pgsql: Add current directory to Perl include path

2018-02-24 Thread Peter Eisentraut
Add current directory to Perl include path Recent Perl versions don't have the current directory in the module include path anymore, so we need to add it here explicitly to make these scripts continue to work. Branch -- master Details --- https://git.postgresql.org/pg/commitdiff/9ee0573e

pgsql: Use croak instead of die in Perl code when appropriate

2018-02-24 Thread Peter Eisentraut
Use croak instead of die in Perl code when appropriate Branch -- master Details --- https://git.postgresql.org/pg/commitdiff/fde03e8b559d0e00bf4acd8cea3bb49411099c34 Modified Files -- src/backend/utils/mb/Unicode/convutils.pm | 5 +++-- src/bin/pg_rewind/RewindTest.pm

pgsql: Update headers of generated files

2018-02-24 Thread Peter Eisentraut
Update headers of generated files The scripts were changed in c98c35cd084a25c6cf9b08c76de8b89facd75fe7, but the output files were not updated to reflect the script changes. Branch -- master Details --- https://git.postgresql.org/pg/commitdiff/c4ba1bee68abe217e441fb81343e5f9e9e2a5353 Mod

pgsql: Fix thinko in in_range_float4_float8.

2018-02-24 Thread Tom Lane
Fix thinko in in_range_float4_float8. I forgot the coding rule for correct use of Float8GetDatumFast. Per buildfarm. Branch -- master Details --- https://git.postgresql.org/pg/commitdiff/32291aed494d425a548e45b3b6ad95f9d5c94e67 Modified Files -- src/backend/utils/adt/float.c

pgsql: Add window RANGE support for float4, float8, numeric.

2018-02-24 Thread Tom Lane
Add window RANGE support for float4, float8, numeric. Commit 0a459cec9 left this for later, but since time's running out, I went ahead and took care of it. There are more data types that somebody might someday want RANGE support for, but this is enough to satisfy all expectations of the SQL stand

pgsql: Check error messages in SSL tests

2018-02-24 Thread Peter Eisentraut
Check error messages in SSL tests In tests that check whether a connection fails, also check the error message. That makes sure that the connection was rejected for the right reason. This discovered that two tests had their connection failing for the wrong reason. One test failed because pg_hba