what is the long-term plan for YP and eglibc?
at the YP resources page for eglibc:
https://www.yoctoproject.org/tools-resources/projects/eglibc
one reads:
"The Yocto Project encourages interested parties to get involved with the EGLIBC project. More information is at http://eglibc.org."
but following that link, one learns:
"EGLIBC is no longer developed and such goals are now being addressed directly in GLIBC."
rday
Signature · 7 lines
========================================================================
Robert P. J. Day Ottawa, Ontario, CANADA
http://crashcourse.ca
Twitter: http://twitter.com/rpjday
LinkedIn: http://ca.linkedin.com/in/rpjday
========================================================================Hi Robert,
Quoted 8 lines · On Friday 01 August 2014 05:38:07 Robert P. J. Day wrote:
at the YP resources page for eglibc:
https://www.yoctoproject.org/tools-resources/projects/eglibc
one reads:
"The Yocto Project encourages interested parties to get involved with the EGLIBC project. More information is at http://eglibc.org."
That might now be out of date in view of upstream developments.
Quoted 4 lines
but following that link, one learns:
"EGLIBC is no longer developed and such goals are now being addressed directly in GLIBC."
The long-term (even medium-term) plan is to switch back to glibc. We do still rely on some of the functionality in eglibc that hasn't been incorporated back into glibc though, e.g. poky-tiny uses eglibc's fine-grained component selection, so it's not necessarily a straightforward switch. Mostly I think we're waiting on some of Khem's (and probably other folks) time to sort this out.
Cheers,
Paul
Signature · 2 lines
Paul Eggleton Intel Open Source Technology Centre
Quoted 19 lines · Paul Eggleton
Hi Robert,
Quoted 8 lines · On Friday 01 August 2014 05:38:07 Robert P. J. Day wrote:
at the YP resources page for eglibc:
https://www.yoctoproject.org/tools-resources/projects/eglibc
one reads:
"The Yocto Project encourages interested parties to get involved with the EGLIBC project. More information is at http://eglibc.org."
That might now be out of date in view of upstream developments.
Quoted 4 lines
but following that link, one learns:
"EGLIBC is no longer developed and such goals are now being addressed directly in GLIBC."
Quoted 6 lines
The long-term (even medium-term) plan is to switch back to glibc. We do still rely on some of the functionality in eglibc that hasn't been incorporated back into glibc though, e.g. poky-tiny uses eglibc's fine-grained component selection, so it's not necessarily a straightforward switch. Mostly I think we're waiting on some of Khem's (and probably other folks) time to sort this out.
The eglibc has had its last release in 2.19 and btw. the release branches in eglibc will get backports and fixes
as long as corresponding glibc branches are open. As of now eglibc trunk has ceased to exist. I am waiting on 2.20 glibc release (which should be happening anytime now)
there is a list of changed in eglibc that are not in glibc the reason being no one carried them forward and partly I am also to blame due to time constraints but I will make up for it :)
- ppc/e500 has few changes around SPE which our FSL/ppc bsp might care
- Option group support - As Paul mentioned we use it and I plan to forward port it and possibly submit it for 2.21
- cross-localedef - We care about it. We have to see whats missing in 2.20 and forward port the missing parts.
- cross testing - There are minor adoptions needed but we dont have ptests for eglibc yet so this should be minor
Rest of items which are left out doesn't bother us.
We are in August. I am hoping that we can include 2.20 in 1.7 release but there might be a risk associated due to above
and that we will change recipes to be called glibc instead of eglibc some metadata changes have to be made as well. I hope to work on it starting second week of august.
Hope that helps
-Khem