[oe] future branch

[Bernhard Reutner-Fischer]

On Feb 8, 2012 8:19 PM, "Khem Raj" <raj.khem at gmail.com> wrote:
>
> On Wed, Feb 8, 2012 at 10:42 AM, Bernhard Reutner-Fischer
> <rep.dot.nop at gmail.com> wrote:
> > Core-image-sato does not build for me due to some buggy package that
> > unconditionally uses *netgrp() or the like.
> > Not amused, way out without going beserk?
>
> Yeah with future branch here is my patch. git am it.
>
>
http://uclibc.org/~kraj/0001-uclibc-git-Update-to-latest-plus-future-branch.patch
>
> apply that on master of oe-core

Fcntl{,64} is a nightmare, yes. But your hunk does not strike me as
immediately helpful.
arm/ioperm hidden def removal is OK, I think, without having sources at
hand.
The socketcall hunk is nonsense, see my previous reply to Peter in this
thread.

+-# provided via pthreads builddir +-CSRC-y := $(filter-out $(libc_a_CSRC)
$(notdir $(libpthread_libc_OBJS:.o=.c)),$(CSRC-y)) +-SSRC := $(filter-out
$(libc_a_SSRC) $(notdir $(libpthread_libc_OBJS:.o=.S)),$(SSRC))
Huh? Sounds inappropriate to me.
Oh, filtering later on makes sense, so OK, good catch!

+-ifeq ($(IS_EABI),y) ++ifeq ($(CONFIG_ARM_EABI),y)
But _that_ is a step backwards, NACK.
>
> then set
>
> UCLIBCVERSION = "0.9.33+gitr%" in local.conf