Ilya Shipitsin

I'm not sure it is reproducible. I ran scan.coverity.com build It might be false positive

do you mind of getting this build on app veyor ?

agreed. I'll prepare pull request. (it will take some time, I'm not sure whether "yc" is available for arm64)

taking into account the complexity of postgresql protocol, wouldn't it be better to implement proxy protocol support on postgresql side ? https://www.postgresql.org/message-id/20190519155903.GI6197%40tamriel.snowman.net

I tried GCC on fedora rawhide, but I do not recall that I tried GCC master branch On Sat, Oct 15, 2022, 4:50 PM Shawn Heisey ***@***.***> wrote: > I...

it is clang "build time" check. we can add "pragma" maybe

``` #pragma clang diagnostic push #pragma clang diagnostic ignored "-Wunused-variable" #pragma clang diagnostic pop ```

you have accidently (?) exposed password in `userlist backoffice_users` not sure how sensitive it is

here's longer report ``` 5889const char *hash_anon(uint32_t scramble, const char *string2hash, const char *prefix, const char *suffix) 5890{ 5891 index_hash++; 1. Condition index_hash > 7, taking false branch. 2. cond_at_most:...

this has raised just few days ago because is was detected by Fedora Rawhide which has switched to clang-15. I guess it is new rule in clang-15