Re: Buildworld fails in 5.1

From: Garance A Drosihn <drosih_at_rpi.edu>
Date: Mon, 21 Jul 2003 14:02:26 -0400
At 9:46 AM -0700 7/21/03, Tim Kientzle wrote:
>Gordon Tetlow wrote:
>>On Mon, Jul 21, 2003, Tim Kientzle wrote:
>>>Hmmm...  Is that what .ORDER is for?  To work around a
>>>parallel make that gratuitously rebuilds things?
>>
>>Right it serializes build dependencies. The problem with
>>crunchgen ...
>
>I would argue "the problem with make"... ;-)  I think it's
>pretty clear that
>
>a b c: foo
>     buildabc
>
>does not require that 'buildabc' be run three times.  Make
>should be able to note that 'buildabc' was already spawned
>for 'a' and just add 'b' and 'c' to the wait list for
>that operation, rather than running additional copies.

It may not be quite that simple, when it gets to splitting
apart the make rules to build things in parallel.  I tracked
down two problems with parallel builds, and in both cases
make was "doing the right thing", given the rules that it
was working with.  On the other hand, I'm certainly no expert
when it comes to writing makefiles for parallel builds.

In any case, the suggested change did not solve the problem
for me, when doing a make -j5 buildworld.  I still get the
failure with:

make: don't know how to make 
/usr/obj/usr/src/rescue/rescue//usr/src/sbin/dhclie
nt/client/clparse.o. Stop

If there's some other way I should run the buildworld to
generate a more helpful output, let me know.  The failure
seems to be 100% reproducible on my machine, at least if I
start out by removing all of /usr/obj/usr/src.

-- 
Garance Alistair Drosehn            =   gad_at_gilead.netel.rpi.edu
Senior Systems Programmer           or  gad_at_freebsd.org
Rensselaer Polytechnic Institute    or  drosih_at_rpi.edu
Received on Mon Jul 21 2003 - 09:02:31 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:37:16 UTC