Re: [PATCH v12 02/27] migration: push Error **errp into vmstate_load_state()

[Arun Menon <armenon@redhat.com>]

Hi,

On Fri, Aug 29, 2025 at 07:37:09PM +0900, Akihiko Odaki wrote:
> On 2025/08/28 17:17, Arun Menon wrote:
> > H Akihiko,
> > 
> > Thanks for the review.
> > 
> > On Thu, Aug 28, 2025 at 10:34:54AM +0900, Akihiko Odaki wrote:
> > > On 2025/08/21 17:43, Arun Menon wrote:
> > > > This is an incremental step in converting vmstate loading
> > > > code to report error via Error objects instead of directly
> > > > printing it to console/monitor.
> > > > It is ensured that vmstate_load_state() must report an error
> > > > in errp, in case of failure.
> > > > 
> > > > The errors are temporarily reported using warn_report_err().
> > > > This is removed in the subsequent patches in this series,
> > > > when we are actually able to propagate the error to the calling
> > > > function using errp. Whereas, if we want the function to exit on
> > > > error, then error_fatal is passed.
> > > > 
> > > > Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
> > > > Signed-off-by: Arun Menon <armenon@redhat.com>
> > > > ---
> > > >    hw/display/virtio-gpu.c     |  2 +-
> > > >    hw/pci/pci.c                |  3 ++-
> > > >    hw/s390x/virtio-ccw.c       |  2 +-
> > > >    hw/scsi/spapr_vscsi.c       |  4 ++-
> > > >    hw/vfio/pci.c               |  5 +++-
> > > >    hw/virtio/virtio-mmio.c     |  3 ++-
> > > >    hw/virtio/virtio-pci.c      |  2 +-
> > > >    hw/virtio/virtio.c          |  7 +++--
> > > >    include/migration/vmstate.h |  2 +-
> > > >    migration/cpr.c             |  3 +--
> > > >    migration/savevm.c          |  8 ++++--
> > > >    migration/vmstate-types.c   | 22 +++++++++++----
> > > >    migration/vmstate.c         | 65 +++++++++++++++++++++++++++++++--------------
> > > >    tests/unit/test-vmstate.c   | 63 ++++++++++++++++++++++++++++++++++++-------
> > > >    ui/vdagent.c                |  5 +++-
> > > >    15 files changed, 147 insertions(+), 49 deletions(-)
> > > > 
> > > > diff --git a/hw/display/virtio-gpu.c b/hw/display/virtio-gpu.c
> > > > index 0a1a625b0ea6cf26cb0d799171a57ed3d3ab2442..5dc31bc6bfb0272e29a4364ab10de2595a4bedf7 100644
> > > > --- a/hw/display/virtio-gpu.c
> > > > +++ b/hw/display/virtio-gpu.c
> > > > @@ -1343,7 +1343,7 @@ static int virtio_gpu_load(QEMUFile *f, void *opaque, size_t size,
> > > >        }
> > > >        /* load & apply scanout state */
> > > > -    vmstate_load_state(f, &vmstate_virtio_gpu_scanouts, g, 1);
> > > > +    vmstate_load_state(f, &vmstate_virtio_gpu_scanouts, g, 1, &error_fatal);
> > > >        return 0;
> > > >    }
> > > > diff --git a/hw/pci/pci.c b/hw/pci/pci.c
> > > > index c70b5ceebaf1f2b10768bd030526cbb518da2b8d..6be932d3bb67ff0c4808707db2a7b6378a90e82b 100644
> > > > --- a/hw/pci/pci.c
> > > > +++ b/hw/pci/pci.c
> > > > @@ -934,7 +934,8 @@ void pci_device_save(PCIDevice *s, QEMUFile *f)
> > > >    int pci_device_load(PCIDevice *s, QEMUFile *f)
> > > >    {
> > > >        int ret;
> > > > -    ret = vmstate_load_state(f, &vmstate_pci_device, s, s->version_id);
> > > > +    ret = vmstate_load_state(f, &vmstate_pci_device, s, s->version_id,
> > > > +                             &error_fatal);
> > > >        /* Restore the interrupt status bit. */
> > > >        pci_update_irq_status(s);
> > > >        return ret;
> > > > diff --git a/hw/s390x/virtio-ccw.c b/hw/s390x/virtio-ccw.c
> > > > index d2f85b39f30f7fc82e0c600144c0a958e1269b2c..6a9641a03d5d3a38a4de7ceb9deffc0cc303bcff 100644
> > > > --- a/hw/s390x/virtio-ccw.c
> > > > +++ b/hw/s390x/virtio-ccw.c
> > > > @@ -1136,7 +1136,7 @@ static void virtio_ccw_save_config(DeviceState *d, QEMUFile *f)
> > > >    static int virtio_ccw_load_config(DeviceState *d, QEMUFile *f)
> > > >    {
> > > >        VirtioCcwDevice *dev = VIRTIO_CCW_DEVICE(d);
> > > > -    return vmstate_load_state(f, &vmstate_virtio_ccw_dev, dev, 1);
> > > > +    return vmstate_load_state(f, &vmstate_virtio_ccw_dev, dev, 1, &error_fatal);
> > > >    }
> > > >    static void virtio_ccw_pre_plugged(DeviceState *d, Error **errp)
> > > > diff --git a/hw/scsi/spapr_vscsi.c b/hw/scsi/spapr_vscsi.c
> > > > index 20f70fb2729de78b9636a6b8c869695dab4f8902..8622419497db650523d51bcb41557bbca254eaa3 100644
> > > > --- a/hw/scsi/spapr_vscsi.c
> > > > +++ b/hw/scsi/spapr_vscsi.c
> > > > @@ -642,15 +642,17 @@ static void *vscsi_load_request(QEMUFile *f, SCSIRequest *sreq)
> > > >        VSCSIState *s = VIO_SPAPR_VSCSI_DEVICE(bus->qbus.parent);
> > > >        vscsi_req *req;
> > > >        int rc;
> > > > +    Error *local_err = NULL;
> > > >        assert(sreq->tag < VSCSI_REQ_LIMIT);
> > > >        req = &s->reqs[sreq->tag];
> > > >        assert(!req->active);
> > > >        memset(req, 0, sizeof(*req));
> > > > -    rc = vmstate_load_state(f, &vmstate_spapr_vscsi_req, req, 1);
> > > > +    rc = vmstate_load_state(f, &vmstate_spapr_vscsi_req, req, 1, &local_err);
> > > >        if (rc) {
> > > >            fprintf(stderr, "VSCSI: failed loading request tag#%u\n", sreq->tag);
> > > > +        warn_report_err(local_err);
> > > >            return NULL;
> > > >        }
> > > >        assert(req->active);
> > > > diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
> > > > index 4fa692c1a32bcfa4e4939e5fcb64f2bf19905b3b..a65e34b6979eadfa0851666aeae7cf731a00fa40 100644
> > > > --- a/hw/vfio/pci.c
> > > > +++ b/hw/vfio/pci.c
> > > > @@ -2790,13 +2790,16 @@ static int vfio_pci_load_config(VFIODevice *vbasedev, QEMUFile *f)
> > > >        PCIDevice *pdev = &vdev->pdev;
> > > >        pcibus_t old_addr[PCI_NUM_REGIONS - 1];
> > > >        int bar, ret;
> > > > +    Error *local_err = NULL;
> > > >        for (bar = 0; bar < PCI_ROM_SLOT; bar++) {
> > > >            old_addr[bar] = pdev->io_regions[bar].addr;
> > > >        }
> > > > -    ret = vmstate_load_state(f, &vmstate_vfio_pci_config, vdev, 1);
> > > > +    ret = vmstate_load_state(f, &vmstate_vfio_pci_config, vdev, 1,
> > > > +                             &local_err);
> > > >        if (ret) {
> > > > +        warn_report_err(local_err);
> > > >            return ret;
> > > >        }
> > > > diff --git a/hw/virtio/virtio-mmio.c b/hw/virtio/virtio-mmio.c
> > > > index 532c67107ba1d2978a76cf49f9cdc1de1dea3e11..0a688909fc606a3c9fde933667ae8c309ab527d0 100644
> > > > --- a/hw/virtio/virtio-mmio.c
> > > > +++ b/hw/virtio/virtio-mmio.c
> > > > @@ -34,6 +34,7 @@
> > > >    #include "qemu/error-report.h"
> > > >    #include "qemu/log.h"
> > > >    #include "trace.h"
> > > > +#include "qapi/error.h"
> > > >    static bool virtio_mmio_ioeventfd_enabled(DeviceState *d)
> > > >    {
> > > > @@ -619,7 +620,7 @@ static int virtio_mmio_load_extra_state(DeviceState *opaque, QEMUFile *f)
> > > >    {
> > > >        VirtIOMMIOProxy *proxy = VIRTIO_MMIO(opaque);
> > > > -    return vmstate_load_state(f, &vmstate_virtio_mmio, proxy, 1);
> > > > +    return vmstate_load_state(f, &vmstate_virtio_mmio, proxy, 1, &error_fatal);
> > > >    }
> > > >    static bool virtio_mmio_has_extra_state(DeviceState *opaque)
> > > > diff --git a/hw/virtio/virtio-pci.c b/hw/virtio/virtio-pci.c
> > > > index 767216d795998708f5716a23ae16c79cd90ff489..b04faa1e5c91b5cef40e54ec41d92422d16bfc13 100644
> > > > --- a/hw/virtio/virtio-pci.c
> > > > +++ b/hw/virtio/virtio-pci.c
> > > > @@ -161,7 +161,7 @@ static int virtio_pci_load_extra_state(DeviceState *d, QEMUFile *f)
> > > >    {
> > > >        VirtIOPCIProxy *proxy = to_virtio_pci_proxy(d);
> > > > -    return vmstate_load_state(f, &vmstate_virtio_pci, proxy, 1);
> > > > +    return vmstate_load_state(f, &vmstate_virtio_pci, proxy, 1, &error_fatal);
> > > >    }
> > > >    static void virtio_pci_save_queue(DeviceState *d, int n, QEMUFile *f)
> > > > diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
> > > > index 9a81ad912e013fc254899c4e55cff1f76a6112a4..419f3516c9f0a3df43831eebc77aa1e6dfcd0f41 100644
> > > > --- a/hw/virtio/virtio.c
> > > > +++ b/hw/virtio/virtio.c
> > > > @@ -3235,6 +3235,7 @@ virtio_load(VirtIODevice *vdev, QEMUFile *f, int version_id)
> > > >        BusState *qbus = qdev_get_parent_bus(DEVICE(vdev));
> > > >        VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
> > > >        VirtioDeviceClass *vdc = VIRTIO_DEVICE_GET_CLASS(vdev);
> > > > +    Error *local_err = NULL;
> > > >        /*
> > > >         * We poison the endianness to ensure it does not get used before
> > > > @@ -3327,15 +3328,17 @@ virtio_load(VirtIODevice *vdev, QEMUFile *f, int version_id)
> > > >        }
> > > >        if (vdc->vmsd) {
> > > > -        ret = vmstate_load_state(f, vdc->vmsd, vdev, version_id);
> > > > +        ret = vmstate_load_state(f, vdc->vmsd, vdev, version_id, &local_err);
> > > >            if (ret) {
> > > > +            warn_report_err(local_err);
> > > 
> > > I think error_report_err() should be used instead because this is replacing
> > > error_report().
> > > 
> > > If I read correctly, warn_report_err() is essentially error_report_err()
> > > prefixing the message with "warning: ". However, the error messages handled
> > > in this series do not have such prefixes so you shouldn't add them in
> > > intemediate changes.
> > 
> > If I am not wrong, we have introduced warn_report_err() not as a replacement to
> > error_report(). They are newly added ones. Besides, some of them are not intermediate
> > changes, because they do end up in the final version. for example in hw/virtio/virtio-pci.c
> 
> I referred to the fact that the messages that were emitted with
> error_report() calls in vmstate_load_state() are now printed with
> warn_report_err(). That causes behavioral changes and I suspect that they
> are accidental.
> 
> I especially want to know reasoning behind them if it is not accidental and
> some of them are going to remain after applying the entire series.
> 
> > In some places in the intermediate patches a warn_report_err() is added right after
> > error_report(), both to be replaced with error_setg()/error_prepend() later in the series.
> 
> The appropriate choice will depend on how Errors generated by error_setg()
> are going to be reported to the user in that case. Using warn_report_err()
> for them is inappropriate if they are going to reported with
> error_report_err(). If they are going to be handled in some special way,
> keeping the original behavior is a sensible option, and I think
> error_report_err() does that in most cases, if not all.
> 
> > 
> > Do you want to replace warn_report_err() everywhere with error_report_err()?
> 
> Perhaps so, but I may have missed some cases where warn_report_err(), so let
> me share some sort of guideline I have in my mind:
> 
> - Keeping the original behavior is a safe choice.
> 
> If "warning: " was not prepended in the past, leaving messages without the
> prefix is at least not a regression, so reviewers can give Reviewed-by: more
> easily.
> 
> - Refer to Markus' email, which provides distinctions among "ordinary"
> erros, programming errors, and warnings and I see it reasonable:
> https://lore.kernel.org/qemu-devel/87h5yijh3b.fsf@pond.sub.org/
> ("Abuse of warnings for unhandled errors and programming errors")
> 

Yes most of them are going to be reported with error_report_err() in the end.
So keeping the warning prefix does not make sense.
I agree. I shall amend and send a new version. Thank you so much.

> Regards,
> Akihiko Odaki
> 

Regards,
Arun Menon