Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Uros Bizjak <ubizjak@gmail.com>
To:     linux-kernel@vger.kernel.org
Cc:     Uros Bizjak <ubizjak@gmail.com>,
        Andrew Morton <akpm@linux-foundation.org>
Subject: [PATCH] lib/genalloc: use try_cmpxchg in {set,clear}_bits_ll
Date:   Wed, 18 Jan 2023 16:07:03 +0100
Message-Id: <20230118150703.4024-1-ubizjak@gmail.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

Use try_cmpxchg instead of cmpxchg (*ptr, old, new) == old in
{set,clear}_bits_ll.  x86 CMPXCHG instruction returns success in ZF
flag, so this change saves a compare after cmpxchg (and related move
instruction in front of cmpxchg).

Also, try_cmpxchg implicitly assigns old *ptr value to "old"
when cmpxchg fails.

Note that the value from *ptr should be read using READ_ONCE to prevent
the compiler from merging, refetching or reordering the read.

The patch also declares these two functions inline, to ensure inlining.

No functional change intended.

Signed-off-by: Uros Bizjak <ubizjak@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
---
 lib/genalloc.c | 18 ++++++++----------
 1 file changed, 8 insertions(+), 10 deletions(-)

diff --git a/lib/genalloc.c b/lib/genalloc.c
index 00fc50d0a640..0c883d6fbd44 100644
--- a/lib/genalloc.c
+++ b/lib/genalloc.c
@@ -40,32 +40,30 @@ static inline size_t chunk_size(const struct gen_pool_chunk *chunk)
 	return chunk->end_addr - chunk->start_addr + 1;
 }
 
-static int set_bits_ll(unsigned long *addr, unsigned long mask_to_set)
+static inline int
+set_bits_ll(unsigned long *addr, unsigned long mask_to_set)
 {
-	unsigned long val, nval;
+	unsigned long val = READ_ONCE(*addr);
 
-	nval = *addr;
 	do {
-		val = nval;
 		if (val & mask_to_set)
 			return -EBUSY;
 		cpu_relax();
-	} while ((nval = cmpxchg(addr, val, val | mask_to_set)) != val);
+	} while (!try_cmpxchg(addr, &val, val | mask_to_set));
 
 	return 0;
 }
 
-static int clear_bits_ll(unsigned long *addr, unsigned long mask_to_clear)
+static inline int
+clear_bits_ll(unsigned long *addr, unsigned long mask_to_clear)
 {
-	unsigned long val, nval;
+	unsigned long val = READ_ONCE(*addr);
 
-	nval = *addr;
 	do {
-		val = nval;
 		if ((val & mask_to_clear) != mask_to_clear)
 			return -EBUSY;
 		cpu_relax();
-	} while ((nval = cmpxchg(addr, val, val & ~mask_to_clear)) != val);
+	} while (!try_cmpxchg(addr, &val, val & ~mask_to_clear));
 
 	return 0;
 }
-- 
2.39.0