| Message ID | 1442389740-29053-1-git-send-email-rmanohar@qti.qualcomm.com (mailing list archive) |
|---|---|
| State | Not Applicable |
| Delegated to: | Kalle Valo |
| Headers | show |
On 16 September 2015 at 09:49, Rajkumar Manoharan <rmanohar@qti.qualcomm.com> wrote: > ldpc is not configured for fixed rates. This blocks auto rate vs fixed > rate performance comparison. Since firmware is considering ldpc vdev > param for fixed rate selection, it has to be configured to enable ldpc > for fixed rates. > > Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com> > --- [...] > @@ -6041,6 +6049,7 @@ static int ath10k_mac_op_set_bitrate_mask(struct ieee80211_hw *hw, > band = def.chan->band; > ht_mcs_mask = mask->control[band].ht_mcs; > vht_mcs_mask = mask->control[band].vht_mcs; > + ldpc = !!(ar->ht_cap_info & WMI_HT_CAP_LDPC); Is still really correct? LDPC must be supported on the Rx side of destination stations hence I would assume you need to check whether all stations associated are LDPC capable before enabling it (e.g. via ieee80211_iterate_stations_atomic). Micha? -- To unsubscribe from this list: send the line "unsubscribe linux-wireless" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, Sep 21, 2015 at 09:05:54AM +0200, Michal Kazior wrote: > On 16 September 2015 at 09:49, Rajkumar Manoharan > <rmanohar@qti.qualcomm.com> wrote: > > ldpc is not configured for fixed rates. This blocks auto rate vs fixed > > rate performance comparison. Since firmware is considering ldpc vdev > > param for fixed rate selection, it has to be configured to enable ldpc > > for fixed rates. > > > > Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com> > > --- > [...] > > @@ -6041,6 +6049,7 @@ static int ath10k_mac_op_set_bitrate_mask(struct ieee80211_hw *hw, > > band = def.chan->band; > > ht_mcs_mask = mask->control[band].ht_mcs; > > vht_mcs_mask = mask->control[band].vht_mcs; > > + ldpc = !!(ar->ht_cap_info & WMI_HT_CAP_LDPC); > > Is still really correct? LDPC must be supported on the Rx side of > destination stations hence I would assume you need to check whether > all stations associated are LDPC capable before enabling it (e.g. via > ieee80211_iterate_stations_atomic). > I already check with firmware on this regard. Below is the clarification got from fw team. So no need to go through sta_iter. "3) how does fixed rate work for non-LDPC clients (ex. legacy clients), if LDPC support is enabled for the given vdev? The rates are selected by ANDng the vdev_rates and client rates. If client does not support LDPC and LDPC is enabled in vdev then, LDPC will not be used." -Rajkumar -- To unsubscribe from this list: send the line "unsubscribe linux-wireless" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Rajkumar Manoharan <rmanohar@qti.qualcomm.com> writes: > ldpc is not configured for fixed rates. This blocks auto rate vs fixed > rate performance comparison. Since firmware is considering ldpc vdev > param for fixed rate selection, it has to be configured to enable ldpc > for fixed rates. > > Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com> Thanks, applied.
diff --git a/drivers/net/wireless/ath/ath10k/mac.c b/drivers/net/wireless/ath/ath10k/mac.c index 80fe43fa..b347b60 100644 --- a/drivers/net/wireless/ath/ath10k/mac.c +++ b/drivers/net/wireless/ath/ath10k/mac.c @@ -5936,7 +5936,7 @@ ath10k_mac_bitrate_mask_get_single_rate(struct ath10k *ar, } static int ath10k_mac_set_fixed_rate_params(struct ath10k_vif *arvif, - u8 rate, u8 nss, u8 sgi) + u8 rate, u8 nss, u8 sgi, u8 ldpc) { struct ath10k *ar = arvif->ar; u32 vdev_param; @@ -5969,6 +5969,13 @@ static int ath10k_mac_set_fixed_rate_params(struct ath10k_vif *arvif, return ret; } + vdev_param = ar->wmi.vdev_param->ldpc; + ret = ath10k_wmi_vdev_set_param(ar, arvif->vdev_id, vdev_param, ldpc); + if (ret) { + ath10k_warn(ar, "failed to set ldpc param %d: %d\n", ldpc, ret); + return ret; + } + return 0; } @@ -6032,6 +6039,7 @@ static int ath10k_mac_op_set_bitrate_mask(struct ieee80211_hw *hw, u8 rate; u8 nss; u8 sgi; + u8 ldpc; int single_nss; int ret; @@ -6041,6 +6049,7 @@ static int ath10k_mac_op_set_bitrate_mask(struct ieee80211_hw *hw, band = def.chan->band; ht_mcs_mask = mask->control[band].ht_mcs; vht_mcs_mask = mask->control[band].vht_mcs; + ldpc = !!(ar->ht_cap_info & WMI_HT_CAP_LDPC); sgi = mask->control[band].gi; if (sgi == NL80211_TXRATE_FORCE_LGI) @@ -6079,7 +6088,7 @@ static int ath10k_mac_op_set_bitrate_mask(struct ieee80211_hw *hw, mutex_lock(&ar->conf_mutex); - ret = ath10k_mac_set_fixed_rate_params(arvif, rate, nss, sgi); + ret = ath10k_mac_set_fixed_rate_params(arvif, rate, nss, sgi, ldpc); if (ret) { ath10k_warn(ar, "failed to set fixed rate params on vdev %i: %d\n", arvif->vdev_id, ret);
ldpc is not configured for fixed rates. This blocks auto rate vs fixed rate performance comparison. Since firmware is considering ldpc vdev param for fixed rate selection, it has to be configured to enable ldpc for fixed rates. Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com> --- drivers/net/wireless/ath/ath10k/mac.c | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-)