Re: [PATCH v3 1/1] Fix the coredump when memory backend id conflicts with default_ram_id

[Li Zhang <zhlcindy@gmail.com>]

On Mon, Jun 13, 2022 at 4:19 PM Igor Mammedov <imammedo@redhat.com> wrote:
>
> On Mon, 13 Jun 2022 12:44:02 +0200
> Li Zhang <lizhang@suse.de> wrote:
>
> > When no memory backend is specified in machine options,
> > a default memory device will be added with default_ram_id.
> > However, if a memory backend object is added in QEMU options
> > and id is the same as default_ram_id, a coredump happens.
> >
> > Command line:
> > qemu-system-x86_64 -name guest=vmtest,debug-threads=on \
> > -machine pc-q35-6.0,accel=kvm,usb=off,vmport=off \
> > -smp 16,sockets=16,cores=1,threads=1 \
> > -m 4G \
> > -object memory-backend-ram,id=pc.ram,size=4G \
> > -no-user-config -nodefaults -nographic
> >
> > Stack trace of thread 16903:
> >     #0  0x00007fb109a9318b raise (libc.so.6 + 0x3a18b)
> >     #1  0x00007fb109a94585 abort (libc.so.6 + 0x3b585)
> >     #2  0x0000558c34bc89be error_handle_fatal (qemu-system-x86_64 + 0x9c89be)
> >     #3  0x0000558c34bc8aee error_setv (qemu-system-x86_64 + 0x9c8aee)
> >     #4  0x0000558c34bc8ccf error_setg_internal (qemu-system-x86_64 + 0x9c8ccf)
> >     #5  0x0000558c349f6899 object_property_try_add (qemu-system-x86_64 + 0x7f6899)
> >     #6  0x0000558c349f7df8 object_property_try_add_child (qemu-system-x86_64 + 0x7f7df8)
> >     #7  0x0000558c349f7e91 object_property_add_child (qemu-system-x86_64 + 0x7f7e91)
> >     #8  0x0000558c3454686d create_default_memdev (qemu-system-x86_64 + 0x34686d)
> >     #9  0x0000558c34546f58 qemu_init_board (qemu-system-x86_64 + 0x346f58)
> >     #10 0x0000558c345471b9 qmp_x_exit_preconfig (qemu-system-x86_64 + 0x3471b9)
> >     #11 0x0000558c345497d9 qemu_init (qemu-system-x86_64 + 0x3497d9)
> >     #12 0x0000558c344e54c2 main (qemu-system-x86_64 + 0x2e54c2)
> >     #13 0x00007fb109a7e34d __libc_start_main (libc.so.6 + 0x2534d)
> >     #14 0x0000558c344e53ba _start (qemu-system-x86_64 + 0x2e53ba)
> >
> > Signed-off-by: Li Zhang <lizhang@suse.de>
>
> Acked-by: Igor Mammedov <imammedo@redhat.com>
>
>
> CCing David as he probably would be the one to merge it
>

Thanks for your review.

> > ---
> >  hw/core/machine.c | 12 ++++++++++++
> >  1 file changed, 12 insertions(+)
> >
> > diff --git a/hw/core/machine.c b/hw/core/machine.c
> > index a673302cce..9ede63b01c 100644
> > --- a/hw/core/machine.c
> > +++ b/hw/core/machine.c
> > @@ -1265,9 +1265,21 @@ MemoryRegion *machine_consume_memdev(MachineState *machine,
> >  static bool create_default_memdev(MachineState *ms, const char *path, Error **errp)
> >  {
> >      Object *obj;
> > +    ObjectProperty *prop;
> >      MachineClass *mc = MACHINE_GET_CLASS(ms);
> >      bool r = false;
> >
> > +    prop = object_property_find(object_get_objects_root(), mc->default_ram_id);
> > +    if (prop) {
> > +        error_report("A memory backend with id '%s' already exists, "
> > +                     "cannot create default RAM backend with the same id. "
> > +                     "Either change id of the offending backend or "
> > +                     "provide system RAM backend explicitly using "
> > +                     "'-machine memory-backend' option. " ,
> > +                     mc->default_ram_id);
> > +        exit(EXIT_FAILURE);
> > +    }
> > +
> >      obj = object_new(path ? TYPE_MEMORY_BACKEND_FILE : TYPE_MEMORY_BACKEND_RAM);
> >      if (path) {
> >          if (!object_property_set_str(obj, "mem-path", path, errp)) {
>


-- 

Best Regards
-Li