Re: [PHP-DEV] One last ldap_connect headache

Le vendredi 6 novembre 2015, 11:43:58 Ferenc Kovacs a écrit : > > on the other hand functions like ftp_connect assumes the default port when > passed NULL: > > I'm fine with changing this for future versions but my understanding is > that this BC break (along with some others) was introduced in a

Re: [PHP-DEV] PHP 7 Segmentation fault

Hi Dmetry, Thanks a lot. That was very helpful. I was not able to run the script without the database, but I was able to create a small test script without the autoloading and narrow it down to one of two funtions in the database extension that causes the problem. I\ll do some more debugging

Re: [PHP-DEV] PHP 7 Segmentation fault

On Wed, Nov 11, 2015 at 11:24 PM, Frank M. Kromann wrote: > Hi Dmitry, > > Here is the output. > > ==28336== Conditional jump or move depends on uninitialised value(s) > ==28336==at 0x64EF568: tzload (FSTimeZones.c:794) > ==28336==by 0x64EFBC0: fstzZoneFromData (FSTimeZones.c:1765) > ==28

RE: [PHP-DEV] PHP 7 Segmentation fault

Hi Frank, > -Original Message- > From: Frank M. Kromann [mailto:f...@webbypixel.com] > Sent: Wednesday, November 11, 2015 9:29 PM > To: Anatol Belski ; 'Dmitry Stogov' > Cc: 'PHP Internals' > Subject: Re: [PHP-DEV] PHP 7 Segmentation fault > > Hi Anatol, > > I created a new clone of th

Re: [PHP-DEV] PHP 7 Segmentation fault

Hi Anatol, I created a new clone of the repository and I do get the _live_range references. frank@dev Source]$ git clone http://git.php.net/repository/php-src.git php-tmp Cloning into 'php-tmp'... remote: Counting objects: 640667, done. remote: Compressing objects: 100% (128774/128774), done

Re: [PHP-DEV] PHP 7 Segmentation fault

Hi Dmitry, Here is the output. ==28336== Conditional jump or move depends on uninitialised value(s) ==28336==at 0x64EF568: tzload (FSTimeZones.c:794) ==28336==by 0x64EFBC0: fstzZoneFromData (FSTimeZones.c:1765) ==28336==by 0x64EA5ED: fbctzTimeZone (FBCTimeZones.c:51) ==28336==by

Re: [PHP-DEV] PHP 7 Segmentation fault

Hi Anatol, That's odd. I switched back to master and the references to _live_range() shows up again. I'll try to do a clean checkout and see what happens. Still on the PHP-7.0 branch with USE_ZEND_alloc=0 set in the environment I get this backtrace. #0 0x74786563 in malloc_consolida

Re: [PHP-DEV] PHP 7 Segmentation fault

I added zend_add_live_range() into master a day ago and replaced it with zend_start_live_range/zend_end_live_range today. Thanks. Dmitry. On Wed, Nov 11, 2015 at 11:02 PM, Anatol Belski wrote: > > > > -Original Message- > > From: Frank M. Kromann [mailto:f...@webbypixel.com] > > Sent: W

Re: [PHP-DEV] PHP 7 Segmentation fault

try to run the same script with valgrind. USE_ZEND_ALLOC=0 valgrind --db-attach=yes php test.php (it is 10 times slower) Show the first reported problem. Thanks. Dmitry. On Wed, Nov 11, 2015 at 10:51 PM, Frank M. Kromann wrote: > Just switched to PHP-7.0 and there is no longer any references

RE: [PHP-DEV] PHP 7 Segmentation fault

> -Original Message- > From: Frank M. Kromann [mailto:f...@webbypixel.com] > Sent: Wednesday, November 11, 2015 8:51 PM > To: Anatol Belski ; 'Dmitry Stogov' > Cc: 'PHP Internals' > Subject: Re: [PHP-DEV] PHP 7 Segmentation fault > > Just switched to PHP-7.0 and there is no longer any

Re: [PHP-DEV] PHP 7 Segmentation fault

Just switched to PHP-7.0 and there is no longer any references to _live_range but the problem with the segfault is still there. Here is a new backtrace. #0 zend_mm_alloc_small (size=, bin_num=, heap=) at /home/frank/Source/php-src-7/Zend/zend_alloc.c:1291 #1 zend_mm_alloc_heap (size=, heap=)

Re: [PHP-DEV] PHP 7 Segmentation fault

Hi Anatol, Sorry, I fogot to include the output of the grep. [frank@dev php-src-7]$ grep -rn _live_range Zend/ Zend/zend_opcode.c:83:op_array->last_live_range = 0; Zend/zend_compile.h:176:typedef struct _zend_live_range { Zend/zend_compile.h:179:} zend_live_range; Zend/zend_compile.h:364:

Re: [PHP-DEV] PHP 7 Segmentation fault

Hi Anatol, Here is what I have [frank@dev php-src-7]$ git show commit d1077f7a897d9267a0cad3d64663fd952dd1c6b2 Merge: e88e3f2 a2e59e7 Author: Anatol Belski Date: Wed Nov 11 16:35:07 2015 +0100 Merge branch 'PHP-7.0' * PHP-7.0: fix format macro names in UPGRADING.INTERNALS [fr

RE: [PHP-DEV] PHP 7 Segmentation fault

Hi Frank, > -Original Message- > From: Frank M. Kromann [mailto:f...@php.net] > Sent: Wednesday, November 11, 2015 8:14 PM > To: Dmitry Stogov > Cc: PHP Internals > Subject: Re: [PHP-DEV] PHP 7 Segmentation fault > > Hi Dmitry, > > Yes this is the master branch. I have previously teste

Re: [PHP-DEV] PHP 7 Segmentation fault

Hi Dmitry, Yes this is the master branch. I have previously tested it on the PHP-7.0 branch with the same result. I will do that again to verify that the problem is the same. I run vcsclean, buildconf, configure, make at least once a day. I do not run the make install as that will break all

[PHP-DEV] Benchmark Results for PHP Master 2015-11-11

Results for project PHP master, build date 2015-11-11 05:25:48+02:00 commit: baf97b1fcc0f8458955f33bcfd325e3130e1161f revision date: 2015-11-10 21:48:03+03:00 environment:Haswell-EP cpu:Intel(R) Xeon(R) CPU E5-2699 v3 @ 2.30GHz 2x18 cores, stepping 2, LLC 45 MB mem

Re: [PHP-DEV] Null bytes in anonymous class names

On Wed, 11 Nov 2015, Nikita Popov wrote: > Using a NUL byte conveniently achieves both of these goals. We did not > however take into account that this would cause issues with 3rd party > tooling that does not support binary data. I withdraw my comment. It does not cause issues for me (anymore)

Re: [PHP-DEV] PHP7 / foreach / references / ugly code / discrepancy to PHP 5.6

Should this be in the UPGRADING notes? > On 10.11.2015, at 20:45, Nikita Popov wrote: > > On Tue, Nov 10, 2015 at 5:48 PM, Philip Hofstetter < > phofstet...@sensational.ch> wrote: > >> Hi, >> >> I'm having a cause of slightly ugly code that runs differently from PHP 5.6 >> to PHP 7 and I don'

[PHP-DEV] Re: Safe Temporary Cleanup on Exception (Variable Liveness)

It looks like the last version of the patch https://github.com/php/php-src/pull/1634 solves all the problems (except for reported by Nikita "unknown liveness in finally" - Zend/tests/temporary_cleaning_010.phpt). Please, analyse the patch and

Re: [PHP-DEV] PHP 7 Segmentation fault

I suppose this is not PHP-7.0, but "master" branch. PHP-7.0 doesn't have zend_add_live_range(). Please, make "git pull" and then full rebuild, "make clean; make; make install" Thanks. Dmitry. On Wed, Nov 11, 2015 at 3:22 AM, Frank M. Kromann wrote: > Hello Everyone, > > I have a large applicat

Re: [PHP-DEV] Null bytes in anonymous class names

Steven Hilder wrote on 11/11/2015 02:13: This behaviour was present in Joe's original implementation, but later removed[3] by Nikita. His commit message says that this was "dead code", but I don't see what was wrong with it. It looks to me like 'decl->name' will never be set for an anonymous

Re: [PHP-DEV] Re: Resource typehint and return type

On Wed, Nov 11, 2015 at 3:29 AM, Ben Scholzen 'DASPRiD' wrote: > On 11.11.2015 03:21, Andrea Faulds wrote: > >> The reason for this is that the resource type is an anachronism from an >> age in which PHP did not have objects, yet still needed to make certain >> types of data opaque. The resource

Re: [PHP-DEV] Null bytes in anonymous class names

On Tue, Nov 10, 2015 at 11:04 PM, Steven Hilder < steven.hilder@sevenpercent.solutions> wrote: On Mon, 09 Nov 2015 16:48:57, Leigh wrote: > >> On 9 November 2015 at 15:27, Steven Hilder >> Can you share your patch? >> > > See > https://github.com/php/php-src/compare/master...stevenhilder:hide-ano