
> come in exotic cases, you might want to check if the cooling seems 
> adequate (e.g. use any bios readings, or feel how hot any external 
> metalwork gets).
> Ken

exotic case, not at the moment, its all exposed in an open draw. it has run for days like this.

i am running a memtest86+, no problems in first 10 mins


How much free photo storage do you get? Store your holiday 
snaps for FREE with Yahoo! Photos

More information about the lfs-support mailing list