This is the mail archive of the
mailing list for the glibc project.
Re: 2.26 freeze in a little over a week
- From: Carlos O'Donell <carlos at redhat dot com>
- To: Joseph Myers <joseph at codesourcery dot com>, Siddhesh Poyarekar <siddhesh at gotplt dot org>
- Cc: Palmer Dabbelt <palmer at dabbelt dot com>, szabolcs dot nagy at arm dot com, nd at arm dot com, Andrew Waterman <andrew at sifive dot com>, Darius Rad <darius at bluespec dot com>, libc-alpha at sourceware dot org
- Date: Mon, 26 Jun 2017 19:34:30 -0400
- Subject: Re: 2.26 freeze in a little over a week
- Authentication-results: sourceware.org; auth=none
- References: <mhng-b49e15b1-af53-4cf6-9246-f8c7c2ab1b14@palmer-si-x1c4> <firstname.lastname@example.org> <alpine.DEB.email@example.com>
On 06/23/2017 03:59 PM, Joseph Myers wrote:
> On Fri, 23 Jun 2017, Siddhesh Poyarekar wrote:
>> On Thursday 22 June 2017 11:18 PM, Palmer Dabbelt wrote:
>>> We're not upstream in Linux yet, we're also in the process of submitting
>>> patches. We've been through two rounds of review of the ABI on the Linux
>>> mailing list and there weren't any comments the second time.
>>> We've committed to maintaining ABI stability with the first one to get released
>>> (which would be glibc if we can get into 2.26).
>> We usually wait for the kernel bits to go in before releasing a glibc
>> with the new ABI, but I'll defer to Carlos, Joseph, etc. on that.
>> Provided that is clear, I think it should be safe to continue reviewing
>> the patchset until about a week or two before the release date
>> (tentatively 1 Aug) and then take a call on whether it is ready or needs
>> to be pushed to 2.27. Given that it does not interact with other
>> architecture code, their testing should not get affected.
> I'm wary of adding new ports without a clear timescale when we can expect
> them to be buildable with build-many-glibcs.py using normal upstream
> sources of other components.
> Has the Linux kernel port been accepted subject to at most minor changes
> relative to the latest version, with good reason to expect it will be in
> Linux 4.13? Is the port in linux-next, or will it be within the next
> month? If the glibc port won't be buildable with Linux 4.13, that's
> leaving a long time with a new port people can't build when making bulk
I agree with Joseph.
We need to see some fairly clear intent from the Linux kernel side before
adopting a new port, and we like the kernel side to land first before
glibc. This ensures the ABI work is done with time.