14 years ago by Pierre Joye — view source — reply

unread

hi,

It seems that 5.4 and trunk has something broken in zlib and maybe
bas64 encoding/decoding.

We have a serie of tests failing on linux or windows:

ext\zlib\tests\bug_52944.phpt
ext\zlib\tests\gzcompress_basic1.phpt
ext\zlib\tests\gzcompress_variation1.phpt
ext\zlib\tests\gzdeflate_basic1.phpt
ext\zlib\tests\gzdeflate_variation1.phpt
ext\zlib\tests\gzencode_basic1.phpt
ext\zlib\tests\gzencode_variation1-win32.phpt (the unix version too)
ext\zlib\tests\gzfile_basic.phpt
ext\zlib\tests\gzfile_basic2.phpt
\tests\gzfilegzreadfile.phpt
ext\zlib\tests\gzuncompress_basic1.phpt
ext\zlib\tests\readgzfile_basic.phpt
ext\zlib\tests\readgzfile_basic2.phpt

cheers,

Pierre

@pierrejoye | http://blog.thepimp.net | http://www.libgd.org

14 years ago by Stas Malyshev — view source — reply

unread

Hi!

hi,

It seems that 5.4 and trunk has something broken in zlib and maybe
bas64 encoding/decoding.

We have a serie of tests failing on linux or windows:

Interestingly enough, none of these fail for me on Darwin. Could you
publish the diffs?

Stanislav Malyshev, Software Architect
SugarCRM: http://www.sugarcrm.com/
(408)454-6900 ext. 227

14 years ago by Laruence — view source — reply

unread

Hi:
ext/zlib/tests/gzencode_variation1.phpt failed in Linux(redhat 64-bit)

test log http://pastebin.com/TzpW8GTK

thanks

2011/9/6 Pierre Joye pierre.php@gmail.com:

hi,

It seems that 5.4 and trunk has something broken in zlib and maybe
bas64 encoding/decoding.

We have a serie of tests failing on linux or windows:

ext\zlib\tests\bug_52944.phpt
ext\zlib\tests\gzcompress_basic1.phpt
ext\zlib\tests\gzcompress_variation1.phpt
ext\zlib\tests\gzdeflate_basic1.phpt
ext\zlib\tests\gzdeflate_variation1.phpt
ext\zlib\tests\gzencode_basic1.phpt
ext\zlib\tests\gzencode_variation1-win32.phpt (the unix version too)
ext\zlib\tests\gzfile_basic.phpt
ext\zlib\tests\gzfile_basic2.phpt
\tests\gzfilegzreadfile.phpt
ext\zlib\tests\gzuncompress_basic1.phpt
ext\zlib\tests\readgzfile_basic.phpt
ext\zlib\tests\readgzfile_basic2.phpt

cheers,

Pierre

@pierrejoye | http://blog.thepimp.net | http://www.libgd.org

--

--
Laruence Xinchen Hui
http://www.laruence.com/

14 years ago by Pierre Joye — view source — reply

unread

hi,

That's the only remaining, the other failed tests were due to LF (was
sure I did the conversion already, incl. external data being read).

However it happens randomly and only on unix.

Hi:
ext/zlib/tests/gzencode_variation1.phpt failed in Linux(redhat 64-bit)

test log http://pastebin.com/TzpW8GTK

thanks

2011/9/6 Pierre Joye pierre.php@gmail.com:

hi,

It seems that 5.4 and trunk has something broken in zlib and maybe
bas64 encoding/decoding.

We have a serie of tests failing on linux or windows:

ext\zlib\tests\bug_52944.phpt
ext\zlib\tests\gzcompress_basic1.phpt
ext\zlib\tests\gzcompress_variation1.phpt
ext\zlib\tests\gzdeflate_basic1.phpt
ext\zlib\tests\gzdeflate_variation1.phpt
ext\zlib\tests\gzencode_basic1.phpt
ext\zlib\tests\gzencode_variation1-win32.phpt (the unix version too)
ext\zlib\tests\gzfile_basic.phpt
ext\zlib\tests\gzfile_basic2.phpt
\tests\gzfilegzreadfile.phpt
ext\zlib\tests\gzuncompress_basic1.phpt
ext\zlib\tests\readgzfile_basic.phpt
ext\zlib\tests\readgzfile_basic2.phpt

cheers,

Pierre

@pierrejoye | http://blog.thepimp.net | http://www.libgd.org

--

--
Laruence Xinchen Hui
http://www.laruence.com/

--
Pierre

@pierrejoye | http://blog.thepimp.net | http://www.libgd.org

14 years ago by Laruence — view source — reply

unread

Hi:
I have linked against zlib-1.2.5 , still failed
and I found php5.3-trunk passed in my environ.

I can not go further till deflateEnd(zlib.c),   and the deflate

result is different between 5.3 and 5.4..

thanks

2011/9/8 Pierre Joye pierre.php@gmail.com:

hi,

That's the only remaining, the other failed tests were due to LF (was
sure I did the conversion already, incl. external data being read).

However it happens randomly and only on unix.

Hi:
ext/zlib/tests/gzencode_variation1.phpt failed in Linux(redhat 64-bit)

test log http://pastebin.com/TzpW8GTK

thanks

2011/9/6 Pierre Joye pierre.php@gmail.com:

hi,

It seems that 5.4 and trunk has something broken in zlib and maybe
bas64 encoding/decoding.

We have a serie of tests failing on linux or windows:

ext\zlib\tests\bug_52944.phpt
ext\zlib\tests\gzcompress_basic1.phpt
ext\zlib\tests\gzcompress_variation1.phpt
ext\zlib\tests\gzdeflate_basic1.phpt
ext\zlib\tests\gzdeflate_variation1.phpt
ext\zlib\tests\gzencode_basic1.phpt
ext\zlib\tests\gzencode_variation1-win32.phpt (the unix version too)
ext\zlib\tests\gzfile_basic.phpt
ext\zlib\tests\gzfile_basic2.phpt
\tests\gzfilegzreadfile.phpt
ext\zlib\tests\gzuncompress_basic1.phpt
ext\zlib\tests\readgzfile_basic.phpt
ext\zlib\tests\readgzfile_basic2.phpt

cheers,

Pierre

@pierrejoye | http://blog.thepimp.net | http://www.libgd.org

--

--
Laruence Xinchen Hui
http://www.laruence.com/

--
Pierre

@pierrejoye | http://blog.thepimp.net | http://www.libgd.org

--
Laruence Xinchen Hui
http://www.laruence.com/

14 years ago by Stas Malyshev — view source — reply

unread

Hi!

here you go (from win)

It fails variously on linux or windows. Laruence reproduces the fails
on linux too, some segfault from time to time.

Hmm... Kind of hard to go on "from time to time" - are you sure there's
no build/old library issue? Do we have a reproduction for the segfaults?
The diffs look like your gzip is producing different data - I wonder if
some version might be doing something slightly different.

Stanislav Malyshev, Software Architect
SugarCRM: http://www.sugarcrm.com/
(408)454-6900 ext. 227

14 years ago by Pierre Joye — view source — reply

unread

Hmm... Kind of hard to go on "from time to time" - are you sure there's no
build/old library issue? Do we have a reproduction for the segfaults? The
diffs look like your gzip is producing different data - I wonder if some
version might be doing something slightly different.

Yes, we use the same libs for 5.3 and 5.4 and same tests pass in 5.3.
Bu it is also reproducible on many unrelated hosts.

Laruence will be back tomorrow, I will ask him to provide his diffs
too (and some bt)

--
Pierre

@pierrejoye | http://blog.thepimp.net | http://www.libgd.org

14 years ago by Stas Malyshev — view source — reply

unread

Hi!

Yes, we use the same libs for 5.3 and 5.4 and same tests pass in 5.3.
Bu it is also reproducible on many unrelated hosts.

Well, maybe somebody that could reproduce it could look into it? Because
I'm unable to reproduce any of it...

Stanislav Malyshev, Software Architect
SugarCRM: http://www.sugarcrm.com/
(408)454-6900 ext. 227

14 years ago by Pierre Joye — view source — reply

unread

Hi!

Yes, we use the same libs for 5.3 and 5.4 and same tests pass in 5.3.
Bu it is also reproducible on many unrelated hosts.

Well, maybe somebody that could reproduce it could look into it? Because I'm
unable to reproduce any of it...

Really? three linux (ubuntu latest and debian 5/6 here) as well on windows.

--
Pierre

@pierrejoye | http://blog.thepimp.net | http://www.libgd.org

14 years ago by Hannes Magnusson — view source — reply

unread

Hi!

Yes, we use the same libs for 5.3 and 5.4 and same tests pass in 5.3.
Bu it is also reproducible on many unrelated hosts.

Well, maybe somebody that could reproduce it could look into it? Because I'm
unable to reproduce any of it...

Really? three linux (ubuntu latest and debian 5/6 here) as well on windows.

What is "Ubuntu latest"? is that 11.04 or the current beta?
I'm running Ubuntu 11.04, and none of these fail for me.
Two skipped (win only), and one expected to fail due to bug #47178 it seems.

Please be a darling and post the diff so we can have something to look
at. Maybe seeing the output could jumpstart someones memory and help
you debug it.

-Hannes

14 years ago by Ferenc Kovacs — view source — reply

unread

On Wed, Sep 7, 2011 at 1:26 AM, Hannes Magnusson
hannes.magnusson@gmail.com wrote:

Hi!

Yes, we use the same libs for 5.3 and 5.4 and same tests pass in 5.3.
Bu it is also reproducible on many unrelated hosts.

Well, maybe somebody that could reproduce it could look into it? Because I'm
unable to reproduce any of it...

Really? three linux (ubuntu latest and debian 5/6 here) as well on windows.

What is "Ubuntu latest"? is that 11.04 or the current beta?
I'm running Ubuntu 11.04, and none of these fail for me.
Two skipped (win only), and one expected to fail due to bug #47178 it seems.

Please be a darling and post the diff so we can have something to look
at. Maybe seeing the output could jumpstart someones memory and help
you debug it.

-Hannes

--

those tests passes on my debian squeeze(amd64)

tyrael@phpize:~/checkouts/php-src/trunk$
TEST_PHP_EXECUTABLE=sapi/cli/php sapi/cli/php run-tests.php -g
SKIP,XFAIL,FAIL -n -c tmp-php.ini ext/zlib/

=====================================================================
PHP : sapi/cli/php
`PHP_SAPI` : cli
`PHP_VERSION` : 5.5.0-dev
ZEND_VERSION: 2.4.0
`PHP_OS` : Linux - Linux phpize 2.6.32-5-amd64 #1 SMP Tue Jun 14
09:42:28 UTC 2011 x86_64
INI actual : /home/tyrael/checkouts/php-src/trunk
More .INIs :
CWD : /home/tyrael/checkouts/php-src/trunk
Extra dirs :
VALGRIND : Not used

Running selected tests.
SKIP Test `gzencode()` function : variation
[ext/zlib/tests/gzencode_variation1-win32.phpt] reason: only for
Windows
SKIP Test `gzencode()` function : variation - verify header contents
with all encoding modes
[ext/zlib/tests/gzencode_variation2-win32.phpt] reason: .. only for
Windows
XFAIL Test `gzencode()` function : variation - verify header contents
with all encoding modes [ext/zlib/tests/gzencode_variation2.phpt]
XFAIL REASON: Test will fail until bug #47178 resolved; missing gzip
headers when `FORCE_DEFLATE` specified

Number of tests : 146 144
Tests skipped : 2 ( 1.4%) --------
Tests warned : 0 ( 0.0%) ( 0.0%)
Tests failed : 0 ( 0.0%) ( 0.0%)
Expected fail : 1 ( 0.7%) ( 0.7%)
Tests passed : 143 ( 97.9%) ( 99.3%)

Time taken : 10 seconds

=====================================================================
EXPECTED FAILED TEST SUMMARY

Test `gzencode()` function : variation - verify header contents with all
encoding modes [ext/zlib/tests/gzencode_variation2.phpt] XFAIL
REASON: Test will fail until bug #47178 resolved; missing gzip headers
when `FORCE_DEFLATE` specified

--
Ferenc Kovács
@Tyr43l - http://tyrael.hu

zlib issues in 5.4 and trunk

cheers,

Interestingly enough, none of these fail for me on Darwin. Could you publish the diffs?

cheers,

cheers,

cheers,

Hmm... Kind of hard to go on "from time to time" - are you sure there's no build/old library issue? Do we have a reproduction for the segfaults? The diffs look like your gzip is producing different data - I wonder if some version might be doing something slightly different.

Well, maybe somebody that could reproduce it could look into it? Because I'm unable to reproduce any of it...

Number of tests : 146 144 Tests skipped : 2 ( 1.4%) -------- Tests warned : 0 ( 0.0%) ( 0.0%) Tests failed : 0 ( 0.0%) ( 0.0%) Expected fail : 1 ( 0.7%) ( 0.7%) Tests passed : 143 ( 97.9%) ( 99.3%)

Time taken : 10 seconds

===================================================================== EXPECTED FAILED TEST SUMMARY

Test gzencode() function : variation - verify header contents with all encoding modes [ext/zlib/tests/gzencode_variation2.phpt] XFAIL REASON: Test will fail until bug #47178 resolved; missing gzip headers when FORCE_DEFLATE specified

Interestingly enough, none of these fail for me on Darwin. Could you
publish the diffs?

Hmm... Kind of hard to go on "from time to time" - are you sure there's
no build/old library issue? Do we have a reproduction for the segfaults?
The diffs look like your gzip is producing different data - I wonder if
some version might be doing something slightly different.

Well, maybe somebody that could reproduce it could look into it? Because
I'm unable to reproduce any of it...

Number of tests : 146 144
Tests skipped : 2 ( 1.4%) --------
Tests warned : 0 ( 0.0%) ( 0.0%)
Tests failed : 0 ( 0.0%) ( 0.0%)
Expected fail : 1 ( 0.7%) ( 0.7%)
Tests passed : 143 ( 97.9%) ( 99.3%)

=====================================================================
EXPECTED FAILED TEST SUMMARY

Test `gzencode()` function : variation - verify header contents with all
encoding modes [ext/zlib/tests/gzencode_variation2.phpt] XFAIL
REASON: Test will fail until bug #47178 resolved; missing gzip headers
when `FORCE_DEFLATE` specified