Mailing-List: contact cygwin-help AT sourceware DOT cygnus DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-subscribe AT sources DOT redhat DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin/>
List-Post: <mailto:cygwin AT sources DOT redhat DOT com>
List-Help: <mailto:cygwin-help AT sources DOT redhat DOT com>, <http://sources.redhat.com/ml/#faqs>
Sender: cygwin-owner AT sources DOT redhat DOT com
Delivered-To: mailing list cygwin AT sources DOT redhat DOT com
Message-ID: <3BF9791F.2070606@enea.se>
Date: Mon, 19 Nov 2001 22:26:55 +0100
From: Mohammad Saleem <mosa AT enea DOT se>
User-Agent: Mozilla/5.0 (Windows; U; WinNT4.0; en-US; rv:0.9.2) Gecko/20010726 Netscape6/6.1
X-Accept-Language: en-us
MIME-Version: 1.0
To: kai DOT ruottu AT luukku DOT com
CC: crossgcc AT sources DOT redhat DOT com, cygwin AT cygwin DOT com
Subject: Re: linking application with a library
References: <21558852 DOT 1006167645775 DOT JavaMail DOT hip-pop AT porkkana> <3BF970D8 DOT B1F67C3B AT luukku DOT com>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit

Hi Kai,

Thanks for answer.
regards

/Mohammad

Kai Ruottu wrote:

>Mohammad Saleem wrote:
>
>>Hi all,
>>
>>I am using h8300-hitachi-hms-gcc cross-complier for building my application.
>>I have a lib mylib.a (20 kb) which I want to link with my application.
>>My application gets 20 kb bigger as soon as I link it with mylib.a
>>without any function being called from mylib.a.
>>
>
> The linker shouldn't take anything from the lib if it doesn't need anything
>from it.
>
>>It seems to me that I have built mylib.a so that the whole lib is linked
>>during linking process even if no one method is being called.
>>
>
> Is the library produced using C++ or ObjC or from where the name 'method'
>comes?
>
>>Now my question is which flag should I use to build mylib.a so that only
>>those functions are linked which are being called.
>>
>
> Compiling each function into its own '.o' module is the normal practice,
>the linker takes only the needed '.o' files from the archive.
> 
>
>>If no function is called then linking mylib.a shouldn't affect the size
>>of my application.
>>
>
> Yes, this should be the case...
>
> BTW, the H8/300 is a architecture which requires tricks like using
>the '__attribute__((eightbit_data))' for the SFRs (I/O-registers) in
>order to produce any effective code...  For instance:
>
>-------------------- clip --------------------------------
>#define SFR       __attribute__((eightbit_data))
>
>extern volatile unsigned char SFR SCI_SSR0;
>extern volatile unsigned char SFR SCI_TDR0;
>
>/*
>#define SCI_TDR0 (*(volatile unsigned char *) (0xffffdb))
>#define SCI_SSR0 (*(volatile unsigned char *) (0xffffdc))
>*/
>
>/* Single char out to the serial port */
>
>void put_ser(unsigned char c)
>{
>        while ((SCI_SSR0 & 0x80) != 0x80)
>                ;
>        SCI_TDR0 = c;
>        SCI_SSR0 &= 0x7F;
>}
>-------------------- clip --------------------------------
>;       GCC For the Hitachi H8/300
>;       By Hitachi America Ltd and Cygnus Support
>;       release F-1
>; -O2
>
>        .h8300s
>        .file   "put_ser.c"
>        .section .text
>        .align 1
>        .global _put_ser
>_put_ser:
>.L5:
>        mov.b   @_SCI_SSR0:8,r2l
>        and     #128,r2l
>        beq     .L5
>        mov.b   r0l,@_SCI_TDR0:8
>        bclr    #7,@_SCI_SSR0:8
>        rts
>        .end
>-------------------- clip --------------------------------
>
> Getting a RISC-like 'load/and/store' operation instead of the
>'bclr #7,@_SCI_SSR0:8' is not rare if not using any of these
>'tricks'... And if the I/O-ports are used much, much 'code
>bloat' will be the result...
>
>Cheers, Kai
>
>
>



--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Bug reporting:         http://cygwin.com/bugs.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/