I'm sorry, I haven't seen that crash.  But I haven't run on the base
7.1.x branch for several years and not with openssl 1.1.1 code.  I
poked around the openssl code and based on the line numbers, it looks
like the v->get_cert_methods method in X509_STORE_add_lookup is
getting messed up.  This is ultimately called from
SSL_CTX_load_verify_locations or X509_STORE_load_locations directly.
I don't think our calls into this have changed from 7.1.x through to
our current master/10.x branch.  So it is unclear what the problem
would be with introducing openssl 1.1.1

What version of openssl 1.1.1 are you using?  Did this work if you ran
it against 1.0.2?  I know that a number of folks are running the 8.0.x
and 8.1.x builds against openssl 1.1.1. You might try moving forward
with ATS.

On Fri, Apr 9, 2021 at 7:26 AM supraja sridhar
<[email protected]> wrote:
>
> Hi,
>
> When I try to send traffic to  ATS 7.1.1 with the system having openssl 1.1.1 
> I am seeing the below crash. Can someone help identify the issue ?
>
> [Thread debugging using libthread_db enabled]
>
> Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
>
> Core was generated by `trafficserver'.
>
> Program terminated with signal SIGSEGV, Segmentation fault.
>
> #0 X509_STORE_add_lookup (v=v@entry=0x7fabe84da600, m=0x7fabf6808620 
> <x509_dir_lookup>) at crypto/x509/x509_lu.c:254
>
> [Current thread is 1 (Thread 0x7fabf42a3700 (LWP 17522))]
>
> #0 X509_STORE_add_lookup (v=v@entry=0x7fabe84da600, m=0x7fabf6808620 
> <x509_dir_lookup>) at crypto/x509/x509_lu.c:254
>
> #1 0x00007fabf653749d in X509_STORE_load_locations (ctx=0x7fabe84da600, 
> file=0x5654faa23130 "/etc/ssl_ca/canonical_ca_roots.pem", path=0x5654faa23110 
> "/etc/ssl_ca/") at crypto/x509/x509_d2.c:48
>
> #2 0x00005654f85b97de in SSLNetVConnection::sslStartHandShake 
> (this=0x7fabec4b9840, event=1, err=@0x7fabf42a2cd8: -197496816) at 
> /workspace/trafficserver/iocore/net/SSLNetVConnection.cc:1005
>
> #3 0x00005654f85d48dd in write_to_net_io (nh=0x7fabf43aae60, 
> vc=0x7fabec4b9840, thread=0x7fabf43a7010) at 
> /workspace/trafficserver/iocore/net/UnixNetVConnection.cc:446
>
> #4 0x00005654f85d47c1 in write_to_net (nh=0x7fabf43aae60, vc=0x7fabec4b9840, 
> thread=0x7fabf43a7010) at 
> /workspace/trafficserver/iocore/net/UnixNetVConnection.cc:424
>
> #5 0x00005654f85ca98e in NetHandler::mainNetEvent (this=0x7fabf43aae60, 
> event=5, e=0x5654fa8c2ca0) at 
> /workspace/trafficserver/iocore/net/UnixNet.cc:516
>
> #6 0x00005654f8349d0c in Continuation::handleEvent (this=0x7fabf43aae60, 
> event=5, data=0x5654fa8c2ca0) at 
> /workspace/trafficserver/iocore/eventsystem/I_Continuation.h:153
>
> #7 0x00005654f85f7225 in EThread::process_event (this=0x7fabf43a7010, 
> e=0x5654fa8c2ca0, calling_code=5) at 
> /workspace/trafficserver/iocore/eventsystem/UnixEThread.cc:143
>
> #8 0x00005654f85f772b in EThread::execute (this=0x7fabf43a7010) at 
> /workspace/trafficserver/iocore/eventsystem/UnixEThread.cc:270
>
> #9 0x00005654f85f68c3 in spawn_thread_internal (a=0x5654fa892810) at 
> /workspace/trafficserver/iocore/eventsystem/Thread.cc:84
>
> #10 0x00007fabf61406ba in start_thread () from 
> /lib/x86_64-linux-gnu/libpthread.so.0
>
> #11 0x00007fabf53d16cd in clone () from /lib/x86_64-linux-gnu/libc.so.6
>
>
> Thanks,
>
> Supraja

Reply via email to