[PATCH V2 1/2] plugins: Fix resource leak in connect_socket()

[PATCH V2 1/2] plugins: Fix resource leak in connect_socket()

[AlexChen]

Close the fd when the connect() fails.

Reported-by: Euler Robot <euler.robot@huawei.com>
Signed-off-by: Alex Chen <alex.chen@huawei.com>
---
 contrib/plugins/lockstep.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/contrib/plugins/lockstep.c b/contrib/plugins/lockstep.c
index a696673dff..319bd44b83 100644
--- a/contrib/plugins/lockstep.c
+++ b/contrib/plugins/lockstep.c
@@ -292,6 +292,7 @@ static bool connect_socket(const char *path)

     if (connect(fd, (struct sockaddr *)&sockaddr, sizeof(sockaddr)) < 0) {
         perror("failed to connect");
+        close(fd);
         return false;
     }

-- 
2.19.1

Re: [PATCH V2 1/2] plugins: Fix resource leak in connect_socket()

[Eric Blake]

On 11/5/20 7:59 PM, AlexChen wrote:
> Close the fd when the connect() fails.
> 
> Reported-by: Euler Robot <euler.robot@huawei.com>
> Signed-off-by: Alex Chen <alex.chen@huawei.com>

Your From: line ("AlexChen") is spelled differently than your S-o-b:
line ("Alex Chen").  While this is not fatal to the patch, it is
confusing, so you may want to update your git settings to produce mail
spelled in the same manner as the S-o-b.

Also, although you did manage to send a 0/2 letter, you did not thread
things:
0/2 Message-ID: <5FA4AE0B.1000007@huawei.com>
1/2 Message-ID: Message-ID: <5FA4AE11.6060701@huawei.com>, but no
In-Reply-To: or References: headers, which means it is a new top-level
thread.  You may want to figure out why your mail setup is not
preserving threading.


-- 
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org

Re: [PATCH V2 1/2] plugins: Fix resource leak in connect_socket()

[AlexChen]

On 2020/11/6 21:17, Eric Blake wrote:
> On 11/5/20 7:59 PM, AlexChen wrote:
>> Close the fd when the connect() fails.
>>
>> Reported-by: Euler Robot <euler.robot@huawei.com>
>> Signed-off-by: Alex Chen <alex.chen@huawei.com>
> 
> Your From: line ("AlexChen") is spelled differently than your S-o-b:
> line ("Alex Chen").  While this is not fatal to the patch, it is
> confusing, so you may want to update your git settings to produce mail
> spelled in the same manner as the S-o-b.
> 

Hi Eric,

Thanks for you suggestion, I will modify the user.name of git to "Alex Chen".

> Also, although you did manage to send a 0/2 letter, you did not thread
> things:
> 0/2 Message-ID: <5FA4AE0B.1000007@huawei.com>
> 1/2 Message-ID: Message-ID: <5FA4AE11.6060701@huawei.com>, but no
> In-Reply-To: or References: headers, which means it is a new top-level
> thread.  You may want to figure out why your mail setup is not
> preserving threading.
> 

This may be my email settings is wrong, I try to modify the setting and send a patch v3.

Thanks,
Alex

Re: [PATCH V2 1/2] plugins: Fix resource leak in connect_socket()

[Eric Blake]

On 11/9/20 2:20 AM, AlexChen wrote:
> On 2020/11/6 21:17, Eric Blake wrote:
>> On 11/5/20 7:59 PM, AlexChen wrote:
>>> Close the fd when the connect() fails.
>>>
>>> Reported-by: Euler Robot <euler.robot@huawei.com>
>>> Signed-off-by: Alex Chen <alex.chen@huawei.com>
>>
>> Your From: line ("AlexChen") is spelled differently than your S-o-b:
>> line ("Alex Chen").  While this is not fatal to the patch, it is
>> confusing, so you may want to update your git settings to produce mail
>> spelled in the same manner as the S-o-b.
>>
> 
> Hi Eric,
> 
> Thanks for you suggestion, I will modify the user.name of git to "Alex Chen".
> 
>> Also, although you did manage to send a 0/2 letter, you did not thread
>> things:
>> 0/2 Message-ID: <5FA4AE0B.1000007@huawei.com>
>> 1/2 Message-ID: Message-ID: <5FA4AE11.6060701@huawei.com>, but no
>> In-Reply-To: or References: headers, which means it is a new top-level
>> thread.  You may want to figure out why your mail setup is not
>> preserving threading.
>>
> 
> This may be my email settings is wrong, I try to modify the setting and send a patch v3.

Thanks for the rapid turnaround - I can confirm that v3 is indeed nicer
to read.

-- 
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org