(no subject)

[Yury Norov]

Subject: [PATCH v2 0/5] lib: rework bitreverse

Cleanup bitreverse-related Kconfig dependency mechanism, and add new
GENERIC_BITREVERSE option to allow architectures to pick the generic
implementation as a fallback for the arch one.

Make RISCV to pick the generic implementation with the absence of ZBKB.

v1: https://lore.kernel.org/all/20260430211351.658193-1-ynorov@nvidia.com/
v2:
 - don't protect headers with the corresponding configs (Arnd, Erik);
 - make HAVE_ARCH_BITREVERSE conditional on BITREVERSE;
 - make GENERIC_BITREVERSE tri-state (sashiko);
 - re-implement GENERIC_BITREVERSE and it's relation to BITREVERSE and
   HAVE_ARCH_BITREVERSE, thus taking over the authorship;
 - RISCV: select GENERIC_BITREVERSE and HAVE_ARCH_BITREVERSE depending
   on BITREVERSE;

Jinjie Ruan (2):
  bitops: Define generic___bitrev8/16/32 for reuse
  arch/riscv: Add bitrev.h file to support rev8 and brev8

Yury Norov (3):
  arch: select HAVE_ARCH_BITREVERSE conditionally on BITREVERSE
  lib/bitrev: Introduce GENERIC_BITREVERSE
  MAINTAINERS: BITOPS: include bitrev.[ch]

 MAINTAINERS                           |  2 ++
 arch/arm/Kconfig                      |  2 +-
 arch/arm64/Kconfig                    |  2 +-
 arch/loongarch/Kconfig                |  2 +-
 arch/mips/Kconfig                     |  2 +-
 arch/riscv/Kconfig                    |  2 ++
 arch/riscv/include/asm/bitrev.h       | 51 +++++++++++++++++++++++++++
 include/asm-generic/bitops/__bitrev.h | 23 ++++++++++++
 include/linux/bitrev.h                | 20 +++--------
 lib/Kconfig                           | 13 +++++++
 lib/Makefile                          |  2 +-
 lib/bitrev.c                          |  3 --
 12 files changed, 100 insertions(+), 24 deletions(-)
 create mode 100644 arch/riscv/include/asm/bitrev.h
 create mode 100644 include/asm-generic/bitops/__bitrev.h

-- 
2.51.0

[PATCH v2 1/5] arch: select HAVE_ARCH_BITREVERSE conditionally on BITREVERSE

[Yury Norov]

Architectures may have bit reversal instructions, but if the API not
needed, the corresponding option should not be selected because it may
lead to generating the unneeded code.

Signed-off-by: Yury Norov <ynorov@nvidia.com>
---
 arch/arm/Kconfig       | 2 +-
 arch/arm64/Kconfig     | 2 +-
 arch/loongarch/Kconfig | 2 +-
 arch/mips/Kconfig      | 2 +-
 lib/Kconfig            | 1 +
 5 files changed, 5 insertions(+), 4 deletions(-)

diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index 71fc5dd4123f..0e963e54fe06 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -83,7 +83,7 @@ config ARM
 	select HARDIRQS_SW_RESEND
 	select HAS_IOPORT
 	select HAVE_ARCH_AUDITSYSCALL if AEABI && !OABI_COMPAT
-	select HAVE_ARCH_BITREVERSE if (CPU_32v7M || CPU_32v7) && !CPU_32v6
+	select HAVE_ARCH_BITREVERSE if (CPU_32v7M || CPU_32v7) && !CPU_32v6 && BITREVERSE
 	select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL && !CPU_ENDIAN_BE32 && MMU && (!PREEMPT_RT || !SMP)
 	select HAVE_ARCH_KFENCE if MMU && !XIP_KERNEL
 	select HAVE_ARCH_KGDB if !CPU_ENDIAN_BE32 && MMU
diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig
index fe60738e5943..f5bb62c2ba9c 100644
--- a/arch/arm64/Kconfig
+++ b/arch/arm64/Kconfig
@@ -150,7 +150,7 @@ config ARM64
 	select HAVE_ACPI_APEI if (ACPI && EFI)
 	select HAVE_ALIGNED_STRUCT_PAGE
 	select HAVE_ARCH_AUDITSYSCALL
-	select HAVE_ARCH_BITREVERSE
+	select HAVE_ARCH_BITREVERSE if BITREVERSE
 	select HAVE_ARCH_COMPILER_H
 	select HAVE_ARCH_HUGE_VMALLOC
 	select HAVE_ARCH_HUGE_VMAP
diff --git a/arch/loongarch/Kconfig b/arch/loongarch/Kconfig
index 3b042dbb2c41..6c3444e31c0e 100644
--- a/arch/loongarch/Kconfig
+++ b/arch/loongarch/Kconfig
@@ -115,7 +115,7 @@ config LOONGARCH
 	select HAS_IOPORT
 	select HAVE_ALIGNED_STRUCT_PAGE if 64BIT
 	select HAVE_ARCH_AUDITSYSCALL
-	select HAVE_ARCH_BITREVERSE if 64BIT
+	select HAVE_ARCH_BITREVERSE if 64BIT && BITREVERSE
 	select HAVE_ARCH_JUMP_LABEL
 	select HAVE_ARCH_JUMP_LABEL_RELATIVE
 	select HAVE_ARCH_KASAN if 64BIT
diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig
index 4364f3dba688..7e1494e0dbfa 100644
--- a/arch/mips/Kconfig
+++ b/arch/mips/Kconfig
@@ -2026,7 +2026,7 @@ config CPU_MIPSR6
 	default y if CPU_MIPS32_R6 || CPU_MIPS64_R6
 	select CPU_HAS_RIXI
 	select CPU_HAS_DIEI if !CPU_DIEI_BROKEN
-	select HAVE_ARCH_BITREVERSE
+	select HAVE_ARCH_BITREVERSE if BITREVERSE
 	select MIPS_ASID_BITS_VARIABLE
 	select MIPS_SPRAM
 
diff --git a/lib/Kconfig b/lib/Kconfig
index 00a9509636c1..d8e7e89ae320 100644
--- a/lib/Kconfig
+++ b/lib/Kconfig
@@ -58,6 +58,7 @@ config BITREVERSE
 config HAVE_ARCH_BITREVERSE
 	bool
 	default n
+	depends on BITREVERSE
 	help
 	  This option enables the use of hardware bit-reversal instructions on
 	  architectures which support such operations.
-- 
2.51.0

[PATCH v2 2/5] lib/bitrev: Introduce GENERIC_BITREVERSE

[Yury Norov]

The generic bit reversal implementation is controlled by
!HAVE_ARCH_BITREVERSE. This makes it difficult for architectures to
provide a hardware-accelerated implementation while being able to
fall back to the generic version if needed.

This patch adds GENERIC_BITREVERSE, so bitreverse API is controlled by
BITREVERSE, GENERIC_BITREVERSE and HAVE_ARCH_BITREVERSE options. The
relationship between them is described as follows:

 - BITREVERSE is selected by user code; it's required to generate the API;
 - Architectures may select HAVE_ARCH_BITREVERSE and provide an arch
   implementation in arch/$(ARCH)/include/asm/bitrev.h.
 - if HAVE_ARCH_BITREVERSE isn't set, BITREVERSE selects GENERIC_BITREVERSE;
 - if GENERIC_BITREVERSE is set and HAVE_ARCH_BITREVERSE is not, the kernel
   provides generic implementation only, and wires bitrevXX() to it.
 - if HAVE_ARCH_BITREVERSE is set and GENERIC_BITREVERSE is not, the arch
   code provides __arch_bitrevXX(), and it is wired to bitrevXX();
 - if both GENERIC_BITREVERSE and HAVE_ARCH_BITREVERSE are selected, the kernel
   generates generic___bitrev(), but wires bitrev() to the __arch_bitrev().

The last option allows architectures to use generic___bitrev() as a
fallback option.

Drivers and core code should never select GENERIC_BITREVERSE or
HAVE_ARCH_BITREVERSE explicitly.

Architectures that require generic bitreverse API as a fallback should
explicitly enable GENERIC_BITREVERSE together with HAVE_ARCH_BITREVERSE.

Signed-off-by: Yury Norov <ynorov@nvidia.com>
---
 lib/Kconfig  | 12 ++++++++++++
 lib/Makefile |  2 +-
 lib/bitrev.c |  3 ---
 3 files changed, 13 insertions(+), 4 deletions(-)

diff --git a/lib/Kconfig b/lib/Kconfig
index d8e7e89ae320..a33988adfaa3 100644
--- a/lib/Kconfig
+++ b/lib/Kconfig
@@ -54,6 +54,7 @@ config PACKING_KUNIT_TEST
 
 config BITREVERSE
 	tristate
+	select GENERIC_BITREVERSE if !HAVE_ARCH_BITREVERSE
 
 config HAVE_ARCH_BITREVERSE
 	bool
@@ -63,6 +64,17 @@ config HAVE_ARCH_BITREVERSE
 	  This option enables the use of hardware bit-reversal instructions on
 	  architectures which support such operations.
 
+config GENERIC_BITREVERSE
+	tristate
+	depends on BITREVERSE
+	help
+	  Generic bit reversal implementation. Drivers should never enable
+	  it explicitly. Instead, enable BITREVERSE.
+
+	  Architectures may want to select it as a fall-back option for
+	  HAVE_ARCH_BITREVERSE, when the hardware-accelerated bit reverse
+	  instruction set is optional, like RISC-V ZBKB extension.
+
 config ARCH_HAS_STRNCPY_FROM_USER
 	bool
 
diff --git a/lib/Makefile b/lib/Makefile
index f33a24bf1c19..23e07d19d01c 100644
--- a/lib/Makefile
+++ b/lib/Makefile
@@ -145,7 +145,7 @@ obj-$(CONFIG_DEBUG_PREEMPT) += smp_processor_id.o
 obj-$(CONFIG_LIST_HARDENED) += list_debug.o
 obj-$(CONFIG_DEBUG_OBJECTS) += debugobjects.o
 
-obj-$(CONFIG_BITREVERSE) += bitrev.o
+obj-$(CONFIG_GENERIC_BITREVERSE) += bitrev.o
 obj-$(CONFIG_LINEAR_RANGES) += linear_ranges.o
 obj-$(CONFIG_PACKING)	+= packing.o
 obj-$(CONFIG_PACKING_KUNIT_TEST) += packing_test.o
diff --git a/lib/bitrev.c b/lib/bitrev.c
index 81b56e0a7f32..05088231f31f 100644
--- a/lib/bitrev.c
+++ b/lib/bitrev.c
@@ -1,5 +1,4 @@
 // SPDX-License-Identifier: GPL-2.0-only
-#ifndef CONFIG_HAVE_ARCH_BITREVERSE
 #include <linux/types.h>
 #include <linux/module.h>
 #include <linux/bitrev.h>
@@ -43,5 +42,3 @@ const u8 byte_rev_table[256] = {
 	0x1f, 0x9f, 0x5f, 0xdf, 0x3f, 0xbf, 0x7f, 0xff,
 };
 EXPORT_SYMBOL_GPL(byte_rev_table);
-
-#endif /* CONFIG_HAVE_ARCH_BITREVERSE */
-- 
2.51.0

[PATCH v2 3/5] bitops: Define generic___bitrev8/16/32 for reuse

[Yury Norov]

From: Jinjie Ruan <ruanjinjie@huawei.com>

Define generic___bitrev8/16/32 using the implementation in
<linux/bitrev.h>, so they can be reused in <asm/bitrev.h>,
such as RISCV.

Reviewed-by: Yury Norov <ynorov@nvidia.com>
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
Signed-off-by: Yury Norov <ynorov@nvidia.com>
---
 include/asm-generic/bitops/__bitrev.h | 23 +++++++++++++++++++++++
 include/linux/bitrev.h                | 20 ++++----------------
 2 files changed, 27 insertions(+), 16 deletions(-)
 create mode 100644 include/asm-generic/bitops/__bitrev.h

diff --git a/include/asm-generic/bitops/__bitrev.h b/include/asm-generic/bitops/__bitrev.h
new file mode 100644
index 000000000000..4addbde14050
--- /dev/null
+++ b/include/asm-generic/bitops/__bitrev.h
@@ -0,0 +1,23 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _ASM_GENERIC_BITOPS___BITREV_H_
+#define _ASM_GENERIC_BITOPS___BITREV_H_
+
+#include <asm/types.h>
+
+extern u8 const byte_rev_table[256];
+static __always_inline __attribute_const__ u8 generic___bitrev8(u8 byte)
+{
+	return byte_rev_table[byte];
+}
+
+static __always_inline __attribute_const__ u16 generic___bitrev16(u16 x)
+{
+	return (generic___bitrev8(x & 0xff) << 8) | generic___bitrev8(x >> 8);
+}
+
+static __always_inline __attribute_const__ u32 generic___bitrev32(u32 x)
+{
+	return (generic___bitrev16(x & 0xffff) << 16) | generic___bitrev16(x >> 16);
+}
+
+#endif /* _ASM_GENERIC_BITOPS___BITREV_H_ */
diff --git a/include/linux/bitrev.h b/include/linux/bitrev.h
index d35b8ec1c485..11620a70e776 100644
--- a/include/linux/bitrev.h
+++ b/include/linux/bitrev.h
@@ -12,22 +12,10 @@
 #define __bitrev8 __arch_bitrev8
 
 #else
-extern u8 const byte_rev_table[256];
-static inline u8 __bitrev8(u8 byte)
-{
-	return byte_rev_table[byte];
-}
-
-static inline u16 __bitrev16(u16 x)
-{
-	return (__bitrev8(x & 0xff) << 8) | __bitrev8(x >> 8);
-}
-
-static inline u32 __bitrev32(u32 x)
-{
-	return (__bitrev16(x & 0xffff) << 16) | __bitrev16(x >> 16);
-}
-
+#include <asm-generic/bitops/__bitrev.h>
+#define __bitrev32 generic___bitrev32
+#define __bitrev16 generic___bitrev16
+#define __bitrev8 generic___bitrev8
 #endif /* CONFIG_HAVE_ARCH_BITREVERSE */
 
 #define __bitrev8x4(x)	(__bitrev32(swab32(x)))
-- 
2.51.0

[PATCH v2 4/5] arch/riscv: Add bitrev.h file to support rev8 and brev8

[Yury Norov]

From: Jinjie Ruan <ruanjinjie@huawei.com>

The RISC-V Bit-manipulation Extension for Cryptography (Zbkb) provides
the 'brev8' instruction, which reverses the bits within each byte.
Combined with the 'rev8' instruction (from Zbb or Zbkb), which reverses
the byte order of a register, we can efficiently implement 16-bit,
32-bit, and (on RV64) 64-bit bit reversal.

This is significantly faster than the default software table-lookup
implementation in lib/bitrev.c, as it replaces memory accesses and
multiple arithmetic operations with just two or three hardware
instructions.

Select HAVE_ARCH_BITREVERSE as well as GENERIC_BITREVERSE,
and provide <asm/bitrev.h> to utilize these instructions when
the Zbkb extension is available at runtime via the alternatives
mechanism.

[Yury: select the options conditionally on BITREVERSE]

Link: https://docs.riscv.org/reference/isa/unpriv/b-st-ext.html
Suggested-by: David Laight <david.laight.linux@gmail.com>
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
Signed-off-by: Yury Norov <ynorov@nvidia.com>
---
 arch/riscv/Kconfig              |  2 ++
 arch/riscv/include/asm/bitrev.h | 51 +++++++++++++++++++++++++++++++++
 2 files changed, 53 insertions(+)
 create mode 100644 arch/riscv/include/asm/bitrev.h

diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
index d235396c4514..a708583f785d 100644
--- a/arch/riscv/Kconfig
+++ b/arch/riscv/Kconfig
@@ -104,6 +104,7 @@ config RISCV
 	select FUNCTION_ALIGNMENT_8B if DYNAMIC_FTRACE_WITH_CALL_OPS
 	select GENERIC_ARCH_TOPOLOGY
 	select GENERIC_ATOMIC64 if !64BIT
+	select GENERIC_BITREVERSE if HAVE_ARCH_BITREVERSE
 	select GENERIC_CLOCKEVENTS_BROADCAST if SMP
 	select GENERIC_CPU_DEVICES
 	select GENERIC_CPU_VULNERABILITIES
@@ -128,6 +129,7 @@ config RISCV
 	select HAS_IOPORT if MMU
 	select HAVE_ALIGNED_STRUCT_PAGE
 	select HAVE_ARCH_AUDITSYSCALL
+	select HAVE_ARCH_BITREVERSE if RISCV_ISA_ZBKB && BITREVERSE
 	select HAVE_ARCH_HUGE_VMALLOC if HAVE_ARCH_HUGE_VMAP
 	select HAVE_ARCH_HUGE_VMAP if MMU && 64BIT
 	select HAVE_ARCH_JUMP_LABEL
diff --git a/arch/riscv/include/asm/bitrev.h b/arch/riscv/include/asm/bitrev.h
new file mode 100644
index 000000000000..4b9b8d34cc3b
--- /dev/null
+++ b/arch/riscv/include/asm/bitrev.h
@@ -0,0 +1,51 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __ASM_BITREV_H
+#define __ASM_BITREV_H
+
+#include <linux/types.h>
+#include <asm/cpufeature-macros.h>
+#include <asm/hwcap.h>
+#include <asm-generic/bitops/__bitrev.h>
+
+static __always_inline __attribute_const__ u32 __arch_bitrev32(u32 x)
+{
+	unsigned long result;
+
+	if (!riscv_has_extension_likely(RISCV_ISA_EXT_ZBKB))
+		return generic___bitrev32(x);
+
+	asm volatile(
+		".option push\n"
+		".option arch,+zbkb\n"
+		"rev8 %0, %1\n"
+		"brev8 %0, %0\n"
+		".option pop"
+		: "=r" (result) : "r" ((long)x)
+	);
+
+	return result >> (__riscv_xlen - 32);
+}
+
+static __always_inline __attribute_const__ u16 __arch_bitrev16(u16 x)
+{
+	return __arch_bitrev32(x) >> 16;
+}
+
+static __always_inline __attribute_const__ u8 __arch_bitrev8(u8 x)
+{
+	unsigned long result;
+
+	if (!riscv_has_extension_likely(RISCV_ISA_EXT_ZBKB))
+		return generic___bitrev8(x);
+
+	asm volatile(
+		".option push\n"
+		".option arch,+zbkb\n"
+		"brev8 %0, %1\n"
+		".option pop"
+		: "=r" (result) : "r" ((long)x)
+	);
+
+	return result;
+}
+#endif
-- 
2.51.0

[PATCH v2 5/5] MAINTAINERS: BITOPS: include bitrev.[ch]

[Yury Norov]

Arch bitrev API is covered in MAINTAINERS under the BITOPS entry,
while generic bitrev is unmaintained. Move it under BITOPS too.

Signed-off-by: Yury Norov <ynorov@nvidia.com>
---
 MAINTAINERS | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/MAINTAINERS b/MAINTAINERS
index 882214b0e7db..30214ac2f06d 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -4549,7 +4549,9 @@ F:	arch/*/lib/bitops.c
 F:	include/asm-generic/bitops
 F:	include/asm-generic/bitops.h
 F:	include/linux/bitops.h
+F:	include/linux/bitrev.h
 F:	include/linux/count_zeros.h
+F:	lib/bitrev.c
 F:	lib/hweight.c
 F:	lib/test_bitops.c
 F:	lib/tests/bitops_kunit.c
-- 
2.51.0