riscv: Add qspinlock support

In order to produce a generic kernel, a user can select
CONFIG_COMBO_SPINLOCKS which will fallback at runtime to the ticket
spinlock implementation if Zabha or Ziccrse are not present.

Note that we can't use alternatives here because the discovery of
extensions is done too late and we need to start with the qspinlock
implementation because the ticket spinlock implementation would pollute
the spinlock value, so let's use static keys.

This is largely based on Guo's work and Leonardo reviews at [1].

Link: https://lore.kernel.org/linux-riscv/20231225125847.2778638-1-guoren@kernel.org/ [1]
Signed-off-by: Guo Ren <guoren@kernel.org>
Signed-off-by: Alexandre Ghiti <alexghiti@rivosinc.com>
Reviewed-by: Andrea Parri <parri.andrea@gmail.com>
Link: https://lore.kernel.org/r/20241103145153.105097-14-alexghiti@rivosinc.com
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>

Documentation/features/locking/queued-spinlocks/arch-support.txt

@@ -20,7 +20,7 @@
     |    openrisc: |  ok  |
     |      parisc: | TODO |
     |     powerpc: |  ok  |
-    |       riscv: | TODO |
+    |       riscv: |  ok  |
     |        s390: | TODO |
     |          sh: | TODO |
     |       sparc: |  ok  |

arch/riscv/Kconfig

@@ -82,6 +82,7 @@ config RISCV
 	select ARCH_WANT_OPTIMIZE_HUGETLB_VMEMMAP
 	select ARCH_WANTS_NO_INSTR
 	select ARCH_WANTS_THP_SWAP if HAVE_ARCH_TRANSPARENT_HUGEPAGE
+	select ARCH_WEAK_RELEASE_ACQUIRE if ARCH_USE_QUEUED_SPINLOCKS
 	select BINFMT_FLAT_NO_DATA_START_OFFSET if !MMU
 	select BUILDTIME_TABLE_SORT if MMU
 	select CLINT_TIMER if RISCV_M_MODE
@@ -507,6 +508,39 @@ config NODES_SHIFT
 	  Specify the maximum number of NUMA Nodes available on the target
 	  system.  Increases memory reserved to accommodate various tables.
 
+choice
+	prompt "RISC-V spinlock type"
+	default RISCV_COMBO_SPINLOCKS
+
+config RISCV_TICKET_SPINLOCKS
+	bool "Using ticket spinlock"
+
+config RISCV_QUEUED_SPINLOCKS
+	bool "Using queued spinlock"
+	depends on SMP && MMU && NONPORTABLE
+	select ARCH_USE_QUEUED_SPINLOCKS
+	help
+	  The queued spinlock implementation requires the forward progress
+	  guarantee of cmpxchg()/xchg() atomic operations: CAS with Zabha or
+	  LR/SC with Ziccrse provide such guarantee.
+
+	  Select this if and only if Zabha or Ziccrse is available on your
+	  platform, RISCV_QUEUED_SPINLOCKS must not be selected for platforms
+	  without one of those extensions.
+
+	  If unsure, select RISCV_COMBO_SPINLOCKS, which will use qspinlocks
+	  when supported and otherwise ticket spinlocks.
+
+config RISCV_COMBO_SPINLOCKS
+	bool "Using combo spinlock"
+	depends on SMP && MMU
+	select ARCH_USE_QUEUED_SPINLOCKS
+	help
+	  Embed both queued spinlock and ticket lock so that the spinlock
+	  implementation can be chosen at runtime.
+
+endchoice
+
 config RISCV_ALTERNATIVE
 	bool
 	depends on !XIP_KERNEL

arch/riscv/include/asm/Kbuild

@@ -6,10 +6,12 @@ generic-y += early_ioremap.h
 generic-y += flat.h
 generic-y += kvm_para.h
 generic-y += mmzone.h
+generic-y += mcs_spinlock.h
 generic-y += parport.h
-generic-y += spinlock.h
 generic-y += spinlock_types.h
+generic-y += ticket_spinlock.h
 generic-y += qrwlock.h
 generic-y += qrwlock_types.h
+generic-y += qspinlock.h
 generic-y += user.h
 generic-y += vmlinux.lds.h

arch/riscv/include/asm/spinlock.h

@@ -0,0 +1,47 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+
+#ifndef __ASM_RISCV_SPINLOCK_H
+#define __ASM_RISCV_SPINLOCK_H
+
+#ifdef CONFIG_RISCV_COMBO_SPINLOCKS
+#define _Q_PENDING_LOOPS	(1 << 9)
+
+#define __no_arch_spinlock_redefine
+#include <asm/ticket_spinlock.h>
+#include <asm/qspinlock.h>
+#include <asm/jump_label.h>
+
+/*
+ * TODO: Use an alternative instead of a static key when we are able to parse
+ * the extensions string earlier in the boot process.
+ */
+DECLARE_STATIC_KEY_TRUE(qspinlock_key);
+
+#define SPINLOCK_BASE_DECLARE(op, type, type_lock)			\
+static __always_inline type arch_spin_##op(type_lock lock)		\
+{									\
+	if (static_branch_unlikely(&qspinlock_key))			\
+		return queued_spin_##op(lock);				\
+	return ticket_spin_##op(lock);					\
+}
+
+SPINLOCK_BASE_DECLARE(lock, void, arch_spinlock_t *)
+SPINLOCK_BASE_DECLARE(unlock, void, arch_spinlock_t *)
+SPINLOCK_BASE_DECLARE(is_locked, int, arch_spinlock_t *)
+SPINLOCK_BASE_DECLARE(is_contended, int, arch_spinlock_t *)
+SPINLOCK_BASE_DECLARE(trylock, bool, arch_spinlock_t *)
+SPINLOCK_BASE_DECLARE(value_unlocked, int, arch_spinlock_t)
+
+#elif defined(CONFIG_RISCV_QUEUED_SPINLOCKS)
+
+#include <asm/qspinlock.h>
+
+#else
+
+#include <asm/ticket_spinlock.h>
+
+#endif
+
+#include <asm/qrwlock.h>
+
+#endif /* __ASM_RISCV_SPINLOCK_H */

arch/riscv/kernel/setup.c

@@ -244,6 +244,42 @@ static void __init parse_dtb(void)
 #endif
 }
 
+#if defined(CONFIG_RISCV_COMBO_SPINLOCKS)
+DEFINE_STATIC_KEY_TRUE(qspinlock_key);
+EXPORT_SYMBOL(qspinlock_key);
+#endif
+
+static void __init riscv_spinlock_init(void)
+{
+	char *using_ext = NULL;
+
+	if (IS_ENABLED(CONFIG_RISCV_TICKET_SPINLOCKS)) {
+		pr_info("Ticket spinlock: enabled\n");
+		return;
+	}
+
+	if (IS_ENABLED(CONFIG_RISCV_ISA_ZABHA) &&
+	    IS_ENABLED(CONFIG_RISCV_ISA_ZACAS) &&
+	    riscv_isa_extension_available(NULL, ZABHA) &&
+	    riscv_isa_extension_available(NULL, ZACAS)) {
+		using_ext = "using Zabha";
+	} else if (riscv_isa_extension_available(NULL, ZICCRSE)) {
+		using_ext = "using Ziccrse";
+	}
+#if defined(CONFIG_RISCV_COMBO_SPINLOCKS)
+	else {
+		static_branch_disable(&qspinlock_key);
+		pr_info("Ticket spinlock: enabled\n");
+		return;
+	}
+#endif
+
+	if (!using_ext)
+		pr_err("Queued spinlock without Zabha or Ziccrse");
+	else
+		pr_info("Queued spinlock %s: enabled\n", using_ext);
+}
+
 extern void __init init_rt_signal_env(void);
 
 void __init setup_arch(char **cmdline_p)
@@ -297,6 +333,7 @@ void __init setup_arch(char **cmdline_p)
 	riscv_set_dma_cache_alignment();
 
 	riscv_user_isa_enable();
+	riscv_spinlock_init();
 }
 
 bool arch_cpu_is_hotpluggable(int cpu)

include/asm-generic/qspinlock.h

@@ -136,6 +136,7 @@ static __always_inline bool virt_spin_lock(struct qspinlock *lock)
 }
 #endif
 
+#ifndef __no_arch_spinlock_redefine
 /*
  * Remapping spinlock architecture specific functions to the corresponding
  * queued spinlock functions.
@@ -146,5 +147,6 @@ static __always_inline bool virt_spin_lock(struct qspinlock *lock)
 #define arch_spin_lock(l)		queued_spin_lock(l)
 #define arch_spin_trylock(l)		queued_spin_trylock(l)
 #define arch_spin_unlock(l)		queued_spin_unlock(l)
+#endif
 
 #endif /* __ASM_GENERIC_QSPINLOCK_H */

include/asm-generic/ticket_spinlock.h

@@ -89,6 +89,7 @@ static __always_inline int ticket_spin_is_contended(arch_spinlock_t *lock)
 	return (s16)((val >> 16) - (val & 0xffff)) > 1;
 }
 
+#ifndef __no_arch_spinlock_redefine
 /*
  * Remapping spinlock architecture specific functions to the corresponding
  * ticket spinlock functions.
@@ -99,5 +100,6 @@ static __always_inline int ticket_spin_is_contended(arch_spinlock_t *lock)
 #define arch_spin_lock(l)		ticket_spin_lock(l)
 #define arch_spin_trylock(l)		ticket_spin_trylock(l)
 #define arch_spin_unlock(l)		ticket_spin_unlock(l)
+#endif
 
 #endif /* __ASM_GENERIC_TICKET_SPINLOCK_H */