This is the mail archive of the libc-alpha@sources.redhat.com mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: malloc question


Thank you for the answer and sorry for the off-topic: I thought this may
have been a bug in malloc().

Best regards,
Sebastien.


Le lun 26/04/2004 à 19:16, Robert Love a écrit :
> On Mon, 2004-04-26 at 18:56 +0200, Sebastien Decugis wrote:
> 
> > The problem is that my process is killed when no memory is left. Is that
> > a wanted behavior? The posix standard requires the function to return
> > NULL and set errno to ENOMEM. Is this behavior dependent on my linux
> > distribution (RH9)?
> 
> This is the intentional behavior of Linux.
> 
> malloc() does return ENOMEM upon memory failure.  Set your SIZEBLK to a
> larger stride and you should see it.  You just do not see the error in
> your case because the Linux kernel's Out Of Memory (OOM) killer killed
> your process first.
> 
> > A last question: Does anybody knows a way to go near memory limit
> > without being killed?
> 
> If your application has the CAP_SYS_RAWIO capability, it is sent SIGTERM
> instead of SIGKILL.  So you can catch the SIGTERM and respond.
> 
> But the real answer is do not consume all of the memory in the system.
> 
> As Uli would say, this email is off-topic for this list, BTW.
> 
> 	Robert Love
-- 
Sébastien DECUGIS
Bull S.A.
NPTL Tests & Trace project
http://nptl.bullopensource.org


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]