Switch branches/tags
Nothing to show
Commits on Nov 21, 2016
  1. Fix memory leak in connection establishment

    Spotted by Coverity.
    committed Nov 21, 2016
Commits on Oct 1, 2014
  1. Run faster!

    There's no reason to wait for an ACK from the server.  Just push all the
    data down its throat as quickly as possible.
    committed Oct 1, 2014
Commits on Jul 16, 2014
  1. Update LICENSE

    committed Jul 16, 2014
  2. Add LICENSE

    committed Jul 16, 2014
Commits on Jul 13, 2014
  1. Hard-code the project name in Travis

    I think the Makefile depends on the directory being in contrib/pgcov,
    so there's no reason to try and be more flexible here.
    committed Jul 13, 2014
  2. Fix typo in Makefile

    committed Jul 13, 2014
  3. Attempt to test 9.1 in Travis

    committed Jul 13, 2014
Commits on Jul 11, 2014
  1. Keep the network connection open across function calls

    Connecting and disconnecting all the time isn't very efficient and might
    have been the cause of recent problems reported by Anders Granlund.
    Instead, keep the connection open, and instead of waiting for the server
    to close the connection, only wait for an acknowledgement and then move
    on.  This will probably keep the connections open a bit longer than
    necessary, but that shouldn't be a huge problem given that you would
    never want to run pgcov on a live environment anyway.  It's probably
    possible to improve on that (by e.g. closing the connection at
    transaction commit/abort), but it's not clear whether that's necessary,
    and depending on when the connection is closed, it might in some cases
    introduce the same problems as the previous approach.
    committed Jul 11, 2014
Commits on Jul 10, 2014
  1. Don't mix declarations and code

    committed Jul 10, 2014
  2. Remove unused variable

    committed Jul 10, 2014
  3. Use MaxBackends as the maximum number of workers

    This should be enough for pretty much everyone.
    committed Jul 10, 2014
  4. Replace spaces with tabs

    committed Jul 10, 2014
Commits on Jun 5, 2014
  1. Fix crash on second call to a non-PL/PgSQL function

    We might have ended up doing an strcmp() call on a NULL pointer if the
    prosrc was NULL in an existing record, or there previously was a prosrc
    but the function was redefined in a language other than PL/PgSQL, and
    called again.
    Per report from Anders Granlund.
    committed Jun 5, 2014
  2. Handle cmd_type PLPGSQL_STMT_DYNEXECUTE

    Reported by Anders Granlund.
    committed Jun 5, 2014
Commits on Jun 4, 2014
  1. Add

    committed Jun 4, 2014
  2. Initial commit

    committed Jun 4, 2014