[Qemu-devel] [PATCH v2] e500: fix memory leak

[Qemu-devel] [PATCH v2] e500: fix memory leak

[arei.gonglei]

From: Gonglei <arei.gonglei@huawei.com>

Signed-off-by: Gonglei <arei.gonglei@huawei.com>
---
v2: fix compilation complaint. (mjt)
---
 hw/ppc/e500.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/hw/ppc/e500.c b/hw/ppc/e500.c
index 7e17d18..c060b50 100644
--- a/hw/ppc/e500.c
+++ b/hw/ppc/e500.c
@@ -309,8 +309,10 @@ static int ppce500_load_device_tree(MachineState *machine,
 
         fdt = load_device_tree(filename, &fdt_size);
         if (!fdt) {
+            g_free(filename);
             goto out;
         }
+        g_free(filename);
         goto done;
     }
 
-- 
1.7.12.4

Re: [Qemu-devel] [PATCH v2] e500: fix memory leak

[Michael Tokarev]

28.02.2015 13:08, arei.gonglei@huawei.com пишет:
> From: Gonglei <arei.gonglei@huawei.com>
> 
> Signed-off-by: Gonglei <arei.gonglei@huawei.com>
> ---
> v2: fix compilation complaint. (mjt)
> ---
>  hw/ppc/e500.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/hw/ppc/e500.c b/hw/ppc/e500.c
> index 7e17d18..c060b50 100644
> --- a/hw/ppc/e500.c
> +++ b/hw/ppc/e500.c
> @@ -309,8 +309,10 @@ static int ppce500_load_device_tree(MachineState *machine,
>  
>          fdt = load_device_tree(filename, &fdt_size);
>          if (!fdt) {
> +            g_free(filename);
>              goto out;
>          }
> +        g_free(filename);
>          goto done;
>      }


How about this?

--- a/hw/ppc/e500.c
+++ b/hw/ppc/e500.c
@@ -308,6 +308,7 @@ static int ppce500_load_device_tree(MachineState *machine,
         }

         fdt = load_device_tree(filename, &fdt_size);
+        g_free(filename);
         if (!fdt) {
             goto out;
         }

Thanks,

/mjt

Re: [Qemu-devel] [PATCH v2] e500: fix memory leak

[Gonglei]

On 2015/2/28 18:13, Michael Tokarev wrote:
> 28.02.2015 13:08, arei.gonglei@huawei.com пишет:
>> From: Gonglei <arei.gonglei@huawei.com>
>>
>> Signed-off-by: Gonglei <arei.gonglei@huawei.com>
>> ---
>> v2: fix compilation complaint. (mjt)
>> ---
>>  hw/ppc/e500.c | 2 ++
>>  1 file changed, 2 insertions(+)
>>
>> diff --git a/hw/ppc/e500.c b/hw/ppc/e500.c
>> index 7e17d18..c060b50 100644
>> --- a/hw/ppc/e500.c
>> +++ b/hw/ppc/e500.c
>> @@ -309,8 +309,10 @@ static int ppce500_load_device_tree(MachineState *machine,
>>  
>>          fdt = load_device_tree(filename, &fdt_size);
>>          if (!fdt) {
>> +            g_free(filename);
>>              goto out;
>>          }
>> +        g_free(filename);
>>          goto done;
>>      }
> 
> 
> How about this?
> 
It's ok on function, but seems oddly, isn't it?

Regards,
-Gonglei
> --- a/hw/ppc/e500.c
> +++ b/hw/ppc/e500.c
> @@ -308,6 +308,7 @@ static int ppce500_load_device_tree(MachineState *machine,
>          }
> 
>          fdt = load_device_tree(filename, &fdt_size);
> +        g_free(filename);
>          if (!fdt) {
>              goto out;
>          }
> 
> Thanks,
> 
> /mjt
> 

Re: [Qemu-devel] [PATCH v2] e500: fix memory leak

[Michael Tokarev]

28.02.2015 13:18, Gonglei wrote:
> On 2015/2/28 18:13, Michael Tokarev wrote:
[]
>> How about this?
>>
> It's ok on function, but seems oddly, isn't it?

I don't see anything odd in it.  The `filename' variable is only
used as an argument for load_device_tree() function and is not used
anywhwere else, so it's logical to free it after use.

>> --- a/hw/ppc/e500.c
>> +++ b/hw/ppc/e500.c
>> @@ -308,6 +308,7 @@ static int ppce500_load_device_tree(MachineState *machine,
>>          }
>>
>>          fdt = load_device_tree(filename, &fdt_size);
>> +        g_free(filename);
>>          if (!fdt) {
>>              goto out;
>>          }
>>
Thanks,

/mjt

Re: [Qemu-devel] [PATCH v2] e500: fix memory leak

[Paolo Bonzini]

On 28/02/2015 11:13, Michael Tokarev wrote:
> 28.02.2015 13:08, arei.gonglei@huawei.com пишет:
>> From: Gonglei <arei.gonglei@huawei.com>
>>
>> Signed-off-by: Gonglei <arei.gonglei@huawei.com>
>> ---
>> v2: fix compilation complaint. (mjt)
>> ---
>>  hw/ppc/e500.c | 2 ++
>>  1 file changed, 2 insertions(+)
>>
>> diff --git a/hw/ppc/e500.c b/hw/ppc/e500.c
>> index 7e17d18..c060b50 100644
>> --- a/hw/ppc/e500.c
>> +++ b/hw/ppc/e500.c
>> @@ -309,8 +309,10 @@ static int ppce500_load_device_tree(MachineState *machine,
>>  
>>          fdt = load_device_tree(filename, &fdt_size);
>>          if (!fdt) {
>> +            g_free(filename);
>>              goto out;
>>          }
>> +        g_free(filename);
>>          goto done;
>>      }
> 
> 
> How about this?
> 
> --- a/hw/ppc/e500.c
> +++ b/hw/ppc/e500.c
> @@ -308,6 +308,7 @@ static int ppce500_load_device_tree(MachineState *machine,
>          }
> 
>          fdt = load_device_tree(filename, &fdt_size);
> +        g_free(filename);
>          if (!fdt) {
>              goto out;
>          }

Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>

See also xilinx_load_device_tree:

        /* Try the local "ppc.dtb" override.  */
        fdt = load_device_tree("ppc.dtb", &fdt_size);
        if (!fdt) {
            path = qemu_find_file(QEMU_FILE_TYPE_BIOS, BINARY_DEVICE_TREE_FILE);
            if (path) {
                fdt = load_device_tree(path, &fdt_size);
                g_free(path);
            }
        }

Paolo

Re: [Qemu-devel] [PATCH v2] e500: fix memory leak

[Stefan Weil]

Am 28.02.2015 um 17:40 schrieb Paolo Bonzini:
> On 28/02/2015 11:13, Michael Tokarev wrote:
>> 28.02.2015 13:08, arei.gonglei@huawei.com пишет:
>>> From: Gonglei <arei.gonglei@huawei.com>
>>>
>>> Signed-off-by: Gonglei <arei.gonglei@huawei.com>
>>> ---
>>> v2: fix compilation complaint. (mjt)
>>> ---
>>>   hw/ppc/e500.c | 2 ++
>>>   1 file changed, 2 insertions(+)
>>>
>>> diff --git a/hw/ppc/e500.c b/hw/ppc/e500.c
>>> index 7e17d18..c060b50 100644
>>> --- a/hw/ppc/e500.c
>>> +++ b/hw/ppc/e500.c
>>> @@ -309,8 +309,10 @@ static int ppce500_load_device_tree(MachineState *machine,
>>>   
>>>           fdt = load_device_tree(filename, &fdt_size);
>>>           if (!fdt) {
>>> +            g_free(filename);
>>>               goto out;
>>>           }
>>> +        g_free(filename);
>>>           goto done;
>>>       }
>>
>> How about this?
>>
>> --- a/hw/ppc/e500.c
>> +++ b/hw/ppc/e500.c
>> @@ -308,6 +308,7 @@ static int ppce500_load_device_tree(MachineState *machine,
>>           }
>>
>>           fdt = load_device_tree(filename, &fdt_size);
>> +        g_free(filename);
>>           if (!fdt) {
>>               goto out;
>>           }
> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>

Which one of the two patch variants did you review?

I suggest using Michael's version (only one g_free as early as 
possible). For that variant, my review may be added:

Reviewed-by: Stefan Weil <sw@weilnetz.de>

Re: [Qemu-devel] [PATCH v2] e500: fix memory leak

[Paolo Bonzini]

On 28/02/2015 17:47, Stefan Weil wrote:
>> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
> 
> Which one of the two patch variants did you review?
> 
> I suggest using Michael's version (only one g_free as early as
> possible). For that variant, my review may be added:
> 
> Reviewed-by: Stefan Weil <sw@weilnetz.de>

I also referred to Michael's.

Paolo

Re: [Qemu-devel] [PATCH v2] e500: fix memory leak

[Gonglei]

On 2015/3/1 18:06, Paolo Bonzini wrote:
> 
> 
> On 28/02/2015 17:47, Stefan Weil wrote:
>>> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
>>
>> Which one of the two patch variants did you review?
>>
>> I suggest using Michael's version (only one g_free as early as
>> possible). For that variant, my review may be added:
>>
>> Reviewed-by: Stefan Weil <sw@weilnetz.de>
> 
> I also referred to Michael's.
> 
OK, thanks.

Michael, need I send another version? Or you own handle it directly?

Regards,
-Gonglei