[PATCH] [Powerpc] fix switch_slb handling of 1T ESID values

[PATCH] [Powerpc] fix switch_slb handling of 1T ESID values

[Will Schmidt]

[Powerpc] fix switch_slb handling of 1T ESID values

Now that we have 1TB segment size support, we need to be using the
GET_ESID_1T macro when comparing ESID values for pc,stack, and
unmapped_base within switch_slb() when we're on a CPU that supports it.

This also happens to fix a duplicate-slb-entry inspired machine-check
exception I was seeing when trying to run java on a power6 partition.

Tested on power6 and power5.

Signed-Off-By:  Will Schmidt <will_schmidt@vnet.ibm.com>

---

There is a similar bit of code in stab.c switch_stab().  Should this change also be made there?
---

 arch/powerpc/mm/slb.c |   19 ++++++++++++++-----
 1 files changed, 14 insertions(+), 5 deletions(-)


diff --git a/arch/powerpc/mm/slb.c b/arch/powerpc/mm/slb.c
index bbd2c51..0c527d7 100644
--- a/arch/powerpc/mm/slb.c
+++ b/arch/powerpc/mm/slb.c
@@ -193,16 +193,25 @@ void switch_slb(struct task_struct *tsk, struct mm_struct *mm)
 		return;
 	slb_allocate(pc);
 
-	if (GET_ESID(pc) == GET_ESID(stack))
-		return;
+	if (cpu_has_feature(CPU_FTR_1T_SEGMENT)) {
+		if (GET_ESID_1T(pc) == GET_ESID_1T(stack))
+			return;
+	} else
+		if (GET_ESID(pc) == GET_ESID(stack))
+			return;
 
 	if (is_kernel_addr(stack))
 		return;
 	slb_allocate(stack);
 
-	if ((GET_ESID(pc) == GET_ESID(unmapped_base))
-	    || (GET_ESID(stack) == GET_ESID(unmapped_base)))
-		return;
+	if (cpu_has_feature(CPU_FTR_1T_SEGMENT)) {
+		if ((GET_ESID_1T(pc) == GET_ESID_1T(unmapped_base))
+		    || (GET_ESID_1T(stack) == GET_ESID_1T(unmapped_base)))
+			return;
+	} else
+		if ((GET_ESID(pc) == GET_ESID(unmapped_base))
+		    || (GET_ESID(stack) == GET_ESID(unmapped_base)))
+			return;
 
 	if (is_kernel_addr(unmapped_base))
 		return;

Re: [PATCH] [Powerpc] fix switch_slb handling of 1T ESID values

[Benjamin Herrenschmidt]

On Fri, 2007-10-26 at 15:46 -0500, Will Schmidt wrote:
> [Powerpc] fix switch_slb handling of 1T ESID values
> 
> Now that we have 1TB segment size support, we need to be using the
> GET_ESID_1T macro when comparing ESID values for pc,stack, and
> unmapped_base within switch_slb() when we're on a CPU that supports it.
> 
> This also happens to fix a duplicate-slb-entry inspired machine-check
> exception I was seeing when trying to run java on a power6 partition.
> 
> Tested on power6 and power5.
> 
> Signed-Off-By:  Will Schmidt <will_schmidt@vnet.ibm.com>

Good catch !

A minor comment is maybe you could factor out the code better doing
something like a ESID_COMPARE() macro ?

> ---
> 
> There is a similar bit of code in stab.c switch_stab().  Should this change also be made there?
> ---

There is no machine that does stab and 1T segments.

Ben.

> 
>  arch/powerpc/mm/slb.c |   19 ++++++++++++++-----
>  1 files changed, 14 insertions(+), 5 deletions(-)
> 
> 
> diff --git a/arch/powerpc/mm/slb.c b/arch/powerpc/mm/slb.c
> index bbd2c51..0c527d7 100644
> --- a/arch/powerpc/mm/slb.c
> +++ b/arch/powerpc/mm/slb.c
> @@ -193,16 +193,25 @@ void switch_slb(struct task_struct *tsk, struct mm_struct *mm)
>  		return;
>  	slb_allocate(pc);
>  
> -	if (GET_ESID(pc) == GET_ESID(stack))
> -		return;
> +	if (cpu_has_feature(CPU_FTR_1T_SEGMENT)) {
> +		if (GET_ESID_1T(pc) == GET_ESID_1T(stack))
> +			return;
> +	} else
> +		if (GET_ESID(pc) == GET_ESID(stack))
> +			return;
>  
>  	if (is_kernel_addr(stack))
>  		return;
>  	slb_allocate(stack);
>  
> -	if ((GET_ESID(pc) == GET_ESID(unmapped_base))
> -	    || (GET_ESID(stack) == GET_ESID(unmapped_base)))
> -		return;
> +	if (cpu_has_feature(CPU_FTR_1T_SEGMENT)) {
> +		if ((GET_ESID_1T(pc) == GET_ESID_1T(unmapped_base))
> +		    || (GET_ESID_1T(stack) == GET_ESID_1T(unmapped_base)))
> +			return;
> +	} else
> +		if ((GET_ESID(pc) == GET_ESID(unmapped_base))
> +		    || (GET_ESID(stack) == GET_ESID(unmapped_base)))
> +			return;
>  
>  	if (is_kernel_addr(unmapped_base))
>  		return;
> 
> 
> _______________________________________________
> Linuxppc-dev mailing list
> Linuxppc-dev@ozlabs.org
> https://ozlabs.org/mailman/listinfo/linuxppc-dev

Re: [PATCH] [Powerpc] fix switch_slb handling of 1T ESID values

[Will Schmidt]

On Sat, 2007-10-27 at 14:19 +1000, Benjamin Herrenschmidt wrote:
> On Fri, 2007-10-26 at 15:46 -0500, Will Schmidt wrote:
> > [Powerpc] fix switch_slb handling of 1T ESID values
> > 
> > Now that we have 1TB segment size support, we need to be using the
> > GET_ESID_1T macro when comparing ESID values for pc,stack, and
> > unmapped_base within switch_slb() when we're on a CPU that supports it.
> > 
> > This also happens to fix a duplicate-slb-entry inspired machine-check
> > exception I was seeing when trying to run java on a power6 partition.
> > 
> > Tested on power6 and power5.
> > 
> > Signed-Off-By:  Will Schmidt <will_schmidt@vnet.ibm.com>
> 
> Good catch !
> 
> A minor comment is maybe you could factor out the code better doing
> something like a ESID_COMPARE() macro ?

Yeah, thats a good idea.   I'll spin up a new patch in the next day or
so.   
It occurred to me that I should continue to use GET_ESID when the user
address is < 1T too. 

> 
> > ---
> > 
> > There is a similar bit of code in stab.c switch_stab().  Should this change also be made there?
> > ---
> 
> There is no machine that does stab and 1T segments.

Ok, thanks for the clarification. 

-Will

> 
> Ben.
> 
> > 
> >  arch/powerpc/mm/slb.c |   19 ++++++++++++++-----
> >  1 files changed, 14 insertions(+), 5 deletions(-)
> > 
> > 
> > diff --git a/arch/powerpc/mm/slb.c b/arch/powerpc/mm/slb.c
> > index bbd2c51..0c527d7 100644
> > --- a/arch/powerpc/mm/slb.c
> > +++ b/arch/powerpc/mm/slb.c
> > @@ -193,16 +193,25 @@ void switch_slb(struct task_struct *tsk, struct mm_struct *mm)
> >  		return;
> >  	slb_allocate(pc);
> >  
> > -	if (GET_ESID(pc) == GET_ESID(stack))
> > -		return;
> > +	if (cpu_has_feature(CPU_FTR_1T_SEGMENT)) {
> > +		if (GET_ESID_1T(pc) == GET_ESID_1T(stack))
> > +			return;
> > +	} else
> > +		if (GET_ESID(pc) == GET_ESID(stack))
> > +			return;
> >  
> >  	if (is_kernel_addr(stack))
> >  		return;
> >  	slb_allocate(stack);
> >  
> > -	if ((GET_ESID(pc) == GET_ESID(unmapped_base))
> > -	    || (GET_ESID(stack) == GET_ESID(unmapped_base)))
> > -		return;
> > +	if (cpu_has_feature(CPU_FTR_1T_SEGMENT)) {
> > +		if ((GET_ESID_1T(pc) == GET_ESID_1T(unmapped_base))
> > +		    || (GET_ESID_1T(stack) == GET_ESID_1T(unmapped_base)))
> > +			return;
> > +	} else
> > +		if ((GET_ESID(pc) == GET_ESID(unmapped_base))
> > +		    || (GET_ESID(stack) == GET_ESID(unmapped_base)))
> > +			return;
> >  
> >  	if (is_kernel_addr(unmapped_base))
> >  		return;
> > 
> > 
> > _______________________________________________
> > Linuxppc-dev mailing list
> > Linuxppc-dev@ozlabs.org
> > https://ozlabs.org/mailman/listinfo/linuxppc-dev
>