Greg Sabino Mullane wrote: > > So, my doubt is: if the return type is int instead of unsigned int, > > is this function testable for negative return values? > > A quick glance at the code in fe-exec.c and fe-protocol3.c shows that > the underlying variable starts at 0 as an int and in incremented by > one every row, so it seems possible that it could wrap around for > very large results sets and/or boxes with a low representation of 'int'. > There may be some other safeguards in place I did not see to prevent this, > but I don't see a reason why we shouldn't use unsigned int or > unsigned long int here, both for ntups and the return value of the > function. I think we need more use cases before we break the API on this one. -- Bruce Momjian <bruce@xxxxxxxxxx> http://momjian.us EnterpriseDB http://www.enterprisedb.com + If your life is a hard drive, Christ can be your backup. + ---------------------------(end of broadcast)--------------------------- TIP 9: In versions below 8.0, the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
