Stas Bekman wrote:
Geoffrey Young wrote:

I haven't been following p5p that well lately, but what is the status of 5.8.1-RC4 wrt mod_perl.

I, for one, cannot get RC4 (or bleedperl) to work with worker on 2.0 or 2.1

I thought the problematic hash stuff was removed in RC3?


I built an exact same perl and tried with the latest httpd-2.0 (haven't tried 2.1), but using patch level 20562 and it worked just fine (both worker and prefork). Can you please try with 20562 (it's post-RC4).

You can get the latest patchlevel, using the following:

rsync -acvz --delete --force rsync://ftp.linux.activestate.com/perl-5.8.x ./perl-5.8.1

ok, I just finished doing that, though perl is now patch level 20581. I still see the problem.


after some investigation, I don't think it's perl. if I set PERL_HASH_SEED to some value, everything is ok, and this patch fixes the segfault

--- src/modules/perl/mod_perl.c 6 Aug 2003 06:46:35 -0000       1.175
+++ src/modules/perl/mod_perl.c 11 Aug 2003 13:52:58 -0000
@@ -33,7 +33,7 @@
     /* check if there is a specific hash seed passed via the env var
      * and if that's the case -- use it */
     apr_status_t rv = apr_env_get(&s, "PERL_HASH_SEED", p);
-    if (rv != APR_ENOTIMPL) {
+    if (rv == APR_SUCCESS) {

which seems like a better check anyway.

so, maybe you have PERL_HASH_SEED set in your env someplace?

--Geoff


--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]



Reply via email to