Update comments to match reality.

diff --git a/sys/dev/sbus/esp_sbus.c b/sys/dev/sbus/esp_sbus.c
index 790493e..2a33a89 100644 (file)
--- a/sys/dev/sbus/esp_sbus.c
+++ b/sys/dev/sbus/esp_sbus.c
@@ -1,4 +1,4 @@
-/*     $OpenBSD: esp_sbus.c,v 1.26 2022/03/13 13:34:54 mpi Exp $       */
+/*     $OpenBSD: esp_sbus.c,v 1.27 2024/05/17 20:03:13 miod Exp $      */
 /*     $NetBSD: esp_sbus.c,v 1.14 2001/04/25 17:53:37 bouyer Exp $     */
 
 /*-
@@ -262,18 +262,14 @@ espattach_sbus(struct device *parent, struct device *self, void *aux)
 
        /*
         * Find the DMA by poking around the dma device structures
-        *
         * What happens here is that if the dma driver has not been
-        * configured, then this returns a NULL pointer. Then when the
-        * dma actually gets configured, it does the opposing test, and
-        * if the sc->sc_esp field in its softc is NULL, then tries to
-        * find the matching esp driver.
+        * configured, then this returns a NULL pointer.
         */
        esc->sc_dma = (struct lsi64854_softc *)
            getdevunit("dma", sc->sc_dev.dv_unit - esp_unit_offset);
 
        /*
-        * and a back pointer to us, for DMA
+        * add a back pointer to us, for DMA
         */
        if (esc->sc_dma)
                esc->sc_dma->sc_client = sc;