Wang Ming <machel@xxxxxxxx> writes: > It is expected that most callers should _ignore_ the errors > return by debugfs_create_dir() in ath9k_htc_init_debug(). > > Signed-off-by: Wang Ming <machel@xxxxxxxx> > --- > drivers/net/wireless/ath/ath9k/htc_drv_debug.c | 2 -- > 1 file changed, 2 deletions(-) > > diff --git a/drivers/net/wireless/ath/ath9k/htc_drv_debug.c b/drivers/net/wireless/ath/ath9k/htc_drv_debug.c > index b3ed65e5c4da..85ad45771b44 100644 > --- a/drivers/net/wireless/ath/ath9k/htc_drv_debug.c > +++ b/drivers/net/wireless/ath/ath9k/htc_drv_debug.c > @@ -491,8 +491,6 @@ int ath9k_htc_init_debug(struct ath_hw *ah) > > priv->debug.debugfs_phy = debugfs_create_dir(KBUILD_MODNAME, > priv->hw->wiphy->debugfsdir); > - if (!priv->debug.debugfs_phy) > - return -ENOMEM; Hmm, so it's true that all the debugfs_create* functions deal correctly with the dir pointer being an error pointer, which means that it's possible to just ignore the return value of debugfs_create_dir() without anything breaking. However, it also seems kinda pointless to have all those calls if we know they're going to fail, so I prefer v1 of this patch that just fixed the IS_ERR check. No need to resend, we can just apply v1 instead... -Toke
