From jean-christophe.barriol at rsautomation.com  Thu Mar  6 17:03:03 2008
From: jean-christophe.barriol at rsautomation.com (Jean Christophe BARRIOL)
Date: Thu, 6 Mar 2008 18:03:03 +0100
Subject: debugger problem with m68k-elf 4.2.47
Message-ID: <004901c87fab$f14b05a0$82bebfc0@JCB2>

Hello,

With m68k-elf 4.1.32  I succed in debugging my application ( I can insert breakpoint and step throuth the code)

But with m68k-elf 4.2.47, It don't work: I can't insert breakpoint, I get an error:
No symbol table is loaded. Use the "file" command.
I work on the same project, with the same options!

Any Idea ?
Is there something new to do?, a new option for GCC or GDB ?






*******************************
Jean Christophe BARRIOL
R.S.A.I.
Rue des Mineurs
ZI de la Vaure
BP 30940
42290 Sorbiers
t?l : +33 (0)4-77-53-30-48
Fax : +33 (0)4-77-53-38-61
Internet: www.rsautomation.com
*******************************
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sourcerytools.com/pipermail/coldfire-gnu-discuss/attachments/20080306/da550a71/attachment.html>

From carlos at codesourcery.com  Thu Mar  6 19:16:34 2008
From: carlos at codesourcery.com (Carlos O'Donell)
Date: Thu, 06 Mar 2008 14:16:34 -0500
Subject: [coldfire-gnu-discuss] debugger problem with m68k-elf 4.2.47
In-Reply-To: <004901c87fab$f14b05a0$82bebfc0@JCB2>
References: <004901c87fab$f14b05a0$82bebfc0@JCB2>
Message-ID: <47D04312.8000601@codesourcery.com>

Jean Christophe BARRIOL wrote:
> Hello,
>  
> With m68k-elf 4.1.32  I succed in debugging my application ( I can 
> insert breakpoint and step throuth the code)
>  
> But with m68k-elf 4.2.47, It don't work: I can't insert breakpoint, I 
> get an error:
> *No symbol table is loaded. Use the "file" command.*
> I work on the same project, with the same options!
>  
> Any Idea ?

How did you invoke the debugger?

Cheers,
Carlos.
-- 
Carlos O'Donell
CodeSourcery
carlos at codesourcery.com
(650) 331-3385 x716


From bob.brusa at gmail.com  Fri Mar  7 09:24:06 2008
From: bob.brusa at gmail.com (Robert Brusa)
Date: Fri, 07 Mar 2008 10:24:06 +0100
Subject: debugging does not work
Message-ID: <1204881846.5427.26.camel@debi27.local.lan>

Hi all,

I have set up the factorial-example as described in chapter 5 of the
document "Getting_Started_Guide.pdf". The program compiles and I also
succeeded to mount the program-folder on my target (using nfs). Then I
start the program there (on the target) using the command
"./factorial<enter>" That works fine. Then I proceed, setting up the
debug configuration, but when I start the debugger by clicking on the
"debug" symbol in the IDE, the perspective does not change to
"debugging", instead, I get an error:

> Error creating session:local host 10000: Connection refused

I am working on a debian etch host. The target is a COBRA5282-Board with
uClinux, kernel 2.6 running. Because of an error in the IDE (the MCF5282
is not available) one must use a trick in the configuration. I applied
this workaround (which I got from CodeSourcery). Could this same
IDE-problem now also be the root of my debug problem? The
Getting-Started_Guide writes about a gdbserver. However, there is no
gdbserver available for the MCF5282. Which one does the IDE use? Or must
it be launched earlier (by the user)? And if yes how?

I hope someone has answers that make my further debugging a success.
Thank you and best regards
Bob
 
-- 
------------------------------
Bob Brusa
BRUSAG Sensorik &
messtechn. Entwicklungen
Chapfwiesenstr. 14
CH-8712 Staefa (Switzerland)
bob.brusa at gmail.com
+41 44 926 74 74
www.brusag.ch



From nathan at codesourcery.com  Fri Mar  7 09:36:00 2008
From: nathan at codesourcery.com (Nathan Sidwell)
Date: Fri, 07 Mar 2008 09:36:00 +0000
Subject: [coldfire-gnu-discuss] debugging does not work
In-Reply-To: <1204881846.5427.26.camel@debi27.local.lan>
References: <1204881846.5427.26.camel@debi27.local.lan>
Message-ID: <47D10C80.6050605@codesourcery.com>

Robert,

> "./factorial<enter>" That works fine. Then I proceed, setting up the
> debug configuration, but when I start the debugger by clicking on the
> "debug" symbol in the IDE, the perspective does not change to
> "debugging", instead, I get an error:

Can you clarify. Are you using Lite with some IDE, or are you using Sourcery G++?

nathan

-- 
Nathan Sidwell    ::   http://www.codesourcery.com   ::         CodeSourcery



From bob.brusa at gmail.com  Fri Mar  7 10:03:22 2008
From: bob.brusa at gmail.com (Robert Brusa)
Date: Fri, 07 Mar 2008 11:03:22 +0100
Subject: [coldfire-gnu-discuss] debugging does not work
In-Reply-To: <47D10C80.6050605@codesourcery.com>
References: <1204881846.5427.26.camel@debi27.local.lan>
	 <47D10C80.6050605@codesourcery.com>
Message-ID: <1204884202.10657.1.camel@debi27.local.lan>

Am Freitag, den 07.03.2008, 09:36 +0000 schrieb Nathan Sidwell:
> Robert,
> 
> > "./factorial<enter>" That works fine. Then I proceed, setting up the
> > debug configuration, but when I start the debugger by clicking on the
> > "debug" symbol in the IDE, the perspective does not change to
> > "debugging", instead, I get an error:
> 
> Can you clarify. Are you using Lite with some IDE, or are you using Sourcery G++?
> 
> nathan
> 
Hi Nathan
I am using sourcery G++. Its help menu provides the following
information:
Sourcery G++ IDE Version: 4.2-54

Regards Bob



From bob.brusa at gmail.com  Fri Mar  7 17:26:34 2008
From: bob.brusa at gmail.com (Robert Brusa)
Date: Fri, 07 Mar 2008 18:26:34 +0100
Subject: Problem Debugging in Sourcery_GXX_IDE
Message-ID: <1204910794.3862.13.camel@debi27.local.lan>

Hi all

further to my previous cry for help, I have have some news.

The uClinux I am running on my COBRA5282-board includes a gdbserver. So
I firstly connect my working directory to the target using ntfs, then I
start the gdbserver on the target. This looks as follows (I am using
minicom on a host running debian etch)

> /> mount -t nfs -o rsize=4096,wsize=4096
> 192.168.0.55:/home/rwb/ws1 /usr
> /> gdbserver localhost:10000 /usr/factorial/Debug/factorial
> Process /usr/factorial/Debug/factorial created; pid = 35
> code at 0x3c8040 - 0x3cbd20, data at 0x3cbd24

Then I go back to my Sourcery_GXX_IDE with the factorial example and
start a debug-session by clicking on the debug-icon and get the message

Error creating session: localhost:10000: Connection refused

What's wrong? The IDE is configured as explained in the documentation
and still this error.

Thanks for any advice and best regards
Bob

-- 
------------------------------
Bob Brusa
BRUSAG Sensorik &
messtechn. Entwicklungen
Chapfwiesenstr. 14
CH-8712 Staefa (Switzerland)
bob.brusa at gmail.com
+41 44 926 74 74
www.brusag.ch



From nathan at codesourcery.com  Fri Mar  7 17:35:03 2008
From: nathan at codesourcery.com (Nathan Sidwell)
Date: Fri, 07 Mar 2008 17:35:03 +0000
Subject: [coldfire-gnu-discuss] Problem Debugging in Sourcery_GXX_IDE
In-Reply-To: <1204910794.3862.13.camel@debi27.local.lan>
References: <1204910794.3862.13.camel@debi27.local.lan>
Message-ID: <47D17CC7.9020504@codesourcery.com>

Robert Brusa wrote:
> Hi all
> 
> further to my previous cry for help, I have have some news.
> 
> The uClinux I am running on my COBRA5282-board includes a gdbserver. So
> I firstly connect my working directory to the target using ntfs, then I
> start the gdbserver on the target. This looks as follows (I am using
> minicom on a host running debian etch)
> 
>> /> mount -t nfs -o rsize=4096,wsize=4096
>> 192.168.0.55:/home/rwb/ws1 /usr
>> /> gdbserver localhost:10000 /usr/factorial/Debug/factorial
>> Process /usr/factorial/Debug/factorial created; pid = 35
>> code at 0x3c8040 - 0x3cbd20, data at 0x3cbd24

gdbserver is listening on localhost:10000 -- the target board

> 
> Then I go back to my Sourcery_GXX_IDE with the factorial example and
> start a debug-session by clicking on the debug-icon and get the message
> 
> Error creating session: localhost:10000: Connection refused

gdb is trying to talk to localhost:10000 -- the host system

nathan
-- 
Nathan Sidwell    ::   http://www.codesourcery.com   ::         CodeSourcery



From bob.brusa at gmail.com  Sat Mar  8 10:36:23 2008
From: bob.brusa at gmail.com (Robert Brusa)
Date: Sat, 08 Mar 2008 11:36:23 +0100
Subject: [coldfire-gnu-discuss] Problem Debugging in Sourcery_GXX_IDE
In-Reply-To: <47D17CC7.9020504@codesourcery.com>
References: <1204910794.3862.13.camel@debi27.local.lan>
	 <47D17CC7.9020504@codesourcery.com>
Message-ID: <1204972583.5842.4.camel@debi27.local.lan>

Am Freitag, den 07.03.2008, 17:35 +0000 schrieb Nathan Sidwell:
> Robert Brusa wrote:
> > Hi all
> > 
> > further to my previous cry for help, I have have some news.
> > 
> > The uClinux I am running on my COBRA5282-board includes a gdbserver. So
> > I firstly connect my working directory to the target using ntfs, then I
> > start the gdbserver on the target. This looks as follows (I am using
> > minicom on a host running debian etch)
> > 
> >> /> mount -t nfs -o rsize=4096,wsize=4096
> >> 192.168.0.55:/home/rwb/ws1 /usr
> >> /> gdbserver localhost:10000 /usr/factorial/Debug/factorial
> >> Process /usr/factorial/Debug/factorial created; pid = 35
> >> code at 0x3c8040 - 0x3cbd20, data at 0x3cbd24
> 
> gdbserver is listening on localhost:10000 -- the target board
> 
> > 
> > Then I go back to my Sourcery_GXX_IDE with the factorial example and
> > start a debug-session by clicking on the debug-icon and get the message
> > 
> > Error creating session: localhost:10000: Connection refused
> 
> gdb is trying to talk to localhost:10000 -- the host system
> 
> nathan
Nathan - sorry, but I do not understand what your comments try to tell
me. What exactly should I do to make it work?
Regards Bob.



From maxim at codesourcery.com  Sat Mar  8 17:36:39 2008
From: maxim at codesourcery.com (Maxim Kuvyrkov)
Date: Sat, 08 Mar 2008 20:36:39 +0300
Subject: [coldfire-gnu-discuss] Problem Debugging in Sourcery_GXX_IDE
In-Reply-To: <1204972583.5842.4.camel@debi27.local.lan>
References: <1204910794.3862.13.camel@debi27.local.lan>	 <47D17CC7.9020504@codesourcery.com> <1204972583.5842.4.camel@debi27.local.lan>
Message-ID: <47D2CEA7.9010509@codesourcery.com>

Robert Brusa wrote:
> Am Freitag, den 07.03.2008, 17:35 +0000 schrieb Nathan Sidwell:
>> Robert Brusa wrote:
>>> Hi all
>>>
>>> further to my previous cry for help, I have have some news.
>>>
>>> The uClinux I am running on my COBRA5282-board includes a gdbserver. So
>>> I firstly connect my working directory to the target using ntfs, then I
>>> start the gdbserver on the target. This looks as follows (I am using
>>> minicom on a host running debian etch)
>>>
>>>> /> mount -t nfs -o rsize=4096,wsize=4096
>>>> 192.168.0.55:/home/rwb/ws1 /usr
>>>> /> gdbserver localhost:10000 /usr/factorial/Debug/factorial
>>>> Process /usr/factorial/Debug/factorial created; pid = 35
>>>> code at 0x3c8040 - 0x3cbd20, data at 0x3cbd24
>> gdbserver is listening on localhost:10000 -- the target board
>>
>>> Then I go back to my Sourcery_GXX_IDE with the factorial example and
>>> start a debug-session by clicking on the debug-icon and get the messag
>>>
>>> Error creating session: localhost:10000: Connection refused
>> gdb is trying to talk to localhost:10000 -- the host system
>>
>> nathan
> Nathan - sorry, but I do not understand what your comments try to tell
> me. What exactly should I do to make it work?

 From what you've posted, I understand that you configured IDE debugger 
to connect to your local system, instead of your board.  You should 
specify your board's name or IP address in the IDE's debugger configuration.

E.g., on the board you should run

$ gdbserver :10000 myprog

and on your host system gdb should connect to myboard:10000 .

--
Maxim



From bob.brusa at gmail.com  Mon Mar 10 09:57:03 2008
From: bob.brusa at gmail.com (Robert Brusa)
Date: Mon, 10 Mar 2008 10:57:03 +0100
Subject: [coldfire-gnu-discuss] Problem Debugging in Sourcery_GXX_IDE
In-Reply-To: <47D2CEA7.9010509@codesourcery.com>
References: <1204910794.3862.13.camel@debi27.local.lan>
	 <47D17CC7.9020504@codesourcery.com>
	 <1204972583.5842.4.camel@debi27.local.lan>
	 <47D2CEA7.9010509@codesourcery.com>
Message-ID: <1205143024.3144.24.camel@debi27.local.lan>

Am Samstag, den 08.03.2008, 20:36 +0300 schrieb Maxim Kuvyrkov:
> Robert Brusa wrote:
> > Am Freitag, den 07.03.2008, 17:35 +0000 schrieb Nathan Sidwell:
> >> Robert Brusa wrote:
> >>> Hi all
> >>>
> >>> further to my previous cry for help, I have have some news.
> >>>
> >>> The uClinux I am running on my COBRA5282-board includes a gdbserver. So
> >>> I firstly connect my working directory to the target using ntfs, then I
> >>> start the gdbserver on the target. This looks as follows (I am using
> >>> minicom on a host running debian etch)
> >>>
> >>>> /> mount -t nfs -o rsize=4096,wsize=4096
> >>>> 192.168.0.55:/home/rwb/ws1 /usr
> >>>> /> gdbserver localhost:10000 /usr/factorial/Debug/factorial
> >>>> Process /usr/factorial/Debug/factorial created; pid = 35
> >>>> code at 0x3c8040 - 0x3cbd20, data at 0x3cbd24
> >> gdbserver is listening on localhost:10000 -- the target board
> >>
> >>> Then I go back to my Sourcery_GXX_IDE with the factorial example and
> >>> start a debug-session by clicking on the debug-icon and get the messag
> >>>
> >>> Error creating session: localhost:10000: Connection refused
> >> gdb is trying to talk to localhost:10000 -- the host system
> >>
> >> nathan
> > Nathan - sorry, but I do not understand what your comments try to tell
> > me. What exactly should I do to make it work?
> 
>  From what you've posted, I understand that you configured IDE debugger 
> to connect to your local system, instead of your board.  You should 
> specify your board's name or IP address in the IDE's debugger configuration.
> 
> E.g., on the board you should run
> 
> $ gdbserver :10000 myprog
> 
> and on your host system gdb should connect to myboard:10000 .
> 
> --
> Maxim
> 
Hi,
Your input helped. Now the two systems connect. But I get an error - see
attachment Error1.txt.

What I have done is the following:
On the serial terminal (mini) connected to my COBRA5282 board I enter
the commands:

> /> mount -t nfs -o rsize=4096,wsize=4096 192.168.0.55:/home/rwb/ws1 /usr
> /> gdbserver :10000 /usr/factorial/Debug/factorial

Then I launch Sourcery_GXX_IDE with the program factorial. There I had
(previously) configured the debug connection - see attachment DebugCfg.png.
When the IDE ist up and ready, I run the program with F11 (debug last launched)
now the error-message pops up.

All this seems overly complicated. I would expect the gdbserver to be capable
of downloading the program from the host. After all, the two systems are connected -
even twice: serial link and LAN. Why do I have to tell the target which program
to debug? Why doesn' the IDE tell this to the target? I doubt that I am on the 
right track. However, the documentation delivered by codesourcery does not give
any clue about these details. Any help around?
Thank you and best regards
Bob

-- 
------------------------------
Bob Brusa
BRUSAG Sensorik &
messtechn. Entwicklungen
Chapfwiesenstr. 14
CH-8712 Staefa (Switzerland)
bob.brusa at gmail.com
+41 44 926 74 74
www.brusag.ch
-------------- next part --------------
Execution is suspended because of error. Remote 'g' packet reply is too long: 003ce3f001000000000000000000000000000001003cbd240025dec80025deb4003c813a003cbcec003cc398003cff6c003cbcf000530004003cff18003cff1000000000003c8140000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
  Remote 'g' packet reply is too long: 003ce3f001000000000000000000000000000001003cbd240025dec80025deb4003c813a003cbcec003cc398003cff6c003cbcf000530004003cff18003cff1000000000003c81400000000000000000000
-------------- next part --------------
A non-text attachment was scrubbed...
Name: DebugCfg.png
Type: image/png
Size: 142158 bytes
Desc: not available
URL: <http://sourcerytools.com/pipermail/coldfire-gnu-discuss/attachments/20080310/be8a2d16/attachment.png>

From nathan at codesourcery.com  Mon Mar 10 10:14:07 2008
From: nathan at codesourcery.com (Nathan Sidwell)
Date: Mon, 10 Mar 2008 10:14:07 +0000
Subject: [coldfire-gnu-discuss] Problem Debugging in Sourcery_GXX_IDE
In-Reply-To: <1205143024.3144.24.camel@debi27.local.lan>
References: <1204910794.3862.13.camel@debi27.local.lan>	 <47D17CC7.9020504@codesourcery.com>	 <1204972583.5842.4.camel@debi27.local.lan>	 <47D2CEA7.9010509@codesourcery.com> <1205143024.3144.24.camel@debi27.local.lan>
Message-ID: <47D509EF.2040602@codesourcery.com>

Robert Brusa wrote:

> All this seems overly complicated. I would expect the gdbserver to be capable
> of downloading the program from the host. After all, the two systems are connected -
> even twice: serial link and LAN. Why do I have to tell the target which program
> to debug? Why doesn' the IDE tell this to the target? I doubt that I am on the 
> right track. However, the documentation delivered by codesourcery does not give
> any clue about these details. Any help around?

This is not functionality that gdbserver supports.

nathan

-- 
Nathan Sidwell    ::   http://www.codesourcery.com   ::         CodeSourcery



From jean-christophe.barriol at rsautomation.com  Wed Mar 12 15:24:58 2008
From: jean-christophe.barriol at rsautomation.com (Jean Christophe BARRIOL)
Date: Wed, 12 Mar 2008 16:24:58 +0100
Subject: Best libraries for mcf5234
Message-ID: <001301c88455$3d140d30$82bebfc0@JCB2>

Hello,

What are the best libraries for Coldfire MCF523x ?
The toolchain proposes m5206e, m5208, m5213 or m5249.

Thanks,

Jean Christophe.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sourcerytools.com/pipermail/coldfire-gnu-discuss/attachments/20080312/9f7285cc/attachment.html>

From nathan at codesourcery.com  Wed Mar 12 15:54:13 2008
From: nathan at codesourcery.com (Nathan Sidwell)
Date: Wed, 12 Mar 2008 15:54:13 +0000
Subject: [coldfire-gnu-discuss] Best libraries for mcf5234
In-Reply-To: <001301c88455$3d140d30$82bebfc0@JCB2>
References: <001301c88455$3d140d30$82bebfc0@JCB2>
Message-ID: <47D7FCA5.6060204@codesourcery.com>

Jean,
> What are the best libraries for Coldfire MCF523x ?
> The toolchain proposes m5206e, m5208, m5213 or m5249.

What do you mean 'the toolchain proposes ...'?  The appropriate libraries will 
be selected when you use the -mcpu=523x compilation option.

nathan

-- 
Nathan Sidwell    ::   http://www.codesourcery.com   ::         CodeSourcery



From bob.brusa at gmail.com  Wed Mar 12 16:08:10 2008
From: bob.brusa at gmail.com (Robert Brusa)
Date: Wed, 12 Mar 2008 17:08:10 +0100
Subject: math-functions: floor, sin, cos, atan, sqrt
Message-ID: <1205338090.28642.7.camel@debi27.local.lan>

Hi
I have written a program that uses the above mentioned math-functions.
At the very top of my source there is the statement:

> #include <stdio.h>
> #include <math.h>

However, I get errors each time I try to call one of these functions.
This triggers two questions?

1) What do I have to import to get these functions in?

2) Where is this type of stuff documented?

Thank you and best regards
Bob

-- 
Bob Brusa
BRUSAG Sensorik &
messtechn. Entwicklungen
Chapfwiesenstr. 14
CH-8712 Staefa (Switzerland)
bob.brusa at gmail.com
+41 44 926 74 74
www.brusag.ch



From nathan at codesourcery.com  Wed Mar 12 16:15:12 2008
From: nathan at codesourcery.com (Nathan Sidwell)
Date: Wed, 12 Mar 2008 16:15:12 +0000
Subject: [coldfire-gnu-discuss] math-functions: floor, sin, cos, atan,
 sqrt
In-Reply-To: <1205338090.28642.7.camel@debi27.local.lan>
References: <1205338090.28642.7.camel@debi27.local.lan>
Message-ID: <47D80190.8050708@codesourcery.com>

Robert Brusa wrote:
> Hi
> I have written a program that uses the above mentioned math-functions.
> At the very top of my source there is the statement:
> 
>> #include <stdio.h>
>> #include <math.h>

have you searched the knowledge base for 'math library'?

nathan

-- 
Nathan Sidwell    ::   http://www.codesourcery.com   ::         CodeSourcery



From bob.brusa at gmail.com  Wed Mar 12 17:22:39 2008
From: bob.brusa at gmail.com (Robert Brusa)
Date: Wed, 12 Mar 2008 18:22:39 +0100
Subject: [coldfire-gnu-discuss] math-functions: floor, sin, cos, atan,
	sqrt
In-Reply-To: <47D80190.8050708@codesourcery.com>
References: <1205338090.28642.7.camel@debi27.local.lan>
	 <47D80190.8050708@codesourcery.com>
Message-ID: <1205342559.32724.10.camel@debi27.local.lan>

Am Mittwoch, den 12.03.2008, 16:15 +0000 schrieb Nathan Sidwell:
> Robert Brusa wrote:
> > Hi
> > I have written a program that uses the above mentioned math-functions.
> > At the very top of my source there is the statement:
> > 
> >> #include <stdio.h>
> >> #include <math.h>
> 
> have you searched the knowledge base for 'math library'?
> 
> nathan

Well no, because I do not know where to enter this search string 'math
library'. When entering this search-string in Sourcery G++ IDE Version:
4.2-54 (via help-menu) I get 0 hits.
Bob



From bob.brusa at gmail.com  Wed Mar 12 17:27:16 2008
From: bob.brusa at gmail.com (Robert Brusa)
Date: Wed, 12 Mar 2008 18:27:16 +0100
Subject: [coldfire-gnu-discuss] math-functions: floor, sin, cos, atan,
	sqrt
In-Reply-To: <47D80190.8050708@codesourcery.com>
References: <1205338090.28642.7.camel@debi27.local.lan>
	 <47D80190.8050708@codesourcery.com>
Message-ID: <1205342836.32724.14.camel@debi27.local.lan>

Am Mittwoch, den 12.03.2008, 16:15 +0000 schrieb Nathan Sidwell:
> Robert Brusa wrote:
> > Hi
> > I have written a program that uses the above mentioned math-functions.
> > At the very top of my source there is the statement:
> > 
> >> #include <stdio.h>
> >> #include <math.h>
> 
> have you searched the knowledge base for 'math library'?
> 
> nathan

Well no - because I do not know where I should enter this search string.
When clicking on the help ? search of the Sourcery G++ IDE Version:
4.2-54 and entering this or similar strings I get 0 hits. Hence: Which
knowledge base do you mean?
Bob



From mark at codesourcery.com  Wed Mar 12 17:34:44 2008
From: mark at codesourcery.com (Mark Mitchell)
Date: Wed, 12 Mar 2008 10:34:44 -0700
Subject: [coldfire-gnu-discuss] math-functions: floor, sin, cos, atan,
 sqrt
In-Reply-To: <1205342836.32724.14.camel@debi27.local.lan>
References: <1205338090.28642.7.camel@debi27.local.lan>	 <47D80190.8050708@codesourcery.com> <1205342836.32724.14.camel@debi27.local.lan>
Message-ID: <47D81434.6020306@codesourcery.com>

Robert Brusa wrote:

> Well no - because I do not know where I should enter this search string.
> When clicking on the help ? search of the Sourcery G++ IDE Version:
> 4.2-54 and entering this or similar strings I get 0 hits. Hence: Which
> knowledge base do you mean?
> Bob

See:

   https://support.codesourcery.com/GNUToolchain/kbentry67

Thanks,

-- 
Mark Mitchell
CodeSourcery
mark at codesourcery.com
(650) 331-3385 x713


From haluongvn at gmail.com  Sat Mar 15 06:56:24 2008
From: haluongvn at gmail.com (Ha Luong)
Date: Sat, 15 Mar 2008 13:56:24 +0700
Subject: About space optimization
Message-ID: <61577c8f0803142356u7ecdbb1ekdf291f28c2ef965a@mail.gmail.com>

Dear Sir,

Could you please suggest the options for the best space optimization? I used
the compiler flag is "-mcpu=5282 -O3 -Os -msoft-float" and the linker flag
is "-Tm5282evb-ram-hosted.ld"

Thank you,
Ha Luong
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sourcerytools.com/pipermail/coldfire-gnu-discuss/attachments/20080315/34353485/attachment.html>

From carlos at codesourcery.com  Sat Mar 15 14:45:31 2008
From: carlos at codesourcery.com (Carlos O'Donell)
Date: Sat, 15 Mar 2008 10:45:31 -0400
Subject: [coldfire-gnu-discuss] About space optimization
In-Reply-To: <61577c8f0803142356u7ecdbb1ekdf291f28c2ef965a@mail.gmail.com>
References: <61577c8f0803142356u7ecdbb1ekdf291f28c2ef965a@mail.gmail.com>
Message-ID: <47DBE10B.2020806@codesourcery.com>

Ha Luong wrote:
> Dear Sir,
>  
> Could you please suggest the options for the best space optimization? I 
> used the compiler flag is "-mcpu=5282 -O3 -Os -msoft-float" and 
> the linker flag is "-Tm5282evb-ram-hosted.ld" 

Remove "-O3" and use only "-Os" for space optimizations.

Cheers,
Carlos.
-- 
Carlos O'Donell
CodeSourcery
carlos at codesourcery.com
(650) 331-3385 x716


From haluongvn at gmail.com  Mon Mar 24 07:32:23 2008
From: haluongvn at gmail.com (Ha Luong)
Date: Mon, 24 Mar 2008 14:32:23 +0700
Subject: The different size is gotten from m68k-elf-nm and m68k-elf-readelf
Message-ID: <61577c8f0803240032h1ee4a8a6o2cde5d9c568c1b0f@mail.gmail.com>

Dear Sir,

I used m68k-elf-readelf and m68k-elf-nm for calculating the elf size. But
the total size from m68k-elf-readelf is different to its from m68k-elf-nm.
I used
The total size from m68k-elf-readelf = the total of size column.
with the command line option "m68k-elf-readelf -S xxx.elf"

And The total size from m68k-elf-nm = the total of column 2
with the command line option "m68k-elf-nm.exe -S --print-size --print-armap
--size-sort xxx.elf"

I double checked by used m68k-elf-objdump by calculating each variable and
each function. I saw its total size is equal The total size from
m68k-elf-nm.

I don't know why the total size from m68k-elf-readelf is different to its
from m68k-elf-nm. Could you please help me?

Thanks,
Ha Luong
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sourcerytools.com/pipermail/coldfire-gnu-discuss/attachments/20080324/5c77af6f/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.zip
Type: application/zip
Size: 210347 bytes
Desc: not available
URL: <http://sourcerytools.com/pipermail/coldfire-gnu-discuss/attachments/20080324/5c77af6f/attachment.zip>

From dan at codesourcery.com  Mon Mar 24 14:00:11 2008
From: dan at codesourcery.com (Daniel Jacobowitz)
Date: Mon, 24 Mar 2008 10:00:11 -0400
Subject: [coldfire-gnu-discuss] The different size is gotten from
	m68k-elf-nm and m68k-elf-readelf
In-Reply-To: <61577c8f0803240032h1ee4a8a6o2cde5d9c568c1b0f@mail.gmail.com>
References: <61577c8f0803240032h1ee4a8a6o2cde5d9c568c1b0f@mail.gmail.com>
Message-ID: <20080324140008.GQ13245@caradoc.them.org>

On Mon, Mar 24, 2008 at 02:32:23PM +0700, Ha Luong wrote:
> Dear Sir,
> 
> I used m68k-elf-readelf and m68k-elf-nm for calculating the elf size. But
> the total size from m68k-elf-readelf is different to its from m68k-elf-nm.
> I used
> The total size from m68k-elf-readelf = the total of size column.
> with the command line option "m68k-elf-readelf -S xxx.elf"
> 
> And The total size from m68k-elf-nm = the total of column 2
> with the command line option "m68k-elf-nm.exe -S --print-size --print-armap
> --size-sort xxx.elf"

The total size of all sections may be larger than the total size of
all symbols.  There are structures in your application, like the
symbol table, which do not have an associated symbol.

If you just want the memory requirements of your application, I
recommend using m68k-elf-size or m68k-elf-readelf -l (lowercase
L).

-- 
Daniel Jacobowitz
CodeSourcery


From dan at codesourcery.com  Tue Mar 25 11:31:07 2008
From: dan at codesourcery.com (Daniel Jacobowitz)
Date: Tue, 25 Mar 2008 07:31:07 -0400
Subject: [coldfire-gnu-discuss] The different size is gotten from
	m68k-elf-nm and m68k-elf-readelf
In-Reply-To: <61577c8f0803250316p522b0f11q28aed6da2a881460@mail.gmail.com>
References: <61577c8f0803240032h1ee4a8a6o2cde5d9c568c1b0f@mail.gmail.com> <20080324140008.GQ13245@caradoc.them.org> <61577c8f0803250316p522b0f11q28aed6da2a881460@mail.gmail.com>
Message-ID: <20080325113102.GN13245@caradoc.them.org>

On Tue, Mar 25, 2008 at 05:16:10PM +0700, Ha Luong wrote:
> Dear Daniel Jacobowitz,
> 
> Could you please told me the file size and mem size mean?

Every allocatable ("A" flag) section, and every LOAD segment, takes a
certain amount of space in the file and a certain amount in memory.
Sometimes the memory size is bigger; the difference is filled with
zeroes for uninitialized variables.

-- 
Daniel Jacobowitz
CodeSourcery


From haluongvn at gmail.com  Tue Mar 25 10:16:10 2008
From: haluongvn at gmail.com (Ha Luong)
Date: Tue, 25 Mar 2008 17:16:10 +0700
Subject: [coldfire-gnu-discuss] The different size is gotten from m68k-elf-nm and m68k-elf-readelf
In-Reply-To: <20080324140008.GQ13245@caradoc.them.org>
References: <61577c8f0803240032h1ee4a8a6o2cde5d9c568c1b0f@mail.gmail.com>
	 <20080324140008.GQ13245@caradoc.them.org>
Message-ID: <61577c8f0803250316p522b0f11q28aed6da2a881460@mail.gmail.com>

Dear Daniel Jacobowitz,

Could you please told me the file size and mem size mean?

Thanks,
Ha Luong


On 3/24/08, Daniel Jacobowitz <dan at codesourcery.com> wrote:
>
> On Mon, Mar 24, 2008 at 02:32:23PM +0700, Ha Luong wrote:
> > Dear Sir,
> >
> > I used m68k-elf-readelf and m68k-elf-nm for calculating the elf size.
> But
> > the total size from m68k-elf-readelf is different to its from
> m68k-elf-nm.
> > I used
> > The total size from m68k-elf-readelf = the total of size column.
> > with the command line option "m68k-elf-readelf -S xxx.elf"
> >
> > And The total size from m68k-elf-nm = the total of column 2
> > with the command line option "m68k-elf-nm.exe -S --print-size
> --print-armap
> > --size-sort xxx.elf"
>
> The total size of all sections may be larger than the total size of
> all symbols.  There are structures in your application, like the
> symbol table, which do not have an associated symbol.
>
> If you just want the memory requirements of your application, I
> recommend using m68k-elf-size or m68k-elf-readelf -l (lowercase
> L).
>
> --
> Daniel Jacobowitz
> CodeSourcery
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sourcerytools.com/pipermail/coldfire-gnu-discuss/attachments/20080325/00a4ea9e/attachment.html>