[PATCH] meta-selftest: Add qemux86copy machine

[PATCH] meta-selftest: Add qemux86copy machine

[Richard Purdie]

I know Martin finds something like this useful for testing sstate signatures,
this adds one to meta-selftest so we can use it from oe-selftest in the sstate
tests there too.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

diff --git a/meta-selftest/conf/machine/qemux86copy.conf b/meta-selftest/conf/machine/qemux86copy.conf
new file mode 100644
index 0000000..bcce2f5
--- /dev/null
+++ b/meta-selftest/conf/machine/qemux86copy.conf
@@ -0,0 +1,3 @@
+require conf/machine/qemux86.conf
+
+MACHINEOVERRIDES .= ":qemux86"

Re: [PATCH] meta-selftest: Add qemux86copy machine

[Martin Jansa]

[-- Attachment #1: Type: text/plain, Size: 1060 bytes --]

On Wed, Sep 30, 2015 at 02:47:13PM +0100, Richard Purdie wrote:
> I know Martin finds something like this useful for testing sstate signatures,
> this adds one to meta-selftest so we can use it from oe-selftest in the sstate
> tests there too.
> 
> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
> 
> diff --git a/meta-selftest/conf/machine/qemux86copy.conf b/meta-selftest/conf/machine/qemux86copy.conf
> new file mode 100644
> index 0000000..bcce2f5
> --- /dev/null
> +++ b/meta-selftest/conf/machine/qemux86copy.conf
> @@ -0,0 +1,3 @@
> +require conf/machine/qemux86.conf
> +
> +MACHINEOVERRIDES .= ":qemux86"

I wasn't changing MACHINEOVERRIDES (in case this would hide some
possible sstate differences), instead I had to duplicate couple qemux86
overrides in linux-yocto, but this is still useful.

I'll drop my version after testing with sstate-diff-machines.sh that
both return the same results.

Acked-by: Martin Jansa <Martin.Jansa@gmail.com>

-- 
Martin 'JaMa' Jansa     jabber: Martin.Jansa@gmail.com

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 188 bytes --]

Re: [PATCH] meta-selftest: Add qemux86copy machine

[Richard Purdie]

On Wed, 2015-09-30 at 16:28 +0200, Martin Jansa wrote:
> On Wed, Sep 30, 2015 at 02:47:13PM +0100, Richard Purdie wrote:
> > I know Martin finds something like this useful for testing sstate signatures,
> > this adds one to meta-selftest so we can use it from oe-selftest in the sstate
> > tests there too.
> > 
> > Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
> > 
> > diff --git a/meta-selftest/conf/machine/qemux86copy.conf b/meta-selftest/conf/machine/qemux86copy.conf
> > new file mode 100644
> > index 0000000..bcce2f5
> > --- /dev/null
> > +++ b/meta-selftest/conf/machine/qemux86copy.conf
> > @@ -0,0 +1,3 @@
> > +require conf/machine/qemux86.conf
> > +
> > +MACHINEOVERRIDES .= ":qemux86"
> 
> I wasn't changing MACHINEOVERRIDES (in case this would hide some
> possible sstate differences),

I think this actually makes it more likely to highlight differences, but
either way, the tests I ran did highlight a raft of issues and it seems
to be good enough to find several problems. I've sent enough fixes to
make the new selftest pass.

>  instead I had to duplicate couple qemux86
> overrides in linux-yocto, but this is still useful.
> 
> I'll drop my version after testing with sstate-diff-machines.sh that
> both return the same results.

Sounds good.

FWIW you can run the new test with:

oe-selftest --run-tests sstatetests.SStateTests.test_sstate_sametune_samesigs

in case that isn't clear. It doesn't rely on building anything so is
comparatively fast.

Cheers,

Richard

Re: [PATCH] meta-selftest: Add qemux86copy machine

[Martin Jansa]

[-- Attachment #1: Type: text/plain, Size: 1789 bytes --]

On Wed, Sep 30, 2015 at 04:59:47PM +0100, Richard Purdie wrote:
> On Wed, 2015-09-30 at 16:28 +0200, Martin Jansa wrote:
> > On Wed, Sep 30, 2015 at 02:47:13PM +0100, Richard Purdie wrote:
> > > I know Martin finds something like this useful for testing sstate signatures,
> > > this adds one to meta-selftest so we can use it from oe-selftest in the sstate
> > > tests there too.
> > > 
> > > Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
> > > 
> > > diff --git a/meta-selftest/conf/machine/qemux86copy.conf b/meta-selftest/conf/machine/qemux86copy.conf
> > > new file mode 100644
> > > index 0000000..bcce2f5
> > > --- /dev/null
> > > +++ b/meta-selftest/conf/machine/qemux86copy.conf
> > > @@ -0,0 +1,3 @@
> > > +require conf/machine/qemux86.conf
> > > +
> > > +MACHINEOVERRIDES .= ":qemux86"
> > 
> > I wasn't changing MACHINEOVERRIDES (in case this would hide some
> > possible sstate differences),
> 
> I think this actually makes it more likely to highlight differences, but
> either way, the tests I ran did highlight a raft of issues and it seems
> to be good enough to find several problems. I've sent enough fixes to
> make the new selftest pass.
> 
> >  instead I had to duplicate couple qemux86
> > overrides in linux-yocto, but this is still useful.
> > 
> > I'll drop my version after testing with sstate-diff-machines.sh that
> > both return the same results.
> 
> Sounds good.
> 
> FWIW you can run the new test with:
> 
> oe-selftest --run-tests sstatetests.SStateTests.test_sstate_sametune_samesigs
> 
> in case that isn't clear. It doesn't rely on building anything so is
> comparatively fast.

same as ssstate-diff-machines.sh --analyze :)

-- 
Martin 'JaMa' Jansa     jabber: Martin.Jansa@gmail.com

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 188 bytes --]