Re: BUG #16162: create index using gist_trgm_ops leads to panic

Hello Heikki,
14.12.2019 1:02, Heikki Linnakangas wrote:
> Like Tom suspected at [1], the bug was that the parent page got split,
> and the stacked information wasn't updated. The code called
> gistFindCorrectParent(), which should've updated the stack, but that
> function checked the LSN of the page and did nothing if it matched. To
> trigger the bug, you needed to have an insertion that split a page
> into three (or more) pages, and inserting the downlink caused the
> parent page to split, too.
>
> Committed a fix. Many thanks for the reproducer scripts, Andreas and
> Alexander!
The script that I presented in [0] still reproduces assertion failure
for me on REL_12_STABLE (fd005e1a):

iteration 6
psql:/tmp/test_gist_intarray.sql:2: NOTICE:  extension "intarray"
already exists, skipping
CREATE EXTENSION
DROP TABLE
CREATE TABLE
COPY 7000
INSERT 0 7000
INSERT 0 14000
INSERT 0 28000
INSERT 0 56000
psql:/tmp/test_gist_intarray.sql:10: server closed the connection
unexpectedly
        This probably means the server terminated abnormally
        before or while processing the request.
psql:/tmp/test_gist_intarray.sql:10: fatal: connection to server was lost

Core was generated by `postgres: law regression [local] CREATE
INDEX                                 '.
Program terminated with signal SIGABRT, Aborted.
#0  __GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:51
51      ../sysdeps/unix/sysv/linux/raise.c: No such file or directory.
(gdb) bt
#0  __GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:51
#1  0x00007f10c0ed4801 in __GI_abort () at abort.c:79
#2  0x000055cfe1d476a6 in ExceptionalCondition (
    conditionName=conditionName@entry=0x55cfe1db42a8 "!(((((ItemId)
(&((PageHeader) (page))->pd_linp[(todelete) - 1])))->lp_len != 0))",
errorType=errorType@entry=0x55cfe1d9e788 "FailedAssertion",
    fileName=fileName@entry=0x55cfe1db4250 "gistutil.c",
lineNumber=lineNumber@entry=70) at assert.c:54
#3  0x000055cfe18e3dc0 in gistnospace (page=page@entry=0x7f10b8705d00
"", itvec=itvec@entry=0x7ffec1b676e0,
    len=len@entry=2, todelete=todelete@entry=5,
freespace=freespace@entry=819) at gistutil.c:70
#4  0x000055cfe18e1301 in gistplacetopage (rel=0x7f10c2006500,
freespace=819,
    giststate=giststate@entry=0x55cfe3934ec0, buffer=920,
itup=itup@entry=0x7ffec1b676e0, ntup=ntup@entry=2,
    oldoffnum=5, newblkno=0x0, leftchildbuf=897,
splitinfo=0x7ffec1b67670, markfollowright=true,
    heapRel=0x7f10c20107a0, is_build=true) at gist.c:257
...
[0]
https://www.postgresql.org/message-id/16134-0423f729671dec64%40postgresql.org
>
> [1] https://www.postgresql.org/message-id/9409.1574617130%40sss.pgh.pa.us
>
> - Heikki
Best regards,
Alexander