Fix o_all_local inserted before set

[sm-shaw]

As PR #356 fixes some inconsistencies inherited from the TPC-C spec example code, there is the opportunity to do the same for the o_all_local set at the same time.

In the Spec o_all_local is inserted into ORDERS before it is set and therefore always inherits the default for the insert.
https://www.tpc.org/tpc_documents_current_versions/pdf/tpc-c_v5.11.0.pdf

o_all_local inserted --> EXEC SQL INSERT INTO ORDERS (o_id , o_d _id , o_w _id , o_c_id ,o_entry_d , o_ol_cnt, o_all_local) VALUES (:o_id , :d _id , :w _id , :c_id ,:datetime, :o_ol_cnt, :o_all_local);
EXEC SQL INSERT INTO NEW_ORDER (no_o_id , no_d_id , no_w _id ) VALUES (:o_id , :d _id , :w _id );
for (ol_number=1; ol_number<=o_ol_cnt; ol_number++)
{
ol_supply_w _id =atol(supware[ol_number -1]);
o_all_local set --> if (ol_supply_w _id != w _id) o_all_local=0;
ol_i_id =atol(item id [ol_number-1]);
ol_qu antity=atol(qty[ol_number-1]);
...

HammerDB inherited this behaviour, however Oracle and PostgreSQL have already been updated to fix this issue. Pull request #372 also corrects this for SQL Server. MySQL, MariaDB and Db2 should be updated for consistency by setting the default to 1 and moving the inserts after the point where o_all_local can be set to 0.