WITH_META_MODE: any effect? Tree built twice!

[O. Hartmann]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

I just updated sources /usr/src to r341879 (13-CURRENT). To have a clean, new build,
performing of "make cleanwolrd cleandir" has been issued and I have set WITH_META_MODE=
in /etc/src-env.conf. Loading filemon.ko and rebuilding world and kernel takes on my
IvyBridge 2-core box aeons. My other box, a XEON  E3-1245 V2 @ 3.40GH based box, once
took ~ 40 minutes to buildword and buildkernel from a clean buildtree (with LLVM/CLANG
extras set to be build!), and that system now is also beyond 90 minutes, starting
earlier this year. You see, there is a reason to cut down build times.

Well, after rebuilding world/kernel and installing those, mergemaster and make
delete-old|-libs afterwards, I started again a build (filemon loaded!). And, surprise,
surprise, compilation of all the long-haul taking LLVM/CLANG stuff starts again! That is
not funny.

Why do I have to rebuild world twice to get WITH_META_MODE in effect?

My in-effect settings in /etc/src.conf are as follows:

#
CPUTYPE?=                       native
#
CFLAGS+=                        -O3
# for the kernel
COPTFLAGS+=                     -O3
#
#CXXFLAGS+=                     -std=c++11
#
WITH_CLANG_EXTRAS=              YES
WITH_LLDB=                              YES
WITH_LLD_IS_LD=                 YES
# eBPF
WITH_LLVM_TARGET_BPF=   YES
#
WITH_IDEA=                              YES
#
#WITH_BSD_GREP=                 YES
#
WITH_OFED_EXTRA=                YES
WITH_NAND=                              YES
#WITH_CTF=                              YES
#
WITH_SVN=                               YES
#
# Enable building openldap support for kerberos.
#WITH_OPENLDAP=                 YES
#
WITH_SORT_THREADS=              YES
#
WITH_EXTRA_TCP_STACKS=  YES
#
WITH_ZONEINFO_LEAPSECONDS_SUPPORT=      YES
#
MALLOC_PRODUCTION=              YES
#
WITHOUT_ASSERT_DEBUG=   YES
#
WITHOUT_TESTS=          YES
#
WITHOUT_DEBUG_FILES=    YES
#
WITHOUT_REPRODUCIBLE_BUILD=     YES
#
#  mitigation for CVE-2017-5715 in the kernel build
#WITH_KERNEL_RETPOLINE= YES


These setting dind't change over the past time, except some WITHOUT_ tags.

Are there any unrevealed secrets?

The boxes in question do have 8 GB RAM (two core/4 threads box, base system residing on
UFS/FFS 256GB Samsung 850 Pro SSD, two ZFS volumes for /home and /usr/ports aboard), and
16 GB RAM (4 cores/8thread XEON box, base system on UFS/FFS 256 GB Samsung 850 Pro SSD,
ZFS RAIDZ volume of 16 TB as data graveyard). 

Thanks in advance.

oh


- -- 
O. Hartmann

Ich widerspreche der Nutzung oder Übermittlung meiner Daten für
Werbezwecke oder für die Markt- oder Meinungsforschung (§ 28 Abs. 4 BDSG).
-----BEGIN PGP SIGNATURE-----

iLUEARMKAB0WIQQZVZMzAtwC2T/86TrS528fyFhYlAUCXBEDpQAKCRDS528fyFhY
lMrDAgCnnh1oIWQuNIEb8Es59SGxW5OwQe1KxnRdwOc1lsXJOuBj4lh1pykHavWG
S7iY0/QrXLFNhDsNPhiz0CRqO2MyAf47nIQ4olxb7qmUx3MtatUk8VV37USqwVTA
gm4pYSgySmH0qKb8BN65VJg19VDkbUEKi1zIwZGMlKQSVq0a3+hO
=XMEQ
-----END PGP SIGNATURE-----