From: Peter Krempa <pkrempa@redhat.com>

Any QEMU we support requires use of NBD for disk migration which is not
supported on tunnelled migration. This is validated in
'qemuMigrationSrcBeginPhase'. Passing the list of disks to migrate is
thus pointless. Remove the argument.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
---
 src/qemu/qemu_migration.c | 14 ++++++--------
 1 file changed, 6 insertions(+), 8 deletions(-)

diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c
index fec808ccfb..4d136e259b 100644
--- a/src/qemu/qemu_migration.c
+++ b/src/qemu/qemu_migration.c
@@ -5428,7 +5428,6 @@ qemuMigrationSrcPerformTunnel(virQEMUDriver *driver,
                               unsigned long bandwidth,
                               virConnectPtr dconn,
                               const char *graphicsuri,
-                              const char **migrate_disks,
                               qemuMigrationParams *migParams)
 {
     int ret = -1;
@@ -5437,10 +5436,10 @@ qemuMigrationSrcPerformTunnel(virQEMUDriver *driver,

     VIR_DEBUG("driver=%p, vm=%p, st=%p, cookiein=%s, cookieinlen=%d, "
               "cookieout=%p, cookieoutlen=%p, flags=0x%x, bandwidth=%lu, "
-              "graphicsuri=%s, migrate_disks=%p",
+              "graphicsuri=%s",
               driver, vm, st, NULLSTR(cookiein), cookieinlen,
               cookieout, cookieoutlen, flags, bandwidth,
-              NULLSTR(graphicsuri), migrate_disks);
+              NULLSTR(graphicsuri));

     spec.fwdType = MIGRATION_FWD_STREAM;
     spec.fwd.stream = st;
@@ -5464,11 +5463,11 @@ qemuMigrationSrcPerformTunnel(virQEMUDriver *driver,
         goto cleanup;
     }

-    /* Migration with NBD is not supported with _TUNNELLED, thus
-     * 'migrate_disks_detect_zeroes' is NULL here */
+    /* Migration with NBD is not supported with _TUNNELLED, thus all corresponding
+     * parameters are NULL here */
     ret = qemuMigrationSrcRun(driver, vm, persist_xml, cookiein, cookieinlen,
                               cookieout, cookieoutlen, flags, bandwidth, &spec,
-                              dconn, graphicsuri, migrate_disks, NULL,
+                              dconn, graphicsuri, NULL, NULL,
                               migParams, NULL);

  cleanup:
@@ -5610,7 +5609,7 @@ qemuMigrationSrcPerformPeer2Peer2(virQEMUDriver *driver,
         ret = qemuMigrationSrcPerformTunnel(driver, vm, st, NULL,
                                             NULL, 0, NULL, NULL,
                                             flags, bandwidth, dconn,
-                                            NULL, NULL, migParams);
+                                            NULL, migParams);
     else
         ret = qemuMigrationSrcPerformNative(driver, vm, NULL, uri_out,
                                             cookie, cookielen,
@@ -5886,7 +5885,6 @@ qemuMigrationSrcPerformPeer2Peer3(virQEMUDriver *driver,
                                                 cookiein, cookieinlen,
                                                 &cookieout, &cookieoutlen,
                                                 flags, bandwidth, dconn, graphicsuri,
-                                                migrate_disks,
                                                 migParams);
         } else {
             ret = qemuMigrationSrcPerformNative(driver, vm, persist_xml, uri,
-- 
2.53.0

On a Wednesday in 2026, Peter Krempa via Devel wrote:
>From: Peter Krempa <pkrempa@redhat.com>
>
>Any QEMU we support requires use of NBD for disk migration which is not
>supported on tunnelled migration. This is validated in

s/on/with/ ?

>'qemuMigrationSrcBeginPhase'. Passing the list of disks to migrate is
>thus pointless. Remove the argument.
>
>Signed-off-by: Peter Krempa <pkrempa@redhat.com>
>---
> src/qemu/qemu_migration.c | 14 ++++++--------
> 1 file changed, 6 insertions(+), 8 deletions(-)
>

Reviewed-by: Ján Tomko <jtomko@redhat.com>

Jano