[ECOS] Still can't build at91

[ECOS] Still can't build at91

[Michael Dorin]

I have upgrade to the most recent snapshot.

I type 
ecosconfig new at91 redboot

then

ecosconfig tree

then make (see below)

Is there something else I should be doing?

-Mike

I get:
 
/lib/libtarget.a src/hal_arm_arch_hal_misc.o
src/hal_arm_arch_context.o src/hal_arm_arch_arm_stub.o
src/hal_arm_arch_hal_syscall.o
arm-elf-gcc -c  -I/home/mike/red7/install/include
-I/home/mike/ecos/packages/hal/arm/arch/current
-I/home/mike/ecos/packages/hal/arm/arch/current/src
-I/home/mike/ecos/packages/hal/arm/arch/current/tests
-I.
-I/home/mike/ecos/packages/hal/arm/arch/current/src/
-finline-limit=7000 -mcpu=arm7tdmi
-mno-short-load-words -Wall -Wpointer-arith
-Wstrict-prototypes -Winline -Wundef  -g -O2
-ffunction-sections -fdata-sections  -fno-exceptions  
-Wp,-MD,src/redboot_linux_exec.tmp -o
src/hal_arm_arch_redboot_linux_exec.o
/home/mike/ecos/packages/hal/arm/arch/current/src/redboot_linux_exec.c
/home/mike/ecos/packages/hal/arm/arch/current/src/redboot_linux_exec.c:80:3:
#error 
make[1]: *** [src/redboot_linux_exec.o.d] Error 1
make[1]: Leaving directory
`/home/mike/red7/hal/arm/arch/current'
make: *** [build] Error 2


-- 
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss

Re: [ECOS] Still can't build at91

[Gary Thomas]

Michael Dorin wrote:
> 
> I have upgrade to the most recent snapshot.
> 
> I type 
> ecosconfig new at91 redboot
> 
> then
> 
> ecosconfig tree
> 
> then make (see below)
> 
> Is there something else I should be doing?

Works for me:

[gthomas@saturn t]$ ecosconfig new at91 redboot
U CYGSEM_HAL_USE_ROM_MONITOR, new inferred value 0
U CYGIMP_HAL_COMMON_INTERRUPTS_USE_INTERRUPT_STACK, new inferred value 0
[gthomas@saturn t]$ ecosconfig import ${ECOS_REPOSITORY}/hal/arm/at91/eb40/current/misc/redboot_RAM.ecm
[gthomas@saturn t]$ ecosconfig tree
[gthomas@saturn t]$ make -s
headers finished
cc1: warning: "-Woverloaded-virtual" is valid for C++ but not for C/ObjC
cc1: warning: "-fno-rtti" is valid for C++ but not for C/ObjC
/work/a_m/keymed/develop/ecos/packages/hal/common/current/src/hal_stub.c: In function 
`handle_exception_cleanup':
/work/a_m/keymed/develop/ecos/packages/hal/common/current/src/hal_stub.c:433: warning: dereferencing 
type-punned pointer will break strict-aliasing rules
/work/a_m/keymed/develop/ecos/packages/hal/common/current/src/hal_stub.c: In function 
`handle_exception_init':
/work/a_m/keymed/develop/ecos/packages/hal/common/current/src/hal_stub.c:489: warning: dereferencing 
type-punned pointer will break strict-aliasing rules
cc1: warning: "-Woverloaded-virtual" is valid for C++ but not for C/ObjC
cc1: warning: "-fno-rtti" is valid for C++ but not for C/ObjC
cc1: warning: "-Woverloaded-virtual" is valid for C++ but not for C/ObjC
cc1: warning: "-fno-rtti" is valid for C++ but not for C/ObjC
build finished

A couple of things to note:

* Make sure you select the correct target.  'at91' is really a
   family of boards.  The target 'at91' specifies ARM's original
   evaluation board the EB40 which came out in around 1998.  I
   doubt that's what you actually have.

* The basic RedBoot configuration often needs some platform specific\
   setups.  That's what the "ecosconfig import XXX.ecm" step is for.
   In particular, the original EB40 was too small to run Linux, so
   all Linux support needed to be disabled (the source of *your* error).
   That's all included in the .../eb40/current/misc/redboot_RAM.ecm file.

* Always start from a clean directory when you run 'ecosconfig new ...'
   Anything left over when you run this is just asking for trouble.

-- 
------------------------------------------------------------
Gary Thomas                 |  Consulting for the
MLB Associates              |    Embedded world
------------------------------------------------------------

-- 
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss