[m-rev.] Upgrade Boehm GC to 7.6.0

Paul Bone paul at bone.id.au
Wed Sep 21 13:51:16 AEST 2016


On Wed, Sep 21, 2016 at 11:56:48AM +1000, Julien Fischer wrote:
> 
> Hi Paul,
> 
> On Wed, 21 Sep 2016, Paul Bone wrote:
> 
> >On Wed, Sep 21, 2016 at 09:29:14AM +1000, Julien Fischer wrote:
> >>
> >>Hi,
> >>
> >>On Tue, 20 Sep 2016, Zoltan Somogyi wrote:
> >>
> >>>On Tue, 20 Sep 2016 18:02:16 +1000 (AEST), Julien Fischer <jfischer at opturion.com> wrote:
> >>>>>I've tested this with hlc.gc, hlc.par.gc, asm_fast.gc, asm_fast.par.gc.stseg
> >>>>>on x86_64 Linux.  If anyone would like to test on other platforms that'd be
> >>>>>good, but it's also okay to test this once its on master.
> >>>>
> >>>>Go and ahead and commit it to master.  I'll take a look at it on OS X
> >>>>and Windows after you do that.
> >>>
> >>>I am afraid this broke my setup. I use Ubuntu 13.10, and it does
> >>>NOT have libtool installed, which is why I get errors such as these
> >>>(extracted from the output of make):
> >>
> >>Paul, I thought you had removed the dependency on libtool; has it crept
> >>back in somewhere?
> >
> >The best I could do last time was to commit the files that libtool
> >generates, so that it's not needed during normal builds.  I'm currently
> >setting up an environment similar to Zoltan's without libtool so that I can
> >see more of what is happening.  Maybe my change didn't go through when
> >Zoltan did his testing or maybe I screwed up, not sure.
> 
> I just "fixed" the problem another way, by taking ltmain.sh and the
> contents of the m4 directory from the libatomic_ops-7.4.4 tarball (from
> the Boehm GC webapage) and copying them into my source tree.  It then
> works.  (In the git version ltmain.sh is currently a symlink to
> somewhere that doesn't exist on my system.)  I think it might be best if
> we just commit the above files in the Mercury version of libatomic_ops.

Agreed,

I can do that.


-- 
Paul Bone
http://paul.bone.id.au


More information about the reviews mailing list