Linux MIPS Architecture development
 help / color / mirror / Atom feed
* insmod problems
@ 2001-05-06 16:13 Shay Deloya
  2001-05-09  2:40 ` Ralf Baechle
  0 siblings, 1 reply; 4+ messages in thread
From: Shay Deloya @ 2001-05-06 16:13 UTC (permalink / raw)
  To: linux-mips

Hi All,

I have an old problem came up again and the old solution aren't helping.
I'm using busybox version 0.50 and with kernel 2.2 , and inserting 
some modules ,especially those with DEBUG macroes e.g:
#define DEBUG_HIGH(args...) {if (debug_level >= HIGH) printk(args);}
causes the message :
Relocation overflow of type 4 for

and insmod fails.

I'm compiling the modules with -mlong-calls and still getting this message.

Is it insmod knowen bugs that the relocation is done in bad way or 
a linker/compiler bug. I'm using compiler: egcs ver 1.0.3a
I'm checking this problem at the moment and looking for insmod bug.

Thanks,

Shay Deloya
______________________________________
Software Developer
Jungo - R&D
email: shayd@jungo.com
web: http://www.jungo.com
Phone: 1-877-514-0537(USA)  +972-9-8859365(Worldwide) ext. 221
Fax:   1-877-514-0538(USA)  +972-9-8859366(Worldwide)

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: insmod problems
  2001-05-06 16:13 insmod problems Shay Deloya
@ 2001-05-09  2:40 ` Ralf Baechle
  2001-05-10  8:53   ` Michael Shmulevich
  0 siblings, 1 reply; 4+ messages in thread
From: Ralf Baechle @ 2001-05-09  2:40 UTC (permalink / raw)
  To: Shay Deloya; +Cc: linux-mips

On Sun, May 06, 2001 at 07:13:43PM +0300, Shay Deloya wrote:

> I have an old problem came up again and the old solution aren't helping.
> I'm using busybox version 0.50 and with kernel 2.2 , and inserting 
> some modules ,especially those with DEBUG macroes e.g:
> #define DEBUG_HIGH(args...) {if (debug_level >= HIGH) printk(args);}
> causes the message :
> Relocation overflow of type 4 for
> 
> and insmod fails.
> 
> I'm compiling the modules with -mlong-calls and still getting this message.
> 
> Is it insmod knowen bugs that the relocation is done in bad way or 
> a linker/compiler bug. I'm using compiler: egcs ver 1.0.3a
> I'm checking this problem at the moment and looking for insmod bug.

You'll have to upgrade to very current binutils which for mips*-linux
targets default to elf32-trad{big,little}mips, not IRIX ELF format.  Also
it seems your modutils are a bit rotten, get the latest from ftp.kernel.org.

  Ralf

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: insmod problems
  2001-05-09  2:40 ` Ralf Baechle
@ 2001-05-10  8:53   ` Michael Shmulevich
  2001-05-10 18:46     ` Ralf Baechle
  0 siblings, 1 reply; 4+ messages in thread
From: Michael Shmulevich @ 2001-05-10  8:53 UTC (permalink / raw)
  To: Ralf Baechle; +Cc: Shay Deloya, linux-mips



Ralf Baechle wrote:

> On Sun, May 06, 2001 at 07:13:43PM +0300, Shay Deloya wrote:
> 
> 
>> I have an old problem came up again and the old solution aren't helping.
>> Relocation overflow of type 4 for
>> Is it insmod knowen bugs that the relocation is done in bad way or 
>> a linker/compiler bug. I'm using compiler: egcs ver 1.0.3a
>> I'm checking this problem at the moment and looking for insmod bug.
> 
> 
> You'll have to upgrade to very current binutils which for mips*-linux
> targets default to elf32-trad{big,little}mips, not IRIX ELF format.  Also
> it seems your modutils are a bit rotten, get the latest from ftp.kernel.org.
> 
>   Ralf

On the same topic: same source (!!) 'insmod' executable compiled with 
new (2.10) binutils doesn't exibit such relocation problem.

What may be causing 'insmod' to be dependant on binutils which compile 
it? The module *.o is exactly same as it was before.


Sincerely yours,
Michael Shmulevich
______________________________________
Software Developer
Jungo - R&D
email: michaels@jungo.com
web: http://www.jungo.com
Phone: 1-877-514-0537(USA)  +972-9-8859365(Worldwide) ext. 233
Fax:   1-877-514-0538(USA)  +972-9-8859366(Worldwide)

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: insmod problems
  2001-05-10  8:53   ` Michael Shmulevich
@ 2001-05-10 18:46     ` Ralf Baechle
  0 siblings, 0 replies; 4+ messages in thread
From: Ralf Baechle @ 2001-05-10 18:46 UTC (permalink / raw)
  To: Michael Shmulevich; +Cc: Shay Deloya, linux-mips

On Thu, May 10, 2001 at 11:53:12AM +0300, Michael Shmulevich wrote:

> On the same topic: same source (!!) 'insmod' executable compiled with 
> new (2.10) binutils doesn't exibit such relocation problem.
> 
> What may be causing 'insmod' to be dependant on binutils which compile 
> it? The module *.o is exactly same as it was before.

Without further explanation this one just sounds bizarre for now.

There always have been differences between the code generated by binutils
in the various versions; the resulting changes may result in different
behaviour of a broken programs.  These are fun to debug to say the least.
The other possibility would be binutils bugs but I'm not aware of one
with the effect you described.

  Ralf

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2001-05-10 22:29 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2001-05-06 16:13 insmod problems Shay Deloya
2001-05-09  2:40 ` Ralf Baechle
2001-05-10  8:53   ` Michael Shmulevich
2001-05-10 18:46     ` Ralf Baechle

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox