2.5.53 : modules_install warnings

2.5.53 : modules_install warnings

[Frank Davis]

Hello all,
  I received the following warnings while a 'make modules_install'. It 
looks like there are a few more locking changes that need to be made. :)
Regards,
Frank

make -rR -f scripts/Makefile.modinst obj=arch/i386/lib
if [ -r System.map ]; then /sbin/depmod -ae -F System.map  2.5.53; fi
WARNING: /lib/modules/2.5.53/kernel/net/atm/lec.ko needs unknown symbol cli
WARNING: /lib/modules/2.5.53/kernel/net/atm/lec.ko needs unknown symbol restore_flags
WARNING: /lib/modules/2.5.53/kernel/net/atm/lec.ko needs unknown symbol save_flags
WARNING: /lib/modules/2.5.53/kernel/drivers/net/wan/x25_asy.ko needs unknown symbol cli
WARNING: /lib/modules/2.5.53/kernel/drivers/net/wan/x25_asy.ko needs unknown symbol restore_flags
WARNING: /lib/modules/2.5.53/kernel/drivers/net/wan/x25_asy.ko needs unknown symbol save_flags
WARNING: /lib/modules/2.5.53/kernel/drivers/net/pcmcia/3c574_cs.ko needs unknown symbol cli
WARNING: /lib/modules/2.5.53/kernel/drivers/net/pcmcia/3c574_cs.ko needs unknown symbol restore_flags
WARNING: /lib/modules/2.5.53/kernel/drivers/net/pcmcia/3c574_cs.ko needs unknown symbol save_flags
WARNING: /lib/modules/2.5.53/kernel/drivers/char/ftape/lowlevel/ftape.ko needs unknown symbol cli
WARNING: /lib/modules/2.5.53/kernel/drivers/char/ftape/lowlevel/ftape.ko needs unknown symbol restore_flags
WARNING: /lib/modules/2.5.53/kernel/drivers/char/ftape/lowlevel/ftape.ko needs unknown symbol save_flags
WARNING: /lib/modules/2.5.53/kernel/drivers/char/ftape/lowlevel/ftape.ko needs unknown symbol sti

Re: 2.5.53 : modules_install warnings

[Arnaldo Carvalho de Melo]

Em Sat, Dec 28, 2002 at 06:01:55PM -0500, Frank Davis escreveu:
> Hello all,
>   I received the following warnings while a 'make modules_install'. It 
> looks like there are a few more locking changes that need to be made. :)

A few? do a make allmodconfig and you rethink your statement 8)

- Arnaldo

Re: 2.5.53 : modules_install warnings

[Rusty Russell]

In message <Pine.LNX.4.44.0212281758230.839-100000@linux-dev> you write:
> Hello all,
>   I received the following warnings while a 'make modules_install'. It 
> looks like there are a few more locking changes that need to be made. :)

This is SMP, right?  Those warnings are perfectly correct (yes, those
files need updating).

Cheers,
Rusty.
--
  Anyone who quotes me in their sig is an idiot. -- Rusty Russell.

Re: 2.5.53 : modules_install warnings

[Bill Davidsen]

On Sun, 29 Dec 2002, Rusty Russell wrote:

> In message <Pine.LNX.4.44.0212281758230.839-100000@linux-dev> you write:
> > Hello all,
> >   I received the following warnings while a 'make modules_install'. It 
> > looks like there are a few more locking changes that need to be made. :)
> 
> This is SMP, right?  Those warnings are perfectly correct (yes, those
> files need updating).

Any guess when you'll get them fixed?

-- 
bill davidsen <davidsen@tmr.com>
  CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.

Re: 2.5.53 : modules_install warnings

[Rusty Russell]

In message <Pine.LNX.3.96.1021230212600.8353B-100000@gatekeeper.tmr.com> you wr
ite:
> On Sun, 29 Dec 2002, Rusty Russell wrote:
> 
> > In message <Pine.LNX.4.44.0212281758230.839-100000@linux-dev> you write:
> > > Hello all,
> > >   I received the following warnings while a 'make modules_install'. It 
> > > looks like there are a few more locking changes that need to be made. :)
> > 
> > This is SMP, right?  Those warnings are perfectly correct (yes, those
> > files need updating).
> 
> Any guess when you'll get them fixed?

I've discovered an interesting (but kinda obvious) phenomenon.  If you
destabilize some part of the kernel, it becomes the natural suspect
for problems.

The corollary is, I'm getting more reports on kernel "module" bugs
which are not actually my fault at all (and some, like erroneous
__init sections, which the new module code just shows up).

This is one: these drivers are actually broken.  They give warnings on
compile, they won't link when compiled in, and they won't insert as
modules.

Hope that clarifies,
Rusty.
--
  Anyone who quotes me in their sig is an idiot. -- Rusty Russell.

Re: 2.5.53 : modules_install warnings

[Bill Davidsen]

On Tue, 31 Dec 2002, Rusty Russell wrote:

> In message <Pine.LNX.3.96.1021230212600.8353B-100000@gatekeeper.tmr.com> you wr
> ite:
> > On Sun, 29 Dec 2002, Rusty Russell wrote:
> > 
> > > In message <Pine.LNX.4.44.0212281758230.839-100000@linux-dev> you write:
> > > > Hello all,
> > > >   I received the following warnings while a 'make modules_install'. It 
> > > > looks like there are a few more locking changes that need to be made. :)
> > > 
> > > This is SMP, right?  Those warnings are perfectly correct (yes, those
> > > files need updating).
> > 
> > Any guess when you'll get them fixed?
> 
> I've discovered an interesting (but kinda obvious) phenomenon.  If you
> destabilize some part of the kernel, it becomes the natural suspect
> for problems.
> 
> The corollary is, I'm getting more reports on kernel "module" bugs
> which are not actually my fault at all (and some, like erroneous
> __init sections, which the new module code just shows up).
> 
> This is one: these drivers are actually broken.  They give warnings on
> compile, they won't link when compiled in, and they won't insert as
> modules.
> 
> Hope that clarifies,

If they didn't work in 2.5.47, before the module change, then clearly they
are broken on their own. If they worked until then, and especially if they
work built-in still, I would certainly suspect that the problem is related
to the module change.

-- 
bill davidsen <davidsen@tmr.com>
  CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.

Re: 2.5.53 : modules_install warnings

[Rusty Russell]

In message <Pine.LNX.3.96.1021231091929.10362B-100000@gatekeeper.tmr.com> you write:
> If they didn't work in 2.5.47, before the module change, then clearly they
> are broken on their own. If they worked until then, and especially if they
> work built-in still, I would certainly suspect that the problem is related
> to the module change.

That's the point: they use cli, sti and save_flags.  All three were
eliminated in SMP completely independently of the module changes.

Hope I'm being clearer?
Rusty.
--
  Anyone who quotes me in their sig is an idiot. -- Rusty Russell.

Re: 2.5.53 : modules_install warnings

[Bill Davidsen]

On Wed, 1 Jan 2003, Rusty Russell wrote:

> In message <Pine.LNX.3.96.1021231091929.10362B-100000@gatekeeper.tmr.com> you write:
> > If they didn't work in 2.5.47, before the module change, then clearly they
> > are broken on their own. If they worked until then, and especially if they
> > work built-in still, I would certainly suspect that the problem is related
> > to the module change.
> 
> That's the point: they use cli, sti and save_flags.  All three were
> eliminated in SMP completely independently of the module changes.
> 
> Hope I'm being clearer?

Okay, so there are two issues, the SMP issue noted and and changes which
might be needed to make them work as modules. Gotit, thanks.

-- 
bill davidsen <davidsen@tmr.com>
  CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.

2.5.x Configuration - about ISA

[Maciej Soltysiak]

Hi,

When CONFIG_ISA is disabled and CONFIG_PNP_CARD is enabled, an option
appears:
CONFIG_ISAPNP

If we disabled ISA bus, shold not ISA Plug and Play be disabled too ?

Regards,
Maciej Soltysiak

Re: 2.5.x Configuration - about ISA

[Alan Cox]

On Thu, 2003-01-02 at 13:06, Maciej Soltysiak wrote:
> Hi,
> 
> When CONFIG_ISA is disabled and CONFIG_PNP_CARD is enabled, an option
> appears:
> CONFIG_ISAPNP
> 
> If we disabled ISA bus, shold not ISA Plug and Play be disabled too ?

ISAPnP yes, PNPBIos no

Re: 2.5.53 : modules_install warnings

[Rusty Russell]

In message <Pine.LNX.3.96.1030102075543.18246D-100000@gatekeeper.tmr.com> you w
rite:
> On Wed, 1 Jan 2003, Rusty Russell wrote:
> 
> > That's the point: they use cli, sti and save_flags.  All three were
> > eliminated in SMP completely independently of the module changes.
> > 
> > Hope I'm being clearer?
> 
> Okay, so there are two issues, the SMP issue noted and and changes which
> might be needed to make them work as modules. Gotit, thanks.

Nope, one issue.  The SMP issue.  There is no module issue here.
These are not the droids you are looking for. 8)

A patch to deprecate those functions might be in order.  But so might
a patch to just rip them out.

Hope that buries this,
Rusty.
--
  Anyone who quotes me in their sig is an idiot. -- Rusty Russell.