On 9/15/21 12:13 AM, Geert Uytterhoeven wrote:

Hi Günter,

On Wed, Sep 15, 2021 at 5:52 AM Guenter Roeck [off-list ref] wrote:

quoted

absolute_pointer() disassociates a pointer from its originating symbol
type and context. Use it to prevent compiler warnings/errors such as

drivers/net/ethernet/i825xx/82596.c: In function 'i82596_probe':
./arch/m68k/include/asm/string.h:72:25: error:
         '__builtin_memcpy' reading 6 bytes from a region of size 0
                 [-Werror=stringop-overread]

Such warnings may be reported by gcc 11.x for string and memory operations
on fixed addresses.

Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
---
v2: No change

  include/linux/compiler.h | 2 ++
  1 file changed, 2 insertions(+)

diff --git a/include/linux/compiler.h b/include/linux/compiler.h
index b67261a1e3e9..3d5af56337bd 100644
--- a/include/linux/compiler.h
+++ b/include/linux/compiler.h

@@ -188,6 +188,8 @@ void ftrace_likely_update(struct ftrace_likely_data *f, int val,
      (typeof(ptr)) (__ptr + (off)); })
  #endif

+#define absolute_pointer(val)  RELOC_HIDE((void *)(val), 0)

I guess we're not worried about "val" being evaluated multiple
times inside RELOC_HIDE(), as this is mainly intended for constants?

No, we are not. It is quite similar to RELOC_HIDE() in that regard.

Guenter