linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: "Liam R. Howlett" <Liam.Howlett@Oracle.com>
To: Peng Zhang <zhangpeng.00@bytedance.com>
Cc: akpm@linux-foundation.org, linux-mm@kvack.org,
	linux-kernel@vger.kernel.org, maple-tree@lists.infradead.org
Subject: Re: [PATCH 04/10] maple_tree: Simplify mas_is_span_wr()
Date: Mon, 15 May 2023 12:06:44 -0400	[thread overview]
Message-ID: <20230515160644.lqhrclw3kcutsafc@revolver> (raw)
In-Reply-To: <20230515131757.60035-5-zhangpeng.00@bytedance.com>

* Peng Zhang <zhangpeng.00@bytedance.com> [230515 09:18]:
> Make the code for detecting spanning writes more concise.
> 
> Signed-off-by: Peng Zhang <zhangpeng.00@bytedance.com>
> ---
>  lib/maple_tree.c | 36 +++++++++---------------------------
>  1 file changed, 9 insertions(+), 27 deletions(-)
> 
> diff --git a/lib/maple_tree.c b/lib/maple_tree.c
> index 43a25d3042c1b..fbb6efc40e576 100644
> --- a/lib/maple_tree.c
> +++ b/lib/maple_tree.c
> @@ -3728,41 +3728,23 @@ static bool mas_is_span_wr(struct ma_wr_state *wr_mas)
>  {
>  	unsigned long max;
>  	unsigned long last = wr_mas->mas->last;
> -	unsigned long piv = wr_mas->r_max;
>  	enum maple_type type = wr_mas->type;
>  	void *entry = wr_mas->entry;
>  
> -	/* Contained in this pivot */
> -	if (piv > last)

I test this first since it is far more likely than the rest of the cases
in this function.  Most writes will land in a leaf node eventually, so
this will be true most of the time.  I'd like to keep it as a fast path.

> +	max = unlikely(ma_is_leaf(type)) ? wr_mas->mas->max : wr_mas->r_max;

Please expand this to multiple lines.  It is better to be more clear
than compact lines of code.

> +	if (last < max) {
> +		/* Contained in this pivot or this leaf node */
>  		return false;
> -
> -	max = wr_mas->mas->max;
> -	if (unlikely(ma_is_leaf(type))) {
> -		/* Fits in the node, but may span slots. */
> -		if (last < max)
> -			return false;
> -
> -		/* Writes to the end of the node but not null. */
> -		if ((last == max) && entry)
> -			return false;
> -
> +	} else if (last == max) {
>  		/*
> -		 * Writing ULONG_MAX is not a spanning write regardless of the
> -		 * value being written as long as the range fits in the node.
> +		 * The last entry of leaf node cannot be NULL unless it is the
> +		 * rightmost node (writing ULONG_MAX), otherwise it spans slots.
> +		 * If this is not leaf node, detect spanning store wr walk.
>  		 */
> -		if ((last == ULONG_MAX) && (last == max))
> -			return false;
> -	} else if (piv == last) {
> -		if (entry)
> -			return false;
> -
> -		/* Detect spanning store wr walk */
> -		if (last == ULONG_MAX)
> +		if (entry || last == ULONG_MAX)
>  			return false;
>  	}
> -
> -	trace_ma_write(__func__, wr_mas->mas, piv, entry);
> -
> +	trace_ma_write(__func__, wr_mas->mas, wr_mas->r_max, entry);
>  	return true;
>  }
>  
> -- 
> 2.20.1
> 


  reply	other threads:[~2023-05-15 16:07 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-05-15 13:17 [PATCH 00/10] Clean ups for maple tree Peng Zhang
2023-05-15 13:17 ` [PATCH 01/10] maple_tree: Drop the test code for mtree_alloc_{range,rrange}() Peng Zhang
2023-05-15 16:52   ` Liam R. Howlett
2023-05-15 13:17 ` [PATCH 02/10] maple_tree: Drop mtree_alloc_{range,rrange}() and related functions Peng Zhang
2023-05-15 16:52   ` Liam R. Howlett
2023-05-15 17:27   ` Matthew Wilcox
2023-05-15 17:35     ` Liam R. Howlett
2023-05-16  0:39       ` Peng Zhang
2023-05-15 13:17 ` [PATCH 03/10] maple_tree: Remove __must_hold() which does not work Peng Zhang
2023-05-15 14:55   ` Matthew Wilcox
2023-05-16  0:42     ` Peng Zhang
2023-05-15 15:00   ` Liam R. Howlett
2023-05-15 13:17 ` [PATCH 04/10] maple_tree: Simplify mas_is_span_wr() Peng Zhang
2023-05-15 16:06   ` Liam R. Howlett [this message]
2023-05-15 13:17 ` [PATCH 05/10] maple_tree: Make the code symmetrical in mas_wr_extend_null() Peng Zhang
2023-05-15 16:54   ` Liam R. Howlett
2023-05-15 13:17 ` [PATCH 06/10] maple_tree: Wrap the replace operation with an inline function Peng Zhang
2023-05-15 17:07   ` Liam R. Howlett
2023-05-16  0:46     ` Peng Zhang
2023-05-16 14:16       ` Liam R. Howlett
2023-05-16 14:22         ` Peng Zhang
2023-05-15 13:17 ` [PATCH 07/10] maple_tree: Add mas_wr_new_end() to calculate new_end accurately Peng Zhang
2023-05-15 13:17 ` [PATCH 08/10] maple_tree: Add comments and some minor cleanups to mas_wr_append() Peng Zhang
2023-05-15 17:29   ` Liam R. Howlett
2023-05-16 10:06     ` Peng Zhang
2023-05-15 13:17 ` [PATCH 09/10] maple_tree: Rework mas_wr_slot_store() to be cleaner and more efficient Peng Zhang
2023-05-15 18:01   ` Liam R. Howlett
2023-05-16  7:27     ` Peng Zhang
2023-05-16 14:17       ` Liam R. Howlett
2023-05-15 13:17 ` [PATCH 10/10] maple_tree: Simplify and clean up mas_wr_node_store() Peng Zhang
2023-05-15 18:58   ` Liam R. Howlett
2023-05-16  0:36     ` Peng Zhang
2023-05-16 10:53     ` Peng Zhang
2023-05-16 15:52       ` Liam R. Howlett
2023-05-16 23:53         ` Peng Zhang
2023-05-17  3:10         ` Peng Zhang

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=20230515160644.lqhrclw3kcutsafc@revolver \
    --to=liam.howlett@oracle.com \
    --cc=akpm@linux-foundation.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=maple-tree@lists.infradead.org \
    --cc=zhangpeng.00@bytedance.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox