On Thu, Mar 17, 2022 at 07:02:06PM +0100, Marcus MERIGHI wrote: > Hello! > > Thanks for your work on this! > > [email protected] (Stefan Sperling), 2022.03.16 (Wed) 16:11 (CET): > > This patch adds initial 11ac support to the iwm(4) driver. > > It allows use of 80 MHz channels and VHT MCS. > [...] > > So far, I have tested this successfully on iwm(4) 8265. > > With this patch a "larger" file transfer will reliably panic my machine. > "large" means even small files. > ICMP and NTP (and probably DNS) seem to work, though. > I've tried it six times, it always crashed. > > This is with: > iwm0 at pci2 dev 0 function 0 "Intel Dual Band Wireless-AC 8265" rev 0x88, msi > iwm0: hw rev 0x230, fw ver 36.ca7b901d.0, address 00:28:f8:xx:yy:zz
If you have not done so yet, could you please try the revised version of the patch which I posted here? https://marc.info/?l=openbsd-tech&m=164746924001340&w=2 I believe this should fix your issue. Below is the difference between the two versions, in case it is easier for you to apply the fix on top of the first version of the patch. diff 0191ab443b6994d9adb38f5f22f66ffea008c0b7 00811a880ed9b9c98b475f172bf5f6fe3e31811f blob - 3d13d9f024f349521cdf890955a4b6ce2e88fd5d blob + dea634bcb86b2e41a7da27dd7be552b751eb1b66 --- sys/net80211/ieee80211_ra_vht.c +++ sys/net80211/ieee80211_ra_vht.c @@ -233,7 +233,7 @@ const struct ieee80211_vht_rateset * ieee80211_ra_vht_next_rateset(struct ieee80211_ra_vht_node *rn, struct ieee80211_node *ni) { - const struct ieee80211_vht_rateset *rs; + const struct ieee80211_vht_rateset *rs, *rsnext; int next; int sgi = ieee80211_ra_vht_use_sgi(ni); int mcs = ni->ni_txmcs; @@ -269,7 +269,11 @@ ieee80211_ra_vht_next_rateset(struct ieee80211_ra_vht_ } else panic("%s: invalid probing mode %d", __func__, rn->probing); - return &ieee80211_std_ratesets_11ac[next]; + rsnext = &ieee80211_std_ratesets_11ac[next]; + if (rn->valid_rates[rsnext->num_ss - 1] == 0) + return NULL; + + return rsnext; } int
