table: Reduce read amplification in view update generation

This commit makes sure that single-partition readers for
read-before-write do not have fast-forwarding enabled,
as it may lead to huge read amplification. The observed case was:
1. Creating an index.
  CREATE INDEX index1  ON myks2.standard1 ("C1");
2. Running cassandra-stress in order to generate view updates.
cassandra-stress write no-warmup n=1000000 cl=ONE -schema \
  'replication(factor=2) compaction(strategy=LeveledCompactionStrategy)' \
  keyspace=myks2 -pop seq=4000000..8000000 -rate threads=100 -errors
  skip-read-validation -node 127.0.0.1;

Without disabling fast-forwarding, single-partition readers
were turned into scanning readers in cache, which resulted
in reading 36GB (sic!) on a workload which generates less
than 1GB of view updates. After applying the fix, the number
dropped down to less than 1GB, as expected.

Refs #5409
Fixes #4615
Fixes #5418

(cherry picked from commit 79c3a508f4)

table.cc

@@ -2543,7 +2543,7 @@ future<row_locker::lock_holder> table::do_push_view_replica_updates(const schema
         std::move(slice),
         std::move(m),
         [base, views = std::move(views), lock = std::move(lock), this, timeout, source = std::move(source), &io_priority] (auto& pk, auto& slice, auto& m) mutable {
-            auto reader = source.make_reader(base, pk, slice, io_priority);
+            auto reader = source.make_reader(base, pk, slice, io_priority, nullptr, streamed_mutation::forwarding::no, mutation_reader::forwarding::no);
             return this->generate_and_propagate_view_updates(base, std::move(views), std::move(m), std::move(reader)).then([lock = std::move(lock)] () mutable {
                 // return the local partition/row lock we have taken so it
                 // remains locked until the caller is done modifying this