[pve-devel] [PATCH qemu-server] fix #6680 (continued): do not use a top throttle node when using scsi-block

[pve-devel] [PATCH qemu-server] fix #6680 (continued): do not use a top throttle node when using scsi-block

[Fiona Ebner]

The throttle block driver does not support issuing ioctls. While the
driver could be patched in QEMU to pass along ioctls to the child,
that seems like a hack, because with scsi-block, throttle limits
already do not apply. This was already the case with '-drive' in
Proxmox VE 8.

Note that live mirroring would require special handling to make the
target be below a throttle node with the correct node name, but it
already is not supported for such disks:
> # qm disk move 103 scsi1 lvm
> unable to parse volume ID '/dev/sdg'

Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
 src/PVE/QemuServer/Blockdev.pm    |  8 ++++++++
 src/test/cfg2cmd/scsiblk.conf     |  5 +++++
 src/test/cfg2cmd/scsiblk.conf.cmd | 32 +++++++++++++++++++++++++++++++
 3 files changed, 45 insertions(+)
 create mode 100644 src/test/cfg2cmd/scsiblk.conf
 create mode 100644 src/test/cfg2cmd/scsiblk.conf.cmd

diff --git a/src/PVE/QemuServer/Blockdev.pm b/src/PVE/QemuServer/Blockdev.pm
index f5798cbf..d4062148 100644
--- a/src/PVE/QemuServer/Blockdev.pm
+++ b/src/PVE/QemuServer/Blockdev.pm
@@ -478,6 +478,14 @@ sub generate_drive_blockdev {
         };
     }
 
+    if ($drive->{scsiblock}) {
+        # When using scsi-block for the front-end device, throttling would not work in any case, and
+        # the throttle block driver doesn't allow doing the necessary ioctls(), so don't attach a
+        # throttle filter. Implementing live mirroring for such disks would require special care!
+        $child->{'node-name'} = top_node_name($drive_id);
+        return $child;
+    }
+
     # for fleecing and TPM backup, this is already the top node
     return $child if $options->{fleecing} || $options->{'tpm-backup'} || $options->{'no-throttle'};
 
diff --git a/src/test/cfg2cmd/scsiblk.conf b/src/test/cfg2cmd/scsiblk.conf
new file mode 100644
index 00000000..c767cef6
--- /dev/null
+++ b/src/test/cfg2cmd/scsiblk.conf
@@ -0,0 +1,5 @@
+boot: order=scsi0
+meta: creation-qemu=10.0.2,ctime=1755189639
+scsi0: /dev/sdg,scsiblock=1,size=32G
+smbios1: uuid=d47e2d5c-7068-46d4-9733-ff94083e28f6
+vmgenid: 0a212e2f-0d9a-4006-971e-17be9fcf3efe
diff --git a/src/test/cfg2cmd/scsiblk.conf.cmd b/src/test/cfg2cmd/scsiblk.conf.cmd
new file mode 100644
index 00000000..3f99b405
--- /dev/null
+++ b/src/test/cfg2cmd/scsiblk.conf.cmd
@@ -0,0 +1,32 @@
+/usr/bin/kvm \
+  -id 8006 \
+  -name 'vm8006,debug-threads=on' \
+  -no-shutdown \
+  -chardev 'socket,id=qmp,path=/var/run/qemu-server/8006.qmp,server=on,wait=off' \
+  -mon 'chardev=qmp,mode=control' \
+  -chardev 'socket,id=qmp-event,path=/var/run/qmeventd.sock,reconnect-ms=5000' \
+  -mon 'chardev=qmp-event,mode=control' \
+  -pidfile /var/run/qemu-server/8006.pid \
+  -daemonize \
+  -smbios 'type=1,uuid=d47e2d5c-7068-46d4-9733-ff94083e28f6' \
+  -smp '1,sockets=1,cores=1,maxcpus=1' \
+  -nodefaults \
+  -boot 'menu=on,strict=on,reboot-timeout=1000,splash=/usr/share/qemu-server/bootsplash.jpg' \
+  -vnc 'unix:/var/run/qemu-server/8006.vnc,password=on' \
+  -cpu kvm64,enforce,+kvm_pv_eoi,+kvm_pv_unhalt,+lahf_lm,+sep \
+  -m 512 \
+  -object '{"id":"throttle-drive-scsi0","limits":{},"qom-type":"throttle-group"}' \
+  -global 'PIIX4_PM.disable_s3=1' \
+  -global 'PIIX4_PM.disable_s4=1' \
+  -device 'pci-bridge,id=pci.1,chassis_nr=1,bus=pci.0,addr=0x1e' \
+  -device 'pci-bridge,id=pci.2,chassis_nr=2,bus=pci.0,addr=0x1f' \
+  -device 'vmgenid,guid=0a212e2f-0d9a-4006-971e-17be9fcf3efe' \
+  -device 'piix3-usb-uhci,id=uhci,bus=pci.0,addr=0x1.0x2' \
+  -device 'usb-tablet,id=tablet,bus=uhci.0,port=1' \
+  -device 'VGA,id=vga,bus=pci.0,addr=0x2' \
+  -device 'virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x3,free-page-reporting=on' \
+  -iscsi 'initiator-name=iqn.1993-08.org.debian:01:aabbccddeeff' \
+  -device 'lsi,id=scsihw0,bus=pci.0,addr=0x5' \
+  -blockdev '{"cache":{"direct":true,"no-flush":false},"detect-zeroes":"on","discard":"ignore","driver":"raw","file":{"aio":"io_uring","cache":{"direct":true,"no-flush":false},"detect-zeroes":"on","discard":"ignore","driver":"host_device","filename":"/dev/sdg","node-name":"e8c91179e47ae3eb292a03618be5cf4","read-only":false},"node-name":"drive-scsi0","read-only":false}' \
+  -device 'scsi-hd,bus=scsihw0.0,scsi-id=0,drive=drive-scsi0,id=scsi0,device_id=drive-scsi0,bootindex=100,write-cache=on' \
+  -machine 'type=pc+pve0'
-- 
2.47.2



_______________________________________________
pve-devel mailing list
pve-devel@lists.proxmox.com
https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-devel

[pve-devel] applied: [PATCH qemu-server] fix #6680 (continued): do not use a top throttle node when using scsi-block

[Thomas Lamprecht]

On Thu, 14 Aug 2025 18:45:07 +0200, Fiona Ebner wrote:
> The throttle block driver does not support issuing ioctls. While the
> driver could be patched in QEMU to pass along ioctls to the child,
> that seems like a hack, because with scsi-block, throttle limits
> already do not apply. This was already the case with '-drive' in
> Proxmox VE 8.
> 
> Note that live mirroring would require special handling to make the
> target be below a throttle node with the correct node name, but it
> already is not supported for such disks:
> > # qm disk move 103 scsi1 lvm
> > unable to parse volume ID '/dev/sdg'
> 
> [...]

Applied, thanks!

[1/1] fix #6680 (continued): do not use a top throttle node when using scsi-block
      commit: c031d373d178c794cc7cb4fde2fd10dfca28a340


_______________________________________________
pve-devel mailing list
pve-devel@lists.proxmox.com
https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-devel