[ECOS] h8300 cannot find libgcc.a

[ECOS] h8300 cannot find libgcc.a

[m mariga]

Hello everyone,

I installed h8300 toolchains in eCos 3.0 and tried to run application using
Akizuki H8/3068 Template on Akizuki Board.
But I was encountered by application's Make-error and in a deadlock
predicament.
Would you please help me and rescue me out ?

I used next sources.
binutils-2.19.1.tar.bz2
gcc-4.4.0.tar.bz2
newlib-1.17.0.tar.gz

As I said, I succeeded to install h8300 toolchains in /gnutools directory
and succeeded to install RedBoot in Akizuki Board.
Next I decided to run eCos-bundled nc_test_slave.c and tried to build
libtarget.a.
In order to build libtarget.a, I selected Akizuki H8/3068 Template and "net"
package.
Here, I've met a serious error problem.

/opt/ecos/ecos-3.0/packages/language/c/libm/v3_0/src/mathincl/fdlibm.h:78:3:
error: #error IEEE-style 64-bit doubles are required to use the math library

Here're extracts from original fdlibm.h.
---
// SANITY CHECKS
// Just check that we support IEEE-style 64-bit doubles. If not, this
// math library will not work
// This check will go away when support for single-precision alternatives
are
// provided

#if DBL_MAX_EXP != 1024
# error IEEE-style 64-bit doubles are required to use the math library
#endif // if DBL_MAX_EXP == 1024
---

Tell the truth I didn't understand but when I searched DBL_MAX_EXP in .c and
.h files couldn't find it, so I escaped by next simplistic evasion.
---
#define DBL_MAX_EXP 1024
#if DBL_MAX_EXP != 1024
# error IEEE-style 64-bit doubles are required to use the math library
#endif // if DBL_MAX_EXP == 1024
---
I inserted #define DBL_MAX_EXP 1024 !!
I could see your scolding faces but I needed to make it,so let me continue
my story.

Anyway I succeeded to make libtarget.a,so next I continued to make
application.
And I've run across next linking error.

$ make
h8300-elf-gcc -c -o
nc_test_slave.o -mh -mint32 -g -Wall -I/home/link/aki-net-0717/untitled_install/include
 -ffunction-s
ections -fdata-sections nc_test_slave.c
nc_test_slave.c: In function 'do_udp_test':
nc_test_slave.c:215: warning: format '%x' expects type 'unsigned int', but
argument 2 has type 'long int'
nc_test_slave.c:215: warning: format '%x' expects type 'unsigned int', but
argument 3 has type 'long int'
nc_test_slave.c: In function 'do_tcp_test':
nc_test_slave.c:391: warning: format '%x' expects type 'unsigned int', but
argument 2 has type 'long int'
nc_test_slave.c:391: warning: format '%x' expects type 'unsigned int', but
argument 3 has type 'long int'
h8300-elf-gcc -nostartfiles -L/home/link/aki-net-0717/untitled_install/lib -Wl,--gc-sections
 -Wl,--Map -Wl,nc_test_slave
.map -o nc_test_slave nc_test_slave.o -Ttarget.ld -nostdlib
/gnutools/lib/gcc/h8300-elf/4.4.0/../../../../h8300-elf/bin/ld: skipping
incompatible /gnutools/lib/gcc/h8300-elf/4.4.0/
libgcc.a when searching for libgcc.a
/gnutools/lib/gcc/h8300-elf/4.4.0/../../../../h8300-elf/bin/ld: cannot find
libgcc.a
collect2: ld returned 1 exit status
make: *** [nc_test_slave] Error 1

I confirmed there is libgcc.a in /gnutools/lib/gcc/h8300-elf/4.4.0/.
I do not know it is directly related to my botched evasion.
Please show me the correct way.

By the way,please let me use an assumed name.Recently I discovered my
original name is used by spams.
May be punished the spam builders !

Thank you in advance for your help.

m mariga


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

[ECOS] Re: h8300 cannot find libgcc.a

[John Dallaway]

m mariga wrote:

> /opt/ecos/ecos-3.0/packages/language/c/libm/v3_0/src/mathincl/fdlibm.h:78:3:
> 
> error: #error IEEE-style 64-bit doubles are required to use the math library

You compiler is providing 32-bit doubles. You need 64-bit doubles in
order to build the eCos math library. There's a patch available for the
KPIT v0401 tools which may help:

  http://www.cetoni.de/develop/develop_ecosh8s_en.html#a_broschueren

> h8300-elf-gcc -c -o nc_test_slave.o -mh -mint32 -g -Wall -I/home/link/aki-net-0717/untitled_install/include -ffunction-sections -fdata-sections nc_test_slave.c
> nc_test_slave.c: In function 'do_udp_test':
> nc_test_slave.c:215: warning: format '%x' expects type 'unsigned int', but argument 2 has type 'long int'
> nc_test_slave.c:215: warning: format '%x' expects type 'unsigned int', but argument 3 has type 'long int'
> nc_test_slave.c: In function 'do_tcp_test':
> nc_test_slave.c:391: warning: format '%x' expects type 'unsigned int', but argument 2 has type 'long int'
> nc_test_slave.c:391: warning: format '%x' expects type 'unsigned int', but argument 3 has type 'long int'
> h8300-elf-gcc -nostartfiles -L/home/link/aki-net-0717/untitled_install/lib -Wl,--gc-sections -Wl,--Map -Wl,nc_test_slave.map -o nc_test_slave nc_test_slave.o -Ttarget.ld -nostdlib
> /gnutools/lib/gcc/h8300-elf/4.4.0/../../../../h8300-elf/bin/ld: skipping incompatible /gnutools/lib/gcc/h8300-elf/4.4.0/libgcc.a when searching for libgcc.a
> /gnutools/lib/gcc/h8300-elf/4.4.0/../../../../h8300-elf/bin/ld: cannot find libgcc.a
> collect2: ld returned 1 exit status

Perhaps your toolchain does not provide the appropriate multilib support
for "-mh -mint32"? You can check using:

   $ h8300-elf-gcc -print-multi-lib

Note that most of the eCos community is using GCC 4.3.2 so you may well
discover a few compiler warnings and other issues if you build with GCC
4.4.0.

John Dallaway

-- 
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] Re: h8300 cannot find libgcc.a

[m mariga]

[-- Attachment #1: Type: text/plain, Size: 2802 bytes --]

Hello,John,and everybody,

Thank you John for your reply.
I think I became a little bit knowledgeable.

> Note that most of the eCos community is using GCC 4.3.2 so you may well
> discover a few compiler warnings and other issues if you build with GCC
> 4.4.0.

First I tried to use GCC 4.3.2 for installing h8300 ToolChain.
But it resulted Segmentation error at configureration.

-- gcc4.3.2 confiure errror
/tmp/build/build_gcc/./gcc/xgcc -B/tmp/build/build_gcc/./gcc/ -nostdinc -B/tmp/build/build_gcc/h8300-elf/h8300s/normal/newlib/
 -isystem /tmp/build/build_gcc/h8300-elf/h8300s/normal/newlib/targ-include
 -isystem
/src/gcc-4.3.2/newlib/libc/include -B/gnutools/h8300-elf/bin/ -B/gnutools/h8300-elf/lib/
 -isystem /gnutools/h8300-elf/include -isystem
/gnutools/h8300-elf/sys-include  -ms -mn -DPACKAGE_NAME
=\"newlib\" -DPACKAGE_TARNAME=\"newlib\" -DPACKAGE_VERSION=\"1.16.0\" -DPACKAGE_STRING=\"newlib\
1.16.0\" -DPACKAGE_BUGREPORT=\"\" -I. -I/src/gcc-4.3.2/newlib/libm/math -I/src/gcc-4.3.2/newlib/libm/ma
th/../common -O2 -DSMALL_DTOA -DSMALL_MEMORY -fno-builtin      -O2 -g -g -O2
    -ms -mn -c -o lib_a-wf_lgamma.o `test -f 'wf_lgamma.c' || echo
'/src/gcc-4.3.2/newlib/libm/math/'`wf_lgamma.c
In file included from /src/gcc-4.3.2/newlib/libc/include/math.h:5,
                 from /src/gcc-4.3.2/newlib/libm/math/../common/fdlibm.h:15,
                 from /src/gcc-4.3.2/newlib/libm/math/wf_lgamma.c:17:
/src/gcc-4.3.2/newlib/libc/include/sys/reent.h:95: internal compiler error:
Segmentation fault
Please submit a full bug report,
---

I asked to gcc mailing list but could not get reply.
I tried to use GCC 4.4.0 and there was no error,so I am determined to use 
GCC 4.4.0.

>> /opt/ecos/ecos-3.0/packages/language/c/libm/v3_0/src/mathincl/fdlibm.h:78:3:
>>
>> error: #error IEEE-style 64-bit doubles are required to use the math 
>> library
>
> You compiler is providing 32-bit doubles. You need 64-bit doubles in
> order to build the eCos math library. There's a patch available for the
> KPIT v0401 tools which may help:
>
>  http://www.cetoni.de/develop/develop_ecosh8s_en.html#a_broschueren

I downloaded the patch file but it is for use of GCC 3.3 source.
I also get the same purpose patch file that is for GCC 3.3.2 source.
I append it in zip file.

I studied it and learned that I could not use it for GCC 4.4.0 source per 
se.
I searched Internet for newer patch corresponding to GCC 4.4.0 or at least 
to GCC 4.3.2 but could not find one.
Couldn't it be possible that you send me the newer patch corresponding to 
GCC 4.4.0 or at least to GCC 4.3.2 ?

If I could not get it I must study the old patch and change the GCC 4.4.0 
source manually.
It is hard, isn't it, because there are too differences between GCC 3.3.2 
and GCC 4.4.0.

Do you have any better idea ?

m mariga

[-- Attachment #2: gcc_diff_new_patch.gz --]
[-- Type: application/x-gzip, Size: 3058 bytes --]

[-- Attachment #3: Type: text/plain, Size: 148 bytes --]

-- 
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] Re: h8300 cannot find libgcc.a

[John Dallaway]

Hi

m mariga wrote:

> I asked to gcc mailing list but could not get reply.
> I tried to use GCC 4.4.0 and there was no error,so I am determined to
> use GCC 4.4.0.

[ snip ]

> I downloaded the patch file but it is for use of GCC 3.3 source.
> I also get the same purpose patch file that is for GCC 3.3.2 source.
> I append it in zip file.
> 
> I studied it and learned that I could not use it for GCC 4.4.0 source
> per se.
> I searched Internet for newer patch corresponding to GCC 4.4.0 or at
> least to GCC 4.3.2 but could not find one.
> Couldn't it be possible that you send me the newer patch corresponding
> to GCC 4.4.0 or at least to GCC 4.3.2 ?

I am not aware of a corresponding patch for GCC 4.x, but the necessary
changes should be reasonably easy to make by editing the source files
manually.

Please post the patch when you get this working.

Thanks

John Dallaway

-- 
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] Re: h8300 cannot find libgcc.a

[m mariga]

[-- Attachment #1: Type: text/plain, Size: 1348 bytes --]

Hello,John,and everybody,

I studied GCC3.3.2 patch and really applied to GCC3.3.2 source.
When I compared differences and tried to convert GCC4.4.0 source,there 
appeared one difficult problem.
Would you please teach me ?
I append before-patch source after-patch source and gcc4.4.0 source in 
compressed form.

The patch changes next four files.
/gcc-3.3.2/gcc/config/h8300/extfloatlib.c
/gcc-3.3.2/gcc/config/h8300/float-h8300.h
/gcc-3.3.2/gcc/config/h8300/h8300.h
/gcc-3.3.2/gcc/config/h8300/t-h8300

First two files(extfloatlib.c,float-h8300.h) are newly made,so I think I 
could use them per ce into GCC4.4.0 source.
h8300.h and t-h8300 are modified and I must apply differences to GCC4.4.0 
source.

The patch adds next line into 143 line of GCC3.3.2's h8300.h.

#define TARGET_MODE (target_flags & 16384)

As I tried to insert above define sentence in GCC4.4.0's h8300.h I was 
shocked to have found that there is no TARGET_MODE or target_flags.
Please teach me whether it is alright that I omit to insert above define 
sentence at all.
Or if it is necessary to add it,could you teach me how to do it ?
I am very obliged.

And about another differences I intend to add the same sentences int 
GCC4.4.0's h8300.h and t-h8300.
If you notice carelessness could you enlighten me.

I am sorry to bother you but you are my only rescue.

m mariga

[-- Attachment #2: sources.tar.bz2 --]
[-- Type: application/octet-stream, Size: 29099 bytes --]

[-- Attachment #3: Type: text/plain, Size: 148 bytes --]

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

[ECOS] Re: h8300 cannot find libgcc.a

[John Dallaway]

m mariga wrote:

> The patch adds next line into 143 line of GCC3.3.2's h8300.h.
> 
> #define TARGET_MODE (target_flags & 16384)
> 
> As I tried to insert above define sentence in GCC4.4.0's h8300.h I was
> shocked to have found that there is no TARGET_MODE or target_flags.
> Please teach me whether it is alright that I omit to insert above define
> sentence at all.
> Or if it is necessary to add it,could you teach me how to do it ?
> I am very obliged.

TARGET_MODE is not referenced in gcc-core-3.3.5 either. Perhaps it is
completely unnecessary.

John Dallaway

-- 
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] Re: h8300 cannot find libgcc.a

[m mariga]

[-- Attachment #1: Type: text/plain, Size: 4145 bytes --]

Hello,John,and everybody,

I am sorry to inform you that I could not succeed to configure GCC4.4.0 to 
install h8300 toolchain.
I omitted TARGET_MODE define sentence,and converted GCC4.4.0 source as I 
previously said.

I succeeded binutils configure.
But when I configured gcc, next error happened.

--- error begin
/tmp/build/build_gcc/./gcc/xgcc -B/tmp/build/build_gcc/./gcc/ -nostdinc -B/tmp/build/build_gcc/h8300-elf/newlib/ 
 -isyste
m /tmp/build/build_gcc/h8300-elf/newlib/targ-include -isystem 
/src/gcc-4.4.0/newlib/libc/include -B/gnutools/h8300-elf/b
in/ -B/gnutools/h8300-elf/lib/ -isystem /gnutools/h8300-elf/include -isystem 
/gnutools/h8300-elf/sys-include -g -O2 -mh
-O2  -g -O2 -DIN_GCC -DCROSS_DIRECTORY_STRUCTURE  -W -Wall -Wwrite-strings -Wstrict-prototypes 
 -Wmissing-prototypes -Wca
st-qual -Wold-style-definition  -isystem 
./include   -g  -DIN_LIBGCC2 -D__GCC_FLOAT_NOT_NEEDED -Dinhibit_libc  -I. -I. 
 -
I../../.././gcc -I/src/gcc-4.4.0/libgcc -I/src/gcc-4.4.0/libgcc/. -I/src/gcc-4.4.0/libgcc/../gcc 
 -I/src/gcc-4.4.0/libgcc
/../include  -DHAVE_CC_TLS -o _bswapsi2.o -MT _bswapsi2.o -MD -MP -MF 
_bswapsi2.dep -DL_bswapsi2 -c /src/gcc-4.4.0/libgc
c/../gcc/libgcc2.c \

/tmp/build/build_gcc/./gcc/xgcc -B/tmp/build/build_gcc/./gcc/ -nostdinc -B/tmp/build/build_gcc/h8300-elf/newlib/ 
 -isyste
m /tmp/build/build_gcc/h8300-elf/newlib/targ-include -isystem 
/src/gcc-4.4.0/newlib/libc/include -B/gnutools/h8300-elf/b
in/ -B/gnutools/h8300-elf/lib/ -isystem /gnutools/h8300-elf/include -isystem 
/gnutools/h8300-elf/sys-include -g -O2 -mh
-O2  -g -O2 -DIN_GCC -DCROSS_DIRECTORY_STRUCTURE  -W -Wall -Wwrite-strings -Wstrict-prototypes 
 -Wmissing-prototypes -Wca
st-qual -Wold-style-definition  -isystem 
./include   -g  -DIN_LIBGCC2 -D__GCC_FLOAT_NOT_NEEDED -Dinhibit_libc  -I. -I. 
 -
I../../.././gcc -I/src/gcc-4.4.0/libgcc -I/src/gcc-4.4.0/libgcc/. -I/src/gcc-4.4.0/libgcc/../gcc 
 -I/src/gcc-4.4.0/libgcc
/../include  -DHAVE_CC_TLS -o _bswapdi2.o -MT _bswapdi2.o -MD -MP -MF 
_bswapdi2.dep -DL_bswapdi2 -c /src/gcc-4.4.0/libgc
c/../gcc/libgcc2.c \

/src/gcc-4.4.0/libgcc/../gcc/libgcc2.c:503: error: expected '=', ',', ';', 
'asm' or '__attribute__' before '__bswapdi2'
/src/gcc-4.4.0/libgcc/../gcc/libgcc2.c:505:19: warning: integer constant is 
too large for its type
/src/gcc-4.4.0/libgcc/../gcc/libgcc2.c:506:14: warning: integer constant is 
too large for its type
/src/gcc-4.4.0/libgcc/../gcc/libgcc2.c:507:14: warning: integer constant is 
too large for its type
/src/gcc-4.4.0/libgcc/../gcc/libgcc2.c:508:14: warning: integer constant is 
too large for its type
make[4]: *** [_bswapdi2.o] Error 1
make[4]: Leaving directory `/tmp/build/build_gcc/h8300-elf/h8300h/libgcc'
make[3]: *** [multi-do] Error 1
make[3]: Leaving directory `/tmp/build/build_gcc/h8300-elf/libgcc'
make[2]: *** [all-multi] Error 2
make[2]: Leaving directory `/tmp/build/build_gcc/h8300-elf/libgcc'
make[1]: *** [all-target-libgcc] Error 2
make[1]: Leaving directory `/tmp/build/build_gcc'
make: *** [all] Error 2
make: Leaving directory `/tmp/build/build_gcc'
--- error end

Here are 501-503 lines of libgcc2.c.
#ifdef L_bswapdi2
DItype
__bswapdi2 (DItype u)

Error is caused by not understanding DItype.
Since when I retrieved original gcc4.4.0 sources then not occurred above 
error,
so I am sure I wrongly converted gcc4.4.0 four files.

I am very sorry to bother you,but I do not understand.
I append original and after-converted four files in compressed form.

I appended gcc3.3.2 original h8300.h,t-h8300.
patch-sources/gcc3.3.2-before-patch-sources/h8300.h,t-h8300

And after-patch-applied gcc3.3.2 four files.
patch-sources/gcc3.3.2-after-patch-sources/h8300.h,t-h8300,extfloatlib.c,float-h8300.h

And gcc4.4.0 original h8300.h,t-h8300.
patch-sources\gcc4.4.0-original-sources/h8300.h,t-h8300

And after-converted gcc4.4.0 four files according to patch.
patch-sources/gcc4.4.0-patch-converted-sources/h8300.h,t-h8300,extfloatlib.c,float-h8300.h

I beseech you to look up wrongfully converted codes in after-converted 
gcc4.4.0 four files.

Would you please let me know the error-causing codes.
I am very much obliged.

m mariga


[-- Attachment #2: patch-sources.tar.bz2 --]
[-- Type: application/octet-stream, Size: 27613 bytes --]

[-- Attachment #3: Type: text/plain, Size: 148 bytes --]

-- 
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] Re: h8300 cannot find libgcc.a

[m mariga]

Hello,

I've studied the GCC3.3.2's patch for building h8300 toolchains and I've 
learned its mechanism
and also I am encountered with new question.
Would you please teach me.

When gcc-configuration makes libgcc.a,double must be 64 bits,and when making 
h8300-elf-gcc double must be 32 bits.
In order to distinguish two targets the patch exhanges next lines.

-#define LONG_LONG_TYPE_SIZE 32
+#define LONG_LONG_TYPE_SIZE (TARGET_INT32 ? 64 : 32)
-#define DOUBLE_TYPE_SIZE 32
+#define DOUBLE_TYPE_SIZE (TARGET_INT32 ? 64 : 32)

And there were next codes in GCC3.3.2's /gcc/config/h8300/h8300.h.

--- from GCC3.3.2's /gcc/config/h8300/h8300.h
/* Run-time compilation parameters selecting different hardware subsets.  */

extern int target_flags;

/* Masks for the -m switches.  */
#define MASK_H8300S  0x00000001
#define MASK_MAC  0x00000002
#define MASK_INT32  0x00000008
#define MASK_ADDRESSES  0x00000040
#define MASK_QUICKCALL  0x00000080
#define MASK_SLOWBYTE  0x00000100
#define MASK_NORMAL_MODE  0x00000200
#define MASK_RELAX  0x00000400
#define MASK_RTL_DUMP  0x00000800
#define MASK_H8300H  0x00001000
#define MASK_ALIGN_300  0x00002000

/* Macros used in the machine description to test the flags.  */

/* Make int's 32 bits.  */
#define TARGET_INT32 (target_flags & MASK_INT32)

/* Dump recorded insn lengths into the output file.  This helps debug the
   md file.  */
#define TARGET_ADDRESSES (target_flags & MASK_ADDRESSES)

/* Pass the first few arguments in registers.  */
#define TARGET_QUICKCALL (target_flags & MASK_QUICKCALL)

/* Pretend byte accesses are slow.  */
#define TARGET_SLOWBYTE (target_flags & MASK_SLOWBYTE)

/* Dump each assembler insn's rtl into the output file.
   This is for debugging the compiler only.  */
#define TARGET_RTL_DUMP (target_flags & MASK_RTL_DUMP)
---

But there are not above codes in GCC4.4.0's /gcc/config/h8300/h8300.h !!!

I tentatively added above codes in GCC4.4.0's /gcc/config/h8300/h8300.h and 
applied patch codes and tried to configure gcc.
But the result was the same error.
/opt/ecos/ecos-3.0/packages/language/c/libm/v3_0/src/mathincl/fdlibm.h:78:3: 
error: #error IEEE-style 64-bit doubles are required to use the math library

Apparently in GCC4.4.0 or newer GCC,target_flags does not function as the 
same as in GCC3.3.2.

Now the question is what mechanism is used in newer GCC versions to 
configure h8300 toolchains.
How should I distinguish targets, libgcc.a and h8300-elf-gcc, in configuring 
?

Would you please enlighten me ?

m mariga


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

[ECOS] Re: h8300 cannot find libgcc.a

[John Dallaway]

Hi

m mariga wrote:

> Apparently in GCC4.4.0 or newer GCC,target_flags does not function as
> the same as in GCC3.3.2.
> 
> Now the question is what mechanism is used in newer GCC versions to
> configure h8300 toolchains.
> How should I distinguish targets, libgcc.a and h8300-elf-gcc, in
> configuring ?
> 
> Would you please enlighten me ?

Unfortunately, I do not have an answer for you from memory. I would have
to delve into the GCC sources and investigate as you are doing.

John Dallaway

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

[ECOS] How to set GDB Stub Options in order to run insight ?

[m mariga]

Hello,

I am working on eCos 3.0 using Akizuki H8/3068 Network board.
That board's CPU is H8/3069F .
I succeeded to install FreeBSD stack and ran nc_test_slave/master test 
programs.

But I cannot run h8300-elf-insight,so please help me.

I start up insight like next line.
$ h8300-elf-insight.exe nc_test_slave.elf

After insight is started up,I entered next commands on console window.
----
(gdb) set remotebau 57600      --- I setted baudrate 57600
(gdb) target remote /dev/com4
Remote debugging using /dev/com4
0x000073a4 in ?? ()

(gdb) load
Loading section .text, size 0x599ce lma 0x410000
Loading section .rodata, size 0x406a lma 0x4699ce
Loading section .data, size 0x1660 lma 0x46da38
Start address 0x410000, load size 389272
Transfer rate: 3 KB/sec, 63 bytes/write.

(gdb) b cyg_start
Note: breakpoint 2 also set at pc 0x411dca.
Breakpoint 3 at 0x411dca: file nc_test_slave.c, line 729.

(gdb) c
Continuing. ...

At this point nothing happens.Although state is "Running" and mouse_cursor 
is running form.
Nothing available(even Ctr-C does not work) except stop button on 
insight-window's main head.
When clicked stop button,appered message window saying,
"No response from target.Detach from target
(and stop debugging it)?"
And clicked YES button, then appeared next line.

Error: Watchdog timeout has expired.  Target detached.

(gdb)

My guessing is GDB stub not functioning properly.

Here are questions.
First of all,couldn't I set "net"and "stubs" Packages together on ConfigTool 
?
By Packages,I mean,it is in Templates dialog-window not in Packages 
dialog-window.
When I selected "net",there's nothing in  Packages dialog-window 
corresponding to "stubs" function.
If I could not set both functions together on ConfigTool conveniently,then I 
must convert cdl files.

I think next Options are related to GDB stubs.
CYGDBG_HAL_DEBUG_GDB_INCLUDE_STUBS
CYGDBG_HAL_DEBUG_GDB_BREAK_SUPPORT
CYGDBG_HAL_DEBUG_GDB_CTRLC_SUPPORT
CYGDBG_HAL_DEBUG_GDB_THREAD_SUPPORT
CYGBLD_BUILD_GDB_STUBS
CYGBLD_BUILD_COMMON_GDB_STUBS

I checked current definitions in cdl files.
cdl_option CYGDBG_HAL_DEBUG_GDB_INCLUDE_STUBS {
    default_value 0

cdl_option CYGDBG_HAL_DEBUG_GDB_BREAK_SUPPORT {
    requires      CYGDBG_HAL_DEBUG_GDB_INCLUDE_STUBS
    default_value CYGDBG_HAL_DEBUG_GDB_INCLUDE_STUBS

cdl_option CYGDBG_HAL_DEBUG_GDB_CTRLC_SUPPORT {
    active_if     { CYGSEM_HAL_USE_ROM_MONITOR || 
CYGDBG_HAL_DEBUG_GDB_INCLUDE_STUBS }
    active_if     { CYGINT_HAL_DEBUG_GDB_CTRLC_UNSUPPORTED == 0 }
    default_value { !CYGDBG_HAL_DEBUG_GDB_INCLUDE_STUBS }

cdl_option CYGDBG_HAL_DEBUG_GDB_THREAD_SUPPORT {
    active_if     { CYGSEM_HAL_ROM_MONITOR || 
CYGDBG_KERNEL_DEBUG_GDB_THREAD_SUPPORT }
    default_value 1

cdl_option CYGBLD_BUILD_GDB_STUBS {
    default_value 0

cdl_option CYGBLD_BUILD_COMMON_GDB_STUBS {
    default_value 0

How should I set these options in order to insatall GDB stub ?

I am very much obliged to your help.

m mariga


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

[ECOS] Re: How to set GDB Stub Options in order to run insight ?

[John Dallaway]

m mariga wrote:

> First of all,couldn't I set "net"and "stubs" Packages together on
> ConfigTool ?
> By Packages,I mean,it is in Templates dialog-window not in Packages
> dialog-window.
> When I selected "net",there's nothing in  Packages dialog-window
> corresponding to "stubs" function.
> If I could not set both functions together on ConfigTool
> conveniently,then I must convert cdl files.

The GDB stubs configuration (configured via the "stubs" template) does
not support networking. If you require GDB remote debugging over a
network connection then you will need to build RedBoot using the
"redboot" template. RedBoot incorporates its own TCP/IP stack so you do
not need to add eCos networking packages.

If you have not already done so, I suggest you proceed by attempting to
configure and build the GDB stubs by selecting the "stubs" template. For
most targets, no further configuration is necessary.

I hope this helps...

John Dallaway

-- 
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] Re: How to set GDB Stub Options in order to run insight ?

[m mariga]

Hello,

Thank you ,John,for your reply.

Am I correct taking you that there is no need to change any options at all ?
Should insight work properly as it is now ?

Do you have any suggestions to investigate the cause of malfunction ?
For example is there any GDB command to help me ?

I appreciate your help.

m mariga


-- 
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] Re: How to set GDB Stub Options in order to run insight ?

[m mariga]

Hello,

I changed <CYGDBG_HAL_DEBUG_GDB_INCLUDE_STUBS> option's defult value from 0 
to 1,
and rebult nc_test_slave.elf and tried to run Insight again.
Then it developed like next log.
{according to Resolve conflicts Winow:
CYGBLD_ISO_FNMATCH_HEADER 
Enabled
CYGDBG_HAL_COMMON_CONTEXT_SAVE_MINIMUM            Diasabled
CYGDBG_HAL_COMMON_INTERRUPTS_SAVE_MINIMUM_CONTEXT Diasabled
CYGBLD_ISO_STRUCTTIMEVAL_HEADER                                    Enabled
}
--- new Insight's cosole window log
(gdb) set remotebaud 57600
(gdb) set remotelogfile remote-4.log
(gdb) target remote /dev/com4
Remote debugging using /dev/com4
0x000073a4 in ?? ()

(gdb) load
Loading section .text, size 0x5d58a lma 0x410000
Loading section .rodata, size 0x431e lma 0x46d58a
Loading section .data, size 0x16f8 lma 0x4718a8
Start address 0x410000, load size 405408
Transfer rate: 3 KB/sec, 63 bytes/write.

(gdb) b cyg_start
Breakpoint 1 at 0x411dda: file nc_test_slave.c, line 729.

(gdb) c
Continuing.
Network stack using 69632 bytes for misc space
                    69632 bytes for mbufs
                    139264 bytes for mbuf clusters
[cyg_net_init] Init: mbinit(0x00000000)
[cyg_net_init] Init: cyg_net_init_devs(0x00000000)
Init device 'dp83902a_eth0'
DP83902A - eeprom ESA: 00:02:cb:01:42:b9
[cyg_net_init] Init: loopattach(0x00000000)
[cyg_net_init] Init: ifinit(0x00000000)
[cyg_net_init] Init: domaininit(0x00000000)
[cyg_net_init] Init: cyg_net_add_domain(0x004728a8)
New domain internet at 0x00000000
[cyg_net_init] Init: cyg_net_add_domain(0x00472298)
New domain route at 0x00000000
[cyg_net_init] Init: call_route_init(0x00000000)
[cyg_net_init] Done

But at this point stuck to silence and no reply like before.

As I compared to previous normal RedBoot Serial log like next, I realized
it stopped at network operation.

--- previous normal RedBoot Serial log
RedBoot> go 0x410000
Network stack using 69632 bytes for misc space
                    69632 bytes for mbufs
                    139264 bytes for mbuf clusters
[cyg_net_init] Init: mbinit(0x00000000)
[cyg_net_init] Init: cyg_net_init_devs(0x00000000)
Init device 'dp83902a_eth0'
DP83902A - eeprom ESA: 00:02:cb:01:42:b9
[cyg_net_init] Init: loopattach(0x00000000)
[cyg_net_init] Init: ifinit(0x00000000)
[cyg_net_init] Init: domaininit(0x00000000)
[cyg_net_init] Init: cyg_net_add_domain(0x0046e9a0)
New domain internet at 0x00000000
[cyg_net_init] Init: cyg_net_add_domain(0x0046e398)
New domain route at 0x00000000
[cyg_net_init] Init: call_route_init(0x00000000)
[cyg_net_init] Done
Start Network Characterization - SLAVE
BOOTP[eth0] op: REPLY
       htype: Ethernet
        hlen: 6
        hops: 0
         xid: 0x0
        secs: 0
       flags: 0x0
       hw_addr: 00:02:cb:01:42:b9
     client IP: 172.16.1.200
         my IP: 172.16.1.200
     server IP: 172.16.1.1
    gateway IP: 172.16.1.1
  options:
        subnet mask: 255.255.255.0
       IP broadcast: 172.16.1.255
            gateway: 172.16.1.1
[eth_drv_ioctl] Warning: Driver can't set multi-cast mode
---

Also after changed <CYGDBG_HAL_DEBUG_GDB_INCLUDE_STUBS> to 1,I loaded 
application and run it on RedBoot,
then where it should appear "Network stack using" like above,it resulted in 
enigmatic numeric codes.

--- after changed <CYGDBG_HAL_DEBUG_GDB_INCLUDE_STUBS> to 1,RedBoot Serial 
run log
RedBoot> load -v -m xmodem -b 0x410000 nc_test_slave.srec
CEntry point: 0x00410000, address range: 0x00410000-0x00472fa0
xyzModem - CRC mode, 8909(SOH)/0(STX)/0(CAN) packets, 5 retries
RedBoot> go 0x410000
$O4E6574776F726B20737461636B207573696E6720363936333220627974657320666F72206D6973632073706163650A#0C

I concluded that according to my changing 
<CYGDBG_HAL_DEBUG_GDB_INCLUDE_STUBS> to 1,network became malfunctioning.

Anyway I could find a way out of the swamp.
I suppose by handling options properly I could run both GDB stub and network 
properly.
I strongly hope I could debug networking application using Insight.
Would you please help me ?

m mariga


-- 
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] Re: How to set GDB Stub Options in order to run insight ?

[John Dallaway]

m mariga wrote:

> But at this point stuck to silence and no reply like before.
> 
> As I compared to previous normal RedBoot Serial log like next, I realized
> it stopped at network operation.

It is hard to follow precisely what you are doing from your mail. I am
assuming that you have switched to using RedBoot rather than the plain
GDB stubs.

> BOOTP[eth0] op: REPLY
>       htype: Ethernet
>        hlen: 6
>        hops: 0
>         xid: 0x0
>        secs: 0
>       flags: 0x0
>       hw_addr: 00:02:cb:01:42:b9
>     client IP: 172.16.1.200
>         my IP: 172.16.1.200

For avoidance of doubt, the RedBoot TCP/IP stack and the eCos
application TCP/IP stack must use different IP addresses. Your eCos
networking stack is using DHCP/BOOTP to obtain an IP address so you must
configure RedBoot to use a unique static IP address (not 172.16.1.200).

John Dallaway

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

[ECOS] cyg_scheduler_lock and DSR's

[Jim Bradleigh]

Hi,

Is the cyg_scheduler_lock() call not supposed to prevent DSR's from running?

I have a alarm, attached to the realtime clock which when activated posts a message to a custom queue system. This queue is also posted to by user land code..

Pushes to the queue are protected by a mutex. Additionally around the mutex I have the cyg_scheduler_lock() and cyg_scheduler_unlock() to prevent the DSR posting a message whilst inside the mutex operation.

However I`m seeing asserts stating the mutex is already locked. A quick breakpoint and I can see that the DSR is still being called even after the lock() function..

Any thoughts?

James


--
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] Re: How to set GDB Stub Options in order to run insight ?

[m mariga]

Hello,

I am very sorry for my ambiguity.
Please make allowance for my poor capacity of English.

What I am doing is only to confirm the operation of Insight debugging 
application(nc_test_slave.elf).
My purpose is that I start up Insight on cygwin and load nc_test_slave.elf 
and know it breaks at top of cyg_start and I can do stepping operation.

Since I changed <CYGDBG_HAL_DEBUG_GDB_INCLUDE_STUBS> option to 1,Insight 
began to display log strings when I entered "c"(continue),
but it still  can't go to break-point at top of cyg_start and bogged in 
silence state after displaying [cyg_net_init] Done.

I only showed normal RedBoot-log for comparing in order to know the cause of 
Insight's stopping.
This was the log that I ran nc_test_slave on RedBoot-Monitor by "go" command 
without Insight.
I deducted that Insight-stopping's cause is related to network operation 
because on RedBoot-log,"Start Network Characterization - SLAVE" was 
displayed after [cyg_net_init] Done.
My guessing is that I perhaps could be able to run through Insight by making 
network-related options available.
I beseech you to teach me that.

After I changed <CYGDBG_HAL_DEBUG_GDB_INCLUDE_STUBS> option to 1,I re-ran 
newly-built nc_test_slave on RedBoot-Monitor without Insight by way of 
trial.
Then it failed by displaying next codes,whereas on previous normal 
RedBoot-log it showed "Network stack using 69632 bytes for misc space" .

RedBoot> go 0x410000
$O4E6574776F726B20737461636B207573696E6720363936333220627974657320666F72206D6973632073706163650A#0C

I strengthened my network-related guessing because network function was 
deteriorated by my changing <CYGDBG_HAL_DEBUG_GDB_INCLUDE_STUBS> option to 
1.

By the way in order to not use BOOTP,I have been making 
CYGSEM_REDBOOT_DEFAULT_NO_BOOTP option available in redboot.cdl.

From here on, my explanation is vagued by bad memory.
About 2 years ago,I succeeded to ran the same nc_test_slave on SH target 
using Insight.
I confirmed I could step through the source codes,so I could debug network 
program on Insight.
At that time,I think I did not change CYGDBG_HAL_DEBUG_GDB_INCLUDE_STUBS to 
1,so maybe there are other ways to run nc_test_slave on Insight not by 
manipulating CYGDBG_HAL_DEBUG_GDB_INCLUDE_STUBS option.
I am not sure.

I am much obliged your help.

m mariga


-- 
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] cyg_scheduler_lock and DSR's

[Bart Veer]

>>>>> "Jim" == Jim Bradleigh <jim.bradleigh1@btinternet.com> writes:

    Jim> Is the cyg_scheduler_lock() call not supposed to prevent
    Jim> DSR's from running?

Yes, and no.    
    
    Jim> I have a alarm, attached to the realtime clock which when
    Jim> activated posts a message to a custom queue system. This
    Jim> queue is also posted to by user land code..

    Jim> Pushes to the queue are protected by a mutex. Additionally
    Jim> around the mutex I have the cyg_scheduler_lock() and
    Jim> cyg_scheduler_unlock() to prevent the DSR posting a message
    Jim> whilst inside the mutex operation.

    Jim> However I`m seeing asserts stating the mutex is already
    Jim> locked. A quick breakpoint and I can see that the DSR is
    Jim> still being called even after the lock() function..

The purpose of cyg_scheduler_lock() is to lock the scheduler, as the
name implies. That means no other threads get to run while the current
thread is runnable (on single-processor systems anyway, it gets a bit
more complicated on SMP systems). A side effect of locking the
scheduler is that DSRs cannot be allowed to run, because a DSR must
run to completion and may manipulate the scheduler.

However, if a thread locks the scheduler and then tries to claim a
mutex that is already owned by another thread, things get more
complicated. A strict interpretation of cyg_scheduler_lock() would
mean that control cannot transfer to that or any other thread, so it
would never get a chance to unlock the mutex, so the system would be
deadlocked.

Instead the correct interpretation of cyg_scheduler_lock() is that it
locks the scheduler while, and only while, the calling thread is
runnable. If the current thread becomes no longer runnable, e.g.
because it tries to lock a mutex that is already owned or because it
tries to wait on a synchronization object, then the current state of
the scheduler lock is saved away, the thread suspends as normal, and
other threads are allowed to run again. DSRs are allowed to run again
as well - there is no reason not to and the alternative would
introduce other possibilities for deadlock. Once the other thread
releases the mutex and your thread gets to run again, the state of the
scheduler lock gets restored so that the subsequent
cyg_scheduler_unlock() does the right thing.

I do not know the full details of your custom queue system. I suspect
there is no need for the mutex lock at all - locking the scheduler
provides all the required inter-thread synchronization to prevent
concurrent access to the queue object, as well as blocking DSRs. If a
mutex is still required for some reason, you'll probably want to lock
the mutex first and then the scheduler, not the other way around.

Bart

-- 
Bart Veer                                   eCos Configuration Architect
eCosCentric Limited    The eCos experts      http://www.ecoscentric.com/
Barnwell House, Barnwell Drive, Cambridge, UK.      Tel: +44 1223 245571
Registered in England and Wales: Reg No 4422071.
 >>>> Visit us at ESC-Boston  http://www.embedded.com/esc/boston <<<<
 >>>> Sep 22-23 on Stand 226  at Hynes Convention Center, Boston <<<<

-- 
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] Re: How to set GDB Stub Options in order to run insight ?

[m mariga]

[-- Attachment #1: Type: text/plain, Size: 3800 bytes --]

Hello,
I have known a little bit about Insight's(or GDB stub) not stopping at first 
breaking point(=cyg_start).
Also I like to ask you a new-born question.

First I re-state the problem.
Insight does not stop at first break point(=cyg_start).
I exchanged GDB commands on console window like these.
---
(gdb) set remotebaud 57600
(gdb) set remotelogfile remote-5.log
(gdb) target remote /dev/com4
Remote debugging using /dev/com4
0x000073a4 in ?? ()

(gdb) load
Loading section .text, size 0x5d58a lma 0x410000
Loading section .rodata, size 0x431e lma 0x46d58a
Loading section .data, size 0x16f8 lma 0x4718a8
Start address 0x410000, load size 405408
Transfer rate: 3 KB/sec, 63 bytes/write.

(gdb) b cyg_start
Note: breakpoints 1 (disabled), 2 (disabled), 3 (disabled) and 4 also set at 
pc 0x411dda.
Breakpoint 5 at 0x411dda: file nc_test_slave.c, line 729.

(gdb) c
Continuing.
Network stack using 69632 bytes for misc space
                    69632 bytes for mbufs
                    139264 bytes for mbuf clusters
[cyg_net_init] Init: mbinit(0x00000000)
[cyg_net_init] Init: cyg_net_init_devs(0x00000000)
Init device 'dp83902a_eth0'
DP83902A - eeprom ESA: 00:02:cb:01:42:b9
[cyg_net_init] Init: loopattach(0x00000000)
[cyg_net_init] Init: ifinit(0x00000000)
[cyg_net_init] Init: domaininit(0x00000000)
[cyg_net_init] Init: cyg_net_add_domain(0x004728a8)
New domain internet at 0x00000000
[cyg_net_init] Init: cyg_net_add_domain(0x00472298)
New domain route at 0x00000000
[cyg_net_init] Init: call_route_init(0x00000000)
[cyg_net_init] Done

At this point it should stop at break point(=cyg_start),but it never reply 
nothing.
---

I perused remote-5.log and learned GDB stub does not reply continue command.
Here are excerpts of remote-5.log.
---
c b cyg_start
c c
w $Z0,411dda,2#d3
r +$#00
w +$m411dda,2#8a
r +$1aa2#25
w +$X411dda,2:\x01\x80#30
r +$OK#9a
w +$vCont?#49
r +$#00
w +$Hc0#db
r +$OK#9a
w +$c#63
r 
+$O4E6574776F726B20737461636B207573696E6720363936333220627974657320666F72206D6973632073706163650A#0C
w +
   -- omitted --
r 
$O5B6379675F6E65745F696E69745D20496E69743A2063616C6C5F726F7574655F696E69742830783030303030303030290A#5D
w +
r $O5B6379675F6E65745F696E69745D20446F6E650A#2D
w +
r +
w \x03                ; I entered Ctr C.
r <Timeout: -1 seconds>
End of log
---

From here my explanation is from my instant knowlegde.Pardon if I 
misunderstood.
GDB stub should cause TRAP exception at break point(cyg_start) 
address(=411dda).
But there's no TRAP reply like T05.
Because normally GDB stub refers to vector table at starting 
address(=410000) and cause TRAP exception.
But there is no vector table at my starting address,so GDB could't cause 
TRAP exception.

My current mlt_h8300_h8300h_aki3068net_ram.ldi is like next.
---
SECTIONS
{
    SECTIONS_BEGIN
    SECTION_text (ram, 0x410000, LMA_EQ_VMA)
    SECTION_fini (ram, ALIGN (0x1), LMA_EQ_VMA)
    SECTION_rodata (ram, ALIGN (0x1), LMA_EQ_VMA)
---

I tentatively changed it like next.
---
SECTIONS
{
    SECTIONS_BEGIN
    SECTION_vectors (ram, 0x410000, LMA_EQ_VMA)
    SECTION_text (ram, ALIGN (0x4), LMA_EQ_VMA)
    SECTION_fini (ram, ALIGN (0x1), LMA_EQ_VMA)
---

I could built libtarget.a.
But Making of application caused next error.
h8300-elf-gcc -mh -mint32 -nostartfiles -L/home/link/aki-net-0831-ALIGN4-2/untitled_install/lib 
 -Wl,--gc-sections -Wl,--
Map -Wl,nc_test_slave.map -o nc_test_slave 
nc_test_slave.o -Ttarget.ld -nostdlib
/gnutools/lib/gcc-lib/h8300-elf/3.3.2/../../../../h8300-elf/bin/ld:target.ld:26: 
parse error
collect2: ld returned 1 exit status
make: *** [nc_test_slave] Error 1

If my deduction is correct,how should I deploy vector table at staring 
address(=0x410000) ?
I sent together log file,map file,and current rom.ldi ram.ldi rom.h ram.h 
files.

Would you please help me ?

m mariga

[-- Attachment #2: sendfiles.tar.bz2 --]
[-- Type: application/octet-stream, Size: 44898 bytes --]

[-- Attachment #3: Type: text/plain, Size: 148 bytes --]

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

[ECOS] About h8000 eCos GDB-Stub

[m mariga]

Hello,

I am still struggling to resolve why not breaking at cyg_start in Insight.
As I am further involved,the problem's aspect became more about eCos's h8000 
GDB-Stub,
so let me change my Mailing list subject title from "How to set GDB Stub 
Options in order to run insight ?" to "About h8000 eCos GDB-Stub".

First I re-state my problem.
I am working on eCos 3.0 using Akizuki H8/3068 Network board.
That board's CPU is H8/3069F.
I succeeded to install FreeBSD stack and ran TCP/IP test programs.
But when I tried to use Insight or GDB,it does not stop at cyg_start 
breaking point.

I start up Insight like next line.
$ h8300-elf-insight.exe nc_test_slave.elf

--- GDB commands log on console window
(gdb) set remotebaud 57600
(gdb) set remotelogfile remote-5.log
(gdb) target remote /dev/com4
Remote debugging using /dev/com4
0x000073a4 in ?? ()

(gdb) load
Loading section .text, size 0x5d58a lma 0x410000
Loading section .rodata, size 0x431e lma 0x46d58a
Loading section .data, size 0x16f8 lma 0x4718a8
Start address 0x410000, load size 405408
Transfer rate: 3 KB/sec, 63 bytes/write.

(gdb) b cyg_start
Note: breakpoints 1 (disabled), 2 (disabled), 3 (disabled) and 4 also set at 
pc 0x411dda.
Breakpoint 5 at 0x411dda: file nc_test_slave.c, line 729.

(gdb) c
Continuing.
Network stack using 69632 bytes for misc space
                    69632 bytes for mbufs
                    139264 bytes for mbuf clusters
[cyg_net_init] Init: mbinit(0x00000000)
[cyg_net_init] Init: cyg_net_init_devs(0x00000000)
Init device 'dp83902a_eth0'
DP83902A - eeprom ESA: 00:02:cb:01:42:b9
[cyg_net_init] Init: loopattach(0x00000000)
[cyg_net_init] Init: ifinit(0x00000000)
[cyg_net_init] Init: domaininit(0x00000000)
[cyg_net_init] Init: cyg_net_add_domain(0x004728a8)
New domain internet at 0x00000000
[cyg_net_init] Init: cyg_net_add_domain(0x00472298)
New domain route at 0x00000000
[cyg_net_init] Init: call_route_init(0x00000000)
[cyg_net_init] Done

At this point it should stop at break point(=cyg_start),but it never reply 
nothing.
---end GDB commands log on console window

I perused remote-5.log and learned GDB-Stub does not operate properly..
---excerpts of remote-5.log
c b cyg_start
c c
w $Z0,411dda,2#d3
r +$#00
w +$m411dda,2#8a
r +$1aa2#25
w +$X411dda,2:\x01\x80#30
r +$OK#9a
w +$vCont?#49
r +$#00
w +$Hc0#db
r +$OK#9a
w +$c#63
r 
+$O4E6574776F726B20737461636B207573696E6720363936333220627974657320666F72206D6973632073706163650A#0C
w +
   -- omitted --
r 
$O5B6379675F6E65745F696E69745D20496E69743A2063616C6C5F726F7574655F696E69742830783030303030303030290A#5D
w +
r $O5B6379675F6E65745F696E69745D20446F6E650A#2D
w +
r +
w \x03                ; I entered Ctr C.
r <Timeout: -1 seconds>
---end of excerpts of remote-5.log

When GDB send $Z0, GDB-Stub should replace address's code with TRAP,but in 
case of h8000 GDB-Stub does nothing.
So GDB try to replace address's code by $X411dda,2:\x01\x80.
But \x01\x80 is never H8's undefined codes but Sleep Command,so CPU sleeps 
at top of cyg_start and never stops.

I studied \ecos-3.0\packages\hal\common\v3_0\src\generic-stub.c,and found 
__process_packet() function looked likely place.
It's 1494-1566 lines read,
#if defined(CYGNUM_HAL_BREAKPOINT_LIST_SIZE) && 
(CYGNUM_HAL_BREAKPOINT_LIST_SIZE > 0)
    case 'Z':
      is_Z = 1;
    case 'z':
      {
 char *ptr = &packet[1];
 target_register_t ztype, addr, length;

There was no CYGNUM_HAL_BREAKPOINT_LIST_SIZE option in h8000 cdl file.
I tentatively commented out defined(CYGNUM_HAL_BREAKPOINT_LIST_SIZE) line 
and ran,but there was no change in GDB command's log.
Apparantly there are something missing.Would you please teach me the best 
way to insert $Z0 operation in h8000 GDB-Stub ?

Also,although I inserted diag_printf() in __process_packet function, nothing 
appeared on console window.
I do not know why diag_printf() did nothing,or am I looking wrong place at 
all ?
Please teach me how to trace codes in generic-stub.c. Are there any way to 
display on Insight console window ?

Thank you in advance.

m mariga


-- 
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] About h8000 eCos GDB-Stub

[m mariga]

Hello,

Since I could not think of any means to trace GDB-Stub program 
(generic-stub.c etc.),
I am in a serious predicament.
Would you please guide me ?

First of all, is it possible to use diag_printf() in generic-stub.c for 
debugging ?

If it is possible, then, there must be some reasons that I could not display 
any string of diag_printf()
on Insight's Console window.
Would you teach me some causes and resolving ways of malfunction ?

If it is not possible to use diag_printf() in generic-stub.c,
then I think there should be any means to debug generic-stub.c etc.
Please teach me that means.

Since my debugging tool is only Insight I would like to display messages on 
Insight's Console window.

I beseech you to help me.

m mariga



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

[ECOS] How to connect RedBoot via Telnet running application

[m mariga]

Hello,

I'd like to connect to RedBoot via Telnet during target's running net 
application.
I tried to run nc_test_slave program as application and connect to RedBoot 
via Telnet from host's cygwin.
The result is below.

$ telnet 172.16.1.200 9000
Trying 172.16.1.200...
telnet: Unable to connect to remote host: Connection refused

I append application running serial consol log below.

I am certain that the next excerpt from Anthony J. Massa's book is related 
to the cause.

--from Anthony J. Massa's book
NOTE:When using the Ethernet port for RedBoot communication
and in your application, it is important to set up the IP
addresses properly. The system distinguishes the destination of
a packet using the IP address. If the IP address for RedBoot and
the application are the same, RedBoot will receive all incoming
Ethernet packets, not passing them on to the application for processing.
To avoid this situation, it is best to use a static IP
address in one image and a separate static or dynamic image in
the other image.

I only set one IP address in CYGHWR_NET_DRIVER_ETH0_ADDRS_IP option like 
below.
cdl_option CYGHWR_NET_DRIVER_ETH0_ADDRS_IP {
    default_value { "172.16.1.200" }
}

But I do not know how to set IP addresses separately.
Right now I disables BOOTP by option mainly because I do not know how to use 
it.
I would rather prefer the two static IPs without using BOOTP or DHCP.
How should I set static IP addresses between RedBoot and application without 
using BOOTP ?
I presumed that RedBoot's IP is setted by fconfig,but I could not rewrite 
Local IP address.
Whenever I tried to rewrite Local IP address,or Gateway IP address,it 
happened 'command abort' message like below.
I do not know what is wrong.Please help me.
Am I correct thinking if I rewite Local IP address by fconfig,I could set 
RedBoot IP address differently with application address ?

RedBoot> fconfig
scan_opts: first=1, argc=1
Run script at boot: true
Boot script:
Enter script, terminate with empty line
>>
Boot script timeout (1000ms resolution): 3
Use BOOTP for network configuration: false
Gateway IP address: 172.16.1.1
Local IP address: 172.16.1.100
** command abort - illegal memory access?
RedBoot>

-- serial consol log begin
My Flash ID is 4:22f9:0:19
eth_drv_init:enaddr=0x8c000600
Ethernet eth0: MAC address 00:40:31:08:01:00
IP: 172.16.1.200/255.255.255.0, Gateway: 172.16.1.1
Default server: 172.16.1.1

RedBoot(tm) bootstrap and debug environment [ROM]
Non-certified release, version UNKNOWN - built 11:52:21, May 15 2008

Platform: inserter (SH 7709S)
Copyright (C) 2000, 2001, 2002, 2003, 2004 Red Hat, Inc.
Copyright (C) 2003, 2004, 2005, 2006 eCosCentric Limited

RAM: 0x8c000000-0x90000000, [0x8c00aa48-0x8ffed000] available
FLASH: 0xa0000000 - 0xa0400000, 64 blocks of 0x00010000 bytes each.
== Executing boot script in 3.000 seconds - enter ^C to abort
RedBoot>
RedBoot> load -v -m xmodem -b 0x8c410000 -c 0
CAddress offset = 0x00400000
Entry point: 0x8c410000, address range: 0x8c410000-0x8c45b180
xyzModem - CRC mode, 7060(SOH)/0(STX)/0(CAN) packets, 6 retries
RedBoot> fis create nc_slave -r 0x8c010000 -f 0xa0080000 -b 0x8c410000 -l 
0x4b180
1 flash_addr=a0080000
An image named 'nc_slave' exists - continue (y/n)? y
... Erase from 0xa0080000-0xa00d0000: .....
... Program from 0x8c410000-0x8c45b180 at 0xa0080000: .....
... Erase from 0xa03f0000-0xa0400000: .
... Program from 0x8fff0000-0x90000000 at 0xa03f0000: .
RedBoot> My Flash ID is 4:22f9:0:19
eth_drv_init:enaddr=0x8c000600
Ethernet eth0: MAC address 00:40:31:08:01:00
IP: 172.16.1.200/255.255.255.0, Gateway: 172.16.1.1
Default server: 172.16.1.1

RedBoot(tm) bootstrap and debug environment [ROM]
Non-certified release, version UNKNOWN - built 11:52:21, May 15 2008

Platform: inserter (SH 7709S)
Copyright (C) 2000, 2001, 2002, 2003, 2004 Red Hat, Inc.
Copyright (C) 2003, 2004, 2005, 2006 eCosCentric Limited

RAM: 0x8c000000-0x90000000, [0x8c00aa48-0x8ffed000] available
FLASH: 0xa0000000 - 0xa0400000, 64 blocks of 0x00010000 bytes each.
== Executing boot script in 3.000 seconds - enter ^C to abort
RedBoot> fis load -b 0x8c010000 nc_slave
RedBoot> channel 0
RedBoot> go 0x8c010000
Network stack using 69632 bytes for misc space
                    69632 bytes for mbufs
                    139264 bytes for mbuf clusters
[cyg_net_init] Init: mbinit(0x00000000)
[cyg_net_init] Init: cyg_net_init_devs(0x00000000)
Init device 'lan91cxx_eth0'
[cyg_net_init] Init: loopattach(0x00000000)
[cyg_net_init] Init: ifinit(0x00000000)
[cyg_net_init] Init: domaininit(0x00000000)
[cyg_net_init] Init: cyg_net_add_domain(0x8c05ab18)
New domain internet at 0x00000000
[cyg_net_init] Init: cyg_net_add_domain(0x8c05a510)
New domain route at 0x00000000
[cyg_net_init] Init: call_route_init(0x00000000)
[cyg_net_init] Done
Start Network Characterization - SLAVE
BOOTP[eth0] op: REPLY
       htype: Ethernet
        hlen: 6
        hops: 0
         xid: 0x0
        secs: 0
       flags: 0x0
       hw_addr: 00:40:31:08:01:00
     client IP: 172.16.1.200
         my IP: 172.16.1.200
     server IP: 172.16.1.1
    gateway IP: 172.16.1.1
  options:
        subnet mask: 255.255.255.0
       IP broadcast: 172.16.1.255
            gateway: 172.16.1.1
[eth_drv_ioctl] Warning: Driver can't set multi-cast mode
[eth_drv_ioctl] Warning: Driver can't set multi-cast mode
[eth_drv_ioctl] Warning: Driver can't set multi-cast mode
No load = 58674
Set background load = 20%
Set background load = 0%
High Load[20] = 37130 => 37%
Set background load = 20%
Set background load = 0%
Load[10] = 47733 => 19%
Set background load = 20%
Set background load = 0%
Final load[10] = 47813 => 19%
Start test for eth0
-- serial consol log end

m mariga


-- 
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] How to connect RedBoot via Telnet running application

[Gary Thomas]

On 10/16/2009 08:33 AM, m mariga wrote:
> Hello,
>
> I'd like to connect to RedBoot via Telnet during target's running net
> application.
> I tried to run nc_test_slave program as application and connect to
> RedBoot via Telnet from host's cygwin.
> The result is below.
>
> $ telnet 172.16.1.200 9000
> Trying 172.16.1.200...
> telnet: Unable to connect to remote host: Connection refused
   <...snip>

Once your application is started, RedBoot is no longer available.

Why are you trying to do this?

-- 
------------------------------------------------------------
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

Re: [ECOS] How to connect RedBoot via Telnet running application

[Gary Thomas]

Please keep your replies on the mailing list so that all may benefit.

On 10/16/2009 09:07 AM, m mariga wrote:
> Hello,
>
> Gary wrote
>>> I'd like to connect to RedBoot via Telnet during target's running net
>>> application.
>>> I tried to run nc_test_slave program as application and connect to
>>> RedBoot via Telnet from host's cygwin.
>>> The result is below.
>>>
>>> $ telnet 172.16.1.200 9000
>>> Trying 172.16.1.200...
>>> telnet: Unable to connect to remote host: Connection refused
>> <...snip>
>>
>> Once your application is started, RedBoot is no longer available.
>>
>> Why are you trying to do this?
>>
>
> I would like to load new application object to target using RedBoot via
> Telnet from host PC during target's application running.
> Is it impossible ?

It is not possible.  As I said, once your application is running,
RedBoot is out of the picture.

> I got an impression from Anthony J. Massa's book that if setted two
> different IP addresses between RedBoot and application,I could connect
> to RedBoot via Telnet during application running. Am I mistaken ?

That's a reference to the ability of RedBoot to provide GDB
services via the network.  The rest of RedBoot's functionality
is no longer available, once your application starts.

-- 
------------------------------------------------------------
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

Re: [ECOS] How to connect RedBoot via Telnet running application

[m mariga]

Hello,Gary,and everyone,

Thank you, Gary, I got it.
Although I think it is too bad not to be able to use RedBoot's monitor 
function like memory dump during application running.

By the way,I do not know why I can't rewrite Local IP address or Gateway IP 
address by fconfig.
Please suggest my mistakes.

RedBoot> fconfig
scan_opts: first=1, argc=1
Run script at boot: true
Boot script:
Enter script, terminate with empty line
>>
Boot script timeout (1000ms resolution): 3
Use BOOTP for network configuration: false
Gateway IP address: 172.16.1.1
Local IP address: 172.16.1.100
** command abort - illegal memory access?
RedBoot>

m mariga


-- 
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] How to connect RedBoot via Telnet running application

[m mariga]

Hello,Gary,and everyone,

I am very sorry.
I dare your anger for my obstination,but I must confirm next point.

Gary wrote,
>> I would like to load new application object to target using RedBoot via
>> Telnet from host PC during target's application running.
>> Is it impossible ?
>
> It is not possible.  As I said, once your application is running,
> RedBoot is out of the picture.
>

Am I right in getting you that even I use BOOTP or DHCP to get  dynamic IP 
address for application and assign other staic IP address for RedBoot,I 
could not connect to RedBoot via Telnet during application running ?

m mariga


-- 
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] How to connect RedBoot via Telnet running application

[Gary Thomas]

On 10/16/2009 10:43 AM, m mariga wrote:
> Hello,Gary,and everyone,
>
> I am very sorry.
> I dare your anger for my obstination,but I must confirm next point.
>
> Gary wrote,
>>> I would like to load new application object to target using RedBoot via
>>> Telnet from host PC during target's application running.
>>> Is it impossible ?
>>
>> It is not possible. As I said, once your application is running,
>> RedBoot is out of the picture.
>>
>
> Am I right in getting you that even I use BOOTP or DHCP to get dynamic
> IP address for application and assign other staic IP address for
> RedBoot,I could not connect to RedBoot via Telnet during application
> running ?

Once your application starts, RedBoot is GONE.  The only way
to return to it is to exit your application, typically by
resetting the hardware.

-- 
------------------------------------------------------------
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