This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: PATCH: Move sysdeps/x86_64/Implies to sysdeps/x86_64/64
- From: Roland McGrath <roland at hack dot frob dot com>
- To: "H.J. Lu" <hjl dot tools at gmail dot com>
- Cc: GNU C Library <libc-alpha at sourceware dot org>
- Date: Fri, 6 Apr 2012 15:47:18 -0700 (PDT)
- Subject: Re: PATCH: Move sysdeps/x86_64/Implies to sysdeps/x86_64/64
- References: <20120322175315.GA9938@intel.com><20120322181118.2F9502C07E@topped-with-meat.com><CAMe9rOpyzLVrfpbZV_OkcZhFE5HH1v+hBxeubO7S_R+Kn8EBUA@mail.gmail.com><20120322195022.F2EF32C07C@topped-with-meat.com><CAMe9rOriXXrVWVLLuES8OUAXvbcVgcJEAYAgJjHWr3YCB2JmaQ@mail.gmail.com><CAMe9rOru6EXRsoqaBKS0a9fGqwey+52DCak0MwGPM1bzWf29NQ@mail.gmail.com><20120322204013.868A12C086@topped-with-meat.com><CAMe9rOq87mKXdoMOGH=cxiq8C6qhK7bTi5bAB9jQk_Eg7x72Vw@mail.gmail.com><20120322210647.330402C07E@topped-with-meat.com><CAMe9rOpiAWE8U2XFxgavJBi9FUomTZh3XEwcdYN59qNBR8OM7g@mail.gmail.com><20120322211655.C5E992C07E@topped-with-meat.com><CAMe9rOp7AVVWn_Bo6sKC7yRuCaRxiqUXGpLCm9wqvg2aSjHx-Q@mail.gmail.com><20120322220155.1AC112C07E@topped-with-meat.com><CAMe9rOqj3J=fw984tRMG-FTgxgxbKbjRZJzrnoaOfqzAi815+g@mail.gmail.com><20120322222523.5F52A2C08D@topped-with-meat.com><CAMe9rOqqKB0g_0xoueMWX-902eOeFqpw3LmO7Ej6p=xERjWtCQ@mail.gmail.com><20120406210447.642DE2C0C4@topped-with-meat.com><CAMe9rOo1gPRVRqhTwYjwdhKWyTgn2-aJMsLnHw7ZYZjPDnGiJw@mail.gmail.com>
I'm too wiped today to think about this very clearly. We'll have to pick
it up next week. I'm starting to tend toward Joseph's line of thinking
that we really need to step back and examine the whole situation and
perhaps do a more thorough revamp. Just adding more arcane options that
interact in ways nobody can really keep track of doesn't feel right at all.
Perhaps to start with, you could show--in detail, and with plenty of
explanation--for x86_64/64 and x86_64/x32, what the order you get now is,
what the order you want is, and any proposals you have for making that
happen and exactly how those would work. (We don't need to see any more
patches for configure.in or suchlike--that would just confuse things
further IMHO. What we need is to work toward actually understanding how
all the pieces would interact in a better arrangement.)
One idea I've had that might simplify things is to drop the implicit use of
parent directories altogether, so that it's just the directories found by
use of the config tuple and the explicit Implies files. But I'm very far
from having thought through how that would go.
I really think we all need to put our heads together to untangle the mess
and figure out what kind of system we actually want to have. It certainly
feels to me like we've reached the tipping point where the accretion of
tweaks has become unmanageable. But perhaps it will seem more tractable to
me next week. I'm simultaneously pretty loath to do anything drastic when
incrementalism has got us this far. Sorry I can't be more help right now.
Thanks,
Roland