All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Vishal Moola (Oracle)" <vishal.moola@gmail.com>
To: Kefeng Wang <wangkefeng.wang@huawei.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
	Muchun Song <muchun.song@linux.dev>,
	Oscar Salvador <osalvador@suse.de>,
	David Hildenbrand <david@redhat.com>,
	linux-mm@kvack.org
Subject: Re: [PATCH 3/7] mm; hugetlb: simpify alloc_buddy_hugetlb_folio()
Date: Mon, 4 Aug 2025 13:56:01 -0700	[thread overview]
Message-ID: <aJEeYY_T7pEhBbll@fedora> (raw)
In-Reply-To: <20250802073107.2787975-4-wangkefeng.wang@huawei.com>

On Sat, Aug 02, 2025 at 03:31:03PM +0800, Kefeng Wang wrote:
> Check folio once instead of three times in alloc_buddy_hugetlb_folio().
> 
> Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
> ---
>  mm/hugetlb.c | 37 +++++++++++++++++--------------------
>  1 file changed, 17 insertions(+), 20 deletions(-)
> 
> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> index afec5a6a8aca..436403fb0bed 100644
> --- a/mm/hugetlb.c
> +++ b/mm/hugetlb.c
> @@ -1949,29 +1949,26 @@ static struct folio *alloc_buddy_hugetlb_folio(struct hstate *h,
>  		nid = numa_mem_id();
>  
>  	folio = (struct folio *)__alloc_frozen_pages(gfp_mask, order, nid, nmask);
> -
> -	/*
> -	 * If we did not specify __GFP_RETRY_MAYFAIL, but still got a
> -	 * folio this indicates an overall state change.  Clear bit so
> -	 * that we resume normal 'try hard' allocations.
> -	 */
> -	if (node_alloc_noretry && folio && !alloc_try_hard)
> -		node_clear(nid, *node_alloc_noretry);
> -
> -	/*
> -	 * If we tried hard to get a folio but failed, set bit so that
> -	 * subsequent attempts will not try as hard until there is an
> -	 * overall state change.
> -	 */
> -	if (node_alloc_noretry && !folio && alloc_try_hard)
> -		node_set(nid, *node_alloc_noretry);
> -
> -	if (!folio) {

I really don't like the way you've done this below. I feel like it makes
the code harder to understand since it feels like we've effectively
divided this function into 2 different ones (depending on if we
successfully got a folio).

I don't have a better idea, but I also don't think eliminating those 2
extra checks is worth increasing the complexity like this.

> +	if (folio) {
> +		/*
> +		 * If we did not specify __GFP_RETRY_MAYFAIL, but still got a
> +		 * folio this indicates an overall state change.  Clear bit so
> +		 * that we resume normal 'try hard' allocations.
> +		 */
> +		if (node_alloc_noretry && !alloc_try_hard)
> +			node_clear(nid, *node_alloc_noretry);
> +		__count_vm_event(HTLB_BUDDY_PGALLOC);
> +	} else {
> +		/*
> +		 * If we tried hard to get a folio but failed, set bit so that
> +		 * subsequent attempts will not try as hard until there is an
> +		 * overall state change.
> +		 */
> +		if (node_alloc_noretry && alloc_try_hard)
> +			node_set(nid, *node_alloc_noretry);
>  		__count_vm_event(HTLB_BUDDY_PGALLOC_FAIL);
> -		return NULL;
>  	}
>  
> -	__count_vm_event(HTLB_BUDDY_PGALLOC);
>  	return folio;
>  }
>  
> -- 
> 2.27.0
> 
> 


  parent reply	other threads:[~2025-08-04 20:56 UTC|newest]

Thread overview: 33+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-08-02  7:31 [PATCH 0/7] mm: hugetlb: cleanup and allocate frozen hugetlb folio Kefeng Wang
2025-08-02  7:31 ` [PATCH 1/7] mm: hugetlb: convert to alloc_fresh_hugetlb_hvo_folio() Kefeng Wang
2025-08-04 15:41   ` Sidhartha Kumar
2025-08-04 20:27   ` Vishal Moola (Oracle)
2025-08-05 14:21     ` Kefeng Wang
2025-08-05 17:56       ` Vishal Moola (Oracle)
2025-08-05 22:18   ` jane.chu
2025-08-06  0:33     ` jane.chu
2025-08-06  0:56       ` Kefeng Wang
2025-08-06 18:16         ` jane.chu
2025-08-02  7:31 ` [PATCH 2/7] mm: hugetlb: convert to prep_account_new_hugetlb_folio() Kefeng Wang
2025-08-04 15:54   ` Sidhartha Kumar
2025-08-04 20:36   ` Vishal Moola (Oracle)
2025-08-05 14:21     ` Kefeng Wang
2025-08-06  0:53     ` jane.chu
2025-08-02  7:31 ` [PATCH 3/7] mm; hugetlb: simpify alloc_buddy_hugetlb_folio() Kefeng Wang
2025-08-04 15:57   ` Sidhartha Kumar
2025-08-04 20:56   ` Vishal Moola (Oracle) [this message]
2025-08-05 14:22     ` Kefeng Wang
2025-08-02  7:31 ` [PATCH 4/7] mm: hugetlb: directly pass order when allocate a hugetlb folio Kefeng Wang
2025-08-04 16:22   ` Sidhartha Kumar
2025-08-06 20:05   ` jane.chu
2025-08-06 20:17     ` jane.chu
2025-08-02  7:31 ` [PATCH 5/7] mm: hugetlb: remove struct hstate from init_new_hugetlb_folio() Kefeng Wang
2025-08-04 16:13   ` Sidhartha Kumar
2025-08-06 20:08   ` jane.chu
2025-08-02  7:31 ` [PATCH 6/7] mm: hugeltb: check NUMA_NO_NODE in only_alloc_fresh_hugetlb_folio() Kefeng Wang
2025-08-04 19:09   ` Sidhartha Kumar
2025-08-06 20:43   ` jane.chu
2025-08-12 12:13     ` Kefeng Wang
2025-08-02  7:31 ` [PATCH 7/7] mm: hugetlb: allocate frozen pages in alloc_gigantic_folio() Kefeng Wang
2025-08-07  1:22   ` jane.chu
2025-08-12 12:11     ` Kefeng Wang

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=aJEeYY_T7pEhBbll@fedora \
    --to=vishal.moola@gmail.com \
    --cc=akpm@linux-foundation.org \
    --cc=david@redhat.com \
    --cc=linux-mm@kvack.org \
    --cc=muchun.song@linux.dev \
    --cc=osalvador@suse.de \
    --cc=wangkefeng.wang@huawei.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.