|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] Re: [PATCH v4 36/49] monitor: tighten monitor_printf*()
Hi
On Wed, Aug 26, 2026 at 2:04 PM Daniel P. Berrangé <berrange@xxxxxxxxxx> wrote:
>
> On Wed, Aug 26, 2026 at 01:41:55PM +0400, Marc-André Lureau wrote:
> > Hi
> >
> > On Wed, Aug 26, 2026 at 12:54 PM Daniel P. Berrangé <berrange@xxxxxxxxxx>
> > wrote:
> > >
> > > On Tue, Aug 25, 2026 at 11:09:43PM +0400, Marc-André Lureau wrote:
> > > > Rename monitor_printf->monitor_hmp_printf, monitor_vprintf->
> > > > monitor_hmp_vprintf, and monitor_printc->monitor_hmp_printc, changing
> > > > the first parameter from Monitor * to MonitorHMP * to enforce type
> > > > safety. The implementation is also simplified: monitor_hmp_vprintf now
> > > > directly calls g_strdup_vprintf + monitor_puts, removing the virtual
> > > > dispatch via moncls->vprintf.
> > > >
> > > > The dev_print() callbacks are temporarily using the MONITOR_HMP(mon)
> > > > cast, they are fixed in the following commits.
> > > >
> > > > Signed-off-by: Marc-André Lureau <marcandre.lureau@xxxxxxxxxx>
> > > > ---
> > > > audio/audio-hmp-cmds.c | 6 +-
> > > > backends/cryptodev-hmp-cmds.c | 9 +-
> > > > block/monitor/block-hmp-cmds.c | 170 +++++++++---------
> > > > chardev/char-hmp-cmds.c | 12 +-
> > > > disas/disas-mon.c | 10 +-
> > > > docs/devel/style.rst | 2 +-
> > > > docs/devel/writing-monitor-commands.rst | 8 +-
> > > > dump/dump-hmp-cmds.c | 5 +-
> > > > hw/char/virtio-serial-bus.c | 10 +-
> > > > hw/core/machine-hmp-cmds.c | 213 +++++++++++-----------
> > > > hw/core/sysbus.c | 5 +-
> > > > hw/hexagon/hexagon_tlb.c | 44 ++---
> > > > hw/i386/kvm/xen-stubs.c | 6 +-
> > > > hw/i386/kvm/xen_evtchn.c | 21 +--
> > > > hw/i386/sgx-hmp-stub.c | 3 +-
> > > > hw/i386/sgx.c | 29 ++-
> > > > hw/misc/auxbus.c | 9 +-
> > > > hw/misc/mos6522-stub.c | 3 +-
> > > > hw/net/rocker/rocker-hmp-cmds.c | 146 ++++++++--------
> > > > hw/pci/pci-hmp-cmds.c | 114 ++++++------
> > > > hw/pci/pci-stub.c | 3 +-
> > > > hw/s390x/s390-skeys.c | 9 +-
> > > > hw/s390x/s390-stattrib.c | 20 +--
> > > > hw/uefi/ovmf-log.c | 5 +-
> > > > hw/usb/bus.c | 11 +-
> > > > hw/usb/host-libusb.c | 21 ++-
> > > > hw/virtio/virtio-hmp-cmds.c | 297
> > > > ++++++++++++++++---------------
> > > > hw/xen/xen-bus.c | 5 +-
> > > > include/disas/disas.h | 4 +-
> > > > include/monitor/hmp.h | 12 +-
> > > > migration/dirtyrate.c | 46 +++--
> > > > migration/migration-hmp-cmds.c | 301
> > > > ++++++++++++++++----------------
> > > > monitor/hmp-cmds.c | 141 +++++++--------
> > > > monitor/hmp.c | 143 +++++++--------
> > > > monitor/monitor-internal.h | 6 -
> > > > monitor/monitor.c | 33 ++--
> > > > net/net-hmp-cmds.c | 31 ++--
> > > > net/slirp.c | 31 ++--
> > > > qom/qom-hmp-cmds.c | 27 ++-
> > > > replay/replay-debugging.c | 5 +-
> > > > stats/stats-hmp-cmds.c | 57 +++---
> > > > stubs/hmp-cmd-info_sev.c | 3 +-
> > > > stubs/monitor-core.c | 2 +-
> > > > system/dirtylimit-hmp-cmds.c | 10 +-
> > > > system/qdev-monitor.c | 19 +-
> > > > system/runstate-hmp-cmds.c | 16 +-
> > > > system/tpm-hmp-cmds.c | 29 ++-
> > > > target/i386/cpu-apic.c | 3 +-
> > > > target/i386/monitor.c | 152 ++++++++--------
> > > > target/i386/sev.c | 35 ++--
> > > > target/m68k/monitor.c | 3 +-
> > > > target/ppc/monitor.c | 3 +-
> > > > target/riscv/monitor.c | 55 +++---
> > > > target/sh4/monitor.c | 29 ++-
> > > > target/sparc/monitor.c | 3 +-
> > > > target/xtensa/monitor.c | 3 +-
> > > > tests/unit/test-util-sockets.c | 2 +-
> > > > tools/qemu-vnc/clipboard.c | 4 +-
> > > > tools/qemu-vnc/stubs.c | 2 +-
> > > > trace/trace-hmp-cmds.c | 12 +-
> > > > ui/ui-hmp-cmds.c | 115 ++++++------
> > > > util/error-report.c | 2 +-
> > > > util/qemu-print.c | 11 +-
> > > > 63 files changed, 1219 insertions(+), 1327 deletions(-)
> > >
> > >
> > >
> > >
> > > > diff --git a/util/qemu-print.c b/util/qemu-print.c
> > > > index 5d4143d425a1..5938f2b6c338 100644
> > > > --- a/util/qemu-print.c
> > > > +++ b/util/qemu-print.c
> > > > @@ -13,6 +13,7 @@
> > > > #include "qemu/osdep.h"
> > > > #include "monitor/monitor.h"
> > > > #include "monitor/hmp.h"
> > > > +#include "qom/object.h"
> > > > #include "qemu/qemu-print.h"
> > > >
> > > > /*
> > > > @@ -23,8 +24,13 @@
> > > > int qemu_vprintf(const char *fmt, va_list ap)
> > > > {
> > > > Monitor *cur_mon = monitor_cur();
> > > > +
> > > > + /* for all monitors: QMP & HMP */
> > > > if (cur_mon) {
> > > > - return monitor_vprintf(cur_mon, fmt, ap);
> > > > + /* don't use monitor_cur_hmp(), to avoid a second lookup */
> > > > + MonitorHMP *hmp = (MonitorHMP *)
> > > > + object_dynamic_cast(OBJECT(cur_mon), TYPE_MONITOR_HMP);
> > > > + return monitor_hmp_vprintf(hmp, fmt, ap);
> > >
> > > This isn't the same semantics AFAICT.
> > >
> > > Original code, if monitor_cur() == QMP, we call monitor_vprintf()
> > > which will return -1.
> > >
> > > New code, if monitor_cur() == QMP, we will get a NULL back from
> > > object_dynamic_cast which we then pass into monitor_hmp_vprintf
> > > which will then crash on monitor_puts() IIUC.
> >
> > monitor_hmp_vprintf() has an early return, if given NULL monitor, it
> > returns -1.
>
> Hmm, I feel like we should be dealing with NULL in this method, as it
> is surprising to be calling a monitor_hmp_XXX method in scenario where
> QMP is a (theoretical) possibility.
"hmp" would be NULL in that case, but I agree it's a bit confusing.
I would go as far and make hmp required for the call, but I can't
review all the code paths to monitor_hmp_(v)printf()
I'll add an early return in qemu-print.c
>
>
> With regards,
> Daniel
> --
> |: https://berrange.com ~~ https://hachyderm.io/@berrange :|
> |: https://libvirt.org ~~ https://entangle-photo.org :|
> |: https://pixelfed.art/berrange ~~ https://fstop138.berrange.com :|
>
|
 |
Lists.xenproject.org is hosted with RackSpace, monitoring our |