[v1] accel: Share CPUState accel context (HAX/NVMM/WHPX/HVF)

[PATCH 09/14] accel: Allocate NVMM vCPU using g_try_FOO()

Posted by Philippe Mathieu-Daudé 2 years, 10 months ago

g_malloc0() can not fail. Use g_try_malloc0() instead.

https://developer-old.gnome.org/glib/stable/glib-Memory-Allocation.html#glib-Memory-Allocation.description

Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
---
 target/i386/nvmm/nvmm-all.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/target/i386/nvmm/nvmm-all.c b/target/i386/nvmm/nvmm-all.c
index 3c7bdd560f..45fd318d23 100644
--- a/target/i386/nvmm/nvmm-all.c
+++ b/target/i386/nvmm/nvmm-all.c
@@ -942,7 +942,7 @@ nvmm_init_vcpu(CPUState *cpu)
         }
     }
 
-    qcpu = g_malloc0(sizeof(*qcpu));
+    qcpu = g_try_malloc0(sizeof(*qcpu));
     if (qcpu == NULL) {
         error_report("NVMM: Failed to allocate VCPU context.");
         return -ENOMEM;
-- 
2.38.1

Re: [PATCH 09/14] accel: Allocate NVMM vCPU using g_try_FOO()

Posted by Alex Bennée 2 years, 10 months ago

Philippe Mathieu-Daudé <philmd@linaro.org> writes:

> g_malloc0() can not fail. Use g_try_malloc0() instead.
>
> https://developer-old.gnome.org/glib/stable/glib-Memory-Allocation.html#glib-Memory-Allocation.description
>
> Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
> ---
>  target/i386/nvmm/nvmm-all.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/target/i386/nvmm/nvmm-all.c b/target/i386/nvmm/nvmm-all.c
> index 3c7bdd560f..45fd318d23 100644
> --- a/target/i386/nvmm/nvmm-all.c
> +++ b/target/i386/nvmm/nvmm-all.c
> @@ -942,7 +942,7 @@ nvmm_init_vcpu(CPUState *cpu)
>          }
>      }
>  
> -    qcpu = g_malloc0(sizeof(*qcpu));
> +    qcpu = g_try_malloc0(sizeof(*qcpu));
>      if (qcpu == NULL) {
>          error_report("NVMM: Failed to allocate VCPU context.");
>          return -ENOMEM;

Why - if we fail to allocate the vCPU context its game over anyway any
established QEMU practice is its ok to assert fail on a malloc when
there isn't enough memory. IOW keep the g_malloc0 and remove the error
handling case.

-- 
Alex Bennée
Virtualisation Tech Lead @ Linaro

Re: [PATCH 09/14] accel: Allocate NVMM vCPU using g_try_FOO()

Posted by Philippe Mathieu-Daudé 2 years, 10 months ago

On 5/4/23 15:55, Alex Bennée wrote:
> 
> Philippe Mathieu-Daudé <philmd@linaro.org> writes:
> 
>> g_malloc0() can not fail. Use g_try_malloc0() instead.
>>
>> https://developer-old.gnome.org/glib/stable/glib-Memory-Allocation.html#glib-Memory-Allocation.description
>>
>> Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
>> ---
>>   target/i386/nvmm/nvmm-all.c | 2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/target/i386/nvmm/nvmm-all.c b/target/i386/nvmm/nvmm-all.c
>> index 3c7bdd560f..45fd318d23 100644
>> --- a/target/i386/nvmm/nvmm-all.c
>> +++ b/target/i386/nvmm/nvmm-all.c
>> @@ -942,7 +942,7 @@ nvmm_init_vcpu(CPUState *cpu)
>>           }
>>       }
>>   
>> -    qcpu = g_malloc0(sizeof(*qcpu));
>> +    qcpu = g_try_malloc0(sizeof(*qcpu));
>>       if (qcpu == NULL) {
>>           error_report("NVMM: Failed to allocate VCPU context.");
>>           return -ENOMEM;
> 
> Why - if we fail to allocate the vCPU context its game over anyway any
> established QEMU practice is its ok to assert fail on a malloc when
> there isn't enough memory. IOW keep the g_malloc0 and remove the error
> handling case.

This was my first approach but then I realized the author took care
to warn / return ENOMEM, so I went for _try_; but you are right,
since this is "game over" let's simply remove the check.