James said:
While you've made a number of important points:
[...]
Note the reference to the C signal handler definition,
which states:
"If the signal occurs other than as the result of calling the
abort or raise function, the behavior is undefined if the
signal handler calls any function in the standard library
other than the signal function itself (with a first argument
of the signal number corresponding to the signal that caused
the invocation of the handler) or refers to any object with
static storage duration other than by assigning a value to a
static storage duration variable of type volatile
sig_atomic_t. Furthermore, if such a call to the signal
function results in a SIG-ERR return, the value of errno is
indeterminate."
Then, in a footnote, the C standard adds:
"If any signal is generated by an asynchronous signal handier,
the behavior is undefined."
In other words, according to the C standard, any signal
generated by SIGALRM is undefined behavior.
What does undefined behaviour in this matter mean exactly?
That the behavior is undefined. Anything the implementation
does is correct. The implementation can refuse to compile it,
cause a core dump or reformat your hard disk when it is
executes, or ... define it to do something specific.
Posix does define some additional things with regards to
signals, including asynchronous signals. Presumable, so does
any non-Posix system which defined SIGALRM; it wouldn't make
much sense otherwise.
It's been raised properly for me. I don't see the problem as
long as I check in the actual timer function what kind of
signal i got it seems to work nicely.
The fact that you have undefined behavior doesn't mean that it
won't work. In this case, there are two aspects to consider:
-- Posix does define asynchronous signals, and SIGALRM. If
you're on a Posix conformant system (or one that is at least
conformant in this regard: any of the *nix, and I think even
Windows), catching the signal in your handler is defined,
and you have a few more defined options than just writing to
a sig_atomic_t (but not really very many). So it's likely
that that part of your code is defined on your system.
-- Undefined behavior doesn't mean that the code is guaranteed
to fail. According to Posix, for example, it is undefined
behavior to call any of the FILE* functions in a signal
handler. In practice, it's likely to work most of the time;
it all depends on exactly when the signal arrives (with
respect to what other things you're doing).
What you can and cannot do here depends on your system, not on C
or C++. So you really have to take it up in detail in a system
specific newsgroup. Before doing that, however, I would suggest
that you read up on signal handling under Unix. The use of
signal() is NOT recommended; you should be using sigaction()
instead. (And it sounds to me like you actually need threads,
and a dedicated thread to handle the signals. This is the most
common solution under Unix, since it means that you actually
handle the signal in a thread, and not in a signal handler, and
thus have access to literally everything the system offers.)
I don't need more than what this is offering me.
Just setting sig_atomic_t? That won't get you very far.
What do you mean by "be careful"? I believe it could end up in
a stack overflow if the actual timer function takes too much
time. Is this what you are referring to?
No. The fact that you're not allowed to call most functions
from a signal handler. See the Signal Concepts section in the
Posix standard
(
http://www.opengroup.org/onlinepubs/009695399/functions/
xsh_chap02_04.html).
Or better yet, get a good, up to date book about Unix
programming (e.g. "Advanced Programming in the UNIX
Environment", by W. Richard Stevens, ISBN 0-201-56317-7).
Handling signals and other asynchronous events is not trivial,
and requires some understanding.
Alright the software I'm writing here is working on an
embedded Linux platform - not POSIX tho.
Close enough in this regard. I'm pretty sure that Linux gives
all of the Posix guarantees with regards to signals. It may
give some more, but I'm not aware of it.
It's just a customized distribution built on ulibC. Why would
having an independent thread make more sense than just raising
a timer signal SIGALRM? I don't quite see the problem...
The problem is that you're trying to respond asynchronously.
Which means that you're interrupting one activity to do
something else, and that you have to be careful using common
state. Except that in many case, you don't see the common
state; it's hidden somewhere in the system routines. (Malloc is
an obvious example; in order to work, malloc needs to maintain
global state concerning what is and is not being used. And
during a call to malloc, there will be moments when that global
state is inconsistent. If your signal handler is called at one
of those times, and calls malloc, the malloc called from the
signal handler will see an inconsistent state, and likely crash,
or do something else disagreeable.)
For my application it makes sense just doing it this way I
believe -> Trying to keep memory usage minimal as well as not
wasting any CPU ressources (thread running in idle).
Well, I've not really understood what you're trying to do. Most
of the time, when I need a keep alive, I'll be waiting on a
socket or a condition. In such cases, I'll just add a timeout
to the wait, and send the keep alive if I return because of the
timeout.
