Fix segfault on non-open primary dimension when compressing

With the new hypertable API hypertables can be created with
primary space partition. In dev-builds this was prevented with
Asserts. This patch removes the Asserts and adds a proper check.

(cherry picked from commit 4dcbc26)

.unreleased/pr_6978

@@ -0,0 +1 @@
+Fixes: #6978 Fix segfault in compress_chunk with primary space partition

src/subspace_store.c

@@ -75,12 +75,6 @@ ts_subspace_store_init(const Hyperspace *space, MemoryContext mcxt, int16 max_it
 	MemoryContext old = MemoryContextSwitchTo(mcxt);
 	SubspaceStore *sst = palloc(sizeof(SubspaceStore));
 
-	/*
-	 * make sure that the first dimension is a time dimension, otherwise the
-	 * tree will grow in a way that makes pruning less effective.
-	 */
-	Assert(space->num_dimensions < 1 || space->dimensions[0].type == DIMENSION_TYPE_OPEN);
-
 	sst->origin = subspace_store_internal_node_create(space->num_dimensions == 1);
 	sst->num_dimensions = space->num_dimensions;
 	/* max_items = 0 is treated as unlimited */

tsl/src/compression/api.c

@@ -256,9 +256,7 @@ find_chunk_to_merge_into(Hypertable *ht, Chunk *current_chunk)
 
 	const Dimension *time_dim = hyperspace_get_open_dimension(ht->space, 0);
 
-	Assert(time_dim != NULL);
-
-	if (time_dim->fd.compress_interval_length == 0)
+	if (!time_dim || time_dim->fd.compress_interval_length == 0)
 		return NULL;
 
 	Assert(current_chunk->cube->num_slices > 0);

tsl/test/expected/compression_merge.out

@@ -820,3 +820,22 @@ NOTICE:  chunk "_hyper_17_344_chunk" is already compressed
 (1 row)
 
 ROLLBACK;
+-- test segfault when compression hypertable with primary space dimension #6977
+CREATE TABLE test_by_hash(id BIGINT, value float8);
+SELECT create_hypertable('test_by_hash', by_hash('id', 8));
+ create_hypertable 
+-------------------
+ (19,t)
+(1 row)
+
+ALTER TABLE test_by_hash SET (timescaledb.compress = true);
+WARNING:  there was some uncertainty picking the default segment by for the hypertable: You do not have any indexes on columns that can be used for segment_by and thus we are not using segment_by for compression. Please make sure you are not missing any indexes
+NOTICE:  default segment by for hypertable "test_by_hash" is set to ""
+NOTICE:  default order by for hypertable "test_by_hash" is set to "id DESC"
+INSERT INTO test_by_hash VALUES (1, 1.0), (2, 2.0), (3, 3.0);
+SELECT compress_chunk('_timescaledb_internal._hyper_19_351_chunk');
+              compress_chunk               
+-------------------------------------------
+ _timescaledb_internal._hyper_19_351_chunk
+(1 row)
+

tsl/test/sql/compression_merge.sql

@@ -308,3 +308,10 @@ BEGIN;
   SELECT hypertable_name, range_start, range_end FROM timescaledb_information.chunks WHERE hypertable_name = 'test9' ORDER BY 2;
 ROLLBACK;
 
+-- test segfault when compression hypertable with primary space dimension #6977
+CREATE TABLE test_by_hash(id BIGINT, value float8);
+SELECT create_hypertable('test_by_hash', by_hash('id', 8));
+ALTER TABLE test_by_hash SET (timescaledb.compress = true);
+INSERT INTO test_by_hash VALUES (1, 1.0), (2, 2.0), (3, 3.0);
+SELECT compress_chunk('_timescaledb_internal._hyper_19_351_chunk');
+