Bug 2248862

Description of problem:
rgw going down executing trim command with integer as "remove chars"

[cephuser@extensa022 ~]$ venv/bin/aws s3api --endpoint-url http://extensa027.ceph.redhat.com:80 select-object-content  --bucket bkt1 --key polarion_data.csv --expression-type 'SQL' --input-serialization '{"CSV": {}, "CompressionType": "NONE"}' --output-serialization '{"CSV": {}}' --expression "select trim( leading 132140533849470 from _3 ) from s3object;" /dev/stdout

Could not connect to the endpoint URL: "http://extensa027.ceph.redhat.com:80/bkt1/polarion_data.csv?select&select-type=2"
[cephuser@extensa022 ~]$


crash info:

     0> 2023-11-09T11:44:01.109+0000 7f941d32b640 -1 *** Caught signal (Segmentation fault) **
 in thread 7f941d32b640 thread_name:radosgw

 ceph version 18.2.0-116.el9cp (6c6dfc1f2b2ce1896bf2696daea444dff96645af) reef (stable)
 1: /lib64/libc.so.6(+0x54df0) [0x7f94b1050df0]
 2: /lib64/libc.so.6(+0xc56fd) [0x7f94b10c16fd]
 3: /usr/bin/radosgw(+0x6326ec) [0x558c60cf56ec]
 4: /usr/bin/radosgw(+0x637272) [0x558c60cfa272]
 5: /usr/bin/radosgw(+0xbba3b9) [0x558c6127d3b9]
 6: /usr/bin/radosgw(+0x644d9c) [0x558c60d07d9c]
 7: (RGWSelectObj_ObjStore_S3::run_s3select_on_csv(char const*, char const*, unsigned long)+0x8cb) [0x558c60d1483b]
 8: (RGWSelectObj_ObjStore_S3::csv_processing(ceph::buffer::v15_2_0::list&, long, long)+0x507) [0x558c60d18377]
 9: (RGWGetObj_BlockDecrypt::process(ceph::buffer::v15_2_0::list&, unsigned long, unsigned long)+0x9a) [0x558c60d4bf7a]
 10: (RGWGetObj_BlockDecrypt::handle_data(ceph::buffer::v15_2_0::list&, long, long)+0x1ae) [0x558c60d53dde]
 11: (RGWRados::get_obj_iterate_cb(DoutPrefixProvider const*, rgw_raw_obj const&, long, long, long, bool, RGWObjState*, void*)+0x11a) [0x558c60e0e7fa]
 12: /usr/bin/radosgw(+0x747846) [0x558c60e0a846]
 13: (RGWRados::iterate_obj(DoutPrefixProvider const*, RGWObjectCtx&, RGWBucketInfo&, rgw_obj const&, long, long, unsigned long, int (*)(DoutPrefixProvider const*, rgw_raw_obj const&, long, long, long, bool, RGWObjState*, void*), void*, optional_yield)+0x428) [0x558c60e0f0f8]
 14: (RGWRados::Object::Read::iterate(DoutPrefixProvider const*, long, long, RGWGetDataCB*, optional_yield)+0x134) [0x558c60e0f8b4]
 15: (RGWGetObj::execute(optional_yield)+0x122c) [0x558c60c304bc]
 16: (RGWSelectObj_ObjStore_S3::execute(optional_yield)+0xc1) [0x558c60d1ac41]
 17: (rgw_process_authenticated(RGWHandler_REST*, RGWOp*&, RGWRequest*, req_state*, optional_yield, rgw::sal::Driver*, bool)+0xa72) [0x558c60ae4c52]
 18: (process_request(RGWProcessEnv const&, RGWRequest*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, RGWRestfulIO*, optional_yield, rgw::dmclock::Scheduler*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::chrono::duration<unsigned long, std::ratio<1l, 1000000000l> >*, int*)+0x1039) [0x558c60ae64d9]
 19: /usr/bin/radosgw(+0xb6ddf6) [0x558c61230df6]
 20: /usr/bin/radosgw(+0x37bc51) [0x558c60a3ec51]
 21: make_fcontext()
 NOTE: a copy of the executable, or `objdump -rdS <executable>` is needed to interpret this.



similar commands which caused rgw down:

executing query-2335: select cast( trim( trailing 205979192405432 from 'kQXLXjV283jqvsSVtUQj' ) as string) from s3object;

executing query-3754: select sum( trim( both 112839306443463 from '2g rL44I0tRJi1tLS1WE' ) ) from s3object;

executing query-4620: select trim( both min( 153401582910812 ) from min( 'Dwj5M28JLOnd2dxm6V0n' ) ) from s3object;

executing query-5605: select extract( year from trim( trailing 172215640942740 from 'NfEZRuJaDY87tLjrrmVL' ) ) from s3object;

executing query-5933: select extract( day from trim( leading 170191654430978 from 'MCPSK3qVHk9fnZx9x6ld' ) ) from s3object;

executing query-7243: select to_string( trim( both 118424875481762 from 'Hb dGpdmYZ486VGWKekF' ) ,  yy) from s3object;


Version-Release number of selected component (if applicable):
ceph version 18.2.0-116.el9cp

How reproducible:
always

Steps to Reproduce:
1.deploy rhcs7.0 ceph cluster
2.upload an csv object using aws-cli
3.execute the query "select trim( leading 132140533849470 from _3 ) from s3object;"

Actual results:
rgw daemon going down

Expected results:
no rgw crashes should be seen

Additional info:
automation script: https://docs.google.com/document/d/1eg2IbKpGWlGD8dG9PX8oDVKC77xNN-URU-A_r68KUmg/edit?usp=sharing