I have been chasing down a memory problem for a few days now and would
appreciate any pointers you might have. We have an application that
allows for uploads of large (>50MB) files. We have a case with Netscape
(works fine with IE) that produces a "Ouch!: malloc failed in
malloc_block" error out of alloc.c with the modssl patches. While
debugging the problem I found that the "is_shm" variable is never set
when entering malloc_block. This routine is called from new_block which
is called by ap_palloc and make_sub_pool_internal. Does is_shm set to 0
mean that we are not running with shared memory on our system? I have
defined EAPI and EAPI_MM and that section of code is being hit (see
below).
Apache is very consistent with the error. We can upload exactly
42,590,208 bytes and the error hits. Am I missing something in the
compilation? Thanks for your help!
Our config is:
apache_1.3.20
Linux 6.2
mod_ssl-2.8.4-1.3.20
openssl-0.95a
mm-1.1.3
Netscape 4.78
#if defined(EAPI) && defined(EAPI_MM)
if (is_shm)
blok = (union block_hdr *)ap_mm_malloc(mm, size + sizeof(union
block_hdr));
else
#endif
blok = (union block_hdr *) malloc(size + sizeof(union block_hdr));
if (blok == NULL)
{
fprintf(stderr, "Ouch!: malloc failed in malloc_block\n");
exit(1);
}
debug_fill(blok, size + sizeof(union block_hdr));
#if defined(EAPI) && defined(EAPI_MM)
blok->h.is_shm = is_shm;
#
______________________________________________________________________
Apache Interface to OpenSSL (mod_ssl) www.modssl.org
User Support Mailing List [EMAIL PROTECTED]
Automated List Manager [EMAIL PROTECTED]
