drm/i915/hdcp: Read Rxcaps for robustibility

We see some monitors and docks report incorrect hdcp version
and capability in first few reads so we read rx_caps three times
before we conclude the monitor's or docks HDCP capability

--v2
-Add comment to justify the 3 time read loop for hdcp capability[Ankit]

Signed-off-by: Suraj Kandpal <suraj.kandpal@intel.com>
Reviewed-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
Signed-off-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240226063051.1685326-7-suraj.kandpal@intel.com

drivers/gpu/drm/i915/display/intel_dp_hdcp.c

@@ -644,18 +644,29 @@ int _intel_dp_hdcp2_get_capability(struct drm_dp_aux *aux,
 				   bool *capable)
 {
 	u8 rx_caps[3];
-	int ret;
+	int ret, i;
 
 	*capable = false;
-	ret = drm_dp_dpcd_read(aux,
-			       DP_HDCP_2_2_REG_RX_CAPS_OFFSET,
-			       rx_caps, HDCP_2_2_RXCAPS_LEN);
-	if (ret != HDCP_2_2_RXCAPS_LEN)
-		return ret >= 0 ? -EIO : ret;
 
-	if (rx_caps[0] == HDCP_2_2_RX_CAPS_VERSION_VAL &&
-	    HDCP_2_2_DP_HDCP_CAPABLE(rx_caps[2]))
-		*capable = true;
+	/*
+	 * Some HDCP monitors act really shady by not giving the correct hdcp
+	 * capability on the first rx_caps read and usually take an extra read
+	 * to give the capability. We read rx_caps three times before we
+	 * declare a monitor not capable of HDCP 2.2.
+	 */
+	for (i = 0; i < 3; i++) {
+		ret = drm_dp_dpcd_read(aux,
+				       DP_HDCP_2_2_REG_RX_CAPS_OFFSET,
+				       rx_caps, HDCP_2_2_RXCAPS_LEN);
+		if (ret != HDCP_2_2_RXCAPS_LEN)
+			return ret >= 0 ? -EIO : ret;
+
+		if (rx_caps[0] == HDCP_2_2_RX_CAPS_VERSION_VAL &&
+		    HDCP_2_2_DP_HDCP_CAPABLE(rx_caps[2])) {
+			*capable = true;
+			break;
+		}
+	}
 
 	return 0;
 }