[FIX] powerpc/prom: Enable Radix GTSE in cpu pa-features

From: Nicholas Piggin <npiggin@gmail.com>

When '029ab30b4c0a ("powerpc/mm: Enable radix GTSE only if supported.")'
made GTSE an MMU feature, it was enabled by default in
powerpc-cpu-features but was missed in pa-features. This causes
random memory corruption during boot of PowerNV kernels where
CONFIG_PPC_DT_CPU_FTRS isn't enabled.

Fixes: 029ab30b4c0a ("powerpc/mm: Enable radix GTSE only if supported.")
Reported-by: Qian Cai <cai@lca.pw>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Bharata B Rao <bharata@linux.ibm.com>
---
 arch/powerpc/kernel/prom.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

Excerpts from Bharata B Rao's message of July 20, 2020 2:42 pm:
> From: Nicholas Piggin <npiggin@gmail.com>
> 
> When '029ab30b4c0a ("powerpc/mm: Enable radix GTSE only if supported.")'
> made GTSE an MMU feature, it was enabled by default in
> powerpc-cpu-features but was missed in pa-features. This causes
> random memory corruption during boot of PowerNV kernels where
> CONFIG_PPC_DT_CPU_FTRS isn't enabled.

Thanks for writing this up, I got a bit bogged down with other things.

> Fixes: 029ab30b4c0a ("powerpc/mm: Enable radix GTSE only if supported.")
> Reported-by: Qian Cai <cai@lca.pw>
> Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
> Signed-off-by: Bharata B Rao <bharata@linux.ibm.com>
> ---
>  arch/powerpc/kernel/prom.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c
> index 9cc49f265c86..a9594bad572a 100644
> --- a/arch/powerpc/kernel/prom.c
> +++ b/arch/powerpc/kernel/prom.c
> @@ -163,7 +163,8 @@ static struct ibm_pa_feature {
>  	{ .pabyte = 0,  .pabit = 6, .cpu_features  = CPU_FTR_NOEXECUTE },
>  	{ .pabyte = 1,  .pabit = 2, .mmu_features  = MMU_FTR_CI_LARGE_PAGE },
>  #ifdef CONFIG_PPC_RADIX_MMU
> -	{ .pabyte = 40, .pabit = 0, .mmu_features  = MMU_FTR_TYPE_RADIX },
> +	{ .pabyte = 40, .pabit = 0,
> +	  .mmu_features  = (MMU_FTR_TYPE_RADIX | MMU_FTR_GTSE) },

It might look better like this:

        { .pabyte = 1,  .pabit = 2, .mmu_features  = MMU_FTR_CI_LARGE_PAGE },
#ifdef CONFIG_PPC_RADIX_MMU
        { .pabyte = 40, .pabit = 0, .mmu_features  = MMU_FTR_TYPE_RADIX },
        { .pabyte = 40, .pabit = 0, .mmu_features  = MMU_FTR_TYPE_RADIX |
                                                     MMU_FTR_GTSE },
#endif
  	{ .pabyte = 1,  .pabit = 1, .invert = 1, .cpu_features = CPU_FTR_NODSISRALIGN },

But that's bikeshedding a bit and the optional bits already put it out 
of alignment.

Thanks,
Nick

On Mon, Jul 20, 2020 at 03:38:29PM +1000, Nicholas Piggin wrote:
> Excerpts from Bharata B Rao's message of July 20, 2020 2:42 pm:
> > diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c
> > index 9cc49f265c86..a9594bad572a 100644
> > --- a/arch/powerpc/kernel/prom.c
> > +++ b/arch/powerpc/kernel/prom.c
> > @@ -163,7 +163,8 @@ static struct ibm_pa_feature {
> >  	{ .pabyte = 0,  .pabit = 6, .cpu_features  = CPU_FTR_NOEXECUTE },
> >  	{ .pabyte = 1,  .pabit = 2, .mmu_features  = MMU_FTR_CI_LARGE_PAGE },
> >  #ifdef CONFIG_PPC_RADIX_MMU
> > -	{ .pabyte = 40, .pabit = 0, .mmu_features  = MMU_FTR_TYPE_RADIX },
> > +	{ .pabyte = 40, .pabit = 0,
> > +	  .mmu_features  = (MMU_FTR_TYPE_RADIX | MMU_FTR_GTSE) },
> 
> It might look better like this:
> 
>         { .pabyte = 1,  .pabit = 2, .mmu_features  = MMU_FTR_CI_LARGE_PAGE },
> #ifdef CONFIG_PPC_RADIX_MMU
>         { .pabyte = 40, .pabit = 0, .mmu_features  = MMU_FTR_TYPE_RADIX },
>         { .pabyte = 40, .pabit = 0, .mmu_features  = MMU_FTR_TYPE_RADIX |
>                                                      MMU_FTR_GTSE },
> #endif
>   	{ .pabyte = 1,  .pabit = 1, .invert = 1, .cpu_features = CPU_FTR_NODSISRALIGN },
> 
> But that's bikeshedding a bit and the optional bits already put it out 
> of alignment.

Here it is...

From 1be7f3f8b43503740431b7bdf585e488ecdeb48f Mon Sep 17 00:00:00 2001
From: Nicholas Piggin <npiggin@gmail.com>
Date: Mon, 20 Jul 2020 09:05:05 +0530
Subject: [FIX PATCH] powerpc/prom: Enable Radix GTSE in cpu pa-features

When '029ab30b4c0a ("powerpc/mm: Enable radix GTSE only if supported.")'
made GTSE an MMU feature, it was enabled by default in
powerpc-cpu-features but was missed in pa-features. This causes
random memory corruption during boot of PowerNV kernels if
CONFIG_PPC_DT_CPU_FTRS isn't enabled.

Fixes: 029ab30b4c0a ("powerpc/mm: Enable radix GTSE only if supported.")
Reported-by: Qian Cai <cai@lca.pw>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Bharata B Rao <bharata@linux.ibm.com>
---
 arch/powerpc/kernel/prom.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c
index 9cc49f265c86..dae30e805e42 100644
--- a/arch/powerpc/kernel/prom.c
+++ b/arch/powerpc/kernel/prom.c
@@ -163,7 +163,8 @@ static struct ibm_pa_feature {
 	{ .pabyte = 0,  .pabit = 6, .cpu_features  = CPU_FTR_NOEXECUTE },
 	{ .pabyte = 1,  .pabit = 2, .mmu_features  = MMU_FTR_CI_LARGE_PAGE },
 #ifdef CONFIG_PPC_RADIX_MMU
-	{ .pabyte = 40, .pabit = 0, .mmu_features  = MMU_FTR_TYPE_RADIX },
+	{ .pabyte = 40, .pabit = 0, .mmu_features  = MMU_FTR_TYPE_RADIX |
+						     MMU_FTR_GTSE },
 #endif
 	{ .pabyte = 1,  .pabit = 1, .invert = 1, .cpu_features = CPU_FTR_NODSISRALIGN },
 	{ .pabyte = 5,  .pabit = 0, .cpu_features  = CPU_FTR_REAL_LE,

On Mon, 20 Jul 2020 10:12:58 +0530, Bharata B Rao wrote:
> When '029ab30b4c0a ("powerpc/mm: Enable radix GTSE only if supported.")'
> made GTSE an MMU feature, it was enabled by default in
> powerpc-cpu-features but was missed in pa-features. This causes
> random memory corruption during boot of PowerNV kernels where
> CONFIG_PPC_DT_CPU_FTRS isn't enabled.

Applied to powerpc/next.

[1/1] powerpc/prom: Enable Radix GTSE in cpu pa-features
      https://git.kernel.org/powerpc/c/9a77c4a0a12597c661be374b8d566516c0341570

cheers