> On Sep 19, 2016, at 2:54 PM, Ben Pfaff <b...@ovn.org> wrote:
> TXN_AGAIN_WAIT and TXN_AGAIN_NOW were combined into a single return code
> TXN_TRY_AGAIN a long time ago, but this comment was not updated.
> Reported-by: Justin Pettit <jpet...@ovn.org>
> Signed-off-by: Ben Pfaff <b...@ovn.org>
> lib/ovsdb-idl.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
> diff --git a/lib/ovsdb-idl.c b/lib/ovsdb-idl.c
> index cdee326..ed4eddc 100644
> --- a/lib/ovsdb-idl.c
> +++ b/lib/ovsdb-idl.c
> @@ -3193,9 +3193,7 @@ ovsdb_idl_txn_write_clone(const struct ovsdb_idl_row
> * prerequisite to completing the transaction. That is, if 'column' in 'row_'
> * changed (or if 'row_' was deleted) between the time that the IDL originally
> * read its contents and the time that the transaction commits, then the
> - * transaction aborts and ovsdb_idl_txn_commit() returns TXN_AGAIN_WAIT or
> - * TXN_AGAIN_NOW (depending on whether the database change has already been
> - * received).
> + * transaction aborts and ovsdb_idl_txn_commit() returns TXN_TRY_AGAIN.
> * The intention is that, to ensure that no transaction commits based on dirty
> * reads, an application should call ovsdb_idl_txn_verify() on each data item
I think you need to update ovsdb/ovsdb-idlc.in, too.
Acked-by: Justin Pettit <jpet...@ovn.org>
dev mailing list