* [PATCH 00/13] iwlwifi: updates intended for v5.10 2020-10-08 part 2
@ 2020-10-08 15:12 Luca Coelho
2020-10-08 15:12 ` [PATCH 01/13] iwlwifi: pcie: fix xtal latency for 9560 devices Luca Coelho
` (12 more replies)
0 siblings, 13 replies; 19+ messages in thread
From: Luca Coelho @ 2020-10-08 15:12 UTC (permalink / raw)
To: kvalo; +Cc: linux-wireless
From: Luca Coelho <luciano.coelho@intel.com>
Hi,
Here's the seventh set of patches intended for v5.10. It's the usual
development, new features, cleanups and bugfixes.
The changes are:
* Some more fixes in some device configurations;
* PNVM implementation continued;
* A few more FW API updates for newer FWs;
* Bump the FW API supported to 59;
* Some other small fixes;
As usual, I'm pushing this to a pending branch, for kbuild bot, and
will send a pull-request later.
Please review.
Cheers,
Luca.
Emmanuel Grumbach (1):
iwlwifi: mvm: don't send a CSA command the firmware doesn't know
Luca Coelho (7):
iwlwifi: pcie: fix xtal latency for 9560 devices
iwlwifi: pcie: fix the xtal latency value for a few qu devices
iwlwifi: move PNVM implementation to common code
iwlwifi: add trans op to set PNVM
iwlwifi: pcie: implement set_pnvm op
iwlwifi: read and parse PNVM file
iwlwifi: bump FW API to 59 for AX devices
Mordechay Goodstein (2):
iwlwifi: stats: add new api fields for statistics cmd/ntfy
iwlwifi: rs: align to new TLC config command API
Naftali Goldstein (1):
iwlwifi: fix sar geo table initialization
Sara Sharon (2):
iwlwifi: mvm: remove memset of kek_kck command
iwlwifi: mvm: fix suspicious rcu usage warnings
drivers/net/wireless/intel/iwlwifi/Makefile | 2 +-
.../net/wireless/intel/iwlwifi/cfg/22000.c | 4 +-
drivers/net/wireless/intel/iwlwifi/cfg/9000.c | 17 +-
drivers/net/wireless/intel/iwlwifi/fw/acpi.c | 20 +-
drivers/net/wireless/intel/iwlwifi/fw/acpi.h | 1 -
.../wireless/intel/iwlwifi/fw/api/commands.h | 6 +-
.../net/wireless/intel/iwlwifi/fw/api/rs.h | 7 +-
.../net/wireless/intel/iwlwifi/fw/api/stats.h | 461 ++++++++++++++++++
drivers/net/wireless/intel/iwlwifi/fw/file.h | 6 +-
drivers/net/wireless/intel/iwlwifi/fw/pnvm.c | 274 +++++++++++
drivers/net/wireless/intel/iwlwifi/fw/pnvm.h | 18 +
.../net/wireless/intel/iwlwifi/iwl-config.h | 1 +
.../intel/iwlwifi/iwl-context-info-gen3.h | 3 +
.../wireless/intel/iwlwifi/iwl-context-info.h | 7 +-
.../net/wireless/intel/iwlwifi/iwl-trans.h | 20 +
drivers/net/wireless/intel/iwlwifi/mvm/d3.c | 21 +-
drivers/net/wireless/intel/iwlwifi/mvm/fw.c | 84 +---
.../net/wireless/intel/iwlwifi/mvm/mac80211.c | 12 +-
.../net/wireless/intel/iwlwifi/mvm/rs-fw.c | 8 +-
drivers/net/wireless/intel/iwlwifi/mvm/rx.c | 195 +++++---
.../wireless/intel/iwlwifi/mvm/time-event.c | 22 +
.../wireless/intel/iwlwifi/mvm/time-event.h | 7 +-
.../intel/iwlwifi/pcie/ctxt-info-gen3.c | 27 +
.../wireless/intel/iwlwifi/pcie/ctxt-info.c | 21 +-
drivers/net/wireless/intel/iwlwifi/pcie/drv.c | 2 +-
.../wireless/intel/iwlwifi/pcie/internal.h | 3 +
.../net/wireless/intel/iwlwifi/pcie/trans.c | 7 +
27 files changed, 1094 insertions(+), 162 deletions(-)
create mode 100644 drivers/net/wireless/intel/iwlwifi/fw/pnvm.c
create mode 100644 drivers/net/wireless/intel/iwlwifi/fw/pnvm.h
--
2.28.0
^ permalink raw reply [flat|nested] 19+ messages in thread
* [PATCH 01/13] iwlwifi: pcie: fix xtal latency for 9560 devices
2020-10-08 15:12 [PATCH 00/13] iwlwifi: updates intended for v5.10 2020-10-08 part 2 Luca Coelho
@ 2020-10-08 15:12 ` Luca Coelho
2020-10-08 17:15 ` Kalle Valo
2020-10-08 15:12 ` [PATCH 02/13] iwlwifi: pcie: fix the xtal latency value for a few qu devices Luca Coelho
` (11 subsequent siblings)
12 siblings, 1 reply; 19+ messages in thread
From: Luca Coelho @ 2020-10-08 15:12 UTC (permalink / raw)
To: kvalo; +Cc: linux-wireless
From: Luca Coelho <luciano.coelho@intel.com>
We were using a very high latency for all 9560 devices so they all
would have time to stabilize. But this causes the system to be
slighly slower, so we can use the best values for each device.
This requires a new trans cfg struct for devices with longer latency
and some adjustments to the other structs.
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
---
drivers/net/wireless/intel/iwlwifi/cfg/9000.c | 17 +++++++++++++----
drivers/net/wireless/intel/iwlwifi/iwl-config.h | 1 +
drivers/net/wireless/intel/iwlwifi/pcie/drv.c | 2 +-
3 files changed, 15 insertions(+), 5 deletions(-)
diff --git a/drivers/net/wireless/intel/iwlwifi/cfg/9000.c b/drivers/net/wireless/intel/iwlwifi/cfg/9000.c
index f84b8e5d3f0b..be4acf4a0e32 100644
--- a/drivers/net/wireless/intel/iwlwifi/cfg/9000.c
+++ b/drivers/net/wireless/intel/iwlwifi/cfg/9000.c
@@ -6,7 +6,7 @@
* GPL LICENSE SUMMARY
*
* Copyright(c) 2015-2017 Intel Deutschland GmbH
- * Copyright (C) 2018 - 2019 Intel Corporation
+ * Copyright(c) 2018 - 2020 Intel Corporation
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of version 2 of the GNU General Public License as
@@ -20,7 +20,7 @@
* BSD LICENSE
*
* Copyright(c) 2015-2017 Intel Deutschland GmbH
- * Copyright (C) 2018 - 2019 Intel Corporation
+ * Copyright(c) 2018 - 2020 Intel Corporation
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -180,7 +180,16 @@ const struct iwl_cfg_trans_params iwl9560_trans_cfg = {
.mq_rx_supported = true,
.rf_id = true,
.integrated = true,
- .xtal_latency = 5000,
+ .xtal_latency = 650,
+};
+
+const struct iwl_cfg_trans_params iwl9560_long_latency_trans_cfg = {
+ .device_family = IWL_DEVICE_FAMILY_9000,
+ .base_params = &iwl9000_base_params,
+ .mq_rx_supported = true,
+ .rf_id = true,
+ .integrated = true,
+ .xtal_latency = 2820,
};
const struct iwl_cfg_trans_params iwl9560_shared_clk_trans_cfg = {
@@ -189,7 +198,7 @@ const struct iwl_cfg_trans_params iwl9560_shared_clk_trans_cfg = {
.mq_rx_supported = true,
.rf_id = true,
.integrated = true,
- .xtal_latency = 5000,
+ .xtal_latency = 670,
.extra_phy_cfg_flags = FW_PHY_CFG_SHARED_CLK
};
diff --git a/drivers/net/wireless/intel/iwlwifi/iwl-config.h b/drivers/net/wireless/intel/iwlwifi/iwl-config.h
index d03f51bf7dfd..ca4967b81d01 100644
--- a/drivers/net/wireless/intel/iwlwifi/iwl-config.h
+++ b/drivers/net/wireless/intel/iwlwifi/iwl-config.h
@@ -519,6 +519,7 @@ struct iwl_dev_info {
*/
extern const struct iwl_cfg_trans_params iwl9000_trans_cfg;
extern const struct iwl_cfg_trans_params iwl9560_trans_cfg;
+extern const struct iwl_cfg_trans_params iwl9560_long_latency_trans_cfg;
extern const struct iwl_cfg_trans_params iwl9560_shared_clk_trans_cfg;
extern const struct iwl_cfg_trans_params iwl_qnj_trans_cfg;
extern const struct iwl_cfg_trans_params iwl_qu_trans_cfg;
diff --git a/drivers/net/wireless/intel/iwlwifi/pcie/drv.c b/drivers/net/wireless/intel/iwlwifi/pcie/drv.c
index 83811fad8431..129021f26791 100644
--- a/drivers/net/wireless/intel/iwlwifi/pcie/drv.c
+++ b/drivers/net/wireless/intel/iwlwifi/pcie/drv.c
@@ -514,7 +514,7 @@ static const struct pci_device_id iwl_hw_card_ids[] = {
{IWL_PCI_DEVICE(0x2526, PCI_ANY_ID, iwl9000_trans_cfg)},
{IWL_PCI_DEVICE(0x271B, PCI_ANY_ID, iwl9000_trans_cfg)},
{IWL_PCI_DEVICE(0x271C, PCI_ANY_ID, iwl9000_trans_cfg)},
- {IWL_PCI_DEVICE(0x30DC, PCI_ANY_ID, iwl9560_trans_cfg)},
+ {IWL_PCI_DEVICE(0x30DC, PCI_ANY_ID, iwl9560_long_latency_trans_cfg)},
{IWL_PCI_DEVICE(0x31DC, PCI_ANY_ID, iwl9560_shared_clk_trans_cfg)},
{IWL_PCI_DEVICE(0x9DF0, PCI_ANY_ID, iwl9560_trans_cfg)},
{IWL_PCI_DEVICE(0xA370, PCI_ANY_ID, iwl9560_trans_cfg)},
--
2.28.0
^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 02/13] iwlwifi: pcie: fix the xtal latency value for a few qu devices
2020-10-08 15:12 [PATCH 00/13] iwlwifi: updates intended for v5.10 2020-10-08 part 2 Luca Coelho
2020-10-08 15:12 ` [PATCH 01/13] iwlwifi: pcie: fix xtal latency for 9560 devices Luca Coelho
@ 2020-10-08 15:12 ` Luca Coelho
2020-10-08 15:47 ` Kalle Valo
2020-10-08 15:12 ` [PATCH 03/13] iwlwifi: mvm: don't send a CSA command the firmware doesn't know Luca Coelho
` (10 subsequent siblings)
12 siblings, 1 reply; 19+ messages in thread
From: Luca Coelho @ 2020-10-08 15:12 UTC (permalink / raw)
To: kvalo; +Cc: linux-wireless
From: Luca Coelho <luciano.coelho@intel.com>
We were using 5 ms for some qu devices, but their xtal stabilizes
much faster than that. Reduce the value to 500 ms.
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
---
drivers/net/wireless/intel/iwlwifi/cfg/22000.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/wireless/intel/iwlwifi/cfg/22000.c b/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
index 60e0640d07dc..8b97c2eb2632 100644
--- a/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
+++ b/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
@@ -253,7 +253,7 @@ const struct iwl_cfg_trans_params iwl_qu_trans_cfg = {
.device_family = IWL_DEVICE_FAMILY_22000,
.base_params = &iwl_22000_base_params,
.integrated = true,
- .xtal_latency = 5000,
+ .xtal_latency = 500,
.ltr_delay = IWL_CFG_TRANS_LTR_DELAY_200US,
};
--
2.28.0
^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 03/13] iwlwifi: mvm: don't send a CSA command the firmware doesn't know
2020-10-08 15:12 [PATCH 00/13] iwlwifi: updates intended for v5.10 2020-10-08 part 2 Luca Coelho
2020-10-08 15:12 ` [PATCH 01/13] iwlwifi: pcie: fix xtal latency for 9560 devices Luca Coelho
2020-10-08 15:12 ` [PATCH 02/13] iwlwifi: pcie: fix the xtal latency value for a few qu devices Luca Coelho
@ 2020-10-08 15:12 ` Luca Coelho
2020-10-08 15:12 ` [PATCH 04/13] iwlwifi: mvm: remove memset of kek_kck command Luca Coelho
` (9 subsequent siblings)
12 siblings, 0 replies; 19+ messages in thread
From: Luca Coelho @ 2020-10-08 15:12 UTC (permalink / raw)
To: kvalo; +Cc: linux-wireless
From: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
We introduced a new flow to remove an on-going CSA but we
assumed the firwmare understands the
CHANNEL_SWITCH_TIME_EVENT_CMD. This is not true for 7265 and down.
Don't send this command for those devices, but rather use the older
command.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
---
.../net/wireless/intel/iwlwifi/mvm/mac80211.c | 12 ++++++----
.../wireless/intel/iwlwifi/mvm/time-event.c | 22 +++++++++++++++++++
.../wireless/intel/iwlwifi/mvm/time-event.h | 7 ++++--
3 files changed, 35 insertions(+), 6 deletions(-)
diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/mac80211.c b/drivers/net/wireless/intel/iwlwifi/mvm/mac80211.c
index 27b2a68c97f6..688c1125e67b 100644
--- a/drivers/net/wireless/intel/iwlwifi/mvm/mac80211.c
+++ b/drivers/net/wireless/intel/iwlwifi/mvm/mac80211.c
@@ -1409,10 +1409,14 @@ static void iwl_mvm_abort_channel_switch(struct ieee80211_hw *hw,
IWL_DEBUG_MAC80211(mvm, "Abort CSA on mac %d\n", mvmvif->id);
mutex_lock(&mvm->mutex);
- WARN_ON(iwl_mvm_send_cmd_pdu(mvm,
- WIDE_ID(MAC_CONF_GROUP,
- CHANNEL_SWITCH_TIME_EVENT_CMD),
- 0, sizeof(cmd), &cmd));
+ if (!fw_has_capa(&mvm->fw->ucode_capa,
+ IWL_UCODE_TLV_CAPA_CHANNEL_SWITCH_CMD))
+ iwl_mvm_remove_csa_period(mvm, vif);
+ else
+ WARN_ON(iwl_mvm_send_cmd_pdu(mvm,
+ WIDE_ID(MAC_CONF_GROUP,
+ CHANNEL_SWITCH_TIME_EVENT_CMD),
+ 0, sizeof(cmd), &cmd));
mutex_unlock(&mvm->mutex);
WARN_ON(iwl_mvm_post_channel_switch(hw, vif));
diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/time-event.c b/drivers/net/wireless/intel/iwlwifi/mvm/time-event.c
index 7ad6c9ab4737..7fce79c1c114 100644
--- a/drivers/net/wireless/intel/iwlwifi/mvm/time-event.c
+++ b/drivers/net/wireless/intel/iwlwifi/mvm/time-event.c
@@ -1016,6 +1016,28 @@ void iwl_mvm_stop_roc(struct iwl_mvm *mvm, struct ieee80211_vif *vif)
iwl_mvm_roc_finished(mvm);
}
+void iwl_mvm_remove_csa_period(struct iwl_mvm *mvm,
+ struct ieee80211_vif *vif)
+{
+ struct iwl_mvm_vif *mvmvif = iwl_mvm_vif_from_mac80211(vif);
+ struct iwl_mvm_time_event_data *te_data = &mvmvif->time_event_data;
+ u32 id;
+
+ lockdep_assert_held(&mvm->mutex);
+
+ if (!te_data->running)
+ return;
+
+ spin_lock_bh(&mvm->time_event_lock);
+ id = te_data->id;
+ spin_unlock_bh(&mvm->time_event_lock);
+
+ if (id != TE_CHANNEL_SWITCH_PERIOD)
+ return;
+
+ iwl_mvm_remove_time_event(mvm, mvmvif, te_data);
+}
+
int iwl_mvm_schedule_csa_period(struct iwl_mvm *mvm,
struct ieee80211_vif *vif,
u32 duration, u32 apply_time)
diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/time-event.h b/drivers/net/wireless/intel/iwlwifi/mvm/time-event.h
index 3186d7e40567..b6bac776f236 100644
--- a/drivers/net/wireless/intel/iwlwifi/mvm/time-event.h
+++ b/drivers/net/wireless/intel/iwlwifi/mvm/time-event.h
@@ -7,7 +7,7 @@
*
* Copyright(c) 2012 - 2014 Intel Corporation. All rights reserved.
* Copyright(c) 2013 - 2014 Intel Mobile Communications GmbH
- * Copyright (C) 2019 Intel Corporation
+ * Copyright (C) 2019 - 2020 Intel Corporation
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of version 2 of the GNU General Public License as
@@ -29,7 +29,7 @@
*
* Copyright(c) 2012 - 2014 Intel Corporation. All rights reserved.
* Copyright(c) 2013 - 2014 Intel Mobile Communications GmbH
- * Copyright (C) 2019 Intel Corporation
+ * Copyright (C) 2019 - 2020 Intel Corporation
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -216,6 +216,9 @@ void iwl_mvm_te_clear_data(struct iwl_mvm *mvm,
void iwl_mvm_cleanup_roc_te(struct iwl_mvm *mvm);
void iwl_mvm_roc_done_wk(struct work_struct *wk);
+void iwl_mvm_remove_csa_period(struct iwl_mvm *mvm,
+ struct ieee80211_vif *vif);
+
/**
* iwl_mvm_schedule_csa_period - request channel switch absence period
* @mvm: the mvm component
--
2.28.0
^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 04/13] iwlwifi: mvm: remove memset of kek_kck command
2020-10-08 15:12 [PATCH 00/13] iwlwifi: updates intended for v5.10 2020-10-08 part 2 Luca Coelho
` (2 preceding siblings ...)
2020-10-08 15:12 ` [PATCH 03/13] iwlwifi: mvm: don't send a CSA command the firmware doesn't know Luca Coelho
@ 2020-10-08 15:12 ` Luca Coelho
2020-10-08 15:12 ` [PATCH 05/13] iwlwifi: mvm: fix suspicious rcu usage warnings Luca Coelho
` (8 subsequent siblings)
12 siblings, 0 replies; 19+ messages in thread
From: Luca Coelho @ 2020-10-08 15:12 UTC (permalink / raw)
To: kvalo; +Cc: linux-wireless
From: Sara Sharon <sara.sharon@intel.com>
iwl_mvm_wowlan_program_keys is now setting data directly in
the KEK_KCK command, and the memset is clearing this later,
causing the data to be incomplete.
Just remove the memory clearing, the structure is local and
cleared in the declaration.
Signed-off-by: Sara Sharon <sara.sharon@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
---
drivers/net/wireless/intel/iwlwifi/mvm/d3.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/d3.c b/drivers/net/wireless/intel/iwlwifi/mvm/d3.c
index e47c0be28656..9987172a9e01 100644
--- a/drivers/net/wireless/intel/iwlwifi/mvm/d3.c
+++ b/drivers/net/wireless/intel/iwlwifi/mvm/d3.c
@@ -870,7 +870,6 @@ static int iwl_mvm_wowlan_config_key_params(struct iwl_mvm *mvm,
else
cmd_size = sizeof(struct iwl_wowlan_kek_kck_material_cmd_v2);
- memset(&kek_kck_cmd, 0, sizeof(kek_kck_cmd));
memcpy(kek_kck_cmd.kck, mvmvif->rekey_data.kck,
mvmvif->rekey_data.kck_len);
kek_kck_cmd.kck_len = cpu_to_le16(mvmvif->rekey_data.kck_len);
--
2.28.0
^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 05/13] iwlwifi: mvm: fix suspicious rcu usage warnings
2020-10-08 15:12 [PATCH 00/13] iwlwifi: updates intended for v5.10 2020-10-08 part 2 Luca Coelho
` (3 preceding siblings ...)
2020-10-08 15:12 ` [PATCH 04/13] iwlwifi: mvm: remove memset of kek_kck command Luca Coelho
@ 2020-10-08 15:12 ` Luca Coelho
2020-10-08 15:12 ` [PATCH 06/13] iwlwifi: stats: add new api fields for statistics cmd/ntfy Luca Coelho
` (7 subsequent siblings)
12 siblings, 0 replies; 19+ messages in thread
From: Luca Coelho @ 2020-10-08 15:12 UTC (permalink / raw)
To: kvalo; +Cc: linux-wireless
From: Sara Sharon <sara.sharon@intel.com>
mvm mutex isn't held by iwl_mvm_set_aes_rx_seq or it caller.
iee80211_local->key_mtx is held, but that is internal to mac80211.
The same applies to iwl_mvm_wowlan_program_keys.
Just hold rcu_read_lock, even though we are protected, the penalty
isn't that bad.
Warnings fixed are:
[ 4143.788196] WARNING: suspicious RCU usage
[ 4143.788211] -----------------------------
[ 4143.788220] suspicious rcu_dereference_protected() usage!
[ 4143.788227]
[ 4143.788234] rcu_scheduler_active = 2, debug_locks = 1
[ 4143.788242] 5 locks held by kworker/u8:9/5921:
[ 4143.788331] #4: ffff88804e69ad08 (&local->key_mtx){+.+.}, at ie80211_iter_keys+0x46/0x380 [mac80211]
[ 4143.788441]
[ 4143.788441] Call Trace:
[ 4143.788455] dump_stack+0xc1/0x11a
[ 4143.788471] lockdep_rcu_suspicious+0x14a/0x153
[ 4143.788515] iwl_mvm_set_aes_rx_seq+0x4a9/0x570 [iwlmvm]
[ 4143.788657] iwl_mvm_d3_update_keys+0x2ac/0x600 [iwlmvm]
[ 4143.788784] ieee80211_iter_keys+0x10e/0x380 [mac80211]
[ 4143.788838] iwl_mvm_setup_connection_keep+0x287/0x8d0 [iwlmvm]
[ 7243.206556] WARNING: suspicious RCU usage
[ 7243.206811] -----------------------------
[ 7243.206926] /suspicious rcu_dereference_protected() usage!
[ 7243.207086]
[ 7243.207204] rcu_scheduler_active = 2, debug_locks = 1
[ 7243.207321] 2 locks held by cat/15952:
[ 7243.207564] #1: ffff888008c8ad08 (&local->key_mtx){+.+.}, at:
ieee80211_iter_keys+0x46/0x380 [mac80211]
[ 7243.207751]
[ 7243.208094] Call Trace:
[ 7243.208211] dump_stack+0xc1/0x11a
[ 7243.208355] lockdep_rcu_suspicious+0x14a/0x153
[ 7243.208509] iwl_mvm_wowlan_program_keys+0x1db7/0x2340 [iwlmvm]
[ 7243.209852] ieee80211_iter_keys+0x10e/0x380 [mac80211]
Signed-off-by: Sara Sharon <sara.sharon@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
---
drivers/net/wireless/intel/iwlwifi/mvm/d3.c | 20 +++++++++++++-------
1 file changed, 13 insertions(+), 7 deletions(-)
diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/d3.c b/drivers/net/wireless/intel/iwlwifi/mvm/d3.c
index 9987172a9e01..d21143495e70 100644
--- a/drivers/net/wireless/intel/iwlwifi/mvm/d3.c
+++ b/drivers/net/wireless/intel/iwlwifi/mvm/d3.c
@@ -344,11 +344,12 @@ static void iwl_mvm_wowlan_program_keys(struct ieee80211_hw *hw,
const u8 *pn;
mvmsta = iwl_mvm_sta_from_mac80211(sta);
- ptk_pn = rcu_dereference_protected(
- mvmsta->ptk_pn[key->keyidx],
- lockdep_is_held(&mvm->mutex));
- if (WARN_ON(!ptk_pn))
+ rcu_read_lock();
+ ptk_pn = rcu_dereference(mvmsta->ptk_pn[key->keyidx]);
+ if (WARN_ON(!ptk_pn)) {
+ rcu_read_unlock();
break;
+ }
for (i = 0; i < IWL_MAX_TID_COUNT; i++) {
pn = iwl_mvm_find_max_pn(key, ptk_pn, &seq, i,
@@ -360,6 +361,8 @@ static void iwl_mvm_wowlan_program_keys(struct ieee80211_hw *hw,
((u64)pn[1] << 32) |
((u64)pn[0] << 40));
}
+
+ rcu_read_unlock();
} else {
for (i = 0; i < IWL_NUM_RSC; i++) {
u8 *pn = seq.ccmp.pn;
@@ -1363,10 +1366,12 @@ static void iwl_mvm_set_aes_rx_seq(struct iwl_mvm *mvm, struct aes_sc *scs,
mvmsta = iwl_mvm_sta_from_mac80211(sta);
- ptk_pn = rcu_dereference_protected(mvmsta->ptk_pn[key->keyidx],
- lockdep_is_held(&mvm->mutex));
- if (WARN_ON(!ptk_pn))
+ rcu_read_lock();
+ ptk_pn = rcu_dereference(mvmsta->ptk_pn[key->keyidx]);
+ if (WARN_ON(!ptk_pn)) {
+ rcu_read_unlock();
return;
+ }
for (tid = 0; tid < IWL_MAX_TID_COUNT; tid++) {
struct ieee80211_key_seq seq = {};
@@ -1378,6 +1383,7 @@ static void iwl_mvm_set_aes_rx_seq(struct iwl_mvm *mvm, struct aes_sc *scs,
memcpy(ptk_pn->q[i].pn[tid],
seq.ccmp.pn, IEEE80211_CCMP_PN_LEN);
}
+ rcu_read_unlock();
} else {
for (tid = 0; tid < IWL_NUM_RSC; tid++) {
struct ieee80211_key_seq seq = {};
--
2.28.0
^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 06/13] iwlwifi: stats: add new api fields for statistics cmd/ntfy
2020-10-08 15:12 [PATCH 00/13] iwlwifi: updates intended for v5.10 2020-10-08 part 2 Luca Coelho
` (4 preceding siblings ...)
2020-10-08 15:12 ` [PATCH 05/13] iwlwifi: mvm: fix suspicious rcu usage warnings Luca Coelho
@ 2020-10-08 15:12 ` Luca Coelho
2020-10-08 15:12 ` [PATCH 07/13] iwlwifi: fix sar geo table initialization Luca Coelho
` (6 subsequent siblings)
12 siblings, 0 replies; 19+ messages in thread
From: Luca Coelho @ 2020-10-08 15:12 UTC (permalink / raw)
To: kvalo; +Cc: linux-wireless
From: Mordechay Goodstein <mordechay.goodstein@intel.com>
The new API uses TLV format for statistics to enable fast changes
and debug on the fly.
Signed-off-by: Mordechay Goodstein <mordechay.goodstein@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
---
.../wireless/intel/iwlwifi/fw/api/commands.h | 6 +-
.../net/wireless/intel/iwlwifi/fw/api/stats.h | 461 ++++++++++++++++++
drivers/net/wireless/intel/iwlwifi/mvm/rx.c | 195 +++++---
3 files changed, 602 insertions(+), 60 deletions(-)
diff --git a/drivers/net/wireless/intel/iwlwifi/fw/api/commands.h b/drivers/net/wireless/intel/iwlwifi/fw/api/commands.h
index 14975a7e7203..8cc36dbb2311 100644
--- a/drivers/net/wireless/intel/iwlwifi/fw/api/commands.h
+++ b/drivers/net/wireless/intel/iwlwifi/fw/api/commands.h
@@ -411,7 +411,8 @@ enum iwl_legacy_cmds {
* one of &struct iwl_statistics_cmd,
* &struct iwl_notif_statistics_v11,
* &struct iwl_notif_statistics_v10,
- * &struct iwl_notif_statistics
+ * &struct iwl_notif_statistics,
+ * &struct iwl_statistics_operational_ntfy
*/
STATISTICS_CMD = 0x9c,
@@ -419,7 +420,8 @@ enum iwl_legacy_cmds {
* @STATISTICS_NOTIFICATION:
* one of &struct iwl_notif_statistics_v10,
* &struct iwl_notif_statistics_v11,
- * &struct iwl_notif_statistics
+ * &struct iwl_notif_statistic,
+ * &struct iwl_statistics_operational_ntfy
*/
STATISTICS_NOTIFICATION = 0x9d,
diff --git a/drivers/net/wireless/intel/iwlwifi/fw/api/stats.h b/drivers/net/wireless/intel/iwlwifi/fw/api/stats.h
index 35265c4e8c19..d41cab4016fe 100644
--- a/drivers/net/wireless/intel/iwlwifi/fw/api/stats.h
+++ b/drivers/net/wireless/intel/iwlwifi/fw/api/stats.h
@@ -464,4 +464,465 @@ struct iwl_statistics_cmd {
__le32 flags;
} __packed; /* STATISTICS_CMD_API_S_VER_1 */
+#define MAX_BCAST_FILTER_NUM 8
+
+/**
+ * enum iwl_fw_statistics_type
+ *
+ * @FW_STATISTICS_OPERATIONAL: operational statistics
+ * @FW_STATISTICS_PHY: phy statistics
+ * @FW_STATISTICS_MAC: mac statistics
+ * @FW_STATISTICS_RX: rx statistics
+ * @FW_STATISTICS_TX: tx statistics
+ * @FW_STATISTICS_DURATION: duration statistics
+ * @FW_STATISTICS_HE: he statistics
+ */
+enum iwl_fw_statistics_type {
+ FW_STATISTICS_OPERATIONAL,
+ FW_STATISTICS_PHY,
+ FW_STATISTICS_MAC,
+ FW_STATISTICS_RX,
+ FW_STATISTICS_TX,
+ FW_STATISTICS_DURATION,
+ FW_STATISTICS_HE,
+}; /* FW_STATISTICS_TYPE_API_E_VER_1 */
+
+/**
+ * struct iwl_statistics_ntfy_hdr
+ *
+ * @type: struct type
+ * @version: version of the struct
+ * @size: size in bytes
+ */
+struct iwl_statistics_ntfy_hdr {
+ u8 type;
+ u8 version;
+ __le16 size;
+}; /* STATISTICS_NTFY_HDR_API_S_VER_1 */
+
+/**
+ * struct iwl_statistics_operational_ntfy
+ *
+ * @hdr: general statistics header
+ * @flags: bitmap of possible notification structures
+ * @mac_id: mac on which the beacon was received
+ * @beacon_filter_average_energy: Average energy [-dBm] of the 2
+ * antennas.
+ * @beacon_filter_reason: beacon filter reason
+ * @radio_temperature: radio temperature
+ * @air_time: air time
+ * @beacon_counter: all beacons (both filtered and not filtered)
+ * @beacon_average_energy: all beacons (both filtered and not
+ * filtered)
+ * @beacon_rssi_a: beacon RSSI on antenna A
+ * @beacon_rssi_b: beacon RSSI on antenna B
+ * @rx_bytes: per MAC RX byte count
+ * @rx_time: rx time
+ * @tx_time: usec the radio is transmitting.
+ * @on_time_rf: The total time in usec the RF is awake.
+ * @on_time_scan: usec the radio is awake due to scan.
+ * @average_energy: in fact it is minus the energy..
+ * @reserved: reserved
+ */
+struct iwl_statistics_operational_ntfy {
+ struct iwl_statistics_ntfy_hdr hdr;
+ __le32 flags;
+ __le32 mac_id;
+ __le32 beacon_filter_average_energy;
+ __le32 beacon_filter_reason;
+ __le32 radio_temperature;
+ __le32 air_time[MAC_INDEX_AUX];
+ __le32 beacon_counter[MAC_INDEX_AUX];
+ __le32 beacon_average_energy[MAC_INDEX_AUX];
+ __le32 beacon_rssi_a;
+ __le32 beacon_rssi_b;
+ __le32 rx_bytes[MAC_INDEX_AUX];
+ __le64 rx_time;
+ __le64 tx_time;
+ __le64 on_time_rf;
+ __le64 on_time_scan;
+ __le32 average_energy[IWL_MVM_STATION_COUNT_MAX];
+ __le32 reserved;
+} __packed; /* STATISTICS_OPERATIONAL_NTFY_API_S_VER_14 */
+
+/**
+ * struct iwl_statistics_phy_ntfy
+ *
+ * @hdr: general statistics header
+ * RX PHY related statistics
+ * @energy_and_config: ???
+ * @rssi_band: @31:24 rssiAllBand_B, 23:16 rssiInBand_B, 15:8
+ * rssiAllBand_A, 7:0 rssiInBand_A
+ * @agc_word: @31:16 agcWord_B, 15:0 agcWord_A
+ * @agc_gain: @19:10 agcGain_B, 9:0 agcGain_A
+ * @dfe_gain: @19:10 dfeGain_B, 9:0 dfeGain_A
+ * @snr_calc_main: @18:0 snrCalcMain
+ * @energy_calc_main: @18:0 energyCalcMain
+ * @snr_calc_aux: @18:0 snrCalcAux
+ * @dsp_dc_estim_a: @27:14 dspDcEstimQA, 13:0 dspDcEstimIA
+ * @dsp_dc_estim_b: @27:14 dspDcEstimQB, 13:0 dspDcEstimIB
+ * @ina_detec_type_and_ofdm_corr_comb: @31:31 inaDetectCckMrc,
+ * 30:27 inaDetectType, 26:0 ofdmCorrComb
+ * @cw_corr_comb: @26:0 cwCorrComb
+ * @rssi_comb: @25:0 rssiComb
+ * @auto_corr_cck: @23:12 autoCck, 11:00 crossCck
+ * @ofdm_fine_freq_and_pina_freq_err: @18:7 ofdmFineFreq, 6:0
+ * ofdmPinaFreqErr
+ * @snrm_evm_main: @31:0 snrmEvmMain
+ * @snrm_evm_aux: @31:0 snrmEvmAux
+ * @rx_rate: @31:0 rate
+ * TX PHY related statistics
+ * @per_chain_enums_and_dsp_atten_a: @perChainEnumsAndDspAtten
+ * (per version)
+ * @target_power_and_power_meas_a: @31:16 targetPower_A, 15:0
+ * powerMeasuredCalc_A
+ * @tx_config_as_i_and_ac_a: @31:16 txConfigAsI_A, 15:0
+ * txConfigAc_A
+ * @predist_dcq_and_dci_a: @31:16 predist_dci_A, 15:0
+ * predist_dcq_A
+ * @per_chain_enums_and_dsp_atten_b: @perChainEnumsAndDspAtten
+ * (per version)
+ * @target_power_and_power_meas_b: @31:16 targetPower_B, 15:0
+ * powerMeasuredCalc_B
+ * @tx_config_as_i_and_ac_b: @31:16 txConfigAsI_B, 15:0
+ * txConfigAc_B
+ * @predist_dcq_and_dci_b: @31:16 predist_dci_B, 15:0
+ * predist_dcq_B
+ * @tx_rate: @31:0 rate
+ * @tlc_backoff: @31:0 tlcBackoff
+ * @mpapd_calib_mode_mpapd_calib_type_a: @31:16
+ * mpapdCalibMode_A, 15:0 mpapdCalibType_A
+ * @psat_and_phy_power_limit_a: @31:16 psat_A, 15:0
+ * phyPowerLimit_A
+ * @sar_and_regulatory_power_limit_a: @31:16 sarPowerLimit_A,
+ * 15:0 regulatoryPowerLimit_A
+ * @mpapd_calib_mode_mpapd_calib_type_b: @31:16
+ * mpapdCalibMode_B, 15:0 mpapdCalibType_B
+ * @psat_and_phy_power_limit_b: @31:16 psat_B, 15:0
+ * phyPowerLimit_B
+ * @sar_and_regulatory_power_limit_b: @31:16 sarPowerLimit_B,
+ * 15:0 regulatoryPowerLimit_B
+ * @srd_and_driver_power_limits: @31:16 srdPowerLimit, 15:0
+ * driverPowerLimit
+ * @reserved: reserved
+ */
+struct iwl_statistics_phy_ntfy {
+ struct iwl_statistics_ntfy_hdr hdr;
+ __le32 energy_and_config;
+ __le32 rssi_band;
+ __le32 agc_word;
+ __le32 agc_gain;
+ __le32 dfe_gain;
+ __le32 snr_calc_main;
+ __le32 energy_calc_main;
+ __le32 snr_calc_aux;
+ __le32 dsp_dc_estim_a;
+ __le32 dsp_dc_estim_b;
+ __le32 ina_detec_type_and_ofdm_corr_comb;
+ __le32 cw_corr_comb;
+ __le32 rssi_comb;
+ __le32 auto_corr_cck;
+ __le32 ofdm_fine_freq_and_pina_freq_err;
+ __le32 snrm_evm_main;
+ __le32 snrm_evm_aux;
+ __le32 rx_rate;
+ __le32 per_chain_enums_and_dsp_atten_a;
+ __le32 target_power_and_power_meas_a;
+ __le32 tx_config_as_i_and_ac_a;
+ __le32 predist_dcq_and_dci_a;
+ __le32 per_chain_enums_and_dsp_atten_b;
+ __le32 target_power_and_power_meas_b;
+ __le32 tx_config_as_i_and_ac_b;
+ __le32 predist_dcq_and_dci_b;
+ __le32 tx_rate;
+ __le32 tlc_backoff;
+ __le32 mpapd_calib_mode_mpapd_calib_type_a;
+ __le32 psat_and_phy_power_limit_a;
+ __le32 sar_and_regulatory_power_limit_a;
+ __le32 mpapd_calib_mode_mpapd_calib_type_b;
+ __le32 psat_and_phy_power_limit_b;
+ __le32 sar_and_regulatory_power_limit_b;
+ __le32 srd_and_driver_power_limits;
+ __le32 reserved;
+} __packed; /* STATISTICS_PHY_NTFY_API_S_VER_1 */
+
+/**
+ * struct iwl_statistics_mac_ntfy
+ *
+ * @hdr: general statistics header
+ * @bcast_filter_passed_per_mac: bcast filter passed per mac
+ * @bcast_filter_dropped_per_mac: bcast filter dropped per mac
+ * @bcast_filter_passed_per_filter: bcast filter passed per filter
+ * @bcast_filter_dropped_per_filter: bcast filter dropped per filter
+ * @reserved: reserved
+ */
+struct iwl_statistics_mac_ntfy {
+ struct iwl_statistics_ntfy_hdr hdr;
+ __le32 bcast_filter_passed_per_mac[NUM_MAC_INDEX_CDB];
+ __le32 bcast_filter_dropped_per_mac[NUM_MAC_INDEX_CDB];
+ __le32 bcast_filter_passed_per_filter[MAX_BCAST_FILTER_NUM];
+ __le32 bcast_filter_dropped_per_filter[MAX_BCAST_FILTER_NUM];
+ __le32 reserved;
+} __packed; /* STATISTICS_MAC_NTFY_API_S_VER_1 */
+
+/**
+ * struct iwl_statistics_rx_ntfy
+ *
+ * @hdr: general statistics header
+ * @rx_agg_mpdu_cnt: aggregation frame count (number of
+ * delimiters)
+ * @rx_agg_cnt: number of RX Aggregations
+ * @unsupported_mcs: number of PLCP headers that have rate which
+ * is unsupported by DSP
+ * @bogus_cts: CTS received when not expecting CTS
+ * @bogus_ack: ACK received when not expecting ACK
+ * @rx_byte_count: ???
+ * @rx_packet_count: ???
+ * @missed_beacons: ???
+ * @unresponded_rts: un-responded RTS, due to NAV not zero
+ * @rxe_frame_limit_overrun: RXE got frame limit overrun
+ * @sent_ba_rsp_cnt: BA response TX count
+ * @late_rx_handle: count the number of times the RX path was
+ * aborted due to late entry
+ * @num_bt_kills: ???
+ * @reserved: reserved
+ */
+struct iwl_statistics_rx_ntfy {
+ struct iwl_statistics_ntfy_hdr hdr;
+ __le32 rx_agg_mpdu_cnt;
+ __le32 rx_agg_cnt;
+ __le32 unsupported_mcs;
+ __le32 bogus_cts;
+ __le32 bogus_ack;
+ __le32 rx_byte_count[MAC_INDEX_AUX];
+ __le32 rx_packet_count[MAC_INDEX_AUX];
+ __le32 missed_beacons;
+ __le32 unresponded_rts;
+ __le32 rxe_frame_limit_overrun;
+ __le32 sent_ba_rsp_cnt;
+ __le32 late_rx_handle;
+ __le32 num_bt_kills;
+ __le32 reserved;
+} __packed; /* STATISTICS_RX_NTFY_API_S_VER_1 */
+
+/**
+ * struct iwl_statistics_tx_ntfy
+ *
+ * @hdr: general statistics header
+ * @cts_timeout: timeout when waiting for CTS
+ * @ack_timeout: timeout when waiting for ACK
+ * @dump_msdu_cnt: number of MSDUs that were dumped due to any
+ * reason
+ * @burst_abort_missing_next_frame_cnt: number of times a burst
+ * was aborted due to missing next frame bytes in txfifo
+ * number of times got timeout when waiting for CTS/ACK/BA and energy was
+ * detected just after sending the RTS/DATA. this statistics may help getting
+ * interesting indicators, like the likelihood of collision (so the benefit of
+ * protection may be estimated Vs. its cost). Or how many of the failures are
+ * due to collision and how many due to SNR.
+ * For Link-quality the CTS collision indication is more reliable then the ACK
+ * collision indication as the RTS frame is short and has more chance that the
+ * frame/s which caused the collision continue after the RTS was sent.
+ * @cts_timeout_collision: ???
+ * ACK/BA failed and energy as detected after DATA
+ * Note: to get the collision ratio need to:
+ * ackOrBaTimeoutCollision / (ack_timeout + ba_timeout)
+ * @ack_or_ba_timeout_collision: ???
+ * @ba_timeout: timeout when waiting for immediate BA response
+ * @ba_reschedule_frames: failed to get BA response and
+ * rescheduled all the non-ACKed frames
+ * gives the avarage number of frames inside aggregation
+ * @scd_query_agg_frame_cnt: ???
+ * @scd_query_no_agg: scheduler query prevented aggregation
+ * @scd_query_agg: scheduler query allowed aggregation
+ * @scd_query_mismatch: scheduler query inaccurate, either too
+ * short or too long
+ * @agg_terminated_underrun: aggregation was terminated due to
+ * underrun
+ * @agg_terminated_bt_prio_kill: aggregation was terminated due
+ * to BT
+ * @tx_kill_on_long_retry: count the tx frames dropped due to
+ * long retry limit (DATA frame failed)
+ * @tx_kill_on_short_retry: count the tx frames dropped due to
+ * short retry limit (RTS frame failed)
+ * TX deffer on energy. This counter is reset on each successful transmit.
+ * When timer exceed TX deffer limit than will be uCode assert.
+ * @tx_deffer_counter: ???
+ * @tx_deffer_base_time: Keep the time of the last successful
+ * transmit
+ * @tx_underrun: TX killed due to underrun
+ * @bt_defer: TX deferred due to BT priority, so probably TX was
+ * not started.
+ * @tx_kill_on_dsp_timeout: TX killed on DSP problem detected
+ * @tx_kill_on_immediate_quiet: TX killed due to immediate quiet
+ * @kill_ba_cnt: number of times sending BA failed
+ * @kill_ack_cnt: number of times sending ACK failed
+ * @kill_cts_cnt: number of times sending CTS failed
+ * @burst_terminated: Count burst or fragmentation termination
+ * occurrence
+ * @late_tx_vec_wr_cnt: ???
+ * TX is not sent because ucode failed to notify the TRM in SIFS-delta from
+ * ON_AIR deassertion.
+ * @late_rx2_tx_cnt: ???
+ * @scd_query_cnt: count the times SCD query was done to check
+ * for TX AGG
+ * @tx_frames_acked_in_agg: count the number of frames
+ * transmitted inside AGG and were successful
+ * @last_tx_ch_width_indx: ???
+ * number of deferred TX per channel width, 0 - 20, 1/2/3 - 40/80/160
+ * @rx_detected_per_ch_width: ???
+ * @success_per_ch_width: ???
+ * @fail_per_ch_width: ???
+ * @reserved: reserved
+ */
+struct iwl_statistics_tx_ntfy {
+ struct iwl_statistics_ntfy_hdr hdr;
+ __le32 cts_timeout;
+ __le32 ack_timeout;
+ __le32 dump_msdu_cnt;
+ __le32 burst_abort_missing_next_frame_cnt;
+ __le32 cts_timeout_collision;
+ __le32 ack_or_ba_timeout_collision;
+ __le32 ba_timeout;
+ __le32 ba_reschedule_frames;
+ __le32 scd_query_agg_frame_cnt;
+ __le32 scd_query_no_agg;
+ __le32 scd_query_agg;
+ __le32 scd_query_mismatch;
+ __le32 agg_terminated_underrun;
+ __le32 agg_terminated_bt_prio_kill;
+ __le32 tx_kill_on_long_retry;
+ __le32 tx_kill_on_short_retry;
+ __le32 tx_deffer_counter;
+ __le32 tx_deffer_base_time;
+ __le32 tx_underrun;
+ __le32 bt_defer;
+ __le32 tx_kill_on_dsp_timeout;
+ __le32 tx_kill_on_immediate_quiet;
+ __le32 kill_ba_cnt;
+ __le32 kill_ack_cnt;
+ __le32 kill_cts_cnt;
+ __le32 burst_terminated;
+ __le32 late_tx_vec_wr_cnt;
+ __le32 late_rx2_tx_cnt;
+ __le32 scd_query_cnt;
+ __le32 tx_frames_acked_in_agg;
+ __le32 last_tx_ch_width_indx;
+ __le32 rx_detected_per_ch_width[4];
+ __le32 success_per_ch_width[4];
+ __le32 fail_per_ch_width[4];
+ __le32 reserved;
+} __packed; /* STATISTICS_TX_NTFY_API_S_VER_1 */
+
+/**
+ * struct iwl_statistics_duration_ntfy
+ *
+ * @hdr: general statistics header
+ * @cont_burst_chk_cnt: number of times continuation or
+ * fragmentation or bursting was checked
+ * @cont_burst_cnt: number of times continuation or fragmentation
+ * or bursting was successful
+ * @wait_for_silence_timeout_cnt: ???
+ * @reserved: reserved
+ */
+struct iwl_statistics_duration_ntfy {
+ struct iwl_statistics_ntfy_hdr hdr;
+ __le32 cont_burst_chk_cnt;
+ __le32 cont_burst_cnt;
+ __le32 wait_for_silence_timeout_cnt;
+ __le32 reserved;
+} __packed; /* STATISTICS_DURATION_NTFY_API_S_VER_1 */
+
+/**
+ * struct iwl_statistics_he_ntfy
+ *
+ * @hdr: general statistics header
+ * received HE frames
+ * @rx_siga_valid_cnt: rx HE SIG-A valid
+ * @rx_siga_invalid_cnt: rx HE SIG-A invalid
+ * received HE frames w/ valid Sig-A
+ * @rx_trig_based_frame_cnt: rx HE-TB (trig-based)
+ * @rx_su_frame_cnt: rx HE-SU
+ * @rx_sigb_invalid_cnt: rx (suspected) HE-MU w/ bad SIG-B
+ * @rx_our_bss_color_cnt: rx valid HE SIG-A w/ our BSS color
+ * @rx_other_bss_color_cnt: rx valid HE SIG-A w/ other BSS color
+ * @rx_zero_bss_color_cnt: ???
+ * received HE-MU frames w/ good Sig-B
+ * @rx_mu_for_us_cnt: match AID
+ * @rx_mu_not_for_us_cnt: no matched AID
+ * received HE-MU frames for us (w/ our AID)
+ * @rx_mu_nss_ar: 0 - SISO, 1 - MIMO2
+ * @rx_mu_mimo_cnt: full BW RU, compressed SIG-B
+ * @rx_mu_ru_bw_ar: MU alloc, MHz: 0 - 2, 1 - 5, 2 - 10, 3 - 20,
+ * 4 - 40, 5 - 80, 6 - 160
+ * received trigger frames
+ * @rx_trig_for_us_cnt: ???
+ * @rx_trig_not_for_us_cnt: ???
+ * trigger for us
+ * @rx_trig_with_cs_req_cnt: ???
+ * @rx_trig_type_ar: ???
+ * @rx_trig_in_agg_cnt: ???
+ * basic trigger for us allocations
+ * @rx_basic_trig_alloc_nss_ar: ???
+ * @rx_basic_trig_alloc_mu_mimo_cnt: ???
+ * @rx_basic_trig_alloc_ru_bw_ar: ???
+ * @rx_basic_trig_total_byte_cnt: ???
+ * trig-based TX
+ * @tx_trig_based_cs_req_fail_cnt: ???
+ * @tx_trig_based_sifs_ok_cnt: ???
+ * @tx_trig_based_sifs_fail_cnt: ???
+ * @tx_trig_based_byte_cnt: ???
+ * @tx_trig_based_pad_byte_cnt: ???
+ * @tx_trig_based_frame_cnt: ???
+ * @tx_trig_based_acked_frame_cnt: ???
+ * @tx_trig_based_ack_timeout_cnt: ???
+ * HE-SU TX
+ * @tx_su_frame_cnt: ???
+ * EDCA <--> MU-EDCA transitions
+ * @tx_edca_to_mu_edca_cnt: ???
+ * @tx_mu_edca_to_edca_by_timeout_cnt: ???
+ * @tx_mu_edca_to_edca_by_ack_fail_cnt: ???
+ * @tx_mu_edca_to_edca_by_small_alloc_cnt: ???
+ * @reserved: reserved
+ */
+struct iwl_statistics_he_ntfy {
+ struct iwl_statistics_ntfy_hdr hdr;
+ __le32 rx_siga_valid_cnt;
+ __le32 rx_siga_invalid_cnt;
+ __le32 rx_trig_based_frame_cnt;
+ __le32 rx_su_frame_cnt;
+ __le32 rx_sigb_invalid_cnt;
+ __le32 rx_our_bss_color_cnt;
+ __le32 rx_other_bss_color_cnt;
+ __le32 rx_zero_bss_color_cnt;
+ __le32 rx_mu_for_us_cnt;
+ __le32 rx_mu_not_for_us_cnt;
+ __le32 rx_mu_nss_ar[2];
+ __le32 rx_mu_mimo_cnt;
+ __le32 rx_mu_ru_bw_ar[7];
+ __le32 rx_trig_for_us_cnt;
+ __le32 rx_trig_not_for_us_cnt;
+ __le32 rx_trig_with_cs_req_cnt;
+ __le32 rx_trig_type_ar[8 + 1];
+ __le32 rx_trig_in_agg_cnt;
+ __le32 rx_basic_trig_alloc_nss_ar[2];
+ __le32 rx_basic_trig_alloc_mu_mimo_cnt;
+ __le32 rx_basic_trig_alloc_ru_bw_ar[7];
+ __le32 rx_basic_trig_total_byte_cnt;
+ __le32 tx_trig_based_cs_req_fail_cnt;
+ __le32 tx_trig_based_sifs_ok_cnt;
+ __le32 tx_trig_based_sifs_fail_cnt;
+ __le32 tx_trig_based_byte_cnt;
+ __le32 tx_trig_based_pad_byte_cnt;
+ __le32 tx_trig_based_frame_cnt;
+ __le32 tx_trig_based_acked_frame_cnt;
+ __le32 tx_trig_based_ack_timeout_cnt;
+ __le32 tx_su_frame_cnt;
+ __le32 tx_edca_to_mu_edca_cnt;
+ __le32 tx_mu_edca_to_edca_by_timeout_cnt;
+ __le32 tx_mu_edca_to_edca_by_ack_fail_cnt;
+ __le32 tx_mu_edca_to_edca_by_small_alloc_cnt;
+ __le32 reserved;
+} __packed; /* STATISTICS_HE_NTFY_API_S_VER_1 */
+
#endif /* __iwl_fw_api_stats_h__ */
diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/rx.c b/drivers/net/wireless/intel/iwlwifi/mvm/rx.c
index fe555986d863..0059c83c2783 100644
--- a/drivers/net/wireless/intel/iwlwifi/mvm/rx.c
+++ b/drivers/net/wireless/intel/iwlwifi/mvm/rx.c
@@ -569,7 +569,8 @@ struct iwl_mvm_stat_data {
__le32 flags;
__le32 mac_id;
u8 beacon_filter_average_energy;
- void *general;
+ __le32 *beacon_counter;
+ u8 *beacon_average_energy;
};
static void iwl_mvm_stat_iterator(void *_data, u8 *mac,
@@ -589,23 +590,10 @@ static void iwl_mvm_stat_iterator(void *_data, u8 *mac,
* data copied into the "data" struct, but rather the data from
* the notification directly.
*/
- if (iwl_mvm_has_new_rx_stats_api(mvm)) {
- struct mvm_statistics_general *general =
- data->general;
-
- mvmvif->beacon_stats.num_beacons =
- le32_to_cpu(general->beacon_counter[vif_id]);
- mvmvif->beacon_stats.avg_signal =
- -general->beacon_average_energy[vif_id];
- } else {
- struct mvm_statistics_general_v8 *general =
- data->general;
-
- mvmvif->beacon_stats.num_beacons =
- le32_to_cpu(general->beacon_counter[vif_id]);
- mvmvif->beacon_stats.avg_signal =
- -general->beacon_average_energy[vif_id];
- }
+ mvmvif->beacon_stats.num_beacons =
+ le32_to_cpu(data->beacon_counter[vif_id]);
+ mvmvif->beacon_stats.avg_signal =
+ -data->beacon_average_energy[vif_id];
/* make sure that beacon statistics don't go backwards with TCM
* request to clear statistics
@@ -701,18 +689,136 @@ iwl_mvm_rx_stats_check_trigger(struct iwl_mvm *mvm, struct iwl_rx_packet *pkt)
iwl_fw_dbg_collect_trig(&mvm->fwrt, trig, NULL);
}
+static void iwl_mvm_update_avg_energy(struct iwl_mvm *mvm,
+ u8 energy[IWL_MVM_STATION_COUNT_MAX])
+{
+ int i;
+
+ if (WARN_ONCE(mvm->fw->ucode_capa.num_stations >
+ IWL_MVM_STATION_COUNT_MAX,
+ "Driver and FW station count mismatch %d\n",
+ mvm->fw->ucode_capa.num_stations))
+ return;
+
+ rcu_read_lock();
+ for (i = 0; i < mvm->fw->ucode_capa.num_stations; i++) {
+ struct iwl_mvm_sta *sta;
+
+ if (!energy[i])
+ continue;
+
+ sta = iwl_mvm_sta_from_staid_rcu(mvm, i);
+ if (!sta)
+ continue;
+ sta->avg_energy = energy[i];
+ }
+ rcu_read_unlock();
+}
+
+static void
+iwl_mvm_update_tcm_from_stats(struct iwl_mvm *mvm, __le32 *air_time_le,
+ __le32 *rx_bytes_le)
+{
+ int i;
+
+ spin_lock(&mvm->tcm.lock);
+ for (i = 0; i < NUM_MAC_INDEX_DRIVER; i++) {
+ struct iwl_mvm_tcm_mac *mdata = &mvm->tcm.data[i];
+ u32 rx_bytes = le32_to_cpu(rx_bytes_le[i]);
+ u32 airtime = le32_to_cpu(air_time_le[i]);
+
+ mdata->rx.airtime += airtime;
+ mdata->uapsd_nonagg_detect.rx_bytes += rx_bytes;
+ if (airtime) {
+ /* re-init every time to store rate from FW */
+ ewma_rate_init(&mdata->uapsd_nonagg_detect.rate);
+ ewma_rate_add(&mdata->uapsd_nonagg_detect.rate,
+ rx_bytes * 8 / airtime);
+ }
+ }
+ spin_unlock(&mvm->tcm.lock);
+}
+
+static void
+iwl_mvm_handle_rx_statistics_tlv(struct iwl_mvm *mvm,
+ struct iwl_rx_packet *pkt)
+{
+ struct iwl_mvm_stat_data data = {
+ .mvm = mvm,
+ };
+ u8 beacon_average_energy[MAC_INDEX_AUX];
+ u8 average_energy[IWL_MVM_STATION_COUNT_MAX];
+ struct iwl_statistics_operational_ntfy *stats;
+ int expected_size;
+ __le32 flags;
+ int i;
+
+ expected_size = sizeof(*stats);
+ if (WARN_ONCE(iwl_rx_packet_payload_len(pkt) < expected_size,
+ "received invalid statistics size (%d)!, expected_size: %d\n",
+ iwl_rx_packet_payload_len(pkt), expected_size))
+ return;
+
+ stats = (void *)&pkt->data;
+
+ if (WARN_ONCE(stats->hdr.type != FW_STATISTICS_OPERATIONAL ||
+ stats->hdr.version != 1,
+ "received unsupported hdr type %d, version %d\n",
+ stats->hdr.type, stats->hdr.version))
+ return;
+
+ flags = stats->flags;
+ mvm->radio_stats.rx_time = le64_to_cpu(stats->rx_time);
+ mvm->radio_stats.tx_time = le64_to_cpu(stats->tx_time);
+ mvm->radio_stats.on_time_rf = le64_to_cpu(stats->on_time_rf);
+ mvm->radio_stats.on_time_scan = le64_to_cpu(stats->on_time_scan);
+
+ iwl_mvm_rx_stats_check_trigger(mvm, pkt);
+
+ data.mac_id = stats->mac_id;
+ data.beacon_filter_average_energy =
+ le32_to_cpu(stats->beacon_filter_average_energy);
+ data.flags = flags;
+ data.beacon_counter = stats->beacon_counter;
+ for (i = 0; i < ARRAY_SIZE(beacon_average_energy); i++)
+ beacon_average_energy[i] =
+ le32_to_cpu(stats->beacon_average_energy[i]);
+
+ data.beacon_average_energy = beacon_average_energy;
+
+ ieee80211_iterate_active_interfaces(mvm->hw,
+ IEEE80211_IFACE_ITER_NORMAL,
+ iwl_mvm_stat_iterator,
+ &data);
+
+ for (i = 0; i < ARRAY_SIZE(average_energy); i++)
+ average_energy[i] = le32_to_cpu(stats->average_energy[i]);
+ iwl_mvm_update_avg_energy(mvm, average_energy);
+
+ /*
+ * Don't update in case the statistics are not cleared, since
+ * we will end up counting twice the same airtime, once in TCM
+ * request and once in statistics notification.
+ */
+ if (le32_to_cpu(flags) & IWL_STATISTICS_REPLY_FLG_CLEAR)
+ iwl_mvm_update_tcm_from_stats(mvm, stats->air_time,
+ stats->rx_bytes);
+}
+
void iwl_mvm_handle_rx_statistics(struct iwl_mvm *mvm,
struct iwl_rx_packet *pkt)
{
struct iwl_mvm_stat_data data = {
.mvm = mvm,
};
+ __le32 *bytes, *air_time, flags;
int expected_size;
- int i;
u8 *energy;
- __le32 *bytes;
- __le32 *air_time;
- __le32 flags;
+
+ /* From ver 14 and up we use TLV statistics format */
+ if (iwl_fw_lookup_notif_ver(mvm->fw, LONG_GROUP,
+ STATISTICS_CMD, 0) >= 14)
+ return iwl_mvm_handle_rx_statistics_tlv(mvm, pkt);
if (!iwl_mvm_has_new_rx_stats_api(mvm)) {
if (iwl_mvm_has_new_rx_api(mvm))
@@ -746,8 +852,9 @@ void iwl_mvm_handle_rx_statistics(struct iwl_mvm *mvm,
mvm->radio_stats.on_time_scan =
le64_to_cpu(stats->general.common.on_time_scan);
- data.general = &stats->general;
-
+ data.beacon_counter = stats->general.beacon_counter;
+ data.beacon_average_energy =
+ stats->general.beacon_average_energy;
flags = stats->flag;
} else {
struct iwl_notif_statistics *stats = (void *)&pkt->data;
@@ -767,8 +874,9 @@ void iwl_mvm_handle_rx_statistics(struct iwl_mvm *mvm,
mvm->radio_stats.on_time_scan =
le64_to_cpu(stats->general.common.on_time_scan);
- data.general = &stats->general;
-
+ data.beacon_counter = stats->general.beacon_counter;
+ data.beacon_average_energy =
+ stats->general.beacon_average_energy;
flags = stats->flag;
}
data.flags = flags;
@@ -797,45 +905,16 @@ void iwl_mvm_handle_rx_statistics(struct iwl_mvm *mvm,
air_time = (void *)&stats->load_stats.air_time;
}
- rcu_read_lock();
- for (i = 0; i < mvm->fw->ucode_capa.num_stations; i++) {
- struct iwl_mvm_sta *sta;
-
- if (!energy[i])
- continue;
-
- sta = iwl_mvm_sta_from_staid_rcu(mvm, i);
- if (!sta)
- continue;
- sta->avg_energy = energy[i];
- }
- rcu_read_unlock();
+ iwl_mvm_update_avg_energy(mvm, energy);
/*
* Don't update in case the statistics are not cleared, since
* we will end up counting twice the same airtime, once in TCM
* request and once in statistics notification.
*/
- if (!(le32_to_cpu(flags) & IWL_STATISTICS_REPLY_FLG_CLEAR))
- return;
+ if (le32_to_cpu(flags) & IWL_STATISTICS_REPLY_FLG_CLEAR)
+ iwl_mvm_update_tcm_from_stats(mvm, air_time, bytes);
- spin_lock(&mvm->tcm.lock);
- for (i = 0; i < NUM_MAC_INDEX_DRIVER; i++) {
- struct iwl_mvm_tcm_mac *mdata = &mvm->tcm.data[i];
- u32 airtime = le32_to_cpu(air_time[i]);
- u32 rx_bytes = le32_to_cpu(bytes[i]);
-
- mdata->uapsd_nonagg_detect.rx_bytes += rx_bytes;
- if (airtime) {
- /* re-init every time to store rate from FW */
- ewma_rate_init(&mdata->uapsd_nonagg_detect.rate);
- ewma_rate_add(&mdata->uapsd_nonagg_detect.rate,
- rx_bytes * 8 / airtime);
- }
-
- mdata->rx.airtime += airtime;
- }
- spin_unlock(&mvm->tcm.lock);
}
void iwl_mvm_rx_statistics(struct iwl_mvm *mvm, struct iwl_rx_cmd_buffer *rxb)
--
2.28.0
^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 07/13] iwlwifi: fix sar geo table initialization
2020-10-08 15:12 [PATCH 00/13] iwlwifi: updates intended for v5.10 2020-10-08 part 2 Luca Coelho
` (5 preceding siblings ...)
2020-10-08 15:12 ` [PATCH 06/13] iwlwifi: stats: add new api fields for statistics cmd/ntfy Luca Coelho
@ 2020-10-08 15:12 ` Luca Coelho
2020-10-08 15:12 ` [PATCH 08/13] iwlwifi: rs: align to new TLC config command API Luca Coelho
` (5 subsequent siblings)
12 siblings, 0 replies; 19+ messages in thread
From: Luca Coelho @ 2020-10-08 15:12 UTC (permalink / raw)
To: kvalo; +Cc: linux-wireless
From: Naftali Goldstein <naftali.goldstein@intel.com>
When adding support for version 3 of the GEO_TX_POWER_LIMIT command,
the table argument of iwl_sar_geo_init was changed from a pointer a 1d
array of now-removed iwl_per_chain_offset_group_v1 to a pointer to a 2d
array of iwl_per_chain_offset (iwl_per_chain_offset_group_v1 was a
struct containing 2 copies of iwl_per_chain_offset).
So even for version 2 where the second dimension is of length 2, which
means that the underlying memory layout of the array didn't change, this
requires a small change in the way we loop over it, and this was missed.
Additionally, for the case of version 3 where the second dimension is now
3, in order to fill the first two elements of each row correctly (lb and
hb), iwl_sar_geo_init must get the true number of bands supported.
But because we don't yet store any values for the 3rd (uhb) band, skip
that band.
Signed-off-by: Naftali Goldstein <naftali.goldstein@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
---
drivers/net/wireless/intel/iwlwifi/fw/acpi.c | 20 ++++++++----
drivers/net/wireless/intel/iwlwifi/fw/acpi.h | 1 -
drivers/net/wireless/intel/iwlwifi/mvm/fw.c | 33 +++++++++++++-------
3 files changed, 36 insertions(+), 18 deletions(-)
diff --git a/drivers/net/wireless/intel/iwlwifi/fw/acpi.c b/drivers/net/wireless/intel/iwlwifi/fw/acpi.c
index e93656e461ea..3e5a35e26ad3 100644
--- a/drivers/net/wireless/intel/iwlwifi/fw/acpi.c
+++ b/drivers/net/wireless/intel/iwlwifi/fw/acpi.c
@@ -662,17 +662,25 @@ int iwl_sar_geo_init(struct iwl_fw_runtime *fwrt,
}
for (i = 0; i < ACPI_NUM_GEO_PROFILES; i++) {
- struct iwl_per_chain_offset *chain =
- (struct iwl_per_chain_offset *)&table[i];
-
for (j = 0; j < n_bands; j++) {
+ struct iwl_per_chain_offset *chain =
+ &table[i * n_bands + j];
u8 *value;
+ if (j * ACPI_GEO_PER_CHAIN_SIZE >=
+ ARRAY_SIZE(fwrt->geo_profiles[0].values))
+ /*
+ * Currently we only store lb an hb values, and
+ * don't have any special ones for uhb. So leave
+ * those empty for the time being
+ */
+ break;
+
value = &fwrt->geo_profiles[i].values[j *
ACPI_GEO_PER_CHAIN_SIZE];
- chain[j].max_tx_power = cpu_to_le16(value[0]);
- chain[j].chain_a = value[1];
- chain[j].chain_b = value[2];
+ chain->max_tx_power = cpu_to_le16(value[0]);
+ chain->chain_a = value[1];
+ chain->chain_b = value[2];
IWL_DEBUG_RADIO(fwrt,
"SAR geographic profile[%d] Band[%d]: chain A = %d chain B = %d max_tx_power = %d\n",
i, j, value[1], value[2], value[0]);
diff --git a/drivers/net/wireless/intel/iwlwifi/fw/acpi.h b/drivers/net/wireless/intel/iwlwifi/fw/acpi.h
index 71ef8647d7b5..bddf8a44e163 100644
--- a/drivers/net/wireless/intel/iwlwifi/fw/acpi.h
+++ b/drivers/net/wireless/intel/iwlwifi/fw/acpi.h
@@ -105,7 +105,6 @@
#define APCI_WTAS_BLACK_LIST_MAX 16
#define ACPI_WTAS_WIFI_DATA_SIZE (3 + APCI_WTAS_BLACK_LIST_MAX)
-#define ACPI_WGDS_NUM_BANDS 2
#define ACPI_WGDS_TABLE_SIZE 3
#define ACPI_PPAG_WIFI_DATA_SIZE ((IWL_NUM_CHAIN_LIMITS * \
diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/fw.c b/drivers/net/wireless/intel/iwlwifi/mvm/fw.c
index 80b52010d47b..9bce25a15df6 100644
--- a/drivers/net/wireless/intel/iwlwifi/mvm/fw.c
+++ b/drivers/net/wireless/intel/iwlwifi/mvm/fw.c
@@ -923,36 +923,47 @@ static int iwl_mvm_sar_geo_init(struct iwl_mvm *mvm)
{
union iwl_geo_tx_power_profiles_cmd cmd;
u16 len;
+ u32 n_bands;
int ret;
u8 cmd_ver = iwl_fw_lookup_cmd_ver(mvm->fw, PHY_OPS_GROUP,
GEO_TX_POWER_LIMIT,
IWL_FW_CMD_VER_UNKNOWN);
- /* the table is also at the same position both in v1 and v2 */
- ret = iwl_sar_geo_init(&mvm->fwrt, &cmd.v1.table[0][0],
- ACPI_WGDS_NUM_BANDS);
-
- /*
- * It is a valid scenario to not support SAR, or miss wgds table,
- * but in that case there is no need to send the command.
- */
- if (ret)
- return 0;
-
+ BUILD_BUG_ON(offsetof(struct iwl_geo_tx_power_profiles_cmd_v1, ops) !=
+ offsetof(struct iwl_geo_tx_power_profiles_cmd_v2, ops) ||
+ offsetof(struct iwl_geo_tx_power_profiles_cmd_v2, ops) !=
+ offsetof(struct iwl_geo_tx_power_profiles_cmd_v3, ops));
/* the ops field is at the same spot for all versions, so set in v1 */
cmd.v1.ops = cpu_to_le32(IWL_PER_CHAIN_OFFSET_SET_TABLES);
if (cmd_ver == 3) {
len = sizeof(cmd.v3);
+ n_bands = ARRAY_SIZE(cmd.v3.table[0]);
cmd.v3.table_revision = cpu_to_le32(mvm->fwrt.geo_rev);
} else if (fw_has_api(&mvm->fwrt.fw->ucode_capa,
IWL_UCODE_TLV_API_SAR_TABLE_VER)) {
len = sizeof(cmd.v2);
+ n_bands = ARRAY_SIZE(cmd.v2.table[0]);
cmd.v2.table_revision = cpu_to_le32(mvm->fwrt.geo_rev);
} else {
len = sizeof(cmd.v1);
+ n_bands = ARRAY_SIZE(cmd.v1.table[0]);
}
+ BUILD_BUG_ON(offsetof(struct iwl_geo_tx_power_profiles_cmd_v1, table) !=
+ offsetof(struct iwl_geo_tx_power_profiles_cmd_v2, table) ||
+ offsetof(struct iwl_geo_tx_power_profiles_cmd_v2, table) !=
+ offsetof(struct iwl_geo_tx_power_profiles_cmd_v3, table));
+ /* the table is at the same position for all versions, so set use v1 */
+ ret = iwl_sar_geo_init(&mvm->fwrt, &cmd.v1.table[0][0], n_bands);
+
+ /*
+ * It is a valid scenario to not support SAR, or miss wgds table,
+ * but in that case there is no need to send the command.
+ */
+ if (ret)
+ return 0;
+
return iwl_mvm_send_cmd_pdu(mvm,
WIDE_ID(PHY_OPS_GROUP, GEO_TX_POWER_LIMIT),
0, len, &cmd);
--
2.28.0
^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 08/13] iwlwifi: rs: align to new TLC config command API
2020-10-08 15:12 [PATCH 00/13] iwlwifi: updates intended for v5.10 2020-10-08 part 2 Luca Coelho
` (6 preceding siblings ...)
2020-10-08 15:12 ` [PATCH 07/13] iwlwifi: fix sar geo table initialization Luca Coelho
@ 2020-10-08 15:12 ` Luca Coelho
2020-10-08 15:12 ` [PATCH 09/13] iwlwifi: move PNVM implementation to common code Luca Coelho
` (4 subsequent siblings)
12 siblings, 0 replies; 19+ messages in thread
From: Luca Coelho @ 2020-10-08 15:12 UTC (permalink / raw)
To: kvalo; +Cc: linux-wireless
From: Mordechay Goodstein <mordechay.goodstein@intel.com>
The new API adds 4 bytes at end of the struct. We just need to make
sure that we don't break compatibility with old FWs.
Signed-off-by: Mordechay Goodstein <mordechay.goodstein@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
---
drivers/net/wireless/intel/iwlwifi/fw/api/rs.h | 7 +++++--
drivers/net/wireless/intel/iwlwifi/mvm/rs-fw.c | 8 +++++++-
2 files changed, 12 insertions(+), 3 deletions(-)
diff --git a/drivers/net/wireless/intel/iwlwifi/fw/api/rs.h b/drivers/net/wireless/intel/iwlwifi/fw/api/rs.h
index 73b839c3cac1..1ea54f643030 100644
--- a/drivers/net/wireless/intel/iwlwifi/fw/api/rs.h
+++ b/drivers/net/wireless/intel/iwlwifi/fw/api/rs.h
@@ -193,6 +193,8 @@ enum IWL_TLC_HT_BW_RATES {
* @sgi_ch_width_supp: bitmap of SGI support per channel width
* use BIT(@enum iwl_tlc_mng_cfg_cw)
* @reserved2: reserved
+ * @max_tx_op: max TXOP in uSecs for all AC (BK, BE, VO, VI),
+ * set zero for no limit.
*/
struct iwl_tlc_config_cmd {
u8 sta_id;
@@ -206,8 +208,9 @@ struct iwl_tlc_config_cmd {
__le16 ht_rates[IWL_TLC_NSS_MAX][2];
__le16 max_mpdu_len;
u8 sgi_ch_width_supp;
- u8 reserved2[1];
-} __packed; /* TLC_MNG_CONFIG_CMD_API_S_VER_2 */
+ u8 reserved2;
+ __le32 max_tx_op;
+} __packed; /* TLC_MNG_CONFIG_CMD_API_S_VER_3 */
/**
* enum iwl_tlc_update_flags - updated fields
diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/rs-fw.c b/drivers/net/wireless/intel/iwlwifi/mvm/rs-fw.c
index 012123268ba9..e0e80906fdc6 100644
--- a/drivers/net/wireless/intel/iwlwifi/mvm/rs-fw.c
+++ b/drivers/net/wireless/intel/iwlwifi/mvm/rs-fw.c
@@ -468,6 +468,12 @@ void rs_fw_rate_init(struct iwl_mvm *mvm, struct ieee80211_sta *sta,
.amsdu = iwl_mvm_is_csum_supported(mvm),
};
int ret;
+ u16 cmd_size = sizeof(cfg_cmd);
+
+ /* In old versions of the API the struct is 4 bytes smaller */
+ if (iwl_fw_lookup_cmd_ver(mvm->fw, DATA_PATH_GROUP,
+ TLC_MNG_CONFIG_CMD, 0) < 3)
+ cmd_size -= 4;
memset(lq_sta, 0, offsetof(typeof(*lq_sta), pers));
@@ -482,7 +488,7 @@ void rs_fw_rate_init(struct iwl_mvm *mvm, struct ieee80211_sta *sta,
*/
sta->max_amsdu_len = max_amsdu_len;
- ret = iwl_mvm_send_cmd_pdu(mvm, cmd_id, CMD_ASYNC, sizeof(cfg_cmd),
+ ret = iwl_mvm_send_cmd_pdu(mvm, cmd_id, CMD_ASYNC, cmd_size,
&cfg_cmd);
if (ret)
IWL_ERR(mvm, "Failed to send rate scale config (%d)\n", ret);
--
2.28.0
^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 09/13] iwlwifi: move PNVM implementation to common code
2020-10-08 15:12 [PATCH 00/13] iwlwifi: updates intended for v5.10 2020-10-08 part 2 Luca Coelho
` (7 preceding siblings ...)
2020-10-08 15:12 ` [PATCH 08/13] iwlwifi: rs: align to new TLC config command API Luca Coelho
@ 2020-10-08 15:12 ` Luca Coelho
2020-10-08 15:12 ` [PATCH 10/13] iwlwifi: add trans op to set PNVM Luca Coelho
` (3 subsequent siblings)
12 siblings, 0 replies; 19+ messages in thread
From: Luca Coelho @ 2020-10-08 15:12 UTC (permalink / raw)
To: kvalo; +Cc: linux-wireless
From: Luca Coelho <luciano.coelho@intel.com>
The PNVM code is generic and can be used by other opmodes. Move it to
a common file and include it in the relevant opmodes.
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
---
drivers/net/wireless/intel/iwlwifi/Makefile | 2 +-
drivers/net/wireless/intel/iwlwifi/fw/pnvm.c | 55 ++++++++++++++++++++
drivers/net/wireless/intel/iwlwifi/fw/pnvm.h | 18 +++++++
drivers/net/wireless/intel/iwlwifi/mvm/fw.c | 51 ++----------------
4 files changed, 78 insertions(+), 48 deletions(-)
create mode 100644 drivers/net/wireless/intel/iwlwifi/fw/pnvm.c
create mode 100644 drivers/net/wireless/intel/iwlwifi/fw/pnvm.h
diff --git a/drivers/net/wireless/intel/iwlwifi/Makefile b/drivers/net/wireless/intel/iwlwifi/Makefile
index 85c6fed28f8e..14b0db28143b 100644
--- a/drivers/net/wireless/intel/iwlwifi/Makefile
+++ b/drivers/net/wireless/intel/iwlwifi/Makefile
@@ -16,7 +16,7 @@ iwlwifi-objs += iwl-trans.o
iwlwifi-objs += queue/tx.o
iwlwifi-objs += fw/img.o fw/notif-wait.o
-iwlwifi-objs += fw/dbg.o
+iwlwifi-objs += fw/dbg.o fw/pnvm.o
iwlwifi-$(CONFIG_IWLMVM) += fw/paging.o fw/smem.o fw/init.o
iwlwifi-$(CONFIG_ACPI) += fw/acpi.o
iwlwifi-$(CONFIG_IWLWIFI_DEBUGFS) += fw/debugfs.o
diff --git a/drivers/net/wireless/intel/iwlwifi/fw/pnvm.c b/drivers/net/wireless/intel/iwlwifi/fw/pnvm.c
new file mode 100644
index 000000000000..f6108fda8915
--- /dev/null
+++ b/drivers/net/wireless/intel/iwlwifi/fw/pnvm.c
@@ -0,0 +1,55 @@
+// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause
+/******************************************************************************
+ *
+ * Copyright(c) 2020 Intel Corporation
+ *
+ *****************************************************************************/
+
+#include "iwl-drv.h"
+#include "pnvm.h"
+#include "iwl-prph.h"
+#include "iwl-io.h"
+#include "fw/api/commands.h"
+#include "fw/api/nvm-reg.h"
+
+static bool iwl_pnvm_complete_fn(struct iwl_notif_wait_data *notif_wait,
+ struct iwl_rx_packet *pkt, void *data)
+{
+ struct iwl_trans *trans = (struct iwl_trans *)data;
+ struct iwl_pnvm_init_complete_ntfy *pnvm_ntf = (void *)pkt->data;
+
+ IWL_DEBUG_FW(trans,
+ "PNVM complete notification received with status %d\n",
+ le32_to_cpu(pnvm_ntf->status));
+
+ return true;
+}
+
+int iwl_pnvm_load(struct iwl_trans *trans,
+ struct iwl_notif_wait_data *notif_wait)
+{
+ struct iwl_notification_wait pnvm_wait;
+ static const u16 ntf_cmds[] = { WIDE_ID(REGULATORY_AND_NVM_GROUP,
+ PNVM_INIT_COMPLETE_NTFY) };
+
+ /* if the SKU_ID is empty, there's nothing to do */
+ if (!trans->sku_id[0] && !trans->sku_id[1] && !trans->sku_id[2])
+ return 0;
+
+ /*
+ * TODO: phase 2: load the pnvm file, find the right section,
+ * load it and set the right DMA pointer.
+ */
+
+ iwl_init_notification_wait(notif_wait, &pnvm_wait,
+ ntf_cmds, ARRAY_SIZE(ntf_cmds),
+ iwl_pnvm_complete_fn, trans);
+
+ /* kick the doorbell */
+ iwl_write_umac_prph(trans, UREG_DOORBELL_TO_ISR6,
+ UREG_DOORBELL_TO_ISR6_PNVM);
+
+ return iwl_wait_notification(notif_wait, &pnvm_wait,
+ MVM_UCODE_PNVM_TIMEOUT);
+}
+IWL_EXPORT_SYMBOL(iwl_pnvm_load);
diff --git a/drivers/net/wireless/intel/iwlwifi/fw/pnvm.h b/drivers/net/wireless/intel/iwlwifi/fw/pnvm.h
new file mode 100644
index 000000000000..e4f91bce222d
--- /dev/null
+++ b/drivers/net/wireless/intel/iwlwifi/fw/pnvm.h
@@ -0,0 +1,18 @@
+/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */
+/******************************************************************************
+ *
+ * Copyright(c) 2020 Intel Corporation
+ *
+ *****************************************************************************/
+
+#ifndef __IWL_PNVM_H__
+#define __IWL_PNVM_H__
+
+#include "fw/notif-wait.h"
+
+#define MVM_UCODE_PNVM_TIMEOUT (HZ / 10)
+
+int iwl_pnvm_load(struct iwl_trans *trans,
+ struct iwl_notif_wait_data *notif_wait);
+
+#endif /* __IWL_PNVM_H__ */
diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/fw.c b/drivers/net/wireless/intel/iwlwifi/mvm/fw.c
index 9bce25a15df6..6385b9641126 100644
--- a/drivers/net/wireless/intel/iwlwifi/mvm/fw.c
+++ b/drivers/net/wireless/intel/iwlwifi/mvm/fw.c
@@ -70,6 +70,7 @@
#include "iwl-io.h" /* for iwl_mvm_rx_card_state_notif */
#include "iwl-prph.h"
#include "fw/acpi.h"
+#include "fw/pnvm.h"
#include "mvm.h"
#include "fw/dbg.h"
@@ -77,9 +78,8 @@
#include "iwl-modparams.h"
#include "iwl-nvm-parse.h"
-#define MVM_UCODE_ALIVE_TIMEOUT HZ
-#define MVM_UCODE_CALIB_TIMEOUT (2*HZ)
-#define MVM_UCODE_PNVM_TIMEOUT (HZ / 10)
+#define MVM_UCODE_ALIVE_TIMEOUT (HZ)
+#define MVM_UCODE_CALIB_TIMEOUT (2 * HZ)
#define UCODE_VALID_OK cpu_to_le32(0x1)
@@ -313,20 +313,6 @@ static bool iwl_alive_fn(struct iwl_notif_wait_data *notif_wait,
return true;
}
-static bool iwl_pnvm_complete_fn(struct iwl_notif_wait_data *notif_wait,
- struct iwl_rx_packet *pkt, void *data)
-{
- struct iwl_mvm *mvm =
- container_of(notif_wait, struct iwl_mvm, notif_wait);
- struct iwl_pnvm_init_complete_ntfy *pnvm_ntf = (void *)pkt->data;
-
- IWL_DEBUG_FW(mvm,
- "PNVM complete notification received with status %d\n",
- le32_to_cpu(pnvm_ntf->status));
-
- return true;
-}
-
static bool iwl_wait_init_complete(struct iwl_notif_wait_data *notif_wait,
struct iwl_rx_packet *pkt, void *data)
{
@@ -350,35 +336,6 @@ static bool iwl_wait_phy_db_entry(struct iwl_notif_wait_data *notif_wait,
return false;
}
-static int iwl_mvm_load_pnvm(struct iwl_mvm *mvm)
-{
- struct iwl_notification_wait pnvm_wait;
- static const u16 ntf_cmds[] = { WIDE_ID(REGULATORY_AND_NVM_GROUP,
- PNVM_INIT_COMPLETE_NTFY) };
-
- /* if the SKU_ID is empty, there's nothing to do */
- if (!mvm->trans->sku_id[0] &&
- !mvm->trans->sku_id[1] &&
- !mvm->trans->sku_id[2])
- return 0;
-
- /*
- * TODO: phase 2: load the pnvm file, find the right section,
- * load it and set the right DMA pointer.
- */
-
- iwl_init_notification_wait(&mvm->notif_wait, &pnvm_wait,
- ntf_cmds, ARRAY_SIZE(ntf_cmds),
- iwl_pnvm_complete_fn, NULL);
-
- /* kick the doorbell */
- iwl_write_umac_prph(mvm->trans, UREG_DOORBELL_TO_ISR6,
- UREG_DOORBELL_TO_ISR6_PNVM);
-
- return iwl_wait_notification(&mvm->notif_wait, &pnvm_wait,
- MVM_UCODE_PNVM_TIMEOUT);
-}
-
static int iwl_mvm_load_ucode_wait_alive(struct iwl_mvm *mvm,
enum iwl_ucode_type ucode_type)
{
@@ -467,7 +424,7 @@ static int iwl_mvm_load_ucode_wait_alive(struct iwl_mvm *mvm,
return -EIO;
}
- ret = iwl_mvm_load_pnvm(mvm);
+ ret = iwl_pnvm_load(mvm->trans, &mvm->notif_wait);
if (ret) {
IWL_ERR(mvm, "Timeout waiting for PNVM load!\n");
iwl_fw_set_current_image(&mvm->fwrt, old_type);
--
2.28.0
^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 10/13] iwlwifi: add trans op to set PNVM
2020-10-08 15:12 [PATCH 00/13] iwlwifi: updates intended for v5.10 2020-10-08 part 2 Luca Coelho
` (8 preceding siblings ...)
2020-10-08 15:12 ` [PATCH 09/13] iwlwifi: move PNVM implementation to common code Luca Coelho
@ 2020-10-08 15:12 ` Luca Coelho
2020-10-08 15:12 ` [PATCH 11/13] iwlwifi: pcie: implement set_pnvm op Luca Coelho
` (2 subsequent siblings)
12 siblings, 0 replies; 19+ messages in thread
From: Luca Coelho @ 2020-10-08 15:12 UTC (permalink / raw)
To: kvalo; +Cc: linux-wireless
From: Luca Coelho <luciano.coelho@intel.com>
Add an op to allow setting the PNVM data in the trans.
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
---
drivers/net/wireless/intel/iwlwifi/iwl-trans.h | 12 ++++++++++++
1 file changed, 12 insertions(+)
diff --git a/drivers/net/wireless/intel/iwlwifi/iwl-trans.h b/drivers/net/wireless/intel/iwlwifi/iwl-trans.h
index caae7b5aa9e6..5dbb5d45304f 100644
--- a/drivers/net/wireless/intel/iwlwifi/iwl-trans.h
+++ b/drivers/net/wireless/intel/iwlwifi/iwl-trans.h
@@ -568,6 +568,8 @@ struct iwl_trans_rxq_dma_data {
* Note that the transport must fill in the proper file headers.
* @debugfs_cleanup: used in the driver unload flow to make a proper cleanup
* of the trans debugfs
+ * @set_pnvm: set the pnvm data in the prph scratch buffer, inside the
+ * context info.
*/
struct iwl_trans_ops {
@@ -640,6 +642,7 @@ struct iwl_trans_ops {
u32 dump_mask);
void (*debugfs_cleanup)(struct iwl_trans *trans);
void (*sync_nmi)(struct iwl_trans *trans);
+ int (*set_pnvm)(struct iwl_trans *trans, const void *data, u32 len);
};
/**
@@ -1449,6 +1452,15 @@ static inline void iwl_trans_sync_nmi(struct iwl_trans *trans)
trans->ops->sync_nmi(trans);
}
+static inline int iwl_trans_set_pnvm(struct iwl_trans *trans,
+ const void *data, u32 len)
+{
+ if (trans->ops->set_pnvm)
+ return trans->ops->set_pnvm(trans, data, len);
+
+ return 0;
+}
+
static inline bool iwl_trans_dbg_ini_valid(struct iwl_trans *trans)
{
return trans->dbg.internal_ini_cfg != IWL_INI_CFG_STATE_NOT_LOADED ||
--
2.28.0
^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 11/13] iwlwifi: pcie: implement set_pnvm op
2020-10-08 15:12 [PATCH 00/13] iwlwifi: updates intended for v5.10 2020-10-08 part 2 Luca Coelho
` (9 preceding siblings ...)
2020-10-08 15:12 ` [PATCH 10/13] iwlwifi: add trans op to set PNVM Luca Coelho
@ 2020-10-08 15:12 ` Luca Coelho
2020-10-08 15:12 ` [PATCH 12/13] iwlwifi: read and parse PNVM file Luca Coelho
2020-10-08 15:12 ` [PATCH 13/13] iwlwifi: bump FW API to 59 for AX devices Luca Coelho
12 siblings, 0 replies; 19+ messages in thread
From: Luca Coelho @ 2020-10-08 15:12 UTC (permalink / raw)
To: kvalo; +Cc: linux-wireless
From: Luca Coelho <luciano.coelho@intel.com>
Implement the set_pnvm op to store the PNVM settings to the context
info and the corresponding code to free the DRAM block when the
context is freed.
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
---
.../intel/iwlwifi/iwl-context-info-gen3.h | 3 ++
.../wireless/intel/iwlwifi/iwl-context-info.h | 7 ++--
.../intel/iwlwifi/pcie/ctxt-info-gen3.c | 33 +++++++++++++++++++
.../wireless/intel/iwlwifi/pcie/ctxt-info.c | 21 +++++++-----
.../wireless/intel/iwlwifi/pcie/internal.h | 3 ++
.../net/wireless/intel/iwlwifi/pcie/trans.c | 2 ++
6 files changed, 58 insertions(+), 11 deletions(-)
diff --git a/drivers/net/wireless/intel/iwlwifi/iwl-context-info-gen3.h b/drivers/net/wireless/intel/iwlwifi/iwl-context-info-gen3.h
index 983e5b2859b3..5624fe42efd9 100644
--- a/drivers/net/wireless/intel/iwlwifi/iwl-context-info-gen3.h
+++ b/drivers/net/wireless/intel/iwlwifi/iwl-context-info-gen3.h
@@ -291,4 +291,7 @@ int iwl_pcie_ctxt_info_gen3_init(struct iwl_trans *trans,
const struct fw_img *fw);
void iwl_pcie_ctxt_info_gen3_free(struct iwl_trans *trans);
+int iwl_trans_pcie_ctx_info_gen3_set_pnvm(struct iwl_trans *trans,
+ const void *data, u32 len);
+
#endif /* __iwl_context_info_file_gen3_h__ */
diff --git a/drivers/net/wireless/intel/iwlwifi/iwl-context-info.h b/drivers/net/wireless/intel/iwlwifi/iwl-context-info.h
index eeaa8cbdddce..76b7bbdf8393 100644
--- a/drivers/net/wireless/intel/iwlwifi/iwl-context-info.h
+++ b/drivers/net/wireless/intel/iwlwifi/iwl-context-info.h
@@ -6,7 +6,7 @@
* GPL LICENSE SUMMARY
*
* Copyright(c) 2017 Intel Deutschland GmbH
- * Copyright(c) 2018 - 2019 Intel Corporation
+ * Copyright(c) 2018 - 2020 Intel Corporation
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of version 2 of the GNU General Public License as
@@ -20,7 +20,7 @@
* BSD LICENSE
*
* Copyright(c) 2017 Intel Deutschland GmbH
- * Copyright(c) 2018 - 2019 Intel Corporation
+ * Copyright(c) 2018 - 2020 Intel Corporation
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -225,5 +225,8 @@ void iwl_pcie_ctxt_info_free_paging(struct iwl_trans *trans);
int iwl_pcie_init_fw_sec(struct iwl_trans *trans,
const struct fw_img *fw,
struct iwl_context_info_dram *ctxt_dram);
+int iwl_pcie_ctxt_info_alloc_dma(struct iwl_trans *trans,
+ const void *data, u32 len,
+ struct iwl_dram_data *dram);
#endif /* __iwl_context_info_file_h__ */
diff --git a/drivers/net/wireless/intel/iwlwifi/pcie/ctxt-info-gen3.c b/drivers/net/wireless/intel/iwlwifi/pcie/ctxt-info-gen3.c
index 1ab136600415..546a24408c63 100644
--- a/drivers/net/wireless/intel/iwlwifi/pcie/ctxt-info-gen3.c
+++ b/drivers/net/wireless/intel/iwlwifi/pcie/ctxt-info-gen3.c
@@ -300,4 +300,37 @@ void iwl_pcie_ctxt_info_gen3_free(struct iwl_trans *trans)
trans_pcie->prph_info_dma_addr);
trans_pcie->prph_info_dma_addr = 0;
trans_pcie->prph_info = NULL;
+ dma_free_coherent(trans->dev, trans_pcie->pnvm_dram.size,
+ trans_pcie->pnvm_dram.block,
+ trans_pcie->pnvm_dram.physical);
+ trans_pcie->pnvm_dram.size = 0;
+ trans_pcie->pnvm_dram.block = NULL;
+ trans_pcie->pnvm_dram.physical = 0;
+}
+
+int iwl_trans_pcie_ctx_info_gen3_set_pnvm(struct iwl_trans *trans,
+ const void *data, u32 len)
+{
+ struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans);
+ struct iwl_prph_scratch_ctrl_cfg *prph_sc_ctrl =
+ &trans_pcie->prph_scratch->ctrl_cfg;
+ int ret;
+
+ if (trans->trans_cfg->device_family < IWL_DEVICE_FAMILY_AX210)
+ return 0;
+
+ ret = iwl_pcie_ctxt_info_alloc_dma(trans, data, len,
+ &trans_pcie->pnvm_dram);
+ if (ret < 0) {
+ IWL_DEBUG_FW(trans, "Failed to allocate PNVM DMA %d.\n",
+ ret);
+ return ret;
+ }
+
+ prph_sc_ctrl->pnvm_cfg.pnvm_base_addr =
+ cpu_to_le64(trans_pcie->pnvm_dram.physical);
+ prph_sc_ctrl->pnvm_cfg.pnvm_size =
+ cpu_to_le32(trans_pcie->pnvm_dram.size);
+
+ return 0;
}
diff --git a/drivers/net/wireless/intel/iwlwifi/pcie/ctxt-info.c b/drivers/net/wireless/intel/iwlwifi/pcie/ctxt-info.c
index 2597faea79c4..13fe9c00d7e8 100644
--- a/drivers/net/wireless/intel/iwlwifi/pcie/ctxt-info.c
+++ b/drivers/net/wireless/intel/iwlwifi/pcie/ctxt-info.c
@@ -93,17 +93,17 @@ static void *iwl_pcie_ctxt_info_dma_alloc_coherent(struct iwl_trans *trans,
return _iwl_pcie_ctxt_info_dma_alloc_coherent(trans, size, phys, 0);
}
-static int iwl_pcie_ctxt_info_alloc_dma(struct iwl_trans *trans,
- const struct fw_desc *sec,
- struct iwl_dram_data *dram)
+int iwl_pcie_ctxt_info_alloc_dma(struct iwl_trans *trans,
+ const void *data, u32 len,
+ struct iwl_dram_data *dram)
{
- dram->block = iwl_pcie_ctxt_info_dma_alloc_coherent(trans, sec->len,
+ dram->block = iwl_pcie_ctxt_info_dma_alloc_coherent(trans, len,
&dram->physical);
if (!dram->block)
return -ENOMEM;
- dram->size = sec->len;
- memcpy(dram->block, sec->data, sec->len);
+ dram->size = len;
+ memcpy(dram->block, data, len);
return 0;
}
@@ -156,7 +156,8 @@ int iwl_pcie_init_fw_sec(struct iwl_trans *trans,
/* initialize lmac sections */
for (i = 0; i < lmac_cnt; i++) {
- ret = iwl_pcie_ctxt_info_alloc_dma(trans, &fw->sec[i],
+ ret = iwl_pcie_ctxt_info_alloc_dma(trans, fw->sec[i].data,
+ fw->sec[i].len,
&dram->fw[dram->fw_cnt]);
if (ret)
return ret;
@@ -169,7 +170,8 @@ int iwl_pcie_init_fw_sec(struct iwl_trans *trans,
for (i = 0; i < umac_cnt; i++) {
/* access FW with +1 to make up for lmac separator */
ret = iwl_pcie_ctxt_info_alloc_dma(trans,
- &fw->sec[dram->fw_cnt + 1],
+ fw->sec[dram->fw_cnt + 1].data,
+ fw->sec[dram->fw_cnt + 1].len,
&dram->fw[dram->fw_cnt]);
if (ret)
return ret;
@@ -192,7 +194,8 @@ int iwl_pcie_init_fw_sec(struct iwl_trans *trans,
/* access FW with +2 to make up for lmac & umac separators */
int fw_idx = dram->fw_cnt + i + 2;
- ret = iwl_pcie_ctxt_info_alloc_dma(trans, &fw->sec[fw_idx],
+ ret = iwl_pcie_ctxt_info_alloc_dma(trans, fw->sec[fw_idx].data,
+ fw->sec[fw_idx].len,
&dram->paging[i]);
if (ret)
return ret;
diff --git a/drivers/net/wireless/intel/iwlwifi/pcie/internal.h b/drivers/net/wireless/intel/iwlwifi/pcie/internal.h
index 033f845a89dc..ff542d2f0054 100644
--- a/drivers/net/wireless/intel/iwlwifi/pcie/internal.h
+++ b/drivers/net/wireless/intel/iwlwifi/pcie/internal.h
@@ -339,6 +339,7 @@ struct cont_rec {
* @trans: pointer to the generic transport area
* @scd_base_addr: scheduler sram base address in SRAM
* @kw: keep warm address
+ * @pnvm_dram: DRAM area that contains the PNVM data
* @pci_dev: basic pci-network driver stuff
* @hw_base: pci hardware address support
* @ucode_write_complete: indicates that the ucode has been copied.
@@ -410,6 +411,8 @@ struct iwl_trans_pcie {
u32 scd_base_addr;
struct iwl_dma_ptr kw;
+ struct iwl_dram_data pnvm_dram;
+
struct iwl_txq *txq_memory;
/* PCI bus related data */
diff --git a/drivers/net/wireless/intel/iwlwifi/pcie/trans.c b/drivers/net/wireless/intel/iwlwifi/pcie/trans.c
index 4ee8b5dd2512..d25a215d5a04 100644
--- a/drivers/net/wireless/intel/iwlwifi/pcie/trans.c
+++ b/drivers/net/wireless/intel/iwlwifi/pcie/trans.c
@@ -81,6 +81,7 @@
#include "fw/api/tx.h"
#include "internal.h"
#include "iwl-fh.h"
+#include "iwl-context-info-gen3.h"
/* extended range in FW SRAM */
#define IWL_FW_MEM_EXTENDED_START 0x40000
@@ -3451,6 +3452,7 @@ static const struct iwl_trans_ops trans_ops_pcie_gen2 = {
.txq_free = iwl_txq_dyn_free,
.wait_txq_empty = iwl_trans_pcie_wait_txq_empty,
.rxq_dma_data = iwl_trans_pcie_rxq_dma_data,
+ .set_pnvm = iwl_trans_pcie_ctx_info_gen3_set_pnvm,
#ifdef CONFIG_IWLWIFI_DEBUGFS
.debugfs_cleanup = iwl_trans_pcie_debugfs_cleanup,
#endif
--
2.28.0
^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 12/13] iwlwifi: read and parse PNVM file
2020-10-08 15:12 [PATCH 00/13] iwlwifi: updates intended for v5.10 2020-10-08 part 2 Luca Coelho
` (10 preceding siblings ...)
2020-10-08 15:12 ` [PATCH 11/13] iwlwifi: pcie: implement set_pnvm op Luca Coelho
@ 2020-10-08 15:12 ` Luca Coelho
2020-10-08 15:12 ` [PATCH 13/13] iwlwifi: bump FW API to 59 for AX devices Luca Coelho
12 siblings, 0 replies; 19+ messages in thread
From: Luca Coelho @ 2020-10-08 15:12 UTC (permalink / raw)
To: kvalo; +Cc: linux-wireless
From: Luca Coelho <luciano.coelho@intel.com>
The driver looks for a PNVM file that contains FW configuration data
for each different HW combination. The FW requests the data for a
certain SKU_ID and the driver tries to find it in the PNVM file.
Read the file, parse its contents and send it to the trans.
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
---
drivers/net/wireless/intel/iwlwifi/fw/file.h | 6 +-
drivers/net/wireless/intel/iwlwifi/fw/pnvm.c | 223 +++++++++++++++++-
.../net/wireless/intel/iwlwifi/iwl-trans.h | 12 +-
.../intel/iwlwifi/pcie/ctxt-info-gen3.c | 6 -
.../net/wireless/intel/iwlwifi/pcie/trans.c | 5 +
5 files changed, 241 insertions(+), 11 deletions(-)
diff --git a/drivers/net/wireless/intel/iwlwifi/fw/file.h b/drivers/net/wireless/intel/iwlwifi/fw/file.h
index 1b7be350828d..02c64b988a13 100644
--- a/drivers/net/wireless/intel/iwlwifi/fw/file.h
+++ b/drivers/net/wireless/intel/iwlwifi/fw/file.h
@@ -146,7 +146,11 @@ enum iwl_ucode_tlv_type {
IWL_UCODE_TLV_UMAC_DEBUG_ADDRS = 54,
IWL_UCODE_TLV_LMAC_DEBUG_ADDRS = 55,
IWL_UCODE_TLV_FW_RECOVERY_INFO = 57,
- IWL_UCODE_TLV_FW_FSEQ_VERSION = 60,
+ IWL_UCODE_TLV_HW_TYPE = 58,
+ IWL_UCODE_TLV_FW_FSEQ_VERSION = 60,
+
+ IWL_UCODE_TLV_PNVM_VERSION = 62,
+ IWL_UCODE_TLV_PNVM_SKU = 64,
IWL_UCODE_TLV_FW_NUM_STATIONS = IWL_UCODE_TLV_CONST_BASE + 0,
diff --git a/drivers/net/wireless/intel/iwlwifi/fw/pnvm.c b/drivers/net/wireless/intel/iwlwifi/fw/pnvm.c
index f6108fda8915..6d8f7bff1243 100644
--- a/drivers/net/wireless/intel/iwlwifi/fw/pnvm.c
+++ b/drivers/net/wireless/intel/iwlwifi/fw/pnvm.c
@@ -11,6 +11,12 @@
#include "iwl-io.h"
#include "fw/api/commands.h"
#include "fw/api/nvm-reg.h"
+#include "fw/api/alive.h"
+
+struct iwl_pnvm_section {
+ __le32 offset;
+ const u8 data[];
+} __packed;
static bool iwl_pnvm_complete_fn(struct iwl_notif_wait_data *notif_wait,
struct iwl_rx_packet *pkt, void *data)
@@ -25,21 +31,234 @@ static bool iwl_pnvm_complete_fn(struct iwl_notif_wait_data *notif_wait,
return true;
}
+static int iwl_pnvm_handle_section(struct iwl_trans *trans, const u8 *data,
+ size_t len)
+{
+ struct iwl_ucode_tlv *tlv;
+ u32 sha1 = 0;
+ u16 mac_type = 0, rf_id = 0;
+ u8 *pnvm_data = NULL, *tmp;
+ u32 size = 0;
+ int ret;
+
+ IWL_DEBUG_FW(trans, "Handling PNVM section\n");
+
+ while (len >= sizeof(*tlv)) {
+ u32 tlv_len, tlv_type;
+
+ len -= sizeof(*tlv);
+ tlv = (void *)data;
+
+ tlv_len = le32_to_cpu(tlv->length);
+ tlv_type = le32_to_cpu(tlv->type);
+
+ if (len < tlv_len) {
+ IWL_ERR(trans, "invalid TLV len: %zd/%u\n",
+ len, tlv_len);
+ ret = -EINVAL;
+ goto out;
+ }
+
+ data += sizeof(*tlv);
+
+ switch (tlv_type) {
+ case IWL_UCODE_TLV_PNVM_VERSION:
+ if (tlv_len < sizeof(__le32)) {
+ IWL_DEBUG_FW(trans,
+ "Invalid size for IWL_UCODE_TLV_PNVM_VERSION (expected %zd, got %d)\n",
+ sizeof(__le32), tlv_len);
+ break;
+ }
+
+ sha1 = le32_to_cpup((__le32 *)data);
+
+ IWL_DEBUG_FW(trans,
+ "Got IWL_UCODE_TLV_PNVM_VERSION %0x\n",
+ sha1);
+ break;
+ case IWL_UCODE_TLV_HW_TYPE:
+ if (tlv_len < 2 * sizeof(__le16)) {
+ IWL_DEBUG_FW(trans,
+ "Invalid size for IWL_UCODE_TLV_HW_TYPE (expected %zd, got %d)\n",
+ 2 * sizeof(__le16), tlv_len);
+ break;
+ }
+
+ mac_type = le16_to_cpup((__le16 *)data);
+ rf_id = le16_to_cpup((__le16 *)(data + sizeof(__le16)));
+
+ IWL_DEBUG_FW(trans,
+ "Got IWL_UCODE_TLV_HW_TYPE mac_type 0x%0x rf_id 0x%0x\n",
+ mac_type, rf_id);
+
+ if (mac_type != CSR_HW_REV_TYPE(trans->hw_rev) ||
+ rf_id != CSR_HW_RFID_TYPE(trans->hw_rf_id)) {
+ IWL_DEBUG_FW(trans,
+ "HW mismatch, skipping PNVM section, mac_type 0x%0x, rf_id 0x%0x.\n",
+ CSR_HW_REV_TYPE(trans->hw_rev), trans->hw_rf_id);
+ ret = -ENOENT;
+ goto out;
+ }
+
+ break;
+ case IWL_UCODE_TLV_SEC_RT: {
+ struct iwl_pnvm_section *section = (void *)data;
+ u32 data_len = tlv_len - sizeof(*section);
+
+ IWL_DEBUG_FW(trans,
+ "Got IWL_UCODE_TLV_SEC_RT len %d\n",
+ tlv_len);
+
+ /* TODO: remove, this is a deprecated separator */
+ if (le32_to_cpup((__le32 *)data) == 0xddddeeee) {
+ IWL_DEBUG_FW(trans, "Ignoring separator.\n");
+ break;
+ }
+
+ IWL_DEBUG_FW(trans, "Adding data (size %d)\n",
+ data_len);
+
+ tmp = krealloc(pnvm_data, size + data_len, GFP_KERNEL);
+ if (!tmp) {
+ IWL_DEBUG_FW(trans,
+ "Couldn't allocate (more) pnvm_data\n");
+
+ ret = -ENOMEM;
+ goto out;
+ }
+
+ pnvm_data = tmp;
+
+ memcpy(pnvm_data + size, section->data, data_len);
+
+ size += data_len;
+
+ break;
+ }
+ case IWL_UCODE_TLV_PNVM_SKU:
+ IWL_DEBUG_FW(trans,
+ "New PNVM section started, stop parsing.\n");
+ goto done;
+ default:
+ IWL_DEBUG_FW(trans, "Found TLV 0x%0x, len %d\n",
+ tlv_type, tlv_len);
+ break;
+ }
+
+ len -= ALIGN(tlv_len, 4);
+ data += ALIGN(tlv_len, 4);
+ }
+
+done:
+ if (!size) {
+ IWL_DEBUG_FW(trans, "Empty PNVM, skipping.\n");
+ ret = -ENOENT;
+ goto out;
+ }
+
+ IWL_INFO(trans, "loaded PNVM version 0x%0x\n", sha1);
+
+ ret = iwl_trans_set_pnvm(trans, pnvm_data, size);
+out:
+ kfree(pnvm_data);
+ return ret;
+}
+
+static int iwl_pnvm_parse(struct iwl_trans *trans, const u8 *data,
+ size_t len)
+{
+ struct iwl_ucode_tlv *tlv;
+
+ IWL_DEBUG_FW(trans, "Parsing PNVM file\n");
+
+ while (len >= sizeof(*tlv)) {
+ u32 tlv_len, tlv_type;
+
+ len -= sizeof(*tlv);
+ tlv = (void *)data;
+
+ tlv_len = le32_to_cpu(tlv->length);
+ tlv_type = le32_to_cpu(tlv->type);
+
+ if (len < tlv_len) {
+ IWL_ERR(trans, "invalid TLV len: %zd/%u\n",
+ len, tlv_len);
+ return -EINVAL;
+ }
+
+ if (tlv_type == IWL_UCODE_TLV_PNVM_SKU) {
+ struct iwl_sku_id *sku_id =
+ (void *)(data + sizeof(*tlv));
+
+ IWL_DEBUG_FW(trans,
+ "Got IWL_UCODE_TLV_PNVM_SKU len %d\n",
+ tlv_len);
+ IWL_DEBUG_FW(trans, "sku_id 0x%0x 0x%0x 0x%0x\n",
+ le32_to_cpu(sku_id->data[0]),
+ le32_to_cpu(sku_id->data[1]),
+ le32_to_cpu(sku_id->data[2]));
+
+ if (trans->sku_id[0] == le32_to_cpu(sku_id->data[0]) &&
+ trans->sku_id[1] == le32_to_cpu(sku_id->data[1]) &&
+ trans->sku_id[2] == le32_to_cpu(sku_id->data[2])) {
+ int ret;
+
+ data += sizeof(*tlv) + ALIGN(tlv_len, 4);
+ len -= ALIGN(tlv_len, 4);
+
+ ret = iwl_pnvm_handle_section(trans, data, len);
+ if (!ret)
+ return 0;
+ } else {
+ IWL_DEBUG_FW(trans, "SKU ID didn't match!\n");
+ }
+ } else {
+ data += sizeof(*tlv) + ALIGN(tlv_len, 4);
+ len -= ALIGN(tlv_len, 4);
+ }
+ }
+
+ return -ENOENT;
+}
+
int iwl_pnvm_load(struct iwl_trans *trans,
struct iwl_notif_wait_data *notif_wait)
{
+ const struct firmware *pnvm;
struct iwl_notification_wait pnvm_wait;
static const u16 ntf_cmds[] = { WIDE_ID(REGULATORY_AND_NVM_GROUP,
PNVM_INIT_COMPLETE_NTFY) };
+ char pnvm_name[64];
+ int ret;
/* if the SKU_ID is empty, there's nothing to do */
if (!trans->sku_id[0] && !trans->sku_id[1] && !trans->sku_id[2])
return 0;
+ /* if we already have it, nothing to do either */
+ if (trans->pnvm_loaded)
+ return 0;
+
/*
- * TODO: phase 2: load the pnvm file, find the right section,
- * load it and set the right DMA pointer.
+ * The prefix unfortunately includes a hyphen at the end, so
+ * don't add the dot here...
*/
+ snprintf(pnvm_name, sizeof(pnvm_name), "%spnvm",
+ trans->cfg->fw_name_pre);
+
+ /* ...but replace the hyphen with the dot here. */
+ if (strlen(trans->cfg->fw_name_pre) < sizeof(pnvm_name))
+ pnvm_name[strlen(trans->cfg->fw_name_pre) - 1] = '.';
+
+ ret = firmware_request_nowarn(&pnvm, pnvm_name, trans->dev);
+ if (ret) {
+ IWL_DEBUG_FW(trans, "PNVM file %s not found %d\n",
+ pnvm_name, ret);
+ } else {
+ iwl_pnvm_parse(trans, pnvm->data, pnvm->size);
+
+ release_firmware(pnvm);
+ }
iwl_init_notification_wait(notif_wait, &pnvm_wait,
ntf_cmds, ARRAY_SIZE(ntf_cmds),
diff --git a/drivers/net/wireless/intel/iwlwifi/iwl-trans.h b/drivers/net/wireless/intel/iwlwifi/iwl-trans.h
index 5dbb5d45304f..b70294e9d07b 100644
--- a/drivers/net/wireless/intel/iwlwifi/iwl-trans.h
+++ b/drivers/net/wireless/intel/iwlwifi/iwl-trans.h
@@ -73,6 +73,7 @@
#include "iwl-config.h"
#include "fw/img.h"
#include "iwl-op-mode.h"
+#include <linux/firmware.h>
#include "fw/api/cmdhdr.h"
#include "fw/api/txq.h"
#include "fw/api/dbg-tlv.h"
@@ -1004,6 +1005,7 @@ struct iwl_trans {
bool pm_support;
bool ltr_enabled;
+ u8 pnvm_loaded:1;
const struct iwl_hcmd_arr *command_groups;
int command_groups_size;
@@ -1455,8 +1457,14 @@ static inline void iwl_trans_sync_nmi(struct iwl_trans *trans)
static inline int iwl_trans_set_pnvm(struct iwl_trans *trans,
const void *data, u32 len)
{
- if (trans->ops->set_pnvm)
- return trans->ops->set_pnvm(trans, data, len);
+ if (trans->ops->set_pnvm) {
+ int ret = trans->ops->set_pnvm(trans, data, len);
+
+ if (ret)
+ return ret;
+ }
+
+ trans->pnvm_loaded = true;
return 0;
}
diff --git a/drivers/net/wireless/intel/iwlwifi/pcie/ctxt-info-gen3.c b/drivers/net/wireless/intel/iwlwifi/pcie/ctxt-info-gen3.c
index 546a24408c63..a0352fa873d9 100644
--- a/drivers/net/wireless/intel/iwlwifi/pcie/ctxt-info-gen3.c
+++ b/drivers/net/wireless/intel/iwlwifi/pcie/ctxt-info-gen3.c
@@ -300,12 +300,6 @@ void iwl_pcie_ctxt_info_gen3_free(struct iwl_trans *trans)
trans_pcie->prph_info_dma_addr);
trans_pcie->prph_info_dma_addr = 0;
trans_pcie->prph_info = NULL;
- dma_free_coherent(trans->dev, trans_pcie->pnvm_dram.size,
- trans_pcie->pnvm_dram.block,
- trans_pcie->pnvm_dram.physical);
- trans_pcie->pnvm_dram.size = 0;
- trans_pcie->pnvm_dram.block = NULL;
- trans_pcie->pnvm_dram.physical = 0;
}
int iwl_trans_pcie_ctx_info_gen3_set_pnvm(struct iwl_trans *trans,
diff --git a/drivers/net/wireless/intel/iwlwifi/pcie/trans.c b/drivers/net/wireless/intel/iwlwifi/pcie/trans.c
index d25a215d5a04..d2e69ad53b27 100644
--- a/drivers/net/wireless/intel/iwlwifi/pcie/trans.c
+++ b/drivers/net/wireless/intel/iwlwifi/pcie/trans.c
@@ -1980,6 +1980,11 @@ void iwl_trans_pcie_free(struct iwl_trans *trans)
iwl_pcie_free_fw_monitor(trans);
+ if (trans_pcie->pnvm_dram.size)
+ dma_free_coherent(trans->dev, trans_pcie->pnvm_dram.size,
+ trans_pcie->pnvm_dram.block,
+ trans_pcie->pnvm_dram.physical);
+
mutex_destroy(&trans_pcie->mutex);
iwl_trans_free(trans);
}
--
2.28.0
^ permalink raw reply related [flat|nested] 19+ messages in thread
* [PATCH 13/13] iwlwifi: bump FW API to 59 for AX devices
2020-10-08 15:12 [PATCH 00/13] iwlwifi: updates intended for v5.10 2020-10-08 part 2 Luca Coelho
` (11 preceding siblings ...)
2020-10-08 15:12 ` [PATCH 12/13] iwlwifi: read and parse PNVM file Luca Coelho
@ 2020-10-08 15:12 ` Luca Coelho
12 siblings, 0 replies; 19+ messages in thread
From: Luca Coelho @ 2020-10-08 15:12 UTC (permalink / raw)
To: kvalo; +Cc: linux-wireless
From: Luca Coelho <luciano.coelho@intel.com>
Start supporting API version 59 for AX devices.
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
---
drivers/net/wireless/intel/iwlwifi/cfg/22000.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/wireless/intel/iwlwifi/cfg/22000.c b/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
index 8b97c2eb2632..d2bbe6a73514 100644
--- a/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
+++ b/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
@@ -57,7 +57,7 @@
#include "iwl-prph.h"
/* Highest firmware API version supported */
-#define IWL_22000_UCODE_API_MAX 57
+#define IWL_22000_UCODE_API_MAX 59
/* Lowest firmware API version supported */
#define IWL_22000_UCODE_API_MIN 39
--
2.28.0
^ permalink raw reply related [flat|nested] 19+ messages in thread
* Re: [PATCH 02/13] iwlwifi: pcie: fix the xtal latency value for a few qu devices
2020-10-08 15:12 ` [PATCH 02/13] iwlwifi: pcie: fix the xtal latency value for a few qu devices Luca Coelho
@ 2020-10-08 15:47 ` Kalle Valo
2020-10-08 19:10 ` Luca Coelho
0 siblings, 1 reply; 19+ messages in thread
From: Kalle Valo @ 2020-10-08 15:47 UTC (permalink / raw)
To: Luca Coelho; +Cc: linux-wireless
Luca Coelho <luca@coelho.fi> writes:
> From: Luca Coelho <luciano.coelho@intel.com>
>
> We were using 5 ms for some qu devices, but their xtal stabilizes
> much faster than that. Reduce the value to 500 ms.
>
> Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
> ---
> drivers/net/wireless/intel/iwlwifi/cfg/22000.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/net/wireless/intel/iwlwifi/cfg/22000.c b/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
> index 60e0640d07dc..8b97c2eb2632 100644
> --- a/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
> +++ b/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
> @@ -253,7 +253,7 @@ const struct iwl_cfg_trans_params iwl_qu_trans_cfg = {
> .device_family = IWL_DEVICE_FAMILY_22000,
> .base_params = &iwl_22000_base_params,
> .integrated = true,
> - .xtal_latency = 5000,
> + .xtal_latency = 500,
> .ltr_delay = IWL_CFG_TRANS_LTR_DELAY_200US,
> };
The commit log doesn't make sense, I'll do s/5 ms/5 s/ during commit.
--
https://patchwork.kernel.org/project/linux-wireless/list/
https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches
^ permalink raw reply [flat|nested] 19+ messages in thread
* Re: [PATCH 01/13] iwlwifi: pcie: fix xtal latency for 9560 devices
2020-10-08 15:12 ` [PATCH 01/13] iwlwifi: pcie: fix xtal latency for 9560 devices Luca Coelho
@ 2020-10-08 17:15 ` Kalle Valo
0 siblings, 0 replies; 19+ messages in thread
From: Kalle Valo @ 2020-10-08 17:15 UTC (permalink / raw)
To: Luca Coelho; +Cc: linux-wireless
Luca Coelho <luca@coelho.fi> wrote:
> From: Luca Coelho <luciano.coelho@intel.com>
>
> We were using a very high latency for all 9560 devices so they all
> would have time to stabilize. But this causes the system to be
> slighly slower, so we can use the best values for each device.
>
> This requires a new trans cfg struct for devices with longer latency
> and some adjustments to the other structs.
>
> Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
13 patches applied to wireless-drivers-next.git, thanks.
24436191a3b4 iwlwifi: pcie: fix xtal latency for 9560 devices
8fac81b50059 iwlwifi: pcie: fix the xtal latency value for a few qu devices
58ddd9b6d194 iwlwifi: mvm: don't send a CSA command the firmware doesn't know
60cbad2b867f iwlwifi: mvm: remove memset of kek_kck command
efa40c7ef689 iwlwifi: mvm: fix suspicious rcu usage warnings
853f4954baf4 iwlwifi: stats: add new api fields for statistics cmd/ntfy
45acebf8d6a6 iwlwifi: fix sar geo table initialization
0fafaa97ef35 iwlwifi: rs: align to new TLC config command API
b3e4c0f34c17 iwlwifi: move PNVM implementation to common code
a182dfab2842 iwlwifi: add trans op to set PNVM
6654cd4eb90f iwlwifi: pcie: implement set_pnvm op
6972592850c0 iwlwifi: read and parse PNVM file
09b3ab8782e7 iwlwifi: bump FW API to 59 for AX devices
--
https://patchwork.kernel.org/patch/11823191/
https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches
^ permalink raw reply [flat|nested] 19+ messages in thread
* Re: [PATCH 02/13] iwlwifi: pcie: fix the xtal latency value for a few qu devices
2020-10-08 15:47 ` Kalle Valo
@ 2020-10-08 19:10 ` Luca Coelho
2020-10-09 7:35 ` Kalle Valo
0 siblings, 1 reply; 19+ messages in thread
From: Luca Coelho @ 2020-10-08 19:10 UTC (permalink / raw)
To: Kalle Valo; +Cc: linux-wireless
On Thu, 2020-10-08 at 18:47 +0300, Kalle Valo wrote:
> Luca Coelho <luca@coelho.fi> writes:
>
> > From: Luca Coelho <luciano.coelho@intel.com>
> >
> > We were using 5 ms for some qu devices, but their xtal stabilizes
> > much faster than that. Reduce the value to 500 ms.
> >
> > Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
> > ---
> > drivers/net/wireless/intel/iwlwifi/cfg/22000.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/drivers/net/wireless/intel/iwlwifi/cfg/22000.c b/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
> > index 60e0640d07dc..8b97c2eb2632 100644
> > --- a/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
> > +++ b/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
> > @@ -253,7 +253,7 @@ const struct iwl_cfg_trans_params iwl_qu_trans_cfg = {
> > .device_family = IWL_DEVICE_FAMILY_22000,
> > .base_params = &iwl_22000_base_params,
> > .integrated = true,
> > - .xtal_latency = 5000,
> > + .xtal_latency = 500,
> > .ltr_delay = IWL_CFG_TRANS_LTR_DELAY_200US,
> > };
>
> The commit log doesn't make sense, I'll do s/5 ms/5 s/ during commit.
Hmmm, indeed it doesn't make sense like that. But I actually think the
5 ms is correct, but the other one should be 500 microseconds.
--
Luca.
^ permalink raw reply [flat|nested] 19+ messages in thread
* Re: [PATCH 02/13] iwlwifi: pcie: fix the xtal latency value for a few qu devices
2020-10-08 19:10 ` Luca Coelho
@ 2020-10-09 7:35 ` Kalle Valo
2020-10-09 7:38 ` Luca Coelho
0 siblings, 1 reply; 19+ messages in thread
From: Kalle Valo @ 2020-10-09 7:35 UTC (permalink / raw)
To: Luca Coelho; +Cc: linux-wireless
Luca Coelho <luca@coelho.fi> writes:
> On Thu, 2020-10-08 at 18:47 +0300, Kalle Valo wrote:
>> Luca Coelho <luca@coelho.fi> writes:
>>
>> > From: Luca Coelho <luciano.coelho@intel.com>
>> >
>> > We were using 5 ms for some qu devices, but their xtal stabilizes
>> > much faster than that. Reduce the value to 500 ms.
>> >
>> > Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
>> > ---
>> > drivers/net/wireless/intel/iwlwifi/cfg/22000.c | 2 +-
>> > 1 file changed, 1 insertion(+), 1 deletion(-)
>> >
>> > diff --git a/drivers/net/wireless/intel/iwlwifi/cfg/22000.c b/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
>> > index 60e0640d07dc..8b97c2eb2632 100644
>> > --- a/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
>> > +++ b/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
>> > @@ -253,7 +253,7 @@ const struct iwl_cfg_trans_params iwl_qu_trans_cfg = {
>> > .device_family = IWL_DEVICE_FAMILY_22000,
>> > .base_params = &iwl_22000_base_params,
>> > .integrated = true,
>> > - .xtal_latency = 5000,
>> > + .xtal_latency = 500,
>> > .ltr_delay = IWL_CFG_TRANS_LTR_DELAY_200US,
>> > };
>>
>> The commit log doesn't make sense, I'll do s/5 ms/5 s/ during commit.
>
> Hmmm, indeed it doesn't make sense like that. But I actually think the
> 5 ms is correct, but the other one should be 500 microseconds.
Oh, didn't think of that. And it's too late to fix it now.
--
https://patchwork.kernel.org/project/linux-wireless/list/
https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches
^ permalink raw reply [flat|nested] 19+ messages in thread
* Re: [PATCH 02/13] iwlwifi: pcie: fix the xtal latency value for a few qu devices
2020-10-09 7:35 ` Kalle Valo
@ 2020-10-09 7:38 ` Luca Coelho
0 siblings, 0 replies; 19+ messages in thread
From: Luca Coelho @ 2020-10-09 7:38 UTC (permalink / raw)
To: Kalle Valo; +Cc: linux-wireless
On Fri, 2020-10-09 at 10:35 +0300, Kalle Valo wrote:
> Luca Coelho <luca@coelho.fi> writes:
>
> > On Thu, 2020-10-08 at 18:47 +0300, Kalle Valo wrote:
> > > Luca Coelho <luca@coelho.fi> writes:
> > >
> > > > From: Luca Coelho <luciano.coelho@intel.com>
> > > >
> > > > We were using 5 ms for some qu devices, but their xtal stabilizes
> > > > much faster than that. Reduce the value to 500 ms.
> > > >
> > > > Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
> > > > ---
> > > > drivers/net/wireless/intel/iwlwifi/cfg/22000.c | 2 +-
> > > > 1 file changed, 1 insertion(+), 1 deletion(-)
> > > >
> > > > diff --git a/drivers/net/wireless/intel/iwlwifi/cfg/22000.c b/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
> > > > index 60e0640d07dc..8b97c2eb2632 100644
> > > > --- a/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
> > > > +++ b/drivers/net/wireless/intel/iwlwifi/cfg/22000.c
> > > > @@ -253,7 +253,7 @@ const struct iwl_cfg_trans_params iwl_qu_trans_cfg = {
> > > > .device_family = IWL_DEVICE_FAMILY_22000,
> > > > .base_params = &iwl_22000_base_params,
> > > > .integrated = true,
> > > > - .xtal_latency = 5000,
> > > > + .xtal_latency = 500,
> > > > .ltr_delay = IWL_CFG_TRANS_LTR_DELAY_200US,
> > > > };
> > >
> > > The commit log doesn't make sense, I'll do s/5 ms/5 s/ during commit.
> >
> > Hmmm, indeed it doesn't make sense like that. But I actually think the
> > 5 ms is correct, but the other one should be 500 microseconds.
>
> Oh, didn't think of that. And it's too late to fix it now.
Yeah, doesn't really matter. FWIW, I double-checked and indeed these
values are in usecs.
--
Cheers,
Luca.
^ permalink raw reply [flat|nested] 19+ messages in thread
end of thread, other threads:[~2020-10-09 7:38 UTC | newest]
Thread overview: 19+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-10-08 15:12 [PATCH 00/13] iwlwifi: updates intended for v5.10 2020-10-08 part 2 Luca Coelho
2020-10-08 15:12 ` [PATCH 01/13] iwlwifi: pcie: fix xtal latency for 9560 devices Luca Coelho
2020-10-08 17:15 ` Kalle Valo
2020-10-08 15:12 ` [PATCH 02/13] iwlwifi: pcie: fix the xtal latency value for a few qu devices Luca Coelho
2020-10-08 15:47 ` Kalle Valo
2020-10-08 19:10 ` Luca Coelho
2020-10-09 7:35 ` Kalle Valo
2020-10-09 7:38 ` Luca Coelho
2020-10-08 15:12 ` [PATCH 03/13] iwlwifi: mvm: don't send a CSA command the firmware doesn't know Luca Coelho
2020-10-08 15:12 ` [PATCH 04/13] iwlwifi: mvm: remove memset of kek_kck command Luca Coelho
2020-10-08 15:12 ` [PATCH 05/13] iwlwifi: mvm: fix suspicious rcu usage warnings Luca Coelho
2020-10-08 15:12 ` [PATCH 06/13] iwlwifi: stats: add new api fields for statistics cmd/ntfy Luca Coelho
2020-10-08 15:12 ` [PATCH 07/13] iwlwifi: fix sar geo table initialization Luca Coelho
2020-10-08 15:12 ` [PATCH 08/13] iwlwifi: rs: align to new TLC config command API Luca Coelho
2020-10-08 15:12 ` [PATCH 09/13] iwlwifi: move PNVM implementation to common code Luca Coelho
2020-10-08 15:12 ` [PATCH 10/13] iwlwifi: add trans op to set PNVM Luca Coelho
2020-10-08 15:12 ` [PATCH 11/13] iwlwifi: pcie: implement set_pnvm op Luca Coelho
2020-10-08 15:12 ` [PATCH 12/13] iwlwifi: read and parse PNVM file Luca Coelho
2020-10-08 15:12 ` [PATCH 13/13] iwlwifi: bump FW API to 59 for AX devices Luca Coelho
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.