[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [gpsd-dev] scons quirks
From: |
Mike Frysinger |
Subject: |
Re: [gpsd-dev] scons quirks |
Date: |
Thu, 21 Nov 2013 13:18:03 -0500 |
User-agent: |
KMail/1.13.7 (Linux/3.12.0; KDE/4.6.5; x86_64; ; ) |
On Monday 18 November 2013 12:50:43 Hal Murray wrote:
> I think the problem I was hitting is that DESTDIR gets wired into the
> executables. Things don't get relinked if I run scons install without a
> DESTDIR or specify a DESTDIR on the "scons" step that does the initial
> build and then scons install with that same DESTDIR.
this is because the SConstruct is wrong. this section:
# Honor the specified installation prefix in link paths.
if not env["sysroot"]:
env.Prepend(LIBPATH=[installdir('libdir')])
else:
env.Prepend(LIBPATH=[env["sysroot"] + installdir('libdir')])
first, the paths where libs are going to be installed should not be used to
automatically construct -L paths. it breaks cross-compile like setting libdir
to /usr/lib and then packaging it up for deployment on another system, and
it's pointless in the non-cross-compile case -- the toolchain already knows
how to finds its libs.
second, it's using installdir which means it prepends DESTDIR which will be
different between compile & install time. and again, is pointless as we don't
want to relink things between the build and install times as it makes no sense
at all.
i imagine this will fix it (i use a similar patch in Gentoo, but i gut more
code to completely disable chrpath):
--- a/SConstruct
+++ b/SConstruct
@@ -285,10 +285,8 @@ def installdir(dir, add_destdir=True):
return wrapped
# Honor the specified installation prefix in link paths.
-if not env["sysroot"]:
- env.Prepend(LIBPATH=[installdir('libdir')])
-else:
- env.Prepend(LIBPATH=[env["sysroot"] + installdir('libdir')])
+if env["sysroot"]:
+ env.Prepend(LIBPATH=[env["sysroot"] + installdir('libdir',
add_destdir=False)])
if env["shared"]:
env.Prepend(RPATH=[installdir('libdir')])
-mike
signature.asc
Description: This is a digitally signed message part.
- Re: [gpsd-dev] scons quirks, (continued)
- Re: [gpsd-dev] scons quirks, Hal Murray, 2013/11/15
- Re: [gpsd-dev] scons quirks, Hal Murray, 2013/11/15
- Re: [gpsd-dev] scons quirks, Eric S. Raymond, 2013/11/15
- Re: [gpsd-dev] scons quirks, Hal Murray, 2013/11/15
- Re: [gpsd-dev] scons quirks, Eric S. Raymond, 2013/11/18
- Re: [gpsd-dev] scons quirks, Hal Murray, 2013/11/18
- Re: [gpsd-dev] scons quirks, Eric S. Raymond, 2013/11/18
- Re: [gpsd-dev] scons quirks, Hal Murray, 2013/11/18
- Re: [gpsd-dev] scons quirks, Eric S. Raymond, 2013/11/19
- Re: [gpsd-dev] scons quirks,
Mike Frysinger <=
- Re: [gpsd-dev] scons quirks, Eric S. Raymond, 2013/11/22
- Re: [gpsd-dev] scons quirks, Mike Frysinger, 2013/11/27
- Re: [gpsd-dev] scons quirks, Eric S. Raymond, 2013/11/27
- Re: [gpsd-dev] scons quirks, Mike Frysinger, 2013/11/27
- Re: [gpsd-dev] scons quirks, Eric S. Raymond, 2013/11/30
- Re: [gpsd-dev] scons quirks, Miroslav Lichvar, 2013/11/27
- Re: [gpsd-dev] scons quirks, Hal Murray, 2013/11/27
- Re: [gpsd-dev] scons quirks, Gary E. Miller, 2013/11/27
- Re: [gpsd-dev] scons quirks, Hal Murray, 2013/11/27
- Re: [gpsd-dev] scons quirks, Gary E. Miller, 2013/11/27