> On Aug 27, 2021, at 2:51 AM, Heiko Schlittermann via Exim-users <exim-users@??? <mailto:exim-users@exim.org>> wrote:
>
> Chad Leigh via Exim-users <exim-users@??? <mailto:exim-users@exim.org>> (Mi 25 Aug 2021 23:30:49 CEST):
>>
>> Hi
>>
>> (I normally am chad@??? <mailto:chad@shire.net> but my smtp server is down at the moment and is the reason for this post)
>>
>> I screwed something on on my SmartOS (Solaris) based system that was running exim4. I updated some system librarues and my existing build was not finding the stuff it was looking for. I wanted to update to the latest anyway and I use my own build due to some site specific stuff I build in. I got the latest 4.94.2 and am trying to do a “make” on it.
>>
>> I am getting a bunch of basic errors that look like language errors… I have tried with gcc49, gcc6, and gcc9
>>
>> This is what I am seeing (and a lot more) and I am not sure what to do next. I would expect that the default codebase should build with gcc.
>>
>>
>> # make
>> /bin/sh scripts/source_checks
>> `Makefile' is up to date.
>>
>> make[1]: Entering directory '/opt/build/exim-4.94.2/build-SunOS5-5.11-i386'
>> gcc -DMACRO_PREDEF macro_predef.c
>> In file included from exim.h:526:0,
>> from macro_predef.c:12:
>> dbstuff.h:693:3: error: unknown type name 'BOOL'
>> BOOL expired; /* Retry time has expired */
>
> After having a short glance, it seems the file src/mytypes.h having the
> typedef for BOOL.
>
I got it running (and did not see the post from my other address, due to my main mail address being down, until Friday)…
The problem was I had made a mistake in copying my local_scan files from the old build directory to the new version build directory and so had accidentally excluded the system local_scan.h, which is where the problem arose. Once I figured out the problem, it built. I had annother issue wth “string_sprintf” as an unresolved symbol, which I will post about separately. I worked around it and got up running but it seems the problem shouldn’t have happened. I will post a different thread when I have time.
But this problem was a faulty move of the local_scan stuff causing local_scan.h to notg be used properly.
