[PATCH for 6.2?] Revert "vga: don't abort when adding a duplicate isa-vga device"

[PATCH for 6.2?] Revert "vga: don't abort when adding a duplicate isa-vga device"

[Alex Bennée]

This reverts commit 7852a77f598635a67a222b6c1463c8b46098aed2.

The check is bogus as it ends up finding itself and falling over.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Resolves: https://gitlab.com/qemu-project/qemu.git/-/issues/733
---
 hw/display/vga-isa.c | 10 ----------
 1 file changed, 10 deletions(-)

diff --git a/hw/display/vga-isa.c b/hw/display/vga-isa.c
index 8cea84f2be..90851e730b 100644
--- a/hw/display/vga-isa.c
+++ b/hw/display/vga-isa.c
@@ -33,7 +33,6 @@
 #include "hw/loader.h"
 #include "hw/qdev-properties.h"
 #include "qom/object.h"
-#include "qapi/error.h"
 
 #define TYPE_ISA_VGA "isa-vga"
 OBJECT_DECLARE_SIMPLE_TYPE(ISAVGAState, ISA_VGA)
@@ -62,15 +61,6 @@ static void vga_isa_realizefn(DeviceState *dev, Error **errp)
     MemoryRegion *vga_io_memory;
     const MemoryRegionPortio *vga_ports, *vbe_ports;
 
-    /*
-     * make sure this device is not being added twice, if so
-     * exit without crashing qemu
-     */
-    if (object_resolve_path_type("", TYPE_ISA_VGA, NULL)) {
-        error_setg(errp, "at most one %s device is permitted", TYPE_ISA_VGA);
-        return;
-    }
-
     s->global_vmstate = true;
     vga_common_init(s, OBJECT(dev));
     s->legacy_address_space = isa_address_space(isadev);
-- 
2.30.2

Re: [PATCH for 6.2?] Revert "vga: don't abort when adding a duplicate isa-vga device"

[Daniel P. Berrangé]

On Thu, Dec 02, 2021 at 04:49:29PM +0000, Alex Bennée wrote:
> This reverts commit 7852a77f598635a67a222b6c1463c8b46098aed2.
> 
> The check is bogus as it ends up finding itself and falling over.
> 
> Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
> Resolves: https://gitlab.com/qemu-project/qemu.git/-/issues/733
> ---
>  hw/display/vga-isa.c | 10 ----------
>  1 file changed, 10 deletions(-)

Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

This turns the qemu-system-mips64el CLI scenario from a
pretty error back into an abort. Either way the user
in question isn't getting a running mips VM, so that's
preferrable to blocking valid usage scenarios.

I agree this should go into 6.2, as it a regression
introduced since 6.1, and this revert is low risk
since it is simply deleting lines.

When a PULL is done we should re-open

  https://gitlab.com/qemu-project/qemu/-/issues/44


> 
> diff --git a/hw/display/vga-isa.c b/hw/display/vga-isa.c
> index 8cea84f2be..90851e730b 100644
> --- a/hw/display/vga-isa.c
> +++ b/hw/display/vga-isa.c
> @@ -33,7 +33,6 @@
>  #include "hw/loader.h"
>  #include "hw/qdev-properties.h"
>  #include "qom/object.h"
> -#include "qapi/error.h"
>  
>  #define TYPE_ISA_VGA "isa-vga"
>  OBJECT_DECLARE_SIMPLE_TYPE(ISAVGAState, ISA_VGA)
> @@ -62,15 +61,6 @@ static void vga_isa_realizefn(DeviceState *dev, Error **errp)
>      MemoryRegion *vga_io_memory;
>      const MemoryRegionPortio *vga_ports, *vbe_ports;
>  
> -    /*
> -     * make sure this device is not being added twice, if so
> -     * exit without crashing qemu
> -     */
> -    if (object_resolve_path_type("", TYPE_ISA_VGA, NULL)) {
> -        error_setg(errp, "at most one %s device is permitted", TYPE_ISA_VGA);
> -        return;
> -    }
> -
>      s->global_vmstate = true;
>      vga_common_init(s, OBJECT(dev));
>      s->legacy_address_space = isa_address_space(isadev);
> -- 
> 2.30.2
> 
> 

Regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|

Re: [PATCH for 6.2?] Revert "vga: don't abort when adding a duplicate isa-vga device"

[Cornelia Huck]

On Thu, Dec 02 2021, Alex Bennée <alex.bennee@linaro.org> wrote:

> This reverts commit 7852a77f598635a67a222b6c1463c8b46098aed2.
>
> The check is bogus as it ends up finding itself and falling over.
>
> Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
> Resolves: https://gitlab.com/qemu-project/qemu.git/-/issues/733
> ---
>  hw/display/vga-isa.c | 10 ----------
>  1 file changed, 10 deletions(-)
>
> diff --git a/hw/display/vga-isa.c b/hw/display/vga-isa.c
> index 8cea84f2be..90851e730b 100644
> --- a/hw/display/vga-isa.c
> +++ b/hw/display/vga-isa.c
> @@ -33,7 +33,6 @@
>  #include "hw/loader.h"
>  #include "hw/qdev-properties.h"
>  #include "qom/object.h"
> -#include "qapi/error.h"
>  
>  #define TYPE_ISA_VGA "isa-vga"
>  OBJECT_DECLARE_SIMPLE_TYPE(ISAVGAState, ISA_VGA)
> @@ -62,15 +61,6 @@ static void vga_isa_realizefn(DeviceState *dev, Error **errp)
>      MemoryRegion *vga_io_memory;
>      const MemoryRegionPortio *vga_ports, *vbe_ports;
>  
> -    /*
> -     * make sure this device is not being added twice, if so
> -     * exit without crashing qemu
> -     */
> -    if (object_resolve_path_type("", TYPE_ISA_VGA, NULL)) {

Don't you simply need to invert the check? You should get NULL back if
there is another device already...

> -        error_setg(errp, "at most one %s device is permitted", TYPE_ISA_VGA);
> -        return;
> -    }
> -
>      s->global_vmstate = true;
>      vga_common_init(s, OBJECT(dev));
>      s->legacy_address_space = isa_address_space(isadev);

Re: [PATCH for 6.2?] Revert "vga: don't abort when adding a duplicate isa-vga device"

[Mark Cave-Ayland]

On 02/12/2021 17:14, Cornelia Huck wrote:

> On Thu, Dec 02 2021, Alex Bennée <alex.bennee@linaro.org> wrote:
> 
>> This reverts commit 7852a77f598635a67a222b6c1463c8b46098aed2.
>>
>> The check is bogus as it ends up finding itself and falling over.
>>
>> Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
>> Resolves: https://gitlab.com/qemu-project/qemu.git/-/issues/733
>> ---
>>   hw/display/vga-isa.c | 10 ----------
>>   1 file changed, 10 deletions(-)
>>
>> diff --git a/hw/display/vga-isa.c b/hw/display/vga-isa.c
>> index 8cea84f2be..90851e730b 100644
>> --- a/hw/display/vga-isa.c
>> +++ b/hw/display/vga-isa.c
>> @@ -33,7 +33,6 @@
>>   #include "hw/loader.h"
>>   #include "hw/qdev-properties.h"
>>   #include "qom/object.h"
>> -#include "qapi/error.h"
>>   
>>   #define TYPE_ISA_VGA "isa-vga"
>>   OBJECT_DECLARE_SIMPLE_TYPE(ISAVGAState, ISA_VGA)
>> @@ -62,15 +61,6 @@ static void vga_isa_realizefn(DeviceState *dev, Error **errp)
>>       MemoryRegion *vga_io_memory;
>>       const MemoryRegionPortio *vga_ports, *vbe_ports;
>>   
>> -    /*
>> -     * make sure this device is not being added twice, if so
>> -     * exit without crashing qemu
>> -     */
>> -    if (object_resolve_path_type("", TYPE_ISA_VGA, NULL)) {
> 
> Don't you simply need to invert the check? You should get NULL back if
> there is another device already...

Yes that's correct - the existing check is simply missing a ! at the front. For an 
existing example take a look at the the fw_cfg device:

https://gitlab.com/qemu-project/qemu/-/blob/master/hw/nvram/fw_cfg.c#L1228
https://gitlab.com/qemu-project/qemu/-/blob/master/hw/nvram/fw_cfg.c#L1125

Unfortunately it seems that I also missed this when I added my R-B tag :(


ATB,

Mark.

Re: [PATCH for 6.2?] Revert "vga: don't abort when adding a duplicate isa-vga device"

[Alex Bennée]

Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> writes:

> On 02/12/2021 17:14, Cornelia Huck wrote:
>
>> On Thu, Dec 02 2021, Alex Bennée <alex.bennee@linaro.org> wrote:
>> 
>>> This reverts commit 7852a77f598635a67a222b6c1463c8b46098aed2.
>>>
>>> The check is bogus as it ends up finding itself and falling over.
>>>
>>> Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
>>> Resolves: https://gitlab.com/qemu-project/qemu.git/-/issues/733
>>> ---
>>>   hw/display/vga-isa.c | 10 ----------
>>>   1 file changed, 10 deletions(-)
>>>
>>> diff --git a/hw/display/vga-isa.c b/hw/display/vga-isa.c
>>> index 8cea84f2be..90851e730b 100644
>>> --- a/hw/display/vga-isa.c
>>> +++ b/hw/display/vga-isa.c
>>> @@ -33,7 +33,6 @@
>>>   #include "hw/loader.h"
>>>   #include "hw/qdev-properties.h"
>>>   #include "qom/object.h"
>>> -#include "qapi/error.h"
>>>     #define TYPE_ISA_VGA "isa-vga"
>>>   OBJECT_DECLARE_SIMPLE_TYPE(ISAVGAState, ISA_VGA)
>>> @@ -62,15 +61,6 @@ static void vga_isa_realizefn(DeviceState *dev, Error **errp)
>>>       MemoryRegion *vga_io_memory;
>>>       const MemoryRegionPortio *vga_ports, *vbe_ports;
>>>   -    /*
>>> -     * make sure this device is not being added twice, if so
>>> -     * exit without crashing qemu
>>> -     */
>>> -    if (object_resolve_path_type("", TYPE_ISA_VGA, NULL)) {
>> Don't you simply need to invert the check? You should get NULL back
>> if
>> there is another device already...
>
> Yes that's correct - the existing check is simply missing a ! at the
> front. For an existing example take a look at the the fw_cfg device:
>
> https://gitlab.com/qemu-project/qemu/-/blob/master/hw/nvram/fw_cfg.c#L1228
> https://gitlab.com/qemu-project/qemu/-/blob/master/hw/nvram/fw_cfg.c#L1125
>
> Unfortunately it seems that I also missed this when I added my R-B tag
> :(

Although that doesn't seem to fix the original problem:

  ➜  ./qemu-system-x86_64 -name ELKS -machine isapc -cpu 486,tsc -m 1M -vga std -rtc base=utc -netdev user,id=mynet,hostfwd=tcp:127.0.0.1:8080-10.0.2.15:80,hostfwd=tcp:127.0.0
  .1:2323-10.0.2.15:23 -device ne2k_isa,irq=12,netdev=mynet -drive if=ide,format=raw,file=$HOME/images/qemu.advent/day23/hd32mbr-fat.bin -drive if=ide,format=qcow2,file=$HOME/images/qemu.advent/day23/scratch.qcow2 -serial mon:stdio
  QEMU 6.1.93 monitor - type 'help' for more information
  (qemu) quit
  🕙10:23:50 alex@zen:qemu.git/builds/all  on  plugins/upgrade-insn-tracking-plugin [$!?] took 5s 
  ➜  ./qemu-system-mips64el -M pica61 -device isa-vga
  RAMBlock "vga.vram" already registered, abort!
  fish: “./qemu-system-mips64el -M pica6…” terminated by signal SIGABRT (Abort)
  🕙10:23:54 alex@zen:qemu.git/builds/all  on  plugins/upgrade-insn-tracking-plugin [$!?] [⚡ IOT] 

>
>
> ATB,
>
> Mark.


-- 
Alex Bennée

Re: [PATCH for 6.2?] Revert "vga: don't abort when adding a duplicate isa-vga device"

[Philippe Mathieu-Daudé]

On 12/2/21 17:49, Alex Bennée wrote:
> This reverts commit 7852a77f598635a67a222b6c1463c8b46098aed2.
> 
> The check is bogus as it ends up finding itself and falling over.
> 
> Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
> Resolves: https://gitlab.com/qemu-project/qemu.git/-/issues/733

This link gives 404. The correct one is:
https://gitlab.com/qemu-project/qemu/-/issues/733

> ---
>  hw/display/vga-isa.c | 10 ----------
>  1 file changed, 10 deletions(-)