Received-SPF: pass (google.com: domain of linux-kernel+bounces-192638-linux.lists.archive=gmail.com@vger.kernel.org designates 139.178.88.99 as permitted sender) client-ip=139.178.88.99;
From: Alexandre Ghiti <alexghiti@rivosinc.com>
To: Jonathan Corbet <corbet@lwn.net>,
	Paul Walmsley <paul.walmsley@sifive.com>,
	Palmer Dabbelt <palmer@dabbelt.com>,
	Albert Ou <aou@eecs.berkeley.edu>,
	Peter Zijlstra <peterz@infradead.org>,
	Ingo Molnar <mingo@redhat.com>,
	Will Deacon <will@kernel.org>,
	Waiman Long <longman@redhat.com>,
	Boqun Feng <boqun.feng@gmail.com>,
	Arnd Bergmann <arnd@arndb.de>,
	Leonardo Bras <leobras@redhat.com>,
	Guo Ren <guoren@kernel.org>,
	linux-doc@vger.kernel.org,
	linux-kernel@vger.kernel.org,
	linux-riscv@lists.infradead.org,
	linux-arch@vger.kernel.org
Cc: Alexandre Ghiti <alexghiti@rivosinc.com>
Subject: [PATCH 3/7] riscv: Implement arch_cmpxchg128() using Zacas
Date: Tue, 28 May 2024 17:10:48 +0200
Message-Id: <20240528151052.313031-4-alexghiti@rivosinc.com>
In-Reply-To: <20240528151052.313031-1-alexghiti@rivosinc.com>
References: <20240528151052.313031-1-alexghiti@rivosinc.com>
Precedence: bulk
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit

Now that Zacas is supported in the kernel, let's use the double word
atomic version of amocas to improve the SLUB allocator.

Note that we have to select fixed registers, otherwise gcc fails to pick
even registers and then produces a reserved encoding which fails to
assemble.

Signed-off-by: Alexandre Ghiti <alexghiti@rivosinc.com>
---
 arch/riscv/Kconfig               |  1 +
 arch/riscv/include/asm/cmpxchg.h | 41 ++++++++++++++++++++++++++++++++
 2 files changed, 42 insertions(+)

diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
index 05597719bb1c..184a9edb04e0 100644
--- a/arch/riscv/Kconfig
+++ b/arch/riscv/Kconfig
@@ -101,6 +101,7 @@ config RISCV
 	select GENERIC_VDSO_TIME_NS if HAVE_GENERIC_VDSO
 	select HARDIRQS_SW_RESEND
 	select HAS_IOPORT if MMU
+	select HAVE_ALIGNED_STRUCT_PAGE
 	select HAVE_ARCH_AUDITSYSCALL
 	select HAVE_ARCH_HUGE_VMALLOC if HAVE_ARCH_HUGE_VMAP
 	select HAVE_ARCH_HUGE_VMAP if MMU && 64BIT && !XIP_KERNEL
diff --git a/arch/riscv/include/asm/cmpxchg.h b/arch/riscv/include/asm/cmpxchg.h
index 65de9771078e..0789fbe38b23 100644
--- a/arch/riscv/include/asm/cmpxchg.h
+++ b/arch/riscv/include/asm/cmpxchg.h
@@ -242,4 +242,45 @@ end:									\
 	arch_cmpxchg_relaxed((ptr), (o), (n));				\
 })
 
+#ifdef CONFIG_RISCV_ISA_ZACAS
+
+#define system_has_cmpxchg128()						\
+			riscv_has_extension_unlikely(RISCV_ISA_EXT_ZACAS)
+
+union __u128_halves {
+	u128 full;
+	struct {
+		u64 low, high;
+	};
+};
+
+#define __arch_cmpxchg128(p, o, n, prepend, append)			\
+({									\
+	__typeof__(*(p)) __o = (o);					\
+	union __u128_halves new = { .full = (n) };			\
+	union __u128_halves old = { .full = (__o) };			\
+	register unsigned long x6 asm ("x6") = new.low;			\
+	register unsigned long x7 asm ("x7") = new.high;		\
+	register unsigned long x28 asm ("x28") = old.low;		\
+	register unsigned long x29 asm ("x29") = old.high;		\
+									\
+	__asm__ __volatile__ (						\
+		prepend							\
+		"	amocas.q %0, %z2, %1\n"				\
+		append							\
+		: "+&r" (x28), "+A" (*(p))				\
+		: "rJ" (x6)						\
+		: "memory");						\
+									\
+	__o;								\
+})
+
+#define arch_cmpxchg128(ptr, o, n)					\
+	__arch_cmpxchg128((ptr), (o), (n), "", "     fence rw, rw\n")
+
+#define arch_cmpxchg128_local(ptr, o, n)				\
+	__arch_cmpxchg128((ptr), (o), (n), "", "")
+
+#endif /* CONFIG_RISCV_ISA_ZACAS */
+
 #endif /* _ASM_RISCV_CMPXCHG_H */
-- 
2.39.2