depend + all vs dependall

[Ruslan Ermilov]

On Mon, Mar 31, 2003 at 05:04:42PM +1000, Bruce Evans wrote:
> On Mon, 31 Mar 2003, Ruslan Ermilov wrote:
> 
> > On Sat, Mar 29, 2003 at 04:33:43PM -0700, M. Warner Losh wrote:
> > ...
> > > @@ -67,7 +67,7 @@
> > >
> > >
> > >  .for __target in all all-man checkdpadd clean cleandepend cleandir \
> > > -    depend distribute lint maninstall \
> > > +    depend dependall distribute lint maninstall \
> > >      obj objlink realinstall regress tags
> > >  ${__target}: _SUBDIR
> > >  .endfor
> >
> > If you want to really try dependall, you should implement it
> > in bsd.subdir.mk like the distribute target, similar to this:
> >
> > .if !target(dependall)
> > dependall:
> >         cd ${.CURDIR}; \
> >             ${MAKE} depend -DNO_SUBDIR; \
> >             ${MAKE} all -DNO_SUBDIR
> > .endif
> 
> That won't work, since it gives a double tree traversal for the
> subdirs and thus defeats the point of dependall.  Any splitting
> up of dependall must not be passed down.
> 
> > Also, your test is not honest because original Makefile.inc1
> > did not parallelize the "all" stage of "buildworld", by not
> > implementing par-all.  Last time I tried par-all, it saved
> > me 16% of time from the -j8 buildworld:
> >
> > 26m8.74s real           26m13.08s user          11m9.70s sys (old)
> > 21m48.52s real          26m20.60s user          11m4.95s sys (new)
> >
> > Attached is the message with the patch.  It has some Russian,
> > but also includes a patch.  Note that par-all only parallelizes
> > top-level bsd.subdir.mk makefiles, as we depend on the ordering
> > of traversing SUBDIRs in a few places.  The plan is to drop
> > this assumption in places that don't need this ordering.
> 
> Please benchmark mainly for the usual case of non-SMP :-).
> 
You mean for the non-SMP case but with -j?

On a Celeron 800 system with ATA100 disk using -j4 -DNOCLEAN buildworld
of RELENG_4 a friend reported the following times:

Without patch		With patch
real    69m43.271s	69m26.722s
user    38m22.009s	38m19.384s
sys     10m45.273s	10m41.596s

Further reports show that on single-CPU systems with large CPU
cache the real time win was near what I have reported for 2-CPU
box, and it had no effect on small cache single-CPU systems and
-j builds.


Cheers,
-- 
Ruslan Ermilov		Sysadmin and DBA,
ru at sunbay.com		Sunbay Software AG,
ru at FreeBSD.org		FreeBSD committer,
+380.652.512.251	Simferopol, Ukraine

http://www.FreeBSD.org	The Power To Serve
http://www.oracle.com	Enabling The Information Age
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/freebsd-arch/attachments/20030331/1066db9d/attachment.bin