linux-next: manual merge of the tip tree with the slab tree

linux-next: manual merge of the tip tree with the slab tree

[Stephen Rothwell]

Hi all,

Today's linux-next merge of the tip tree got a conflict in mm/slub.c
between commit 964cf35c88f93b4927dbc4e950dfa4d880c7f9d1 ("SLUB: Fix early
boot GFP_DMA allocations") from the slab tree and commit
18fd427debcf37c06917b55295df682fd05fee76 ("slub: add hooks for
kmemcheck") from the tip tree.

I fixed it up (see below) and can carry the fix as necessary.
-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

diff --cc mm/slub.c
index 3976098,73fb7d1..0000000
--- a/mm/slub.c
+++ b/mm/slub.c
@@@ -2700,18 -2657,10 +2726,18 @@@ static noinline struct kmem_cache *dma_
  			 (unsigned int)realsize);
  	s = kmalloc(kmem_size, flags & ~SLUB_DMA);
  
 +	/*
 +	 * Must defer sysfs creation to a workqueue because we don't know
 +	 * what context we are called from. Before sysfs comes up, we don't
 +	 * need to do anything because our sysfs initcall will start by
 +	 * adding all existing slabs to sysfs.
 +	 */
- 	slabflags = SLAB_CACHE_DMA;
++	slabflags = SLAB_CACHE_DMA | SLAB_NOTRACK;
 +	if (slab_state >= SYSFS)
 +		slabflags |= __SYSFS_ADD_DEFERRED;
 +
  	if (!s || !text || !kmem_cache_open(s, flags, text,
 -			realsize, ARCH_KMALLOC_MINALIGN,
 -			SLAB_CACHE_DMA|SLAB_NOTRACK|__SYSFS_ADD_DEFERRED,
 -			NULL)) {
 +			realsize, ARCH_KMALLOC_MINALIGN, slabflags, NULL)) {
  		kfree(s);
  		kfree(text);
  		goto unlock_out;

Re: linux-next: manual merge of the tip tree with the slab tree

[Pekka Enberg]

On Tue, 2009-06-16 at 15:46 +1000, Stephen Rothwell wrote:
> Hi all,
> 
> Today's linux-next merge of the tip tree got a conflict in mm/slub.c
> between commit 964cf35c88f93b4927dbc4e950dfa4d880c7f9d1 ("SLUB: Fix early
> boot GFP_DMA allocations") from the slab tree and commit
> 18fd427debcf37c06917b55295df682fd05fee76 ("slub: add hooks for
> kmemcheck") from the tip tree.
> 
> I fixed it up (see below) and can carry the fix as necessary.

Looks good! Thanks.

Re: linux-next: manual merge of the tip tree with the slab tree

[Ingo Molnar]

* Pekka Enberg <penberg@cs.helsinki.fi> wrote:

> On Tue, 2009-06-16 at 15:46 +1000, Stephen Rothwell wrote:
> > Hi all,
> > 
> > Today's linux-next merge of the tip tree got a conflict in mm/slub.c
> > between commit 964cf35c88f93b4927dbc4e950dfa4d880c7f9d1 ("SLUB: Fix early
> > boot GFP_DMA allocations") from the slab tree and commit
> > 18fd427debcf37c06917b55295df682fd05fee76 ("slub: add hooks for
> > kmemcheck") from the tip tree.
> > 
> > I fixed it up (see below) and can carry the fix as necessary.
> 
> Looks good! Thanks.

Btw., this SLAB commit is for this merge window, right?

	Ingo

Re: linux-next: manual merge of the tip tree with the slab tree

[Pekka Enberg]

On Tue, 2009-06-16 at 09:44 +0200, Ingo Molnar wrote:
> * Pekka Enberg <penberg@cs.helsinki.fi> wrote:
> 
> > On Tue, 2009-06-16 at 15:46 +1000, Stephen Rothwell wrote:
> > > Hi all,
> > > 
> > > Today's linux-next merge of the tip tree got a conflict in mm/slub.c
> > > between commit 964cf35c88f93b4927dbc4e950dfa4d880c7f9d1 ("SLUB: Fix early
> > > boot GFP_DMA allocations") from the slab tree and commit
> > > 18fd427debcf37c06917b55295df682fd05fee76 ("slub: add hooks for
> > > kmemcheck") from the tip tree.
> > > 
> > > I fixed it up (see below) and can carry the fix as necessary.
> > 
> > Looks good! Thanks.
> 
> Btw., this SLAB commit is for this merge window, right?

Yes, it's a fix for SLUB that's going to be part of the next pull
request (probably out today or tomorrow).

			Pekka

Re: linux-next: manual merge of the tip tree with the slab tree

[Pekka Enberg]

On Mon, 2009-07-06 at 14:40 +1000, Stephen Rothwell wrote:
> Today's linux-next merge of the tip tree got a conflict in
> include/linux/rcupdate.h between commit
> 249acdd3fd13d1ec579ff9b1c5273cf5515cd352 ("SLQB slab allocator (try 2)")
> from the slab tree and commit 03b042bf1dc14a268a3d65d38b4ec2a4261e8477
> ("rcu: Add synchronize_sched_expedited() primitive") from the tip tree.
> 
> Just context changes. I fixed it up (see below) and can carry the fix as
> necessary.

Yup, looks good to me. Thanks, Stephen!

Re: linux-next: manual merge of the tip tree with the slab tree

[Pekka Enberg]

On Mon, 2009-09-14 at 15:52 +1000, Stephen Rothwell wrote:
> Hi Pekka,
> 
> On Mon, 6 Jul 2009 14:40:51 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> >
> > Today's linux-next merge of the tip tree got a conflict in
> > include/linux/rcupdate.h between commit
> > 249acdd3fd13d1ec579ff9b1c5273cf5515cd352 ("SLQB slab allocator (try 2)")
> > from the slab tree and commit 03b042bf1dc14a268a3d65d38b4ec2a4261e8477
> > ("rcu: Add synchronize_sched_expedited() primitive") from the tip tree.
> > 
> > Just context changes. I fixed it up (see below) and can carry the fix as
> > necessary.
> 
> This conflict is now between the slab tree and Linus' tree.

OK, thanks Stephen, I'll fix it up in my tree.

linux-next: manual merge of the tip tree with the slab tree

[Stephen Rothwell]

Hi all,

Today's linux-next merge of the tip tree got a conflict in
include/linux/percpu.h between commit
6fc80ef491b981f59233beaf6aeaccc0c947031d ("percpu: use percpu allocator
on UP too") from the slab tree and commit
8b8e2ec1eeca7f6941bc81cefc9663018d6ceb57 ("percpu: Add {get,put}
_cpu_ptr") from the tip tree.

Just context changes.  I fixed it up (see below) and can carry the fix as
necessary.
-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

diff --cc include/linux/percpu.h
index aeeeef1,0eb5083..0000000
--- a/include/linux/percpu.h
+++ b/include/linux/percpu.h
@@@ -39,8 -39,19 +39,17 @@@
  	preempt_enable();				\
  } while (0)
  
+ #define get_cpu_ptr(var) ({				\
+ 	preempt_disable();				\
+ 	this_cpu_ptr(var); })
+ 
+ #define put_cpu_ptr(var) do {				\
+ 	(void)(var);					\
+ 	preempt_enable();				\
+ } while (0)
+ 
 -#ifdef CONFIG_SMP
 -
  /* minimum unit size, also is the maximum supported allocation size */
 -#define PCPU_MIN_UNIT_SIZE		PFN_ALIGN(64 << 10)
 +#define PCPU_MIN_UNIT_SIZE		PFN_ALIGN(32 << 10)
  
  /*
   * Percpu allocator can serve percpu allocations before slab is

Re: linux-next: manual merge of the tip tree with the slab tree

[Pekka Enberg]

On Thu, Sep 23, 2010 at 6:44 AM, Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> Today's linux-next merge of the tip tree got a conflict in
> include/linux/percpu.h between commit
> 6fc80ef491b981f59233beaf6aeaccc0c947031d ("percpu: use percpu allocator
> on UP too") from the slab tree and commit
> 8b8e2ec1eeca7f6941bc81cefc9663018d6ceb57 ("percpu: Add {get,put}
> _cpu_ptr") from the tip tree.
>
> Just context changes.  I fixed it up (see below) and can carry the fix as
> necessary.

Why are we seeing a merge conflict here? I cherry-picked patches from
Tejun's for-next branch but didn't modify them.

> diff --cc include/linux/percpu.h
> index aeeeef1,0eb5083..0000000
> --- a/include/linux/percpu.h
> +++ b/include/linux/percpu.h
> @@@ -39,8 -39,19 +39,17 @@@
>        preempt_enable();                               \
>  } while (0)
>
> + #define get_cpu_ptr(var) ({                           \
> +       preempt_disable();                              \
> +       this_cpu_ptr(var); })
> +
> + #define put_cpu_ptr(var) do {                         \
> +       (void)(var);                                    \
> +       preempt_enable();                               \
> + } while (0)
> +
>  -#ifdef CONFIG_SMP
>  -
>  /* minimum unit size, also is the maximum supported allocation size */
>  -#define PCPU_MIN_UNIT_SIZE            PFN_ALIGN(64 << 10)
>  +#define PCPU_MIN_UNIT_SIZE            PFN_ALIGN(32 << 10)
>
>  /*
>   * Percpu allocator can serve percpu allocations before slab is
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>

Re: linux-next: manual merge of the tip tree with the slab tree

[Tejun Heo]

On 09/23/2010 11:33 AM, Pekka Enberg wrote:
> On Thu, Sep 23, 2010 at 6:44 AM, Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>> Today's linux-next merge of the tip tree got a conflict in
>> include/linux/percpu.h between commit
>> 6fc80ef491b981f59233beaf6aeaccc0c947031d ("percpu: use percpu allocator
>> on UP too") from the slab tree and commit
>> 8b8e2ec1eeca7f6941bc81cefc9663018d6ceb57 ("percpu: Add {get,put}
>> _cpu_ptr") from the tip tree.
>>
>> Just context changes.  I fixed it up (see below) and can carry the fix as
>> necessary.
> 
> Why are we seeing a merge conflict here? I cherry-picked patches from
> Tejun's for-next branch but didn't modify them.

That's a different one coming from the perf tree.  get_cpu_ptr() is
only used by perf at this point so it got routed through there, so the
conflict.  Nothing to worry about.

Thanks.

-- 
tejun

linux-next: manual merge of the tip tree with the slab tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1452 bytes --]

Hi all,

Today's linux-next merge of the tip tree got conflicts in mm/slab.c and
mm/slub.c between commit e7b691b085fd ("slab/mempolicy: always use local
policy from interrupt context") from the slab tree and commit
2ab41dd59922 ("mm: Optimize put_mems_allowed() usage") from the tip tree.

Just context changes.  I fixed it up (see below) and can carry the fix as
necessary.
-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

diff --cc mm/slab.c
index dd607a8,5070f96..0000000
--- a/mm/slab.c
+++ b/mm/slab.c
@@@ -3341,8 -3367,8 +3341,8 @@@ static void *fallback_alloc(struct kmem
  	local_flags = flags & (GFP_CONSTRAINT_MASK|GFP_RECLAIM_MASK);
  
  retry_cpuset:
- 	cpuset_mems_cookie = get_mems_allowed();
+ 	cpuset_mems_cookie = read_mems_allowed_begin();
 -	zonelist = node_zonelist(slab_node(current->mempolicy), flags);
 +	zonelist = node_zonelist(slab_node(), flags);
  
  retry:
  	/*
diff --cc mm/slub.c
index f96d8bc,fb2ef09..0000000
--- a/mm/slub.c
+++ b/mm/slub.c
@@@ -1616,8 -1616,8 +1616,8 @@@ static void *get_any_partial(struct kme
  		return NULL;
  
  	do {
- 		cpuset_mems_cookie = get_mems_allowed();
+ 		cpuset_mems_cookie = read_mems_allowed_begin();
 -		zonelist = node_zonelist(slab_node(current->mempolicy), flags);
 +		zonelist = node_zonelist(slab_node(), flags);
  		for_each_zone_zonelist(zone, z, zonelist, high_zoneidx) {
  			struct kmem_cache_node *n;
  

[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]

Re: linux-next: manual merge of the tip tree with the slab tree

[David Rientjes]

On Thu, 21 Jun 2012, Stephen Rothwell wrote:

> diff --cc mm/slab.c
> index dd607a8,5070f96..0000000
> --- a/mm/slab.c
> +++ b/mm/slab.c
> @@@ -3341,8 -3367,8 +3341,8 @@@ static void *fallback_alloc(struct kmem
>   	local_flags = flags & (GFP_CONSTRAINT_MASK|GFP_RECLAIM_MASK);
>   
>   retry_cpuset:
> - 	cpuset_mems_cookie = get_mems_allowed();
> + 	cpuset_mems_cookie = read_mems_allowed_begin();
>  -	zonelist = node_zonelist(slab_node(current->mempolicy), flags);
>  +	zonelist = node_zonelist(slab_node(), flags);
>   
>   retry:
>   	/*
> diff --cc mm/slub.c
> index f96d8bc,fb2ef09..0000000
> --- a/mm/slub.c
> +++ b/mm/slub.c
> @@@ -1616,8 -1616,8 +1616,8 @@@ static void *get_any_partial(struct kme
>   		return NULL;
>   
>   	do {
> - 		cpuset_mems_cookie = get_mems_allowed();
> + 		cpuset_mems_cookie = read_mems_allowed_begin();
>  -		zonelist = node_zonelist(slab_node(current->mempolicy), flags);
>  +		zonelist = node_zonelist(slab_node(), flags);
>   		for_each_zone_zonelist(zone, z, zonelist, high_zoneidx) {
>   			struct kmem_cache_node *n;
>   

Looks good.