You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
qemu-kvm/SOURCES/kvm-Revert-globally-limit-t...

59 lines
2.2 KiB

From 5ab8613582fd56b847fe75750acb5b7255900b35 Mon Sep 17 00:00:00 2001
From: Vitaly Kuznetsov <vkuznets@redhat.com>
Date: Thu, 9 Jun 2022 11:55:15 +0200
Subject: [PATCH 15/16] Revert "globally limit the maximum number of CPUs"
RH-Author: Vitaly Kuznetsov <vkuznets@redhat.com>
RH-MergeRequest: 99: Revert "globally limit the maximum number of CPUs"
RH-Commit: [1/1] 13100d4a2209b2190a3654c1f9cf4ebade1e8d24 (vkuznets/qemu-kvm-c9s)
RH-Bugzilla: 2094270
RH-Acked-by: Andrew Jones <drjones@redhat.com>
RH-Acked-by: Cornelia Huck <cohuck@redhat.com>
RH-Acked-by: Thomas Huth <thuth@redhat.com>
Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=2094270
Brew: https://brewweb.engineering.redhat.com/brew/taskinfo?taskID=45871149
Upstream Status: RHEL-only
Tested: with upstream kernel
Downstream QEMU carries a patch that sets the hard limit of possible vCPUs
to the value that the KVM code of the kernel recommends as soft limit.
Upstream KVM code has been changed recently to not use an arbitrary soft
limit anymore, but to cap the value on the amount of available physical
CPUs of the host. This defeats the purpose of the downstream change in
QEMU completely. Drop the downstream-only patch to allow CPU overcommit.
This reverts commit 6669f6fa677d43144f39d6ad59725b7ba622f1c2.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
---
accel/kvm/kvm-all.c | 12 ------------
1 file changed, 12 deletions(-)
diff --git a/accel/kvm/kvm-all.c b/accel/kvm/kvm-all.c
index fdf0e4d429..5f1377ca04 100644
--- a/accel/kvm/kvm-all.c
+++ b/accel/kvm/kvm-all.c
@@ -2430,18 +2430,6 @@ static int kvm_init(MachineState *ms)
soft_vcpus_limit = kvm_recommended_vcpus(s);
hard_vcpus_limit = kvm_max_vcpus(s);
-#ifdef HOST_PPC64
- /*
- * On POWER, the kernel advertises a soft limit based on the
- * number of CPU threads on the host. We want to allow exceeding
- * this for testing purposes, so we don't want to set hard limit
- * to soft limit as on x86.
- */
-#else
- /* RHEL doesn't support nr_vcpus > soft_vcpus_limit */
- hard_vcpus_limit = soft_vcpus_limit;
-#endif
-
while (nc->name) {
if (nc->num > soft_vcpus_limit) {
warn_report("Number of %s cpus requested (%d) exceeds "
--
2.31.1