[OE-core] [PATCH] glibc: 'yes' within the path sets wrong variables

2015-02-23 Thread Benjamin Esquivel
The pre-processed output of conftest.c contains an include dir and if the path of it contains a 'yes' it will cause some variables to be wrongly set to yes because of the way it greps for it. The fix is to use a more elaborated string instead of just 'yes'. This has to be done at the configure and

Re: [OE-core] [PATCH] glibc: 'yes' within the path sets wrong variables

2015-02-24 Thread Saul Wold
On 02/23/2015 01:54 AM, Benjamin Esquivel wrote: The pre-processed output of conftest.c contains an include dir and if the path of it contains a 'yes' it will cause some variables to be wrongly set to yes because of the way it greps for it. The fix is to use a more elaborated string instead of ju

Re: [OE-core] [PATCH] glibc: 'yes' within the path sets wrong variables

2015-02-24 Thread Benjamin Esquivel
On Tue, 2015-02-24 at 11:40 -0800, Saul Wold wrote: > On 02/23/2015 01:54 AM, Benjamin Esquivel wrote: > > The pre-processed output of conftest.c contains an include dir > > and if the path of it contains a 'yes' it will cause some variables > > to be wrongly set to yes because of the way it greps

Re: [OE-core] [PATCH] glibc: 'yes' within the path sets wrong variables

2015-02-25 Thread Burton, Ross
On 24 February 2015 at 22:17, Benjamin Esquivel < benjamin.esqui...@linux.intel.com> wrote: > > So the configure file does not get regenerated as part of the standard > > auto-reconf? > > > No, if the fix is only applied at the configure.ac file the unwanted > behavior still shows up from the conf

Re: [OE-core] [PATCH] glibc: 'yes' within the path sets wrong variables

2015-02-25 Thread Benjamin Esquivel
On Wed, 2015-02-25 at 12:23 +, Burton, Ross wrote: > > On 24 February 2015 at 22:17, Benjamin Esquivel > wrote: > > So the configure file does not get regenerated as part of > the standard > > auto-reconf? > > > No, if the fix is only applied at the con

Re: [OE-core] [PATCH] glibc: 'yes' within the path sets wrong variables

2015-02-26 Thread Burton, Ross
On 25 February 2015 at 18:14, Benjamin Esquivel < benjamin.esqui...@linux.intel.com> wrote: > Saul, is there a specific part of the glibc's run.do_configure that > should re-conf? > The glibc recipe explicitly avoids autoreconfig: "# override this function to avoid the autoconf/automake/aclocal/

Re: [OE-core] [PATCH] glibc: 'yes' within the path sets wrong variables

2015-02-26 Thread Khem Raj
On Mon, Feb 23, 2015 at 1:54 AM, Benjamin Esquivel wrote: > The pre-processed output of conftest.c contains an include dir > and if the path of it contains a 'yes' it will cause some variables > to be wrongly set to yes because of the way it greps for it. > The fix is to use a more elaborated stri

Re: [OE-core] [PATCH] glibc: 'yes' within the path sets wrong variables

2015-02-26 Thread Khem Raj
On Thu, Feb 26, 2015 at 4:22 AM, Burton, Ross wrote: > > On 25 February 2015 at 18:14, Benjamin Esquivel > wrote: >> >> Saul, is there a specific part of the glibc's run.do_configure that >> should re-conf? > > > The glibc recipe explicitly avoids autoreconfig: > > "# override this function to av

Re: [OE-core] [PATCH] glibc: 'yes' within the path sets wrong variables

2015-03-02 Thread Benjamin Esquivel
On Thu, 2015-02-26 at 08:18 -0800, Khem Raj wrote: > On Mon, Feb 23, 2015 at 1:54 AM, Benjamin Esquivel > wrote: > > The pre-processed output of conftest.c contains an include dir > > and if the path of it contains a 'yes' it will cause some variables > > to be wrongly set to yes because of the wa

Re: [OE-core] [PATCH] glibc: 'yes' within the path sets wrong variables

2015-03-02 Thread Benjamin Esquivel
On Thu, 2015-02-26 at 08:17 -0800, Khem Raj wrote: > On Thu, Feb 26, 2015 at 4:22 AM, Burton, Ross wrote: > > > > On 25 February 2015 at 18:14, Benjamin Esquivel > > wrote: > >> > >> Saul, is there a specific part of the glibc's run.do_configure that > >> should re-conf? > > > > > > The glibc rec