pg_bulkload on PG13 fails to handle large data duplicates about pg_bulkload HOT 8 OPEN

# current master
$ pg_bulkload -V
pg_bulkload 3.1.20

# generate data
$ for i in `seq 1 500`; do echo "$i,$((i+1)),$((i+2))"; done > /tmp/in.csv

# create table
$ psql -c "SELECT version();"
                                                version
--------------------------------------------------------------------------------------------------------
 PostgreSQL 13.11 on x86_64-pc-linux-gnu, compiled by gcc (Ubuntu 9.4.0-1ubuntu1~20.04.1) 9.4.0, 64-bit
(1 row)
$ psql -c "CREATE EXTENSION pg_bulkload;"
CREATE EXTENSION
$ psql -c "CREATE TABLE test (id1 int, id2 int, id3 int, UNIQUE(id1, id2, id3));" 
CREATE TABLE

# pg_bulkload control file
$ cat /tmp/test.ctl
# output table
OUTPUT = test

# input data
WRITER = DIRECT
INPUT = /tmp/in.csv
TYPE = CSV
QUOTE = "\""
ESCAPE = \
DELIMITER = ","

# error handling
CHECK_CONSTRAINTS = YES
PARSE_ERRORS = INFINITE
DUPLICATE_ERRORS = INFINITE

# logging
LOGFILE = /tmp/pg_bulkload.log
PARSE_BADFILE = /tmp/parse_badfile.csv
DUPLICATE_BADFILE = /tmp/duplicate_badfile.csv

# test
$ pg_bulkload /tmp/test.ctl 
NOTICE: BULK LOAD START
NOTICE: BULK LOAD END
        0 Rows skipped.
        500 Rows successfully loaded.
        0 Rows not loaded due to parse errors.
        0 Rows not loaded due to duplicate errors.
        0 Rows replaced with new rows.

$ pg_bulkload /tmp/test.ctl 
NOTICE: BULK LOAD START
ERROR: query failed: ERROR:  could not create unique index "test_id1_id2_id3_key"                              
DETAIL:  Key (id1, id2, id3)=(126, 127, 128) is duplicated. 
DETAIL: query was: SELECT * FROM pgbulkload.pg_bulkload($1)  

$ for i in `seq 1 200`; do echo "$i,$((i+1)),$((i+2))"; done > /tmp/in.csv
$ psql -c "DROP TABLE test; CREATE TABLE test (id1 int, id2 int, id3 int, UNIQUE(id1, id2, id3));" 

$ pg_bulkload /tmp/test.ctl
NOTICE: BULK LOAD START
NOTICE: BULK LOAD END
        0 Rows skipped.
        200 Rows successfully loaded.
        0 Rows not loaded due to parse errors.
        0 Rows not loaded due to duplicate errors.
        0 Rows replaced with new rows.

$ pg_bulkload /tmp/test.ctl
NOTICE: BULK LOAD START
NOTICE: BULK LOAD END
        0 Rows skipped.
        200 Rows successfully loaded.
        0 Rows not loaded due to parse errors.
        0 Rows not loaded due to duplicate errors.
        200 Rows replaced with new rows.

$ wc -l /tmp/duplicate_badfile.csv
200 /tmp/duplicate_badfile.csv

$ head -n 10 /tmp/duplicate_badfile.csv
1,2,3
2,3,4
3,4,5
4,5,6
5,6,7
6,7,8
7,8,9
8,9,10
9,10,11
10,11,12

		firstid = PageGetItemId(reader->page, P_FIRSTDATAKEY(opaque));
		itup = (IndexTuple) PageGetItem(reader->page, firstid);

		if ((itup->t_tid).ip_posid == 0)
		{
			elog(DEBUG1, "pg_bulkload: failded in BTReaderInit for \"%s\"",
				RelationGetRelationName(rel));
			return -1;
		}
``` inside BTReaderInit.

I did a quick test with that **condition removed**, which sounds like a very naive solution of course... (I believe it wasn't there in the older version of pg_bulkload as it was added in 2020), then the program performs as expected. It properly identifies duplicates, removes them etc and index gets created properly.
That check there is there because of some other issue(s). It is supposed to address crashes. I did a test with memory lowered to 1MB in postgres to try to make it crash with that change, but that didn't happen. 
That PR with that check, had also few other changes, so perhaps those other changes address it? or perhaps my dataset was too small to make it crash with