> > > > one comment: instead of having block variables onexit, in the while > > loops in shmem_exit, can we have a single local variable defined at > > the start of the shmem_exit function > > and reuse them in the while loops? same comment for onexit block > > variable in proc_exit_prepare() function. > > > > If you are worried about the declaration and initialization of the variable > will > happen with every loop cycle then you shouldn't because that happens only > once before the loop-block is entered. >
thanks. understood (just for info [1]) . Is there a test case covering this part of the code(I'm not sure if one exists in the regression test suite)? If no, can we add one? [1] - https://stackoverflow.com/questions/29785789/why-do-the-objects-created-in-a-loop-have-the-same-address/29785868 With Regards, Bharath Rupireddy. EnterpriseDB: http://www.enterprisedb.com
