netdev - [PATCH net-next 17/18] rds: Acquire per-CPU pointer within BH disabled section.

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20250309144653.825351-18-bigeasy@linutronix.de>
Date: Sun,  9 Mar 2025 15:46:52 +0100
From: Sebastian Andrzej Siewior <bigeasy@...utronix.de>
To: netdev@...r.kernel.org,
	linux-rt-devel@...ts.linux.dev
Cc: "David S. Miller" <davem@...emloft.net>,
	Eric Dumazet <edumazet@...gle.com>,
	Jakub Kicinski <kuba@...nel.org>,
	Paolo Abeni <pabeni@...hat.com>,
	Simon Horman <horms@...nel.org>,
	Thomas Gleixner <tglx@...utronix.de>,
	Sebastian Andrzej Siewior <bigeasy@...utronix.de>,
	Allison Henderson <allison.henderson@...cle.com>,
	linux-rdma@...r.kernel.org
Subject: [PATCH net-next 17/18] rds: Acquire per-CPU pointer within BH disabled section.

rds_page_remainder_alloc() obtains the current CPU with get_cpu() while
disabling preemption. Then the CPU number is used to access the per-CPU
data structure via per_cpu().

This can be optimized by relying on local_bh_disable() to provide a
stable CPU number/ avoid migration and then using this_cpu_ptr() to
retrieve the data structure.

Cc: Allison Henderson <allison.henderson@...cle.com>
Cc: linux-rdma@...r.kernel.org
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@...utronix.de>
---
 net/rds/page.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/net/rds/page.c b/net/rds/page.c
index e0dd4f62ea47a..58a8548a915a9 100644
--- a/net/rds/page.c
+++ b/net/rds/page.c
@@ -86,8 +86,8 @@ int rds_page_remainder_alloc(struct scatterlist *scat, unsigned long bytes,
 		goto out;
 	}
 
-	rem = &per_cpu(rds_page_remainders, get_cpu());
 	local_bh_disable();
+	rem = this_cpu_ptr(&rds_page_remainders);
 
 	while (1) {
 		/* avoid a tiny region getting stuck by tossing it */
@@ -116,12 +116,11 @@ int rds_page_remainder_alloc(struct scatterlist *scat, unsigned long bytes,
 
 		/* alloc if there is nothing for us to use */
 		local_bh_enable();
-		put_cpu();
 
 		page = alloc_page(gfp);
 
-		rem = &per_cpu(rds_page_remainders, get_cpu());
 		local_bh_disable();
+		rem = this_cpu_ptr(&rds_page_remainders);
 
 		if (!page) {
 			ret = -ENOMEM;
@@ -140,7 +139,6 @@ int rds_page_remainder_alloc(struct scatterlist *scat, unsigned long bytes,
 	}
 
 	local_bh_enable();
-	put_cpu();
 out:
 	rdsdebug("bytes %lu ret %d %p %u %u\n", bytes, ret,
 		 ret ? NULL : sg_page(scat), ret ? 0 : scat->offset,
-- 
2.47.2