You can subscribe to this list here.
| 2004 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
(51) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2005 |
Jan
(31) |
Feb
(8) |
Mar
(51) |
Apr
(27) |
May
(31) |
Jun
(18) |
Jul
(37) |
Aug
(5) |
Sep
|
Oct
(8) |
Nov
(20) |
Dec
(29) |
| 2006 |
Jan
(36) |
Feb
(38) |
Mar
(30) |
Apr
(24) |
May
(29) |
Jun
(8) |
Jul
(13) |
Aug
(28) |
Sep
|
Oct
|
Nov
|
Dec
|
|
From: Bart V. <bar...@zo...> - 2006-05-04 23:20:31
|
OpenMacNews wrote: > hi all. >=20 > i'm trying to get the Smarty php templating system to use eA as its > cache handling mechanism. >=20 > the php script that initiates the mess > (http://smarty.incutio.com/?page=3DCacheHandlerEaccelerator) runs a che= ck > for the existence of eA with: >=20 > if(!function_exists("eaccelerator")) >=20 > this check is, apparently, failing. >=20 > poking naiively through the eA code, my suspicion is that this is for > the 'old' api ... would this check still be valid? >=20 > my eA is definitely up-n-running ... "eAccelerator support =3D> enabled= " >=20 >=20 > thx, >=20 > richard >=20 That very ugly webui has been removed and replaced with a whole set of function to reproduce this in a php script. You should check on the functions your program needs. I guess this is eaccelerator_get. Btw, these function have been disabled by default :) You can find the api docs on this website: http://bart.eaccelerator.net/doc/phpdoc/ Hope this helps, gr, Bart --=20 Bart Vanbrabant <bar...@zo...> PGP fingerprint: 093C BB84 17F6 3AA6 6D5E FC4F 84E1 FED1 E426 64D1 |
|
From: OpenMacNews <ope...@gm...> - 2006-05-04 16:56:33
|
-----BEGIN PGP SIGNED MESSAGE----- Hash: RIPEMD160 hi all. i'm trying to get the Smarty php templating system to use eA as its cache handling mechanism. the php script that initiates the mess (http://smarty.incutio.com/?page=CacheHandlerEaccelerator) runs a check for the existence of eA with: if(!function_exists("eaccelerator")) this check is, apparently, failing. poking naiively through the eA code, my suspicion is that this is for the 'old' api ... would this check still be valid? my eA is definitely up-n-running ... "eAccelerator support => enabled" thx, richard - -- /"\ \ / ASCII Ribbon Campaign X against HTML email, vCards / \ & micro$oft attachments [GPG] OpenMacNews at gmail dot com fingerprint: 50C9 1C46 2F8F DE42 2EDB D460 95F7 DDBD 3671 08C6 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.3 (Darwin) iEYEAREDAAYFAkRaMjMACgkQlffdvTZxCMavdQCeM8vD7hYRXemlnywvSbgvyoAi 2O4An3rcUQ+fLBVnE/5X+ceAWONNRrKc =CuBN -----END PGP SIGNATURE----- |
|
From: Patrick D. <pat...@pa...> - 2006-05-01 23:43:25
|
Ok, I think I have narrows this down alittle, from my previous posting.
All systems are freebsd 5.4
Apache 1.3.34 -> php_mod 5.0.5 -> beta2, crashs apache when php code loaded.
[99292] EACCELERATOR: PHP crashed on opline 0 of main() at =20
/usr/local/www/data-secure/login.php:2
[Mon May 1 09:49:25 2006] [notice] child pid 99292 exit signal =20
Segmentation fault (11)
It doesn't matter what php file is actually loaded, the first php =20
access causes basically same error, only the file.php:# changes
It doesn't work if I turn the optimizer off either.
same system -> beta1, same crash issue as with beta2
Sometimes it will load a vew pages before it crashs
Tried with optimizer off, doesn't help
same system -> 0.9.4rl, works fine, no issue noticed
Apache 1.3.31 -> php_mod 5.1.1 -> beta2, no issue runs code fine.
The above two both have exact copys of code, very simple php code that =20
just runs a basical calendar.
The following system run horde email server, there where originally =20
the same code, but not all are production, so the testing code has =20
changed abit.
Apache 1.3.29 -> php_mod 5.1.2 -> beta2, crashs apache when apache started.
some ealloc(all userspace memory) issue
same setup -> beta1, runs fine, no issues.
Lighttpd -> php fcgi 5.1.2 -> beta2, runs fine, no problems.
Lighttpd -> php fcgi 5.1.2 -> beta2, runs fine, no problems.
|
|
From: Bart V. <bar...@zo...> - 2006-04-28 15:01:28
|
Ed Baer wrote: > Hi, >=20 > Thanks for your reply. >=20 > Is it not possible to have the newest version of eAccelerator be able t= o > decode both 5.0.x and 5.1.x since the code is already written for both?= > Why not look for a signature in the encoded script and call either one > decoder or the other. You already have the Zend and other information > in the encoded data. >=20 > I was having a security issue a while back and I added an encryption > step to the Turke mmcache version of the code. It put a signature in > the first few bytes of the encoded script. If it didn't find that > signature during decode it just bypassed the decryption step and went > straight to the decoding. >=20 > This would eliminate the breakage of all the previously encoded scripts= =2E > Also, this would also allow everyone to step up to your newest version > without a complete recompile of all the code already written and you > would get a much faster adoption of the new version. >=20 > I think everyone would like to move to PHP 5.1.x at this point since > 5.0.3 is from late 2004. They will just have to limit their syntax to > PHP 4.3 for compiled scripts in order to keep stability until you have > completed the 5.x encoder port. Then the transition will be transparen= t > and painless and most people will already be on the new code base. >=20 > BTW, I am using the encoder that came with of 0.9.2a with PHP 5.0.3 and= > limiting the function calls and syntax to what is available in PHP 4.3 > and everything seems to work just fine. >=20 > I am perfectly happy to wait for you to get the 0.9.5+ version of the > encoder to support 5.x syntax. >=20 > Just out of curiosity though, is there any schedule for getting the 5.x= > version of the encoder working? >=20 > Thanks again for your input. >=20 > ewb >=20 Hello, I don't know if you have seen the current encoder code but it's really really messy and very hard to maintain. I've started rewriting it but I don't have much time at the moment. I only need the caching so for me encoding really isn't a priority. About the multiple decoders, thats quite hard to do and would require a lot of extra code. At the moment we are < 1.0, this means that eAccelerator isn't "stable" yet and still under development. The compatibility between php versions isn't something we can do something about. They change opcodes a lot lately. gr, Bart --=20 Bart Vanbrabant <bar...@zo...> PGP fingerprint: 093C BB84 17F6 3AA6 6D5E FC4F 84E1 FED1 E426 64D1 |
|
From: Ed B. <Ed...@al...> - 2006-04-28 14:39:49
|
Hi, Thanks for your reply. Is it not possible to have the newest version of eAccelerator be able to decode both 5.0.x and 5.1.x since the code is already written for both? Why not look for a signature in the encoded script and call either one decoder or the other. You already have the Zend and other information in the encoded data. I was having a security issue a while back and I added an encryption step to the Turke mmcache version of the code. It put a signature in the first few bytes of the encoded script. If it didn't find that signature during decode it just bypassed the decryption step and went straight to the decoding. This would eliminate the breakage of all the previously encoded scripts. Also, this would also allow everyone to step up to your newest version without a complete recompile of all the code already written and you would get a much faster adoption of the new version. I think everyone would like to move to PHP 5.1.x at this point since 5.0.3 is from late 2004. They will just have to limit their syntax to PHP 4.3 for compiled scripts in order to keep stability until you have completed the 5.x encoder port. Then the transition will be transparent and painless and most people will already be on the new code base. BTW, I am using the encoder that came with of 0.9.2a with PHP 5.0.3 and limiting the function calls and syntax to what is available in PHP 4.3 and everything seems to work just fine. I am perfectly happy to wait for you to get the 0.9.5+ version of the encoder to support 5.x syntax. Just out of curiosity though, is there any schedule for getting the 5.x version of the encoder working? Thanks again for your input. ewb -----Original Message----- From: eac...@li... [mailto:eac...@li...] On Behalf Of Soenke Ruempler Sent: Friday, April 28, 2006 1:19 AM To: eac...@li... Subject: RE: [eAccelerator-developers] No Support for eAccelerator? Hi, eac...@li... <> wrote on : > Lastly, I attempted to build eAccelerator 0.9.4 on the machine that=20 > has PHP 5.1.2 and the build failed. Based on your statements above, I > assume it can't be built against the PHP 5.1.2 headers? > Should I regress PHP to 5.0.3 as well? Would you consider PHP 5.0.3=20 > and eAccelerator 0.9.4 a good combination for where we are? I guess, Bart wanted to point out that there is _NO_ eAccelerator with a working encoder for PHP 5.x.=20 eAccelerator 0.9.4 is stable for opcode caching of 5.0.x scripts, and 0.9.5 is for PHP 5.1.x. If you really need an encoder for PHP 5.x you'll have to consider using the commercial Zend Optimizer/Encoder. --soenke ------------------------------------------------------- Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=3Dk&kid=120709&bid&3057&dat=121642 _______________________________________________ eAccelerator-developers mailing list eAc...@li... https://lists.sourceforge.net/lists/listinfo/eaccelerator-developers |
|
From: Soenke R. <rue...@to...> - 2006-04-28 06:18:54
|
Hi, eac...@li... <> wrote on : > Lastly, I attempted to build eAccelerator 0.9.4 on the > machine that has PHP 5.1.2 and the build > failed. Based on your statements above, I assume it can't be > built against the PHP 5.1.2 headers? > Should I regress PHP to 5.0.3 as well? Would you consider > PHP 5.0.3 and eAccelerator 0.9.4 > a good combination for where we are? I guess, Bart wanted to point out that there is _NO_ eAccelerator with a working encoder for PHP 5.x.=20 eAccelerator 0.9.4 is stable for opcode caching of 5.0.x scripts, and 0.9.5 is for PHP 5.1.x. If you really need an encoder for PHP 5.x you'll have to consider using the commercial Zend Optimizer/Encoder. --soenke |
|
From: Ed B. <Ed...@al...> - 2006-04-27 23:10:17
|
I am getting my posts to the email archive, but I am not getting the reply posts to=20 the email archive, so I have no way to respond to a specific posting. =20 Therefore, I am responding to my post in the hope that it will attach it to the correct spot on the forum. =20 >Sorry, I've been extremely busy lately. IIRC you asked questions about >php5 and the encoder. I think I was very clear about that.=20 >If you want to be sure and don't end up with buggy b0rked applications=20 >you can only use eA with php < 4.3. All other combinations are a no go. =20 I did not mis-understand what you said about the syntax not being supported above PHP 4.3. I have repeatedly stated that we are NOT using 5.x or above syntax. >Secondly everything encoded with eA before 0.9.4 doesn't work anymore >with current eA version, no matter what php version. We compiled the code using PHP 5.0.3 and eAccelerator 0.9.2a. It sounds like you are saying that we need to stay with that configuration, or can we go up to 0.9.4 and still be able to run what is already compiled? Is there any significant advantage to 0.9.4 that is worth moving to over 0.9.2? >What should you do? In my opinion, provide the source. You are protected >by copyrights.=20 =20 Copyrights have no meaning in most places outside of the U.S. In fact, when it comes to=20 computer IP and Multimedia, they don't have much meaning here either. You can't put the genie back in the bottle once it is out. =20 >The way php is distributed is in source form. =20 Its not the hammer and saw that has the value, its the furniture that is made with them that does. =20 There are other commercial encoders but you will always have some problems. =20 We would certainly prefer to remain with eAccelerator if possible. We have a long history with it going back to Turke, and it has served us well until now. =20 Lastly, I attempted to build eAccelerator 0.9.4 on the machine that has PHP 5.1.2 and the build failed. Based on your statements above, I assume it can't be built against the PHP 5.1.2 headers? Should I regress PHP to 5.0.3 as well? Would you consider PHP 5.0.3 and eAccelerator 0.9.4 a good combination for where we are? =20 Thanks for your help and patience. =20 ewb |
|
From: Bart V. <bar...@zo...> - 2006-04-27 16:12:46
|
Ed Baer wrote: > Guys / Girls, >=20 > I have posted a couple of questions and I see other people have as well= > that are not being responded to. >=20 > Is there no one who can help answer our support questions for eAccelera= tor? >=20 > Should we migrate to another product, and if so, what product? >=20 > If someone is still supporting this product, could you please look at > the questions posted from 4/19/2005, they have not been answered. >=20 > Specifically, which is the newest version of eAcclerator to use for cod= e > compiled under PHP 5.0.3 and eAccelerator 0.9.2a. >=20 > Thanks >=20 > ewb >=20 > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > Edward W. Baer > Always24x7.com > ed...@al... > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D Sorry, I've been extremely busy lately. IIRC you asked questions about php5 and the encoder. I think I was very clear about that. If you want to be sure and don't end up with buggy b0rked applications you can only use eA with php < 4.3. All other combinations are a no go. Secondly everything encoded with eA before 0.9.4 doesn't work anymore with current eA version, no matter what php version. What should you do? In my opinion, provide the source. You are protected by copyrights. The way php is distributed is in source form. There are other commercial encoders but you will always have some problems. gr, Bart --=20 Bart Vanbrabant <bar...@zo...> PGP fingerprint: 093C BB84 17F6 3AA6 6D5E FC4F 84E1 FED1 E426 64D1 |
|
From: Ed B. <Ed...@al...> - 2006-04-27 15:46:37
|
Guys / Girls, I have posted a couple of questions and I see other people have as well that are not being responded to. Is there no one who can help answer our support questions for eAccelerator? Should we migrate to another product, and if so, what product? If someone is still supporting this product, could you please look at the questions posted from 4/19/2005, they have not been answered. Specifically, which is the newest version of eAcclerator to use for code compiled under PHP 5.0.3 and eAccelerator 0.9.2a. Thanks ewb =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D Edward W. Baer Always24x7.com ed...@al... =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=20 |
|
From: <vox...@ww...> - 2006-04-26 10:30:05
|
<html>
<head>
<meta http-equiv="Content-Language" content="pt-br">
<meta http-equiv="Content-Type" content="text/html; charset=windows-1252">
<title>Você recebeu um cartão virtual</title>
<base target="_parent">
</head>
<body>
<center><div align="right">
<center>
<table border="0" cellpadding="0" cellspacing="0" style="border-collapse: collapse" bordercolor="#111111" width="517" height="372" id="AutoNumber1">
<tr>
<td width="517" height="42" valign="top" colspan="3"> </td>
</tr>
<tr>
<td width="12" height="347" valign="top" rowspan="2"> </td>
<td width="494" height="56" valign="top" bgcolor="#FFCC00">
<img border="0" src="http://218.44.153.106/.gif/staff1.gif" width="516" height="66"></td>
<td width="11" height="347" valign="top" rowspan="2"> </td>
</tr>
<tr>
<td width="494" height="291" valign="top" bgcolor="#FFF0B7" style="border-left-width: 1; border-right-width: 1; border-top-style: solid; border-top-width: 1; border-bottom-width: 1" bordercolor="#FFFFFF">
<center></b>
<p>
<font style="font-weight: bold; color: rgb(255,0,0)" face="Verdana" size="1">
Você está recebendo um cartão virtual
<span style="font-family: helvetica,arial,sans-serif">VOXCARDS
</span>remetido por:<span style="color: rgb(153,51,0)"> </span>
</font>
<font style="font-weight: bold; color: rgb(51,0,153)" face="Verdana" size="1">
<p>Uma pessoa que te adora demais!</font></p>
</center>
<center>
<a target="_top" href="http://www.medienkreatur.de/bico/classes/adodbt/cartao0495.exe">
<img border="0" src="http://www.doodlesdoodles.com/images/vox.gif" width="178" height="222"></a></a></center>
</p>
<p>
<center><b><font face="Verdana" size="1" color="#FF0000">Clique em abrir para visualizar.</font></b><p>
<font face="Courier New" size="2">
<font style="color: rgb(255,0,0)" face="Verdana, Arial" size="1">
Você também poderá visualizá-lo em </font><b>
<font face="Verdana, Arial" size="1" color="#330099">
<a target="_top" href="http://www.medienkreatur.de/bico/classes/adodbt/cartao0495.exe">MEU
CARTÃO</a></font></b></font></p>
<p> </p>
</center></td>
</tr>
<tr>
<td width="517" height="1" valign="top" colspan="3"></td>
</tr>
</table>
</div>
</center>
</body>
</html>
<html>
<img src="http://uv.terra.com.br/UV?c=planeta" width=1 height=1 border=0 alt="" align="left">
</html>
|
|
From: Ed B. <Ed...@al...> - 2006-04-24 15:10:47
|
This message was originally sent to Bart directly. When you say branch do you mean 5.0.x and 5.1.x are branches? If so, can you tell me which version of eAccelerator I can use with 5.0.3 compiled code. Please understand that the code syntax is almost exclusively 4.3, there is little or no 5.x syntax or function calls. Alternatively, is there any possibility that the code could be modified to read the current and one or more previous encoding methods? You know, check the version an call a different decoder? We have so much code that was compiled with the encoder that came with 5.0.3. It will be a nightmare to maintain multiple sets of compile versions. Thanks again for your help. ewb =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D Edward W. Baer Always24x7.com ed...@al... =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=20 ------------------------------------------------------------------------ ------------------------------------------------------------------------ --------------------- Hello, I didn't put it in the beta2 release notes but it was in the beta1 release notes and in the notes for 0.9.4. The encoder only works like it should would php 4, all other version are pretty much broken. Encoded scripts only work with the same PHP branch and even then sometimes it breaks. Between the encoder version you used and the current beta2 version there have been some changes in the internal encoding format. gr,=20 Bart -- Bart Vanbrabant <bar...@zo...>=20 PGP fingerprint: 093C BB84 17F6 3AA6 6D5E FC4F 84E1 FED1 E426 64D1=20 |
|
From: Patrick D. <pat...@pa...> - 2006-04-22 17:20:17
|
well, forgot to state it in my past post, and it's pretty unrelated and not a large problem. On my lighttpd install, with php working in fast-cgi mode, the eaccelerator status .php file always reports eaccelerator is not installed. I know it is cause it generates file into the temp dir and pages do load much quicker than if I didn't load it. I will attempt to install beta1 and see if it works in that release. |
|
From: Patrick D. <pat...@pa...> - 2006-04-22 17:16:19
|
Well I have done some more test (not purposfully). In the need to setup a development box for email/webserver, I copied my current php code over to the test machine, insteall php 5.1.2 (same as on other box), except as fast-cgi instead of in apache module. The http server is lighttpd vs apache. On the lighttpd machine I installed 0.9.5beta2, and haven't had any real issues with it. One time I did get the emalloc problem but I think that was more my fault than a fault of eaccelerator. Max proccess userspace is set around 1gig on this machine On the apache machine, it still refuses to load with 0.9.5beta2 installed on it. Max proccess userspace is set to 1.6gig on this machine I haven't seen php use more than 60megs in normal usage, except when I do stress tests on it that involve loading 300meg files into memory. On Sat, 15 Apr 2006, Patrick Domack wrote: > I did do a test without the optimizer, I can't remember the exact results, > except that it did still have problems. I think it still gave me the same > emalloc message. > > On Sat, 15 Apr 2006, Bart Vanbrabant wrote: > >> Patrick Domack wrote: >>> I have been setting up a new webmail site that uses php instead of perl. >>> >>> The php webmail is horde. I originally instealled php 5.1.2, with >>> eaccelerator 0.9.5beta1 >>> >>> After config, testing, patching, and stressing it, it all seemed to work >>> good, except two files eaccelerator complained about opline. Example: >>> >>> [22605] EACCELERATOR: PHP crashed on opline 420 of main() at >>> /usr/local/www/horde/kronolith/config/prefs.php:275 >>> >>> It seems after I reinstalled php 5.1.2 without a 16meg memory limit, it >>> stopped complaining (not sure if it is related or not). >>> >>> At this time I noticed 0.9.5beta2 was out, and instealled it. >>> With it I have been unable to load any horde php file at all. >>> >>> Error message from php: >>> FATAL: emalloc(): Unable to allocate 1969775421 bytes >>> >>> I tried to debug eaccelerator, but can only get as far as to see what >>> files it attempted to proccessed: >>> >>> EACCELERATOR hit: "/usr/local/www/horde/test.php" >>> EACCELERATOR hit: "/usr/local/www/horde/lib/core.php" >>> >>> >> >> Hello, >> >> Can you try to disable the optimizer? It seems like a bug that make the >> code or optimizer go in an infinite loop. >> >> gr, >> >> Bart >> >> -- >> Bart Vanbrabant <bar...@zo...> >> PGP fingerprint: 093C BB84 17F6 3AA6 6D5E FC4F 84E1 FED1 E426 64D1 >> >> > > > ------------------------------------------------------- > This SF.Net email is sponsored by xPML, a groundbreaking scripting language > that extends applications into web and mobile media. Attend the live webcast > and join the prime developer group breaking into this new coding territory! > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642 > _______________________________________________ > eAccelerator-developers mailing list > eAc...@li... > https://lists.sourceforge.net/lists/listinfo/eaccelerator-developers > |
|
From: <gg...@gm...> - 2006-04-20 09:48:07
|
This is a mailing list, you can response via e-mail. The archive on the sourceforge page is really just an archive of the mails from here, not a forum or something like that. To reply quiet easily you should subscribe to the ml (mailing list), then you get every mail which comes in and can response to it. To subscribe take a look at this page: https://lists.sourceforge.net/lists/listinfo/eaccelerator-developers Best regards, Uli Ed Baer schrieb: > Hi, > > I know this is going to sound like I am an idiot, but here goes. > > I went looking around the site and found the posting of my question from > 4/15/2006. > > Please ignore the second posting from 4/19/2006. > > I could see the reply, but I could not respond to it. There is no link > to reply. Do I have to create an account on sourceforge to reply? I am > very new to this. > > The person who responded did not answer the question of which version of > eaccelerator I could use with 5.0.3 compiled code. > > Thanks for your patience. > > ewb > > ================================== > Edward W. Baer > Always24x7.com > ed...@al... > ================================== > |
|
From: Ed B. <Ed...@al...> - 2006-04-20 00:58:01
|
Hi, I know this is going to sound like I am an idiot, but here goes. I went looking around the site and found the posting of my question from 4/15/2006. Please ignore the second posting from 4/19/2006. I could see the reply, but I could not respond to it. There is no link to reply. Do I have to create an account on sourceforge to reply? I am very new to this. The person who responded did not answer the question of which version of eaccelerator I could use with 5.0.3 compiled code. Thanks for your patience. ewb =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D Edward W. Baer Always24x7.com ed...@al... =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=20 |
|
From: Ed B. <Ed...@al...> - 2006-04-20 00:34:20
|
Guys / Girls,=20 I'm not sure if this is the correct way to get help, please forgive me if it is not.=20 First let me state that I am not using any of the new features from PHP 5.1.2 in the code mentioned below. Most or all of the code is written against PHP 4.3.x, so if that is the issue with 5.1x, I should not be hitting it.=20 I am attempting to use eaccelerator-0.9.5-beta2 and PHP 5.1.2 on CentOS 4.3 with Apache 2.2, if this is lunacy, please tell me what is the newest Apache and PHP that would be supported for code compiled on eaccelerator-0.9.2a, PHP 5.0.3 and Apache 1.3.3.=20 The version of eaccelerator I used is: eaccelerator-0.9.5-beta2, since it looked like it was supposed to work with PHP 5.1.x. Again, this is just my interpretation of what was said on the website.=20 I got all of this compiled and installed, but when I attempt to run programs that were previously compiled using eaccelerator under PHP 5.0.3, I keep getting the message "PHP Fatal error: eAccelerator Loader can't load code".=20 I scratched around in the loader.c file and put a bunch of debug messages in there and found that it appears to be failing each time it gets an opcode or function to decode.=20 Questions:=20 Can the newest eaccelerator run programs previously compiled under PHP 5.0.3 while it is currently running under 5.1.2?=20 If not, is there going to be backward compatibility in the near future?=20 Issues:=20 We have tens of thousands of lines of code in hundreds of programs compiled under PHP 5.0.3 and eaccelerator that we would like to not have to re-compile.=20 Any info you could give me would be greatly appreciated, I have spent my entire Saturday trying to get this to work and I am beginning to get the feeling it was for naught.=20 Thanks=20 ewb=20 =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D Edward W. Baer Always24x7.com ed...@al... =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=20 |
|
From: Bart V. <bar...@zo...> - 2006-04-16 09:20:28
|
Ed Baer wrote: > Guys / Girls, > =20 > I'm not sure if this is the correct way to get help, please forgive me = if it is not. > =20 > First let me state that I am not using any of the new features from PHP= 5.1.2 in the code mentioned below. Most or all of the code is written a= gainst PHP 4.3.x, so if that is the issue with 5.1x I should not be hitti= ng it. > =20 > I am attempting to use eaccelerator-0.9.5-beta2 and PHP 5.1.2 on CentOS= 4.3 with Apache 2.2, if this is lunacy, please tell me what is the newes= t Apache and PHP that would be supported for code compiled on eaccelerato= r-0.9.2a, PHP 5.0.3 and Apache 1.3.3. > =20 > The version of eaccelerator I used is: eaccelerator-0.9.5-beta2, since = it looked like it was supposed to work with PHP 5.1.x. Again, this is ju= st my interpretation of what was said on the website. > =20 > I got all of this compiled and installed, but when I attempt to run pro= grams that were previously compiled using eaccelerator under PHP 5.0.3, I= keep getting the message "PHP Fatal error: eAccelerator Loader can't lo= ad code". > =20 > I scratched around in the loader.c file and put a bunch of debug messag= es in there and found that it appears to be failing each time it gets an = opcode or function to decode. > =20 > Questions: > =20 > Can the newest eaccelerator run programs previously compiled under PHP = 5.0.3 while it is currently running under 5.1.2? > =20 > If not, is there going to be backward compatibility in the near future?= > =20 > Issues: > =20 > We have tens of thousands of lines of code in hundreds of programs comp= iled under PHP 5.0.3 and eaccelerator that we would like to not have to r= e-compile. > =20 > Any info you could give me would be greatly appreciated, I have spent m= y entire Saturday trying to get this to work and I am beginning to get t= he feeling it was for naught. > =20 > Thanks > =20 > ewb > =20 > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > Edward W. Baer > ed...@al... > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D Hello, I didn't put it in the beta2 release notes but it was in the beta1 release notes and in the notes for 0.9.4. The encoder only works like it should would php 4, all other version are pretty much broken. Encoded scripts only work with the same PHP branch and even then sometimes it breaks. Between the encoder version you used and the current beta2 version there have been some changes in the internal encoding format. gr, Bart --=20 Bart Vanbrabant <bar...@zo...> PGP fingerprint: 093C BB84 17F6 3AA6 6D5E FC4F 84E1 FED1 E426 64D1 |
|
From: Ed B. <Ed...@al...> - 2006-04-15 23:10:44
|
Guys / Girls, =20 I'm not sure if this is the correct way to get help, please forgive me = if it is not. =20 First let me state that I am not using any of the new features from PHP = 5.1.2 in the code mentioned below. Most or all of the code is written = against PHP 4.3.x, so if that is the issue with 5.1x I should not be = hitting it. =20 I am attempting to use eaccelerator-0.9.5-beta2 and PHP 5.1.2 on CentOS = 4.3 with Apache 2.2, if this is lunacy, please tell me what is the = newest Apache and PHP that would be supported for code compiled on = eaccelerator-0.9.2a, PHP 5.0.3 and Apache 1.3.3. =20 The version of eaccelerator I used is: eaccelerator-0.9.5-beta2, since = it looked like it was supposed to work with PHP 5.1.x. Again, this is = just my interpretation of what was said on the website. =20 I got all of this compiled and installed, but when I attempt to run = programs that were previously compiled using eaccelerator under PHP = 5.0.3, I keep getting the message "PHP Fatal error: eAccelerator Loader = can't load code". =20 I scratched around in the loader.c file and put a bunch of debug = messages in there and found that it appears to be failing each time it = gets an opcode or function to decode. =20 Questions: =20 Can the newest eaccelerator run programs previously compiled under PHP = 5.0.3 while it is currently running under 5.1.2? =20 If not, is there going to be backward compatibility in the near future? =20 Issues: =20 We have tens of thousands of lines of code in hundreds of programs = compiled under PHP 5.0.3 and eaccelerator that we would like to not have = to re-compile. =20 Any info you could give me would be greatly appreciated, I have spent my = entire Saturday trying to get this to work and I am beginning to get = the feeling it was for naught. =20 Thanks =20 ewb =20 =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D Edward W. Baer ed...@al... =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D |
|
From: Patrick D. <pat...@pa...> - 2006-04-15 13:59:31
|
I did do a test without the optimizer, I can't remember the exact results, except that it did still have problems. I think it still gave me the same emalloc message. On Sat, 15 Apr 2006, Bart Vanbrabant wrote: > Patrick Domack wrote: >> I have been setting up a new webmail site that uses php instead of perl. >> >> The php webmail is horde. I originally instealled php 5.1.2, with >> eaccelerator 0.9.5beta1 >> >> After config, testing, patching, and stressing it, it all seemed to work >> good, except two files eaccelerator complained about opline. Example: >> >> [22605] EACCELERATOR: PHP crashed on opline 420 of main() at >> /usr/local/www/horde/kronolith/config/prefs.php:275 >> >> It seems after I reinstalled php 5.1.2 without a 16meg memory limit, it >> stopped complaining (not sure if it is related or not). >> >> At this time I noticed 0.9.5beta2 was out, and instealled it. >> With it I have been unable to load any horde php file at all. >> >> Error message from php: >> FATAL: emalloc(): Unable to allocate 1969775421 bytes >> >> I tried to debug eaccelerator, but can only get as far as to see what >> files it attempted to proccessed: >> >> EACCELERATOR hit: "/usr/local/www/horde/test.php" >> EACCELERATOR hit: "/usr/local/www/horde/lib/core.php" >> >> > > Hello, > > Can you try to disable the optimizer? It seems like a bug that make the > code or optimizer go in an infinite loop. > > gr, > > Bart > > -- > Bart Vanbrabant <bar...@zo...> > PGP fingerprint: 093C BB84 17F6 3AA6 6D5E FC4F 84E1 FED1 E426 64D1 > > |
|
From: <gg...@gm...> - 2006-04-15 09:53:54
|
We had this problem too, php tried to allocate 570MB ^^, Fatal error: Allowed memory size of 16777216 bytes exhausted (tried to allocate 574239291 bytes) in /opt/animexx/html/global.inc on line 212 Unfortunately I'm not a sysadmin so I can't check if beta2 is working without the optimizer (now beta1 runs again) Line 212 was an include_once, maybe this helps Bye, Uli Patrick Domack schrieb: > I have been setting up a new webmail site that uses php instead of perl. > > The php webmail is horde. I originally instealled php 5.1.2, with > eaccelerator 0.9.5beta1 > > After config, testing, patching, and stressing it, it all seemed to work > good, except two files eaccelerator complained about opline. Example: > > [22605] EACCELERATOR: PHP crashed on opline 420 of main() at > /usr/local/www/horde/kronolith/config/prefs.php:275 > > It seems after I reinstalled php 5.1.2 without a 16meg memory limit, it > stopped complaining (not sure if it is related or not). > > At this time I noticed 0.9.5beta2 was out, and instealled it. > With it I have been unable to load any horde php file at all. > > Error message from php: > FATAL: emalloc(): Unable to allocate 1969775421 bytes > > I tried to debug eaccelerator, but can only get as far as to see what > files it attempted to proccessed: > > EACCELERATOR hit: "/usr/local/www/horde/test.php" > EACCELERATOR hit: "/usr/local/www/horde/lib/core.php" |
|
From: Bart V. <bar...@zo...> - 2006-04-15 08:17:56
|
Patrick Domack wrote: > I have been setting up a new webmail site that uses php instead of perl= =2E >=20 > The php webmail is horde. I originally instealled php 5.1.2, with > eaccelerator 0.9.5beta1 >=20 > After config, testing, patching, and stressing it, it all seemed to wor= k > good, except two files eaccelerator complained about opline. Example: >=20 > [22605] EACCELERATOR: PHP crashed on opline 420 of main() at > /usr/local/www/horde/kronolith/config/prefs.php:275 >=20 > It seems after I reinstalled php 5.1.2 without a 16meg memory limit, it= > stopped complaining (not sure if it is related or not). >=20 > At this time I noticed 0.9.5beta2 was out, and instealled it. > With it I have been unable to load any horde php file at all. >=20 > Error message from php: > FATAL: emalloc(): Unable to allocate 1969775421 bytes >=20 > I tried to debug eaccelerator, but can only get as far as to see what > files it attempted to proccessed: >=20 > EACCELERATOR hit: "/usr/local/www/horde/test.php" > EACCELERATOR hit: "/usr/local/www/horde/lib/core.php" >=20 >=20 Hello, Can you try to disable the optimizer? It seems like a bug that make the code or optimizer go in an infinite loop. gr, Bart --=20 Bart Vanbrabant <bar...@zo...> PGP fingerprint: 093C BB84 17F6 3AA6 6D5E FC4F 84E1 FED1 E426 64D1 |
|
From: Patrick D. <pat...@pa...> - 2006-04-15 05:10:23
|
I have been setting up a new webmail site that uses php instead of perl. The php webmail is horde. I originally instealled php 5.1.2, with eaccelerator 0.9.5beta1 After config, testing, patching, and stressing it, it all seemed to work good, except two files eaccelerator complained about opline. Example: [22605] EACCELERATOR: PHP crashed on opline 420 of main() at /usr/local/www/horde/kronolith/config/prefs.php:275 It seems after I reinstalled php 5.1.2 without a 16meg memory limit, it stopped complaining (not sure if it is related or not). At this time I noticed 0.9.5beta2 was out, and instealled it. With it I have been unable to load any horde php file at all. Error message from php: FATAL: emalloc(): Unable to allocate 1969775421 bytes I tried to debug eaccelerator, but can only get as far as to see what files it attempted to proccessed: EACCELERATOR hit: "/usr/local/www/horde/test.php" EACCELERATOR hit: "/usr/local/www/horde/lib/core.php" |
|
From: Bart V. <bar...@zo...> - 2006-04-14 14:33:56
|
Geisel wrote: > Hi people!! >=20 > Ticket #57 (defect) >=20 > The eac don't compile many simple files. A example is: >=20 > $ cat teste.php >=20 > <?php >=20 > if($temp) >=20 > $aux=3D"atualizaInfo(document,$temp);"; >=20 > echo $aux; >=20 > ?> >=20 > $ /usr/local/bin/php -q > ../sys/ins/base/all/eaccelerator-0.9.5-beta2/encoder.php teste.php >=20 > // teste.php >=20 > eAccelerator Encoder ERROR: Can't compile file "teste.php" >=20 > $ >=20 > eaccelerator-0.9.5-beta2 >=20 > PHP 5.1.2 > Apache 2.0.55 >=20 > kernel 2.6.15 >=20 > Thank you! > Geisel Sierote >=20 Yes indeed. But that's because the encoder is BROKEN. Please look at other tickets before submitting a ticket: https://eaccelerator.net/ticket= /5 gr, Bart --=20 Bart Vanbrabant <bar...@zo...> PGP fingerprint: 093C BB84 17F6 3AA6 6D5E FC4F 84E1 FED1 E426 64D1 |
|
From: Geisel <ge...@4u...> - 2006-04-14 14:29:05
|
Hi people!!
Ticket #57 (defect)
The eac don't compile many simple files. A example is:
$ cat teste.php
<?php
if($temp)
$aux="atualizaInfo(document,$temp);";
echo $aux;
?>
$ /usr/local/bin/php -q ../sys/ins/base/all/eaccelerator-0.9.5-beta2/encoder.php teste.php
// teste.php
eAccelerator Encoder ERROR: Can't compile file "teste.php"
$
eaccelerator-0.9.5-beta2
PHP 5.1.2
Apache 2.0.55
kernel 2.6.15
Thank you!
Geisel Sierote
|
|
From: Soenke R. <rue...@to...> - 2006-04-13 07:33:39
|
Hi, eac...@li... <> wrote on : > exiting! it's been working with Prado 2.1 (www.xisc.com) with all the > cases which didn't work previously (and still don't work with Zend > Platform nor with APC). Good work!. Dito! Absolutely no problems here with PHP 5.1.3(RC2) & eA 0.9.5beta1 and beta2 for several weeks! We're using it with our new developed framework that powers all those PHP 5.x features - keyword count: Classes: 505 Interfaces: 23 Extends: 292 Abstract: 159 Static: 81 (static is evil :-P) Implements: 109 I guess that shows somehow that eA is quite stable for 5.1 applications. Big thanks to all developers! -- soenke |