Hi Eugene,
this should work...
We need more of your code and/or logging to figure out what's going on.
Maybe could you also log r and r-notes (with format %pp) to see if
the two modules really share the same request?
Regards,
Yann.
On Sat, May 9, 2015 at 6:40 PM, Eugene Teslenko
I don't know much about mod_fastcgi (third-party module), but it seems
that it uses ap_internal_redirect[_handler]().
This creates a new (redirect) request which is linked with the
original one through the r-next (= newr) and newr-prev (= r).
Unlike sub-requests, internal redirects are fully
mod_a handler function returns declined both times it's being called.
Here is the source of mod's
= mod_a.c =
module AP_MODULE_DECLARE_DATA mod_a;
static int handler(request_rec *r) {
// if (strcmp(r-handler, fastcgi-script)) return DECLINED;
apr_table_set(r-notes, test-var,
Thanks for your replies!
After adding pointers to log and disabling handler check inside mod_a, I've
noted the problem.
Error log has three lines for single http-request now:
mod_a[49059]: test-var=test-string, r_pp=8450610a0, notes_pp=8450621a8,
handler=application/x-httpd-php56
What does mod_a handler return?
How do you register it (ap_hook_handler() line)?
On Sun, May 10, 2015 at 3:59 PM, Eugene Teslenko
eugene.tesle...@gmail.com wrote:
Thanks for your replies!
After adding pointers to log and disabling handler check inside mod_a, I've
noted the problem.
Error log
It may be any handler module. FastCGI is just for example.
In apache's config it's just Action + SetHandler options.
Thank you very much for your replies and suggestions, it was really an
internal redirect and could get an original request rec through r-prev.
2015-05-10 21:30 GMT+03:00 Yann
Hi,
There is a way to do this, it would involve having the request variables changed
apr_table_set(rec-subprocess_env, KEY_OF_MY_DATA, data);
where the second argument would be the key you intend to pass and thirdbeing
its contents, you will also need to ensure APR_HOOK_MIDDLE is given to the
