Fix parallel building zlib

[jdemeyer]

See http://sage.math.washington.edu/home/palmieri/misc/zlib-1.2.5.log

New spkg: http://boxen.math.washington.edu/home/jdemeyer/spkg/zlib-1.2.5.p0.spkg

Reported upstream by personal email to the address given on http://zlib.net/

Upstream: Reported upstream. Little or no feedback.

CC: @jhpalmieri

Component: packages: standard

Author: Jeroen Demeyer

Reviewer: John Palmieri

Merged: sage-4.8.alpha4

Issue created by migration from https://trac.sagemath.org/ticket/12138

[jdemeyer]

Upstream: Not yet reported upstream; Will do shortly.

[jdemeyer]

Author: Jeroen Demeyer

[jdemeyer]

comment:3

John, could you please test both the old and new zlib spkgs using

MAKEFLAGS="j8" ./sage -f zlib...

(or some other value of j)

On Linux, I do not manage to reproduce the issue.

[jdemeyer]

comment:4

Managed to reproduce anyway with the old zlib on sage.math, not with the new one here. So looks good, but needs formal review.

[jdemeyer]

Changed upstream from Not yet reported upstream; Will do shortly. to Reported upstream. Little or no feedback.

[jdemeyer]

Attachment: zlib-1.2.5-p0.diff.gz

Diff for the zlib spkg, for review only

[jhpalmieri]

comment:6

Looks good to me and builds consistently on sage.math and on OS X, under circumstances where the old spkg failed. (For example, on OS X, I got the old spkg to fail on 7 out of 10 builds, while this one built 20 times in row with no failures, and similarly on sage.math, using MAKEFLAGS='j -l2' ./sage -f ....)

[jhpalmieri]

Reviewer: John Palmieri

[jdemeyer]

Merged: sage-4.8.alpha4