pkgsrc-Users archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: shells/ast-ksh fails to build on NetBSD



Tests are also not too bad:
---
OK:       175
FAIL:      21
SKIP:       0
TIMEOUT:   10
---

This is on 8.99.24 amd64. Apart from the few removed math functions, no other attempt for porting was done; some of the failures are in maths tests, likely due to the removal of these functions.  

On Sat, 1 Sep 2018 at 22:05 Chavdar Ivanov <ci4ic4%gmail.com@localhost> wrote:
As strange as this may seem, the github version compiles almost from the box. The required build system works straight, without any problems, using python3.6. The 'ninja -C build' link phase fails with missing references to the l(for long)  versions of some 6-7 floating point functions, which I just took away from the math.c module. It appears we have all these functions, but not the l variants.

I guess the best would be to completely rework the package making use of the latest version. I've tried on numerous occasions to compile ast-ksh with the same result (  I tried to figure out what was going on, but the 'iffe' script was beyond my reckoning... ).  

On Sat, 1 Sep 2018 at 19:51 Benny Siegert <bsiegert%gmail.com@localhost> wrote:
As far as I can tell from the bulk results in bulktracker.appspot.com,
the last release on which ast-ksh worked was NetBSD 6. My build on -8
fails in exactly the same way as the build log at
http://ftp.netbsd.org/pub/pkgsrc/packages/reports/2018Q1/NetBSD-8.0-x86_64/20180621.1349//ast-ksh-20120801nb2/build.log.

Any ideas?

I saw that our ksh version is from 2012, while there seems to be a
newer one (from 2017) available on https://github.com/att/ast.

--
Benny


Home | Main Index | Thread Index | Old Index