[dpdk-dev] [Bug 377] CRYPTODEV: set_sym_session_private_data() line 489: Set private data for driver 0 not allowed

* [dpdk-dev] [Bug 377] CRYPTODEV: set_sym_session_private_data() line 489: Set private data for driver 0 not allowed
@ 2019-12-10 17:17 bugzilla
  2020-04-26  0:40 ` bugzilla
  0 siblings, 1 reply; 2+ messages in thread
From: bugzilla @ 2019-12-10 17:17 UTC (permalink / raw)
  To: dev

https://bugs.dpdk.org/show_bug.cgi?id=377

            Bug ID: 377
           Summary: CRYPTODEV: set_sym_session_private_data() line 489:
                    Set private data for driver 0 not allowed
           Product: DPDK
           Version: unspecified
          Hardware: x86
                OS: All
            Status: UNCONFIRMED
          Severity: normal
          Priority: Normal
         Component: cryptodev
          Assignee: dev@dpdk.org
          Reporter: sunnylandh@gmail.com
  Target Milestone: ---

Environment:  Xeon E5-2640, Ubuntu 16.04, DPDK 19.11, Intel AESNI-MB library
0.53.0

When OpenSSL or AESNI-MB cryptodev is being used in sessionless mode for
symmetric crypto operation (e.g. SHA256 hash), the driver prints error message:

    CRYPTODEV: set_sym_session_private_data() line 489: Set private data for
driver 0 not allowed

This error is caused by:

1. set_sym_session_private_data() checks sess->nb_drivers > driver_id, or it
will print the above error.
2. In session-based mode, rte_cryptodev_sym_session_create() allocates a
session structure from the session pool, then sets sess->nb_drivers =
pool_priv->nb_drivers, so that there wouldn't be an error.
3. In sessionless mode, get_session() in rte_openssl_pmd.c or
rte_aesni_mb_pmd.c allocates a session structure from the session pool, but
does not set sess->nb_drivers, causing the error.

Due to this error, AESNI-MB driver segfaults in post_process_mb_job() after
get_sym_session_private_data() returns NULL.
OpenSSL driver can still compute SHA256 digest successfully, apart from the
error message, because it does not actually use get_sym_session_private_data().

-- 
You are receiving this mail because:
You are the assignee for the bug.

^ permalink raw reply	[flat|nested] 2+ messages in thread