Re: [PATCHv2 3/8] mm: Accept memory in __alloc_pages_bulk().

Re: [PATCHv2 3/8] mm: Accept memory in __alloc_pages_bulk().

[Tom Lendacky]

On 8/9/24 06:48, Kirill A. Shutemov wrote:
> Currently, the kernel only accepts memory in get_page_from_freelist(),
> but there is another path that directly takes pages from free lists -
> __alloc_page_bulk(). This function can consume all accepted memory and
> will resort to __alloc_pages_noprof() if necessary.
> 
> Conditionally accepted in __alloc_pages_bulk().
> 
> The same issue may arise due to deferred page initialization. Kick the
> deferred initialization machinery before abandoning the zone, as the
> kernel does in get_page_from_freelist().

Is the deferred page init issue an existing problem? In other words,
should it be a separate patch with a Fixes: tag so it can go back to
stable kernels?

Thanks,
Tom

> 
> Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
> Acked-by: David Hildenbrand <david@redhat.com>
> ---
>  mm/page_alloc.c | 11 +++++++++++
>  1 file changed, 11 insertions(+)
> 
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index f7bb885aab07..ed62ecd6775f 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -4613,12 +4613,23 @@ unsigned long alloc_pages_bulk_noprof(gfp_t gfp, int preferred_nid,
>  			goto failed;
>  		}
>  
> +		cond_accept_memory(zone, 0);
> +retry_this_zone:
>  		mark = wmark_pages(zone, alloc_flags & ALLOC_WMARK_MASK) + nr_pages;
>  		if (zone_watermark_fast(zone, 0,  mark,
>  				zonelist_zone_idx(ac.preferred_zoneref),
>  				alloc_flags, gfp)) {
>  			break;
>  		}
> +
> +		if (cond_accept_memory(zone, 0))
> +			goto retry_this_zone;
> +
> +		/* Try again if zone has deferred pages */
> +		if (deferred_pages_enabled()) {
> +			if (_deferred_grow_zone(zone, 0))
> +				goto retry_this_zone;
> +		}
>  	}
>  
>  	/*

Re: [PATCHv2 3/8] mm: Accept memory in __alloc_pages_bulk().

[Kirill A. Shutemov]

On Fri, Aug 09, 2024 at 10:31:03AM -0500, Tom Lendacky wrote:
> On 8/9/24 06:48, Kirill A. Shutemov wrote:
> > Currently, the kernel only accepts memory in get_page_from_freelist(),
> > but there is another path that directly takes pages from free lists -
> > __alloc_page_bulk(). This function can consume all accepted memory and
> > will resort to __alloc_pages_noprof() if necessary.
> > 
> > Conditionally accepted in __alloc_pages_bulk().
> > 
> > The same issue may arise due to deferred page initialization. Kick the
> > deferred initialization machinery before abandoning the zone, as the
> > kernel does in get_page_from_freelist().
> 
> Is the deferred page init issue an existing problem? In other words,
> should it be a separate patch with a Fixes: tag so it can go back to
> stable kernels?

Yes, it is existing problem, but the problem is transient (deferred thread
will add more pages) and there's fallback to __alloc_pages_noprof() which
knowns how to handle deferred pages. I don't think it is worth
backporting.

-- 
  Kiryl Shutsemau / Kirill A. Shutemov

Re: [PATCHv2 1/8] mm: Fix endless reclaim on machines with unaccepted memory

[Jianxiong Gao]

On Fri, Aug 9, 2024 at 4:49 AM Kirill A. Shutemov
<kirill.shutemov@linux.intel.com> wrote:
>
> Unaccepted memory is considered unusable free memory, which is not
> counted as free on the zone watermark check. This causes
> get_page_from_freelist() to accept more memory to hit the high
> watermark, but it creates problems in the reclaim path.
>
> The reclaim path encounters a failed zone watermark check and attempts
> to reclaim memory. This is usually successful, but if there is little or
> no reclaimable memory, it can result in endless reclaim with little to
> no progress. This can occur early in the boot process, just after start
> of the init process when the only reclaimable memory is the page cache
> of the init executable and its libraries.
>
> Make unaccepted memory free from watermark check point of view. This way
> unaccepted memory will never be the trigger of memory reclaim.
> Accept more memory in the get_page_from_freelist() if needed.
>
> Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
> Reported-by: Jianxiong Gao <jxgao@google.com>
> Acked-by: David Hildenbrand <david@redhat.com>
> Fixes: dcdfdd40fa82 ("mm: Add support for unaccepted memory")
> Cc: stable@vger.kernel.org # v6.5+

Tested-by: Jianxiong Gao <jxgao@google.com>
I have verified that the patch fixes the systemd issue reported.