From: Andrey Drobyshev <andrey.drobyshev@virtuozzo.com>
To: svt-core@virtuozzo.com
Cc: Andrey Drobyshev <andrey.drobyshev@virtuozzo.com>
Subject: [PATCH QEMU HCI-8.1 17/21] vfio/listener: skip dirty page tracking during CPR
Date: Fri, 09 Oct 2026 17:56:16 +0300 [thread overview]
Message-ID: <20261009-vfio_v1-v1-17-2ae91d5b8048@virtuozzo.com> (raw)
In-Reply-To: <20261009-vfio_v1-v1-0-2ae91d5b8048@virtuozzo.com>
During cpr-transfer and cpr-exec guest RAM is preserved in place, not
copied, so there's no need to track which pages the device writes to.
Yet starting the migration also starts VFIO dirty tracking, on the
container or on the device itself (DMA logging). With cpr-exec it isn't
stopped before exec, so e.g. the DMA page tracker of an NVIDIA vGPU is
left running after the update.
Don't start or stop dirty tracking in CPR modes.
Signed-off-by: Andrey Drobyshev <andrey.drobyshev@virtuozzo.com>
---
hw/vfio/listener.c | 10 ++++++++++
1 file changed, 10 insertions(+)
diff --git a/hw/vfio/listener.c b/hw/vfio/listener.c
index afcf518f567..5585d656c4a 100644
--- a/hw/vfio/listener.c
+++ b/hw/vfio/listener.c
@@ -41,6 +41,7 @@
#include "trace.h"
#include "qapi/error.h"
#include "migration/misc.h"
+#include "migration/cpr.h"
#include "migration/qemu-file.h"
#include "system/tcg.h"
#include "system/tpm.h"
@@ -1012,6 +1013,11 @@ static bool vfio_listener_log_global_start(MemoryListener *listener,
listener);
bool ret;
+ /* Preserved devices keep DMA-ing into preserved RAM, nothing to track */
+ if (cpr_incoming_needed(NULL)) {
+ return true;
+ }
+
if (vfio_container_devices_dirty_tracking_is_supported(bcontainer)) {
ret = vfio_devices_dma_logging_start(bcontainer, errp);
} else {
@@ -1031,6 +1037,10 @@ static void vfio_listener_log_global_stop(MemoryListener *listener)
Error *local_err = NULL;
int ret = 0;
+ if (cpr_incoming_needed(NULL)) {
+ return;
+ }
+
if (vfio_container_devices_dirty_tracking_is_supported(bcontainer)) {
vfio_devices_dma_logging_stop(bcontainer);
} else {
--
2.47.1
next prev parent reply other threads:[~2026-10-09 14:56 UTC|newest]
Thread overview: 43+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-10-09 14:55 [PATCH QEMU HCI-8.1 00/21] vfio: support qemu-update with passed through GPU's and vGPU's #VSTOR-103860 Andrey Drobyshev
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 01/21] migration: Plug memory leaks after migrate_set_error() Andrey Drobyshev
2026-10-09 15:09 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 02/21] hw/vfio: Fix liveness check in vfio_connect_kvm_msi_virq() Andrey Drobyshev
2026-10-09 15:09 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 03/21] vfio/pci: Reject invalid MSI-X Table and PBA BIR values Andrey Drobyshev
2026-10-09 15:10 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 04/21] vfio/pci: Fix information leak in vfio_rom_read() Andrey Drobyshev
2026-10-09 15:10 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 05/21] vfio/pci: Initialize rom_read_failed in vfio_pci_load_rom() Andrey Drobyshev
2026-10-09 15:13 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 06/21] vfio/pci: Grow buffer in vfio_pci_host_match() Andrey Drobyshev
2026-10-09 15:10 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 07/21] q35: Fix migration of SMRAM state Andrey Drobyshev
2026-10-09 15:12 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 08/21] migration: ensure APIC is loaded prior to VFIO PCI devices Andrey Drobyshev
2026-10-09 15:13 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 09/21] migration: Fix regression of passing error_fatal into vmstate_load_state() Andrey Drobyshev
2026-10-09 15:17 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 10/21] migration: yank: Move register instance earlier Andrey Drobyshev
2026-10-09 15:16 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 11/21] migration: Free cpr-transfer MigrationAddress along with gsource Andrey Drobyshev
2026-10-09 15:17 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 12/21] migration: Move CPR HUP watch to cpr-transfer.c Andrey Drobyshev
2026-10-09 15:18 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 13/21] migration: Fix crash on second migration when cancel early Andrey Drobyshev
2026-10-09 15:20 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 14/21] kvm: tolerate ioctls after kvm_close() Andrey Drobyshev
2026-10-09 15:22 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 15/21] kvm: fix coalesced MMIO flush racing with kvm_close() Andrey Drobyshev
2026-10-09 15:22 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 16/21] vfio/migration: don't migrate device state during CPR Andrey Drobyshev
2026-10-09 15:29 ` no-reply
2026-10-09 14:56 ` Andrey Drobyshev [this message]
2026-10-09 15:27 ` [PATCH QEMU HCI-8.1 17/21] vfio/listener: skip dirty page tracking " no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 18/21] VZ: physmem: align fd-backed RAM like anonymous RAM #VSTOR-103860 Andrey Drobyshev
2026-10-09 15:31 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 19/21] VZ: migration/cpr-exec: abort cpr-exec if memory can't be preserved Andrey Drobyshev
2026-10-09 15:31 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 20/21] VZ: vfio/region: preserve BAR mappings across cpr-exec #VSTOR-103860 Andrey Drobyshev
2026-10-09 15:36 ` no-reply
2026-10-09 14:56 ` [PATCH QEMU HCI-8.1 21/21] VZ: migration/cpr-exec: allow cpr-exec with VFIO devices #VSTOR-103860 Andrey Drobyshev
2026-10-09 15:34 ` no-reply
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20261009-vfio_v1-v1-17-2ae91d5b8048@virtuozzo.com \
--to=andrey.drobyshev@virtuozzo.com \
--cc=svt-core@virtuozzo.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox