* [PATCH AUTOSEL 4.9 83/87] wlcore: Fix memory leak in case wl12xx_fetch_firmware failure
From: Sasha Levin @ 2019-03-27 18:20 UTC (permalink / raw)
To: linux-kernel, stable
Cc: Zumeng Chen, Kalle Valo, Sasha Levin, linux-wireless, netdev
In-Reply-To: <20190327182040.17444-1-sashal@kernel.org>
From: Zumeng Chen <zumeng.chen@gmail.com>
[ Upstream commit ba2ffc96321c8433606ceeb85c9e722b8113e5a7 ]
Release fw_status, raw_fw_status, and tx_res_if when wl12xx_fetch_firmware
failed instead of meaningless goto out to avoid the following memory leak
reports(Only the last one listed):
unreferenced object 0xc28a9a00 (size 512):
comm "kworker/0:4", pid 31298, jiffies 2783204 (age 203.290s)
hex dump (first 32 bytes):
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
backtrace:
[<6624adab>] kmemleak_alloc+0x40/0x74
[<500ddb31>] kmem_cache_alloc_trace+0x1ac/0x270
[<db4d731d>] wl12xx_chip_wakeup+0xc4/0x1fc [wlcore]
[<76c5db53>] wl1271_op_add_interface+0x4a4/0x8f4 [wlcore]
[<cbf30777>] drv_add_interface+0xa4/0x1a0 [mac80211]
[<65bac325>] ieee80211_reconfig+0x9c0/0x1644 [mac80211]
[<2817c80e>] ieee80211_restart_work+0x90/0xc8 [mac80211]
[<7e1d425a>] process_one_work+0x284/0x42c
[<55f9432e>] worker_thread+0x2fc/0x48c
[<abb582c6>] kthread+0x148/0x160
[<63144b13>] ret_from_fork+0x14/0x2c
[< (null)>] (null)
[<1f6e7715>] 0xffffffff
Signed-off-by: Zumeng Chen <zumeng.chen@gmail.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
drivers/net/wireless/ti/wlcore/main.c | 7 +++++--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/drivers/net/wireless/ti/wlcore/main.c b/drivers/net/wireless/ti/wlcore/main.c
index 5438975c7ff2..17d32ce5d16b 100644
--- a/drivers/net/wireless/ti/wlcore/main.c
+++ b/drivers/net/wireless/ti/wlcore/main.c
@@ -1058,8 +1058,11 @@ static int wl12xx_chip_wakeup(struct wl1271 *wl, bool plt)
goto out;
ret = wl12xx_fetch_firmware(wl, plt);
- if (ret < 0)
- goto out;
+ if (ret < 0) {
+ kfree(wl->fw_status);
+ kfree(wl->raw_fw_status);
+ kfree(wl->tx_res_if);
+ }
out:
return ret;
--
2.19.1
^ permalink raw reply related
* [PATCH] cfg80211: remove redundant zero check on variable 'changed'
From: Colin King @ 2019-03-27 18:35 UTC (permalink / raw)
To: Johannes Berg, David S . Miller, linux-wireless, netdev
Cc: kernel-janitors, linux-kernel
From: Colin Ian King <colin.king@canonical.com>
The zero check on variable changed is redundant as it must be
between 1 and 3 at the end of the proceeding if statement block.
Remove the redundant check.
Signed-off-by: Colin Ian King <colin.king@canonical.com>
---
net/wireless/wext-compat.c | 3 ---
1 file changed, 3 deletions(-)
diff --git a/net/wireless/wext-compat.c b/net/wireless/wext-compat.c
index d522787c7354..46e4d69db845 100644
--- a/net/wireless/wext-compat.c
+++ b/net/wireless/wext-compat.c
@@ -353,9 +353,6 @@ static int cfg80211_wext_siwretry(struct net_device *dev,
changed |= WIPHY_PARAM_RETRY_SHORT;
}
- if (!changed)
- return 0;
-
err = rdev_set_wiphy_params(rdev, changed);
if (err) {
wdev->wiphy->retry_short = oshort;
--
2.20.1
^ permalink raw reply related
* [PATCH] wireless-regdb: Update regulatory rules for South Korea
From: Peter Oh @ 2019-03-27 18:34 UTC (permalink / raw)
To: wireless-regdb@lists.infradead.org
Cc: Peter Oh, linux-wireless@vger.kernel.org
From: Peter Oh <peter.oh@bowerswilkins.com>
Update power limit as documented in:
https://www.rra.go.kr/FileDownSvl?file_type=LAWKR&file_parentseq=54&file_seq=27
which revised on November 30, 2016.
Signed-off-by: Peter Oh <peter.oh@bowerswilkins.com>
---
db.txt | 11 ++++++-----
1 file changed, 6 insertions(+), 5 deletions(-)
diff --git a/db.txt b/db.txt
index 4fb1948..52c17e9 100644
--- a/db.txt
+++ b/db.txt
@@ -695,11 +695,12 @@ country KP: DFS-JP
(5735 - 5815 @ 20), (30)
country KR: DFS-JP
- (2402 - 2482 @ 40), (13)
- (5170 - 5250 @ 80), (20), AUTO-BW
- (5250 - 5330 @ 80), (20), DFS, AUTO-BW
- (5490 - 5710 @ 160), (30), DFS
- (5735 - 5835 @ 80), (30)
+ # ref: https://www.rra.go.kr/ko/reference/lawList_view.do?lw_seq=54&lw_type=3&searchCon=&lw_select=ALL&lw_model=&searchTxt=
+ (2402 - 2482 @ 40), (23)
+ (5170 - 5250 @ 80), (17), AUTO-BW
+ (5250 - 5330 @ 80), (23), DFS, AUTO-BW
+ (5490 - 5710 @ 160), (23), DFS
+ (5735 - 5835 @ 80), (23)
# 60 GHz band channels 1-4,
# ref: http://www.law.go.kr/%ED%96%89%EC%A0%95%EA%B7%9C%EC%B9%99/%EB%AC%B4%EC%84%A0%EC%84%A4%EB%B9%84%EA%B7%9C%EC%B9%99
(57000 - 66000 @ 2160), (43)
--
2.7.4
^ permalink raw reply related
* [PATCH AUTOSEL 4.4 48/63] mt7601u: bump supported EEPROM version
From: Sasha Levin @ 2019-03-27 18:23 UTC (permalink / raw)
To: linux-kernel, stable
Cc: Stanislaw Gruszka, Kalle Valo, Sasha Levin, linux-wireless,
netdev
In-Reply-To: <20190327182323.18577-1-sashal@kernel.org>
From: Stanislaw Gruszka <sgruszka@redhat.com>
[ Upstream commit 3bd1505fed71d834f45e87b32ff07157fdda47e0 ]
As reported by Michael eeprom 0d is supported and work with the driver.
Dump of /sys/kernel/debug/ieee80211/phy1/mt7601u/eeprom_param
with 0d EEPORM looks like this:
RSSI offset: 0 0
Reference temp: f9
LNA gain: 8
Reg channels: 1-14
Per rate power:
raw:05 bw20:05 bw40:05
raw:05 bw20:05 bw40:05
raw:03 bw20:03 bw40:03
raw:03 bw20:03 bw40:03
raw:04 bw20:04 bw40:04
raw:00 bw20:00 bw40:00
raw:00 bw20:00 bw40:00
raw:00 bw20:00 bw40:00
raw:02 bw20:02 bw40:02
raw:00 bw20:00 bw40:00
Per channel power:
tx_power ch1:09 ch2:09
tx_power ch3:0a ch4:0a
tx_power ch5:0a ch6:0a
tx_power ch7:0b ch8:0b
tx_power ch9:0b ch10:0b
tx_power ch11:0b ch12:0b
tx_power ch13:0b ch14:0b
Reported-and-tested-by: Michael <ZeroBeat@gmx.de>
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
Acked-by: Jakub Kicinski <kubakici@wp.pl>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
drivers/net/wireless/mediatek/mt7601u/eeprom.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/wireless/mediatek/mt7601u/eeprom.h b/drivers/net/wireless/mediatek/mt7601u/eeprom.h
index 662d12703b69..57b503ae63f1 100644
--- a/drivers/net/wireless/mediatek/mt7601u/eeprom.h
+++ b/drivers/net/wireless/mediatek/mt7601u/eeprom.h
@@ -17,7 +17,7 @@
struct mt7601u_dev;
-#define MT7601U_EE_MAX_VER 0x0c
+#define MT7601U_EE_MAX_VER 0x0d
#define MT7601U_EEPROM_SIZE 256
#define MT7601U_DEFAULT_TX_POWER 6
--
2.19.1
^ permalink raw reply related
* [PATCH AUTOSEL 3.18 38/41] wlcore: Fix memory leak in case wl12xx_fetch_firmware failure
From: Sasha Levin @ 2019-03-27 18:25 UTC (permalink / raw)
To: linux-kernel, stable
Cc: Zumeng Chen, Kalle Valo, Sasha Levin, linux-wireless, netdev
In-Reply-To: <20190327182518.19394-1-sashal@kernel.org>
From: Zumeng Chen <zumeng.chen@gmail.com>
[ Upstream commit ba2ffc96321c8433606ceeb85c9e722b8113e5a7 ]
Release fw_status, raw_fw_status, and tx_res_if when wl12xx_fetch_firmware
failed instead of meaningless goto out to avoid the following memory leak
reports(Only the last one listed):
unreferenced object 0xc28a9a00 (size 512):
comm "kworker/0:4", pid 31298, jiffies 2783204 (age 203.290s)
hex dump (first 32 bytes):
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
backtrace:
[<6624adab>] kmemleak_alloc+0x40/0x74
[<500ddb31>] kmem_cache_alloc_trace+0x1ac/0x270
[<db4d731d>] wl12xx_chip_wakeup+0xc4/0x1fc [wlcore]
[<76c5db53>] wl1271_op_add_interface+0x4a4/0x8f4 [wlcore]
[<cbf30777>] drv_add_interface+0xa4/0x1a0 [mac80211]
[<65bac325>] ieee80211_reconfig+0x9c0/0x1644 [mac80211]
[<2817c80e>] ieee80211_restart_work+0x90/0xc8 [mac80211]
[<7e1d425a>] process_one_work+0x284/0x42c
[<55f9432e>] worker_thread+0x2fc/0x48c
[<abb582c6>] kthread+0x148/0x160
[<63144b13>] ret_from_fork+0x14/0x2c
[< (null)>] (null)
[<1f6e7715>] 0xffffffff
Signed-off-by: Zumeng Chen <zumeng.chen@gmail.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
drivers/net/wireless/ti/wlcore/main.c | 7 +++++--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/drivers/net/wireless/ti/wlcore/main.c b/drivers/net/wireless/ti/wlcore/main.c
index 575c8f6d4009..16b69b026f8f 100644
--- a/drivers/net/wireless/ti/wlcore/main.c
+++ b/drivers/net/wireless/ti/wlcore/main.c
@@ -1110,8 +1110,11 @@ static int wl12xx_chip_wakeup(struct wl1271 *wl, bool plt)
goto out;
ret = wl12xx_fetch_firmware(wl, plt);
- if (ret < 0)
- goto out;
+ if (ret < 0) {
+ kfree(wl->fw_status);
+ kfree(wl->raw_fw_status);
+ kfree(wl->tx_res_if);
+ }
out:
return ret;
--
2.19.1
^ permalink raw reply related
* [PATCH AUTOSEL 4.4 60/63] wlcore: Fix memory leak in case wl12xx_fetch_firmware failure
From: Sasha Levin @ 2019-03-27 18:23 UTC (permalink / raw)
To: linux-kernel, stable
Cc: Zumeng Chen, Kalle Valo, Sasha Levin, linux-wireless, netdev
In-Reply-To: <20190327182323.18577-1-sashal@kernel.org>
From: Zumeng Chen <zumeng.chen@gmail.com>
[ Upstream commit ba2ffc96321c8433606ceeb85c9e722b8113e5a7 ]
Release fw_status, raw_fw_status, and tx_res_if when wl12xx_fetch_firmware
failed instead of meaningless goto out to avoid the following memory leak
reports(Only the last one listed):
unreferenced object 0xc28a9a00 (size 512):
comm "kworker/0:4", pid 31298, jiffies 2783204 (age 203.290s)
hex dump (first 32 bytes):
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
backtrace:
[<6624adab>] kmemleak_alloc+0x40/0x74
[<500ddb31>] kmem_cache_alloc_trace+0x1ac/0x270
[<db4d731d>] wl12xx_chip_wakeup+0xc4/0x1fc [wlcore]
[<76c5db53>] wl1271_op_add_interface+0x4a4/0x8f4 [wlcore]
[<cbf30777>] drv_add_interface+0xa4/0x1a0 [mac80211]
[<65bac325>] ieee80211_reconfig+0x9c0/0x1644 [mac80211]
[<2817c80e>] ieee80211_restart_work+0x90/0xc8 [mac80211]
[<7e1d425a>] process_one_work+0x284/0x42c
[<55f9432e>] worker_thread+0x2fc/0x48c
[<abb582c6>] kthread+0x148/0x160
[<63144b13>] ret_from_fork+0x14/0x2c
[< (null)>] (null)
[<1f6e7715>] 0xffffffff
Signed-off-by: Zumeng Chen <zumeng.chen@gmail.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
drivers/net/wireless/ti/wlcore/main.c | 7 +++++--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/drivers/net/wireless/ti/wlcore/main.c b/drivers/net/wireless/ti/wlcore/main.c
index 7b27c7e23af2..cc10b72607c6 100644
--- a/drivers/net/wireless/ti/wlcore/main.c
+++ b/drivers/net/wireless/ti/wlcore/main.c
@@ -1123,8 +1123,11 @@ static int wl12xx_chip_wakeup(struct wl1271 *wl, bool plt)
goto out;
ret = wl12xx_fetch_firmware(wl, plt);
- if (ret < 0)
- goto out;
+ if (ret < 0) {
+ kfree(wl->fw_status);
+ kfree(wl->raw_fw_status);
+ kfree(wl->tx_res_if);
+ }
out:
return ret;
--
2.19.1
^ permalink raw reply related
* [PATCH AUTOSEL 4.19 110/192] ath10k: fix shadow register implementation for WCN3990
From: Sasha Levin @ 2019-03-27 18:09 UTC (permalink / raw)
To: linux-kernel, stable
Cc: Rakesh Pillai, Kalle Valo, Sasha Levin, ath10k, linux-wireless,
netdev
In-Reply-To: <20190327181025.13507-1-sashal@kernel.org>
From: Rakesh Pillai <pillair@codeaurora.org>
[ Upstream commit 1863008369ae0407508033b4b00f98b985adeb15 ]
WCN3990 supports shadow registers write operation support
for copy engine for regular operation in powersave mode.
Since WCN3990 is a 64-bit target, the shadow register
implementation needs to be done in the copy engine handlers
for 64-bit target. Currently the shadow register implementation
is present in the 32-bit target handlers of copy engine.
Fix the shadow register copy engine write operation
implementation for 64-bit target(WCN3990).
Tested HW: WCN3990
Tested FW: WLAN.HL.2.0-01188-QCAHLSWMTPLZ-1
Fixes: b7ba83f7c414 ("ath10k: add support for shadow register for WNC3990")
Signed-off-by: Rakesh Pillai <pillair@codeaurora.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
drivers/net/wireless/ath/ath10k/ce.c | 26 +++++++++++++-------------
drivers/net/wireless/ath/ath10k/ce.h | 2 +-
2 files changed, 14 insertions(+), 14 deletions(-)
diff --git a/drivers/net/wireless/ath/ath10k/ce.c b/drivers/net/wireless/ath/ath10k/ce.c
index 18c709c484e7..f761d651c16e 100644
--- a/drivers/net/wireless/ath/ath10k/ce.c
+++ b/drivers/net/wireless/ath/ath10k/ce.c
@@ -500,14 +500,8 @@ static int _ath10k_ce_send_nolock(struct ath10k_ce_pipe *ce_state,
write_index = CE_RING_IDX_INCR(nentries_mask, write_index);
/* WORKAROUND */
- if (!(flags & CE_SEND_FLAG_GATHER)) {
- if (ar->hw_params.shadow_reg_support)
- ath10k_ce_shadow_src_ring_write_index_set(ar, ce_state,
- write_index);
- else
- ath10k_ce_src_ring_write_index_set(ar, ctrl_addr,
- write_index);
- }
+ if (!(flags & CE_SEND_FLAG_GATHER))
+ ath10k_ce_src_ring_write_index_set(ar, ctrl_addr, write_index);
src_ring->write_index = write_index;
exit:
@@ -581,8 +575,14 @@ static int _ath10k_ce_send_nolock_64(struct ath10k_ce_pipe *ce_state,
/* Update Source Ring Write Index */
write_index = CE_RING_IDX_INCR(nentries_mask, write_index);
- if (!(flags & CE_SEND_FLAG_GATHER))
- ath10k_ce_src_ring_write_index_set(ar, ctrl_addr, write_index);
+ if (!(flags & CE_SEND_FLAG_GATHER)) {
+ if (ar->hw_params.shadow_reg_support)
+ ath10k_ce_shadow_src_ring_write_index_set(ar, ce_state,
+ write_index);
+ else
+ ath10k_ce_src_ring_write_index_set(ar, ctrl_addr,
+ write_index);
+ }
src_ring->write_index = write_index;
exit:
@@ -1394,12 +1394,12 @@ static int ath10k_ce_alloc_shadow_base(struct ath10k *ar,
u32 nentries)
{
src_ring->shadow_base_unaligned = kcalloc(nentries,
- sizeof(struct ce_desc),
+ sizeof(struct ce_desc_64),
GFP_KERNEL);
if (!src_ring->shadow_base_unaligned)
return -ENOMEM;
- src_ring->shadow_base = (struct ce_desc *)
+ src_ring->shadow_base = (struct ce_desc_64 *)
PTR_ALIGN(src_ring->shadow_base_unaligned,
CE_DESC_RING_ALIGN);
return 0;
@@ -1453,7 +1453,7 @@ ath10k_ce_alloc_src_ring(struct ath10k *ar, unsigned int ce_id,
ret = ath10k_ce_alloc_shadow_base(ar, src_ring, nentries);
if (ret) {
dma_free_coherent(ar->dev,
- (nentries * sizeof(struct ce_desc) +
+ (nentries * sizeof(struct ce_desc_64) +
CE_DESC_RING_ALIGN),
src_ring->base_addr_owner_space_unaligned,
base_addr);
diff --git a/drivers/net/wireless/ath/ath10k/ce.h b/drivers/net/wireless/ath/ath10k/ce.h
index b8fb5382dede..8088f7a66426 100644
--- a/drivers/net/wireless/ath/ath10k/ce.h
+++ b/drivers/net/wireless/ath/ath10k/ce.h
@@ -118,7 +118,7 @@ struct ath10k_ce_ring {
u32 base_addr_ce_space;
char *shadow_base_unaligned;
- struct ce_desc *shadow_base;
+ struct ce_desc_64 *shadow_base;
/* keep last */
void *per_transfer_context[0];
--
2.19.1
^ permalink raw reply related
* [PATCH AUTOSEL 4.19 100/192] iwlwifi: pcie: fix emergency path
From: Sasha Levin @ 2019-03-27 18:08 UTC (permalink / raw)
To: linux-kernel, stable
Cc: Sara Sharon, Luca Coelho, Sasha Levin, linux-wireless, netdev
In-Reply-To: <20190327181025.13507-1-sashal@kernel.org>
From: Sara Sharon <sara.sharon@intel.com>
[ Upstream commit c6ac9f9fb98851f47b978a9476594fc3c477a34d ]
Allocator swaps the pending requests with 0 when it starts
working. This means that relying on it n RX path to decide if
to move to emergency is not always a good idea, since it may
be zero, but there are still a lot of unallocated RBs in the
system. Change allocator to decrement the pending requests on
real time. It is more expensive since it accesses the atomic
variable more times, but it gives the RX path a better idea
of the system's status.
Reported-by: Ilan Peer <ilan.peer@intel.com>
Signed-off-by: Sara Sharon <sara.sharon@intel.com>
Fixes: 868a1e863f95 ("iwlwifi: pcie: avoid empty free RB queue")
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
drivers/net/wireless/intel/iwlwifi/pcie/rx.c | 11 ++++++++---
1 file changed, 8 insertions(+), 3 deletions(-)
diff --git a/drivers/net/wireless/intel/iwlwifi/pcie/rx.c b/drivers/net/wireless/intel/iwlwifi/pcie/rx.c
index d4a31e014c82..b2905f01b7df 100644
--- a/drivers/net/wireless/intel/iwlwifi/pcie/rx.c
+++ b/drivers/net/wireless/intel/iwlwifi/pcie/rx.c
@@ -502,7 +502,7 @@ static void iwl_pcie_rx_allocator(struct iwl_trans *trans)
struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
struct iwl_rb_allocator *rba = &trans_pcie->rba;
struct list_head local_empty;
- int pending = atomic_xchg(&rba->req_pending, 0);
+ int pending = atomic_read(&rba->req_pending);
IWL_DEBUG_RX(trans, "Pending allocation requests = %d\n", pending);
@@ -557,11 +557,13 @@ static void iwl_pcie_rx_allocator(struct iwl_trans *trans)
i++;
}
+ atomic_dec(&rba->req_pending);
pending--;
+
if (!pending) {
- pending = atomic_xchg(&rba->req_pending, 0);
+ pending = atomic_read(&rba->req_pending);
IWL_DEBUG_RX(trans,
- "Pending allocation requests = %d\n",
+ "Got more pending allocation requests = %d\n",
pending);
}
@@ -573,12 +575,15 @@ static void iwl_pcie_rx_allocator(struct iwl_trans *trans)
spin_unlock(&rba->lock);
atomic_inc(&rba->req_ready);
+
}
spin_lock(&rba->lock);
/* return unused rbds to the allocator empty list */
list_splice_tail(&local_empty, &rba->rbd_empty);
spin_unlock(&rba->lock);
+
+ IWL_DEBUG_RX(trans, "%s, exit.\n", __func__);
}
/*
--
2.19.1
^ permalink raw reply related
* [PATCH AUTOSEL 5.0 116/262] mwifiex: don't advertise IBSS features without FW support
From: Sasha Levin @ 2019-03-27 17:59 UTC (permalink / raw)
To: linux-kernel, stable
Cc: Brian Norris, Kalle Valo, Sasha Levin, linux-wireless, netdev
In-Reply-To: <20190327180158.10245-1-sashal@kernel.org>
From: Brian Norris <briannorris@chromium.org>
[ Upstream commit 6f21ab30469d670de620f758330aca9f3433f693 ]
As it is, doing something like
# iw phy phy0 interface add foobar type ibss
on a firmware that doesn't have ad-hoc support just yields failures of
HostCmd_CMD_SET_BSS_MODE, which happened to return a '-1' error code
(-EPERM? not really right...) and sometimes may even crash the firmware
along the way.
Let's parse the firmware capability flag while registering the wiphy, so
we don't allow attempting IBSS at all, and we get a proper -EOPNOTSUPP
from nl80211 instead.
Fixes: e267e71e68ae ("mwifiex: Disable adhoc feature based on firmware capability")
Signed-off-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
drivers/net/wireless/marvell/mwifiex/cfg80211.c | 10 +++++++---
1 file changed, 7 insertions(+), 3 deletions(-)
diff --git a/drivers/net/wireless/marvell/mwifiex/cfg80211.c b/drivers/net/wireless/marvell/mwifiex/cfg80211.c
index 1467af22e394..883752f640b4 100644
--- a/drivers/net/wireless/marvell/mwifiex/cfg80211.c
+++ b/drivers/net/wireless/marvell/mwifiex/cfg80211.c
@@ -4310,11 +4310,13 @@ int mwifiex_register_cfg80211(struct mwifiex_adapter *adapter)
wiphy->mgmt_stypes = mwifiex_mgmt_stypes;
wiphy->max_remain_on_channel_duration = 5000;
wiphy->interface_modes = BIT(NL80211_IFTYPE_STATION) |
- BIT(NL80211_IFTYPE_ADHOC) |
BIT(NL80211_IFTYPE_P2P_CLIENT) |
BIT(NL80211_IFTYPE_P2P_GO) |
BIT(NL80211_IFTYPE_AP);
+ if (ISSUPP_ADHOC_ENABLED(adapter->fw_cap_info))
+ wiphy->interface_modes |= BIT(NL80211_IFTYPE_ADHOC);
+
wiphy->bands[NL80211_BAND_2GHZ] = &mwifiex_band_2ghz;
if (adapter->config_bands & BAND_A)
wiphy->bands[NL80211_BAND_5GHZ] = &mwifiex_band_5ghz;
@@ -4374,11 +4376,13 @@ int mwifiex_register_cfg80211(struct mwifiex_adapter *adapter)
wiphy->available_antennas_tx = BIT(adapter->number_of_antenna) - 1;
wiphy->available_antennas_rx = BIT(adapter->number_of_antenna) - 1;
- wiphy->features |= NL80211_FEATURE_HT_IBSS |
- NL80211_FEATURE_INACTIVITY_TIMER |
+ wiphy->features |= NL80211_FEATURE_INACTIVITY_TIMER |
NL80211_FEATURE_LOW_PRIORITY_SCAN |
NL80211_FEATURE_NEED_OBSS_SCAN;
+ if (ISSUPP_ADHOC_ENABLED(adapter->fw_cap_info))
+ wiphy->features |= NL80211_FEATURE_HT_IBSS;
+
if (ISSUPP_RANDOM_MAC(adapter->fw_cap_info))
wiphy->features |= NL80211_FEATURE_SCAN_RANDOM_MAC_ADDR |
NL80211_FEATURE_SCHED_SCAN_RANDOM_MAC_ADDR |
--
2.19.1
^ permalink raw reply related
* Re: [PATCH] ath10k: remove iteration in wake_tx_queue
From: Peter Oh @ 2019-03-27 17:49 UTC (permalink / raw)
To: Erik Stromdahl, kvalo@qca.qualcomm.com,
linux-wireless@vger.kernel.org, ath10k@lists.infradead.org
Cc: yiboz@codeaurora.org
In-Reply-To: <20190327162906.6010-1-erik.stromdahl@gmail.com>
On 03/27/2019 09:29 AM, Erik Stromdahl wrote:
> Iterating the TX queue and thereby dequeuing all available packets in the
> queue could result in performance penalties on some SMP systems.
>
Please share the test results and numbers you've run to help others
thoughts.
Thanks,
Peter
^ permalink raw reply
* [PATCH] ath10k: remove iteration in wake_tx_queue
From: Erik Stromdahl @ 2019-03-27 16:29 UTC (permalink / raw)
To: kvalo, linux-wireless, ath10k; +Cc: yiboz, Erik Stromdahl
Iterating the TX queue and thereby dequeuing all available packets in the
queue could result in performance penalties on some SMP systems.
The reason for this is most likely that the per-ac lock (active_txq_lock)
in mac80211 will be held by the CPU iterating the current queue.
This will lock up other CPUs trying to push new messages on the TX
queue.
Instead of iterating the queue we fetch just one packet at the time,
resulting in minimal starvation of the other CPUs.
Reported-by: Yibo Zhao <yiboz@codeaurora.org>
Signed-off-by: Erik Stromdahl <erik.stromdahl@gmail.com>
---
drivers/net/wireless/ath/ath10k/mac.c | 9 +++------
1 file changed, 3 insertions(+), 6 deletions(-)
diff --git a/drivers/net/wireless/ath/ath10k/mac.c b/drivers/net/wireless/ath/ath10k/mac.c
index b73c23d4ce86..c9e700b844f8 100644
--- a/drivers/net/wireless/ath/ath10k/mac.c
+++ b/drivers/net/wireless/ath/ath10k/mac.c
@@ -4356,7 +4356,6 @@ static void ath10k_mac_op_wake_tx_queue(struct ieee80211_hw *hw,
struct ieee80211_txq *txq)
{
struct ath10k *ar = hw->priv;
- int ret;
u8 ac;
ath10k_htt_tx_txq_update(hw, txq);
@@ -4369,11 +4368,9 @@ static void ath10k_mac_op_wake_tx_queue(struct ieee80211_hw *hw,
if (!txq)
goto out;
- while (ath10k_mac_tx_can_push(hw, txq)) {
- ret = ath10k_mac_tx_push_txq(hw, txq);
- if (ret < 0)
- break;
- }
+ if (ath10k_mac_tx_can_push(hw, txq))
+ ath10k_mac_tx_push_txq(hw, txq);
+
ieee80211_return_txq(hw, txq);
ath10k_htt_tx_txq_update(hw, txq);
out:
--
2.19.1
^ permalink raw reply related
* [PATCH] mt76: introduce mt76_free_device routine
From: Lorenzo Bianconi @ 2019-03-27 11:41 UTC (permalink / raw)
To: nbd; +Cc: linux-wireless, lorenzo.bianconi, sgruszka
In-Reply-To: <cover.1553682658.git.lorenzo@kernel.org>
Move mt76_tx_free in mt76_free_device routine in order to
unmap all txwi descriptors at module unload
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
---
drivers/net/wireless/mediatek/mt76/mac80211.c | 8 +++++++-
drivers/net/wireless/mediatek/mt76/mt76.h | 1 +
drivers/net/wireless/mediatek/mt76/mt7603/init.c | 2 +-
drivers/net/wireless/mediatek/mt76/mt76x0/pci.c | 2 +-
drivers/net/wireless/mediatek/mt76/mt76x2/pci.c | 2 +-
5 files changed, 11 insertions(+), 4 deletions(-)
diff --git a/drivers/net/wireless/mediatek/mt76/mac80211.c b/drivers/net/wireless/mediatek/mt76/mac80211.c
index 15825e9a458f..f0d418b751ad 100644
--- a/drivers/net/wireless/mediatek/mt76/mac80211.c
+++ b/drivers/net/wireless/mediatek/mt76/mac80211.c
@@ -369,10 +369,16 @@ void mt76_unregister_device(struct mt76_dev *dev)
mt76_tx_status_check(dev, NULL, true);
ieee80211_unregister_hw(hw);
- mt76_tx_free(dev);
}
EXPORT_SYMBOL_GPL(mt76_unregister_device);
+void mt76_free_device(struct mt76_dev *dev)
+{
+ mt76_tx_free(dev);
+ ieee80211_free_hw(dev->hw);
+}
+EXPORT_SYMBOL_GPL(mt76_free_device);
+
void mt76_rx(struct mt76_dev *dev, enum mt76_rxq_id q, struct sk_buff *skb)
{
if (!test_bit(MT76_STATE_RUNNING, &dev->state)) {
diff --git a/drivers/net/wireless/mediatek/mt76/mt76.h b/drivers/net/wireless/mediatek/mt76/mt76.h
index b432da3f55c7..f0d34901c825 100644
--- a/drivers/net/wireless/mediatek/mt76/mt76.h
+++ b/drivers/net/wireless/mediatek/mt76/mt76.h
@@ -605,6 +605,7 @@ struct mt76_dev *mt76_alloc_device(struct device *pdev, unsigned int size,
int mt76_register_device(struct mt76_dev *dev, bool vht,
struct ieee80211_rate *rates, int n_rates);
void mt76_unregister_device(struct mt76_dev *dev);
+void mt76_free_device(struct mt76_dev *dev);
struct dentry *mt76_register_debugfs(struct mt76_dev *dev);
void mt76_seq_puts_array(struct seq_file *file, const char *str,
diff --git a/drivers/net/wireless/mediatek/mt76/mt7603/init.c b/drivers/net/wireless/mediatek/mt76/mt7603/init.c
index e17b6b89dfe2..a84b239887ca 100644
--- a/drivers/net/wireless/mediatek/mt76/mt7603/init.c
+++ b/drivers/net/wireless/mediatek/mt76/mt7603/init.c
@@ -574,5 +574,5 @@ void mt7603_unregister_device(struct mt7603_dev *dev)
mt76_unregister_device(&dev->mt76);
mt7603_mcu_exit(dev);
mt7603_dma_cleanup(dev);
- ieee80211_free_hw(mt76_hw(dev));
+ mt76_free_device(&dev->mt76);
}
diff --git a/drivers/net/wireless/mediatek/mt76/mt76x0/pci.c b/drivers/net/wireless/mediatek/mt76/mt76x0/pci.c
index a7ff7e32b9b3..156d3d064ba0 100644
--- a/drivers/net/wireless/mediatek/mt76/mt76x0/pci.c
+++ b/drivers/net/wireless/mediatek/mt76/mt76x0/pci.c
@@ -234,7 +234,7 @@ mt76x0e_remove(struct pci_dev *pdev)
mt76_unregister_device(mdev);
mt76x0e_cleanup(dev);
- ieee80211_free_hw(mdev->hw);
+ mt76_free_device(mdev);
}
static const struct pci_device_id mt76x0e_device_table[] = {
diff --git a/drivers/net/wireless/mediatek/mt76/mt76x2/pci.c b/drivers/net/wireless/mediatek/mt76/mt76x2/pci.c
index 4747f782417a..e84d5c5911ea 100644
--- a/drivers/net/wireless/mediatek/mt76/mt76x2/pci.c
+++ b/drivers/net/wireless/mediatek/mt76/mt76x2/pci.c
@@ -107,7 +107,7 @@ mt76pci_remove(struct pci_dev *pdev)
mt76_unregister_device(mdev);
mt76x2_cleanup(dev);
- ieee80211_free_hw(mdev->hw);
+ mt76_free_device(mdev);
}
MODULE_DEVICE_TABLE(pci, mt76pci_device_table);
--
2.20.1
^ permalink raw reply related
* Re: [PATCH v8 00/14] rtw88: mac80211 driver for Realtek 802.11ac wireless network chips
From: Stanislaw Gruszka @ 2019-03-27 10:26 UTC (permalink / raw)
To: Tony Chuang
Cc: Brian Norris, Kalle Valo, Johannes Berg, linux-wireless,
Greg Kroah-Hartman, Pkshih, Andy Huang, Larry Finger
In-Reply-To: <F7CD281DE3E379468C6D07993EA72F84D17CB940@RTITMBSVM04.realtek.com.tw>
On Wed, Mar 27, 2019 at 04:59:44AM +0000, Tony Chuang wrote:
> > > I also tested on RTL8822CE we generously got from Tony. I replaced wifi
> > > card on my laptop by the 8822CE device. I have warning on resume from
> > > suspend (see below, it could be mac80211 issue though). But network
> > > connection was established after the warning and in general device
> > > works stable for 2 days now.
> > >
> > > Tested-by: Stanislaw Gruszka <sgruszka@redhat.com>
> > >
> > > [16158.826873] rtw_pci 0000:02:00.0: start vif 00:e0:4c:09:94:0f on port 0
> > > [16158.827693] ------------[ cut here ]------------
> > > [16158.827694] wlan0: invalid CW_min/CW_max: 0/0
> >
> > I didn't test system suspend/resume with this driver yet (oops), but
> > FWIW I've seen reports of this warning on every other mac80211 driver
> > I've supported. I don't think we've ever totally resolved them and I
> > haven't looked closely into the reason for the WARNING.
It looks as mac80211 issue for me as well.
> And we haven't submit patches related to suspend yet.
> Like something in CONFIG_PM.
> Should be added later.
Suspend/resume works for me anyway :-)
Stanislaw
^ permalink raw reply
* [PATCH 5.1] rt2x00: do not increment sequence number while re-transmitting
From: Stanislaw Gruszka @ 2019-03-27 10:03 UTC (permalink / raw)
To: linux-wireless
Cc: Tomislav Požega, Daniel Golle, Felix Fietkau, Mathias Kresin,
Vijayakumar Durai, stable, Stanislaw Gruszka
From: Vijayakumar Durai <vijayakumar.durai1@vivint.com>
Currently rt2x00 devices retransmit the management frames with
incremented sequence number if hardware is assigning the sequence.
This is HW bug fixed already for non-QOS data frames, but it should
be fixed for management frames except beacon.
Without fix retransmitted frames have wrong SN:
AlphaNet_e8:fb:36 Vivotek_52:31:51 Authentication, SN=1648, FN=0, Flags=........C Frame is not being retransmitted 1648 1
AlphaNet_e8:fb:36 Vivotek_52:31:51 Authentication, SN=1649, FN=0, Flags=....R...C Frame is being retransmitted 1649 1
AlphaNet_e8:fb:36 Vivotek_52:31:51 Authentication, SN=1650, FN=0, Flags=....R...C Frame is being retransmitted 1650 1
With the fix SN stays correctly the same:
88:6a:e3:e8:f9:a2 8c:f5:a3:88:76:87 Authentication, SN=1450, FN=0, Flags=........C
88:6a:e3:e8:f9:a2 8c:f5:a3:88:76:87 Authentication, SN=1450, FN=0, Flags=....R...C
88:6a:e3:e8:f9:a2 8c:f5:a3:88:76:87 Authentication, SN=1450, FN=0, Flags=....R...C
Cc: stable@vger.kernel.org
Signed-off-by: Vijayakumar Durai <vijayakumar.durai1@vivint.com>
[sgruszka: simplify code, change comments and changelog]
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
---
drivers/net/wireless/ralink/rt2x00/rt2x00.h | 1 -
drivers/net/wireless/ralink/rt2x00/rt2x00mac.c | 10 ----------
drivers/net/wireless/ralink/rt2x00/rt2x00queue.c | 15 +++++++++------
3 files changed, 9 insertions(+), 17 deletions(-)
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2x00.h b/drivers/net/wireless/ralink/rt2x00/rt2x00.h
index 4b1744e9fb78..50b92ca92bd7 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2x00.h
+++ b/drivers/net/wireless/ralink/rt2x00/rt2x00.h
@@ -673,7 +673,6 @@ enum rt2x00_state_flags {
CONFIG_CHANNEL_HT40,
CONFIG_POWERSAVING,
CONFIG_HT_DISABLED,
- CONFIG_QOS_DISABLED,
CONFIG_MONITORING,
/*
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2x00mac.c b/drivers/net/wireless/ralink/rt2x00/rt2x00mac.c
index 2825560e2424..e8462f25d252 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2x00mac.c
+++ b/drivers/net/wireless/ralink/rt2x00/rt2x00mac.c
@@ -642,19 +642,9 @@ void rt2x00mac_bss_info_changed(struct ieee80211_hw *hw,
rt2x00dev->intf_associated--;
rt2x00leds_led_assoc(rt2x00dev, !!rt2x00dev->intf_associated);
-
- clear_bit(CONFIG_QOS_DISABLED, &rt2x00dev->flags);
}
/*
- * Check for access point which do not support 802.11e . We have to
- * generate data frames sequence number in S/W for such AP, because
- * of H/W bug.
- */
- if (changes & BSS_CHANGED_QOS && !bss_conf->qos)
- set_bit(CONFIG_QOS_DISABLED, &rt2x00dev->flags);
-
- /*
* When the erp information has changed, we should perform
* additional configuration steps. For all other changes we are done.
*/
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2x00queue.c b/drivers/net/wireless/ralink/rt2x00/rt2x00queue.c
index 92ddc19e7bf7..4834b4eb0206 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2x00queue.c
+++ b/drivers/net/wireless/ralink/rt2x00/rt2x00queue.c
@@ -201,15 +201,18 @@ static void rt2x00queue_create_tx_descriptor_seq(struct rt2x00_dev *rt2x00dev,
if (!rt2x00_has_cap_flag(rt2x00dev, REQUIRE_SW_SEQNO)) {
/*
* rt2800 has a H/W (or F/W) bug, device incorrectly increase
- * seqno on retransmited data (non-QOS) frames. To workaround
- * the problem let's generate seqno in software if QOS is
- * disabled.
+ * seqno on retransmitted data (non-QOS) and management frames.
+ * To workaround the problem let's generate seqno in software.
+ * Except for beacons which are transmitted periodically by H/W
+ * hence hardware has to assign seqno for them.
*/
- if (test_bit(CONFIG_QOS_DISABLED, &rt2x00dev->flags))
- __clear_bit(ENTRY_TXD_GENERATE_SEQ, &txdesc->flags);
- else
+ if (ieee80211_is_beacon(hdr->frame_control)) {
+ __set_bit(ENTRY_TXD_GENERATE_SEQ, &txdesc->flags);
/* H/W will generate sequence number */
return;
+ }
+
+ __clear_bit(ENTRY_TXD_GENERATE_SEQ, &txdesc->flags);
}
/*
--
2.7.5
^ permalink raw reply related
* [PATCH 8/8] rt2x00mmio: remove legacy comment
From: Stanislaw Gruszka @ 2019-03-27 9:58 UTC (permalink / raw)
To: linux-wireless
Cc: Tomislav Požega, Daniel Golle, Felix Fietkau, Mathias Kresin
In-Reply-To: <1553680711-15730-1-git-send-email-sgruszka@redhat.com>
Remove comment about fields that were removed.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
---
drivers/net/wireless/ralink/rt2x00/rt2x00mmio.h | 2 --
1 file changed, 2 deletions(-)
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2x00mmio.h b/drivers/net/wireless/ralink/rt2x00/rt2x00mmio.h
index 184a4148b2f8..03e6cdb0b5a4 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2x00mmio.h
+++ b/drivers/net/wireless/ralink/rt2x00/rt2x00mmio.h
@@ -80,8 +80,6 @@ int rt2x00mmio_regbusy_read(struct rt2x00_dev *rt2x00dev,
*
* @desc: Pointer to device descriptor
* @desc_dma: DMA pointer to &desc.
- * @data: Pointer to device's entry memory.
- * @data_dma: DMA pointer to &data.
*/
struct queue_entry_priv_mmio {
__le32 *desc;
--
2.7.5
^ permalink raw reply related
* [PATCH 7/8] rt2x00: remove not used entry field
From: Stanislaw Gruszka @ 2019-03-27 9:58 UTC (permalink / raw)
To: linux-wireless
Cc: Tomislav Požega, Daniel Golle, Felix Fietkau, Mathias Kresin
In-Reply-To: <1553680711-15730-1-git-send-email-sgruszka@redhat.com>
Remove not used any longer queue_entry field and flag.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
---
drivers/net/wireless/ralink/rt2x00/rt2x00queue.h | 3 ---
1 file changed, 3 deletions(-)
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2x00queue.h b/drivers/net/wireless/ralink/rt2x00/rt2x00queue.h
index a15bae29917b..20113f861b9e 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2x00queue.h
+++ b/drivers/net/wireless/ralink/rt2x00/rt2x00queue.h
@@ -361,7 +361,6 @@ enum queue_entry_flags {
ENTRY_DATA_PENDING,
ENTRY_DATA_IO_FAILED,
ENTRY_DATA_STATUS_PENDING,
- ENTRY_DATA_STATUS_SET,
};
/**
@@ -387,8 +386,6 @@ struct queue_entry {
unsigned int entry_idx;
- u32 status;
-
void *priv_data;
};
--
2.7.5
^ permalink raw reply related
* [PATCH 6/8] rt2x00: remove last_nostatus_check
From: Stanislaw Gruszka @ 2019-03-27 9:58 UTC (permalink / raw)
To: linux-wireless
Cc: Tomislav Požega, Daniel Golle, Felix Fietkau, Mathias Kresin
In-Reply-To: <1553680711-15730-1-git-send-email-sgruszka@redhat.com>
We do not any longer check txstatus timeout from tasklet, so do not need
this optimization.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
---
drivers/net/wireless/ralink/rt2x00/rt2800lib.c | 9 ---------
drivers/net/wireless/ralink/rt2x00/rt2x00.h | 2 --
drivers/net/wireless/ralink/rt2x00/rt2x00queue.c | 1 -
3 files changed, 12 deletions(-)
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2800lib.c b/drivers/net/wireless/ralink/rt2x00/rt2800lib.c
index e338b5d67a57..cb290b5a6438 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2800lib.c
+++ b/drivers/net/wireless/ralink/rt2x00/rt2800lib.c
@@ -1164,15 +1164,6 @@ bool rt2800_txstatus_timeout(struct rt2x00_dev *rt2x00dev)
struct data_queue *queue;
struct queue_entry *entry;
- if (!test_bit(DEVICE_STATE_FLUSHING, &rt2x00dev->flags)) {
- unsigned long tout = msecs_to_jiffies(1000);
-
- if (time_before(jiffies, rt2x00dev->last_nostatus_check + tout))
- return false;
- }
-
- rt2x00dev->last_nostatus_check = jiffies;
-
tx_queue_for_each(rt2x00dev, queue) {
entry = rt2x00queue_get_entry(queue, Q_INDEX_DONE);
if (rt2800_entry_txstatus_timeout(rt2x00dev, entry))
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2x00.h b/drivers/net/wireless/ralink/rt2x00/rt2x00.h
index c1519c98ad41..207d92ddc77c 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2x00.h
+++ b/drivers/net/wireless/ralink/rt2x00/rt2x00.h
@@ -981,8 +981,6 @@ struct rt2x00_dev {
*/
DECLARE_KFIFO_PTR(txstatus_fifo, u32);
- unsigned long last_nostatus_check;
-
/*
* Timer to ensure tx status reports are read (rt2800usb).
*/
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2x00queue.c b/drivers/net/wireless/ralink/rt2x00/rt2x00queue.c
index 947fc8964e9a..1ab507c75c30 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2x00queue.c
+++ b/drivers/net/wireless/ralink/rt2x00/rt2x00queue.c
@@ -1039,7 +1039,6 @@ void rt2x00queue_start_queues(struct rt2x00_dev *rt2x00dev)
*/
tx_queue_for_each(rt2x00dev, queue)
rt2x00queue_start_queue(queue);
- rt2x00dev->last_nostatus_check = jiffies;
rt2x00queue_start_queue(rt2x00dev->rx);
}
--
2.7.5
^ permalink raw reply related
* [PATCH 5/8] rt2800mmio: use timer and work for handling tx statuses timeouts
From: Stanislaw Gruszka @ 2019-03-27 9:58 UTC (permalink / raw)
To: linux-wireless
Cc: Tomislav Požega, Daniel Golle, Felix Fietkau, Mathias Kresin
In-Reply-To: <1553680711-15730-1-git-send-email-sgruszka@redhat.com>
Sometimes we can get into situation when there are pending statuses,
but we do not get INT_SOURCE_CSR_TX_FIFO_STATUS. Handle this situation
by arming timeout timer and read statuses (it will fix case when
we just do not have irq) and queue work to handle case we missed
statues from hardware FIFO.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
---
drivers/net/wireless/ralink/rt2x00/rt2800mmio.c | 81 +++++++++++++++++++++++--
drivers/net/wireless/ralink/rt2x00/rt2800mmio.h | 1 +
drivers/net/wireless/ralink/rt2x00/rt2800pci.c | 2 +-
drivers/net/wireless/ralink/rt2x00/rt2800soc.c | 2 +-
drivers/net/wireless/ralink/rt2x00/rt2x00dev.c | 4 ++
5 files changed, 82 insertions(+), 8 deletions(-)
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2800mmio.c b/drivers/net/wireless/ralink/rt2x00/rt2800mmio.c
index 3d4e37a9854e..ecc4c9332ec7 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2800mmio.c
+++ b/drivers/net/wireless/ralink/rt2x00/rt2800mmio.c
@@ -426,6 +426,9 @@ void rt2800mmio_start_queue(struct data_queue *queue)
}
EXPORT_SYMBOL_GPL(rt2800mmio_start_queue);
+/* 200 ms */
+#define TXSTATUS_TIMEOUT 200000000
+
void rt2800mmio_kick_queue(struct data_queue *queue)
{
struct rt2x00_dev *rt2x00dev = queue->rt2x00dev;
@@ -440,6 +443,8 @@ void rt2800mmio_kick_queue(struct data_queue *queue)
entry = rt2x00queue_get_entry(queue, Q_INDEX);
rt2x00mmio_register_write(rt2x00dev, TX_CTX_IDX(queue->qid),
entry->entry_idx);
+ hrtimer_start(&rt2x00dev->txstatus_timer,
+ TXSTATUS_TIMEOUT, HRTIMER_MODE_REL);
break;
case QID_MGMT:
entry = rt2x00queue_get_entry(queue, Q_INDEX);
@@ -484,12 +489,8 @@ void rt2800mmio_flush_queue(struct data_queue *queue, bool drop)
* For TX queues schedule completion tasklet to catch
* tx status timeouts, othewise just wait.
*/
- if (tx_queue) {
- tasklet_disable(&rt2x00dev->txstatus_tasklet);
- rt2800_txdone(rt2x00dev, UINT_MAX);
- rt2800_txdone_nostatus(rt2x00dev);
- tasklet_enable(&rt2x00dev->txstatus_tasklet);
- }
+ if (tx_queue)
+ queue_work(rt2x00dev->workqueue, &rt2x00dev->txdone_work);
/*
* Wait for a little while to give the driver
@@ -627,6 +628,10 @@ void rt2800mmio_clear_entry(struct queue_entry *entry)
word = rt2x00_desc_read(entry_priv->desc, 1);
rt2x00_set_field32(&word, TXD_W1_DMA_DONE, 1);
rt2x00_desc_write(entry_priv->desc, 1, word);
+
+ /* If last entry stop txstatus timer */
+ if (entry->queue->length == 1)
+ hrtimer_cancel(&rt2x00dev->txstatus_timer);
}
}
EXPORT_SYMBOL_GPL(rt2800mmio_clear_entry);
@@ -759,6 +764,70 @@ int rt2800mmio_enable_radio(struct rt2x00_dev *rt2x00dev)
}
EXPORT_SYMBOL_GPL(rt2800mmio_enable_radio);
+static void rt2800mmio_work_txdone(struct work_struct *work)
+{
+ struct rt2x00_dev *rt2x00dev =
+ container_of(work, struct rt2x00_dev, txdone_work);
+
+ if (!test_bit(DEVICE_STATE_ENABLED_RADIO, &rt2x00dev->flags))
+ return;
+
+ while (!kfifo_is_empty(&rt2x00dev->txstatus_fifo) ||
+ rt2800_txstatus_timeout(rt2x00dev)) {
+
+ tasklet_disable(&rt2x00dev->txstatus_tasklet);
+ rt2800_txdone(rt2x00dev, UINT_MAX);
+ rt2800_txdone_nostatus(rt2x00dev);
+ tasklet_enable(&rt2x00dev->txstatus_tasklet);
+ }
+
+ if (rt2800_txstatus_pending(rt2x00dev))
+ hrtimer_start(&rt2x00dev->txstatus_timer,
+ TXSTATUS_TIMEOUT, HRTIMER_MODE_REL);
+}
+
+static enum hrtimer_restart rt2800mmio_tx_sta_fifo_timeout(struct hrtimer *timer)
+{
+ struct rt2x00_dev *rt2x00dev =
+ container_of(timer, struct rt2x00_dev, txstatus_timer);
+
+ if (!test_bit(DEVICE_STATE_ENABLED_RADIO, &rt2x00dev->flags))
+ goto out;
+
+ if (!rt2800_txstatus_pending(rt2x00dev))
+ goto out;
+
+ rt2800mmio_fetch_txstatus(rt2x00dev);
+ if (!kfifo_is_empty(&rt2x00dev->txstatus_fifo))
+ tasklet_schedule(&rt2x00dev->txstatus_tasklet);
+ else
+ queue_work(rt2x00dev->workqueue, &rt2x00dev->txdone_work);
+out:
+ return HRTIMER_NORESTART;
+}
+
+int rt2800mmio_probe_hw(struct rt2x00_dev *rt2x00dev)
+{
+ int retval;
+
+ retval = rt2800_probe_hw(rt2x00dev);
+ if (retval)
+ return retval;
+
+ /*
+ * Set txstatus timer function.
+ */
+ rt2x00dev->txstatus_timer.function = rt2800mmio_tx_sta_fifo_timeout;
+
+ /*
+ * Overwrite TX done handler
+ */
+ INIT_WORK(&rt2x00dev->txdone_work, rt2800mmio_work_txdone);
+
+ return 0;
+}
+EXPORT_SYMBOL_GPL(rt2800mmio_probe_hw);
+
MODULE_AUTHOR(DRV_PROJECT);
MODULE_VERSION(DRV_VERSION);
MODULE_DESCRIPTION("rt2800 MMIO library");
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2800mmio.h b/drivers/net/wireless/ralink/rt2x00/rt2800mmio.h
index 3a513273f414..ca58e6c3a4e5 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2800mmio.h
+++ b/drivers/net/wireless/ralink/rt2x00/rt2800mmio.h
@@ -153,6 +153,7 @@ void rt2800mmio_stop_queue(struct data_queue *queue);
void rt2800mmio_queue_init(struct data_queue *queue);
/* Initialization functions */
+int rt2800mmio_probe_hw(struct rt2x00_dev *rt2x00dev);
bool rt2800mmio_get_entry_state(struct queue_entry *entry);
void rt2800mmio_clear_entry(struct queue_entry *entry);
int rt2800mmio_init_queues(struct rt2x00_dev *rt2x00dev);
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2800pci.c b/drivers/net/wireless/ralink/rt2x00/rt2800pci.c
index 0291441ac548..43e1b1ee96bf 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2800pci.c
+++ b/drivers/net/wireless/ralink/rt2x00/rt2800pci.c
@@ -346,7 +346,7 @@ static const struct rt2x00lib_ops rt2800pci_rt2x00_ops = {
.tbtt_tasklet = rt2800mmio_tbtt_tasklet,
.rxdone_tasklet = rt2800mmio_rxdone_tasklet,
.autowake_tasklet = rt2800mmio_autowake_tasklet,
- .probe_hw = rt2800_probe_hw,
+ .probe_hw = rt2800mmio_probe_hw,
.get_firmware_name = rt2800pci_get_firmware_name,
.check_firmware = rt2800_check_firmware,
.load_firmware = rt2800_load_firmware,
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2800soc.c b/drivers/net/wireless/ralink/rt2x00/rt2800soc.c
index 05a801774713..cc9c2a1d5916 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2800soc.c
+++ b/drivers/net/wireless/ralink/rt2x00/rt2800soc.c
@@ -185,7 +185,7 @@ static const struct rt2x00lib_ops rt2800soc_rt2x00_ops = {
.tbtt_tasklet = rt2800mmio_tbtt_tasklet,
.rxdone_tasklet = rt2800mmio_rxdone_tasklet,
.autowake_tasklet = rt2800mmio_autowake_tasklet,
- .probe_hw = rt2800_probe_hw,
+ .probe_hw = rt2800mmio_probe_hw,
.get_firmware_name = rt2800soc_get_firmware_name,
.check_firmware = rt2800soc_check_firmware,
.load_firmware = rt2800soc_load_firmware,
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2x00dev.c b/drivers/net/wireless/ralink/rt2x00/rt2x00dev.c
index 357c0941aaad..bdc55d649a88 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2x00dev.c
+++ b/drivers/net/wireless/ralink/rt2x00/rt2x00dev.c
@@ -1391,6 +1391,8 @@ int rt2x00lib_probe_dev(struct rt2x00_dev *rt2x00dev)
mutex_init(&rt2x00dev->conf_mutex);
INIT_LIST_HEAD(&rt2x00dev->bar_list);
spin_lock_init(&rt2x00dev->bar_list_lock);
+ hrtimer_init(&rt2x00dev->txstatus_timer, CLOCK_MONOTONIC,
+ HRTIMER_MODE_REL);
set_bit(DEVICE_STATE_PRESENT, &rt2x00dev->flags);
@@ -1515,6 +1517,8 @@ void rt2x00lib_remove_dev(struct rt2x00_dev *rt2x00dev)
cancel_delayed_work_sync(&rt2x00dev->autowakeup_work);
cancel_work_sync(&rt2x00dev->sleep_work);
+ hrtimer_cancel(&rt2x00dev->txstatus_timer);
+
/*
* Kill the tx status tasklet.
*/
--
2.7.5
^ permalink raw reply related
* [PATCH 4/8] rt2800mmio: fetch tx status changes
From: Stanislaw Gruszka @ 2019-03-27 9:58 UTC (permalink / raw)
To: linux-wireless
Cc: Tomislav Požega, Daniel Golle, Felix Fietkau, Mathias Kresin
In-Reply-To: <1553680711-15730-1-git-send-email-sgruszka@redhat.com>
Prepare to use rt2800mmio_fetch_txstatus() in concurrent manner and drop
return value since is not longer needed.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
---
drivers/net/wireless/ralink/rt2x00/rt2800mmio.c | 17 +++++++++--------
1 file changed, 9 insertions(+), 8 deletions(-)
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2800mmio.c b/drivers/net/wireless/ralink/rt2x00/rt2800mmio.c
index f7635a47dbcd..3d4e37a9854e 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2800mmio.c
+++ b/drivers/net/wireless/ralink/rt2x00/rt2800mmio.c
@@ -255,12 +255,12 @@ void rt2800mmio_autowake_tasklet(unsigned long data)
}
EXPORT_SYMBOL_GPL(rt2800mmio_autowake_tasklet);
-static bool rt2800mmio_fetch_txstatus(struct rt2x00_dev *rt2x00dev)
+static void rt2800mmio_fetch_txstatus(struct rt2x00_dev *rt2x00dev)
{
u32 status;
- bool more = false;
+ unsigned long flags;
- /* FIXEME: rewrite this comment
+ /*
* The TX_FIFO_STATUS interrupt needs special care. We should
* read TX_STA_FIFO but we should do it immediately as otherwise
* the register can overflow and we would lose status reports.
@@ -271,20 +271,21 @@ static bool rt2800mmio_fetch_txstatus(struct rt2x00_dev *rt2x00dev)
* because we can schedule the tasklet multiple times (when the
* interrupt fires again during tx status processing).
*
- * txstatus tasklet is called with INT_SOURCE_CSR_TX_FIFO_STATUS
- * disabled so have only one producer and one consumer - we don't
- * need to lock the kfifo.
+ * We also read statuses from tx status timeout timer, use
+ * lock to prevent concurent writes to fifo.
*/
+
+ spin_lock_irqsave(&rt2x00dev->irqmask_lock, flags);
+
while (!kfifo_is_full(&rt2x00dev->txstatus_fifo)) {
status = rt2x00mmio_register_read(rt2x00dev, TX_STA_FIFO);
if (!rt2x00_get_field32(status, TX_STA_FIFO_VALID))
break;
kfifo_put(&rt2x00dev->txstatus_fifo, status);
- more = true;
}
- return more;
+ spin_unlock_irqrestore(&rt2x00dev->irqmask_lock, flags);
}
void rt2800mmio_txstatus_tasklet(unsigned long data)
--
2.7.5
^ permalink raw reply related
* [PATCH 3/8] rt2800: move txstatus pending routine
From: Stanislaw Gruszka @ 2019-03-27 9:58 UTC (permalink / raw)
To: linux-wireless
Cc: Tomislav Požega, Daniel Golle, Felix Fietkau, Mathias Kresin
In-Reply-To: <1553680711-15730-1-git-send-email-sgruszka@redhat.com>
Move rt2800usb_txstatus_pending routine to rt2800lib. It will be reused
by rt2800mmio code.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
---
drivers/net/wireless/ralink/rt2x00/rt2800lib.c | 17 +++++++++++++++++
drivers/net/wireless/ralink/rt2x00/rt2800lib.h | 1 +
drivers/net/wireless/ralink/rt2x00/rt2800usb.c | 22 +++-------------------
3 files changed, 21 insertions(+), 19 deletions(-)
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2800lib.c b/drivers/net/wireless/ralink/rt2x00/rt2800lib.c
index 635aa6729529..e338b5d67a57 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2800lib.c
+++ b/drivers/net/wireless/ralink/rt2x00/rt2800lib.c
@@ -1183,6 +1183,23 @@ bool rt2800_txstatus_timeout(struct rt2x00_dev *rt2x00dev)
}
EXPORT_SYMBOL_GPL(rt2800_txstatus_timeout);
+/*
+ * test if there is an entry in any TX queue for which DMA is done
+ * but the TX status has not been returned yet
+ */
+bool rt2800_txstatus_pending(struct rt2x00_dev *rt2x00dev)
+{
+ struct data_queue *queue;
+
+ tx_queue_for_each(rt2x00dev, queue) {
+ if (rt2x00queue_get_entry(queue, Q_INDEX_DMA_DONE) !=
+ rt2x00queue_get_entry(queue, Q_INDEX_DONE))
+ return true;
+ }
+ return false;
+}
+EXPORT_SYMBOL_GPL(rt2800_txstatus_pending);
+
void rt2800_txdone_nostatus(struct rt2x00_dev *rt2x00dev)
{
struct data_queue *queue;
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2800lib.h b/drivers/net/wireless/ralink/rt2x00/rt2800lib.h
index 506c319070dd..759eab2b70c3 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2800lib.h
+++ b/drivers/net/wireless/ralink/rt2x00/rt2800lib.h
@@ -198,6 +198,7 @@ void rt2800_txdone_entry(struct queue_entry *entry, u32 status, __le32 *txwi,
void rt2800_txdone(struct rt2x00_dev *rt2x00dev, unsigned int quota);
void rt2800_txdone_nostatus(struct rt2x00_dev *rt2x00dev);
bool rt2800_txstatus_timeout(struct rt2x00_dev *rt2x00dev);
+bool rt2800_txstatus_pending(struct rt2x00_dev *rt2x00dev);
void rt2800_write_beacon(struct queue_entry *entry, struct txentry_desc *txdesc);
void rt2800_clear_beacon(struct queue_entry *entry);
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2800usb.c b/drivers/net/wireless/ralink/rt2x00/rt2800usb.c
index 0bdcab374203..46f2239698a3 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2800usb.c
+++ b/drivers/net/wireless/ralink/rt2x00/rt2800usb.c
@@ -100,22 +100,6 @@ static void rt2800usb_stop_queue(struct data_queue *queue)
}
}
-/*
- * test if there is an entry in any TX queue for which DMA is done
- * but the TX status has not been returned yet
- */
-static bool rt2800usb_txstatus_pending(struct rt2x00_dev *rt2x00dev)
-{
- struct data_queue *queue;
-
- tx_queue_for_each(rt2x00dev, queue) {
- if (rt2x00queue_get_entry(queue, Q_INDEX_DMA_DONE) !=
- rt2x00queue_get_entry(queue, Q_INDEX_DONE))
- return true;
- }
- return false;
-}
-
#define TXSTATUS_READ_INTERVAL 1000000
static bool rt2800usb_tx_sta_fifo_read_completed(struct rt2x00_dev *rt2x00dev,
@@ -145,7 +129,7 @@ static bool rt2800usb_tx_sta_fifo_read_completed(struct rt2x00_dev *rt2x00dev,
if (rt2800_txstatus_timeout(rt2x00dev))
queue_work(rt2x00dev->workqueue, &rt2x00dev->txdone_work);
- if (rt2800usb_txstatus_pending(rt2x00dev)) {
+ if (rt2800_txstatus_pending(rt2x00dev)) {
/* Read register after 1 ms */
hrtimer_start(&rt2x00dev->txstatus_timer,
TXSTATUS_READ_INTERVAL,
@@ -160,7 +144,7 @@ static bool rt2800usb_tx_sta_fifo_read_completed(struct rt2x00_dev *rt2x00dev,
* clear_bit someone could do rt2x00usb_interrupt_txdone, so recheck
* here again if status reading is needed.
*/
- if (rt2800usb_txstatus_pending(rt2x00dev) &&
+ if (rt2800_txstatus_pending(rt2x00dev) &&
!test_and_set_bit(TX_STATUS_READING, &rt2x00dev->flags))
return true;
else
@@ -489,7 +473,7 @@ static void rt2800usb_work_txdone(struct work_struct *work)
* if the medium is busy, thus the TX_STA_FIFO entry is
* also delayed -> use a timer to retrieve it.
*/
- if (rt2800usb_txstatus_pending(rt2x00dev))
+ if (rt2800_txstatus_pending(rt2x00dev))
rt2800usb_async_read_tx_status(rt2x00dev);
}
}
--
2.7.5
^ permalink raw reply related
* [PATCH 2/8] rt2800: new flush implementation for SoC devices
From: Stanislaw Gruszka @ 2019-03-27 9:58 UTC (permalink / raw)
To: linux-wireless
Cc: Tomislav Požega, Daniel Golle, Felix Fietkau, Mathias Kresin
In-Reply-To: <1553680711-15730-1-git-send-email-sgruszka@redhat.com>
Use new flush_queue() callback for SoC devices, what was already done for
PCIe devices.
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
---
drivers/net/wireless/ralink/rt2x00/rt2800soc.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2800soc.c b/drivers/net/wireless/ralink/rt2x00/rt2800soc.c
index a502816214ab..05a801774713 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2800soc.c
+++ b/drivers/net/wireless/ralink/rt2x00/rt2800soc.c
@@ -203,7 +203,7 @@ static const struct rt2x00lib_ops rt2800soc_rt2x00_ops = {
.start_queue = rt2800mmio_start_queue,
.kick_queue = rt2800mmio_kick_queue,
.stop_queue = rt2800mmio_stop_queue,
- .flush_queue = rt2x00mmio_flush_queue,
+ .flush_queue = rt2800mmio_flush_queue,
.write_tx_desc = rt2800mmio_write_tx_desc,
.write_tx_data = rt2800_write_tx_data,
.write_beacon = rt2800_write_beacon,
--
2.7.5
^ permalink raw reply related
* [PATCH 1/8] rt2800: partially restore old mmio txstatus behaviour
From: Stanislaw Gruszka @ 2019-03-27 9:58 UTC (permalink / raw)
To: linux-wireless
Cc: Tomislav Požega, Daniel Golle, Felix Fietkau, Mathias Kresin
In-Reply-To: <1553680711-15730-1-git-send-email-sgruszka@redhat.com>
Do not disable txstatus interrupt and add quota of processed tx statuses in
one tasklet. Quota is needed to allow to fed device with new frames during
processing of tx statuses.
Patch fixes about 15% performance degradation on some scenarios caused by
0b0d556e0ebb ("rt2800mmio: use txdone/txstatus routines from lib").
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
---
drivers/net/wireless/ralink/rt2x00/rt2800lib.c | 4 ++--
drivers/net/wireless/ralink/rt2x00/rt2800lib.h | 2 +-
drivers/net/wireless/ralink/rt2x00/rt2800mmio.c | 30 +++++++------------------
drivers/net/wireless/ralink/rt2x00/rt2800usb.c | 2 +-
4 files changed, 12 insertions(+), 26 deletions(-)
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2800lib.c b/drivers/net/wireless/ralink/rt2x00/rt2800lib.c
index a03b5284a050..635aa6729529 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2800lib.c
+++ b/drivers/net/wireless/ralink/rt2x00/rt2800lib.c
@@ -1100,7 +1100,7 @@ void rt2800_txdone_entry(struct queue_entry *entry, u32 status, __le32 *txwi,
}
EXPORT_SYMBOL_GPL(rt2800_txdone_entry);
-void rt2800_txdone(struct rt2x00_dev *rt2x00dev)
+void rt2800_txdone(struct rt2x00_dev *rt2x00dev, unsigned int quota)
{
struct data_queue *queue;
struct queue_entry *entry;
@@ -1108,7 +1108,7 @@ void rt2800_txdone(struct rt2x00_dev *rt2x00dev)
u8 qid;
bool match;
- while (kfifo_get(&rt2x00dev->txstatus_fifo, ®)) {
+ while (quota-- > 0 && kfifo_get(&rt2x00dev->txstatus_fifo, ®)) {
/*
* TX_STA_FIFO_PID_QUEUE is a 2-bit field, thus qid is
* guaranteed to be one of the TX QIDs .
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2800lib.h b/drivers/net/wireless/ralink/rt2x00/rt2800lib.h
index 0dff2c7b3010..506c319070dd 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2800lib.h
+++ b/drivers/net/wireless/ralink/rt2x00/rt2800lib.h
@@ -195,7 +195,7 @@ void rt2800_process_rxwi(struct queue_entry *entry, struct rxdone_entry_desc *tx
void rt2800_txdone_entry(struct queue_entry *entry, u32 status, __le32 *txwi,
bool match);
-void rt2800_txdone(struct rt2x00_dev *rt2x00dev);
+void rt2800_txdone(struct rt2x00_dev *rt2x00dev, unsigned int quota);
void rt2800_txdone_nostatus(struct rt2x00_dev *rt2x00dev);
bool rt2800_txstatus_timeout(struct rt2x00_dev *rt2x00dev);
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2800mmio.c b/drivers/net/wireless/ralink/rt2x00/rt2800mmio.c
index ddb88cfeace2..f7635a47dbcd 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2800mmio.c
+++ b/drivers/net/wireless/ralink/rt2x00/rt2800mmio.c
@@ -255,20 +255,6 @@ void rt2800mmio_autowake_tasklet(unsigned long data)
}
EXPORT_SYMBOL_GPL(rt2800mmio_autowake_tasklet);
-static void rt2800mmio_txdone(struct rt2x00_dev *rt2x00dev)
-{
- bool timeout = false;
-
- while (!kfifo_is_empty(&rt2x00dev->txstatus_fifo) ||
- (timeout = rt2800_txstatus_timeout(rt2x00dev))) {
-
- rt2800_txdone(rt2x00dev);
-
- if (timeout)
- rt2800_txdone_nostatus(rt2x00dev);
- }
-}
-
static bool rt2800mmio_fetch_txstatus(struct rt2x00_dev *rt2x00dev)
{
u32 status;
@@ -305,14 +291,11 @@ void rt2800mmio_txstatus_tasklet(unsigned long data)
{
struct rt2x00_dev *rt2x00dev = (struct rt2x00_dev *)data;
- do {
- rt2800mmio_txdone(rt2x00dev);
+ rt2800_txdone(rt2x00dev, 16);
- } while (rt2800mmio_fetch_txstatus(rt2x00dev));
+ if (!kfifo_is_empty(&rt2x00dev->txstatus_fifo))
+ tasklet_schedule(&rt2x00dev->txstatus_tasklet);
- if (test_bit(DEVICE_STATE_ENABLED_RADIO, &rt2x00dev->flags))
- rt2800mmio_enable_interrupt(rt2x00dev,
- INT_SOURCE_CSR_TX_FIFO_STATUS);
}
EXPORT_SYMBOL_GPL(rt2800mmio_txstatus_tasklet);
@@ -339,8 +322,10 @@ irqreturn_t rt2800mmio_interrupt(int irq, void *dev_instance)
mask = ~reg;
if (rt2x00_get_field32(reg, INT_SOURCE_CSR_TX_FIFO_STATUS)) {
+ rt2x00_set_field32(&mask, INT_MASK_CSR_TX_FIFO_STATUS, 1);
rt2800mmio_fetch_txstatus(rt2x00dev);
- tasklet_schedule(&rt2x00dev->txstatus_tasklet);
+ if (!kfifo_is_empty(&rt2x00dev->txstatus_fifo))
+ tasklet_schedule(&rt2x00dev->txstatus_tasklet);
}
if (rt2x00_get_field32(reg, INT_SOURCE_CSR_PRE_TBTT))
@@ -500,7 +485,8 @@ void rt2800mmio_flush_queue(struct data_queue *queue, bool drop)
*/
if (tx_queue) {
tasklet_disable(&rt2x00dev->txstatus_tasklet);
- rt2800mmio_txdone(rt2x00dev);
+ rt2800_txdone(rt2x00dev, UINT_MAX);
+ rt2800_txdone_nostatus(rt2x00dev);
tasklet_enable(&rt2x00dev->txstatus_tasklet);
}
diff --git a/drivers/net/wireless/ralink/rt2x00/rt2800usb.c b/drivers/net/wireless/ralink/rt2x00/rt2800usb.c
index 19eabf16147b..0bdcab374203 100644
--- a/drivers/net/wireless/ralink/rt2x00/rt2800usb.c
+++ b/drivers/net/wireless/ralink/rt2x00/rt2800usb.c
@@ -480,7 +480,7 @@ static void rt2800usb_work_txdone(struct work_struct *work)
while (!kfifo_is_empty(&rt2x00dev->txstatus_fifo) ||
rt2800_txstatus_timeout(rt2x00dev)) {
- rt2800_txdone(rt2x00dev);
+ rt2800_txdone(rt2x00dev, UINT_MAX);
rt2800_txdone_nostatus(rt2x00dev);
--
2.7.5
^ permalink raw reply related
* [PATCH 0/8] rt2x00: performance fixes for mmio
From: Stanislaw Gruszka @ 2019-03-27 9:58 UTC (permalink / raw)
To: linux-wireless
Cc: Tomislav Požega, Daniel Golle, Felix Fietkau, Mathias Kresin
Patch series should fix performance regression and do not reintroduce
tx queue hungs.
Patchset is based on:
'[PATCH v3 0/4] rt2x00 random fixes'
https://lore.kernel.org/linux-wireless/1552384303-29529-1-git-send-email-sgruszka@redhat.com/
RFC -> v1
- fix some spelling mistakes in changlogs
Stanislaw Gruszka (8):
rt2800: partially restore old mmio txstatus behaviour
rt2800: new flush implementation for SoC devices
rt2800: move txstatus pending routine
rt2800mmio: fetch tx status changes
rt2800mmio: use timer and work for handling tx statuses timeouts
rt2x00: remove last_nostatus_check
rt2x00: remove not used entry field
rt2x00mmio: remove legacy comment
drivers/net/wireless/ralink/rt2x00/rt2800lib.c | 30 ++++--
drivers/net/wireless/ralink/rt2x00/rt2800lib.h | 3 +-
drivers/net/wireless/ralink/rt2x00/rt2800mmio.c | 124 ++++++++++++++++-------
drivers/net/wireless/ralink/rt2x00/rt2800mmio.h | 1 +
drivers/net/wireless/ralink/rt2x00/rt2800pci.c | 2 +-
drivers/net/wireless/ralink/rt2x00/rt2800soc.c | 4 +-
drivers/net/wireless/ralink/rt2x00/rt2800usb.c | 24 +----
drivers/net/wireless/ralink/rt2x00/rt2x00.h | 2 -
drivers/net/wireless/ralink/rt2x00/rt2x00dev.c | 4 +
drivers/net/wireless/ralink/rt2x00/rt2x00mmio.h | 2 -
drivers/net/wireless/ralink/rt2x00/rt2x00queue.c | 1 -
drivers/net/wireless/ralink/rt2x00/rt2x00queue.h | 3 -
12 files changed, 123 insertions(+), 77 deletions(-)
--
2.7.5
^ permalink raw reply
* mwifiex: update set_mac_address logic
From: Ganapathi Bhat @ 2019-03-27 7:29 UTC (permalink / raw)
To: linux-wireless@vger.kernel.org
Cc: Cathy Luo, Zhiyuan Yang, James Cao, Rakesh Parmar, Brian Norris
From: Sharvari Harisangam <sharvari@marvell.com>
Date: Wed, 27 Mar 2019 11:47:03 +0530
Subject: mwifiex: update set_mac_address logic
In set_mac_address, driver check for interfaces with same bss_type
For first STA entry, this would return 3 interfaces since all priv's have
bss_type as 0 due to kzalloc. Thus mac address gets changed for STA
unexpected. This patch adds check for first STA and avoids mac address
change. This patch also adds mac_address change for p2p based on bss_num
type.
Signed-off-by: Sharvari Harisangam <sharvari@marvell.com>
Signed-off-by: Ganapathi Bhat <gbhat@marvell.com>
---
drivers/net/wireless/marvell/mwifiex/main.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/net/wireless/marvell/mwifiex/main.c b/drivers/net/wireless/marvell/mwifiex/main.c
index 5ed2d9b..4eabd94 100644
--- a/drivers/net/wireless/marvell/mwifiex/main.c
+++ b/drivers/net/wireless/marvell/mwifiex/main.c
@@ -965,10 +965,10 @@ int mwifiex_set_mac_address(struct mwifiex_private *priv,
mac_addr = old_mac_addr;
- if (priv->bss_type == MWIFIEX_BSS_TYPE_P2P)
+ if (priv->bss_type == MWIFIEX_BSS_TYPE_P2P) {
mac_addr |= BIT_ULL(MWIFIEX_MAC_LOCAL_ADMIN_BIT);
-
- if (mwifiex_get_intf_num(priv->adapter, priv->bss_type) > 1) {
+ mac_addr += priv->bss_num;
+ } else if (priv->adapter->priv[0] != priv) {
/* Set mac address based on bss_type/bss_num */
mac_addr ^= BIT_ULL(priv->bss_type + 8);
mac_addr += priv->bss_num;
--
1.9.1
^ permalink raw reply related
* Re: [PATCH] iwlwifi: mvm: no need to check return value of debugfs_create functions
From: Luca Coelho @ 2019-03-27 6:35 UTC (permalink / raw)
To: Greg Kroah-Hartman, Laura Abbott
Cc: linux-kernel, Johannes Berg, Emmanuel Grumbach,
Intel Linux Wireless, Kalle Valo, linux-wireless
In-Reply-To: <20190327015346.GA17979@kroah.com>
On Wed, 2019-03-27 at 10:53 +0900, Greg Kroah-Hartman wrote:
> On Tue, Mar 26, 2019 at 06:47:33PM -0700, Laura Abbott wrote:
> > On 3/26/19 6:26 PM, Greg Kroah-Hartman wrote:
> > > On Tue, Mar 26, 2019 at 04:55:54PM -0700, Laura Abbott wrote:
> > > > On 1/22/19 7:21 AM, Greg Kroah-Hartman wrote:
> > > > > diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/debugfs-vif.c b/drivers/net/wireless/intel/iwlwifi/mvm/debugfs-vif.c
> > > > > index 33b0af24a537..c52cdc538678 100644
> > > > > --- a/drivers/net/wireless/intel/iwlwifi/mvm/debugfs-vif.c
> > > > > +++ b/drivers/net/wireless/intel/iwlwifi/mvm/debugfs-vif.c
> > > > > @@ -1446,9 +1446,8 @@ static ssize_t iwl_dbgfs_quota_min_read(struct file *file,
> > > > > #define MVM_DEBUGFS_READ_WRITE_FILE_OPS(name, bufsz) \
> > > > > _MVM_DEBUGFS_READ_WRITE_FILE_OPS(name, bufsz, struct ieee80211_vif)
> > > > > #define MVM_DEBUGFS_ADD_FILE_VIF(name, parent, mode) do { \
> > > > > - if (!debugfs_create_file(#name, mode, parent, vif, \
> > > > > - &iwl_dbgfs_##name##_ops)) \
> > > > > - goto err; \
> > > > > + debugfs_create_file(#name, mode, parent, vif, \
> > > > > + &iwl_dbgfs_##name##_ops); \
> > > > > } while (0)
> > > > > MVM_DEBUGFS_READ_FILE_OPS(mac_params);
> > > > > @@ -1483,12 +1482,6 @@ void iwl_mvm_vif_dbgfs_register(struct iwl_mvm *mvm, struct ieee80211_vif *vif)
> > > > > mvmvif->dbgfs_dir = debugfs_create_dir("iwlmvm", dbgfs_dir);
> > > > > - if (!mvmvif->dbgfs_dir) {
> > > > > - IWL_ERR(mvm, "Failed to create debugfs directory under %pd\n",
> > > > > - dbgfs_dir);
> > > > > - return;
> > > > > - }
> > > > > -
> > > > > if (iwlmvm_mod_params.power_scheme != IWL_POWER_SCHEME_CAM &&
> > > > > ((vif->type == NL80211_IFTYPE_STATION && !vif->p2p) ||
> > > > > (vif->type == NL80211_IFTYPE_STATION && vif->p2p)))
> > > > > @@ -1537,12 +1530,6 @@ void iwl_mvm_vif_dbgfs_register(struct iwl_mvm *mvm, struct ieee80211_vif *vif)
> > > > > mvmvif->dbgfs_slink = debugfs_create_symlink(dbgfs_dir->d_name.name,
> > > > > mvm->debugfs_dir, buf);
> > > > > - if (!mvmvif->dbgfs_slink)
> > > > > - IWL_ERR(mvm, "Can't create debugfs symbolic link under %pd\n",
> > > > > - dbgfs_dir);
> > > > > - return;
> > > > > -err:
> > > > > - IWL_ERR(mvm, "Can't create debugfs entity\n");
> > > > > }
> > > >
> > > > Fedora got a bug report https://bugzilla.redhat.com/show_bug.cgi?id=1691034
> > > > of a crash with 5.0 and the user did a bisect which pointed to ff9fb72bc077
> > > > ("debugfs: return error values, not NULL") because the error checking is
> > > > no longer correct in this driver.
> > > >
> > > > Based on https://patchwork.kernel.org/patch/10865839/, it looks like
> > > > this is supposed to go in for 5.2 but this needs to go in now as
> > > > the error checking is currently broken without it. Can this get queued
> > > > for Linus so we can get it in 5.0 stable?
> > >
> > > That's odd, I can't see how the error checking is wrong here. If the
> > > directory is not created, an error will be returned, which should be
> > > able to be handled by debugfs_create_file().
> > >
> > > So with this patch does the error go away?
> > >
> >
> > The full patch didn't apply cleanly and I didn't try to backport it
> > for the reporter to test. I was going off of the theory that if the
> > patch was there it would fix the problem.
> >
> > What I _think_ is going wrong is dbgfs_dir is actually an errno value:
> >
> >
> > struct dentry *dbgfs_dir = vif->debugfs_dir;
> > struct iwl_mvm_vif *mvmvif = iwl_mvm_vif_from_mac80211(vif);
> > char buf[100];
> >
> > /*
> > * Check if debugfs directory already exist before creating it.
> > * This may happen when, for example, resetting hw or suspend-resume
> > */
> > if (!dbgfs_dir || mvmvif->dbgfs_dir)
> > return;
> >
> >
> > so this blows up in the snprintf
> >
> > snprintf(buf, 100, "../../../%pd3/%pd",
> > dbgfs_dir,
> > mvmvif->dbgfs_dir);
>
> Ah, yeah, that's horrible. They had the name before, why pull it out of
> the dentry again? That will blow up hard, but maybe printk should check
> to see if the pointer really is a pointer first.
I agree this is ugly. But do you mean we could use
("../../../%pd3/%s", dbgfs_dir, "iwlmvm")? Or how did we have the name?
Also, this would solve the sprintf() problem, but still wouldn't solve
the real issue, which is not check for ERR in dbgfs_dir.
> > Unless I misunderstood what the debugfs error change did. I think this
> > also means the if check needs to look for IS_ERR and not just !dbgfs_dir.
>
> Yes, that is correct.
Yeah, we can do that. So this patch doesn't need to be sent for
v5.1-rc* and v5.0, right? At least I don't see how it would fix the
issue. What we need is a new patch with the IS_ERR check.
--
Cheers,
Luca.
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox