summaryrefslogtreecommitdiffstats
path: root/sys/dev
diff options
context:
space:
mode:
authorimp <imp@FreeBSD.org>2009-06-01 16:22:01 +0000
committerimp <imp@FreeBSD.org>2009-06-01 16:22:01 +0000
commit36075f2a9a28015328775bf2e599fd600579ac09 (patch)
treeec2a2ba3ddf0b496b529a5ff6dbceb1cb4915c5e /sys/dev
parentb850b8553439e2dd1340a5518e8b22c5303276ab (diff)
downloadFreeBSD-src-36075f2a9a28015328775bf2e599fd600579ac09.zip
FreeBSD-src-36075f2a9a28015328775bf2e599fd600579ac09.tar.gz
Add a comment about what may be happening when we get certain
messages. No change to actual code.
Diffstat (limited to 'sys/dev')
-rw-r--r--sys/dev/bwi/if_bwi.c17
1 files changed, 17 insertions, 0 deletions
diff --git a/sys/dev/bwi/if_bwi.c b/sys/dev/bwi/if_bwi.c
index 8405e94..83f6d10 100644
--- a/sys/dev/bwi/if_bwi.c
+++ b/sys/dev/bwi/if_bwi.c
@@ -1628,6 +1628,23 @@ bwi_intr(void *xsc)
/* Disable all interrupts */
bwi_disable_intrs(sc, BWI_ALL_INTRS);
+ /*
+ * http://bcm-specs.sipsolutions.net/Interrupts
+ * Says for this bit (0x800):
+ * "Fatal Error
+ *
+ * We got this one while testing things when by accident the
+ * template ram wasn't set to big endian when it should have
+ * been after writing the initial values. It keeps on being
+ * triggered, the only way to stop it seems to shut down the
+ * chip."
+ *
+ * Suggesting that we should never get it and if we do we're not
+ * feeding TX packets into the MAC correctly if we do... Apparently,
+ * it is valid only on mac version 5 and higher, but I couldn't
+ * find a reference for that... Since I see them from time to time
+ * on my card, this suggests an error in the tx path still...
+ */
if (intr_status & BWI_INTR_PHY_TXERR) {
if (mac->mac_flags & BWI_MAC_F_PHYE_RESET) {
if_printf(ifp, "%s: intr PHY TX error\n", __func__);
OpenPOWER on IntegriCloud