Skip to content
  1. Jun 11, 2018
  3. Jun 09, 2018
  5. Jun 07, 2018
  7. Jun 06, 2018
  9. Jun 05, 2018
  11. Jun 04, 2018
  13. May 31, 2018
  15. May 29, 2018
  17. May 28, 2018
  19. May 27, 2018
  21. May 26, 2018
  23. May 25, 2018
  25. May 24, 2018
  27. May 23, 2018
    • Luca Toscano's avatar
      http_protocol.c: avoid duplicate headers when using · 21d05348
      Luca Toscano authored 
                       ap_send_error_response

While debugging PR 61860 I found a chain of events
that leads to duplicate headers in the HTTP response
if Header always set is used in the httpd's config.

As far as can understand, mod_headers uses err_headers_out
when dealing with the 'always' setting, since it is expected
to be preserved even on error. The current code seems
to correctly preserve the important headers like Location
(happening before the bit of code that changed), but then
it swaps headers_out with err_headers_out and clears
err_headers_out. My understanding is that this will cause
mod_headers, if called again, to re-insert the headers
set via 'always' again in err_headers_out, leading to a
duplication in the response. So far I managed to reproduce this
only with the specific use case outlined by the PR, but
there might be more.

r1831585 was added to the test suite for PR 61860,
(marked as TODO), and now it seems to pass as expected.

Since thi...
      21d05348
    • Joe Orton's avatar
      * modules/ssl/ssl_engine_config.c: Fix typos, but isn't this first TODO · bade60cb
      Joe Orton authored 
        actually done?


git-svn-id: https://svn.apache.org/repos/asf/httpd/httpd/trunk@1832078 13f79535-47bb-0310-9956-ffa450edef68
      bade60cb
  29. May 22, 2018
  31. May 20, 2018
    • Yann Ylavic's avatar
      mod_proxy_balancer: clear slotmem slots' inuse flags from persisted files. · c2f503de
      Yann Ylavic authored 
      Otherwise, when BalancerPersist is enabled, we might fail (re)starting with a
saved slotmem if a BalancerMember was renamed (e.g. PR 62308's port change).
This is because the renamed member is considered a new one, while the old name
still reserves a slot, so we end up missing one room. The overall number of
members does not change so the sizes check succeeds thus the persisted slotmem
is not invalidated. Yet the slomem is not really compatible.

By clearing inuse flags, we still allow for slots to be reused if their index
did not change (thanks to ap_proxy_find_balancershm() and slotmem_fgrab() at
the given index), but will also invalidate renamed slots which don't match the
index. We now have the correct behaviour by slot and the server (re)starts in
any case.



git-svn-id: https://svn.apache.org/repos/asf/httpd/httpd/trunk@1831938 13f79535-47bb-0310-9956-ffa450edef68
      c2f503de