Re: [PATCH 4/7] mm, pagemap: remove SLOB and SLQB from comments and documentation

[Vlastimil Babka]

On 3/14/23 09:19, Hyeonggon Yoo wrote:
> On Fri, Mar 10, 2023 at 11:32:06AM +0100, Vlastimil Babka wrote:
>> SLOB has been removed and SLQB never merged, so remove their mentions
>> from comments and documentation of pagemap.
>> 
>> Signed-off-by: Vlastimil Babka <vbabka@xxxxxxx>
>> ---
>>  Documentation/admin-guide/mm/pagemap.rst | 6 +++---
>>  fs/proc/page.c                           | 5 ++---
>>  2 files changed, 5 insertions(+), 6 deletions(-)
>> 
>> diff --git a/Documentation/admin-guide/mm/pagemap.rst b/Documentation/admin-guide/mm/pagemap.rst
>> index b5f970dc91e7..bb4aa897a773 100644
>> --- a/Documentation/admin-guide/mm/pagemap.rst
>> +++ b/Documentation/admin-guide/mm/pagemap.rst
>> @@ -91,9 +91,9 @@ Short descriptions to the page flags
>>     The page is being locked for exclusive access, e.g. by undergoing read/write
>>     IO.
>>  7 - SLAB
>> -   The page is managed by the SLAB/SLOB/SLUB/SLQB kernel memory allocator.
>> -   When compound page is used, SLUB/SLQB will only set this flag on the head
>> -   page; SLOB will not flag it at all.
>> +   The page is managed by the SLAB/SLUB kernel memory allocator.
>> +   When compound page is used, either will only set this flag on the head
>> +   page..
>>  10 - BUDDY
>>      A free memory block managed by the buddy system allocator.
>>      The buddy system organizes free memory in blocks of various orders.
>> diff --git a/fs/proc/page.c b/fs/proc/page.c
>> index 6249c347809a..1356aeffd8dc 100644
>> --- a/fs/proc/page.c
>> +++ b/fs/proc/page.c
>> @@ -125,7 +125,7 @@ u64 stable_page_flags(struct page *page)
>>  	/*
>>  	 * pseudo flags for the well known (anonymous) memory mapped pages
>>  	 *
>> -	 * Note that page->_mapcount is overloaded in SLOB/SLUB/SLQB, so the
>> +	 * Note that page->_mapcount is overloaded in SLAB/SLUB, so the
> 
> SLUB does not overload _mapcount.

True, I overlooked that, thanks.

>>  	 * simple test in page_mapped() is not enough.
>>  	 */
>>  	if (!PageSlab(page) && page_mapped(page))
>> @@ -166,8 +166,7 @@ u64 stable_page_flags(struct page *page)
>>  
>>  	/*
>>  	 * Caveats on high order pages: page->_refcount will only be set
>> -	 * -1 on the head page; SLUB/SLQB do the same for PG_slab;
>> -	 * SLOB won't set PG_slab at all on compound pages.
>> +	 * -1 on the head page; SLAB/SLUB do the same for PG_slab;
> 
> I think this comment could be just saying that PG_buddy is only set on
> head page, not saying
> 
> _refcount is set to -1 on head page (is it even correct?)

It's not, that scheme is outdated. So I'll have it mention PG_buddy as you
suggest, but PG_slab also needs special care as it's not set on tail pages.
But I noticed the compound_head() is unnecessary as that's covered by
PageSlab() which is defined as PF_NO_TAIL. So the sum of modifications to
this patch:

diff --git a/Documentation/admin-guide/mm/pagemap.rst b/Documentation/admin-guide/mm/pagemap.rst
index bb4aa897a773..c8f380271cad 100644
--- a/Documentation/admin-guide/mm/pagemap.rst
+++ b/Documentation/admin-guide/mm/pagemap.rst
@@ -93,7 +93,7 @@ Short descriptions to the page flags
 7 - SLAB
    The page is managed by the SLAB/SLUB kernel memory allocator.
    When compound page is used, either will only set this flag on the head
-   page..
+   page.
 10 - BUDDY
     A free memory block managed by the buddy system allocator.
     The buddy system organizes free memory in blocks of various orders.
diff --git a/fs/proc/page.c b/fs/proc/page.c
index 1356aeffd8dc..195b077c0fac 100644
--- a/fs/proc/page.c
+++ b/fs/proc/page.c
@@ -125,7 +125,7 @@ u64 stable_page_flags(struct page *page)
 	/*
 	 * pseudo flags for the well known (anonymous) memory mapped pages
 	 *
-	 * Note that page->_mapcount is overloaded in SLAB/SLUB, so the
+	 * Note that page->_mapcount is overloaded in SLAB, so the
 	 * simple test in page_mapped() is not enough.
 	 */
 	if (!PageSlab(page) && page_mapped(page))
@@ -165,8 +165,8 @@ u64 stable_page_flags(struct page *page)
 
 
 	/*
-	 * Caveats on high order pages: page->_refcount will only be set
-	 * -1 on the head page; SLAB/SLUB do the same for PG_slab;
+	 * Caveats on high order pages: PG_buddy and PG_slab will only be set
+	 * on the head page.
 	 */
 	if (PageBuddy(page))
 		u |= 1 << KPF_BUDDY;
@@ -184,7 +184,7 @@ u64 stable_page_flags(struct page *page)
 	u |= kpf_copy_bit(k, KPF_LOCKED,	PG_locked);
 
 	u |= kpf_copy_bit(k, KPF_SLAB,		PG_slab);
-	if (PageTail(page) && PageSlab(compound_head(page)))
+	if (PageTail(page) && PageSlab(page))
 		u |= 1 << KPF_SLAB;
 
 	u |= kpf_copy_bit(k, KPF_ERROR,		PG_error);