ecpg assertion on windows - Mailing list pgsql-hackers

Hi,

I occasionally

Running the ecpg regression tests interactively (to try to find a different
issue), triggered a crash on windows due to an uninitialized variable (after
pressing "ignore" in that stupid gui window that we've only disabled for the
backend).

"The variable 'replace_val' is being used without being initialized."

Child-SP          RetAddr               Call Site
000000b3`3bcfe140 00007ff9`03f9cd74     libpgtypes!failwithmessage(
                        void * retaddr = 0x00007ff9`03f96133,
                        int crttype = 0n1,
                        int errnum = 0n3,
                        char * msg = 0x000000b3`3bcff050 "The variable 'replace_val' is being used without being
initialized.")+0x234[d:\a01\_work\12\s\src\vctools\crt\vcstartup\src\rtc\error.cpp @ 213]
 
000000b3`3bcff030 00007ff9`03f96133     libpgtypes!_RTC_UninitUse(
                        char * varname = 0x00007ff9`03fa8a90 "replace_val")+0xa4
[d:\a01\_work\12\s\src\vctools\crt\vcstartup\src\rtc\error.cpp@ 362]
 
000000b3`3bcff470 00007ff9`03f94acd     libpgtypes!dttofmtasc_replace(
                        int64 * ts = 0x000000b3`3bcff778,
                        long dDate = 0n0,
                        int dow = 0n6,
                        struct tm * tm = 0x000000b3`3bcff598,
                        char * output = 0x0000026e`9c223de0 "abc-00:00:00",
                        int * pstr_len = 0x000000b3`3bcff620,
                        char * fmtstr = 0x00007ff7`b01ae5c0 "abc-%X-def-%x-ghi%%")+0xe53
[C:\dev\postgres-meson\src\interfaces\ecpg\pgtypeslib\timestamp.c@ 759]
 
*** WARNING: Unable to verify checksum for
C:\dev\postgres-meson\build-msbuild\src\interfaces\ecpg\test\pgtypeslib\dt_test.exe
000000b3`3bcff550 00007ff7`b01a23c9     libpgtypes!PGTYPEStimestamp_fmt_asc(
                        int64 * ts = 0x000000b3`3bcff778,
                        char * output = 0x0000026e`9c223de0 "abc-00:00:00",
                        int str_len = 0n19,
                        char * fmtstr = 0x00007ff7`b01ae5c0 "abc-%X-def-%x-ghi%%")+0xed
[C:\dev\postgres-meson\src\interfaces\ecpg\pgtypeslib\timestamp.c@ 794]
 
000000b3`3bcff610 00007ff7`b01a4499     dt_test!main(void)+0xe59
[C:\dev\postgres-meson\src\interfaces\ecpg\test\pgtypeslib\dt_test.pgc@ 200]
 
000000b3`3bcff860 00007ff7`b01a433e     dt_test!invoke_main(void)+0x39
[d:\a01\_work\12\s\src\vctools\crt\vcstartup\src\startup\exe_common.inl@ 79]
 
000000b3`3bcff8b0 00007ff7`b01a41fe     dt_test!__scrt_common_main_seh(void)+0x12e
[d:\a01\_work\12\s\src\vctools\crt\vcstartup\src\startup\exe_common.inl@ 288]
 
000000b3`3bcff920 00007ff7`b01a452e     dt_test!__scrt_common_main(void)+0xe
[d:\a01\_work\12\s\src\vctools\crt\vcstartup\src\startup\exe_common.inl@ 331]
 
000000b3`3bcff950 00007ff9`1d987034     dt_test!mainCRTStartup(
                        void * __formal = 0x000000b3`3bbe8000)+0xe
[d:\a01\_work\12\s\src\vctools\crt\vcstartup\src\startup\exe_main.cpp@ 17]
 
000000b3`3bcff980 00007ff9`1f842651     KERNEL32!BaseThreadInitThunk+0x14
000000b3`3bcff9b0 00000000`00000000     ntdll!RtlUserThreadStart+0x21

I haven't analyzed this further.


CI also shows ecpg itself occasionally crashing, but I haven't managed to
catch it in the act.

Greetings,

Andres Freund