Re: [OE-Core][master][kirkstone][PATCH] numactl: skip test case when target platform doesn't have 2 CPU node

Re: [OE-Core][master][kirkstone][PATCH] numactl: skip test case when target platform doesn't have 2 CPU node

[Xiangyu Chen]

On 1/19/23 13:24, Xiangyu Chen wrote:
> From: Xiangyu Chen <xiangyu.chen@windriver.com>
>
> when current test platform doesn't have 2 or more CPU node, the test
> case would report FAIL, according to numa test script and numademo
> code, when return code=77 should be skip test, so using SKIP instead
> of FAIL in test script.
>
> Signed-off-by: Xiangyu Chen <xiangyu.chen@windriver.com>

Hi Steve,

Friendly ping, this fix has already merged to master and it also can be 
applied on kirkstone, could you please help to merge it on kirkstone? 
thanks!


Br,

Xiangyu

> ---
>   .../numactl/numactl/Fix-the-test-output-format.patch        | 3 ++-
>   meta/recipes-support/numactl/numactl/run-ptest              | 6 +++++-
>   2 files changed, 7 insertions(+), 2 deletions(-)
>
> diff --git a/meta/recipes-support/numactl/numactl/Fix-the-test-output-format.patch b/meta/recipes-support/numactl/numactl/Fix-the-test-output-format.patch
> index 9812ecc8b3..a7bc8d322e 100644
> --- a/meta/recipes-support/numactl/numactl/Fix-the-test-output-format.patch
> +++ b/meta/recipes-support/numactl/numactl/Fix-the-test-output-format.patch
> @@ -7,6 +7,7 @@ Upstream-Status: Pending
>   
>   Signed-off-by: Roy Li <rongqing.li@windriver.com>
>   Signed-off-by: Li Xin <lixin.fnst@cn.fujitsu.com>
> +Signed-off-by: Xiangyu Chen <xiangyu.chen@windriver.com>
>   ---
>    test/regress  |  6 +++---
>    test/regress2 | 11 +++++------
> @@ -20,7 +21,7 @@ index 2ce1705..d086a47 100755
>    	if [ $numnodes -lt 2 ] ; then
>    	    echo "need at least two nodes with at least $NEEDPAGES each of"
>    	    echo "free memory for mempolicy regression tests"
> -+	    echo "FAIL: numa regress"
> ++	    echo "SKIP: numa regress"
>   	    exit 77  # Skip test
>    	fi
>    }
> diff --git a/meta/recipes-support/numactl/numactl/run-ptest b/meta/recipes-support/numactl/numactl/run-ptest
> index bf269da755..e019b0d364 100755
> --- a/meta/recipes-support/numactl/numactl/run-ptest
> +++ b/meta/recipes-support/numactl/numactl/run-ptest
> @@ -8,7 +8,11 @@ if ! numactl -s | grep -q "No NUMA support available on this system."; then
>   	if  numademo -t -e 10M; then
>   		echo "PASS: numademo"
>   	else
> -		echo "FAIL: numademo"
> +		if [ "$?" = 77 ] ; then
> +			echo "SKIP: numademo"
> +		else
> +			echo "FAIL: numademo"
> +		fi
>   	fi
>   else
>   	echo "SKIP: ./../test/bind_range"
>
> -=-=-=-=-=-=-=-=-=-=-=-
> Links: You receive all messages sent to this group.
> View/Reply Online (#176122): https://lists.openembedded.org/g/openembedded-core/message/176122
> Mute This Topic: https://lists.openembedded.org/mt/96372135/7175143
> Group Owner: openembedded-core+owner@lists.openembedded.org
> Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [xiangyu.chen@eng.windriver.com]
> -=-=-=-=-=-=-=-=-=-=-=-
>

Re: [OE-Core][master][kirkstone][PATCH] numactl: skip test case when target platform doesn't have 2 CPU node

[Steve Sakoman]

On Sun, Jan 29, 2023 at 5:18 PM Xiangyu Chen
<xiangyu.chen@eng.windriver.com> wrote:
>
>
> On 1/19/23 13:24, Xiangyu Chen wrote:
> > From: Xiangyu Chen <xiangyu.chen@windriver.com>
> >
> > when current test platform doesn't have 2 or more CPU node, the test
> > case would report FAIL, according to numa test script and numademo
> > code, when return code=77 should be skip test, so using SKIP instead
> > of FAIL in test script.
> >
> > Signed-off-by: Xiangyu Chen <xiangyu.chen@windriver.com>
>
> Friendly ping, this fix has already merged to master and it also can be
> applied on kirkstone, could you please help to merge it on kirkstone?
> thanks!

This patch is in my current test queue, so if all goes well it will be
in the next patchset for both langdale and kirkstone.

Steve

[OE-Core][master][kirkstone][PATCH] numactl: skip test case when target platform doesn't have 2 CPU node

[Xiangyu Chen]

From: Xiangyu Chen <xiangyu.chen@windriver.com>

when current test platform doesn't have 2 or more CPU node, the test
case would report FAIL, according to numa test script and numademo
code, when return code=77 should be skip test, so using SKIP instead
of FAIL in test script.

Signed-off-by: Xiangyu Chen <xiangyu.chen@windriver.com>
---
 .../numactl/numactl/Fix-the-test-output-format.patch        | 3 ++-
 meta/recipes-support/numactl/numactl/run-ptest              | 6 +++++-
 2 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/meta/recipes-support/numactl/numactl/Fix-the-test-output-format.patch b/meta/recipes-support/numactl/numactl/Fix-the-test-output-format.patch
index 9812ecc8b3..a7bc8d322e 100644
--- a/meta/recipes-support/numactl/numactl/Fix-the-test-output-format.patch
+++ b/meta/recipes-support/numactl/numactl/Fix-the-test-output-format.patch
@@ -7,6 +7,7 @@ Upstream-Status: Pending
 
 Signed-off-by: Roy Li <rongqing.li@windriver.com>
 Signed-off-by: Li Xin <lixin.fnst@cn.fujitsu.com>
+Signed-off-by: Xiangyu Chen <xiangyu.chen@windriver.com>
 ---
  test/regress  |  6 +++---
  test/regress2 | 11 +++++------
@@ -20,7 +21,7 @@ index 2ce1705..d086a47 100755
  	if [ $numnodes -lt 2 ] ; then
  	    echo "need at least two nodes with at least $NEEDPAGES each of"
  	    echo "free memory for mempolicy regression tests"
-+	    echo "FAIL: numa regress"
++	    echo "SKIP: numa regress"
 	    exit 77  # Skip test
  	fi
  }
diff --git a/meta/recipes-support/numactl/numactl/run-ptest b/meta/recipes-support/numactl/numactl/run-ptest
index bf269da755..e019b0d364 100755
--- a/meta/recipes-support/numactl/numactl/run-ptest
+++ b/meta/recipes-support/numactl/numactl/run-ptest
@@ -8,7 +8,11 @@ if ! numactl -s | grep -q "No NUMA support available on this system."; then
 	if  numademo -t -e 10M; then
 		echo "PASS: numademo"
 	else
-		echo "FAIL: numademo"
+		if [ "$?" = 77 ] ; then
+			echo "SKIP: numademo"
+		else
+			echo "FAIL: numademo"
+		fi
 	fi
 else
 	echo "SKIP: ./../test/bind_range"
-- 
2.34.1