I just added a definition for IMPAIR(BEHAVIOUR). This is so that code can start to get away from DBGP() and the assumption that impair flags live in cur_debugging.
As defined, code would use it as: if (IMPAIR(RETRANSMITS)) ... I.e., not IMPAIR(IMPAIR_RETRANSMITS); this seemed redundant; and: if (IMPAIR(RETRANSMITS) || IMPAIR(TIMEOUTS)) ... i.e., not IMPAIR(IMAPIR_RETRANSMITS | IMPAIR_TIMEOUTS); with lset_t, I'm never sure if that is OR vs AND ---------- Forwarded message ---------- From: Andrew Cagney <[email protected]> Date: 20 November 2017 at 11:47 Subject: [Swan-commit] Changes to ref refs/heads/master To: [email protected] New commits: commit b14d01dff8330cc85ddeb54077271ad4bd10e694 Author: Andrew Cagney <[email protected]> Date: Thu Nov 16 10:17:56 2017 -0500 logging: #define IMPAIR(BEHAVIOUR) So that code can start to get away from DBGP() and the assumption that impair flags live in cur_debugging. Note that IMPAIR(BEHAVIOUR) tests for IMPAIR_BEHAVIOUR. _______________________________________________ Swan-commit mailing list [email protected] https://lists.libreswan.org/mailman/listinfo/swan-commit _______________________________________________ Swan-dev mailing list [email protected] https://lists.libreswan.org/mailman/listinfo/swan-dev
