[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Assistance Request

From: Joel Sherrill
Subject: Re: Assistance Request
Date: Tue, 12 Feb 2008 10:35:05 -0600
User-agent: Thunderbird (X11/20071115)

Dave Korn wrote:
On 12 February 2008 15:48, Joel Sherrill wrote:

 From the log I see this which would seem to indicate that I am
not getting any compiler flags passed in.

  BTW, although you've already fixed this: I wouldn't expect to see any of the
RUNTESTFLAGS you pass on the commandline ending up in the site.exp in any

I am getting arguments through now.  I am on Fedora 8
and the DejaGNU RPM doesn't seem to support ~/.dejagnurc.

  Ow, that's a bit "quirky and individualistic" :)

I moved my .exp file to the /usr/share/dejagnu/baseboard and
am using this to drive things:

make check-gcc RUNTESTFLAGS="\

--target=powerpc-rtems4.9 \
--target_board=rtems-powerpc-sim --all \

I am sure I won't be able to execute anything correctly yet
because I haven't linked with an RTEMS startup file in
the process.  But things are looking better now that I can
see my settings having an impact.  I hope to post in a little
while that I am running real executable running.

  Looks like it's progressing nicely.
I think so.  I actually got an executable which
I could run but DejaGNU didn't know how to. So that's half the battle. ;)

PASS: gcc.c-torture/execute/builtins/memset-chk.c compilation, -O3 -fomit-frame-pointer -funroll-loops core_find_mapping() - access to unmaped address, attach a default map to handle this - addr=0x7c9c60 nr_bytes=0x4 processor=0x40017008 cia=0x14980

FAIL: gcc.c-torture/execute/builtins/memset-chk.c execution, -O3 -fomit-frame-pointer -funroll-loops

But when I run it by hand...

$ psim-4.9 ./memset-chk.x5

*** EXIT code 0

How can I see what is invoked for the simulator run?
FWIW I just want to run the external script "psim-4.9" which takes
care of everything.  Clearly "setup_sim psim4.9" isn't the right thing.
And the "set_board_info sim "psim-4.9" isn't working either. :(

I have attached rtems-powerpc-sim.exp in case you can spot
something small and stupid.  I am very very close.

Is there an easy way to skip to the executable tests?
I am tired of watching the compile only ones pass. :-D
RTEMS has fixed size task stacks.  Any idea on how much
stack space the test main's will take?

  Most of the tests in the gcc testsuite are pretty lightweight on stack
usage, but there are a few big ones.  I use this in my board.exp:

# We only have a small stack available to us
set_board_info gcc,stack_size 2048

to try and avoid running the excessive ones.
That's a good option.  I added it and hope that Fedora's DejaGNU
knows it.
  I also have an old patch kicking around that adds a whole load of extra
"set_board_info no_XXXXX" options that are useful in small embedded systems
where you don't have a full runtime, stdio, signals, etc. etc.; it cuts out a
whole load of the noise from the list of FAILs.  Let me know if you'd like a

I don't know that we will need this for RTEMS.  We have all of
POSIX 1003.1b except processes, shared memory, and memory
mapping.  You can turn it all of if you like and go with a pSOS+
like "Classic API".  But for testing, we can build with everything
enabled. the Ada run-time expects it anyway.
  One random other thing, I have this line in my board.exp:

# The simulator isn't really remote.
set_board_info isremote 0

  I think you'll need that as well.
That sounds quite important as well.

I feel we will have results posted soon. :D

Can't think of a witty .sigline today....

Joel Sherrill, Ph.D.             Director of Research & Development
address@hidden        On-Line Applications Research
Ask me about RTEMS: a free RTOS  Huntsville AL 35805
  Support Available             (256) 722-9985

# Copyright (C) 1997-2008 Free Software
# Foundation, Inc.
# This file is part of DejaGnu.
# DejaGnu is free software; you can redistribute it and/or modify it
# under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
# DejaGnu is distributed in the hope that it will be useful, but
# WITHOUT ANY WARRANTY; without even the implied warranty of
# General Public License for more details.
# You should have received a copy of the GNU General Public License
# along with DejaGnu; if not, write to the Free Software Foundation,
# Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.

# This is a list of toolchains that are supported on this board.
set_board_info target_install {powerpc-rtems4.9}

# Load the generic configuration for this board, This will define a basic
# set of routines needed by the tool to communicate with the board.
load_generic_config "sim"

# basic-sim.exp is a basic description for the standard Cygnus simulator.
load_base_board_description "basic-sim"

# The name of the simulator is "ppc".
setup_sim psim-4.9

# No multilib flags needed by default.
process_multilib_options ""

# The compiler used to build for this board. This has *nothing* to do
# with what compiler is tested if we're testing gcc.
set_board_info compiler  "[find_gcc]"

set_board_info cflags  "-B${RTEMS_MAKEFILE_PATH}/lib/ -specs bsp_specs -qrtems 
set_board_info ldflags ""

# The simulator isn't really remote.
set_board_info isremote 0

# We only have a small stack available to us
set_board_info gcc,stack_size 2048

# psim-4.9 is a wrapper script which handles the device tree
set_board_info sim "psim-4.9 "

# No support for signals on this target.
set_board_info gdb,nosignals 1

# The simulator doesn't return exit statuses and we need to indicate this.
set_board_info needs_status_wrapper 1

# Can't call functions from GDB.
# set_board_info gdb,cannot_call_functions 1

reply via email to

[Prev in Thread] Current Thread [Next in Thread]