Hi,
I do see a persistent glibc segfault on my current build
(glibc-2.26, gcc-7.2.0, flex-2.6.4) that did not appear on the previous
one (glibc-2.25, gcc-7.1.0, flex-2.6.4) when trying to execute
flex -i -s acpica-unix-20170728/source/compiler/aslcompiler.l
from https://acpica.org/sites/acpica/files/acpica-unix-20170728.tar.gz
Program received signal SIGSEGV, Segmentation fault.
#0 0x00007f8484d8d4e4 in __strcpy_ssse3 () from /lib/libc.so.6
#1 0x000000000040d31b in add_action (new_text=0x67d207
"PARSEOP_SERIALIZERULE_NOTSERIAL") at misc.c:134
#2 0x000000000041abfa in flexscan () at ./scan.l:939
#3 0x0000000000414dcd in yylex () at yylex.c:51
#4 0x000000000040f8f5 in yyparse () at parse.c:1405
#5 0x000000000040b135 in readin () at main.c:1475
#6 0x000000000040cd20 in flex_main (argc=6, argv=0x7fffffffebb8) at
main.c:170
#7 0x00007f8484c6af2a in __libc_start_main () from /lib/libc.so.6
#8 0x0000000000401cca in _start ()
I tend to see that as a glibc bug. Can anyone confirm or disprove? There
is little difference between working and non-working: same machine, pretty
much same build scripts close but not identical to the book, so glibc (or
gcc) version seems to be the likely source. But before blaming any of
those, I'd like to ensure it is not just me who fucked up the build.
Therefore: Does the above work/fail on a "properly built" LFS-8.1?
Thanks in advance
Uwe
--
http://lists.linuxfromscratch.org/listinfo/lfs-dev
FAQ: http://www.linuxfromscratch.org/faq/
Unsubscribe: See the above information page
