diff mbox series

util/module: fix a memory leak

Message ID 1576805650-16380-1-git-send-email-pannengyuan@huawei.com (mailing list archive)
State New, archived
Headers show
Series util/module: fix a memory leak | expand

Commit Message

Pan Nengyuan Dec. 20, 2019, 1:34 a.m. UTC
From: Pan Nengyuan <pannengyuan@huawei.com>

spotted by ASAN

Fixes: 81d8ccb1bea4fb9eaaf4c8e30bd4021180a9a39f
Reported-by: Euler Robot <euler.robot@huawei.com>
Signed-off-by: Pan Nengyuan <pannengyuan@huawei.com>
---
 util/module.c | 1 +
 1 file changed, 1 insertion(+)

Comments

Markus Armbruster Dec. 20, 2019, 7:37 a.m. UTC | #1
Cc: qemu-trivial

<pannengyuan@huawei.com> writes:

> From: Pan Nengyuan <pannengyuan@huawei.com>
>
> spotted by ASAN
>
> Fixes: 81d8ccb1bea4fb9eaaf4c8e30bd4021180a9a39f
> Reported-by: Euler Robot <euler.robot@huawei.com>
> Signed-off-by: Pan Nengyuan <pannengyuan@huawei.com>
> ---
>  util/module.c | 1 +
>  1 file changed, 1 insertion(+)
>
> diff --git a/util/module.c b/util/module.c
> index e9fe3e5..8c5315a 100644
> --- a/util/module.c
> +++ b/util/module.c
> @@ -214,6 +214,7 @@ bool module_load_one(const char *prefix, const char *lib_name)
>  
>      if (!success) {
>          g_hash_table_remove(loaded_modules, module_name);
> +        g_free(module_name);
>      }
>  
>      for (i = 0; i < n_dirs; i++) {

Reviewed-by: Markus Armbruster <armbru@redhat.com>

Reviewing this made me wonder: @loaded_modules is global state, but
there appears to be no synchronization.  What's the safety argument?
Laurent Vivier Dec. 20, 2019, 7:55 a.m. UTC | #2
Le 20/12/2019 à 08:37, Markus Armbruster a écrit :
> Cc: qemu-trivial
> 
> <pannengyuan@huawei.com> writes:
> 
>> From: Pan Nengyuan <pannengyuan@huawei.com>
>>
>> spotted by ASAN
>>
>> Fixes: 81d8ccb1bea4fb9eaaf4c8e30bd4021180a9a39f
>> Reported-by: Euler Robot <euler.robot@huawei.com>
>> Signed-off-by: Pan Nengyuan <pannengyuan@huawei.com>
>> ---
>>  util/module.c | 1 +
>>  1 file changed, 1 insertion(+)
>>
>> diff --git a/util/module.c b/util/module.c
>> index e9fe3e5..8c5315a 100644
>> --- a/util/module.c
>> +++ b/util/module.c
>> @@ -214,6 +214,7 @@ bool module_load_one(const char *prefix, const char *lib_name)
>>  
>>      if (!success) {
>>          g_hash_table_remove(loaded_modules, module_name);
>> +        g_free(module_name);
>>      }
>>  
>>      for (i = 0; i < n_dirs; i++) {
> 
> Reviewed-by: Markus Armbruster <armbru@redhat.com>
> 
> Reviewing this made me wonder: @loaded_modules is global state, but
> there appears to be no synchronization.  What's the safety argument?
> 
> 

Applied to my trivial-patches branch.

Thanks,
Laurent
diff mbox series

Patch

diff --git a/util/module.c b/util/module.c
index e9fe3e5..8c5315a 100644
--- a/util/module.c
+++ b/util/module.c
@@ -214,6 +214,7 @@  bool module_load_one(const char *prefix, const char *lib_name)
 
     if (!success) {
         g_hash_table_remove(loaded_modules, module_name);
+        g_free(module_name);
     }
 
     for (i = 0; i < n_dirs; i++) {