glibc: Do not use thumb1 ISA on armv6

Submitted by Khem Raj on Oct. 12, 2018, 8:28 p.m. | Patch ID: 155638

Details

Message ID 20181012202823.4003-1-raj.khem@gmail.com
State New
Headers show

Commit Message

Khem Raj Oct. 12, 2018, 8:28 p.m.
This does not work and is unsupported so lets compile glibc in ARM mode
always on armv6 SOCs

Fixes [YOCTO #12929]

Signed-off-by: Khem Raj <raj.khem@gmail.com>
---
 meta/recipes-core/glibc/glibc.inc | 1 +
 1 file changed, 1 insertion(+)

Patch hide | download patch | download mbox

diff --git a/meta/recipes-core/glibc/glibc.inc b/meta/recipes-core/glibc/glibc.inc
index ddf0c6b80e..99c92ce46e 100644
--- a/meta/recipes-core/glibc/glibc.inc
+++ b/meta/recipes-core/glibc/glibc.inc
@@ -56,6 +56,7 @@  INHIBIT_DEFAULT_DEPS = "1"
 # http://lists.openembedded.org/pipermail/openembedded-core/2015-January/100679.html
 ARM_INSTRUCTION_SET_armv4 = "arm"
 ARM_INSTRUCTION_SET_armv5 = "arm"
+ARM_INSTRUCTION_SET_armv6 = "arm"
 
 # glibc uses PARALLELMFLAGS variable to pass parallel build info so transfer
 # PARALLEL_MAKE into PARALLELMFLAGS and empty out PARALLEL_MAKE

Comments

Andre McCurdy Oct. 12, 2018, 8:56 p.m.
On Fri, Oct 12, 2018 at 1:28 PM, Khem Raj <raj.khem@gmail.com> wrote:
> This does not work and is unsupported so lets compile glibc in ARM mode
> always on armv6 SOCs

Note the comments in glibc-collateral.inc:

  # This needs to match with glibc.inc, otherwise glibc-scripts and glibc-locale
  # will fail to find main glibc, for details see
  # http://lists.openembedded.org/pipermail/openembedded-core/2015-January/100679.html
  ARM_INSTRUCTION_SET_armv4 = "arm"
  ARM_INSTRUCTION_SET_armv5 = "arm"

Either the two recipes need to be kept in sync or the comment no longer applies?

> Fixes [YOCTO #12929]
>
> Signed-off-by: Khem Raj <raj.khem@gmail.com>
> ---
>  meta/recipes-core/glibc/glibc.inc | 1 +
>  1 file changed, 1 insertion(+)
>
> diff --git a/meta/recipes-core/glibc/glibc.inc b/meta/recipes-core/glibc/glibc.inc
> index ddf0c6b80e..99c92ce46e 100644
> --- a/meta/recipes-core/glibc/glibc.inc
> +++ b/meta/recipes-core/glibc/glibc.inc
> @@ -56,6 +56,7 @@ INHIBIT_DEFAULT_DEPS = "1"
>  # http://lists.openembedded.org/pipermail/openembedded-core/2015-January/100679.html
>  ARM_INSTRUCTION_SET_armv4 = "arm"
>  ARM_INSTRUCTION_SET_armv5 = "arm"
> +ARM_INSTRUCTION_SET_armv6 = "arm"
>
>  # glibc uses PARALLELMFLAGS variable to pass parallel build info so transfer
>  # PARALLEL_MAKE into PARALLELMFLAGS and empty out PARALLEL_MAKE
> --
> 2.19.1
>
> --
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-core
Khem Raj Oct. 12, 2018, 9:13 p.m.
On Fri, Oct 12, 2018 at 1:56 PM Andre McCurdy <armccurdy@gmail.com> wrote:
>
> On Fri, Oct 12, 2018 at 1:28 PM, Khem Raj <raj.khem@gmail.com> wrote:
> > This does not work and is unsupported so lets compile glibc in ARM mode
> > always on armv6 SOCs
>
> Note the comments in glibc-collateral.inc:
>
>   # This needs to match with glibc.inc, otherwise glibc-scripts and glibc-locale
>   # will fail to find main glibc, for details see
>   # http://lists.openembedded.org/pipermail/openembedded-core/2015-January/100679.html
>   ARM_INSTRUCTION_SET_armv4 = "arm"
>   ARM_INSTRUCTION_SET_armv5 = "arm"
>
> Either the two recipes need to be kept in sync or the comment no longer applies?

I think we can delete comment in a followup

>
> > Fixes [YOCTO #12929]
> >
> > Signed-off-by: Khem Raj <raj.khem@gmail.com>
> > ---
> >  meta/recipes-core/glibc/glibc.inc | 1 +
> >  1 file changed, 1 insertion(+)
> >
> > diff --git a/meta/recipes-core/glibc/glibc.inc b/meta/recipes-core/glibc/glibc.inc
> > index ddf0c6b80e..99c92ce46e 100644
> > --- a/meta/recipes-core/glibc/glibc.inc
> > +++ b/meta/recipes-core/glibc/glibc.inc
> > @@ -56,6 +56,7 @@ INHIBIT_DEFAULT_DEPS = "1"
> >  # http://lists.openembedded.org/pipermail/openembedded-core/2015-January/100679.html
> >  ARM_INSTRUCTION_SET_armv4 = "arm"
> >  ARM_INSTRUCTION_SET_armv5 = "arm"
> > +ARM_INSTRUCTION_SET_armv6 = "arm"
> >
> >  # glibc uses PARALLELMFLAGS variable to pass parallel build info so transfer
> >  # PARALLEL_MAKE into PARALLELMFLAGS and empty out PARALLEL_MAKE
> > --
> > 2.19.1
> >
> > --
> > _______________________________________________
> > Openembedded-core mailing list
> > Openembedded-core@lists.openembedded.org
> > http://lists.openembedded.org/mailman/listinfo/openembedded-core
Richard Purdie Oct. 12, 2018, 9:42 p.m.
On Fri, 2018-10-12 at 14:13 -0700, Khem Raj wrote:
> On Fri, Oct 12, 2018 at 1:56 PM Andre McCurdy <armccurdy@gmail.com>
> wrote:
> > 
> > On Fri, Oct 12, 2018 at 1:28 PM, Khem Raj <raj.khem@gmail.com>
> > wrote:
> > > This does not work and is unsupported so lets compile glibc in
> > > ARM mode
> > > always on armv6 SOCs
> > 
> > Note the comments in glibc-collateral.inc:
> > 
> >   # This needs to match with glibc.inc, otherwise glibc-scripts and
> > glibc-locale
> >   # will fail to find main glibc, for details see
> >   # 
> > http://lists.openembedded.org/pipermail/openembedded-core/2015-January/100679.html
> >   ARM_INSTRUCTION_SET_armv4 = "arm"
> >   ARM_INSTRUCTION_SET_armv5 = "arm"
> > 
> > Either the two recipes need to be kept in sync or the comment no
> > longer applies?
> 
> I think we can delete comment in a followup

Looking at the link I'd have said that issue still applies?

Cheers,

Richard
Khem Raj Oct. 12, 2018, 9:47 p.m.
On Fri, Oct 12, 2018 at 2:42 PM Richard Purdie
<richard.purdie@linuxfoundation.org> wrote:
>
> On Fri, 2018-10-12 at 14:13 -0700, Khem Raj wrote:
> > On Fri, Oct 12, 2018 at 1:56 PM Andre McCurdy <armccurdy@gmail.com>
> > wrote:
> > >
> > > On Fri, Oct 12, 2018 at 1:28 PM, Khem Raj <raj.khem@gmail.com>
> > > wrote:
> > > > This does not work and is unsupported so lets compile glibc in
> > > > ARM mode
> > > > always on armv6 SOCs
> > >
> > > Note the comments in glibc-collateral.inc:
> > >
> > >   # This needs to match with glibc.inc, otherwise glibc-scripts and
> > > glibc-locale
> > >   # will fail to find main glibc, for details see
> > >   #
> > > http://lists.openembedded.org/pipermail/openembedded-core/2015-January/100679.html
> > >   ARM_INSTRUCTION_SET_armv4 = "arm"
> > >   ARM_INSTRUCTION_SET_armv5 = "arm"
> > >
> > > Either the two recipes need to be kept in sync or the comment no
> > > longer applies?
> >
> > I think we can delete comment in a followup
>
> Looking at the link I'd have said that issue still applies?
>

I am not 100% sure, thats why I wanted
to check and come back with followup
if update was needed.

> Cheers,
>
> Richard
>
Andre McCurdy Oct. 12, 2018, 10:10 p.m.
On Fri, Oct 12, 2018 at 2:47 PM, Khem Raj <raj.khem@gmail.com> wrote:
> On Fri, Oct 12, 2018 at 2:42 PM Richard Purdie
> <richard.purdie@linuxfoundation.org> wrote:
>>
>> On Fri, 2018-10-12 at 14:13 -0700, Khem Raj wrote:
>> > On Fri, Oct 12, 2018 at 1:56 PM Andre McCurdy <armccurdy@gmail.com>
>> > wrote:
>> > >
>> > > On Fri, Oct 12, 2018 at 1:28 PM, Khem Raj <raj.khem@gmail.com>
>> > > wrote:
>> > > > This does not work and is unsupported so lets compile glibc in
>> > > > ARM mode
>> > > > always on armv6 SOCs
>> > >
>> > > Note the comments in glibc-collateral.inc:
>> > >
>> > >   # This needs to match with glibc.inc, otherwise glibc-scripts and
>> > > glibc-locale
>> > >   # will fail to find main glibc, for details see
>> > >   #
>> > > http://lists.openembedded.org/pipermail/openembedded-core/2015-January/100679.html
>> > >   ARM_INSTRUCTION_SET_armv4 = "arm"
>> > >   ARM_INSTRUCTION_SET_armv5 = "arm"
>> > >
>> > > Either the two recipes need to be kept in sync or the comment no
>> > > longer applies?
>> >
>> > I think we can delete comment in a followup
>>
>> Looking at the link I'd have said that issue still applies?
>>
>
> I am not 100% sure, thats why I wanted
> to check and come back with followup
> if update was needed.

If you're not sure then it's better to respect the comment and keep
the two recipes in sync.