[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#747519: bibtool: "Symbol does not start with a letter" random errors with pipe



-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi Again,

On 02/03/15 14:46, Vincent Lefevre wrote:
> On 2015-03-02 13:34:30 +0100, Jerome BENOIT wrote:
>> I can reproduce the issue on my laptop too,
>> but the upstream could not reproduce it on his virtual box.
>> Have you try your material on a virtual box ?
> 
> I have a virtual machine with Ubuntu 14.04, but I can't reproduce the
> bug on it (the bibtool version is not the cause, since I can reproduce
> the bug on a Debian 7.8 machine, which has the same bibtool version).
> 
> Note also that I can't reproduce the bug when bibtool is run under
> valgrind (by adding "valgrind" in front of the bibtool line in my
> testcase) on machines where I could reproduce the bug.
> 

I observed the same.

> Perhaps the code runs too slowly on the virtual machine and under
> valgrind.
> 
> BTW, valgrind signals a read error:
> 
> ==4060== Invalid read of size 1
> ==4060==    at 0x114836: ??? (in /usr/bin/bibtool)
> ==4060==    by 0x115110: ??? (in /usr/bin/bibtool)
> ==4060==    by 0x10B304: ??? (in /usr/bin/bibtool)
> ==4060==    by 0x10B945: ??? (in /usr/bin/bibtool)
> ==4060==    by 0x109FAE: ??? (in /usr/bin/bibtool)
> ==4060==    by 0x506FB44: (below main) (libc-start.c:287)
> ==4060==  Address 0x7f59f1c is 0 bytes after a block of size 60 alloc'd
> ==4060==    at 0x4C28C20: malloc (vg_replace_malloc.c:296)
> ==4060==    by 0x1196B6: ??? (in /usr/bin/bibtool)
> ==4060==    by 0x1197E0: ??? (in /usr/bin/bibtool)
> ==4060==    by 0x1135BB: ??? (in /usr/bin/bibtool)
> ==4060==    by 0x113734: ??? (in /usr/bin/bibtool)
> ==4060==    by 0x113DB9: ??? (in /usr/bin/bibtool)
> ==4060==    by 0x10BEBB: ??? (in /usr/bin/bibtool)
> ==4060==    by 0x109DE4: ??? (in /usr/bin/bibtool)
> ==4060==    by 0x506FB44: (below main) (libc-start.c:287)
> 

Ok. I will resubmit the bug to the upstream maintainer.
The last time I submitted it, I could not reproduce the issue.
And on my side, I could hardly trace it.

Thanks,
Jerome



-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)

iQEcBAEBAgAGBQJU9GvpAAoJEIC/w4IMSybjR4MIAKdj5I1QxLXtaapx5JOJH/K7
YH4B3R5rMx33B1vbQSV0dWQdDQPq/D2LfuKFmY0hLAwec+dRbpgovxI1sqmxrtiE
2jIHGtxKBu31hJSHWl0Cvd7R9R8Vl8T4P9PhRUYou+GaPGMOEKICGiuyD2aM9FML
g1dLkva1XGMOmwYD1lQiaffMjUvcFBt3/kASto1Cj/VVWVAKDT/nWTNTZlyaXeGs
ALc9DGVs95b+A4C/QUW5k6m0uYMTsgJoaxilZBgvfru165MKfNMYVXctQVVaLEaz
UicjFdkVNAvRvLbYOL/qb1FQp1RcNBL+sOyY0x8ryBEIohJlGCjyXWK2f/VO4ck=
=ay/1
-----END PGP SIGNATURE-----


Reply to: