mac80211: fix and optimize MCS mask handling

Currently the code always copies the configured MCS mask (even if it is
set to default), but only uses it if legacy rates were also masked out.
Fix this by adding a flag that tracks whether the configured MCS mask is
set to default or not.
Optimize the code further by storing a pointer to the configured rate
mask in txrc instead of using memcpy.

Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
diff --git a/include/net/mac80211.h b/include/net/mac80211.h
index bc5d818..05dbb97 100644
--- a/include/net/mac80211.h
+++ b/include/net/mac80211.h
@@ -4107,7 +4107,7 @@
  *	(deprecated; this will be removed once drivers get updated to use
  *	rate_idx_mask)
  * @rate_idx_mask: user-requested (legacy) rate mask
- * @rate_idx_mcs_mask: user-requested MCS rate mask
+ * @rate_idx_mcs_mask: user-requested MCS rate mask (NULL if not in use)
  * @bss: whether this frame is sent out in AP or IBSS mode
  */
 struct ieee80211_tx_rate_control {
@@ -4119,7 +4119,7 @@
 	bool rts, short_preamble;
 	u8 max_rate_idx;
 	u32 rate_idx_mask;
-	u8 rate_idx_mcs_mask[IEEE80211_HT_MCS_MASK_LEN];
+	u8 *rate_idx_mcs_mask;
 	bool bss;
 };
 
diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c
index fdd95bd..72ab1c0 100644
--- a/net/mac80211/cfg.c
+++ b/net/mac80211/cfg.c
@@ -2417,9 +2417,22 @@
 	}
 
 	for (i = 0; i < IEEE80211_NUM_BANDS; i++) {
+		struct ieee80211_supported_band *sband = wiphy->bands[i];
+		int j;
+
 		sdata->rc_rateidx_mask[i] = mask->control[i].legacy;
 		memcpy(sdata->rc_rateidx_mcs_mask[i], mask->control[i].mcs,
 		       sizeof(mask->control[i].mcs));
+
+		sdata->rc_has_mcs_mask[i] = false;
+		if (!sband)
+			continue;
+
+		for (j = 0; j < IEEE80211_HT_MCS_MASK_LEN; j++)
+			if (~sdata->rc_rateidx_mcs_mask[i][j]) {
+				sdata->rc_has_mcs_mask[i] = true;
+				break;
+			}
 	}
 
 	return 0;
diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h
index f4a65a3..21c1720 100644
--- a/net/mac80211/ieee80211_i.h
+++ b/net/mac80211/ieee80211_i.h
@@ -739,6 +739,8 @@
 
 	/* bitmap of allowed (non-MCS) rate indexes for rate control */
 	u32 rc_rateidx_mask[IEEE80211_NUM_BANDS];
+
+	bool rc_has_mcs_mask[IEEE80211_NUM_BANDS];
 	u8  rc_rateidx_mcs_mask[IEEE80211_NUM_BANDS][IEEE80211_HT_MCS_MASK_LEN];
 
 	union {
diff --git a/net/mac80211/rate.c b/net/mac80211/rate.c
index dd88381..5d545dd2 100644
--- a/net/mac80211/rate.c
+++ b/net/mac80211/rate.c
@@ -460,9 +460,12 @@
 	 * the common case.
 	 */
 	mask = sdata->rc_rateidx_mask[info->band];
-	memcpy(mcs_mask, sdata->rc_rateidx_mcs_mask[info->band],
-	       sizeof(mcs_mask));
-	if (mask != (1 << txrc->sband->n_bitrates) - 1) {
+	if (mask != (1 << txrc->sband->n_bitrates) - 1 || txrc->rate_idx_mcs_mask) {
+		if (txrc->rate_idx_mcs_mask)
+			memcpy(mcs_mask, txrc->rate_idx_mcs_mask, sizeof(mcs_mask));
+		else
+			memset(mcs_mask, 0xff, sizeof(mcs_mask));
+
 		if (sta) {
 			/* Filter out rates that the STA does not support */
 			mask &= sta->sta.supp_rates[info->band];
diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c
index bb82c87..15c1b28 100644
--- a/net/mac80211/tx.c
+++ b/net/mac80211/tx.c
@@ -642,9 +642,11 @@
 		txrc.max_rate_idx = -1;
 	else
 		txrc.max_rate_idx = fls(txrc.rate_idx_mask) - 1;
-	memcpy(txrc.rate_idx_mcs_mask,
-	       tx->sdata->rc_rateidx_mcs_mask[info->band],
-	       sizeof(txrc.rate_idx_mcs_mask));
+
+	if (tx->sdata->rc_has_mcs_mask[info->band])
+		txrc.rate_idx_mcs_mask =
+			tx->sdata->rc_rateidx_mcs_mask[info->band];
+
 	txrc.bss = (tx->sdata->vif.type == NL80211_IFTYPE_AP ||
 		    tx->sdata->vif.type == NL80211_IFTYPE_MESH_POINT ||
 		    tx->sdata->vif.type == NL80211_IFTYPE_ADHOC);
@@ -2508,8 +2510,6 @@
 		txrc.max_rate_idx = -1;
 	else
 		txrc.max_rate_idx = fls(txrc.rate_idx_mask) - 1;
-	memcpy(txrc.rate_idx_mcs_mask, sdata->rc_rateidx_mcs_mask[band],
-	       sizeof(txrc.rate_idx_mcs_mask));
 	txrc.bss = true;
 	rate_control_get_rate(sdata, NULL, &txrc);