[[!meta copyright="Copyright © 2010, 2013, 2014 Free Software Foundation, Inc."]] [[!meta license="""[[!toggle id="license" text="GFDL 1.2+"]][[!toggleable id="license" text="Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the license is included in the section entitled [[GNU Free Documentation License|/fdl]]."]]"""]] [[!tag open_issue_libpthread open_issue_glibc]] [[!toc]] # IRC, freenode, #hurd, 2010-07-31 <tschwinge> Other question: how difficult is a NPTL port? Futexes and some kernel interfaces for scheduling stuff etc. -- what else? <youpi> actually NPTL doesn't _require_ futexes <youpi> it just requires low-level locks <youpi> Mmm, it seems to be so only in principle <youpi> I can see futex names here and there in the generic code <youpi> looks like Drepper isn't disciplined enough in that area either <tschwinge> (well, why would he...) <youpi> I'm not sure we really want to port NPTL <tschwinge> OK. <youpi> Drepper will keep finding things to add <youpi> while the interface between glibc and libpthread isn't increasing _so_ much <tschwinge> ... and even less so the interfavce that actual applications are using. <tschwinge> We'd need to evaluate which benefits NPTL would bring. # Debian GNU/kFreeBSD: FBTL ## IRC, freenode, #hurd, 2013-08-05 <gnu_srs> Hi, looks like kfreebsd are now using an NPTL-based pthread library: FBTL, http://lists.debian.org/debian-bsd/2013/07/msg00060.html [[!message-id "alpine.LNX.2.00.1307102021050.4232@contest.felk.cvut.cz"]]. <gnu_srs> Anything of interest for porting to Hurd? See also http://lists.debian.org/debian-hurd/2013/08/msg00000.html <azeem> Petr could've been more verbose in his announcements <pinotree> and there's http://www.gnu.org/software/hurd/open_issues/nptl.html in our wiki <azeem> well, it seems to work fine for kFreeBSD: http://lists.debian.org/debian-bsd/2013/07/msg00134.html <azeem> and http://lists.debian.org/debian-bsd/2013/07/msg00138.html ## [[!message-id "alpine.LNX.2.00.1308021035160.5570@contest.felk.cvut.cz"]] ## [[!message-id "alpine.LNX.2.00.1405082034530.8707@contest.felk.cvut.cz"]] ## [[!message-id "87wqdv1314.fsf@kepler.schwinge.homeip.net"]] # IRC, freenode, #hurd, 2013-12-26 <nalaginrut> hm? has NPTL already supported for Hurd? <braunr> probably won't ever be <nalaginrut> so no plan for it? <braunr> what for ? <nalaginrut> no one interested in it, or no necessary adding it? <braunr> why would you want nptl ? <braunr> ntpl was created to overcome the defficiencies of linuxthreads <braunr> we have our own libpthread <braunr> (with its own defficiencies) <braunr> supporting nptl would probably force us to implement something a la clone <nalaginrut> well, just inertia, now that Linux/kFreebsd has it <braunr> are you sure kfreebsd has it ? * teythoon thought we have clone <nalaginrut> http://www.gnu.org/software/hurd/open_issues/nptl.html <nalaginrut> seems someone mentioned it <braunr> it's a "nptl-like implementation" <nalaginrut> yes, I don't think it should be the same with Linux one, but something like it <braunr> but what for ? <braunr> as mentioned in the link you just gave, "<tschwinge> We'd need to evaluate which benefits NPTL would bring." <nalaginrut> well, it's the note of 2010, I don't know if it's relative now <braunr> relevant* <nalaginrut> ah thanks <braunr> but that still doesn't answer anything <braunr> why are *you* talking about nptl ? <nalaginrut> just saw pthread, then recall nptl, dunno <nalaginrut> just asking <braunr> :) <nalaginrut> but you mentioned that Hurd has its own thread implementation, is it similar or better than Linux NPTL? <nalaginrut> or there's no benchmark yet? <braunr> it's inferior in performance <braunr> almost everything in the hurd is inferior performance-wise because of the lack of optimizations <braunr> currently we care more about correctness <nalaginrut> speak the NPTL, I ever argued with a friend since I saw drepper mentioned NPTL should be m:n, then I thought it is...But finally I was failed, he didn't implement it yet... <braunr> what ? <braunr> nptl was always 1:1 <nalaginrut> but in nptl-design draft, I thought it's m:n <nalaginrut> anyway, it's draft <nalaginrut> and seems being a draft for long time <braunr> never read anything like that <nalaginrut> I think it's my misread <nalaginrut> I have to go, see you guys tomorrow <braunr> The consensus among the kernel developers was that an M-on-N implementation <braunr> would not fit into the Linux kernel concept. The necessary infrastructure which would <braunr> have to be added comes with a cost which is too high. --- # Resources * <http://www.akkadia.org/drepper/nptl-design.pdf> * <http://nptltracetool.sourceforge.net/> * <http://posixtest.sourceforge.net/>