diff mbox

[v2,10/10] qdev: Do not let the user try to device_add when it cannot work

Message ID 1383062897-30084-11-git-send-email-armbru@redhat.com
State New
Headers show

Commit Message

Markus Armbruster Oct. 29, 2013, 4:08 p.m. UTC 
From: Markus Armbruster <armbru@redhat.com>

Such devices have always been unavailable and omitted from the list of
available devices shown by device_add help.  Until commit 18b6dad
silently broke the former, setting up nasty traps for unwary users,
like this one:

    $ qemu-system-x86_64 -nodefaults -monitor stdio -display none
    QEMU 1.6.50 monitor - type 'help' for more information
    (qemu) device_add apic
    Segmentation fault (core dumped)

I call that a regression.  Fix it.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
---
 qdev-monitor.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Comments

Marcel Apfelbaum Oct. 30, 2013, 9:45 a.m. UTC | #1 
On Tue, 2013-10-29 at 17:08 +0100, armbru@redhat.com wrote:
> From: Markus Armbruster <armbru@redhat.com>
> 
> Such devices have always been unavailable and omitted from the list of
> available devices shown by device_add help.  Until commit 18b6dad
> silently broke the former, setting up nasty traps for unwary users,
> like this one:
> 
>     $ qemu-system-x86_64 -nodefaults -monitor stdio -display none
>     QEMU 1.6.50 monitor - type 'help' for more information
>     (qemu) device_add apic
>     Segmentation fault (core dumped)
> 
> I call that a regression.  Fix it.
> 
> Signed-off-by: Markus Armbruster <armbru@redhat.com>
> ---
>  qdev-monitor.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/qdev-monitor.c b/qdev-monitor.c
> index 36f6f09..c538fec 100644
> --- a/qdev-monitor.c
> +++ b/qdev-monitor.c
> @@ -477,7 +477,7 @@ DeviceState *qdev_device_add(QemuOpts *opts)
>          }
>      }
>  
> -    if (!obj) {
> +    if (!obj || DEVICE_CLASS(obj)->cannot_instantiate_with_device_add_yet) {
>          qerror_report(QERR_INVALID_PARAMETER_VALUE, "driver", "device type");
>          return NULL;
>      }
Minor comment, if you move the if statement after
 k = DEVICE_CLASS(obj);
you don't need the cast anymore.


Seems OK to me.
Reviewed-by: Marcel Apfelbaum <marcel.a@redhat.com>
Markus Armbruster Oct. 30, 2013, 12:15 p.m. UTC | #2 
Marcel Apfelbaum <marcel.a@redhat.com> writes:

> On Tue, 2013-10-29 at 17:08 +0100, armbru@redhat.com wrote:
>> From: Markus Armbruster <armbru@redhat.com>
>> 
>> Such devices have always been unavailable and omitted from the list of
>> available devices shown by device_add help.  Until commit 18b6dad
>> silently broke the former, setting up nasty traps for unwary users,
>> like this one:
>> 
>>     $ qemu-system-x86_64 -nodefaults -monitor stdio -display none
>>     QEMU 1.6.50 monitor - type 'help' for more information
>>     (qemu) device_add apic
>>     Segmentation fault (core dumped)
>> 
>> I call that a regression.  Fix it.
>> 
>> Signed-off-by: Markus Armbruster <armbru@redhat.com>
>> ---
>>  qdev-monitor.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>> 
>> diff --git a/qdev-monitor.c b/qdev-monitor.c
>> index 36f6f09..c538fec 100644
>> --- a/qdev-monitor.c
>> +++ b/qdev-monitor.c
>> @@ -477,7 +477,7 @@ DeviceState *qdev_device_add(QemuOpts *opts)
>>          }
>>      }
>>  
>> -    if (!obj) {
>> +    if (!obj || DEVICE_CLASS(obj)->cannot_instantiate_with_device_add_yet) {
>>          qerror_report(QERR_INVALID_PARAMETER_VALUE, "driver", "device type");
>>          return NULL;
>>      }
> Minor comment, if you move the if statement after
>  k = DEVICE_CLASS(obj);
> you don't need the cast anymore.

Ignorant question: does DEVICE_CLASS(NULL) work and return NULL?

> Seems OK to me.
> Reviewed-by: Marcel Apfelbaum <marcel.a@redhat.com>

Thanks!
Marcel Apfelbaum Oct. 30, 2013, 2:20 p.m. UTC | #3 
On Wed, 2013-10-30 at 13:15 +0100, Markus Armbruster wrote:
> Marcel Apfelbaum <marcel.a@redhat.com> writes:
> 
> > On Tue, 2013-10-29 at 17:08 +0100, armbru@redhat.com wrote:
> >> From: Markus Armbruster <armbru@redhat.com>
> >> 
> >> Such devices have always been unavailable and omitted from the list of
> >> available devices shown by device_add help.  Until commit 18b6dad
> >> silently broke the former, setting up nasty traps for unwary users,
> >> like this one:
> >> 
> >>     $ qemu-system-x86_64 -nodefaults -monitor stdio -display none
> >>     QEMU 1.6.50 monitor - type 'help' for more information
> >>     (qemu) device_add apic
> >>     Segmentation fault (core dumped)
> >> 
> >> I call that a regression.  Fix it.
> >> 
> >> Signed-off-by: Markus Armbruster <armbru@redhat.com>
> >> ---
> >>  qdev-monitor.c | 2 +-
> >>  1 file changed, 1 insertion(+), 1 deletion(-)
> >> 
> >> diff --git a/qdev-monitor.c b/qdev-monitor.c
> >> index 36f6f09..c538fec 100644
> >> --- a/qdev-monitor.c
> >> +++ b/qdev-monitor.c
> >> @@ -477,7 +477,7 @@ DeviceState *qdev_device_add(QemuOpts *opts)
> >>          }
> >>      }
> >>  
> >> -    if (!obj) {
> >> +    if (!obj || DEVICE_CLASS(obj)->cannot_instantiate_with_device_add_yet) {
> >>          qerror_report(QERR_INVALID_PARAMETER_VALUE, "driver", "device type");
> >>          return NULL;
> >>      }
> > Minor comment, if you move the if statement after
> >  k = DEVICE_CLASS(obj);
> > you don't need the cast anymore.
> 
> Ignorant question: does DEVICE_CLASS(NULL) work and return NULL?
Checked it, yes, it will return NULL.

Marcel

> 
> > Seems OK to me.
> > Reviewed-by: Marcel Apfelbaum <marcel.a@redhat.com>
> 
> Thanks!
>
Markus Armbruster Oct. 30, 2013, 2:49 p.m. UTC | #4 
Marcel Apfelbaum <marcel.a@redhat.com> writes:

> On Wed, 2013-10-30 at 13:15 +0100, Markus Armbruster wrote:
>> Marcel Apfelbaum <marcel.a@redhat.com> writes:
>> 
>> > On Tue, 2013-10-29 at 17:08 +0100, armbru@redhat.com wrote:
>> >> From: Markus Armbruster <armbru@redhat.com>
>> >> 
>> >> Such devices have always been unavailable and omitted from the list of
>> >> available devices shown by device_add help.  Until commit 18b6dad
>> >> silently broke the former, setting up nasty traps for unwary users,
>> >> like this one:
>> >> 
>> >>     $ qemu-system-x86_64 -nodefaults -monitor stdio -display none
>> >>     QEMU 1.6.50 monitor - type 'help' for more information
>> >>     (qemu) device_add apic
>> >>     Segmentation fault (core dumped)
>> >> 
>> >> I call that a regression.  Fix it.
>> >> 
>> >> Signed-off-by: Markus Armbruster <armbru@redhat.com>
>> >> ---
>> >>  qdev-monitor.c | 2 +-
>> >>  1 file changed, 1 insertion(+), 1 deletion(-)
>> >> 
>> >> diff --git a/qdev-monitor.c b/qdev-monitor.c
>> >> index 36f6f09..c538fec 100644
>> >> --- a/qdev-monitor.c
>> >> +++ b/qdev-monitor.c
>> >> @@ -477,7 +477,7 @@ DeviceState *qdev_device_add(QemuOpts *opts)
>> >>          }
>> >>      }
>> >>  
>> >> -    if (!obj) {
>> >> +    if (!obj || DEVICE_CLASS(obj)->cannot_instantiate_with_device_add_yet) {
>> >>          qerror_report(QERR_INVALID_PARAMETER_VALUE, "driver", "device type");
>> >>          return NULL;
>> >>      }
>> > Minor comment, if you move the if statement after
>> >  k = DEVICE_CLASS(obj);
>> > you don't need the cast anymore.
>> 
>> Ignorant question: does DEVICE_CLASS(NULL) work and return NULL?
> Checked it, yes, it will return NULL.

I'll give it a try when I respin.  Thanks!
diff mbox

Patch

diff --git a/qdev-monitor.c b/qdev-monitor.c
index 36f6f09..c538fec 100644
--- a/qdev-monitor.c
+++ b/qdev-monitor.c
@@ -477,7 +477,7 @@  DeviceState *qdev_device_add(QemuOpts *opts)
         }
     }
 
-    if (!obj) {
+    if (!obj || DEVICE_CLASS(obj)->cannot_instantiate_with_device_add_yet) {
         qerror_report(QERR_INVALID_PARAMETER_VALUE, "driver", "device type");
         return NULL;
     }