You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@perl.apache.org by Stas Bekman <st...@stason.org> on 2002/03/15 19:20:15 UTC
modperl 1.x and perl 5.7.3 aren't playing nice?
Somebody at p5p was asking about problems with 5.7.3 and modperl 1.x.
I've run make test and got a bunch of failures, though didn't get a
chance to investigate.
Failed Test Stat Wstat Total Fail Failed List of Failed
-------------------------------------------------------------------------------
internal/api.t 86 84 97.67% 2-69 71-86
internal/dirmagic.t ?? ?? % ??
internal/headers.t 13 8 61.54% 1-7 13
internal/http-get.t 255 65280 16 12 75.00% 5-16
internal/http-post.t 255 65280 7 5 71.43% 3-7
internal/redirect.t 6 5 83.33% 1-4 6
internal/rwrite.t 2 2 100.00% 1-2
modules/cgi.t 5 3 60.00% 2-4
modules/constants.t ?? ?? % ??
11 tests skipped.
httpd terminated
5.6.1 passes all tests.
_____________________________________________________________________
Stas Bekman JAm_pH -- Just Another mod_perl Hacker
http://stason.org/ mod_perl Guide http://perl.apache.org/guide
mailto:stas@stason.org http://ticketmaster.com http://apacheweek.com
http://singlesheaven.com http://perl.apache.org http://perlmonth.com/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@perl.apache.org
For additional commands, e-mail: dev-help@perl.apache.org
Re: modperl 1.x and perl 5.7.3 aren't playing nice?
Posted by Stephen Clouse <st...@theiqgroup.com>.
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Sat, Mar 16, 2002 at 02:20:15AM +0800, Stas Bekman wrote:
> Somebody at p5p was asking about problems with 5.7.3 and modperl 1.x.
> I've run make test and got a bunch of failures, though didn't get a
> chance to investigate.
>
> Failed Test Stat Wstat Total Fail Failed List of Failed
> -------------------------------------------------------------------------------
> internal/api.t 86 84 97.67% 2-69 71-86
> internal/dirmagic.t ?? ?? % ??
> internal/headers.t 13 8 61.54% 1-7 13
> internal/http-get.t 255 65280 16 12 75.00% 5-16
> internal/http-post.t 255 65280 7 5 71.43% 3-7
> internal/redirect.t 6 5 83.33% 1-4 6
> internal/rwrite.t 2 2 100.00% 1-2
> modules/cgi.t 5 3 60.00% 2-4
> modules/constants.t ?? ?? % ??
> 11 tests skipped.
> httpd terminated
I've been meaning to bring the topic up but I wanted to see if I could discover
the cause and offer a patch first. Since I haven't had any luck yet, I'll just
add a "me too" to this for now. I am experiencing the exact same failures under
bleedperl.
The output with TEST_VERBOSE on is bizarre...it looks like whole sections of the
output are getting skipped (example: internal/api.t outputs "ok 2" followed
immediately by "ok 76") , but I haven't figured out what's different between the
parts getting skipped and the parts that aren't. I'll keep looking....
- --
Stephen Clouse <st...@theiqgroup.com>
Senior Programmer, IQ Coordinator Project Lead
The IQ Group, Inc. <http://www.theiqgroup.com/>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org
iEYEARECAAYFAjySZMEACgkQA4aoazQ9p2cwLACfeBfb7xr/ZxRI7KqV0D2EkhRb
ao4AnR7eF03CZy4sRFJeAmjcKlBu0Cuy
=Xq7+
-----END PGP SIGNATURE-----
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@perl.apache.org
For additional commands, e-mail: dev-help@perl.apache.org
Re: modperl 1.x and perl 5.7.3 aren't playing nice?
Posted by Stas Bekman <st...@stason.org>.
Stas Bekman wrote:
> Geoffrey Young wrote:
>
>> Stas Bekman wrote:
>>
>>> Somebody at p5p was asking about problems with 5.7.3 and modperl 1.x.
>>> I've run make test and got a bunch of failures, though didn't get a
>>> chance to investigate.
>>>
>>
>> with the AUTOLOAD stuff fixed, I'm back to compiling bleedperl and
>> mod_perl CVS nightly, though I don't play with it every day as I used
>> to.
>>
>> regardless of the failing tests it seems to work ok - I can complile,
>> start the server, and do mod_perl stuff.
>> the only thing I've noticed is that print "foo" doesn't seem to work
>> but $r->print("foo") does - I figured something about the TIEHANDLE
>> interface changed, but I didn't track it down. it might have fixed
>> itself since I last looked.
>
>
> Yup, here are some insights from N I-S:
>
> From: Nick Ing-Simmons <ni...@ing-simmons.net>
> Subject: Re: Does mod_perl work with 5.7.3?
>
> Mike Guy <mj...@cam.ac.uk> writes:
>
> [snip]
>
> >Before leaving, my colleague said it looked like the problem was
> >something to do with tieing *STDIN and *STDOUT, which mod_perl does
> >from C code.
>
> Tk's fileevent does likewise and also needed tweeking for 5.7.*
> The problem is that 5.6.* applied tie magic to the GV * , and 5.7.*
> applies it to the IO *. So if - like TK - your C code did io ops
> on the IO expecting to access the "inner" untied IO with 5.7.* you instead
> recurse back into the tie code.
>
Now fixed in cvs, thanks to Charles Jardine
--
_____________________________________________________________________
Stas Bekman JAm_pH -- Just Another mod_perl Hacker
http://stason.org/ mod_perl Guide http://perl.apache.org/guide
mailto:stas@stason.org http://ticketmaster.com http://apacheweek.com
http://singlesheaven.com http://perl.apache.org http://perlmonth.com/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@perl.apache.org
For additional commands, e-mail: dev-help@perl.apache.org
Re: modperl 1.x and perl 5.7.3 aren't playing nice?
Posted by Stas Bekman <st...@stason.org>.
Geoffrey Young wrote:
> Stas Bekman wrote:
>
>>Somebody at p5p was asking about problems with 5.7.3 and modperl 1.x.
>>I've run make test and got a bunch of failures, though didn't get a
>>chance to investigate.
>>
>
> with the AUTOLOAD stuff fixed, I'm back to compiling bleedperl and
> mod_perl CVS nightly, though I don't play with it every day as I used
> to.
>
> regardless of the failing tests it seems to work ok - I can complile,
> start the server, and do mod_perl stuff.
>
> the only thing I've noticed is that print "foo" doesn't seem to work
> but $r->print("foo") does - I figured something about the TIEHANDLE
> interface changed, but I didn't track it down. it might have fixed
> itself since I last looked.
Yup, here are some insights from N I-S:
From: Nick Ing-Simmons <ni...@ing-simmons.net>
Subject: Re: Does mod_perl work with 5.7.3?
Mike Guy <mj...@cam.ac.uk> writes:
[snip]
>Before leaving, my colleague said it looked like the problem was
>something to do with tieing *STDIN and *STDOUT, which mod_perl does
>from C code.
Tk's fileevent does likewise and also needed tweeking for 5.7.*
The problem is that 5.6.* applied tie magic to the GV * , and 5.7.*
applies it to the IO *. So if - like TK - your C code did io ops
on the IO expecting to access the "inner" untied IO with 5.7.* you instead
recurse back into the tie code.
--
Nick Ing-Simmons
http://www.ni-s.u-net.com/
_____________________________________________________________________
Stas Bekman JAm_pH -- Just Another mod_perl Hacker
http://stason.org/ mod_perl Guide http://perl.apache.org/guide
mailto:stas@stason.org http://ticketmaster.com http://apacheweek.com
http://singlesheaven.com http://perl.apache.org http://perlmonth.com/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@perl.apache.org
For additional commands, e-mail: dev-help@perl.apache.org
Re: modperl 1.x and perl 5.7.3 aren't playing nice?
Posted by Geoffrey Young <ge...@modperlcookbook.org>.
Stas Bekman wrote:
>
> Somebody at p5p was asking about problems with 5.7.3 and modperl 1.x.
> I've run make test and got a bunch of failures, though didn't get a
> chance to investigate.
with the AUTOLOAD stuff fixed, I'm back to compiling bleedperl and
mod_perl CVS nightly, though I don't play with it every day as I used
to.
regardless of the failing tests it seems to work ok - I can complile,
start the server, and do mod_perl stuff.
the only thing I've noticed is that print "foo" doesn't seem to work
but $r->print("foo") does - I figured something about the TIEHANDLE
interface changed, but I didn't track it down. it might have fixed
itself since I last looked.
I'm hoping to start hacking away at the STATUS file next week a bit at
a time, so I'll investigate what I can.
--Geoff
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@perl.apache.org
For additional commands, e-mail: dev-help@perl.apache.org