[PATCH] zsmalloc: zspage sanity check

[PATCH] zsmalloc: zspage sanity check

[Minchan Kim]

On Thu, Jun 02, 2016 at 09:25:19AM +0900, Minchan Kim wrote:
> On Wed, Jun 01, 2016 at 04:09:26PM +0200, Vlastimil Babka wrote:
> > On 06/01/2016 01:21 AM, Minchan Kim wrote:
> > 
> > [...]
> > 
> > > 
> > > Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
> > > Signed-off-by: Minchan Kim <minchan@kernel.org>
> > 
> > I'm not that familiar with zsmalloc, so this is not a full review. I was
> > just curious how it's handling the movable migration API, and stumbled
> > upon some things pointed out below.
> > 
> > > @@ -252,16 +276,23 @@ struct zs_pool {
> > >   */
> > >  #define FULLNESS_BITS	2
> > >  #define CLASS_BITS	8
> > > +#define ISOLATED_BITS	3
> > > +#define MAGIC_VAL_BITS	8
> > >  
> > >  struct zspage {
> > >  	struct {
> > >  		unsigned int fullness:FULLNESS_BITS;
> > >  		unsigned int class:CLASS_BITS;
> > > +		unsigned int isolated:ISOLATED_BITS;
> > > +		unsigned int magic:MAGIC_VAL_BITS;
> > 
> > This magic seems to be only tested via VM_BUG_ON, so it's presence
> > should be also guarded by #ifdef DEBUG_VM, no?
> 
> Thanks for the point.
> 
> Then, I want to change it to BUG_ON because struct zspage corruption
> is really risky to work rightly and want to catch on it in real product
> which disable CONFIG_DEBUG_VM for a while until make the feature stable.

Andrew,

Please fold this patch into zsmalloc: page migration support.
Thanks!

[PATCH] zsmalloc: zspage sanity check

[Minchan Kim]

On Thu, Jun 02, 2016 at 09:25:19AM +0900, Minchan Kim wrote:
> On Wed, Jun 01, 2016 at 04:09:26PM +0200, Vlastimil Babka wrote:
> > On 06/01/2016 01:21 AM, Minchan Kim wrote:
> > 
> > [...]
> > 
> > > 
> > > Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
> > > Signed-off-by: Minchan Kim <minchan@kernel.org>
> > 
> > I'm not that familiar with zsmalloc, so this is not a full review. I was
> > just curious how it's handling the movable migration API, and stumbled
> > upon some things pointed out below.
> > 
> > > @@ -252,16 +276,23 @@ struct zs_pool {
> > >   */
> > >  #define FULLNESS_BITS	2
> > >  #define CLASS_BITS	8
> > > +#define ISOLATED_BITS	3
> > > +#define MAGIC_VAL_BITS	8
> > >  
> > >  struct zspage {
> > >  	struct {
> > >  		unsigned int fullness:FULLNESS_BITS;
> > >  		unsigned int class:CLASS_BITS;
> > > +		unsigned int isolated:ISOLATED_BITS;
> > > +		unsigned int magic:MAGIC_VAL_BITS;
> > 
> > This magic seems to be only tested via VM_BUG_ON, so it's presence
> > should be also guarded by #ifdef DEBUG_VM, no?
> 
> Thanks for the point.
> 
> Then, I want to change it to BUG_ON because struct zspage corruption
> is really risky to work rightly and want to catch on it in real product
> which disable CONFIG_DEBUG_VM for a while until make the feature stable.

Andrew,

Please fold this patch into zsmalloc: page migration support.
Thanks!

>From 6b8fa9889b7892bdf5b2f01698310cfc766c4d17 Mon Sep 17 00:00:00 2001
From: Minchan Kim <minchan@kernel.org>
Date: Fri, 3 Jun 2016 09:54:03 +0900
Subject: [PATCH] zsmalloc: zspage sanity check

The zsmalloc page migration is new feature so I want to catch any
corruption of zspage struct which is very critical to work
regardless of CONFIG_DEBUG_VM.

Signed-off-by: Minchan Kim <minchan@kernel.org>
---
 mm/zsmalloc.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/mm/zsmalloc.c b/mm/zsmalloc.c
index a80100db16d6..f75d2d134741 100644
--- a/mm/zsmalloc.c
+++ b/mm/zsmalloc.c
@@ -523,7 +523,7 @@ static void get_zspage_mapping(struct zspage *zspage,
 				unsigned int *class_idx,
 				enum fullness_group *fullness)
 {
-	VM_BUG_ON(zspage->magic != ZSPAGE_MAGIC);
+	BUG_ON(zspage->magic != ZSPAGE_MAGIC);
 
 	*fullness = zspage->fullness;
 	*class_idx = zspage->class;
@@ -857,7 +857,7 @@ static struct zspage *get_zspage(struct page *page)
 {
 	struct zspage *zspage = (struct zspage *)page->private;
 
-	VM_BUG_ON(zspage->magic != ZSPAGE_MAGIC);
+	BUG_ON(zspage->magic != ZSPAGE_MAGIC);
 	return zspage;
 }
 
-- 
1.9.1