[PATCH] m68k: hash: Use lower_16_bits() helper

Geert Uytterhoeven posted 1 patch 4 days ago
arch/m68k/include/asm/hash.h | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
[PATCH] m68k: hash: Use lower_16_bits() helper
Posted by Geert Uytterhoeven 4 days ago
When building for m68k with CONFIG_M68000=y and C=1:

    drivers/clk/rockchip/clk-rk3528.c: note: in included file (through include/linux/hash.h, include/linux/slab.h):
    arch/m68k/include/asm/hash.h:57:24: warning: cast truncates bits from constant value (18720 becomes 8720)
    arch/m68k/include/asm/hash.h:57:24: warning: cast truncates bits from constant value (1e8e8 becomes e8e8)

Sparse does not realize the truncation is intentional.
Make this explicit by using the lower_16_bits() helper instead, which
also masks the unwanted bits.

Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202605191434.PQkj2Rki-lkp@intel.com/
Reported-by: Heiko Stuebner <heiko@sntech.de>
Closes: https://lore.kernel.org/20260603213726.1025094-1-heiko@sntech.de/
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
---
No change in generated code.

 arch/m68k/include/asm/hash.h | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/arch/m68k/include/asm/hash.h b/arch/m68k/include/asm/hash.h
index 6d0d0c893f16dd49..6dadf493706632f5 100644
--- a/arch/m68k/include/asm/hash.h
+++ b/arch/m68k/include/asm/hash.h
@@ -2,6 +2,8 @@
 #ifndef _ASM_HASH_H
 #define _ASM_HASH_H
 
+#include <linux/wordpart.h>
+
 /*
  * If CONFIG_M68000=y (original mc68000/010), this file is #included
  * to work around the lack of a MULU.L instruction.
@@ -54,7 +56,7 @@ static inline u32 __attribute_const__ __hash_32(u32 x)
 	: "=&d,d" (a), "=&r,r" (b)
 	: "r,roi?" (x));	/* a+b = x*0x8647 */
 
-	return ((u16)(x*0x61c8) << 16) + a + b;
+	return (lower_16_bits(x * 0x61c8) << 16) + a + b;
 }
 
 #endif	/* _ASM_HASH_H */
-- 
2.43.0
Re: [PATCH] m68k: hash: Use lower_16_bits() helper
Posted by Daniel Palmer 3 days, 23 hours ago
Hi Geert,

On Thu, 4 Jun 2026 at 17:00, Geert Uytterhoeven <geert@linux-m68k.org> wrote:
> No change in generated code.

To be expected since the generated code hasn't been changed but I
build and boot tested in my 68000 virt machine and it still works:

smolsh / > uname -a
uClinux smol 7.1.0-rc6-00147-g6896dcea2588 #140 Thu Jun  4 17:54:03
JST 2026 m68k nolibc/Linux
[    2.890000] uname (26) used greatest stack depth: 4744 bytes left
smolsh / > cat /proc/cpuinfo
CPU:            68000
MMU:            none
FPU:            none
Clocking:       3169.5MHz
BogoMips:       933.06
Calibration:    466534400 loops

FWIW:

Tested-by: Daniel Palmer <daniel@thingy.jp>