This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH] change GLIBC PPC64/ELF2 ABI default to 2.17
- From: "Carlos O'Donell" <carlos at redhat dot com>
- To: Roland McGrath <roland at hack dot frob dot com>
- Cc: Adam Conrad <adconrad at 0c3 dot net>, munroesj at us dot ibm dot com, libc-alpha at sourceware dot org, Adhemerval Zanella <azanella at linux dot vnet dot ibm dot com>, "Joseph S. Myers" <joseph at codesourcery dot com>, Jeff Law <law at redhat dot com>
- Date: Thu, 30 Jan 2014 22:48:01 -0500
- Subject: Re: [PATCH] change GLIBC PPC64/ELF2 ABI default to 2.17
- Authentication-results: sourceware.org; auth=none
- References: <1391008726 dot 16702 dot 105 dot camel at spokane1 dot rchland dot ibm dot com> <52E92E7C dot 1040707 at redhat dot com> <20140129172158 dot GT15976 at 0c3 dot net> <20140129181118 dot 9F85174441 at topped-with-meat dot com> <52E953C4 dot 40503 at redhat dot com> <20140130212010 dot 8E5D974441 at topped-with-meat dot com>
On 01/30/2014 04:20 PM, Roland McGrath wrote:
>> The problem with that policy is that it isn't what is best for our users,
>> and I'm talking with both FSF and Red Hat hats on.
>
> I don't buy this argument at all. Though I don't doubt your intentions to
> consider it objectively, I think your perception is colored by your
> personal involvement with distro ports based on unreleased glibc code.
>
> But I think most everyone who has expressed any opinions is biased in a
> similar fashion, so it's no surprise that I'm alone in taking the pure
> position that is concerned only with the long-term effects on all the users
> rather than some myopic concern with the surely tiny number of early users
> of hardware that does not yet exist.
What long-term benefits does setting the ABI baseline to 2.19 have for our users?
I would like to understand your position in detail.
I could see an argument to lower the ABI baseline to the oldest version possible.
That would allow users to pick any glibc version to build and install on the new
hardware (with appropriate backports). The downside being that it increases the
long term maintenance burden.
The obvious benefit from choosing an ABI baseline of 2.19 is that old compat
symbols are removed. This simplifies binaries and maintenance for the port
going forward. The problem I have with this somewhat egotistical position is
that it's developer centric not user centric.
Am I missing something?
>> The bickering is reduced to: Get it released upstream first.
>
> This I agree with wholeheartedly. Red Hat, IBM, and whoever else gave
> users a distribution based on unreleased glibc code without also giving
> them the clear understanding that no binaries built from such a
> distribution would be usable in the long run did a disservice to their
> users. It's a real shame if the GNU Project does an (albeit milder)
> disservice to all the future users just to mitigate the harm already done
> by these for-profit actors who knowingly flouted established glibc policy.
Agreed.
Cheers,
Carlos.