Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 28 Dec 2005 03:04:11 +0200
From:      Giorgos Keramidas <keramida@ceid.upatras.gr>
To:        Doug Hardie <bc979@lafn.org>
Cc:        freebsd-questions <freebsd-questions@freebsd.org>
Subject:   Re: make buildworld
Message-ID:  <20051228010411.GB8568@flame.pc>
In-Reply-To: <03683319-0CC1-4367-BCBC-29D0D4B97D41@lafn.org>
References:  <03683319-0CC1-4367-BCBC-29D0D4B97D41@lafn.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On 2005-12-26 23:49, Doug Hardie <bc979@lafn.org> wrote:
> I am upgrading a server to 6.0 and encountered an error in make
> buildworld.  However, I don't know what the error was as I piped
> stdout to a file, but not stderr.

I usually keep them both, with something like:

    # cd /usr/src/
    # cvs -q up -APd | logfile.cvs
    # make buildworld buildkernel 2>&1 | tee logfile.build

> It was fairly near the end so I really hate to restart from the
> beginning again.  The master server is a fairly slow machine.

Then, someone could argue that the problem is not a build that's
failing, but the fact that you're using a slow machine as the build
server :-/

I'm afraid there's no way to recover data that has scrolled off the
scrollback buffer of syscons or screen(1), when the same data wasn't
saved in a file.

> When something like this happens, is there a way to restart the make
> where it died?

You can try using -DNO_CLEAN, but this will do a fair bit of work too.

> Is there an easy way to build the specific module that failed to get
> the complete errors?  In this case the module was /usr/libexec/telnet.
> I went to /usr/src/ libexec/telnet and did a make.

Then the error was somewhere else.  Were you using make -j XXX with a
non-default XXX number?




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20051228010411.GB8568>