[14/30] memblock: add align parameter to memblock_alloc_node()

Message ID 1536927045-23536-15-git-send-email-rppt@linux.vnet.ibm.com
State New
Headers show
Series
  • mm: remove bootmem allocator
Related show

Commit Message

Mike Rapoport Sept. 14, 2018, 12:10 p.m.
With the align parameter memblock_alloc_node() can be used as drop in
replacement for alloc_bootmem_pages_node() and __alloc_bootmem_node(),
which is done in the following patches.

Signed-off-by: Mike Rapoport <rppt@linux.vnet.ibm.com>
---
 include/linux/bootmem.h | 4 ++--
 mm/sparse.c             | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

Comments

Michal Hocko Sept. 26, 2018, 9:31 a.m. | #1
On Fri 14-09-18 15:10:29, Mike Rapoport wrote:
> With the align parameter memblock_alloc_node() can be used as drop in
> replacement for alloc_bootmem_pages_node() and __alloc_bootmem_node(),
> which is done in the following patches.

/me confused. Why do we need this patch at all? Maybe it should be
folded into the later patch you are refereing here?

> Signed-off-by: Mike Rapoport <rppt@linux.vnet.ibm.com>
> ---
>  include/linux/bootmem.h | 4 ++--
>  mm/sparse.c             | 2 +-
>  2 files changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/include/linux/bootmem.h b/include/linux/bootmem.h
> index 7d91f0f..3896af2 100644
> --- a/include/linux/bootmem.h
> +++ b/include/linux/bootmem.h
> @@ -157,9 +157,9 @@ static inline void * __init memblock_alloc_from_nopanic(
>  }
>  
>  static inline void * __init memblock_alloc_node(
> -						phys_addr_t size, int nid)
> +		phys_addr_t size, phys_addr_t align, int nid)
>  {
> -	return memblock_alloc_try_nid(size, 0, BOOTMEM_LOW_LIMIT,
> +	return memblock_alloc_try_nid(size, align, BOOTMEM_LOW_LIMIT,
>  					    BOOTMEM_ALLOC_ACCESSIBLE, nid);
>  }
>  
> diff --git a/mm/sparse.c b/mm/sparse.c
> index 04e97af..509828f 100644
> --- a/mm/sparse.c
> +++ b/mm/sparse.c
> @@ -68,7 +68,7 @@ static noinline struct mem_section __ref *sparse_index_alloc(int nid)
>  	if (slab_is_available())
>  		section = kzalloc_node(array_size, GFP_KERNEL, nid);
>  	else
> -		section = memblock_alloc_node(array_size, nid);
> +		section = memblock_alloc_node(array_size, 0, nid);
>  
>  	return section;
>  }
> -- 
> 2.7.4
>
Michal Hocko Sept. 26, 2018, 9:36 a.m. | #2
On Wed 26-09-18 11:31:27, Michal Hocko wrote:
> On Fri 14-09-18 15:10:29, Mike Rapoport wrote:
> > With the align parameter memblock_alloc_node() can be used as drop in
> > replacement for alloc_bootmem_pages_node() and __alloc_bootmem_node(),
> > which is done in the following patches.
> 
> /me confused. Why do we need this patch at all? Maybe it should be
> folded into the later patch you are refereing here?

OK, I can see 1536927045-23536-17-git-send-email-rppt@linux.vnet.ibm.com
now. If you are going to repost for whatever reason please merge those
two. Also I would get rid of the implicit "0 implies SMP_CACHE_BYTES"
behavior. It is subtle and you have to dig deep to find that out. Why
not make it explicit?
Mike Rapoport Sept. 26, 2018, 1:43 p.m. | #3
On Wed, Sep 26, 2018 at 11:36:48AM +0200, Michal Hocko wrote:
> On Wed 26-09-18 11:31:27, Michal Hocko wrote:
> > On Fri 14-09-18 15:10:29, Mike Rapoport wrote:
> > > With the align parameter memblock_alloc_node() can be used as drop in
> > > replacement for alloc_bootmem_pages_node() and __alloc_bootmem_node(),
> > > which is done in the following patches.
> > 
> > /me confused. Why do we need this patch at all? Maybe it should be
> > folded into the later patch you are refereing here?
> 
> OK, I can see 1536927045-23536-17-git-send-email-rppt@linux.vnet.ibm.com
> now. If you are going to repost for whatever reason please merge those
> two. Also I would get rid of the implicit "0 implies SMP_CACHE_BYTES"
> behavior. It is subtle and you have to dig deep to find that out. Why
> not make it explicit?

Agree. I'd just prefer to make it a separate patch rather then resend the
whole series. 

> -- 
> Michal Hocko
> SUSE Labs
>
Michal Hocko Sept. 26, 2018, 2:23 p.m. | #4
On Wed 26-09-18 16:43:35, Mike Rapoport wrote:
> On Wed, Sep 26, 2018 at 11:36:48AM +0200, Michal Hocko wrote:
> > On Wed 26-09-18 11:31:27, Michal Hocko wrote:
> > > On Fri 14-09-18 15:10:29, Mike Rapoport wrote:
> > > > With the align parameter memblock_alloc_node() can be used as drop in
> > > > replacement for alloc_bootmem_pages_node() and __alloc_bootmem_node(),
> > > > which is done in the following patches.
> > > 
> > > /me confused. Why do we need this patch at all? Maybe it should be
> > > folded into the later patch you are refereing here?
> > 
> > OK, I can see 1536927045-23536-17-git-send-email-rppt@linux.vnet.ibm.com
> > now. If you are going to repost for whatever reason please merge those
> > two. Also I would get rid of the implicit "0 implies SMP_CACHE_BYTES"
> > behavior. It is subtle and you have to dig deep to find that out. Why
> > not make it explicit?
> 
> Agree. I'd just prefer to make it a separate patch rather then resend the
> whole series. 

Sure, no objection from me.

Patch

diff --git a/include/linux/bootmem.h b/include/linux/bootmem.h
index 7d91f0f..3896af2 100644
--- a/include/linux/bootmem.h
+++ b/include/linux/bootmem.h
@@ -157,9 +157,9 @@  static inline void * __init memblock_alloc_from_nopanic(
 }
 
 static inline void * __init memblock_alloc_node(
-						phys_addr_t size, int nid)
+		phys_addr_t size, phys_addr_t align, int nid)
 {
-	return memblock_alloc_try_nid(size, 0, BOOTMEM_LOW_LIMIT,
+	return memblock_alloc_try_nid(size, align, BOOTMEM_LOW_LIMIT,
 					    BOOTMEM_ALLOC_ACCESSIBLE, nid);
 }
 
diff --git a/mm/sparse.c b/mm/sparse.c
index 04e97af..509828f 100644
--- a/mm/sparse.c
+++ b/mm/sparse.c
@@ -68,7 +68,7 @@  static noinline struct mem_section __ref *sparse_index_alloc(int nid)
 	if (slab_is_available())
 		section = kzalloc_node(array_size, GFP_KERNEL, nid);
 	else
-		section = memblock_alloc_node(array_size, nid);
+		section = memblock_alloc_node(array_size, 0, nid);
 
 	return section;
 }