[PATCH v1 12/15] cpus-common: ensure auto-assigned cpu_indexes don't clash

Alex Bennée posted 15 patches 5 years, 8 months ago
There is a newer version of this series
[PATCH v1 12/15] cpus-common: ensure auto-assigned cpu_indexes don't clash
Posted by Alex Bennée 5 years, 8 months ago
Basing the cpu_index on the number of currently allocated vCPUs fails
when vCPUs aren't removed in a LIFO manner. This is especially true
when we are allocating a cpu_index for each guest thread in
linux-user where there is no ordering constraint on their allocation
and de-allocation.

[I've dropped the assert which is there to guard against out-of-order
removal as this should probably be caught higher up the stack. Maybe
we could just ifdef CONFIG_SOFTTMU it?]

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Cc: Nikolay Igotti <igotti@gmail.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Igor Mammedov <imammedo@redhat.com>
Cc: Eduardo Habkost <ehabkost@redhat.com>

---
v2
  - slightly tweak the index algorithm to preserve cpu_index = 0
---
 cpus-common.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/cpus-common.c b/cpus-common.c
index 55d5df89237..70a9d12981a 100644
--- a/cpus-common.c
+++ b/cpus-common.c
@@ -61,13 +61,15 @@ static bool cpu_index_auto_assigned;
 static int cpu_get_free_index(void)
 {
     CPUState *some_cpu;
-    int cpu_index = 0;
+    int max_cpu_index = 0;
 
     cpu_index_auto_assigned = true;
     CPU_FOREACH(some_cpu) {
-        cpu_index++;
+        if (some_cpu->cpu_index >= max_cpu_index) {
+            max_cpu_index = some_cpu->cpu_index + 1;
+        }
     }
-    return cpu_index;
+    return max_cpu_index;
 }
 
 void cpu_list_add(CPUState *cpu)
@@ -90,8 +92,6 @@ void cpu_list_remove(CPUState *cpu)
         return;
     }
 
-    assert(!(cpu_index_auto_assigned && cpu != QTAILQ_LAST(&cpus)));
-
     QTAILQ_REMOVE_RCU(&cpus, cpu, node);
     cpu->cpu_index = UNASSIGNED_CPU_INDEX;
 }
-- 
2.20.1


Re: [PATCH v1 12/15] cpus-common: ensure auto-assigned cpu_indexes don't clash
Posted by Richard Henderson 5 years, 8 months ago
On 5/20/20 7:05 AM, Alex Bennée wrote:
> Basing the cpu_index on the number of currently allocated vCPUs fails
> when vCPUs aren't removed in a LIFO manner. This is especially true
> when we are allocating a cpu_index for each guest thread in
> linux-user where there is no ordering constraint on their allocation
> and de-allocation.
> 
> [I've dropped the assert which is there to guard against out-of-order
> removal as this should probably be caught higher up the stack. Maybe
> we could just ifdef CONFIG_SOFTTMU it?]
> 
> Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
> Cc: Nikolay Igotti <igotti@gmail.com>
> Cc: Paolo Bonzini <pbonzini@redhat.com>
> Cc: Igor Mammedov <imammedo@redhat.com>
> Cc: Eduardo Habkost <ehabkost@redhat.com>
> 
> ---
> v2
>   - slightly tweak the index algorithm to preserve cpu_index = 0
> ---
>  cpus-common.c | 10 +++++-----
>  1 file changed, 5 insertions(+), 5 deletions(-)

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>


r~