Bug 583030

Summary:	app-misc/strigi-0.7.8-r1: building w/gcc-6: CMake Error at libstreams/cmake/MacroEnsureVersion.cmake:27 (MATH): math cannot parse the expression: "x86_64610000 + x86_641100 + x86_640":
Product:	Gentoo Linux	Reporter:	Peter Levine <plevine457>
Component:	[OLD] GCC Porting	Assignee:	Gentoo KDE team <kde>
Status:	RESOLVED FIXED
Severity:	normal	CC:	plevine457
Priority:	Normal	Keywords:	UPSTREAM
Version:	unspecified
Hardware:	All
OS:	Linux
URL:	https://sourceforge.net/p/strigi/patches/4/
Whiteboard:
Package list:		Runtime testing required:	---
Bug Depends on:
Bug Blocks:	582084
Attachments:	strigi-0.7.8-gcc6.patch

Description Peter Levine 2016-05-14 16:16:18 UTC

Building app-misc/strigi-0.7.8-r1 with gcc-6 fails in configure stage with:

> CMake Error at libstreams/cmake/MacroEnsureVersion.cmake:27 (MATH):
>   math cannot parse the expression: "x86_646*10000 + x86_641*100 + x86_640":
>   syntax error, unexpected exp_NUMBER, expecting $end (8)
> Call Stack (most recent call first):
>   libstreams/cmake/MacroCheckGccVisibility.cmake:24 (macro_ensure_version)
>   libstreams/CMakeLists.txt:47 (macro_check_gcc_visibility)

> CMake Error at libstreamanalyzer/cmake/MacroEnsureVersion.cmake:27 (MATH):
>   math cannot parse the expression: "x86_646*10000 + x86_641*100 + x86_640":
>   syntax error, unexpected exp_NUMBER, expecting $end (8)
> Call Stack (most recent call first):
>   libstreamanalyzer/cmake/MacroCheckGccVisibility.cmake:24 (macro_ensure_version)
>   libstreamanalyzer/CMakeLists.txt:70 (macro_check_gcc_visibility)

Comment 1 Peter Levine 2016-05-14 16:19:18 UTC

Created attachment 434280 [details, diff]
strigi-0.7.8-gcc6.patch

Pending approval upstream: https://sourceforge.net/p/strigi/patches/4/

Comment 2 Andreas Sturmlechner gentoo-dev

2016-05-14 20:34:55 UTC

I see that fix was applied, but this is bound to break with GCC 7 again. Why not fix it once and for all?

Comment 3 Peter Levine 2016-05-15 20:59:48 UTC

(In reply to Andreas Sturmlechner from comment #2)
> I see that fix was applied, but this is bound to break with GCC 7 again. Why
> not fix it once and for all?

I agree it's not ideal but I don't maintain the cmake module.  IIRC, it has other problems waiting to happen (a bump in the clang version for instance).  I just wanted to demonstrate the simplest fix and allow upstream to decide if/how it will restructure it.

Comment 4 Johannes Huber (RETIRED) gentoo-dev

2016-05-19 13:34:46 UTC

Thanks. 

https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=19fe7ff6b906d04e9641f6229a5166d7b5c006dc

(In reply to Andreas Sturmlechner from comment #2)
> I see that fix was applied, but this is bound to break with GCC 7 again. Why
> not fix it once and for all?

This is correct but i think at this time strigi is already removed ;).