[gnutls-devel] GnuTLS | mini-dtls-eagain has a chance to hang under valgrind (#1274)

[Read-only notification of GnuTLS library development activities]

Alexander Sosedkin created an issue: https://gitlab.com/gnutls/gnutls/-/issues/1274



## Description of problem:
mini-dtls-eagain has a chance to hang under valgrind

## Version of gnutls used:
4989ed12f8

## Distributor of gnutls (e.g., Ubuntu, Fedora, RHEL):
observed on Fedora 33 x86_64

## How reproducible
probabilistically, can take hundreds of iterations to manifest

Steps to Reproduce:
`while true; do valgrind --leak-check=full ./tests/mini-eagain-dtls -v 2>&1 | tee mini-eagain-dtls.log; sleep 2; done`

## Actual results:
execution sometimes hangs way past the normal execution times with

    server|<9>| ENC[0x50c41e0]: cipher: AES-256-CBC, MAC: SHA1, Epoch: 1
    server|<11>| WRITE: enqueued 77 bytes for 0x50c41e0. Total 91 bytes.
    server|<5>| REC[0x50c41e0]: Sent Packet[281474976710657] Handshake(22) in epoch 1 and length: 77
    server|<11>| WRITE FLUSH: 91 bytes in buffer.
    server|<3>| ASSERT: buffers.c[_gnutls_writev_emu]:464
    server|<2>| WRITE: -1 returned from 0x50c41e0, errno: 11
    server|<11>| WRITE interrupted: 91 bytes left.
    server|<3>| ASSERT: dtls.c[_dtls_transmit]:348
    server|<3>| ASSERT: handshake.c[send_handshake_final]:3340
    client|<11>| WRITE FLUSH: 0 bytes in buffer.
    client|<3>| ASSERT: buffers.c[_gnutls_io_write_flush]:696
    client|<3>| ASSERT: dtls.c[_dtls_transmit]:263
    client|<3>| ASSERT: dtls.c[_dtls_transmit]:420
    client|<3>| ASSERT: dtls.c[_dtls_wait_and_retransmit]:440
    client|<3>| ASSERT: handshake.c[recv_handshake_final]:3383
    server|<11>| HWRITE FLUSH: 25 bytes in buffer.
    server|<11>| WRITE FLUSH: 91 bytes in buffer.
    server|<3>| ASSERT: buffers.c[_gnutls_writev_emu]:464
    -- keeps repeating from here on --

## Expected results:
execution doesn't hang

-- 
Reply to this email directly or view it on GitLab: https://gitlab.com/gnutls/gnutls/-/issues/1274
You're receiving this email because of your account on gitlab.com.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.gnupg.org/pipermail/gnutls-devel/attachments/20210903/4151b09f/attachment.html>