- 16 6月, 2016 2 次提交
-
-
由 Heikki Linnakangas 提交于
The other FIXME comment that the removed comment refers to was reverted earlier, before we pushed the 8.3 merge. Should've uncommented this back then, but I missed the need for that because I've been building with --enable-cassert. This fixes the regression failure on gpdtm_plpgsql test case, when built without assertions.
-
由 Daniel Gustafsson 提交于
Commit 5dc7dda9 added a forward declaration for PartitionSelector and PartitionSelectorState, the redefenition of which is only legal in C11 which we are yet to use. +1 from hlinnakangas@
-
- 15 6月, 2016 10 次提交
-
-
由 Daniel Gustafsson 提交于
-
由 Heikki Linnakangas 提交于
These queries were backported from later PostgreSQL releases earlier, but had to be disabled when we started the 8.3 merge, because we didn't have all the catalog changes from 8.3 yet. Now we do, so re-enable the queries.
-
由 Heikki Linnakangas 提交于
The PostgreSQL 8.3 merge included a band-aid fix for this, but since we haven't yet made a GPDB 5.0 release, and are therefore free to still modify the catalogs, let's apply the proper fix that went into PostgreSQL 9.3 at the time. commit 71627f3d Author: Tom Lane <tgl@sss.pgh.pa.us> Date: Wed Feb 13 16:20:01 2013 -0500 Fix CVE-2013-0255 properly. Revert commit ab0f7b60 (in HEAD only) in favor of the proper solution, which is to declare enum_recv() correctly in the system catalogs. It should be declared to take type "internal" not "cstring". Also improve the type_sanity regression test, which should have caught this typo, so that it actually would. Most of the relevant checks on the signature of type I/O functions should not have been restricted to basetypes/pseudotypes, as they should apply to any type's I/O functions.
-
由 Heikki Linnakangas 提交于
Everything in PostgreSQL 8.3 is now in Greenplum. PostgreSQL 8.3 has been end-of-lifed in upstream, so there will be no new upstream minor releases of this series anymore either. This includes a lot of new features from PostgreSQL 8.3, and a ton of bug fixes. See upstream release notes for details. One big user-visible change included is the removal of implicit cast from other datatypes to text. When this was done in PostgreSQL, it caused a lot of sloppily written application queries to break. Which is a good thing in the long run, but it caused a lot of pain on upgrade. A few features work slightly differently in GPDB: * Lazy XIDs feature in upstream reduces XID consumption, by not assigning XIDs to read-only transactions. That optimization has not been implemented for GPDB distributed transactions, however, so practically all queries in GPDB still consume XIDs, whether they're read-only or not. * temp_tablespaces GUC was added in upstream, but it has been disabled in GPDB, in favor of GPDB-specific system to decide which filespace to use for temporary files. * B-tree indexes can now be used to answer "col IS NULL" queries. That support has not been implemented for bitmap indexes. * Support was added for "top N" sorting, speeding up queries with ORDER BY and LIMIT. That support was not implemented for tuplesort_mk.c, however, so you only get that benefit with enable_mk_sort=off. * Multi-worker autovacuum support was merged, but autovacuum as whole is still disabled in GPDB. * Planner hook was added. Nothing special there, but we should consider refactoring the ORCA glue code to use it. * Plan cache invalidation was added upstream, which made the gp_plpgsql_clear_cache_always option obsolete. We also backported the further invalidation improvements from 8.4. See below for more information. In addition to those, there were some big refactorings that were not that interesting from user point of view, but caused a lot of code churn: InitPlans are now initialized separately at executor startup, the executor range table was changed to be "flat", and code related to parse analysis of DDL commands was moved around. The way UPDATE WHERE CURRENT OF was implemented in GPDB was refactored to match the new upstream support for the same (see below for more information). Plan cache invalidation ----------------------- One notable feature included in this merge is the support for plan cache invalidation. GPDB contained a hack to invalidate all plans in master between transactions, for the same purpose, but the proper plan cache invalidation is a much better solution. However, because the GPDB hack also handled dropped/recreated functions in some cases, if we just leave it out and replace with the 8.3 plan cache invalidation, there will be a regression in those cases. To fix, this merge commit includes a backport the rest of the plan cache invalidation support from PostgreSQL 8.4, which handles functions and operators too. The upstream commit for that was: commit ee33b95d Author: Tom Lane <tgl@sss.pgh.pa.us> Date: Tue Sep 9 18:58:09 2008 +0000 Improve the plan cache invalidation mechanism to make it invalidate plans when user-defined functions used in a plan are modified. Also invalidate plans when schemas, operators, or operator classes are modified; but for these cases we just invalidate everything rather than tracking exact dependencies, since these types of objects seldom change in a production database. Tom Lane; loosely based on a patch by Martin Pihlak. As a result of that, the gp_plpgsql_clear_cache_always GUC is removed, as it's no longer needed. This also includes new test cases in the plpgsql cache regression test, to demonstrate the improvements. CURRENT OF code changes ----------------------- This merge includes the upstream support for UPDATE WHERE CURRENT OF. GPDB had support for that already, it's been refactored to use the upstream code as much as possible (execCurrentOf()). Now that we have the upstream code available, this also refactors the way current position of a cursor is dispatched. Instead of modifying every CurrentOfExpr node in the plan tree before dispatching it, store the current positions of each cursor mentioned in a CurrentOfExpr in a separate CursorPosInfo node, and dispatch those along with the query plan, in QueryDispatchDesc. This was a joint effort between me (Heikki) and Daniel Gustafsson.
-
由 Heikki Linnakangas 提交于
The printablePredicate of a static PartitionSelector node contains Var nodes with varno=INNER. That's bogus, because the PartitionSelector node doesn't actually have any child nodes, but works at execution time because the printablePredicate is not only used by EXPLAIN. In most cases, it still worked, because most Var nodes carry a varnoold field, which is used by EXPLAIN for the lookup, but there was one case of "bogus varno" error even memorized in the expected output of the regression suite. (PostgreSQL 8.3 changed the way EXPLAIN resolves the printable name so that varnoold no longer saves the bacon, and you would get a lot more of those errors) To fix, teach the EXPLAIN of a Sequence node to also reach into the static PartitionSelector node, and print the printablePredicate as if that qual was part of the Sequence node directly. The user-visible effect of this is that the static Partition Selector expression now appears in EXPLAIN output as a direct attribute of the Sequence node, not as a separate child node. Also, if a static Partition Selector doesn't have a "printablePredicate", i.e. it doesn't actually do any selection, it's not printed at all.
-
由 Heikki Linnakangas 提交于
This is mostly copy-pasted from CoerceToDomain. Note: This requires an up-to-date version of ORCA to compile, older versions of the ORCA library itself don't know about CoerceViaIO nodes either.
-
由 Heikki Linnakangas 提交于
This change in commit af5e4bcf was surely not intentional, and made the test to fail.
-
Port `dml/triggers` Port `olap/groupingfunction` Port `functions/builtin` Port `dml/functional` Port `dml/functional/sql_partiton` Port `functions/functionProperty` Update optfunctional_schedule with ported tests Update the Makefile for optimizer functional tests
-
由 Nikos Armenatzoglou 提交于
When SharedScan requests a tuple from the underlying Sort, and the Sort has spilled to disk, the mk_sort code (also the regular sort) makes a copy of the memtuple, but places it in the sortcontext. When hitting a QueryFinishPending (e.g., master received enough tuples, so stop the execution), EagerFree is called for the entire tree. When reaching the sort node, the sortcontext is deleted, even though the SharedScan above it still has a pointer to a memtuple allocated in this context. The solution is: pass a memory context for the mk_sort and regular sort to use (change the API and implementation of the following two functions: tuplesort_gettupleslot_pos_mk and tuplesort_gettupleslot_pos). Those functions should use the context provided by the caller to make a copy of the memtuple. Signed-off-by: NGeorge Caragea <gcaragea@pivotal.io>
-
由 Jamie McAtamney 提交于
Removed extraneous argument to _prompt_continue that would cause gpcrondump to fail if run without -a.
-
- 14 6月, 2016 4 次提交
-
-
-
由 Jamie McAtamney 提交于
In commit f79e98fa, which enabled the backup and restoration of roles containing special characters, the new Behave test did not remove the created test role at the end, so running the test a second time would fail. This commit updates that test so that it does not leave the test role around between runs.
-
由 Daniel Gustafsson 提交于
Fixed a few spelling errors, and minor whitespace issues, in the Resource Scheduling README while reading it. Also updated the file markers on the README, Makefile and code to match upstream CVS to Git conversion. No semantic changes made to the text.
-
由 Jamie McAtamney 提交于
Previously, the cdatabase file created by gpcrondump did not correctly quote a role containing special characters in its CREATE DATABASE statement. Such roles are now handled correctly. Authors: Jamie McAtamney and Chumki Roy
-
- 13 6月, 2016 3 次提交
-
-
由 Andreas Scherbaum 提交于
Fix provided by @akon-dey Closes: #780
-
由 Kenan Yao 提交于
Include a map from sliceIndex to gang_id in the dispatched string, and remove the localSlice field, hence QE should get the localSlice from the map now. By this way, we avoid duplicating and modifying the dispatch text string slice by slice, and each QE of a sliced dispatch would get same contents now. The extra space cost is sizeof(int) * SliceNumber bytes, and the extra computing cost is iterating the SliceNumber-size array. Compared with memcpy of text string for each slice in previous implementation, this way is much cheaper, because SliceNumber is much smaller than the size of dispatch text string. Also, since SliceNumber is so small, we just use an array for the map instead of a hash table. Also, clean up some dead code in dispatcher, including: (1) Remove primary_gang_id field of Slice struct and DispatchCommandDtxProtocolParms struct, since dispatch agent is deprecated now; (2) Remove redundant logic in cdbdisp_dispatchX; (3) Clean up buildGpDtxProtocolCommand;
-
由 Pengzhou Tang 提交于
In commit d2725929, GPDB marked all allocatedReaderGangs with noReuse flag. When plan contains init plan and a SET command executed within it, GPDB will mark pre-assigned gangs to noReuse and destroy them which make query crash
-
- 11 6月, 2016 4 次提交
-
-
由 Foyzur Rahman 提交于
The GPDB Vmem is the lowest layer of memory allocator that supports higher allocators such as AllocSet. This layer (mostly defined in memprot.c) is in charge of actually calling malloc/realloc/free to allocate/reallocate/free memory. In this process this layer is also in charge of reserving "virtual" memory or Vmem, which is a GPDB specific shared memory counter to track per-segment combined allocations across all the GPDB processes under Vmem umbrella. The Vmem counter is managed by a separate module Vmem_Tracker, and the memprot functions (such as gp_malloc, gp_free2 and gp_realloc) call the APIs provided by VmemTracker. Previously the memprot allocators (gp_alloc/gp_realloc/gp_free) were only allocating/freeing memory but were not adding any additional metadata in the header (and there was no header) to track the size of allocations. Therefore, there was no gp_free as freeing memory requires the size of the free to adjust Vmem counter inside VmemTracker. This was patched by explicitly passing size info in gp_free2. In this PR we do the following: * We add allocation size in Vmem header (along with checksums which are only available in debug build to detect header and footer boundary, and buffer overruns). * We remove size information from the block header of AllocSet. * We rename gp_free2 to gp_free as the second parameter (size information) is now obtained from the header and therefore no longer necessary * We modify all the consumers of memprot.c APIs to use the new APIs * We add unit tests to test the metadata and the correctness of the new Vmem allocators This is the first step to integrate external modules and third party allocations with Vmem. A long running issue in GPDB is its inability to track allocations by external components including libraries such as ORCA. Therefore, the central Vmem counter is often way off from the underlying allocations, and this may run the system out of memory. By maintaining the size information in the Vmem header, we now have a self-contained allocator that can be exposed to external allocators such as GPOS allocators, without forcing them to manage size information separately. This fixes #117269929. Signed-off-by: NMarc Spehlmann <marc.spehlmann@gmail.com>
-
This GUC will be used to control the MEMO size as well as optimization time for large IN list or large array comparison expressions. Only the Array with less number of elements than the GUC will be expanded and participate in constraint derivation. Trade-off of using this GUC is loss of potential benefits from the constraint derivation (e.g. conflict detection, partition elimination) with shorter optimization time and less memory utilization.
-
由 Nikos Armenatzoglou 提交于
-
由 Nikos Armenatzoglou 提交于
This reverts commit d9edb869. g_dataSourceCtx should not be reset in AtAbort_ExtTables, since external tables is not the only component that uses it.
-
- 10 6月, 2016 2 次提交
-
-
由 Jamie McAtamney 提交于
A previous commit added the capability to gprecoverseg to copy all backup files from the mirror to the primary during a full recovery, but assumed that the backup directory would always exist and so gprecoverseg -F would fail if it was not present. This commit adds a check to ensure that gprecoverseg -F will finish successfully if there is no backup directory.
-
由 Haisheng Yuan 提交于
This patch updates minirepro utility to support the following functions: 1. Dump ddl and stats for multiple queries in a query file 2. Dump ddl of relation that is used in CTE 3. Dump ddl of function that is used in the query 4. Add 2 options: relation-oids and function-oids into pg_dump command line tool
-
- 09 6月, 2016 4 次提交
-
-
由 Heikki Linnakangas 提交于
Backport this patch from PostgreSQL 9.0, which replaces the tuple table array with a linked list of individually palloc'd slots. With that, we don't need to know the size of the array beforehand, and don't need to count the slots. The counting was especially funky for subplans in GPDB, and it was about to change with the upcoming PostgreSQL 8.3 merge again. This makes it a lot simpler. I don't plan to backport the follow-up patch to remove the ExecCountSlots infrastructure. We'll get that later, when we merge with PostgreSQL 9.0. commit f92e8a4b Author: Tom Lane <tgl@sss.pgh.pa.us> Date: Sun Sep 27 20:09:58 2009 +0000 Replace the array-style TupleTable data structure with a simple List of TupleTableSlot nodes. This eliminates the need to count in advance how many Slots will be needed, which seems more than worth the small increase in the amount of palloc traffic during executor startup. The ExecCountSlots infrastructure is now all dead code, but I'll remove it in a separate commit for clarity. Per a comment from Robert Haas.
-
由 Heikki Linnakangas 提交于
This reverts much of the changes vs. upstream, related to temp schema creation. Instead of using the normal CREATE SCHEMA processing to also create the temporary schema, let InitTempTableNameSpace() to do that like in the upstream. But in addition to creating the the temp schema locally, it dispatches a special CreateSchemaStmt command to the executor nodes, which instructs the executor nodes to also call InitTempTableNameSpace().
-
由 Heikki Linnakangas 提交于
The regression database has grown over time, so that it's just above the 1 GB size that the regression test used as a "sanity check". I think the new zlib regression test broke the camel's back. Bump it up to 5 GB, giving us about 4 GB of headroom to grow.
-
由 Shreedhar Hardikar 提交于
-
- 08 6月, 2016 4 次提交
-
-
由 Asim R P 提交于
If found inconsistent, the free tid list will be rebuilt automatically during recovery. During normal operation, super user may invoke the function gp_persistent_freelist_rebuild(OID) to rebuild the free list. A basic test case is added to verify sanity of a free tid list rebuilt using the function.
-
由 Asim R P 提交于
Break long elog messages into multiple lines, remove trailing whitespace and start elog messages with lower case.
-
由 Nikos Armenatzoglou 提交于
-
This closes #815
-
- 07 6月, 2016 7 次提交
-
-
由 Heikki Linnakangas 提交于
This allows removing the weird pfree() of the resultTupleSlot's tuple descriptor. What would've happened without the pfree() is that the old slot was allocated in the first ExecAssignScanProjectionInfo() call, in partitionMemoryContext, and then immediately destroyed when the memory context was reset. The second call to ExecAssignScanProjectionInfo() tries to free the slot, again, causing the segfault. But we can avoid that by this rearrangement of the calls in a cleaner way. In the passing, clean up the code a bit. I found having separate variables, indexState and scanState, which point to the same struct, to be confusing.
-
由 Heikki Linnakangas 提交于
ss_ScanTupleSlot is not an array, it's a single slot. The slot is allocated from a bigger array, however, so this trampled over some other slot that was allocated right after the scan slot. This has apparently been harmless, as no-one's noticed, but it's surely wrong. I bumped into this in the PostgreSQL 8.3 merge branch, where I had changed the way the slots are allocated so that they're not stored in one big array anymore. This bug led to segfaults in that case.
-
由 Heikki Linnakangas 提交于
We had later added code that allows "DEFAULT NULL", by doing a table rewrite. DEFAULT NULL is really the same as no default, so we might as well do a table rewrite for that case too, and save the code needed to handle them differently.
-
由 Heikki Linnakangas 提交于
A function might legitimately assume that an argument's varlen datum always has a 4-byte header, if the datatype is marked as 'plain', and crash if we then pass it a datum with 1-byte header, because it was packed in a memtuple. I bumped into this while working on the PostgreSQL 8.3 merge, because the merge brought us one such function: ts_rewrite().
-
由 Heikki Linnakangas 提交于
The one in nodeBitmapHeapScan.c is what the upstream has. The one in execBitmapHeapScan.c was copied from it in GPDB. No need for the duplication. It's not clear to me why we have the execBitmapHeapScan.c file at all, why not just use all the functions in nodeBitmapHeapScan.c. But I'll leave investigating that for another day.
-
由 Heikki Linnakangas 提交于
This has the effect that gpmapreduce is exempt from logging also when it uses the extended query protocol (we only to only perform the checks in exec_simple_query()). That makes no difference in practice, though, because gpmapreduce doesn't actually use the extended query protocol.
-
由 Daniel Gustafsson 提交于
The redefinitions broke the win32 build on Pulse. This was part of 3bc25384 which was backported from upstream but left this in. This makes the win32 Pulse build green.
-