Reinhard Tartler wrote:
> Glenn Saberton <gsaber...@foomagic.org> writes:
>
>> Reinhard Tartler wrote:
>>> Glenn Saberton <gsaber...@foomagic.org> writes:
>>>
>>>> +static int start_wpa_daemon(struct debconfclient *client)
>>>> +{
>>>> + wpa_supplicant_pid = fork();
>>>> +
>>>> + if (wpa_supplicant_pid == 0) {
>>>> + fclose(client->out);
>>>> + if (execlp("wpa_supplicant", "wpa_supplicant", "-i", interface,
>>>> "-C",
>>>> + WPASUPP_CTRL, "-P", WPAPID, "-B", NULL) == -1) {
>>>> + di_error("could not exec wpasupplicant: %s", strerror(errno));
>>>> + return 1;
>>>> + }
>>>> + else
>>>> + return 0;
>>>> + }
>>>> + else {
>>>> + waitpid(wpa_supplicant_pid, NULL, 0);
>>>> + return 0;
>>>> + }
>>>> +}
>>> this looks fishy. Are you sure you want to return if the child has
>>> failed to exec?
>>>
>> Can you explain what you mean by looks fishy? Suggestions and
>> improvements are more than welcome, and I still have a few things I want
>> to do yet myself. I'm a bit of a C novice, so I am more than likely to
>> have done a few things the less than perfect way, but I'm not sure what
>> you mean by the above comment. As is, if the exec fails, then we return
>> back, which could really do with an error template as well. I think I
>> still need to do a bit more work on the error handling side of things.
>
> fork() clones the current process. The exec then tries to replace the
> "child" process with a new one, in this case wpasupplicant. If
> wpasupplicant can be executed and the exec call succeeds, it never
> returns.
>
> However if exec fails, you generally don't want to continue execution,
> but exit with an error code.
>
> If done like you did here, you'll have 2 netcfg processes in the
> system. From the first look, it doesn't look that this is intended.
>
>
>
Ah, I see what you mean. I hadn't thought of that. Thanks!
Cheers
Glenn
--
To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
