On Fri, Oct 12, 2018 at 04:58:17PM +0200, Philipp Tomsich wrote:
On 12.10.2018, at 16:34, Tom Rini trini@konsulko.com wrote:
We need to #ifdef some variables to avoid warning about them being unused.
Fixes: 1a048cd65645 ("driver: net: fsl-mc: Add support of multiple phys for dpmac") Signed-off-by: Tom Rini trini@konsulko.com
drivers/net/ldpaa_eth/ldpaa_eth.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ldpaa_eth/ldpaa_eth.c b/drivers/net/ldpaa_eth/ldpaa_eth.c index fe1c03e9e436..73b7ba29dfdc 100644 --- a/drivers/net/ldpaa_eth/ldpaa_eth.c +++ b/drivers/net/ldpaa_eth/ldpaa_eth.c @@ -402,10 +402,12 @@ error: static int ldpaa_get_dpmac_state(struct ldpaa_eth_priv *priv, struct dpmac_link_state *state) {
- struct phy_device *phydev = NULL; phy_interface_t enet_if;
- int phy_num, phys_detected;
- int err;
- int phys_detected;
+#ifdef CONFIG_PHYLIB
- struct phy_device *phydev = NULL;
- int err, phy_num;
+#endif
Reading this, I started to wonder whether we want to use __maybe_unused to suppress these warnings w/o having to add these #if defined(…) blocks?
The drawback is that GCC won’t warn about lingering unused variables is a ‘maybe’ unused becomes a permanently unused variable...
Yeah, I'm not sure sometimes if __maybe_unused is cleaner looking or not. I think it depends on the number of things that are or aren't needing to get blocked off.