]> git.ipfire.org Git - thirdparty/qemu.git/commit
vhost: Don't set vring call if guest notifier is unused
authorHuaitong Han <hanht2@chinatelecom.cn>
Thu, 22 May 2025 10:05:48 +0000 (18:05 +0800)
committerMichael S. Tsirkin <mst@redhat.com>
Mon, 2 Jun 2025 18:29:15 +0000 (14:29 -0400)
commita9403bfcd93025df7b1924d0cf34fbc408955b33
treedf88014dfbbc632cea4f798e7cae0b86e687dd5c
parent82acc8c83f93814bb01669412a0a226c39fa8d77
vhost: Don't set vring call if guest notifier is unused

The vring call fd is set even when the guest does not use MSI-X (e.g., in the
case of virtio PMD), leading to unnecessary CPU overhead for processing
interrupts.

The commit 96a3d98d2c("vhost: don't set vring call if no vector") optimized the
case where MSI-X is enabled but the queue vector is unset. However, there's an
additional case where the guest uses INTx and the INTx_DISABLED bit in the PCI
config is set, meaning that no interrupt notifier will actually be used.

In such cases, the vring call fd should also be cleared to avoid redundant
interrupt handling.

Fixes: 96a3d98d2c("vhost: don't set vring call if no vector")
Reported-by: Zhiyuan Yuan <yuanzhiyuan@chinatelecom.cn>
Signed-off-by: Jidong Xia <xiajd@chinatelecom.cn>
Signed-off-by: Huaitong Han <hanht2@chinatelecom.cn>
Message-Id: <20250522100548.212740-1-hanht2@chinatelecom.cn>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
hw/pci/pci.c
hw/virtio/virtio-pci.c
include/hw/pci/pci.h