[aarch64-port-dev ] 8173339: AArch64: Fix minimum stack size computations

[Lindenmaier, Goetz]

Hi Andrew, 

You need to adapt these for aarch in os_cpu_linux_aarch.cpp
size_t os::Posix::_compiler_thread_min_stack_allowed = 32 * K;
size_t os::Posix::_java_thread_min_stack_allowed = 32 * K;
size_t os::Posix::_vm_internal_thread_min_stack_allowed = 64 * K;

See also http://hg.openjdk.java.net/jdk9/hs/hotspot/rev/c7a256349729
Try with 
-XX:ThreadStackSize
-XX:CompilerThreadStackSize
-XX:VMThreadStackSize
You probably need to increase the values by 40K as you pass 320 but need 360K.
Guard pages are added to these values.

Best regards,
  Goetz

> -----Original Message-----
> From: Andrew Dinn [mailto:adinn at redhat.com]
> Sent: Donnerstag, 26. Januar 2017 22:18
> To: Lindenmaier, Goetz <goetz.lindenmaier at sap.com>
> Cc: aarch64-port-dev at openjdk.java.net; jdk9-dev at openjdk.java.net
> Subject: Re: 8173339: AArch64: Fix minimum stack size computations
> 
> On 26/01/17 19:05, Lindenmaier, Goetz wrote:
> > Hi Andrew,
> >
> > Yes this should go to 9. Looks Good.
> >
> > Does the TooSmallStackSize Test pass?
> > It might be affected by my previous changes, too.
> 
> Yes, I am afraid it does indeed fail.
> 
> When passed -Xss16K the VM reports that it needs a minimum stack size of
> 320K. However running the program fails with a StackOverflow when you
> provide that value.
> 
> n.b. execution does work with a stack size of 360K. Do you have an
> explanation as to why the value reported by -Xss hits a StackOverflow
> (while 360K does not)?
> 
> regards,
> 
> 
> Andrew Dinn
> -----------
> Senior Principal Software Engineer
> Red Hat UK Ltd
> Registered in England and Wales under Company Registration No. 03798903
> Directors: Michael Cunningham, Michael ("Mike") O'Neill, Eric Shander