Looks like this is "fallout" from:
http://svn.apache.org/viewvc?view=revision&revision=1707230
> On Jan 31, 2016, at 9:41 PM, Joachim Achtzehnter <[email protected]> wrote:
>
> After upgrading from 2.4.12 to 2.4.18 we find that some requests for files
> cause a lock-up when HTTPS is used, but not with plain HTTP. After some
> debugging it seems the problem is that the mod_ssl no longer explicitly
> flushes the output it produced before going back to reading a new request on
> the same connection.
>
> This is a 32-bit Intel Linux platform, and by stepping through the mod_ssl
> output filter in the debugger I can see that the files that cause problems
> are being read by Apache using the mmap approach.
>
> Restoring the explicit flushing from 2.4.12 gets the 2.4.18 version of
> mod-ssl working for us. Here is the patch:
>
>
> --- ssl_engine_io.c.orig 2015-11-19 11:55:25.000000000 -0800
> +++ ssl_engine_io.c 2016-01-31 17:22:54.586227441 -0800
> @@ -466,6 +466,21 @@
> return -1;
> }
>
> + /* In theory, OpenSSL should flush as necessary, but it is known
> + * not to do so correctly in some cases; see PR 46952.
> + *
> + * Historically, this flush call was performed only for an SSLv2
> + * connection or for a proxy connection. Calling _out_flush
> + * should be very cheap in cases where it is unnecessary (and no
> + * output is buffered) so the performance impact of doing it
> + * unconditionally should be minimal.
> + */
> + if (bio_filter_out_flush(inctx->bio_out) < 0) {
> + bio_filter_out_ctx_t *outctx = inctx->bio_out->ptr;
> + inctx->rc = outctx->rc;
> + return -1;
> + }
> +
> BIO_clear_retry_flags(bio);
>
> if (!inctx->bb) {
>
>
>
