I did some more research... the problem I'm seeing is in 5.6.1 as
compiled on OS X 10.1.
 
Compiling 5.6.0 on the same operating system with useperlio works like a
charm...
 
Anyone know what's changed? (Otherwise, I'm going to stick with 5.6.0
out of fear :)

-----Original Message-----
From: Nathan Herring [mailto:[EMAIL PROTECTED]] 
Sent: Thursday, December 13, 2001 12:44 PM
To: [EMAIL PROTECTED]
Subject: PerlIO on OS X 10.1?



Anyone try building using PerlIO (perlapio)? (i.e., setting
-Duseperlio=true)? 

I'm seeing the following problems: 
* I'm getting a metric ton of "Setting cnt to ###, ptr implies ###"
statements in miniperl during the build process and in perl proper
during the test. 

* Miniperl periodically emits "Running ../../miniperl XXX" exits with
status # at (eval #) line #. e.g. "Running '../../miniperl
.../../lib/ExtUtils/xsubpp temp0002>&1' exits with status 255 at (eval
33) line 43.

* It fails 23 test scripts of 245, 90.61% okay. 

The scripts it fails are: 
run/runenv -- FAILED at test 2 
op/anonsub -- FAILED at test 1 
op/attrs -- FAILED at test 1 
op/closure -- FAILED at test 172 
op/fork -- FAILED at test 1 
op/hashwarn -- FAILED at test 1 
op/magic -- fails to compile require at ../lib/warnings.pm line 127
(FAILED at test 0) 
op/pat -- FAILED at test 223 
op/runlevel -- FAILED at test 1 
op/tie -- WARNING: FAILED at test 0 
pragma/constant -- FAILED at test 36 
pragma/strict -- FAILED at test 1 
pragma/subs -- FAILED at test 1 
*pragma/warnings -- FAILED at test 1 
lib/b -- FAILED at test 8 
*lib/db-btree -- FAILED at test 0 
*lib/db-recno -- FAILED at test 51 
lib/open2 -- FAILED at test 0 
lib/open3 -- FAILED at test 0 
*lib/posix -- FAILED at test 11 

* denotes a failure that happens with the normal Mac OS X build of Perl.


I thought that PerlIO, if turned on, still defaulted to use stdio until
you actually rewrite perlio.c to use your particular remapping of the
IO. In which case, it doesn't make sense that all of these extra errors
would show up.

Anyone know better? 

Also, is there a similar idea for memory management? I haven't found it
yet... 

Thanks, 
nh 

Reply via email to