Mailing List Archive: Python-1.5.2 testing zlib

Python-1.5.2 testing zlib

Apr 17, 1999, 6:03 PM

Post #1 of 4 (946 views)

My "make check" fails with a core dump after "test_zlib".
Runng that test by hand shows:

: s Python-1.5.2 ; ./python Lib/test/test_zlib.py
0xe5c1a120 0x43b6aa94
0xbd602f7 0xbd602f7
expecting Bad compression level
expecting Invalid initialization option
expecting Invalid initialization option
normal compression/decompression succeeded
compress/decompression obj succeeded
decompress with init options succeeded
decompressobj with init options succeeded
Bus error - core dumped

Adding a print at the start of the last loop:
: s Python-1.5.2 ; ./python Lib/test/test_zlib.py
0xe5c1a120 0x43b6aa94
0xbd602f7 0xbd602f7
expecting Bad compression level
expecting Invalid initialization option
expecting Invalid initialization option
normal compression/decompression succeeded
compress/decompression obj succeeded
decompress with init options succeeded
decompressobj with init options succeeded
Decompress: flush mode=0, level=0
Memory fault - core dumped

Anyone else seen this?

(My system:
: s Python-1.5.2 ; uname -a
SunOS staff 5.6 Generic_105181-05 sun4u sparc SUNW,Ultra-2
: s Python-1.5.2 ; ./python
Python 1.5.2 (#2, Apr 17 1999, 20:08:31) [GCC 2.8.1] on sunos5
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
)

Python-1.5.2 testing zlib [ In reply to ]

arcege at shore

Apr 18, 1999, 5:17 AM

Post #2 of 4 (928 views)

Permalink

Piers Lauder <piers@cs.su.oz.au> wrote:
: My "make check" fails with a core dump after "test_zlib".
: Runng that test by hand shows:

: : s Python-1.5.2 ; ./python Lib/test/test_zlib.py
: 0xe5c1a120 0x43b6aa94
: 0xbd602f7 0xbd602f7
: expecting Bad compression level
: expecting Invalid initialization option
: expecting Invalid initialization option
: normal compression/decompression succeeded
: compress/decompression obj succeeded
: decompress with init options succeeded
: decompressobj with init options succeeded
: Bus error - core dumped

: Adding a print at the start of the last loop:
: : s Python-1.5.2 ; ./python Lib/test/test_zlib.py
: 0xe5c1a120 0x43b6aa94
: 0xbd602f7 0xbd602f7
: expecting Bad compression level
: expecting Invalid initialization option
: expecting Invalid initialization option
: normal compression/decompression succeeded
: compress/decompression obj succeeded
: decompress with init options succeeded
: decompressobj with init options succeeded
: Decompress: flush mode=0, level=0
: Memory fault - core dumped

: Anyone else seen this?

: (My system:
: : s Python-1.5.2 ; uname -a
: SunOS staff 5.6 Generic_105181-05 sun4u sparc SUNW,Ultra-2
: : s Python-1.5.2 ; ./python
: Python 1.5.2 (#2, Apr 17 1999, 20:08:31) [GCC 2.8.1] on sunos5
: Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
: )

Yes, I found this on Solaris for Intel 2.6 as well. I hadn't gotten
time to look into it more. For now, I've removed the zlib module from
the build and continued on with my own system tests. I was thinking I
just had a _very_ old version of libz (Oct 97).

Since I didn't get time to download the libz, I couldn't test this on
the AIX 4.2 systems at work.

SunOS golem 5.6 Generic_105182-05 i86pc i386 i86pc
Python 1.5.2 (#5, Apr 17 1999, 14:07:59) [GCC 2.7.2.2] on sunos5

-Arcege

Python-1.5.2 testing zlib [ In reply to ]

mark at chem

Apr 19, 1999, 1:31 AM

Post #3 of 4 (924 views)

Permalink

piers@cs.su.oz.au (Piers Lauder) writes:

>My "make check" fails with a core dump after "test_zlib".
>Runng that test by hand shows:

> Bus error - core dumped

>(My system:
> : s Python-1.5.2 ; uname -a
> SunOS staff 5.6 Generic_105181-05 sun4u sparc SUNW,Ultra-2
> : s Python-1.5.2 ; ./python
> Python 1.5.2 (#2, Apr 17 1999, 20:08:31) [GCC 2.8.1] on sunos5

Works just fine for me on both DEC Alpha
(uname -a: OSF1 hostname V4.0 878 alpha,
python: Python 1.5.2 (#1, Apr 16 1999, 15:49:36) [C] on osf1V4)
and Solaris
(uname -a: SunOS hostname 5.6 Generic sun4m sparc SUNW,SPARCstation-10,
python: 1.5.2 (@1, Apr 16 1999, 17:38:58) [GCC egcs-2.90.29 980515 (egcs-1.0.3
re on sinos5)

These are both with zlib-1.1.3.

--
Email - mark@chem.uwa.edu.au ,-_|\ Mark C Favas
Phone - +61 9 380 3482 / \ Department of Chemistry
Fax - +61 9 380 1005 ---> *_,-._/ The University of Western Australia
v Nedlands
Loc - 31.97 S, 115.81 E Western Australia 6009

Python-1.5.2 testing zlib [ In reply to ]

guido at eric

Apr 19, 1999, 8:13 AM

Post #4 of 4 (928 views)

Permalink

piers@cs.su.oz.au (Piers Lauder) writes:
> My "make check" fails with a core dump after "test_zlib".

mark@chem.uwa.edu.au (Mark C Favas) replies:
> Works just fine for me on both DEC Alpha
[..]
> These are both with zlib-1.1.3.

Yes, the new zlib uses features of zlib 1.1.3 that aren't supported by
zlib 1.0.4, so you must use 1.1.3 or later.

This is documented in the source but easily overseen; unfortunately we
didn't think of putting in a version number test.

--Guido van Rossum (home page: http://www.python.org/~guido/)