| Message ID | 20220302222634.22185-11-quic_alokad@quicinc.com |
|---|---|
| State | Changes Requested |
| Headers | show |
| Series | hostapd: MBSSID and EMA support | expand |
On Wed, Mar 02, 2022 at 02:26:31PM -0800, Aloka Dixit wrote: > Set the DTIM periods of non-transmitted profiles equal to the EMA > profile periodicity if those are not a multiple of the latter > already as recommended in IEEE P802.11ax/D8.0, October 2020, > Multiple BSSID configuration examples, AA.1 Introduction. Please reference the published amendment instead of an older draft. > diff --git a/src/ap/ieee802_11.c b/src/ap/ieee802_11.c > static u8 * hostapd_eid_mbssid_elem(struct hostapd_data *hapd, u8 *eid, u8 *end, > u32 frame_type, u8 max_bssid_indicator, > - size_t *bss_index) > + size_t *bss_index, u8 elem_count) > { > struct hostapd_data *tx_bss = hostapd_mbssid_get_tx_bss(hapd); > size_t i; > @@ -7601,6 +7601,9 @@ static u8 * hostapd_eid_mbssid_elem(struct hostapd_data *hapd, u8 *eid, u8 *end, > if (frame_type == WLAN_FC_STYPE_BEACON) { > *eid++ = 3; > *eid++ = i; > + if (hapd->iconf->ema && > + (conf->dtim_period % elem_count)) > + conf->dtim_period = elem_count; > *eid++ = conf->dtim_period; > *eid++ = 0xFF; > } else { > @@ -7686,7 +7689,7 @@ u8 * hostapd_eid_mbssid(struct hostapd_data *hapd, u8 *eid, u8 *end, > } > eid = hostapd_eid_mbssid_elem(hapd, eid, end, frame_type, > hostapd_max_bssid_indicator(hapd), > - &bss_index); > + &bss_index, elem_count); Is there a reason for having these in a separate patch instead of merging the changes into patch 5/13 that introduced this function in this same patchset?
On 4/7/2022 1:46 PM, Jouni Malinen wrote: > On Wed, Mar 02, 2022 at 02:26:31PM -0800, Aloka Dixit wrote: >> @@ -7686,7 +7689,7 @@ u8 * hostapd_eid_mbssid(struct hostapd_data *hapd, u8 *eid, u8 *end, >> } >> eid = hostapd_eid_mbssid_elem(hapd, eid, end, frame_type, >> hostapd_max_bssid_indicator(hapd), >> - &bss_index); >> + &bss_index, elem_count); > > Is there a reason for having these in a separate patch instead of > merging the changes into patch 5/13 that introduced this function in > this same patchset? > This part was added by me, a different author than the original function added by John, hence a separate patch.
On Thu, Apr 21, 2022 at 11:16:31AM -0700, Aloka Dixit wrote: > This part was added by me, a different author than the original function > added by John, hence a separate patch. Since neither patch is in the hostap.git repository yet, I'd prefer to see the patchset modified to be clearer and not try to split clearly related functionality into independent patches.
diff --git a/src/ap/ieee802_11.c b/src/ap/ieee802_11.c index 33db7b411a8a..3365b48fa639 100644 --- a/src/ap/ieee802_11.c +++ b/src/ap/ieee802_11.c @@ -7560,7 +7560,7 @@ size_t hostapd_eid_mbssid_len(struct hostapd_data *hapd, u32 frame_type, static u8 * hostapd_eid_mbssid_elem(struct hostapd_data *hapd, u8 *eid, u8 *end, u32 frame_type, u8 max_bssid_indicator, - size_t *bss_index) + size_t *bss_index, u8 elem_count) { struct hostapd_data *tx_bss = hostapd_mbssid_get_tx_bss(hapd); size_t i; @@ -7601,6 +7601,9 @@ static u8 * hostapd_eid_mbssid_elem(struct hostapd_data *hapd, u8 *eid, u8 *end, if (frame_type == WLAN_FC_STYPE_BEACON) { *eid++ = 3; *eid++ = i; + if (hapd->iconf->ema && + (conf->dtim_period % elem_count)) + conf->dtim_period = elem_count; *eid++ = conf->dtim_period; *eid++ = 0xFF; } else { @@ -7686,7 +7689,7 @@ u8 * hostapd_eid_mbssid(struct hostapd_data *hapd, u8 *eid, u8 *end, } eid = hostapd_eid_mbssid_elem(hapd, eid, end, frame_type, hostapd_max_bssid_indicator(hapd), - &bss_index); + &bss_index, elem_count); } return eid; }
Set the DTIM periods of non-transmitted profiles equal to the EMA profile periodicity if those are not a multiple of the latter already as recommended in IEEE P802.11ax/D8.0, October 2020, Multiple BSSID configuration examples, AA.1 Introduction. Signed-off-by: Aloka Dixit <quic_alokad@quicinc.com> --- src/ap/ieee802_11.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-)