netdev - Re: [EXTERNAL] Re: [PATCH iproute2-next v2] arpd: create /var/lib/arpd on first use

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date: Mon, 18 Mar 2024 12:37:47 +0300
From: Denis Kirjanov <dkirjanov@...e.de>
To: Max Gautier <mg@....gautier.name>, Ratheesh Kannoth <rkannoth@...vell.com>
Cc: "netdev@...r.kernel.org" <netdev@...r.kernel.org>
Subject: Re: [EXTERNAL] Re: [PATCH iproute2-next v2] arpd: create
 /var/lib/arpd on first use



On 3/18/24 12:26, Max Gautier wrote:
> On Mon, Mar 18, 2024 at 09:18:59AM +0000, Ratheesh Kannoth wrote:
>>>>>>> +	if (strcmp(default_dbname, dbname) == 0
>>>>>>> +			&& mkdir(ARPDDIR, 0755) != 0
>>>>>>> +			&& errno != EEXIST
>>>>>> why do you need errno != EEXIST case ? mkdir() will return error
>>>>>> in this case
>>>>> as well.
>>>>>
>>>>> EEXIST is not an error in this case: if the default location already
>>>>> exist, all is good. mkdir would still return -1 in this case, so we
>>>>> need to exclude it manually.
>>>>
>>>> ACK. IMO, it would make a more readable code if you consider splitting the
>>> "if" loop.
>>>
>>> Something like this ? I tend to pack conditions unless branching is necessary,
>>> but no problem if this form is preferred.
>>>
>>> if (strcmp(default_dbname, dbname) == 0) {
>>>     if (mkdir(ARPDDIR, 0755) != 0 && errno != EEXIST) {
>>>    ...
>>>    }
>>> }
>> ACK.   
>> instead of errno != EXIST ,  you may consider stat() before mkdir() call. Just my way thinking(please ignore it, if you don't like). 
>> My thinking is --> you need to execute mkdir () only first time, second time onwards, stat() call will return 0.
> 
> That's racy: we can stat and have a non existing folder, then have
> another arpd instance (or anything else, really) create the directory,

Agreed ^^

> and we would hit EEXIST anyway when we call mkdir.
> Also, that needs two syscalls instead of one.
>