[2.6 patch] x8664_ksyms.c: unexport register_die_notifier

[2.6 patch] x8664_ksyms.c: unexport register_die_notifier

[Adrian Bunk]

The only user of register_die_notifier (kernel/kprobes.c) can't be 
built modular. Therefore, it's the EXPORT_SYMBOL is superfluous.


Signed-off-by: Adrian Bunk <bunk@stusta.de>

--- linux-2.6.11-rc1-mm1-full/arch/x86_64/kernel/x8664_ksyms.c.old	2005-01-16 05:38:17.000000000 +0100
+++ linux-2.6.11-rc1-mm1-full/arch/x86_64/kernel/x8664_ksyms.c	2005-01-16 08:45:54.000000000 +0100
@@ -198,7 +198,6 @@
 #endif
 
 EXPORT_SYMBOL(die_chain);
-EXPORT_SYMBOL(register_die_notifier);
 
 #ifdef CONFIG_SMP
 EXPORT_SYMBOL(cpu_sibling_map);

Re: [2.6 patch] x8664_ksyms.c: unexport register_die_notifier

[Andi Kleen]

On Sun, Jan 16, 2005 at 08:46:49AM +0100, Adrian Bunk wrote:
> The only user of register_die_notifier (kernel/kprobes.c) can't be 
> built modular. Therefore, it's the EXPORT_SYMBOL is superfluous.

Please don't apply this, it was especially intended for modular debuggers.
There is already a hacvked kdb around that uses it.

-Andi

Re: [2.6 patch] x8664_ksyms.c: unexport register_die_notifier

[Arjan van de Ven]

On Mon, 2005-01-17 at 10:26 +0100, Andi Kleen wrote:
> On Sun, Jan 16, 2005 at 08:46:49AM +0100, Adrian Bunk wrote:
> > The only user of register_die_notifier (kernel/kprobes.c) can't be 
> > built modular. Therefore, it's the EXPORT_SYMBOL is superfluous.
> 
> Please don't apply this, it was especially intended for modular debuggers.
> There is already a hacvked kdb around that uses it.

eh didn't Tigran just mail lkml claiming that kdb and x86-64 really
don't mix ??

Re: [discuss] Re: [2.6 patch] x8664_ksyms.c: unexport register_die_notifier

[Tigran Aivazian]

On Mon, 17 Jan 2005, Arjan van de Ven wrote:

> On Mon, 2005-01-17 at 10:26 +0100, Andi Kleen wrote:
>> On Sun, Jan 16, 2005 at 08:46:49AM +0100, Adrian Bunk wrote:
>>> The only user of register_die_notifier (kernel/kprobes.c) can't be
>>> built modular. Therefore, it's the EXPORT_SYMBOL is superfluous.
>>
>> Please don't apply this, it was especially intended for modular debuggers.
>> There is already a hacvked kdb around that uses it.
>
> eh didn't Tigran just mail lkml claiming that kdb and x86-64 really
> don't mix ??

No, I emailed saying that kdb on x86_64 does NOT show the values of 
parameters passed to functions in the backtrace. And, ok, it has some 
other bugs (e.g. reboot doesn't work on SMP) too. But generally speaking 
kdb does work on x86_64. It just still has bugs, but so does every piece 
of software. We should just fix them all, that's all :)

Kind regards
Tigran

Re: [2.6 patch] x8664_ksyms.c: unexport register_die_notifier

[Andi Kleen]

On Mon, Jan 17, 2005 at 10:54:18AM +0100, Arjan van de Ven wrote:
> On Mon, 2005-01-17 at 10:26 +0100, Andi Kleen wrote:
> > On Sun, Jan 16, 2005 at 08:46:49AM +0100, Adrian Bunk wrote:
> > > The only user of register_die_notifier (kernel/kprobes.c) can't be 
> > > built modular. Therefore, it's the EXPORT_SYMBOL is superfluous.
> > 
> > Please don't apply this, it was especially intended for modular debuggers.
> > There is already a hacvked kdb around that uses it.
> 
> eh didn't Tigran just mail lkml claiming that kdb and x86-64 really
> don't mix ??

It cannot display function arguments and uses imprecise backtrace right now
(like normal oopses), other than that it works just fine. 

-Andi

Re: [2.6 patch] x8664_ksyms.c: unexport register_die_notifier

[Arjan van de Ven]

On Mon, 2005-01-17 at 11:13 +0100, Andi Kleen wrote:
> On Mon, Jan 17, 2005 at 10:54:18AM +0100, Arjan van de Ven wrote:
> > On Mon, 2005-01-17 at 10:26 +0100, Andi Kleen wrote:
> > > On Sun, Jan 16, 2005 at 08:46:49AM +0100, Adrian Bunk wrote:
> > > > The only user of register_die_notifier (kernel/kprobes.c) can't be 
> > > > built modular. Therefore, it's the EXPORT_SYMBOL is superfluous.
> > > 
> > > Please don't apply this, it was especially intended for modular debuggers.
> > > There is already a hacvked kdb around that uses it.
> > 
> > eh didn't Tigran just mail lkml claiming that kdb and x86-64 really
> > don't mix ??
> 
> It cannot display function arguments and uses imprecise backtrace right now
> (like normal oopses), other than that it works just fine. 

does it need any other kernel patches? Because if it does it might as
well patch this export in for it's specialist use

Re: [2.6 patch] x8664_ksyms.c: unexport register_die_notifier

[Andi Kleen]

On Mon, Jan 17, 2005 at 12:19:28PM +0100, Arjan van de Ven wrote:
> On Mon, 2005-01-17 at 11:13 +0100, Andi Kleen wrote:
> > On Mon, Jan 17, 2005 at 10:54:18AM +0100, Arjan van de Ven wrote:
> > > On Mon, 2005-01-17 at 10:26 +0100, Andi Kleen wrote:
> > > > On Sun, Jan 16, 2005 at 08:46:49AM +0100, Adrian Bunk wrote:
> > > > > The only user of register_die_notifier (kernel/kprobes.c) can't be 
> > > > > built modular. Therefore, it's the EXPORT_SYMBOL is superfluous.
> > > > 
> > > > Please don't apply this, it was especially intended for modular debuggers.
> > > > There is already a hacvked kdb around that uses it.
> > > 
> > > eh didn't Tigran just mail lkml claiming that kdb and x86-64 really
> > > don't mix ??
> > 
> > It cannot display function arguments and uses imprecise backtrace right now
> > (like normal oopses), other than that it works just fine. 
> 
> does it need any other kernel patches? Because if it does it might as
> well patch this export in for it's specialist use

The modular kdb needs other patches right now yes, but I would still
like to keep it for other debuggers. Previously we always had tons of 
ugly patchkit comming with every debugging add on and the exception 
notifiers were exactly designed to provide a clean interface for 
everybody for this.  I definitely don't want to go back to the old 
"own patch for everbody" mess again on x86-64.

-Andi
>