Merge tag 'riscv-for-linus-4.15-arch-v9-premerge' of git://git.kernel.org/pub/scm/linux/kernel/git/palmer/linux

Pull RISC-V architecture support from Palmer Dabbelt:
 "This contains the core RISC-V Linux port, which has been through nine
  rounds of review on various mailing lists. The port is not complete:
  there's some cleanup patches moving through the review process, a
  whole bunch of drivers that need some work, and a lot of feature
  additions that will be needed.

  The patches contained in this tag have been through nine rounds of
  review on the various mailing lists. I have some outstanding cleanup
  patches, but since there's been so much review on these patches I
  thought it would be best to submit them as-is and then submit explicit
  cleanup patches so everyone can review them. This first patch set is
  big enough that it's a bit of a pain to constantly rewrite, and it's
  caused a few headaches with various contributors.

  The port is definately a work in progress. While what's there builds
  and boots with 4.14, it's a bit hard to actually see anything happen
  because there are no device drivers yet. I maintain a staging branch
  that contains all the device drivers and cleanup that actually works,
  but those patches won't all be ready for a while. I'd like to get what
  we currently have into your tree so everyone can start working from a
  single base -- of particular importance is allowing the glibc
  upstreaming process to proceed so we can sort out any possibly
  lingering user-visible ABI problems we might have.

  Copied below is the ChangeLog that contains the history of this patch
  set:

   (v9) As per suggestions on our v8 patch set, I've split the core
        architecture code out from our drivers and would like to submit
        this patch set to be included into linux-next, with the goal
        being to be merged in during the next merge window. This patch
        set is based on 4.14-rc2, but if it's better to have it based on
        something else then I can change it around.

        This patch set contains just the core arch code for RISC-V, so
        while it builds an nominally boots, you can't print or take an
        interrupt so it's not that useful. If you're looking to actually
        boot a system it would probably be better to use the full patch
        set listed below.

        We've collected a handful of tags from reviewers, and the
        remainder of the patch set only got minimal feedback last time.
        Here's what changed:

         - We now use the device tree to initialize the timer driver so
           it's less tighly coupled with the arch port.

         - I cleaned up the defconfigs -- there's actually now just one,
           and it's empty. For now I think we're OK with what the kernel
           sets as defaults, but I anticipate we'll begin to expand this
           as people start to use the port more.

         - The VDSO symbols version is sane.

         - We WFI while spinning in the boot loop.

         - A handful of comments have been added.

        While there are still a handful of FIXMEs in this patch set,
        we've started to get enough interest from various users and
        contributors that maintaining an out of tree patch set is
        starting to become a big burden. Hopefully the patches are good
        enough to merge now, which will at least get everyone working in
        a more reasonable manner as we clean up the remaining issues.

   (v8) I know it may not be the ideal time to submit a patch set right
        now, as it's the middle of the merge window, but things have
        calmed down quite a bit in the last month so I thought it would
        be good to get everyone on the same page. There's been a handful
        of changes since the last patch set, but most of them are fairly
        minor:

         - We changed PAGE_OFFSET to allowing mapping more physical
           memory on 64-bit systems. This is user configurable, as it
           triggers a different code model that generates slightly less
           efficient code.

         - The device tree binding documentation is back, I'd managed to
           lose it at some point.

         - We now pass the atomic64 test suite

         - The SBI timer driver has been refactored.

   (v7) It's been a while since my last patch set, but the changes han
        been fairly minimal:

         - The PCI cleanup patches have been dropped, we'll do them as a
           separate patch set later.

         - We've the Kconfig entries from CONFIG_ISA_* to
           CONFIG_RISCV_ISA_*, to make grep easier.

         - There have been a handful of memory model related tweaks in
           I/O land, particularly relating the PCI and the upcoming
           platform specification. There are significant comments in the
           relevant files. This is still a WIP, but I think we're close
           to getting as good as we're going to get until we end up with
           some more specifications.

   (v6) As it's been only a day since the v5 patch set, the changes are
        pretty minimal:

         - The patch set is now based on linux-next/master, which I
           believe is a better base now that we're getting closer to
           upstream.

         - EARLY_PRINTK is no longer an option. Since the SBI console is
           reasonable, there's no penalty to enabling it (and thus no
           benefit to disabling it).

         - The mmap syscalls were refactored a bit.

   (v5) Things have really started to calm down, so this is fairly
        similar to the v4 patch set. The most interesting changes
        include:

         - We've moved back to a single patch set.

         - SMP support has been fixed, I was accidentally running on a
           non-SMP configuration. There were various mistakes all over
           the tree as a result of this.

         - The cmpxchg syscalls have been removed, as they were deemed a
           bad idea. As a result, RISC-V Linux systems mandate the A
           extension. The corresponding Kconfig entry to enable builds
           on non-A systems has been removed.

         - A few more atomic fixes: mostly fence changes, but those
           resulted in a handful of additional macros that were no
           longer necessary.

         - riscv_early_sie has been removed.

   (v4) There have only been a few changes since the v3 patch set:

         - The cmpxchg64 syscall is no longer enabled on 32-bit systems.
           It's not possible to provide this on SMP systems, and it's
           not necessary as glibc knows not to call it.

         - We provide a ELF_HWCAP so users can determine the ISA of the
           machine the kernel is running on.

         - The multi-line comments are in a better form.

         - There were a handful of headers that could be replaced with
           the asm-generic versions, and a few unnecessary definitions.

         - We no longer use printk, but instead use pr_*.

         - A few Kconfig and defconfig entries have been cleaned up.

   (v3) A highlight of the changes since the v2 patch set includes:

         - We've split out all our drivers into separate patch sets,
           which I've already sent out to the relevant maintainers. I
           haven't included those patches in this patch set, but some of
           them are necessary to build our port.

         - The patch set is now split up differently: rather than being
           split per directory it is split per topic. Hopefully this
           will make it easier to review the port on the mailing list.
           The split is a bit rough, so you probably still want to look
           at the patch set as a whole.

         - atomic.h has been completely rewritten and is hopefully now
           correct. I've attempted to sanitize the various other memory
           model related code as well, and I think it should all be sane
           now aside from a handful of FIXMEs commented in the code.

         - We've changed the cmpexchg syscall to always exist and to not
           be multiplexed. There is also a VDSO entry for compare and
           exchange, which allows kernels with the A extension to
           execute user code without the A extension reasonably fast.

         - Our user-visible register state now contains enough space for
           the Q extension for 128-bit floating point, as well as a few
           words to allow extensibility to future ISA extensions like
           the eventual V extension for vectors.

         - A handful of driver cleanups, but these have been split into
           separate patch sets now so I won't duplicate them here.

   (v2) A highlight of the changes since the v1 patch set includes:

         - We've split out our drivers into the right places, which
           means now there's a lot more patches. I'll be submitting
           these patches to various subsystem maintainers and including
           them in any future RISC-V patch sets until they've been
           merged.

         - The SBI console driver has been completely rewritten to use
           the HVC helpers and is now significantly smaller.

         - We've begun to use weaker barriers as opposed to just the big
           "fence". There's still some work to do here, specifically:
            - We need fences in the relaxed MMIO functions.
            - The non-relaxed MMIO functions are missing R/W bits on their fences.
            - Many AMOs need the aq and rl bits set.

         - We now have thread_info in task_struct. As a result, sscratch
           now contains TP instead of SP. This was necessary because
           thread_info is no longer on the stack.

         - A few shared routines have been added that we use instead of
           creating another arch copy"

Reviewed-by: Arnd Bergmann <arnd@arndb.de>

* tag 'riscv-for-linus-4.15-arch-v9-premerge' of git://git.kernel.org/pub/scm/linux/kernel/git/palmer/linux:
  RISC-V: Build Infrastructure
  RISC-V: User-facing API
  RISC-V: Paging and MMU
  RISC-V: Device, timer, IRQs, and the SBI
  RISC-V: Task implementation
  RISC-V: ELF and module implementation
  RISC-V: Generic library routines and assembly
  RISC-V: Atomic and Locking Code
  RISC-V: Init and Halt Code
  dt-bindings: RISC-V CPU Bindings
  lib: Add shared copies of some GCC library routines
  MAINTAINERS: Add RISC-V

1 files changed, 375 insertions, 0 deletions

diff --git a/arch/riscv/include/asm/atomic.h b/arch/riscv/include/asm/atomic.h
new file mode 100644
index 000000000000..e2e37c57cbeb
--- /dev/null
+++ b/arch/riscv/include/asm/atomic.h
@@ -0,0 +1,375 @@
+/*
+ * Copyright (C) 2007 Red Hat, Inc. All Rights Reserved.
+ * Copyright (C) 2012 Regents of the University of California
+ * Copyright (C) 2017 SiFive
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public Licence
+ * as published by the Free Software Foundation; either version
+ * 2 of the Licence, or (at your option) any later version.
+ */
+
+#ifndef _ASM_RISCV_ATOMIC_H
+#define _ASM_RISCV_ATOMIC_H
+
+#ifdef CONFIG_GENERIC_ATOMIC64
+# include <asm-generic/atomic64.h>
+#else
+# if (__riscv_xlen < 64)
+#  error "64-bit atomics require XLEN to be at least 64"
+# endif
+#endif
+
+#include <asm/cmpxchg.h>
+#include <asm/barrier.h>
+
+#define ATOMIC_INIT(i)	{ (i) }
+static __always_inline int atomic_read(const atomic_t *v)
+{
+	return READ_ONCE(v->counter);
+}
+static __always_inline void atomic_set(atomic_t *v, int i)
+{
+	WRITE_ONCE(v->counter, i);
+}
+
+#ifndef CONFIG_GENERIC_ATOMIC64
+#define ATOMIC64_INIT(i) { (i) }
+static __always_inline long atomic64_read(const atomic64_t *v)
+{
+	return READ_ONCE(v->counter);
+}
+static __always_inline void atomic64_set(atomic64_t *v, long i)
+{
+	WRITE_ONCE(v->counter, i);
+}
+#endif
+
+/*
+ * First, the atomic ops that have no ordering constraints and therefor don't
+ * have the AQ or RL bits set.  These don't return anything, so there's only
+ * one version to worry about.
+ */
+#define ATOMIC_OP(op, asm_op, c_op, I, asm_type, c_type, prefix)				\
+static __always_inline void atomic##prefix##_##op(c_type i, atomic##prefix##_t *v)		\
+{												\
+	__asm__ __volatile__ (									\
+		"amo" #asm_op "." #asm_type " zero, %1, %0"					\
+		: "+A" (v->counter)								\
+		: "r" (I)									\
+		: "memory");									\
+}
+
+#ifdef CONFIG_GENERIC_ATOMIC64
+#define ATOMIC_OPS(op, asm_op, c_op, I)			\
+        ATOMIC_OP (op, asm_op, c_op, I, w,  int,   )
+#else
+#define ATOMIC_OPS(op, asm_op, c_op, I)			\
+        ATOMIC_OP (op, asm_op, c_op, I, w,  int,   )	\
+        ATOMIC_OP (op, asm_op, c_op, I, d, long, 64)
+#endif
+
+ATOMIC_OPS(add, add, +,  i)
+ATOMIC_OPS(sub, add, +, -i)
+ATOMIC_OPS(and, and, &,  i)
+ATOMIC_OPS( or,  or, |,  i)
+ATOMIC_OPS(xor, xor, ^,  i)
+
+#undef ATOMIC_OP
+#undef ATOMIC_OPS
+
+/*
+ * Atomic ops that have ordered, relaxed, acquire, and relese variants.
+ * There's two flavors of these: the arithmatic ops have both fetch and return
+ * versions, while the logical ops only have fetch versions.
+ */
+#define ATOMIC_FETCH_OP(op, asm_op, c_op, I, asm_or, c_or, asm_type, c_type, prefix)			\
+static __always_inline c_type atomic##prefix##_fetch_##op##c_or(c_type i, atomic##prefix##_t *v)	\
+{													\
+	register c_type ret;										\
+	__asm__ __volatile__ (										\
+		"amo" #asm_op "." #asm_type #asm_or " %1, %2, %0"					\
+		: "+A" (v->counter), "=r" (ret)								\
+		: "r" (I)										\
+		: "memory");										\
+	return ret;											\
+}
+
+#define ATOMIC_OP_RETURN(op, asm_op, c_op, I, asm_or, c_or, asm_type, c_type, prefix)			\
+static __always_inline c_type atomic##prefix##_##op##_return##c_or(c_type i, atomic##prefix##_t *v)	\
+{													\
+        return atomic##prefix##_fetch_##op##c_or(i, v) c_op I;						\
+}
+
+#ifdef CONFIG_GENERIC_ATOMIC64
+#define ATOMIC_OPS(op, asm_op, c_op, I, asm_or, c_or)				\
+        ATOMIC_FETCH_OP (op, asm_op, c_op, I, asm_or, c_or, w,  int,   )	\
+        ATOMIC_OP_RETURN(op, asm_op, c_op, I, asm_or, c_or, w,  int,   )
+#else
+#define ATOMIC_OPS(op, asm_op, c_op, I, asm_or, c_or)				\
+        ATOMIC_FETCH_OP (op, asm_op, c_op, I, asm_or, c_or, w,  int,   )	\
+        ATOMIC_OP_RETURN(op, asm_op, c_op, I, asm_or, c_or, w,  int,   )	\
+        ATOMIC_FETCH_OP (op, asm_op, c_op, I, asm_or, c_or, d, long, 64)	\
+        ATOMIC_OP_RETURN(op, asm_op, c_op, I, asm_or, c_or, d, long, 64)
+#endif
+
+ATOMIC_OPS(add, add, +,  i,      , _relaxed)
+ATOMIC_OPS(add, add, +,  i, .aq  , _acquire)
+ATOMIC_OPS(add, add, +,  i, .rl  , _release)
+ATOMIC_OPS(add, add, +,  i, .aqrl,         )
+
+ATOMIC_OPS(sub, add, +, -i,      , _relaxed)
+ATOMIC_OPS(sub, add, +, -i, .aq  , _acquire)
+ATOMIC_OPS(sub, add, +, -i, .rl  , _release)
+ATOMIC_OPS(sub, add, +, -i, .aqrl,         )
+
+#undef ATOMIC_OPS
+
+#ifdef CONFIG_GENERIC_ATOMIC64
+#define ATOMIC_OPS(op, asm_op, c_op, I, asm_or, c_or)				\
+        ATOMIC_FETCH_OP(op, asm_op, c_op, I, asm_or, c_or, w,  int,   )
+#else
+#define ATOMIC_OPS(op, asm_op, c_op, I, asm_or, c_or)				\
+        ATOMIC_FETCH_OP(op, asm_op, c_op, I, asm_or, c_or, w,  int,   )		\
+        ATOMIC_FETCH_OP(op, asm_op, c_op, I, asm_or, c_or, d, long, 64)
+#endif
+
+ATOMIC_OPS(and, and, &,  i,      , _relaxed)
+ATOMIC_OPS(and, and, &,  i, .aq  , _acquire)
+ATOMIC_OPS(and, and, &,  i, .rl  , _release)
+ATOMIC_OPS(and, and, &,  i, .aqrl,         )
+
+ATOMIC_OPS( or,  or, |,  i,      , _relaxed)
+ATOMIC_OPS( or,  or, |,  i, .aq  , _acquire)
+ATOMIC_OPS( or,  or, |,  i, .rl  , _release)
+ATOMIC_OPS( or,  or, |,  i, .aqrl,         )
+
+ATOMIC_OPS(xor, xor, ^,  i,      , _relaxed)
+ATOMIC_OPS(xor, xor, ^,  i, .aq  , _acquire)
+ATOMIC_OPS(xor, xor, ^,  i, .rl  , _release)
+ATOMIC_OPS(xor, xor, ^,  i, .aqrl,         )
+
+#undef ATOMIC_OPS
+
+#undef ATOMIC_FETCH_OP
+#undef ATOMIC_OP_RETURN
+
+/*
+ * The extra atomic operations that are constructed from one of the core
+ * AMO-based operations above (aside from sub, which is easier to fit above).
+ * These are required to perform a barrier, but they're OK this way because
+ * atomic_*_return is also required to perform a barrier.
+ */
+#define ATOMIC_OP(op, func_op, comp_op, I, c_type, prefix)			\
+static __always_inline bool atomic##prefix##_##op(c_type i, atomic##prefix##_t *v) \
+{										\
+	return atomic##prefix##_##func_op##_return(i, v) comp_op I;		\
+}
+
+#ifdef CONFIG_GENERIC_ATOMIC64
+#define ATOMIC_OPS(op, func_op, comp_op, I)			\
+        ATOMIC_OP (op, func_op, comp_op, I,  int,   )
+#else
+#define ATOMIC_OPS(op, func_op, comp_op, I)			\
+        ATOMIC_OP (op, func_op, comp_op, I,  int,   )		\
+        ATOMIC_OP (op, func_op, comp_op, I, long, 64)
+#endif
+
+ATOMIC_OPS(add_and_test, add, ==, 0)
+ATOMIC_OPS(sub_and_test, sub, ==, 0)
+ATOMIC_OPS(add_negative, add,  <, 0)
+
+#undef ATOMIC_OP
+#undef ATOMIC_OPS
+
+#define ATOMIC_OP(op, func_op, c_op, I, c_type, prefix)				\
+static __always_inline void atomic##prefix##_##op(atomic##prefix##_t *v)	\
+{										\
+	atomic##prefix##_##func_op(I, v);					\
+}
+
+#define ATOMIC_FETCH_OP(op, func_op, c_op, I, c_type, prefix)				\
+static __always_inline c_type atomic##prefix##_fetch_##op(atomic##prefix##_t *v)	\
+{											\
+	return atomic##prefix##_fetch_##func_op(I, v);					\
+}
+
+#define ATOMIC_OP_RETURN(op, asm_op, c_op, I, c_type, prefix)				\
+static __always_inline c_type atomic##prefix##_##op##_return(atomic##prefix##_t *v)	\
+{											\
+        return atomic##prefix##_fetch_##op(v) c_op I;					\
+}
+
+#ifdef CONFIG_GENERIC_ATOMIC64
+#define ATOMIC_OPS(op, asm_op, c_op, I)						\
+        ATOMIC_OP       (op, asm_op, c_op, I,  int,   )				\
+        ATOMIC_FETCH_OP (op, asm_op, c_op, I,  int,   )				\
+        ATOMIC_OP_RETURN(op, asm_op, c_op, I,  int,   )
+#else
+#define ATOMIC_OPS(op, asm_op, c_op, I)						\
+        ATOMIC_OP       (op, asm_op, c_op, I,  int,   )				\
+        ATOMIC_FETCH_OP (op, asm_op, c_op, I,  int,   )				\
+        ATOMIC_OP_RETURN(op, asm_op, c_op, I,  int,   )				\
+        ATOMIC_OP       (op, asm_op, c_op, I, long, 64)				\
+        ATOMIC_FETCH_OP (op, asm_op, c_op, I, long, 64)				\
+        ATOMIC_OP_RETURN(op, asm_op, c_op, I, long, 64)
+#endif
+
+ATOMIC_OPS(inc, add, +,  1)
+ATOMIC_OPS(dec, add, +, -1)
+
+#undef ATOMIC_OPS
+#undef ATOMIC_OP
+#undef ATOMIC_FETCH_OP
+#undef ATOMIC_OP_RETURN
+
+#define ATOMIC_OP(op, func_op, comp_op, I, prefix)				\
+static __always_inline bool atomic##prefix##_##op(atomic##prefix##_t *v)	\
+{										\
+	return atomic##prefix##_##func_op##_return(v) comp_op I;		\
+}
+
+ATOMIC_OP(inc_and_test, inc, ==, 0,   )
+ATOMIC_OP(dec_and_test, dec, ==, 0,   )
+#ifndef CONFIG_GENERIC_ATOMIC64
+ATOMIC_OP(inc_and_test, inc, ==, 0, 64)
+ATOMIC_OP(dec_and_test, dec, ==, 0, 64)
+#endif
+
+#undef ATOMIC_OP
+
+/* This is required to provide a barrier on success. */
+static __always_inline int __atomic_add_unless(atomic_t *v, int a, int u)
+{
+       int prev, rc;
+
+	__asm__ __volatile__ (
+		"0:\n\t"
+		"lr.w.aqrl  %[p],  %[c]\n\t"
+		"beq        %[p],  %[u], 1f\n\t"
+		"add       %[rc],  %[p], %[a]\n\t"
+		"sc.w.aqrl %[rc], %[rc], %[c]\n\t"
+		"bnez      %[rc], 0b\n\t"
+		"1:"
+		: [p]"=&r" (prev), [rc]"=&r" (rc), [c]"+A" (v->counter)
+		: [a]"r" (a), [u]"r" (u)
+		: "memory");
+	return prev;
+}
+
+#ifndef CONFIG_GENERIC_ATOMIC64
+static __always_inline long __atomic64_add_unless(atomic64_t *v, long a, long u)
+{
+       long prev, rc;
+
+	__asm__ __volatile__ (
+		"0:\n\t"
+		"lr.d.aqrl  %[p],  %[c]\n\t"
+		"beq        %[p],  %[u], 1f\n\t"
+		"add       %[rc],  %[p], %[a]\n\t"
+		"sc.d.aqrl %[rc], %[rc], %[c]\n\t"
+		"bnez      %[rc], 0b\n\t"
+		"1:"
+		: [p]"=&r" (prev), [rc]"=&r" (rc), [c]"+A" (v->counter)
+		: [a]"r" (a), [u]"r" (u)
+		: "memory");
+	return prev;
+}
+
+static __always_inline int atomic64_add_unless(atomic64_t *v, long a, long u)
+{
+	return __atomic64_add_unless(v, a, u) != u;
+}
+#endif
+
+/*
+ * The extra atomic operations that are constructed from one of the core
+ * LR/SC-based operations above.
+ */
+static __always_inline int atomic_inc_not_zero(atomic_t *v)
+{
+        return __atomic_add_unless(v, 1, 0);
+}
+
+#ifndef CONFIG_GENERIC_ATOMIC64
+static __always_inline long atomic64_inc_not_zero(atomic64_t *v)
+{
+        return atomic64_add_unless(v, 1, 0);
+}
+#endif
+
+/*
+ * atomic_{cmp,}xchg is required to have exactly the same ordering semantics as
+ * {cmp,}xchg and the operations that return, so they need a barrier.  We just
+ * use the other implementations directly.
+ */
+#define ATOMIC_OP(c_t, prefix, c_or, size, asm_or)						\
+static __always_inline c_t atomic##prefix##_cmpxchg##c_or(atomic##prefix##_t *v, c_t o, c_t n) 	\
+{												\
+	return __cmpxchg(&(v->counter), o, n, size, asm_or, asm_or);				\
+}												\
+static __always_inline c_t atomic##prefix##_xchg##c_or(atomic##prefix##_t *v, c_t n) 		\
+{												\
+	return __xchg(n, &(v->counter), size, asm_or);						\
+}
+
+#ifdef CONFIG_GENERIC_ATOMIC64
+#define ATOMIC_OPS(c_or, asm_or)			\
+	ATOMIC_OP( int,   , c_or, 4, asm_or)
+#else
+#define ATOMIC_OPS(c_or, asm_or)			\
+	ATOMIC_OP( int,   , c_or, 4, asm_or)		\
+	ATOMIC_OP(long, 64, c_or, 8, asm_or)
+#endif
+
+ATOMIC_OPS(        , .aqrl)
+ATOMIC_OPS(_acquire,   .aq)
+ATOMIC_OPS(_release,   .rl)
+ATOMIC_OPS(_relaxed,      )
+
+#undef ATOMIC_OPS
+#undef ATOMIC_OP
+
+static __always_inline int atomic_sub_if_positive(atomic_t *v, int offset)
+{
+       int prev, rc;
+
+	__asm__ __volatile__ (
+		"0:\n\t"
+		"lr.w.aqrl  %[p],  %[c]\n\t"
+		"sub       %[rc],  %[p], %[o]\n\t"
+		"bltz      %[rc],    1f\n\t"
+		"sc.w.aqrl %[rc], %[rc], %[c]\n\t"
+		"bnez      %[rc],    0b\n\t"
+		"1:"
+		: [p]"=&r" (prev), [rc]"=&r" (rc), [c]"+A" (v->counter)
+		: [o]"r" (offset)
+		: "memory");
+	return prev - offset;
+}
+
+#define atomic_dec_if_positive(v)	atomic_sub_if_positive(v, 1)
+
+#ifndef CONFIG_GENERIC_ATOMIC64
+static __always_inline long atomic64_sub_if_positive(atomic64_t *v, int offset)
+{
+       long prev, rc;
+
+	__asm__ __volatile__ (
+		"0:\n\t"
+		"lr.d.aqrl  %[p],  %[c]\n\t"
+		"sub       %[rc],  %[p], %[o]\n\t"
+		"bltz      %[rc],    1f\n\t"
+		"sc.d.aqrl %[rc], %[rc], %[c]\n\t"
+		"bnez      %[rc],    0b\n\t"
+		"1:"
+		: [p]"=&r" (prev), [rc]"=&r" (rc), [c]"+A" (v->counter)
+		: [o]"r" (offset)
+		: "memory");
+	return prev - offset;
+}
+
+#define atomic64_dec_if_positive(v)	atomic64_sub_if_positive(v, 1)
+#endif
+
+#endif /* _ASM_RISCV_ATOMIC_H */