[arch-dev-public] NOTICE glibc 2.14
Hi, stop compiling packages against testing and push them in extra before glibc is moved out. glibc 2.14 is a bitch https://bugs.archlinux.org/task/24640 https://bugs.archlinux.org/task/24632 https://bugs.archlinux.org/task/24657 thanks -- Ionuț
Am 09.06.2011 22:41, schrieb Ionut Biru:
Hi,
stop compiling packages against testing and push them in extra before glibc is moved out.
glibc 2.14 is a bitch
Let's be a more precise here. For the first time since glibc 2.7, there is an ABI change that is not backward compatible. Thus, programs compiled with the new toolchain require a GLIBC_2.14 symbol. This has happened before, but not for several years.
On 10/06/11 06:46, Thomas Bächler wrote:
Am 09.06.2011 22:41, schrieb Ionut Biru:
Hi,
stop compiling packages against testing and push them in extra before glibc is moved out.
glibc 2.14 is a bitch
Let's be a more precise here. For the first time since glibc 2.7, there is an ABI change that is not backward compatible. Thus, programs compiled with the new toolchain require a GLIBC_2.14 symbol.
This has happened before, but not for several years.
This ususally happens with gcc and c++ programs so it is not really a new thing. Anyway, I have found the issue in glibc and will hopefully have some time to push the fixed packages some time tomorrow. Then I will be taking some signoffs and we can move this. Allan
Am 10.06.2011 11:30, schrieb Allan McRae:
This ususally happens with gcc and c++ programs so it is not really a new thing.
It almost never happens with C programs though.
On 10/06/11 19:34, Thomas Bächler wrote:
Am 10.06.2011 11:30, schrieb Allan McRae:
This ususally happens with gcc and c++ programs so it is not really a new thing.
It almost never happens with C programs though.
If you look at the output of "readelf -s /lib/libc.so.6" it happens every glibc release... e.g. https://bbs.archlinux.org/viewtopic.php?id=103820 https://bbs.archlinux.org/viewtopic.php?id=103820 Allan
Am 10.06.2011 11:43, schrieb Allan McRae:
On 10/06/11 19:34, Thomas Bächler wrote:
Am 10.06.2011 11:30, schrieb Allan McRae:
This ususally happens with gcc and c++ programs so it is not really a new thing.
It almost never happens with C programs though.
If you look at the output of "readelf -s /lib/libc.so.6" it happens every glibc release...
e.g. https://bbs.archlinux.org/viewtopic.php?id=103820 https://bbs.archlinux.org/viewtopic.php?id=103820
This is weird, the binaries I checked only required the GLIBC_2.7 and GLIBC_2.14 symbols.
participants (3)
-
Allan McRae
-
Ionut Biru
-
Thomas Bächler