diff mbox series

test: Use nodeInfo content for virNodeGetCPUMap

Message ID 407cb2b56fea2a9621e5dbabe38fdb45764eb691.1556196423.git.crobinso@redhat.com
State New
Headers show
Series test: Use nodeInfo content for virNodeGetCPUMap | expand

Commit Message

Cole Robinson April 25, 2019, 12:57 p.m. UTC
Right now the test driver implementation of virNodeGetCPUMap uses
hardcoded data. Instead use cpu data from our nodeInfo, which can be
overridden with input test XML. This helps to emulate some virsh
commands (vcpupin, hostcpumap) with a large number of host cpus. [1]

The content reported for test:///default is now slightly different:
previously it would report 8 total cpus with 0,2,4 online. Now it
reports 16 cpus with all online, which is the pre-existing nodeInfo
default.

[1] https://www.redhat.com/archives/libvir-list/2019-March/msg02042.html

Signed-off-by: Cole Robinson <crobinso@redhat.com>

---
John, in trying to understand and test your bitmap patch, I added
this to the test driver. I think there's a bigger issue at play
but I'm still trying to work it out, I'll send a follow up email
when I've collected my thoughts

 src/test/test_driver.c | 35 +++++++++++++++++++++++++++++------
 1 file changed, 29 insertions(+), 6 deletions(-)

-- 
2.21.0

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list

Comments

Michal Prívozník April 29, 2019, 2:57 p.m. UTC | #1
On 4/25/19 2:57 PM, Cole Robinson wrote:
> Right now the test driver implementation of virNodeGetCPUMap uses

> hardcoded data. Instead use cpu data from our nodeInfo, which can be

> overridden with input test XML. This helps to emulate some virsh

> commands (vcpupin, hostcpumap) with a large number of host cpus. [1]

> 

> The content reported for test:///default is now slightly different:

> previously it would report 8 total cpus with 0,2,4 online. Now it

> reports 16 cpus with all online, which is the pre-existing nodeInfo

> default.

> 

> [1] https://www.redhat.com/archives/libvir-list/2019-March/msg02042.html

> 

> Signed-off-by: Cole Robinson <crobinso@redhat.com>

> ---

> John, in trying to understand and test your bitmap patch, I added

> this to the test driver. I think there's a bigger issue at play

> but I'm still trying to work it out, I'll send a follow up email

> when I've collected my thoughts

> 

>   src/test/test_driver.c | 35 +++++++++++++++++++++++++++++------

>   1 file changed, 29 insertions(+), 6 deletions(-)


ACK

Michal

--
libvir-list mailing list
libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-list
diff mbox series

Patch

diff --git a/src/test/test_driver.c b/src/test/test_driver.c
index d5eecf4b7f..f00b6eb1c8 100644
--- a/src/test/test_driver.c
+++ b/src/test/test_driver.c
@@ -5811,23 +5811,46 @@  static int testConnectListAllDomains(virConnectPtr conn,
 }
 
 static int
-testNodeGetCPUMap(virConnectPtr conn ATTRIBUTE_UNUSED,
+testNodeGetCPUMap(virConnectPtr conn,
                   unsigned char **cpumap,
                   unsigned int *online,
                   unsigned int flags)
 {
+    testDriverPtr privconn = conn->privateData;
+    int maxcpus = VIR_NODEINFO_MAXCPUS(privconn->nodeInfo);
+    virBitmapPtr cpus = NULL;
+    int ret = -1;
+    int dummy;
+    size_t i;
+
     virCheckFlags(0, -1);
 
+    if (!cpumap && !online)
+        return maxcpus;
+
+    if (!(cpus = virBitmapNew(maxcpus)))
+        goto cleanup;
+
     if (cpumap) {
-        if (VIR_ALLOC_N(*cpumap, 1) < 0)
-            return -1;
-        *cpumap[0] = 0x15;
+        for (i = 0; i < privconn->nodeInfo.cpus; i++) {
+            if (virBitmapSetBit(cpus, i) < 0)
+                goto cleanup;
+        }
+
+        if (virBitmapToData(cpus, cpumap, &dummy) < 0)
+            goto cleanup;
     }
 
     if (online)
-        *online = 3;
+        *online = privconn->nodeInfo.cpus;
 
-    return  8;
+    ret = maxcpus;
+
+ cleanup:
+    if (ret < 0 && cpumap)
+        VIR_FREE(*cpumap);
+    VIR_FREE(cpus);
+    return ret;
 }
 
 static char *