[Qemu-devel] [PATCH 0/3] Some fixes for make check-acceptance on a POWER host

[Qemu-devel] [PATCH 0/3] Some fixes for make check-acceptance on a POWER host

[David Gibson]

Here are several patches which address failures I see when attempting
to run "make check-acceptance" on a POWER host machine.

David Gibson (3):
  tests/acceptance: Specify arch for QueryCPUModelExpansion
  tests/acceptance: Handle ppc64le host arch correctly
  tests/acceptance: Increase timeout for LinuxSSH test

 tests/acceptance/avocado_qemu/__init__.py | 2 ++
 tests/acceptance/cpu_queries.py           | 3 +++
 tests/acceptance/linux_ssh_mips_malta.py  | 2 +-
 3 files changed, 6 insertions(+), 1 deletion(-)

-- 
2.21.0

[Qemu-devel] [PATCH 1/3] tests/acceptance: Specify arch for QueryCPUModelExpansion

[David Gibson]

At the moment this test runs on whatever the host arch is.  But it looks
for 'unavailable-features' which is an x86 specific cpu property.  Tag it
to always use qemu-system-x86_64.

Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
 tests/acceptance/cpu_queries.py | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/tests/acceptance/cpu_queries.py b/tests/acceptance/cpu_queries.py
index e71edec39f..af47d2795a 100644
--- a/tests/acceptance/cpu_queries.py
+++ b/tests/acceptance/cpu_queries.py
@@ -18,6 +18,9 @@ class QueryCPUModelExpansion(Test):
     """
 
     def test(self):
+        """
+        :avocado: tags=arch:x86_64
+        """
         self.vm.set_machine('none')
         self.vm.add_args('-S')
         self.vm.launch()
-- 
2.21.0

[Qemu-devel] [PATCH 2/3] tests/acceptance: Handle ppc64le host arch correctly

[David Gibson]

ppc64 and ppc64le are different archs from the host kernel point of view
and are advertised as such in uname.  But these cover the same set of CPUs,
just in different endianness modes.  qemu-system-ppc64 handles both modes,
so make sure we select the correct binary when running on ppc64le host
architecture.

Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
 tests/acceptance/avocado_qemu/__init__.py | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/tests/acceptance/avocado_qemu/__init__.py b/tests/acceptance/avocado_qemu/__init__.py
index 2b236a1cf0..0ba9c536f4 100644
--- a/tests/acceptance/avocado_qemu/__init__.py
+++ b/tests/acceptance/avocado_qemu/__init__.py
@@ -39,6 +39,8 @@ def pick_default_qemu_bin(arch=None):
     """
     if arch is None:
         arch = os.uname()[4]
+        if arch == 'ppc64le':
+            arch = 'ppc64'
     qemu_bin_relative_path = os.path.join("%s-softmmu" % arch,
                                           "qemu-system-%s" % arch)
     if is_readable_executable_file(qemu_bin_relative_path):
-- 
2.21.0

[Qemu-devel] [PATCH 3/3] tests/acceptance: Increase timeout for LinuxSSH test

[David Gibson]

At least on the hosts I use (a POWER8 Tuleta, a POWER9 Boston and an 8 cpu
x86 instance in OpenStack) the linux_ssh_mips_malta tests in make
check-acceptance frequently timeout.  Increasing the timeout from 2.5 to
5 minutes makes that much less common.

Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
 tests/acceptance/linux_ssh_mips_malta.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/acceptance/linux_ssh_mips_malta.py b/tests/acceptance/linux_ssh_mips_malta.py
index aafb0c39f6..7ed6c6c90f 100644
--- a/tests/acceptance/linux_ssh_mips_malta.py
+++ b/tests/acceptance/linux_ssh_mips_malta.py
@@ -20,7 +20,7 @@ from avocado.utils import archive
 
 class LinuxSSH(Test):
 
-    timeout = 150 # Not for 'configure --enable-debug --enable-debug-tcg'
+    timeout = 300 # Not for 'configure --enable-debug --enable-debug-tcg'
 
     KERNEL_COMMON_COMMAND_LINE = 'printk.time=0 '
     VM_IP = '127.0.0.1'
-- 
2.21.0

Re: [Qemu-devel] [PATCH 2/3] tests/acceptance: Handle ppc64le host arch correctly

[Philippe Mathieu-Daudé]

On 6/13/19 8:07 AM, David Gibson wrote:
> ppc64 and ppc64le are different archs from the host kernel point of view
> and are advertised as such in uname.  But these cover the same set of CPUs,
> just in different endianness modes.  qemu-system-ppc64 handles both modes,
> so make sure we select the correct binary when running on ppc64le host
> architecture.
> 
> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
> ---
>  tests/acceptance/avocado_qemu/__init__.py | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/tests/acceptance/avocado_qemu/__init__.py b/tests/acceptance/avocado_qemu/__init__.py
> index 2b236a1cf0..0ba9c536f4 100644
> --- a/tests/acceptance/avocado_qemu/__init__.py
> +++ b/tests/acceptance/avocado_qemu/__init__.py
> @@ -39,6 +39,8 @@ def pick_default_qemu_bin(arch=None):
>      """
>      if arch is None:
>          arch = os.uname()[4]
> +        if arch == 'ppc64le':
> +            arch = 'ppc64'

I prefer the generic patch from Cleber:
https://lists.gnu.org/archive/html/qemu-devel/2018-10/msg03418.html
(I guess remember another version with a json file)

>      qemu_bin_relative_path = os.path.join("%s-softmmu" % arch,
>                                            "qemu-system-%s" % arch)
>      if is_readable_executable_file(qemu_bin_relative_path):
> 

Re: [Qemu-devel] [PATCH 2/3] tests/acceptance: Handle ppc64le host arch correctly

[David Gibson]

[-- Attachment #1: Type: text/plain, Size: 1730 bytes --]

On Thu, Jun 13, 2019 at 11:01:19AM +0200, Philippe Mathieu-Daudé wrote:
> On 6/13/19 8:07 AM, David Gibson wrote:
> > ppc64 and ppc64le are different archs from the host kernel point of view
> > and are advertised as such in uname.  But these cover the same set of CPUs,
> > just in different endianness modes.  qemu-system-ppc64 handles both modes,
> > so make sure we select the correct binary when running on ppc64le host
> > architecture.
> > 
> > Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
> > ---
> >  tests/acceptance/avocado_qemu/__init__.py | 2 ++
> >  1 file changed, 2 insertions(+)
> > 
> > diff --git a/tests/acceptance/avocado_qemu/__init__.py b/tests/acceptance/avocado_qemu/__init__.py
> > index 2b236a1cf0..0ba9c536f4 100644
> > --- a/tests/acceptance/avocado_qemu/__init__.py
> > +++ b/tests/acceptance/avocado_qemu/__init__.py
> > @@ -39,6 +39,8 @@ def pick_default_qemu_bin(arch=None):
> >      """
> >      if arch is None:
> >          arch = os.uname()[4]
> > +        if arch == 'ppc64le':
> > +            arch = 'ppc64'
> 
> I prefer the generic patch from Cleber:
> https://lists.gnu.org/archive/html/qemu-devel/2018-10/msg03418.html
> (I guess remember another version with a json file)

I hadn't seen that, that does indeed look better.

> 
> >      qemu_bin_relative_path = os.path.join("%s-softmmu" % arch,
> >                                            "qemu-system-%s" % arch)
> >      if is_readable_executable_file(qemu_bin_relative_path):
> > 
> 

-- 
David Gibson			| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you.  NOT _the_ _other_
				| _way_ _around_!
http://www.ozlabs.org/~dgibson

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [Qemu-devel] [PATCH 2/3] tests/acceptance: Handle ppc64le host arch correctly

[David Gibson]

[-- Attachment #1: Type: text/plain, Size: 2017 bytes --]

On Fri, Jun 14, 2019 at 11:40:56AM +1000, David Gibson wrote:
> On Thu, Jun 13, 2019 at 11:01:19AM +0200, Philippe Mathieu-Daudé wrote:
> > On 6/13/19 8:07 AM, David Gibson wrote:
> > > ppc64 and ppc64le are different archs from the host kernel point of view
> > > and are advertised as such in uname.  But these cover the same set of CPUs,
> > > just in different endianness modes.  qemu-system-ppc64 handles both modes,
> > > so make sure we select the correct binary when running on ppc64le host
> > > architecture.
> > > 
> > > Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
> > > ---
> > >  tests/acceptance/avocado_qemu/__init__.py | 2 ++
> > >  1 file changed, 2 insertions(+)
> > > 
> > > diff --git a/tests/acceptance/avocado_qemu/__init__.py b/tests/acceptance/avocado_qemu/__init__.py
> > > index 2b236a1cf0..0ba9c536f4 100644
> > > --- a/tests/acceptance/avocado_qemu/__init__.py
> > > +++ b/tests/acceptance/avocado_qemu/__init__.py
> > > @@ -39,6 +39,8 @@ def pick_default_qemu_bin(arch=None):
> > >      """
> > >      if arch is None:
> > >          arch = os.uname()[4]
> > > +        if arch == 'ppc64le':
> > > +            arch = 'ppc64'
> > 
> > I prefer the generic patch from Cleber:
> > https://lists.gnu.org/archive/html/qemu-devel/2018-10/msg03418.html
> > (I guess remember another version with a json file)
> 
> I hadn't seen that, that does indeed look better.

Hrm.  Well, it is better but it's been outstanding for like 8 months.
I'd just like to get this working on a ppc host in the near future.

> 
> > 
> > >      qemu_bin_relative_path = os.path.join("%s-softmmu" % arch,
> > >                                            "qemu-system-%s" % arch)
> > >      if is_readable_executable_file(qemu_bin_relative_path):
> > > 
> > 
> 



-- 
David Gibson			| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you.  NOT _the_ _other_
				| _way_ _around_!
http://www.ozlabs.org/~dgibson

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [Qemu-devel] [PATCH 2/3] tests/acceptance: Handle ppc64le host arch correctly

[Philippe Mathieu-Daudé]

[-- Attachment #1.1: Type: text/plain, Size: 1815 bytes --]

On 6/14/19 3:53 AM, David Gibson wrote:
> On Fri, Jun 14, 2019 at 11:40:56AM +1000, David Gibson wrote:
>> On Thu, Jun 13, 2019 at 11:01:19AM +0200, Philippe Mathieu-Daudé wrote:
>>> On 6/13/19 8:07 AM, David Gibson wrote:
>>>> ppc64 and ppc64le are different archs from the host kernel point of view
>>>> and are advertised as such in uname.  But these cover the same set of CPUs,
>>>> just in different endianness modes.  qemu-system-ppc64 handles both modes,
>>>> so make sure we select the correct binary when running on ppc64le host
>>>> architecture.
>>>>
>>>> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
>>>> ---
>>>>  tests/acceptance/avocado_qemu/__init__.py | 2 ++
>>>>  1 file changed, 2 insertions(+)
>>>>
>>>> diff --git a/tests/acceptance/avocado_qemu/__init__.py b/tests/acceptance/avocado_qemu/__init__.py
>>>> index 2b236a1cf0..0ba9c536f4 100644
>>>> --- a/tests/acceptance/avocado_qemu/__init__.py
>>>> +++ b/tests/acceptance/avocado_qemu/__init__.py
>>>> @@ -39,6 +39,8 @@ def pick_default_qemu_bin(arch=None):
>>>>      """
>>>>      if arch is None:
>>>>          arch = os.uname()[4]
>>>> +        if arch == 'ppc64le':
>>>> +            arch = 'ppc64'
>>>
>>> I prefer the generic patch from Cleber:
>>> https://lists.gnu.org/archive/html/qemu-devel/2018-10/msg03418.html
>>> (I guess remember another version with a json file)
>>
>> I hadn't seen that, that does indeed look better.
> 
> Hrm.  Well, it is better but it's been outstanding for like 8 months.
> I'd just like to get this working on a ppc host in the near future.

True :S I wanted you to know Cleber was working on this, but I don't
have any problem to get your fix merged and Cleber work added on top of
it, so:
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [Qemu-devel] [PATCH 1/3] tests/acceptance: Specify arch for QueryCPUModelExpansion

[Wainer dos Santos Moschetta]

On 06/13/2019 03:07 AM, David Gibson wrote:
> At the moment this test runs on whatever the host arch is.  But it looks
> for 'unavailable-features' which is an x86 specific cpu property.  Tag it
> to always use qemu-system-x86_64.
>
> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
> ---
>   tests/acceptance/cpu_queries.py | 3 +++
>   1 file changed, 3 insertions(+)

Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>

>
> diff --git a/tests/acceptance/cpu_queries.py b/tests/acceptance/cpu_queries.py
> index e71edec39f..af47d2795a 100644
> --- a/tests/acceptance/cpu_queries.py
> +++ b/tests/acceptance/cpu_queries.py
> @@ -18,6 +18,9 @@ class QueryCPUModelExpansion(Test):
>       """
>   
>       def test(self):
> +        """
> +        :avocado: tags=arch:x86_64
> +        """
>           self.vm.set_machine('none')
>           self.vm.add_args('-S')
>           self.vm.launch()