| Summary: | dev-libs/boost-1.34.x - fix function library bug in a multi-threaded context (from boost web page) | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Christian Kotz <christian.kotz> |
| Component: | [OLD] Library | Assignee: | Tiziano Müller (RETIRED) <dev-zero> |
| Status: | RESOLVED DUPLICATE | ||
| Severity: | minor | CC: | christian.kotz, cpp+disabled |
| Priority: | High | ||
| Version: | unspecified | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
| Attachments: |
1 file to replace
1 file to replace 2nd file to replace ebuild that replaces the two buggy files ebuild + patchfile for boost-1.34.1-r1 |
||
|
Description
Christian Kotz
2007-12-24 11:23:13 UTC
Created attachment 139242 [details, diff]
1 file to replace
Created attachment 139244 [details]
1 file to replace
Created attachment 139246 [details]
2nd file to replace
Created attachment 139247 [details]
ebuild that replaces the two buggy files
The website that announces the bug rated critical for the boost function library is http://boost.org section latest news dec 19 2007 Unfortunately after patching boost with the 2 files there is a problem with the ebuild and boost. It seems that the interface has changed in 'function_template.hpp' and the change has yet not been propagated to other boost libraries. The build of 'read_graphviz_spirit.o' from the boost graph library fails due to cast away constness (g++ 4.2.2p1 gentoo on my machine). I haven't had time to look at the code yet, but will do so during the x-mas days. Maybe patching should be postponed until a new 1.34.x or 1.35 revision from boost.org is released. @Christian: Which use-flags do you use for building boost? My use="doc -debug -icu -pyste -tools" Created attachment 139392 [details]
ebuild + patchfile for boost-1.34.1-r1
alternate ebuild that uses a patch instead of copying the 2 files
em, obviously you didn't see that I already released boost-1.34.1-r1 a couple of days ago with the two files patched. But if you could come up with a patch which solves the issue with the patch you mentioned, that would be great. At the moment it seems that upstream isn't even aware of that problem (at least, there aren't any messages on the mailinglist). Cheers. *** This bug has been marked as a duplicate of bug 202964 *** |
