Re: [OE-core] [PATCH 18/24] procps: update 3.3.16 -> 3.3.17

["Richard Purdie" <richard.purdie@linuxfoundation.org>]

On Sun, 2021-02-28 at 11:10 -0800, Khem Raj wrote:
> 
> On 2/28/21 7:36 AM, Alexander Kanavin wrote:
> > Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
> > ---
> >   .../procps/0001-w.c-correct-musl-builds.patch | 44 +++++++++++++++++++
> >   ...02-proc-escape.c-add-missing-include.patch | 23 ++++++++++
> >   .../{procps_3.3.16.bb => procps_3.3.17.bb}    |  4 +-
> >   3 files changed, 70 insertions(+), 1 deletion(-)
> >   create mode 100644 meta/recipes-extended/procps/procps/0001-w.c-correct-musl-builds.patch
> >   create mode 100644 meta/recipes-extended/procps/procps/0002-proc-escape.c-add-missing-include.patch
> >   rename meta/recipes-extended/procps/{procps_3.3.16.bb => procps_3.3.17.bb} (94%)
> > 
> > diff --git a/meta/recipes-extended/procps/procps/0001-w.c-correct-musl-builds.patch b/meta/recipes-extended/procps/procps/0001-w.c-correct-musl-builds.patch
> > new file mode 100644
> > index 0000000000..c92ad28e4f
> > --- /dev/null
> > +++ b/meta/recipes-extended/procps/procps/0001-w.c-correct-musl-builds.patch
> > @@ -0,0 +1,44 @@
> > +From 22f8d25567b8d64bdbab0fb0b4915b4362561d9b Mon Sep 17 00:00:00 2001
> > +From: Alexander Kanavin <alex.kanavin@gmail.com>
> > +Date: Wed, 24 Feb 2021 21:14:31 +0000
> > +Subject: [PATCH] w.c: correct musl builds
> > +
> > +No need to redefine UT_ stuff to something that does not exist.
> > +
> > +UT_ is already provided in musl but via utmp.h header, so include
> > +it always.
> > +
> > +Upstream-Status: Submitted [https://gitlab.com/procps-ng/procps/-/merge_requests/126]
> > +Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
> > +---
> > + w.c | 9 +--------
> > + 1 file changed, 1 insertion(+), 8 deletions(-)
> > +
> > +diff --git a/w.c b/w.c
> > +index 9d07ac9..d10639b 100644
> > +--- a/w.c
> > ++++ b/w.c
> > +@@ -57,9 +57,8 @@
> > + #include <unistd.h>
> > + #ifdef HAVE_UTMPX_H
> > + #	include <utmpx.h>
> > +-#else
> > +-#	include <utmp.h>
> > + #endif
> > ++#include <utmp.h>
> 
> this would mean when utmpx.h is provided ( e.g. ) it will now include 
> both, which is a change perhaps to look more into, it might want them to 
> be exclusive.

This one is worrying me a bit as it does seem to change behaviour on glibc.
Was anyone able to look into it further?

Cheers,

Richard