> > "emsg" is leaked in some cases when running rpmbuild, because "emsg"
> > may be alloced and be set when checkDep returns RPMRC_OK.
>
> If emsg is allocated when checkDep() returns without an error, then perhaps
> _that_ should be fixed instead I think. What's the case where that happens, I
> didn't see anything immediately obvious there?
When `checkSep(EVR, '-', emsg)` or `checkSep(EVR, ':', emsg)` or
`checkEpoch(EVR, emsg)` returns RPMRC_FAIL, and
`_wrong_version_format_terminate_build` is defined as 0, the checkDep() will
returns RPMRC_OK and emsg is set.
```
if (checkSep(EVR, '-', emsg) != RPMRC_OK ||
checkSep(EVR, ':', emsg) != RPMRC_OK ||
checkEpoch(EVR, emsg) != RPMRC_OK) {
if (rpmExpandNumeric("%{?_wrong_version_format_terminate_build}"))
return RPMRC_FAIL;
}
```
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/2293#issuecomment-1327336296
You are receiving this because you are subscribed to this thread.
Message ID: <rpm-software-management/rpm/pull/2293/c1327336...@github.com>_______________________________________________
Rpm-maint mailing list
Rpm-maint@lists.rpm.org
http://lists.rpm.org/mailman/listinfo/rpm-maint
