Re: Fix __ASSUME_RECVMMSG issues (bug 16610)

[Chris Metcalf <cmetcalf at tilera dot com>]

On 2/19/2014 7:42 PM, Joseph S. Myers wrote:
> Architecture-specific note: Tile's kernel-features.h says "TILE glibc
> support starts with 2.6.36", which is accurate in that 2.6.36 was the
> first kernel version with Tile support, and on that basis I've made
> that header define __ASSUME_RECVMMSG_SYSCALL unconditionally.
> However, Tile's configure.ac has arch_minimum_kernel=2.6.32.  Since
> arch_minimum_kernel is meant to reflect only kernel.org kernel
> versions, I think that should change to 2.6.36.  (If using glibc with
> kernel versions from before a port went in kernel.org, it's your
> responsibility to change arch_minimum_kernel in a local patch, and at
> the same time to adjust any __ASSUME_* definitions that may not be
> correct for your older kernel; for developing the official glibc it
> should only ever be necessary to consider what official kernel.org
> releases support.)

The 2.6.32 support is because we ship a suitably-modified version of RedHat 6
(really CentOS 6) to our customers in our Multicore Development Environment,
and wanted to be able to ship a userspace that was compatible with the
kernel version normally used for RHEL 6.

This is similar to the rationale for including support for glibc 2.12 in the
ABI, even though tile support was only added to glibc in version 2.15.

You make a reasonable case for fixing arch_minimum_kernel in terms of
upstream support, but I'm not sure it rises to the level of actually changing
it at this point; my instinct would be to leave it alone for tile, i.e. wrap
the __ASSUME_RECVMMSG_SYSCALL in a conditional.  If this makes sense I'm happy
to expand on the comment about the first supported kernel version to clarify
the use of conditionals.  But I'm also open to modifying the minimum version
if there's a consensus that we should strictly abide by the
upstream-availability rule.

-- 
Chris Metcalf, Tilera Corp.
http://www.tilera.com