We have observed that the persistent TID and serial number associated with a
relfilenode in a relation descriptor may be updated to a new value before
relfilenode is updated.  If an XLog record is written with such a relation
descriptor, then the XLog record fails to apply during recovery.

This commit adds a check to validate sanity of persistent information in a
relation descriptor by fetching persistent information using relfilenode from
gp_relation_node.

The validation is controlled by a new GUC, gp_validate_pt_info_relcache.

[#130666969]
Internally this is the shared memory variable used_segspace in
the view gp_toolkit.gp_workfile_mgr_used_diskspace.
Signed-off-by: NXin Zhang <xzhang@pivotal.io>

This is an old version of PL/Java and the newer version is maintained in a separate
repo (https://github.com/greenplum-db/pljava). This removes the PL/Java directory, 
PL/Java tests, and all references to PL/Java.

This should have been part of commits
458a6745 and
37380b5f

When a DML `UPDATE`s the distribution column, the postgres planner will
bail. ORCA can theoretically do this, and indeed it does on some
platforms, e.g. Darwin (OS X 10.11). It's unclear to us why it fails to
do so on some Linux builds. We are working hard to resolve that which is
tracked by #1139.

Before #1139 is resolved we want to remove these two SQL statements to
expedite getting all of `installcheck` to green.

https://gporca.ci.pivotalci.info/pipelines/gpdb_perfSigned-off-by: NJesse Zhang <jzhang@pivotal.io>

Enhance GenerateAdvanceAggregates function and support int8inc built-in function to generate code for COUNT aggregate function.
Signed-off-by: NShreedhar Hardikar <shardikar@pivotal.io>

The references to cdb-pg are from quite long ago when the source tree
structure was different. Remove these references and clean up various
file headers that were affected along the way.

This borrows a page from the playbook of other unittests in order to
regain control after an ereport({ERROR|FATAL} ..) call in order to be
able to perform assertions. By trapping and siglongjmping we can gain
control even after an ERROR/FATAL.

In commit db41b9b9, to unify the error message when connection timeout in both sync and async codes, i modify the implement of gp_segment_connect_timeout checking. That fix has some defects and i want to revert it.

When process_startup_packets is triggered, gp_debug_linger was not set to 0 which cause anoying message "HINT: process xxxx" exist in output file and make tests unstable. This commit change the fault injection location to send_qe_details_init_backend where gp_debug_linger has been set to 0, so no hint message is generated in output file.

Apply the same fix as in 48983e3c which missed these invocations of
zero derefs. Dereferencing address zero is undefined behavior in C
and the compiler is free to optimize any such calls away (even though
that is believed to be rare). Ensure that any optimizations are
avoided by declaring the variable volatile and at the same time get
rid of a pesky clang compiler warning.

The current .travis.yml is no longer special condition the osx, but we still build on osx.
Just remove this osx, since we don't depends on travis to verify the build.

PR pipeline is at: https://gporca.ci.pivotalci.info/pipelines/gpdb_pr

- Add new pullrequest_resource gpdb_pr through docker image at `pivotaldata/pr-resource:latest`
- Add github pull request pipeline `pr_pipeline.yml`
Signed-off-by: NKarthikeyan Jambu Rajaraman <krajaraman@pivotal.io>

Shared libraries must be compiled with the fPIC flag
(position-independent code).

Signed-off-by: NTom Meyer <tmeyer@pivotal.io>
Signed-off-by: NAndrew Edstrom <aedstrom@pivotal.io>
Signed-off-by: NTom Meyer <tmeyer@pivotal.io>

Thanks to @edespino for paring up on this seemingly daunting task.

CDB_MOCKING is an old leftover that isn't set anymore, remove the
comment and add whitespace to make the code like upstream in order
to further minimize the risk of merge conflicts.

use asynchronous way to create gang and dispatch command by default.

Clean up some whitespace issues, correct spelling in comments, move
bracing to be uniform with codebase style (as well the lions part of
this file) and remove dead commented out code.

The file pointer is being freed by url_fclose() so we must copy the
members of the struct we need for error reporting before invoking,
the current coding causes a NULL pointer dereference.

Reported by Coverity scanning.

The ereport() macro will now expand to a do { .. } while(0); type
construction which is incompatible with (void) casting of fprintf()
done in yy_fatal_error() by default. Add a level of indirection to
cope with the new construction.

The elog() calls will now be expanded into a do {..} while(0); type
construction, fix up calls which were incompatible with this.

This prevents the postmaster from unexpectedly croaking if postgresql.conf
contains something like:

include 'invalid_directory_name'

Noah Misch. Reviewed by Tom Lane and myself.

Conflicts with GPDB code:
	src/backend/utils/misc/guc-file.l

In commit 71450d7f, we added code to inform
suitably-intelligent compilers that ereport() doesn't return if the elevel
is ERROR or higher.  This patch extends that to elog(), and also fixes a
double-evaluation hazard that the previous commit created in ereport(),
as well as reducing the emitted code size.

The elog() improvement requires the compiler to support __VA_ARGS__, which
should be available in just about anything nowadays since it's required by
C99.  But our minimum language baseline is still C89, so add a configure
test for that.

The previous commit assumed that ereport's elevel could be evaluated twice,
which isn't terribly safe --- there are already counterexamples in xlog.c.
On compilers that have __builtin_constant_p, we can use that to protect the
second test, since there's no possible optimization gain if the compiler
doesn't know the value of elevel.  Otherwise, use a local variable inside
the macros to prevent double evaluation.  The local-variable solution is
inferior because (a) it leads to useless code being emitted when elevel
isn't constant, and (b) it increases the optimization level needed for the
compiler to recognize that subsequent code is unreachable.  But it seems
better than not teaching non-gcc compilers about unreachability at all.

Lastly, if the compiler has __builtin_unreachable(), we can use that
instead of abort(), resulting in a noticeable code savings since no
function call is actually emitted.  However, it seems wise to do this only
in non-assert builds.  In an assert build, continue to use abort(), so that
the behavior will be predictable and debuggable if the "impossible"
happens.

These changes involve making the ereport and elog macros emit do-while
statement blocks not just expressions, which forces small changes in
a few call sites.

Andres Freund, Tom Lane, Heikki Linnakangas

Conflicts with GPDB codebase:
	config/c-compiler.m4
	configure
	configure.in
	contrib/cube/cubescan.l
	contrib/seg/segscan.l
	src/include/pg_config.h.in
	src/include/pg_config.h.win32

When elevel >= ERROR, we add an abort() call to the ereport() macro to
give the compiler a hint that the ereport() expansion will not return,
but the abort() isn't actually reached because the longjmp happens in
errfinish().

Because the effect of ereport() varies with the elevel, we cannot use
standard compiler attributes such as noreturn for this.