From: Sujith Manoharan <sujith@msujith.org>
To: John Linville <linville@tuxdriver.com>
Cc: linux-wireless@vger.kernel.org, ath9k-devel@qca.qualcomm.com
Subject: [PATCH v2 22/23] ath9k: Improve flush() in mcc mode
Date: Fri, 17 Oct 2014 07:40:29 +0530 [thread overview]
Message-ID: <1413511830-5536-23-git-send-email-sujith@msujith.org> (raw)
In-Reply-To: <1413511830-5536-1-git-send-email-sujith@msujith.org>
From: Sujith Manoharan <c_manoha@qca.qualcomm.com>
The flush timeout in MCC mode is very small, since
we are constrained by the time slice for each
channel context, but since only the HW queues are
flushed when switching contexts, it is acceptable.
Since the SW queues are also emptied in the mac80211 flush()
callback, a larger duration is needed. Add an override
argument to __ath9k_flush() and set it when flush()
is called in MCC mode. This allows the driver to
drain both the SW and HW queues.
Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com>
---
drivers/net/wireless/ath/ath9k/ath9k.h | 2 +-
drivers/net/wireless/ath/ath9k/channel.c | 4 ++--
drivers/net/wireless/ath/ath9k/main.c | 28 +++++++++++++++++++++++++---
3 files changed, 28 insertions(+), 6 deletions(-)
diff --git a/drivers/net/wireless/ath/ath9k/ath9k.h b/drivers/net/wireless/ath/ath9k/ath9k.h
index 7ac1b06..85d74ff 100644
--- a/drivers/net/wireless/ath/ath9k/ath9k.h
+++ b/drivers/net/wireless/ath/ath9k/ath9k.h
@@ -719,7 +719,7 @@ void ath_update_survey_nf(struct ath_softc *sc, int channel);
void ath9k_queue_reset(struct ath_softc *sc, enum ath_reset_type type);
void ath_ps_full_sleep(unsigned long data);
void __ath9k_flush(struct ieee80211_hw *hw, u32 queues, bool drop,
- bool sw_pending);
+ bool sw_pending, bool timeout_override);
/**********/
/* BTCOEX */
diff --git a/drivers/net/wireless/ath/ath9k/channel.c b/drivers/net/wireless/ath/ath9k/channel.c
index 25b898e..c7234d5 100644
--- a/drivers/net/wireless/ath/ath9k/channel.c
+++ b/drivers/net/wireless/ath/ath9k/channel.c
@@ -1232,11 +1232,11 @@ void ath_chanctx_set_next(struct ath_softc *sc, bool force)
ath9k_chanctx_stop_queues(sc, sc->cur_chan);
queues_stopped = true;
- __ath9k_flush(sc->hw, ~0, true, false);
+ __ath9k_flush(sc->hw, ~0, true, false, false);
if (ath_chanctx_send_ps_frame(sc, true))
__ath9k_flush(sc->hw, BIT(IEEE80211_AC_VO),
- false, false);
+ false, false, false);
send_ps = true;
spin_lock_bh(&sc->chan_lock);
diff --git a/drivers/net/wireless/ath/ath9k/main.c b/drivers/net/wireless/ath/ath9k/main.c
index a524eeb..961a388 100644
--- a/drivers/net/wireless/ath/ath9k/main.c
+++ b/drivers/net/wireless/ath/ath9k/main.c
@@ -2031,14 +2031,33 @@ static void ath9k_flush(struct ieee80211_hw *hw, struct ieee80211_vif *vif,
u32 queues, bool drop)
{
struct ath_softc *sc = hw->priv;
+ struct ath_common *common = ath9k_hw_common(sc->sc_ah);
+
+ if (ath9k_is_chanctx_enabled()) {
+ if (!test_bit(ATH_OP_MULTI_CHANNEL, &common->op_flags))
+ goto flush;
+ /*
+ * If MCC is active, extend the flush timeout
+ * and wait for the HW/SW queues to become
+ * empty. This needs to be done outside the
+ * sc->mutex lock to allow the channel scheduler
+ * to switch channel contexts.
+ *
+ * The vif queues have been stopped in mac80211,
+ * so there won't be any incoming frames.
+ */
+ __ath9k_flush(hw, queues, drop, true, true);
+ return;
+ }
+flush:
mutex_lock(&sc->mutex);
- __ath9k_flush(hw, queues, drop, true);
+ __ath9k_flush(hw, queues, drop, true, false);
mutex_unlock(&sc->mutex);
}
void __ath9k_flush(struct ieee80211_hw *hw, u32 queues, bool drop,
- bool sw_pending)
+ bool sw_pending, bool timeout_override)
{
struct ath_softc *sc = hw->priv;
struct ath_hw *ah = sc->sc_ah;
@@ -2059,7 +2078,10 @@ void __ath9k_flush(struct ieee80211_hw *hw, u32 queues, bool drop,
}
spin_lock_bh(&sc->chan_lock);
- timeout = sc->cur_chan->flush_timeout;
+ if (timeout_override)
+ timeout = HZ / 5;
+ else
+ timeout = sc->cur_chan->flush_timeout;
spin_unlock_bh(&sc->chan_lock);
ath_dbg(common, CHAN_CTX,
--
2.1.2
next prev parent reply other threads:[~2014-10-17 2:11 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-10-17 2:10 [PATCH v2 00/23] ath9k patches Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 01/23] ath9k: Use sta_state() callback Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 02/23] ath9k: Enable multi-channel properly Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 03/23] ath9k: Process beacons properly Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 04/23] ath9k: Unify reset API Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 05/23] ath9k: Set ATH_OP_HW_RESET before HW reset Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 06/23] ath9k: Disable beacon tasklet during reset Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 07/23] ath9k: Clear NoA schedule properly Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 08/23] ath9k: Use configurable timeout for flush Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 09/23] ath9k: Fix MCC flush timeout Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 10/23] ath9k: Fix offchannel " Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 11/23] ath9k: Check for pending frames properly Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 12/23] ath9k: Send AUTHORIZED event only for station mode Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 13/23] ath9k: Fix address management Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 14/23] ath9k: Add a function to check for an active GO Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 15/23] ath9k: Check for active GO in mgd_prepare_tx() Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 16/23] ath9k: Use a helper function for offchannel NoA Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 17/23] ath9k: Use a helper function to set NoA Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 18/23] ath9k: Use a helper function for bmiss Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 19/23] ath9k: Fix RoC expiration Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 20/23] ath9k: Send oneshot NoA Sujith Manoharan
2014-10-17 2:10 ` [PATCH v2 21/23] ath9k: Fix HW scan abort Sujith Manoharan
2014-10-17 2:10 ` Sujith Manoharan [this message]
2014-10-17 2:10 ` [PATCH v2 23/23] ath9k: Do not start BA when scanning Sujith Manoharan
2014-10-20 9:30 ` Johannes Berg
2014-10-20 9:39 ` Sujith Manoharan
2014-10-20 10:38 ` Johannes Berg
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1413511830-5536-23-git-send-email-sujith@msujith.org \
--to=sujith@msujith.org \
--cc=ath9k-devel@qca.qualcomm.com \
--cc=linux-wireless@vger.kernel.org \
--cc=linville@tuxdriver.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox