Date: Sat, 4 Apr 2026 07:50:00 -0700 From: bob prohaska <fbsd@www.zefox.net> To: Mark Millard <marklmi@yahoo.com> Cc: freebsd-current@freebsd.org Subject: Re: Etcupdate runtime has more than doubled....Why? Message-ID: <adElGE1DtwBG9QXp@www.zefox.net> In-Reply-To: <3580ce8f-3500-4386-a35a-f1b49dbc7d44@yahoo.com> References: <ac1SLQ4rS9tScr5-@www.zefox.net> <626584dc-ea9c-4a63-902c-4de4e10bba00@yahoo.com> <ac1lP2IqDJgiqHx4@www.zefox.net> <c82fcae8-94e6-4a97-9d44-2ab14e9b1351@yahoo.com> <3580ce8f-3500-4386-a35a-f1b49dbc7d44@yahoo.com>
index | next in thread | previous in thread | raw e-mail
On Wed, Apr 01, 2026 at 12:20:11PM -0700, Mark Millard wrote: > On 4/1/26 12:10, Mark Millard wrote: > > On 4/1/26 11:34, bob prohaska wrote: > >> On Wed, Apr 01, 2026 at 11:00:56AM -0700, Mark Millard wrote: > >>>> > >>> > >>> etcupdate has the option: > >>> > >>> > >>> -B Do not build generated files in a private object tree. > >>> Instead, reuse the generated files from a previously built > >>> object tree that matches the source tree. This can be > >>> useful to avoid gratuitous conflicts in sendmail(8) > >>> configuration files when bootstrapping. It can also be > >>> useful for building a tarball that matches a specific > >>> world build. > >>> I misunderstood the gist of the description, thinking -B was intended only for special cases of multiple source and object trees. In fact -B seems to solve my problem entirely, etcupdate finishes in minutes now, not hours. Thanks to all who responded! bob prohaska > >>> > >>> So, if one already has the required context, its use might be relevant. > >>> Of course, without the proper context, it should not be used. > >>> > >>> What was the context-status and use vs. not for -B ? > >> > >> There's only one /usr/obj on the system. > >> > >> The buildworld command used was > >> make -j3 KERNCONF=GENERIC TARGET=arm TARGET_ARCH=armv7 buildworld > >> and the etcupdate command following buildworld was > >> etcupdate -p > >> > >> After installing world and kernel, etcupdate runs again as > >> etcupdate > >> > >> Might using etcupdate -B in the second invocation be a remedy? > > > > Extracted/simplified from my more overall scripts full of local details: > > > > # make WITH_META_MODE=yes buildworld > > # make WITH_META_MODE=yes buildkernel > > # etcupdate -p > > # etcupdate resolve -p > > # make WITH_META_MODE=yes installkernel > > Note for live updating the booted file system: > > # # Possibly: vi /boot/loader.conf > # shutdown -r now > . . . > > (I have contexts that do not involve that.) > > > > # make WITH_META_MODE=yes installworld > > # etcupdate -B > > # etcupdate resolve > > # make delete-old check-old > > # #AS NEEDED for timing?: make delete-old-libs > > # etcupdate status > > Note for live updates the booted file system or just wanting to: > > # # Possibly: vi /boot/loader.conf > # shutdown -r now > . . . > > (I have contexts that do not involve that. For those I may deal with > booting at some later time for some other reason.) > > > > >> > >> Thanks for writing! > >> > >> bob prohaska > >> > >> > >> > > > > > > > -- > === > Mark Millard > marklmi at yahoo.comhome | help
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?adElGE1DtwBG9QXp>