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: [RFC] Add TIMEOUT environment variable support to all tests, usable via <test>-ENV= in Makefile.


On Thu, Jan 08, 2004 at 01:04:38PM -0800, Ulrich Drepper wrote:
> > As such I don't want a linear scaling factor to *all* my timeouts,
> 
> To achieve the test you want use some real hardware, not junk.  I'm not
> interested in the least in a change different from what I outlined.  Why
> should I miss a slowdown on reasonable hardware just because you choose
> to use junk?

You wouldn't miss a slowdown. Your tests still run at the same timeout.

o Single test architecture slowdowns are handled with:
  sysdeps/<arch>/Makefile => xxxxx-ENV = TIMEOUT=YYYY

o Scaling the *entire* testsuite timeout defaults are done with
  a global environment variable. Such is suitable if you decide
  to run on a box that is 2x faster, or 4x slower, and you want 
  to scale the timeouts accordingly.
 
http://www.hp.com/products1/servers/entry_level/index.html
For future reference and timing considerations, it's an rp2430.
                                                                                                
Linux's idea of user<->page mappings doesn't match the architectures
hardware implementation, thus munmap is slow because we need to flush
all the aliased address mappings.

c.


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