The NVEC_PHD macro is a debugging helper that is only enabled when
NVEC_PS2_DEBUG is defined. As this flag is never defined in the kernel,
the macro and all of its call sites are dead code.

As suggested by Greg Kroah-Hartman, removing the code is the cleanest
solution. This also fixes a build error discovered by the kernel test
robot after the v1 patch modified the macro's body.

Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202508182005.4PiKxXcN-lkp@intel.com/
Signed-off-by: Mohammed Guermoud <mohammed.guermoud@gmail.com>
---
 drivers/staging/nvec/nvec_ps2.c | 13 -------------
 1 file changed, 13 deletions(-)

diff --git a/drivers/staging/nvec/nvec_ps2.c b/drivers/staging/nvec/nvec_ps2.c
index 7accdcafeeac..2e36893e0008 100644
--- a/drivers/staging/nvec/nvec_ps2.c
+++ b/drivers/staging/nvec/nvec_ps2.c
@@ -23,14 +23,6 @@
 #define DISABLE_MOUSE	0xf5
 #define PSMOUSE_RST	0xff
 
-#ifdef NVEC_PS2_DEBUG
-#define NVEC_PHD(str, buf, len) \
-	print_hex_dump(KERN_DEBUG, str, DUMP_PREFIX_NONE, \
-			16, 1, buf, len, false)
-#else
-#define NVEC_PHD(str, buf, len) do { (void)str; (void)buf; (void)len; } while (0)
-#endif
-
 enum ps2_subcmds {
 	SEND_COMMAND = 1,
 	RECEIVE_N,
@@ -70,18 +62,13 @@ static int nvec_ps2_notifier(struct notifier_block *nb,
 	case NVEC_PS2_EVT:
 		for (i = 0; i < msg[1]; i++)
 			serio_interrupt(ps2_dev.ser_dev, msg[2 + i], 0);
-		NVEC_PHD("ps/2 mouse event: ", &msg[2], msg[1]);
 		return NOTIFY_STOP;
 
 	case NVEC_PS2:
 		if (msg[2] == 1) {
 			for (i = 0; i < (msg[1] - 2); i++)
 				serio_interrupt(ps2_dev.ser_dev, msg[i + 4], 0);
-			NVEC_PHD("ps/2 mouse reply: ", &msg[4], msg[1] - 2);
 		}
-
-		else if (msg[1] != 2) /* !ack */
-			NVEC_PHD("unhandled mouse event: ", msg, msg[1] + 2);
 		return NOTIFY_STOP;
 	}
 
-- 
2.50.1

On Mon, Aug 18, 2025 at 09:38:55PM +0100, Mohammed Guermoud wrote:
> The NVEC_PHD macro is a debugging helper that is only enabled when
> NVEC_PS2_DEBUG is defined. As this flag is never defined in the kernel,
> the macro and all of its call sites are dead code.
> 
> As suggested by Greg Kroah-Hartman, removing the code is the cleanest
> solution. This also fixes a build error discovered by the kernel test
> robot after the v1 patch modified the macro's body.

We don't reference v1 here.  We would do it under the --- cut off.

> 
> Reported-by: kernel test robot <lkp@intel.com>
> Closes: https://lore.kernel.org/oe-kbuild-all/202508182005.4PiKxXcN-lkp@intel.com/

Don't add these tags.  They're just confusing because we never merged v1.

> Signed-off-by: Mohammed Guermoud <mohammed.guermoud@gmail.com>
> ---
Here you would put a comment like:

v2: version caused a build error.  It's better to just delete the
    code instead of trying to silence the warning.

Please send v3 with a comment:

v3: Fix commit message.
v2: version caused a build error.  It's better to just delete the
    code instead of trying to silence the warning.

https://staticthinking.wordpress.com/2022/07/27/how-to-send-a-v2-patch/

regards,
dan carpenter