make universe errors..

[Julian Elischer]

Poul-Henning Kamp wrote:
> In message <4825CF8D.8020202 at elischer.org>, Julian Elischer writes:
>> Poul-Henning Kamp wrote:
>>>> how many errors in a make universe are expected?
>>> None.
>>>
>>> But we do see some warnings that people probably should pay attention to. 
>>>
>>> I suggest you compare to the tinderboxes as a first sanity-check.
>>>
>>> It would be a good addition to the tinderboxes with an installworld
>>> test, even if only on one arch, and a universe test on at least
>>> i386 and amd64.
>> as I said, my make universe test dies pretty quiclky at libc
>> building shared library libc.so.7
>> /usr/obj/amd64/usr/src/tmp/usr/lib/libgcc.a: could not read symbols: 
>> Archive has
>>  no index; run ranlib to add one
>> *** Error code 1
> 
> Try nuking /usr/obj/* and make sure you have no spurious stuff
> in your src tree.
> 


Trying to make an amd64 build with "make buildworld" on an
amd64 works just fine.
Trying to make an amd64 build with make TARGET=amd64 buildworld on an
i386  fails at the end of libc.

Same source tree checked out. /usr/obj starting from scratch.