bug#17245: GREP BUG: grep -P and binary files

[damon]

Hi there -

Bingo, that does it.  I have LANG set to en_CA.utf-8.  If i run:

LANG=en_CA grep-2.18/src/grep -P '\[.?max' test-image.png

It works fine, reporting no match.

Same for every other version i have compiled.

So definitely utf-8 related.  Let me know if i can provide anything
else.

-damon

On 13 Apr, Norihiro Tanaka wrote:

> This bug is similar to bug#16586.
>
> It seems that the pointer `eptr' for a current position in a text
> exceeded the starting position in backword searching.  I seem that PCRE
> library may assume that a text doesn't have invalid sequence in UTF-8.
>
> Could you re-try in them non-UTF8 locales?
>
> Norihiro

> $ gdb src/grep core.1430
> GNU gdb (GDB) 7.6.2
> Copyright (C) 2013 Free Software Foundation, Inc.
> License GPLv3+: GNU GPL version 3 or later
> <http://gnu.org/licenses/gpl.html>
> This is free software: you are free to change and redistribute it.
> There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
> and "show warranty" for details.
> This GDB was configured as "i386-pc-linux-gnu".
> For bug reporting instructions, please see:
> <http://www.gnu.org/software/gdb/bugs/>...
> Reading symbols from /home/staff/b/grep-2.18/src/grep...done.
> [New LWP 1430]
>
> warning: Can't read pathname for load map: Input/output error.
> Core was generated by `src/grep -P .?ma test-image.png'.
> Program terminated with signal 11, Segmentation fault.
> #0  0x001612ca in match (eptr=0x9a24fff <Address 0x9a24fff out of bounds>,
> ecode=0x9a25e65 "\035m\035ax",
>     mstart=0x9a26e9d
> "\272\374;\017\233\323\230:\364\005+\373a&\367\032X\304\216
> \342y\274\301\357\361\005",
>     offset_top=2, md=0xbfe18a64, eptrb=0x0, rdepth=0) at pcre_exec.c:5943
> 5943              BACKCHAR(eptr);
> (gdb) bt
> #0  0x001612ca in match (eptr=0x9a24fff <Address 0x9a24fff out of bounds>,
> ecode=0x9a25e65 "\035m\035ax",
>     mstart=0x9a26e9d
> "\272\374;\017\233\323\230:\364\005+\373a&\367\032X\304\216
> \342y\274\301\357\361\005",
>     offset_top=2, md=0xbfe18a64, eptrb=0x0, rdepth=0) at pcre_exec.c:5943
> #1  0x0016308a in pcre_exec (argument_re=0x9a25e28, extra_data=0x9a25e78,
>     subject=0x9a26e9d
> "\272\374;\017\233\323\230:\364\005+\373a&\367\032X\304\216
> \342y\274\301\357\361\005",
>     length=101, start_offset=0, options=8192, offsets=0xbfe18bdc,
> offsetcount=300) at pcre_exec.c:6941
> #2  0x0805a472 in Pexecute (buf=0x9a26000 "\211PNG\r\n\032\n", size=6568,
> match_size=0xbfe19114, start_ptr=0x0)
>     at pcresearch.c:174
> #3  0x0804ba07 in do_execute (buf=0x9a26000 "\211PNG\r\n\032\n", size=6568,
> match_size=0xbfe19114, start_ptr=0x0)
>     at grep.c:1073
> #4  0x0804bc98 in grepbuf (beg=0x9a26000 "\211PNG\r\n\032\n",
>     lim=0x9a279a8
> "\217\222(\016\001c\025R\221c\233S\250\327\177m\002\344Q\022\362$\320\066\3
> 76\327\245{\f\035D\001\260\251\326a\247{T\200_\bj8\274") at grep.c:1109
> #5  0x0804bfb3 in grep (fd=3, st=0xbfe19200) at grep.c:1220
> #6  0x0804c9ab in grepdesc (desc=3, command_line=1) at grep.c:1474
> #7  0x0804c650 in grepfile (dirdesc=-100, name=0xbfe19889 "test-image.png",
> follow=1, command_line=1) at grep.c:1375
> #8  0x0804cc22 in grep_command_line_arg (arg=0xbfe19889 "test-image.png") at
> grep.c:1526
> #9  0x0804e358 in main (argc=4, argv=0xbfe194a4) at grep.c:2362



--
Damon Harper           _/\_    Nothing is as simple as it seems at
address@hidden      __\  /__  first, as hopeless as it seems in
                     \      /  the middle, or as finished as it
www.usrbin.ca/damon   |/||\|   seems in the end.