On 11/19/25 1:50 PM, Christophe Leroy wrote:

Le 19/11/2025 à 09:08, Shrikanth Hegde a écrit :

quoted

Hi Greg.

On 11/19/25 1:12 PM, Greg KH wrote:

quoted

On Wed, Nov 19, 2025 at 11:50:59AM +0530, Shrikanth Hegde wrote:

quoted

This is helper patch which could be used to set the range of CPUs as
paravirt. One could make use of this for quick testing of this infra
instead of writing arch specific code.

This is currently not meant be merged, since paravirt sysfs file is 
meant
to be Read-Only.

echo 100-200,600-700 >  /sys/devices/system/cpu/paravirt
cat /sys/devices/system/cpu/paravirt
100-200,600-700

echo > /sys/devices/system/cpu/paravirt
cat /sys/devices/system/cpu/paravirt

Signed-off-by: Shrikanth Hegde <redacted>
---
  drivers/base/cpu.c | 48 ++++++++++++++++++++++++++++++++++++++++++ 
++--
  1 file changed, 46 insertions(+), 2 deletions(-)

diff --git a/drivers/base/cpu.c b/drivers/base/cpu.c
index 59ceae217b22..043e4f4ce1a9 100644
--- a/drivers/base/cpu.c
+++ b/drivers/base/cpu.c

@@ -375,12 +375,57 @@ static int cpu_uevent(const struct device 

*dev, struct kobj_uevent_env *env)
  #endif
  #ifdef CONFIG_PARAVIRT
+static ssize_t store_paravirt_cpus(struct device *dev,
+                   struct device_attribute *attr,
+                   const char *buf, size_t count)
+{
+    cpumask_var_t temp_mask;
+    int retval = 0;
+
+    if (!alloc_cpumask_var(&temp_mask, GFP_KERNEL))
+        return -ENOMEM;
+
+    retval = cpulist_parse(buf, temp_mask);
+    if (retval)
+        goto free_mask;
+
+    /* ALL cpus can't be marked as paravirt */
+    if (cpumask_equal(temp_mask, cpu_online_mask)) {
+        retval = -EINVAL;
+        goto free_mask;
+    }
+    if (cpumask_weight(temp_mask) > num_online_cpus()) {
+        retval = -EINVAL;
+        goto free_mask;
+    }
+
+    /* No more paravirt cpus */
+    if (cpumask_empty(temp_mask)) {
+        cpumask_copy((struct cpumask *)&__cpu_paravirt_mask, 
temp_mask);
+    } else {
+        cpumask_copy((struct cpumask *)&__cpu_paravirt_mask, 
temp_mask);
+
+        /* Enable tick on nohz_full cpu */
+        int cpu;
+        for_each_cpu(cpu, temp_mask) {
+            if (tick_nohz_full_cpu(cpu))
+                tick_nohz_dep_set_cpu(cpu, TICK_DEP_BIT_SCHED);
+        }
+    }
+
+    retval = count;
+
+free_mask:
+    free_cpumask_var(temp_mask);
+    return retval;
+}
+
  static ssize_t print_paravirt_cpus(struct device *dev,
                     struct device_attribute *attr, char *buf)
  {
      return sysfs_emit(buf, "%*pbl\n", 
cpumask_pr_args(cpu_paravirt_mask));
  }
-static DEVICE_ATTR(paravirt, 0444, print_paravirt_cpus, NULL);
+static DEVICE_ATTR(paravirt, 0644, print_paravirt_cpus, 
store_paravirt_cpus);

DEVICE_ATTR_RW()?

ok.

quoted

And where is the documentation update for this sysfs file change?

[RFC PATCH v4 11/17] has the documentation of this sysfs file.

There is a problem in the way you sent this patch and the other helper 
patch. They appear in the cover letter of your series are part of it but 
at the end the series is only sent with 15 patches over 17, and the last 
two patches appear as independent from the series:

Series at https://patchwork.ozlabs.org/project/linuxppc-dev/list/? 
series=482680

Other patches are on their own: https://patchwork.ozlabs.org/project/ 
linuxppc-dev/list/?submitter=87866

I edited the patch header before sending. Wanted to say they are debug patches.
Thought "helper" maybe a name. My bad.

I didn't realize that it could seen an separate patches :(
So sorry. I thought it would come up as thread of the series. Like it showed up in
https://lore.kernel.org/all/20251119062100.1112520-12-sshegde@linux.ibm.com/#r (local)

Should I resend the series?