Re: [PATCH v2 RESEND 1/2] arm, arm64: change_memory_common with numpages == 0 should be no-op.

[Will Deacon <will.deacon@arm.com>]

Hi Mika,

On Tue, Jan 26, 2016 at 04:59:52PM +0200, mika.penttila@nextfour.com wrote:
> From: Mika Penttilä <mika.penttila@nextfour.com>
> 
> This makes the caller set_memory_xx() consistent with x86.
> 
> arm64 part is rebased on 4.5.0-rc1 with Ard's patch
>  lkml.kernel.org/g/<1453125665-26627-1-git-send-email-ard.biesheuvel@linaro.org>
> applied.
> 
> Signed-off-by: Mika Penttilä mika.penttila@nextfour.com
> Reviewed-by: Laura Abbott <labbott@redhat.com>
> Acked-by: David Rientjes <rientjes@google.com>
> 
> ---
>  arch/arm/mm/pageattr.c   | 3 +++
>  arch/arm64/mm/pageattr.c | 3 +++
>  2 files changed, 6 insertions(+)
> 
> diff --git a/arch/arm/mm/pageattr.c b/arch/arm/mm/pageattr.c
> index cf30daf..d19b1ad 100644
> --- a/arch/arm/mm/pageattr.c
> +++ b/arch/arm/mm/pageattr.c
> @@ -49,6 +49,9 @@ static int change_memory_common(unsigned long addr, int numpages,
>  		WARN_ON_ONCE(1);
>  	}
>  
> +	if (!numpages)
> +		return 0;
> +
>  	if (start < MODULES_VADDR || start >= MODULES_END)
>  		return -EINVAL;
>  
> diff --git a/arch/arm64/mm/pageattr.c b/arch/arm64/mm/pageattr.c
> index 1360a02..b582fc2 100644
> --- a/arch/arm64/mm/pageattr.c
> +++ b/arch/arm64/mm/pageattr.c
> @@ -53,6 +53,9 @@ static int change_memory_common(unsigned long addr, int numpages,
>  		WARN_ON_ONCE(1);
>  	}
>  
> +	if (!numpages)
> +		return 0;
> +

Thanks for this. I can reproduce the failure on my Juno board, so I'd
like to queue this for 4.5 since it fixes a real issue. I've taken the
liberty of rebasing the arm64 part to my fixes branch and writing a
commit message. Does the patch below look ok to you?

Will

--->8

>From 57adec866c0440976c96a4b8f5b59fb411b1cacb Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Mika=20Penttil=C3=A4?= <mika.penttila@nextfour.com>
Date: Tue, 26 Jan 2016 15:47:25 +0000
Subject: [PATCH] arm64: mm: avoid calling apply_to_page_range on empty range
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Calling apply_to_page_range with an empty range results in a BUG_ON
from the core code. This can be triggered by trying to load the st_drv
module with CONFIG_DEBUG_SET_MODULE_RONX enabled:

  kernel BUG at mm/memory.c:1874!
  Internal error: Oops - BUG: 0 [#1] PREEMPT SMP
  Modules linked in:
  CPU: 3 PID: 1764 Comm: insmod Not tainted 4.5.0-rc1+ #2
  Hardware name: ARM Juno development board (r0) (DT)
  task: ffffffc9763b8000 ti: ffffffc975af8000 task.ti: ffffffc975af8000
  PC is at apply_to_page_range+0x2cc/0x2d0
  LR is at change_memory_common+0x80/0x108

This patch fixes the issue by making change_memory_common (called by the
set_memory_* functions) a NOP when numpages == 0, therefore avoiding the
erroneous call to apply_to_page_range and bringing us into line with x86
and s390.

Cc: <stable@vger.kernel.org>
Reviewed-by: Laura Abbott <labbott@redhat.com>
Acked-by: David Rientjes <rientjes@google.com>
Signed-off-by: Mika Penttilä <mika.penttila@nextfour.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
 arch/arm64/mm/pageattr.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/arch/arm64/mm/pageattr.c b/arch/arm64/mm/pageattr.c
index 3571c7309c5e..cf6240741134 100644
--- a/arch/arm64/mm/pageattr.c
+++ b/arch/arm64/mm/pageattr.c
@@ -57,6 +57,9 @@ static int change_memory_common(unsigned long addr, int numpages,
 	if (end < MODULES_VADDR || end >= MODULES_END)
 		return -EINVAL;
 
+	if (!numpages)
+		return 0;
+
 	data.set_mask = set_mask;
 	data.clear_mask = clear_mask;
 
-- 
2.1.4