[PATCH v2 0/3] Ensure inX() is ordered wrt delay() routines

[PATCH v2 0/3] Ensure inX() is ordered wrt delay() routines

[Will Deacon]

Hi all,

This is version two of the patches I previously posted here:

  https://lore.kernel.org/lkml/20190211174544.4302-1-will.deacon@arm.com/T/#u

Changes since v2 include:

  * Incorporate riscv changes from Palmer
  * Update macro definitions as suggested by Geert
  * Extend to cover non-port reads via __io_ar() as well

Feedback welcome,

Will

--->8

Will Deacon (3):
  asm-generic/io: Pass result of I/O accessor to __io_[p]ar()
  riscv: io: Update __io_[p]ar() macros to take an argument
  arm64: io: Hook up __io_par() for inX() ordering

 arch/arm64/include/asm/io.h |  1 +
 arch/riscv/include/asm/io.h | 36 ++++++++++++++++++------------------
 include/asm-generic/io.h    | 20 ++++++++++----------
 3 files changed, 29 insertions(+), 28 deletions(-)

-- 
2.11.0

[PATCH v2 1/3] asm-generic/io: Pass result of I/O accessor to __io_[p]ar()

[Will Deacon]

The inX() and readX() I/O accessors must enforce ordering against
subsequent calls to the delay() routines, so that a read-back from a
device can be used to postpone a subsequent write to the same device.

On some architectures, including arm64, this ordering can only be
achieved by creating a dependency on the value returned by the I/O
accessor operation, so we need to pass the value we read to the
__io_par() and __io_ar() macros in these cases.

Acked-by: Arnd Bergmann <arnd@arndb.de>
Reported-by: Andrew Murray <andrew.murray@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
 include/asm-generic/io.h | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)

diff --git a/include/asm-generic/io.h b/include/asm-generic/io.h
index d356f802945a..303871651f8a 100644
--- a/include/asm-generic/io.h
+++ b/include/asm-generic/io.h
@@ -32,9 +32,9 @@
 /* prevent prefetching of coherent DMA data ahead of a dma-complete */
 #ifndef __io_ar
 #ifdef rmb
-#define __io_ar()      rmb()
+#define __io_ar(v)      rmb()
 #else
-#define __io_ar()      barrier()
+#define __io_ar(v)      barrier()
 #endif
 #endif
 
@@ -65,7 +65,7 @@
 #endif
 
 #ifndef __io_par
-#define __io_par()     __io_ar()
+#define __io_par(v)     __io_ar(v)
 #endif
 
 
@@ -158,7 +158,7 @@ static inline u8 readb(const volatile void __iomem *addr)
 
 	__io_br();
 	val = __raw_readb(addr);
-	__io_ar();
+	__io_ar(val);
 	return val;
 }
 #endif
@@ -171,7 +171,7 @@ static inline u16 readw(const volatile void __iomem *addr)
 
 	__io_br();
 	val = __le16_to_cpu(__raw_readw(addr));
-	__io_ar();
+	__io_ar(val);
 	return val;
 }
 #endif
@@ -184,7 +184,7 @@ static inline u32 readl(const volatile void __iomem *addr)
 
 	__io_br();
 	val = __le32_to_cpu(__raw_readl(addr));
-	__io_ar();
+	__io_ar(val);
 	return val;
 }
 #endif
@@ -198,7 +198,7 @@ static inline u64 readq(const volatile void __iomem *addr)
 
 	__io_br();
 	val = __le64_to_cpu(__raw_readq(addr));
-	__io_ar();
+	__io_ar(val);
 	return val;
 }
 #endif
@@ -471,7 +471,7 @@ static inline u8 inb(unsigned long addr)
 
 	__io_pbr();
 	val = __raw_readb(PCI_IOBASE + addr);
-	__io_par();
+	__io_par(val);
 	return val;
 }
 #endif
@@ -484,7 +484,7 @@ static inline u16 inw(unsigned long addr)
 
 	__io_pbr();
 	val = __le16_to_cpu(__raw_readw(PCI_IOBASE + addr));
-	__io_par();
+	__io_par(val);
 	return val;
 }
 #endif
@@ -497,7 +497,7 @@ static inline u32 inl(unsigned long addr)
 
 	__io_pbr();
 	val = __le32_to_cpu(__raw_readl(PCI_IOBASE + addr));
-	__io_par();
+	__io_par(val);
 	return val;
 }
 #endif
-- 
2.11.0

[PATCH v2 2/3] riscv: io: Update __io_[p]ar() macros to take an argument

[Will Deacon]

The definitions of the __io_[p]ar() macros in asm-generic/io.h take the
value returned by the preceding I/O read as an argument so that
architectures can use this to create order with a subsequent delayX()
routine using a dependency.

Update the riscv barrier definitions to match, although the argument
is currently unused.

Suggested-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Palmer Dabbelt <palmer@sifive.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
 arch/riscv/include/asm/io.h | 36 ++++++++++++++++++------------------
 1 file changed, 18 insertions(+), 18 deletions(-)

diff --git a/arch/riscv/include/asm/io.h b/arch/riscv/include/asm/io.h
index b269451e7e85..1d9c1376dc64 100644
--- a/arch/riscv/include/asm/io.h
+++ b/arch/riscv/include/asm/io.h
@@ -163,20 +163,20 @@ static inline u64 __raw_readq(const volatile void __iomem *addr)
  * doesn't define any ordering between the memory space and the I/O space.
  */
 #define __io_br()	do {} while (0)
-#define __io_ar()	__asm__ __volatile__ ("fence i,r" : : : "memory");
+#define __io_ar(v)	__asm__ __volatile__ ("fence i,r" : : : "memory");
 #define __io_bw()	__asm__ __volatile__ ("fence w,o" : : : "memory");
 #define __io_aw()	do {} while (0)
 
-#define readb(c)	({ u8  __v; __io_br(); __v = readb_cpu(c); __io_ar(); __v; })
-#define readw(c)	({ u16 __v; __io_br(); __v = readw_cpu(c); __io_ar(); __v; })
-#define readl(c)	({ u32 __v; __io_br(); __v = readl_cpu(c); __io_ar(); __v; })
+#define readb(c)	({ u8  __v; __io_br(); __v = readb_cpu(c); __io_ar(__v); __v; })
+#define readw(c)	({ u16 __v; __io_br(); __v = readw_cpu(c); __io_ar(__v); __v; })
+#define readl(c)	({ u32 __v; __io_br(); __v = readl_cpu(c); __io_ar(__v); __v; })
 
 #define writeb(v,c)	({ __io_bw(); writeb_cpu((v),(c)); __io_aw(); })
 #define writew(v,c)	({ __io_bw(); writew_cpu((v),(c)); __io_aw(); })
 #define writel(v,c)	({ __io_bw(); writel_cpu((v),(c)); __io_aw(); })
 
 #ifdef CONFIG_64BIT
-#define readq(c)	({ u64 __v; __io_br(); __v = readq_cpu(c); __io_ar(); __v; })
+#define readq(c)	({ u64 __v; __io_br(); __v = readq_cpu(c); __io_ar(__v); __v; })
 #define writeq(v,c)	({ __io_bw(); writeq_cpu((v),(c)); __io_aw(); })
 #endif
 
@@ -198,20 +198,20 @@ static inline u64 __raw_readq(const volatile void __iomem *addr)
  * writes.
  */
 #define __io_pbr()	__asm__ __volatile__ ("fence io,i"  : : : "memory");
-#define __io_par()	__asm__ __volatile__ ("fence i,ior" : : : "memory");
+#define __io_par(v)	__asm__ __volatile__ ("fence i,ior" : : : "memory");
 #define __io_pbw()	__asm__ __volatile__ ("fence iow,o" : : : "memory");
 #define __io_paw()	__asm__ __volatile__ ("fence o,io"  : : : "memory");
 
-#define inb(c)		({ u8  __v; __io_pbr(); __v = readb_cpu((void*)(PCI_IOBASE + (c))); __io_par(); __v; })
-#define inw(c)		({ u16 __v; __io_pbr(); __v = readw_cpu((void*)(PCI_IOBASE + (c))); __io_par(); __v; })
-#define inl(c)		({ u32 __v; __io_pbr(); __v = readl_cpu((void*)(PCI_IOBASE + (c))); __io_par(); __v; })
+#define inb(c)		({ u8  __v; __io_pbr(); __v = readb_cpu((void*)(PCI_IOBASE + (c))); __io_par(__v); __v; })
+#define inw(c)		({ u16 __v; __io_pbr(); __v = readw_cpu((void*)(PCI_IOBASE + (c))); __io_par(__v); __v; })
+#define inl(c)		({ u32 __v; __io_pbr(); __v = readl_cpu((void*)(PCI_IOBASE + (c))); __io_par(__v); __v; })
 
 #define outb(v,c)	({ __io_pbw(); writeb_cpu((v),(void*)(PCI_IOBASE + (c))); __io_paw(); })
 #define outw(v,c)	({ __io_pbw(); writew_cpu((v),(void*)(PCI_IOBASE + (c))); __io_paw(); })
 #define outl(v,c)	({ __io_pbw(); writel_cpu((v),(void*)(PCI_IOBASE + (c))); __io_paw(); })
 
 #ifdef CONFIG_64BIT
-#define inq(c)		({ u64 __v; __io_pbr(); __v = readq_cpu((void*)(c)); __io_par(); __v; })
+#define inq(c)		({ u64 __v; __io_pbr(); __v = readq_cpu((void*)(c)); __io_par(__v); __v; })
 #define outq(v,c)	({ __io_pbw(); writeq_cpu((v),(void*)(c)); __io_paw(); })
 #endif
 
@@ -254,16 +254,16 @@ static inline u64 __raw_readq(const volatile void __iomem *addr)
 		afence;								\
 	}
 
-__io_reads_ins(reads,  u8, b, __io_br(), __io_ar())
-__io_reads_ins(reads, u16, w, __io_br(), __io_ar())
-__io_reads_ins(reads, u32, l, __io_br(), __io_ar())
+__io_reads_ins(reads,  u8, b, __io_br(), __io_ar(addr))
+__io_reads_ins(reads, u16, w, __io_br(), __io_ar(addr))
+__io_reads_ins(reads, u32, l, __io_br(), __io_ar(addr))
 #define readsb(addr, buffer, count) __readsb(addr, buffer, count)
 #define readsw(addr, buffer, count) __readsw(addr, buffer, count)
 #define readsl(addr, buffer, count) __readsl(addr, buffer, count)
 
-__io_reads_ins(ins,  u8, b, __io_pbr(), __io_par())
-__io_reads_ins(ins, u16, w, __io_pbr(), __io_par())
-__io_reads_ins(ins, u32, l, __io_pbr(), __io_par())
+__io_reads_ins(ins,  u8, b, __io_pbr(), __io_par(addr))
+__io_reads_ins(ins, u16, w, __io_pbr(), __io_par(addr))
+__io_reads_ins(ins, u32, l, __io_pbr(), __io_par(addr))
 #define insb(addr, buffer, count) __insb((void __iomem *)(long)addr, buffer, count)
 #define insw(addr, buffer, count) __insw((void __iomem *)(long)addr, buffer, count)
 #define insl(addr, buffer, count) __insl((void __iomem *)(long)addr, buffer, count)
@@ -283,10 +283,10 @@ __io_writes_outs(outs, u32, l, __io_pbw(), __io_paw())
 #define outsl(addr, buffer, count) __outsl((void __iomem *)(long)addr, buffer, count)
 
 #ifdef CONFIG_64BIT
-__io_reads_ins(reads, u64, q, __io_br(), __io_ar())
+__io_reads_ins(reads, u64, q, __io_br(), __io_ar(addr))
 #define readsq(addr, buffer, count) __readsq(addr, buffer, count)
 
-__io_reads_ins(ins, u64, q, __io_pbr(), __io_par())
+__io_reads_ins(ins, u64, q, __io_pbr(), __io_par(addr))
 #define insq(addr, buffer, count) __insq((void __iomem *)addr, buffer, count)
 
 __io_writes_outs(writes, u64, q, __io_bw(), __io_aw())
-- 
2.11.0

[PATCH v2 3/3] arm64: io: Hook up __io_par() for inX() ordering

[Will Deacon]

Ensure that inX() provides the same ordering guarantees as readX()
by hooking up __io_par() so that it maps directly to __iormb().

Reported-by: Andrew Murray <andrew.murray@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
 arch/arm64/include/asm/io.h | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/arm64/include/asm/io.h b/arch/arm64/include/asm/io.h
index ee723835c1f4..8bb7210ac286 100644
--- a/arch/arm64/include/asm/io.h
+++ b/arch/arm64/include/asm/io.h
@@ -121,6 +121,7 @@ static inline u64 __raw_readq(const volatile void __iomem *addr)
 		     : "memory");					\
 })
 
+#define __io_par(v)		__iormb(v)
 #define __iowmb()		wmb()
 
 #define mmiowb()		do { } while (0)
-- 
2.11.0

Re: [PATCH v2 0/3] Ensure inX() is ordered wrt delay() routines

[Palmer Dabbelt]

On Fri, 22 Feb 2019 10:04:51 PST (-0800), Will Deacon wrote:
> Hi all,
>
> This is version two of the patches I previously posted here:
>
>   https://lore.kernel.org/lkml/20190211174544.4302-1-will.deacon@arm.com/T/#u
>
> Changes since v2 include:
>
>   * Incorporate riscv changes from Palmer
>   * Update macro definitions as suggested by Geert
>   * Extend to cover non-port reads via __io_ar() as well
>
> Feedback welcome,
>
> Will
>
> --->8
>
> Will Deacon (3):
>   asm-generic/io: Pass result of I/O accessor to __io_[p]ar()
>   riscv: io: Update __io_[p]ar() macros to take an argument
>   arm64: io: Hook up __io_par() for inX() ordering
>
>  arch/arm64/include/asm/io.h |  1 +
>  arch/riscv/include/asm/io.h | 36 ++++++++++++++++++------------------
>  include/asm-generic/io.h    | 20 ++++++++++----------
>  3 files changed, 29 insertions(+), 28 deletions(-)

Reviewed-by: Palmer Dabbelt <palmer@sifive.com>

Thanks!

Re: [PATCH v2 0/3] Ensure inX() is ordered wrt delay() routines

[Catalin Marinas]

On Fri, Feb 22, 2019 at 06:04:51PM +0000, Will Deacon wrote:
> Will Deacon (3):
>   asm-generic/io: Pass result of I/O accessor to __io_[p]ar()
>   riscv: io: Update __io_[p]ar() macros to take an argument
>   arm64: io: Hook up __io_par() for inX() ordering

Since we have the acks in place, I plan to queue these patches via the
arm64 tree. They don't conflict with -next (as of today) but if anyone
expects other issues, please let me know.

-- 
Catalin