make test fails on IRIX 6.5 with perl 5.8.9 when building for irix-mips3-cc
and irix64-mips4-cc (the only platforms I've tried so far).
TOP=.. PERL=/usr/nekoware/bin/perl /usr/nekoware/bin/perl run_tests.pl alltests
../test/recipes/01-test_ordinals.t ok
../test/recipes/05-test_bf.t
Building openssl-1.1.0pre3 for irix-mips3-cc fails because
struct timeval is undefined:
cc-1070 cc: ERROR File = s_server.c, Line = 2092
The indicated type is incomplete.
struct timeval timeout;
With the patch below, openssl builds, and aside from a few possibly
spurious errors in
I'm trying to build openssl 1.1.0pre3 on IRIX 6.5 with the new unified
building system using perl 5.8.9. This fails with the error message
below. Building without --unified works as expected. Any hints
how to debug this or extract better logs would be appreciated.
$ perl
On Apr 1, 2010, at 17:44 , Stephen Henson via RT wrote:
That is most peculiar. If an ENGINE isn't in use I wouldn't expect that
to happen unless something somewhere is loading one.
OK, next step is to see what ENGINE it is. Can you in the debugger print
out the engine structure for
On Apr 2, 2010, at 15:30 , Stephen Henson via RT wrote:
This looks suspicious:
#3 0x414a0787 in engine_pkey_asn1_meths_free (e=0x39) at
tb_asnmth.c:165
165 EVP_PKEY_asn1_free(pkm);
Not sure where the 0x39 is coming from. In frame#2:
#2
A clean build of openssl-1.0.0 (./config run without parameters) does not
use the hashed certificate files from -CApath and fails certificate
verification. The same behavior is seen when using openssl via the openssl
s_client command as well as the library with curl 7.20.0. The system
used here
On Apr 6, 2010, at 17:08 , Stephen Henson via RT wrote:
[...]
Please try rehashing the directories using the 1.0.0 version of OpenSSL.
The certificate file hash algorithm changed in OpenSSL 1.0.0 and is not
compatible with 0.9.8: this was needed to support some additional
functionality.
The
I'm using an apache 2.2.15 with php-5.2.13 and curl-7.20.0. When using
openssl-1.0.0, I get easy to reproduce Aborts in CRYPTO_free if I reload
a page while it is still loading. The problem reproducably goes away when
I replace openssl-1.0.0 with 0.9.8m (and recompile the whole stack).
OSes are
On Apr 1, 2010, at 12:26 , Stephen Henson via RT wrote:
Well first question is... are you compiling against the 0.9.8 headers
instead of the 1.0.0 ones?
No. I've had those problems with builds from scratch, i.e. completely empty
source and destination directories as well. I also don't have
On Apr 1, 2010, at 13:04 , Stephen Henson via RT wrote:
OK, I'm trying to see where that ENGINE call comes from. Are you using
an ENGINE to load a private key?
This is just a normal vhost in apache. Apparently, it's sufficient
to just point a browser to an http URL of a large object (I'm
On Apr 1, 2010, at 17:44 , Stephen Henson via RT wrote:
That is most peculiar. If an ENGINE isn't in use I wouldn't expect that
to happen unless something somewhere is loading one.
OK, next step is to see what ENGINE it is. Can you in the debugger print
out the engine structure for
11 matches
Mail list logo
