Re: setting 'ulimit -s' within /etc/system


On 24 Mar 2006 10:55:08 GMT, Casper H.S. Dik <Casper.Dik@xxxxxxx>
wrote:

jsmith665 <jsmith665@xxxxxxxxxxx> writes:

Is there an /etc/system parameter that is equivalent to the value set
by 'ulimit -s' (i.e., maximum size of stack segment) for Solaris 8/9?

The "Solaris Tunable Parameters Reference Manual" references both
'lwp_default_stksize' and 'rpcmod:svc_default_stksize'. From the
description, "Default value of size of stack to be used when a kernel
thread is created," the former would appear to be the desired
parameter.

However, testing with either and both (rebooting the system between
settings) appears to have no effect on the desired setting---'ulimit
-s' reports the default of 8192 regardless of the settings of the two
aforementioned parameters.

Those two parameters are for kernel threads; and they are in bytes.

Please change them back because you may now be wasting tons of memory.

You are interested in the per-process stack; this cannot be set
system wide except through, e.g., /etc/profile or /etc/.login.

Users are free to change that limit in anyway.

Casper

Thanks, for the definitive answer.

We did make note of the fact that both parameters set the kernel
thread stack size (one for normal threads, the other for rpc). But,
in the absence of information to the contrary, either seemed a
possible (if unreasonable) candidate. And, we did notice the spike in
memory utilization, as well. So, not having the desired effect, the
parameters have already been returned to their pre-testing values. We
will revert to our original plan, of adding the 'ulimit -s' command to
the process scripts requiring the change.

Thanks, again.
.

Relevant Pages

  • Re: setting ulimit -s within /etc/system
    ... The "Solaris Tunable Parameters Reference Manual" references both ... "Default value of size of stack to be used when a kernel ... -s' reports the default of 8192 regardless of the settings of the two ...
    (comp.unix.solaris)
  • [Full-disclosure] PHRACK 64: ATTACKING THE CORE
    ... - The Slab Allocator ... - Slab overflow exploiting: ... - Forcing a kernel path to sleep ... - Stack Frame Flow Recovery ...
    (Full-Disclosure)
  • Re: [Bug #11342] Linux 2.6.27-rc3: kernel BUG at mm/vmalloc.c - bisected
    ... stack usages for you is that they contain a 'cpumask_t' on the stack. ... We can enable MAXSMP and raise the CPU limits some time in the future. ... not accept a specially built kernel, but only a kernel that has been ... know how extensively these distributions test and certify for many known ...
    (Linux-Kernel)
  • Re: Interrupt context...
    ... > gone through most of the posts on interrupt in usenet. ... > kernel stack and ISR is executed. ... More may be saved depending on the architecture. ... Here the kernel have assembler code to save all general ...
    (comp.os.linux.development.system)
  • Re: The Linux Staging tree, what it is and is not.
    ... if you compromize them and write the wrong ones to the stack, ... Load two user space fieldbus stacks on the same network card (remembering ... right, but that's the same, no matter if it runs in kernel or user ... need an objdict API to userspace...? ...
    (Linux-Kernel)