that is yet another multilib-related structual cleanup patch:
Changes:
* Make RTEMS_TEST_NO_PAUSE a tests/ subpackage specific option.
- Remove RTEMS_TEST_NO_PAUSE from custom/*.cfg, targopts.h and
cpuopts.h.
- Add autoconf macros RTEMS_*_RTEMS_TEST_NO_PAUSE
(aclocal/rtems-test-no-pause.m4).
- Add RTEMS_*_RTEMS_TEST_NO_PAUSE support to sptests/configure.ins
and tmtests/configure.in. These are the only subdirectories which
currently apply RTEMS_TEST_NO_PAUSE.
- Add autoconf-DEFS support to all test subpackages' configure.ins
below tests/. I.e. AC_DEFINES now get explicitly propagated as
preprocessor defines into Makefiles, cf. AM_CPPFLAGS in
tests/*/*.am, instead of using a global config-files.
- Remove NDEBUG from custom/*.cfg.
* AC_DEFINE POSIX_API, ITRON_API and MULTIPROCESSING in
exec/configure.in, only.
- All other sources now should relay on the values from cpuopts.h
and should not define them themselves.
- Several related changes to many configure.ins
* Bug-fixes to RTEMS_*_RTEMS_DEBUG macros (Actually workarounds to
quoting bugs in autoconf).
Notes:
* This patch is rather immature and only tested for a small subset
of BSPs (requires the tests to be enabled and therefore takes an
tremendous amount of disc space and time.)
* The patches to *cfg were generated by a script. Expect file
formating changes :)
that is yet another multilib-related structual cleanup patch:
Changes:
* Make RTEMS_DEBUG a global per-cpu configuration option
* Remove RTEMS_DEBUG from targopts.h
* Add a global --enable-rtems-debug option disabled by default.
* Add RTEMS_DEBUG to cpuopts.h
* Remove all references to RTEMS_DEBUG from custom/*.cfg
Notes:
* RTEMS_DEBUG is set in c/src/exec/configure.in only
(RTEMS_CHECK_RTEMS_DEBUG) and should be defined in cpuopts.h only.
BSPs should not redefine it, but use the value being provided by
cpuopts.h.
=> With multilibs, users have to choose: Either enable RTEMS_DEBUG
for all BSPs and CPU_MODELs of a cpu or not.
* Only few BSPs had RTEMS_DEBUG enabled, therefore I set the default
to disabled.
* This patch influences the per-BSP building scheme. Existing BSPs
which set RTEMS_DEBUG in their make-target-options rule might have
problems at runtime.
that decouples exec/ for the sh, m68k and i960 from targopts.h.
NOTE: The change to system.h is a hack to enable cpuopts.h
for some targets, but keep using targopts.h for others - I know it
does *not* work for sparc, mips, i386 and ppc. This will have
to be addressed as work continues on multilibing.
<charles.gauthier@iit.nrc.ca>, and Darlene A. Stewart
<Darlene.Stewart@nrc.ca> to add support for a number of very
significant things:
+ BSPs for many variations on the Motorola MBX8xx board series
+ Cache Manager including initial support for m68040
and PowerPC
+ Rework of mpc8xx libcpu code so all mpc8xx CPUs now use
same code base.
+ Rework of eth_comm BSP to utiltize above.
John reports this works on the 821 and 860
FP issues on this target:
The default variants of libc, libm and libgcc assume that a 68881
coprocessor is present. Without the FPSP, any floating point operation,
including printf() with a "%f" format specifier, is likely to cause an
unimplemented instruction exception.
The FPSP works with the default variants of libc, libm and libgcc. It does not
work in conjunction with the msoft-float variants. The paranoia test goes into
an infinite loop at milestone 40. I am guessing that floor() is returning an
incorrect value.
The msoft-float variants of libc, libm and libgcc appear to do floating point
I/O properly. They only failed in paranoia. Offhand, I can't think of why they
would conflict with the FPSP, so I think that there is something wrong with the
msoft-float code. It might be my installation.
Given my experiences, I decided to install the FPSP in bsp_start(), and to link
against the default variants of libc, libm and libgcc. This causes the
executables to increase in size by about 60 KB. The README file and the
mvme167.cfg specify how to remove the FPSP, and how to link against the
msoft-float variants of the libraries. This is not what Eric Norum had done: on
my host, his gen68360_040 port links RTEMS code with the msoft-float variants
of libc and libm, and the default variant of libgcc. In this configuration, the
output of printf() with "%f" is garbage on my target.
