not working

.dockerignore

@@ -21,4 +21,5 @@
 !workspace_hack/
 !neon_local/
 !scripts/ninstall.sh
+!scripts/combine_control_files.py
 !vm-cgconfig.conf

.github/workflows/build_and_test.yml

@@ -917,9 +917,9 @@ jobs:
         run: rm -rf ~/.ecr
 
   upload-postgres-extensions-to-s3:
-    if: |
-      (github.ref_name == 'main' || github.ref_name == 'release') &&
-       github.event_name != 'workflow_dispatch'
+    # if: |
+    #   (github.ref_name == 'main' || github.ref_name == 'release') &&
+    #    github.event_name != 'workflow_dispatch'
     runs-on: ${{ github.ref_name == 'release' && fromJSON('["self-hosted", "prod", "x64"]') || fromJSON('["self-hosted", "gen3", "small"]') }}
     needs: [ tag, promote-images ]
     strategy:
@@ -929,7 +929,7 @@ jobs:
 
     env:
       # While on transition period we extract public extensions from compute-node image and custom extensions from extensions image.
-      # Later all the extensions will be moved to extensions image.
+      # Later all the extensions will be moved to extensions image. (unless that is too slow)
       EXTENSIONS_IMAGE: ${{ github.ref_name == 'release' && '093970136003' || '369495373322'}}.dkr.ecr.eu-central-1.amazonaws.com/extensions-${{ matrix.version }}:latest
       COMPUTE_NODE_IMAGE: ${{ github.ref_name == 'release' && '093970136003' || '369495373322'}}.dkr.ecr.eu-central-1.amazonaws.com/compute-node-${{ matrix.version }}:latest
       AWS_ACCESS_KEY_ID: ${{ github.ref_name == 'release' && secrets.AWS_ACCESS_KEY_PROD || secrets.AWS_ACCESS_KEY_DEV }}
@@ -956,33 +956,23 @@ jobs:
 
       - name: Extract postgres-extensions from container
         run: |
-          rm -rf ./extensions-to-upload ./custom-extensions # Just in case
+          rm -rf ./extensions ./control_files # Just in case
+          mkdir extensions
+          mkdir control_files
 
-          # In compute image we have a bit different directory layout
-          mkdir -p extensions-to-upload/share
-          docker cp ${{ steps.create-container.outputs.CID }}:/usr/local/share/extension ./extensions-to-upload/share/extension
-          docker cp ${{ steps.create-container.outputs.CID }}:/usr/local/lib             ./extensions-to-upload/lib
+          # TODO: Delete Neon extensitons (they always present on compute-node image)
+          # rm -rf ./extensions/share/extension/neon*
+          # rm -rf ./extensions/lib/neon*
 
-          # Delete Neon extensitons (they always present on compute-node image)
-          rm -rf ./extensions-to-upload/share/extension/neon*
-          rm -rf ./extensions-to-upload/lib/neon*
-
-          # Delete leftovers from the extension build step
-          rm -rf ./extensions-to-upload/lib/pgxs
-          rm -rf ./extensions-to-upload/lib/pkgconfig
-
-          docker cp ${{ steps.create-container.outputs.EID }}:/extensions ./custom-extensions
-          for EXT_NAME in $(ls ./custom-extensions); do
-            mkdir -p ./extensions-to-upload/${EXT_NAME}/share
-
-            mv ./custom-extensions/${EXT_NAME}/share/extension ./extensions-to-upload/${EXT_NAME}/share/extension
-            mv ./custom-extensions/${EXT_NAME}/lib             ./extensions-to-upload/${EXT_NAME}/lib
-          done
+          docker cp ${{ steps.create-container.outputs.EID }}:/ ./
+          # sh ./scripts/combine_control_files.sh
+          echo '{ "enabled_extensions": { "123454321": [ "anon" ], "public": [ "embedding" ] }, "control_data": { "embedding": "comment = 'hnsw index' \ndefault_version = '0.1.0' \nmodule_pathname = '$libdir/embedding' \nrelocatable = true \ntrusted = true", "anon": "# PostgreSQL Anonymizer (anon) extension \ncomment = 'Data anonymization tools' \ndefault_version = '1.1.0' \ndirectory='extension/anon' \nrelocatable = false \nrequires = 'pgcrypto' \nsuperuser = false \nmodule_pathname = '$libdir/anon' \ntrusted = true \n" } }' > ext_index.json
 
       - name: Upload postgres-extensions to S3
         run: |
           for BUCKET in $(echo ${S3_BUCKETS}); do
-            aws s3 cp --recursive --only-show-errors ./extensions-to-upload s3://${BUCKET}/${{ needs.tag.outputs.build-tag }}/${{ matrix.version }}
+            aws s3 cp --recursive --only-show-errors ./extensions s3://${BUCKET}/${{ needs.tag.outputs.build-tag }}/${{ matrix.version }}
+            aws s3 cp --only-show-errors ./ext_index.json s3://${BUCKET}/${{ needs.tag.outputs.build-tag }}/${{ matrix.version }}
           done
 
       - name: Cleanup

Dockerfile.compute-node

@@ -530,9 +530,15 @@ ENV PATH "/usr/local/pgsql/bin/:$PATH"
 RUN wget https://github.com/neondatabase/pg_embedding/archive/2465f831ea1f8d49c1d74f8959adb7fc277d70cd.tar.gz -O pg_embedding.tar.gz && \
     echo "047af2b1f664a1e6e37867bd4eeaf5934fa27d6ba3d6c4461efa388ddf7cd1d5 pg_embedding.tar.gz" | sha256sum --check && \
     mkdir pg_embedding-src && cd pg_embedding-src && tar xvzf ../pg_embedding.tar.gz --strip-components=1 -C . && \
+    find /usr/local/pgsql -type f | sort  > /before_embedding.txt && \
     make -j $(getconf _NPROCESSORS_ONLN) && \
     make -j $(getconf _NPROCESSORS_ONLN) install && \
-    echo 'trusted = true' >> /usr/local/pgsql/share/extension/embedding.control
+    echo 'trusted = true' >> /usr/local/pgsql/share/extension/embedding.control &&\
+    find /usr/local/pgsql -type f | sort  > /after_embedding.txt &&\
+    /bin/bash -c 'for from in $(comm -13 /before_embedding.txt /after_embedding.txt); do to=/extensions/embedding/${from:17} && mkdir -p $(dirname ${to}) && cp -a ${from} ${to}; done' && \
+    tar -zcvf /extensions/embedding.tar.gz /extensions/embedding && \
+    mkdir -p /control_files &&\
+    cp /usr/local/pgsql/share/extension/embedding.control /control_files/embedding.control
 
 #########################################################################################
 #
@@ -552,7 +558,15 @@ RUN wget https://gitlab.com/dalibo/postgresql_anonymizer/-/archive/1.1.0/postgre
     make -j $(getconf _NPROCESSORS_ONLN) install PG_CONFIG=/usr/local/pgsql/bin/pg_config && \
     echo 'trusted = true' >> /usr/local/pgsql/share/extension/anon.control && \
     find /usr/local/pgsql -type f | sort  > /after.txt && \
-    /bin/bash -c 'for from in $(comm -13 /before.txt /after.txt); do to=/extensions/anon/${from:17} && mkdir -p $(dirname ${to}) && cp -a ${from} ${to}; done'
+    /bin/bash -c 'for from in $(comm -13 /before.txt /after.txt); do to=/extensions/anon/${from:17} && mkdir -p $(dirname ${to}) && cp -a ${from} ${to}; done' && \
+    mkdir /control_files &&\
+    tar -zcvf /extensions/anon.tar.gz /extensions/anon && \
+    mkdir -p /control_files &&\
+    cp /usr/local/pgsql/share/extension/anon.control /control_files/anon.control
+
+# # TODO: Delete leftovers from the extension build step
+# rm -rf ./extensions/lib/pgxs
+# rm -rf ./extensions/lib/pkgconfig
 
 #########################################################################################
 #
@@ -741,19 +755,24 @@ RUN rm /usr/local/pgsql/lib/lib*.a
 
 #########################################################################################
 #
-# Extenstion only
+# Extension only
 #
 #########################################################################################
 FROM scratch AS postgres-extensions
 # After the transition this layer will include all extensitons.
 # As for now, it's only for new custom ones
 #
-# # Default extensions
-# COPY --from=postgres-cleanup-layer /usr/local/pgsql/share/extension /usr/local/pgsql/share/extension
-# COPY --from=postgres-cleanup-layer /usr/local/pgsql/lib             /usr/local/pgsql/lib
-# Custom extensions
-COPY --from=pg-anon-pg-build /extensions/anon/lib/ /extensions/anon/lib
-COPY --from=pg-anon-pg-build /extensions/anon/share/extension /extensions/anon/share/extension
+# extensions
+COPY --from=pg-anon-pg-build /extensions/anon.tar.gz /extensions/anon.tar.gz
+COPY --from=pg-anon-pg-build /control_files/anon.control /control_files/anon.control
+
+COPY --from=pg-embedding-pg-build /extensions/embedding.tar.gz /extensions/embedding.tar.gz
+COPY --from=pg-embedding-pg-build /control_files/embedding.control /control_files/embedding.control
+
+FROM python:3.11-slim-bookworm AS alekpython
+COPY scripts/combine_control_files.py combine_control_files.py
+COPY --from=postgres-extensions /control_files /control_files
+CMD [ "python3", "./combine_control_files.py"]
 
 #########################################################################################
 #

libs/postgres_ffi/src/nonrelfile_utils.rs

@@ -57,9 +57,9 @@ pub fn slru_may_delete_clogsegment(segpage: u32, cutoff_page: u32) -> bool {
 // Multixact utils
 
 pub fn mx_offset_to_flags_offset(xid: MultiXactId) -> usize {
-    ((xid / pg_constants::MULTIXACT_MEMBERS_PER_MEMBERGROUP as u32)
-        % pg_constants::MULTIXACT_MEMBERGROUPS_PER_PAGE as u32
-        * pg_constants::MULTIXACT_MEMBERGROUP_SIZE as u32) as usize
+    ((xid / pg_constants::MULTIXACT_MEMBERS_PER_MEMBERGROUP as u32) as u16
+        % pg_constants::MULTIXACT_MEMBERGROUPS_PER_PAGE
+        * pg_constants::MULTIXACT_MEMBERGROUP_SIZE) as usize
 }
 
 pub fn mx_offset_to_flags_bitshift(xid: MultiXactId) -> u16 {

pgxn/hnsw/hnsw.c

@@ -149,7 +149,7 @@ hnsw_check_available_memory(Size requested)
 	struct sysinfo si;
 	Size total;
 	if (sysinfo(&si) < 0)
-		elog(ERROR, "Failed to get amount of RAM: %m");
+		elog(ERROR, "Failed to get amount of RAM: %n");
 
 	total = si.totalram*si.mem_unit;
 	if ((Size)NBuffers*BLCKSZ + requested >= total)
@@ -580,7 +580,6 @@ l2_distance(PG_FUNCTION_ARGS)
 				 errmsg("different array dimensions %d and %d", a_dim, b_dim)));
 	}
 
-	#pragma clang loop vectorize(enable)
 	for (int i = 0; i < a_dim; i++)
 	{
 		diff = ax[i] - bx[i];

pgxn/hnsw/hnswalg.cpp

@@ -223,7 +223,6 @@ dist_t fstdistfunc_scalar(const coord_t *x, const coord_t *y, size_t n)
 {
     dist_t 	distance = 0.0;
 
-    #pragma clang loop vectorize(enable)
     for (size_t i = 0; i < n; i++)
     {
         dist_t diff = x[i] - y[i];

scripts/combine_control_files.py

@@ -0,0 +1,20 @@
+import json
+import os
+
+with open("SOMETHING", "w") as f:
+    f.write("SOMETHING")
+
+# enable custom extensions for specific tenants
+enabled_extensions = {"123454321": ["anon"], "public": ["embedding"]}
+
+control_data = {}
+os.chdir("control_files")
+for control_file in os.listdir("."):
+    ext_name = control_file.replace(".control", "")
+    with open(control_file, "r") as f:
+        control_data[ext_name] = f.read()
+
+all_data = {"enabled_extensions": enabled_extensions, "control_data": control_data}
+
+with open("ext_index.json", "w") as f:
+    json.dump(all_data, f)

test_runner/regress/test_multixact.py

@@ -1,7 +1,3 @@
-import random
-import threading
-from threading import Thread
-
 from fixtures.log_helper import log
 from fixtures.neon_fixtures import NeonEnv, check_restored_datadir_content
 from fixtures.utils import query_scalar
@@ -19,17 +15,11 @@ def test_multixact(neon_simple_env: NeonEnv, test_output_dir):
     endpoint = env.endpoints.create_start("test_multixact")
 
     log.info("postgres is running on 'test_multixact' branch")
-
-    n_records = 100
-    n_threads =   5
-    n_iters =  1000
-    n_restarts = 10
-
     cur = endpoint.connect().cursor()
     cur.execute(
-        f"""
-        CREATE TABLE t1(pk int primary key, val integer);
-        INSERT INTO t1 values (generate_series(1, {n_records}), 0);
+        """
+        CREATE TABLE t1(i int primary key);
+        INSERT INTO t1 select * from generate_series(1, 100);
     """
     )
 
@@ -38,32 +28,26 @@ def test_multixact(neon_simple_env: NeonEnv, test_output_dir):
     )
 
     # Lock entries using parallel connections in a round-robin fashion.
-    def do_updates():
+    nclients = 20
+    connections = []
+    for i in range(nclients):
+        # Do not turn on autocommit. We want to hold the key-share locks.
         conn = endpoint.connect(autocommit=False)
-        for i in range(n_iters):
-            pk = random.randrange(1, n_records)
-            conn.cursor().execute(f"update t1 set val=val+1 where pk={pk}")
-            conn.cursor().execute("select * from t1 for key share")
-            conn.commit()
-        conn.close()
+        connections.append(conn)
 
-    for iter in range(n_restarts):
-        threads: List[threading.Thread] = []
-        for i in range(n_threads):
-            threads.append(threading.Thread(target=do_updates, args=(), daemon=False))
-            threads[-1].start()
+    # On each iteration, we commit the previous transaction on a connection,
+    # and issue antoher select. Each SELECT generates a new multixact that
+    # includes the new XID, and the XIDs of all the other parallel transactions.
+    # This generates enough traffic on both multixact offsets and members SLRUs
+    # to cross page boundaries.
+    for i in range(5000):
+        conn = connections[i % nclients]
+        conn.commit()
+        conn.cursor().execute("select * from t1 for key share")
 
-        for thread in threads:
-            thread.join()
-
-        # Restart endpoint
-        endpoint.stop()
-        endpoint.start()
-
-        conn = endpoint.connect()
-        cur = conn.cursor()
-        cur.execute("select count(*) from t1")
-        assert cur.fetchone() == (n_records,)
+    # We have multixacts now. We can close the connections.
+    for c in connections:
+        c.close()
 
     # force wal flush
     cur.execute("checkpoint")
@@ -90,3 +74,6 @@ def test_multixact(neon_simple_env: NeonEnv, test_output_dir):
 
     # Check that we restored pg_controlfile correctly
     assert next_multixact_id_new == next_multixact_id
+
+    # Check that we can restore the content of the datadir correctly
+    check_restored_datadir_content(test_output_dir, env, endpoint)