ath10k: Allow setting coverage class and rx cleanups
v3 cover ath10k: Allow setting coverage class and rx cleanups
OK, another revision adding the fast path as suggested by Ben Greear. Only other change is that I am submitting it together with Vasanthakumar's cleanup patches to allow merging both into ath-next in one go.
Changes in v4:
* Only queue register setting work if coverage class is non-zero
* Add cleanup patches from Vasanthakumar into series
Changes in v3:
* Only update register values if hardware is initialized
* Use the SM/MS macros instead of manual bitshifts and masks
* Use the hardware phyclk register instead of assuming a clock speed
* Force dbglog_mask of ~0 and dbglog_level of at least WARN if active
* Only call code for selected WMI events
The last two items are a bit of a guess for me. In particular it looks like the default log level is already WARN, but I suppose that this should be good enough to get reports for internal chip resets. I have no way of verifying this myself though.
Benjamin Berg (1):
ath10k: Allow setting coverage class
Vasanthakumar Thiagarajan (4):
ath10k: Move ath10k_hw_params definition to hw.h
ath10k: Add provision for Rx descriptor abstraction
ath10k: Properly remove padding from the start of rx payload
ath10k: Remove 4-addr padding related hw_param configuration
drivers/net/wireless/ath/ath10k/core.c | 32 +++++-- drivers/net/wireless/ath/ath10k/core.h | 61 +++--------- drivers/net/wireless/ath/ath10k/htt_rx.c | 36 +++++--- drivers/net/wireless/ath/ath10k/hw.c | 154 +++++++++++++++++++++++++++++++ drivers/net/wireless/ath/ath10k/hw.h | 99 ++++++++++++++++++-- drivers/net/wireless/ath/ath10k/mac.c | 19 ++++ drivers/net/wireless/ath/ath10k/wmi.c | 48 ++++++++++ 7 files changed, 373 insertions(+), 76 deletions(-)
Signature · 7 lines
2.9.3 _______________________________________________ ath10k mailing list [email protected] http://lists.infradead.org/mailman/listinfo/ath10k
PATCH v3 1/5 ath10k: Move ath10k_hw_params definition to hw.h
From: Vasanthakumar Thiagarajan <[email protected]>
This is to prepare for rx descriptor abstraction where we'll be dereferencing ath10k_hw_params member in hw.h. Moreover hw.h looks more suitable to house ath10k_hw_params definition than core.h
--- drivers/net/wireless/ath/ath10k/core.h | 48 +--------------------------------- drivers/net/wireless/ath/ath10k/hw.h | 48 ++++++++++++++++++++++++++++++++++ 2 files changed, 49 insertions(+), 47 deletions(-)
drivers/net/wireless/ath/ath10k/core.h +1 -47
@@ -715,53 +715,7 @@ struct ath10k {struct ath10k_htc htc;struct ath10k_htt htt; -struct ath10k_hw_params { -u32 id; -u16 dev_id; -const char *name; -u32 patch_load_addr; -int uart_pin; -u32 otp_exe_param; - -/* Type of hw cycle counter wraparound logic, for more info -* refer enum ath10k_hw_cc_wraparound_type. -*/ -enum ath10k_hw_cc_wraparound_type cc_wraparound_type; - -/* Some of chip expects fragment descriptor to be continuous -* memory for any TX operation. Set continuous_frag_desc flag -* for the hardware which have such requirement. -*/ -bool continuous_frag_desc; - -/* CCK hardware rate table mapping for the newer chipsets -* like QCA99X0, QCA4019 got revised. The CCK h/w rate values -* are in a proper order with respect to the rate/preamble -*/ -bool cck_rate_map_rev2; - -u32 channel_counters_freq_hz; - -/* Mgmt tx descriptors threshold for limiting probe response -* frames. -*/ -u32 max_probe_resp_desc_thres; - -/* The padding bytes's location is different on various chips */ -enum ath10k_hw_4addr_pad hw_4addr_pad; - -u32 tx_chain_mask; -u32 rx_chain_mask; -u32 max_spatial_stream; -u32 cal_data_len; - -struct ath10k_hw_params_fw { -const char *dir; -const char *board; -size_t board_size; -size_t board_ext_size; - } fw; - } hw_params; +struct ath10k_hw_params hw_params;/* contains the firmware images used with ATH10K_FIRMWARE_MODE_NORMAL */ struct ath10k_fw_components normal_mode_fw;
drivers/net/wireless/ath/ath10k/hw.h +48 -0
54 lines not on this page · raw message
Signature · 7 lines
2.9.3 _______________________________________________ ath10k mailing list [email protected] http://lists.infradead.org/mailman/listinfo/ath10k
Re: [PATCHv3,1/5] ath10k: Move ath10k_hw_params definition to hw.h
Quoted 8 lines · Benjamin Berg
From: Vasanthakumar Thiagarajan <[email protected]>
This is to prepare for rx descriptor abstraction where we'll be dereferencing
ath10k_hw_paramsmember inhw.h. Moreoverhw.hlooks more suitable to houseath10k_hw_paramsdefinition thancore.hSigned-off-by Vasanthakumar Thiagarajan <[email protected]>
Thanks, 4 patches applied to ath-next branch of ath.git:
43d923e2c192 ath10k: move ath10k_hw_params definition to hw.h ae02c8719aab ath10k: add provision for Rx descriptor abstraction 9e19e1326142 ath10k: properly remove padding from the start of rx payload 95b5bf7ccca0 ath10k: remove 4-addr padding related hw_param configuration
Signature · 8 lines
Sent by pwcli https://patchwork.kernel.org/patch/9304059/ _______________________________________________ ath10k mailing list [email protected] http://lists.infradead.org/mailman/listinfo/ath10k
PATCH v3 2/5 ath10k: Add provision for Rx descriptor abstraction
In reply to Benjamin BergFrom: Vasanthakumar Thiagarajan <[email protected]>
There are slight differences in Rx hw descriptor information among different chips. So far driver does not use those new information for any functionalities, but there is one important information which is available from QCA99X0 onwards to indicate the number of bytes that hw padded at the begining of the rx payload and this information is needed to undecap the rx packet. Add an abstraction for Rx desc to make use of the new desc information available. The callback that this patch defines to retrieve the padding bytes will be used in follow-up patch.
--- drivers/net/wireless/ath/ath10k/core.c | 12 ++++++++++++ drivers/net/wireless/ath/ath10k/hw.c | 13 +++++++++++++ drivers/net/wireless/ath/ath10k/hw.h | 12 ++++++++++++ 3 files changed, 37 insertions(+)
drivers/net/wireless/ath/ath10k/core.c +12 -0
@@ -68,6 +68,7 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .board_size = QCA988X_BOARD_DATA_SZ, .board_ext_size = QCA988X_BOARD_EXT_DATA_SZ, }, + .hw_ops = &qca988x_ops, }, { .id = QCA9887_HW_1_0_VERSION,
@@ -87,6 +88,7 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .board_size = QCA9887_BOARD_DATA_SZ, .board_ext_size = QCA9887_BOARD_EXT_DATA_SZ, }, + .hw_ops = &qca988x_ops, }, { .id = QCA6174_HW_2_1_VERSION,
@@ -104,6 +106,7 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .board_size = QCA6174_BOARD_DATA_SZ, .board_ext_size = QCA6174_BOARD_EXT_DATA_SZ, }, + .hw_ops = &qca988x_ops, }, { .id = QCA6174_HW_2_1_VERSION,
@@ -122,6 +125,7 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .board_size = QCA6174_BOARD_DATA_SZ, .board_ext_size = QCA6174_BOARD_EXT_DATA_SZ, }, + .hw_ops = &qca988x_ops, }, { .id = QCA6174_HW_3_0_VERSION,
@@ -140,6 +144,7 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .board_size = QCA6174_BOARD_DATA_SZ, .board_ext_size = QCA6174_BOARD_EXT_DATA_SZ, }, + .hw_ops = &qca988x_ops, }, { .id = QCA6174_HW_3_2_VERSION,
@@ -159,6 +164,7 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .board_size = QCA6174_BOARD_DATA_SZ, .board_ext_size = QCA6174_BOARD_EXT_DATA_SZ, }, + .hw_ops = &qca988x_ops, }, { .id = QCA99X0_HW_2_0_DEV_VERSION,
@@ -182,6 +188,7 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .board_size = QCA99X0_BOARD_DATA_SZ, .board_ext_size = QCA99X0_BOARD_EXT_DATA_SZ, }, + .hw_ops = &qca99x0_ops, }, { .id = QCA9984_HW_1_0_DEV_VERSION,
@@ -205,6 +212,7 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .board_size = QCA99X0_BOARD_DATA_SZ, .board_ext_size = QCA99X0_BOARD_EXT_DATA_SZ, }, + .hw_ops = &qca99x0_ops, }, { .id = QCA9888_HW_2_0_DEV_VERSION,
@@ -227,6 +235,7 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .board_size = QCA99X0_BOARD_DATA_SZ, .board_ext_size = QCA99X0_BOARD_EXT_DATA_SZ, }, + .hw_ops = &qca99x0_ops, }, { .id = QCA9377_HW_1_0_DEV_VERSION,
@@ -244,6 +253,7 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .board_size = QCA9377_BOARD_DATA_SZ, .board_ext_size = QCA9377_BOARD_EXT_DATA_SZ, }, + .hw_ops = &qca988x_ops, }, { .id = QCA9377_HW_1_1_DEV_VERSION,
@@ -261,6 +271,7 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .board_size = QCA9377_BOARD_DATA_SZ, .board_ext_size = QCA9377_BOARD_EXT_DATA_SZ, }, + .hw_ops = &qca988x_ops, }, { .id = QCA4019_HW_1_0_DEV_VERSION,
@@ -285,6 +296,7 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .board_size = QCA4019_BOARD_DATA_SZ, .board_ext_size = QCA4019_BOARD_EXT_DATA_SZ, }, + .hw_ops = &qca99x0_ops, }, };
drivers/net/wireless/ath/ath10k/hw.c +13 -0
@@ -219,3 +219,16 @@ void ath10k_hw_fill_survey_time(struct ath10k *ar, struct survey_info *survey, survey->time =CCNT_TO_MSEC (ar, cc); survey->time_busy =CCNT_TO_MSEC (ar, rcc); } + +const struct ath10k_hw_ops qca988x_ops = { +}; + +static int ath10k_qca99x0_rx_desc_get_l3_pad_bytes (struct htt_rx_desc *rxd) +{ +return MS (__le32_to_cpu (rxd->msdu_end.qca99x0.info1), +RX_MSDU_END_INFO1_L3_HDR_PAD ); +} + +const struct ath10k_hw_ops qca99x0_ops = { + .rx_desc_get_l3_pad_bytes = ath10k_qca99x0_rx_desc_get_l3_pad_bytes, +};
drivers/net/wireless/ath/ath10k/hw.h +12 -0
@@ -409,8 +409,20 @@ struct ath10k_hw_params {size_t board_size;size_t board_ext_size; } fw; + +const struct ath10k_hw_ops *hw_ops; }; +struct htt_rx_desc ; + +/* Defines needed for Rx descriptor abstraction */ +struct ath10k_hw_ops { +int (*rx_desc_get_l3_pad_bytes)(struct htt_rx_desc *rxd); +}; + +extern const struct ath10k_hw_ops qca988x_ops; +extern const struct ath10k_hw_ops qca99x0_ops; +/* Target specific defines for MAIN firmware */ #define TARGET_NUM_VDEVS 8#define TARGET_NUM_PEER_AST 2
Signature · 7 lines
2.9.3 _______________________________________________ ath10k mailing list [email protected] http://lists.infradead.org/mailman/listinfo/ath10k
PATCH v3 3/5 ath10k: Properly remove padding from the start of rx payload
In reply to Benjamin BergFrom: Vasanthakumar Thiagarajan <[email protected]>
In QCA99X0 (QCA99X0, QCA9984, QCA9888 and QCA4019) family chips, hw adds padding at the begining of the rx payload to make L3 header 4-byte aligned. In the chips doing this type of padding, the number of bytes padded will be indicated through msdu_end:info1. Define a hw_rx_desc_ops wrapper to retrieve the number of padded bytes and use this while doing undecap. This should fix padding related issues with ethernt decap format with QCA99X0, QCA9984, QCA9888 and QCA4019 hw.
--- drivers/net/wireless/ath/ath10k/htt_rx.c | 36 +++++++++++++++++++------------- drivers/net/wireless/ath/ath10k/hw.h | 9 ++++++++ 2 files changed, 31 insertions(+), 14 deletions(-)
drivers/net/wireless/ath/ath10k/htt_rx.c +22 -14
@@ -1056,9 +1056,11 @@ static void ath10k_htt_rx_h_undecap_nwifi(struct ath10k *ar, const u8 first_hdr[64]) {struct ieee80211_hdr *hdr; +struct htt_rx_desc *rxd;size_t hdr_len;u8 da[ETH_ALEN ];u8 sa[ETH_ALEN ]; +int l3_pad_bytes;/* Delivered decapped frame: * [nwifi 802.11 header] <-- replaced with 802.11 hdr
@@ -1072,19 +1074,12 @@ static void ath10k_htt_rx_h_undecap_nwifi(struct ath10k *ar,*/ /* pull decapped header and copy SA & DA */ -if ((ar->hw_params.hw_4addr_pad ==ATH10K_HW_4ADDR_PAD_BEFORE ) && -ieee80211_has_a4 (((struct ieee80211_hdr *)first_hdr)->frame_control)) { -/* The QCA99X0 4 address mode pad 2 bytes at the -* beginning of MSDU -*/ - hdr = (struct ieee80211_hdr *)(msdu->data +2 ); -/* The skb length need be extended 2 as the 2 bytes at the tail -* be excluded due to the padding -*/ -skb_put (msdu,2 ); - }else { - hdr = (struct ieee80211_hdr *)(msdu->data); - } + rxd = (void *)msdu->data -sizeof (*rxd); + + l3_pad_bytes =ath10k_rx_desc_get_l3_pad_bytes (&ar->hw_params, rxd); +skb_put (msdu, l3_pad_bytes); + + hdr = (struct ieee80211_hdr *)(msdu->data + l3_pad_bytes); hdr_len =ath10k_htt_rx_nwifi_hdrlen (ar, hdr);ether_addr_copy (da,ieee80211_get_DA (hdr));
@@ -1151,6 +1146,8 @@ static void ath10k_htt_rx_h_undecap_eth(struct ath10k *ar,void *rfc1042;u8 da[ETH_ALEN ];u8 sa[ETH_ALEN ]; +int l3_pad_bytes; +struct htt_rx_desc *rxd;/* Delivered decapped frame: * [eth header] <-- replaced with 802.11 hdr & rfc1042/llc
@@ -1161,6 +1158,11 @@ static void ath10k_htt_rx_h_undecap_eth(struct ath10k *ar,if (WARN_ON_ONCE (!rfc1042))return ; + rxd = (void *)msdu->data -sizeof (*rxd); + l3_pad_bytes =ath10k_rx_desc_get_l3_pad_bytes (&ar->hw_params, rxd); +skb_put (msdu, l3_pad_bytes); +skb_pull (msdu, l3_pad_bytes); +/* pull decapped header and copy SA & DA */ eth = (struct ethhdr *)msdu->data;ether_addr_copy (da, eth->h_dest);
@@ -1191,6 +1193,8 @@ static void ath10k_htt_rx_h_undecap_snap(struct ath10k *ar, {struct ieee80211_hdr *hdr;size_t hdr_len; +int l3_pad_bytes; +struct htt_rx_desc *rxd;/* Delivered decapped frame: * [amsdu header] <-- replaced with 802.11 hdr
@@ -1198,7 +1202,11 @@ static void ath10k_htt_rx_h_undecap_snap(struct ath10k *ar,* [payload] */ -skb_pull (msdu,sizeof (struct amsdu_subframe_hdr )); + rxd = (void *)msdu->data -sizeof (*rxd); + l3_pad_bytes =ath10k_rx_desc_get_l3_pad_bytes (&ar->hw_params, rxd); + +skb_put (msdu, l3_pad_bytes); +skb_pull (msdu,sizeof (struct amsdu_subframe_hdr ) + l3_pad_bytes); hdr = (struct ieee80211_hdr *)first_hdr; hdr_len =ieee80211_hdrlen (hdr->frame_control);
drivers/net/wireless/ath/ath10k/hw.h +9 -0
@@ -423,6 +423,15 @@ struct ath10k_hw_ops {extern const struct ath10k_hw_ops qca988x_ops;extern const struct ath10k_hw_ops qca99x0_ops; +static inline int +ath10k_rx_desc_get_l3_pad_bytes (struct ath10k_hw_params *hw, +struct htt_rx_desc *rxd) +{ +if (hw->hw_ops->rx_desc_get_l3_pad_bytes) +return hw->hw_ops->rx_desc_get_l3_pad_bytes (rxd); +return 0 ; +} +/* Target specific defines for MAIN firmware */ #define TARGET_NUM_VDEVS 8#define TARGET_NUM_PEER_AST 2
Signature · 7 lines
2.9.3 _______________________________________________ ath10k mailing list [email protected] http://lists.infradead.org/mailman/listinfo/ath10k
PATCH v3 4/5 ath10k: Remove 4-addr padding related hw_param configuration
In reply to Benjamin BergFrom: Vasanthakumar Thiagarajan <[email protected]>
hw_4addr_pad was added to handle different types of padding in 4-address rx frame. But this padding is not very specific to 4-address, it can happen even with three address + ethernet decap mode. Since the padding information can be obtained through Rx desc for QCA99X0 and newer chips, this hw_param is not needed any more.
--- drivers/net/wireless/ath/ath10k/core.c | 9 --------- drivers/net/wireless/ath/ath10k/hw.h | 8 -------- 2 files changed, 17 deletions(-)
drivers/net/wireless/ath/ath10k/core.c +0 -9
@@ -60,7 +60,6 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .otp_exe_param = 0, .channel_counters_freq_hz = 88000, .max_probe_resp_desc_thres = 0, - .hw_4addr_pad = ATH10K_HW_4ADDR_PAD_AFTER, .cal_data_len = 2116, .fw = { .dir = QCA988X_HW_2_0_FW_DIR,
@@ -80,7 +79,6 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .otp_exe_param = 0, .channel_counters_freq_hz = 88000, .max_probe_resp_desc_thres = 0, - .hw_4addr_pad = ATH10K_HW_4ADDR_PAD_AFTER, .cal_data_len = 2116, .fw = { .dir = QCA9887_HW_1_0_FW_DIR,
@@ -117,7 +115,6 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .otp_exe_param = 0, .channel_counters_freq_hz = 88000, .max_probe_resp_desc_thres = 0, - .hw_4addr_pad = ATH10K_HW_4ADDR_PAD_AFTER, .cal_data_len = 8124, .fw = { .dir = QCA6174_HW_2_1_FW_DIR,
@@ -136,7 +133,6 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .otp_exe_param = 0, .channel_counters_freq_hz = 88000, .max_probe_resp_desc_thres = 0, - .hw_4addr_pad = ATH10K_HW_4ADDR_PAD_AFTER, .cal_data_len = 8124, .fw = { .dir = QCA6174_HW_3_0_FW_DIR,
@@ -155,7 +151,6 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .otp_exe_param = 0, .channel_counters_freq_hz = 88000, .max_probe_resp_desc_thres = 0, - .hw_4addr_pad = ATH10K_HW_4ADDR_PAD_AFTER, .cal_data_len = 8124, .fw = {/* uses same binaries as hw3.0 */
@@ -177,7 +172,6 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .cck_rate_map_rev2 = true, .channel_counters_freq_hz = 150000, .max_probe_resp_desc_thres = 24, - .hw_4addr_pad = ATH10K_HW_4ADDR_PAD_BEFORE, .tx_chain_mask = 0xf, .rx_chain_mask = 0xf, .max_spatial_stream = 4,
@@ -201,7 +195,6 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .cck_rate_map_rev2 = true, .channel_counters_freq_hz = 150000, .max_probe_resp_desc_thres = 24, - .hw_4addr_pad = ATH10K_HW_4ADDR_PAD_BEFORE, .tx_chain_mask = 0xf, .rx_chain_mask = 0xf, .max_spatial_stream = 4,
@@ -224,7 +217,6 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .continuous_frag_desc = true, .channel_counters_freq_hz = 150000, .max_probe_resp_desc_thres = 24, - .hw_4addr_pad = ATH10K_HW_4ADDR_PAD_BEFORE, .tx_chain_mask = 3, .rx_chain_mask = 3, .max_spatial_stream = 2,
@@ -285,7 +277,6 @@ static const struct ath10k_hw_params ath10k_hw_params_list[] = { .cck_rate_map_rev2 = true, .channel_counters_freq_hz = 125000, .max_probe_resp_desc_thres = 24, - .hw_4addr_pad = ATH10K_HW_4ADDR_PAD_BEFORE, .tx_chain_mask = 0x3, .rx_chain_mask = 0x3, .max_spatial_stream = 2,
drivers/net/wireless/ath/ath10k/hw.h +0 -8
@@ -338,11 +338,6 @@ enum ath10k_hw_rate_rev2_cck { ATH10K_HW_RATE_REV2_CCK_SP_11M, }; -enum ath10k_hw_4addr_pad { -ATH10K_HW_4ADDR_PAD_AFTER , -ATH10K_HW_4ADDR_PAD_BEFORE , -}; -enum ath10k_hw_cc_wraparound_type {ATH10K_HW_CC_WRAP_DISABLED =0 ,
@@ -395,9 +390,6 @@ struct ath10k_hw_params {*/ u32 max_probe_resp_desc_thres; -/* The padding bytes's location is different on various chips */ -enum ath10k_hw_4addr_pad hw_4addr_pad; -u32 tx_chain_mask;u32 rx_chain_mask;u32 max_spatial_stream;
Signature · 7 lines
2.9.3 _______________________________________________ ath10k mailing list [email protected] http://lists.infradead.org/mailman/listinfo/ath10k
PATCH v3 5/5 ath10k: Allow setting coverage class
In reply to Benjamin BergUnfortunately ath10k does not generally allow modifying the coverage class with the stock firmware and Qualcomm has so far refused to implement this feature so that it can be properly supported in ath10k. If we however know the registers that need to be modified for proper operation with a higher coverage class, then we can do these modifications from the driver.
This patch implements this hack for first generation cards which are based on a core that is similar to ath9k. The registers are modified in place and need to be re-written every time the firmware sets them. To achieve this the register status is verified after certain WMI events from the firmware.
The coverage class may not be modified temporarily right after the card re-initializes the registers. This is for example the case during scanning.
Thanks to Sebastian Gottschall <[email protected]> for initially working on a userspace support for this. This patch wouldn't have been possible without this documentation.
--- drivers/net/wireless/ath/ath10k/core.c | 11 +++ drivers/net/wireless/ath/ath10k/core.h | 13 +++ drivers/net/wireless/ath/ath10k/hw.c | 141 +++++++++++++++++++++++++++++++++ drivers/net/wireless/ath/ath10k/hw.h | 28 ++++++- drivers/net/wireless/ath/ath10k/mac.c | 19 +++++ drivers/net/wireless/ath/ath10k/wmi.c | 48 +++++++++++ 6 files changed, 259 insertions(+), 1 deletion(-)
drivers/net/wireless/ath/ath10k/core.c +11 -0
@@ -1543,6 +1543,15 @@ static void ath10k_core_restart(struct work_struct *work)mutex_unlock (&ar->conf_mutex); } +static void ath10k_core_set_coverage_class_work (struct work_struct *work) +{ + struct ath10k *ar = container_of(work, struct ath10k, + set_coverage_class_work); + +if (ar->hw_params.hw_ops->set_coverage_class) + ar->hw_params.hw_ops->set_coverage_class (ar,-1 ); +} +static int ath10k_core_init_firmware_features (struct ath10k *ar) {struct ath10k_fw_file *fw_file = &ar->normal_mode_fw.fw_file;
@@ -2251,6 +2260,8 @@ struct ath10k *ath10k_core_create(size_t priv_size, struct device *dev,INIT_WORK (&ar->register_work, ath10k_core_register_work);INIT_WORK (&ar->restart_work, ath10k_core_restart); +INIT_WORK (&ar->set_coverage_class_work, + ath10k_core_set_coverage_class_work); ret =ath10k_debug_create (ar);if (ret)
drivers/net/wireless/ath/ath10k/core.h +13 -0
@@ -890,6 +890,19 @@ struct ath10k {struct ath10k_thermal thermal;struct ath10k_wow wow; +struct work_struct set_coverage_class_work; +/* protected by conf_mutex */ +struct { +/* protected by data_lock */ +s16 coverage_class; + +u32 reg_phyclk; +u32 reg_slottime_conf; +u32 reg_slottime_orig; +u32 reg_ack_cts_timeout_conf; +u32 reg_ack_cts_timeout_orig; + } fw_coverage; +/* must be last */ u8 drv_priv[0 ]__aligned(sizeof(void *)) ; };
drivers/net/wireless/ath/ath10k/hw.c +141 -0
177 lines not on this page · raw message
drivers/net/wireless/ath/ath10k/hw.h +27 -1
@@ -230,6 +230,7 @@ struct ath10k_hw_regs {u32 rtc_soc_base_address;u32 rtc_wmac_base_address;u32 soc_core_base_address; +u32 wlan_mac_base_address;u32 ce_wrapper_base_address;u32 ce0_base_address;u32 ce1_base_address;
@@ -410,6 +411,7 @@ struct htt_rx_desc;/* Defines needed for Rx descriptor abstraction */ struct ath10k_hw_ops {int (*rx_desc_get_l3_pad_bytes)(struct htt_rx_desc *rxd); +void (*set_coverage_class)(struct ath10k *ar,s16 value); };extern const struct ath10k_hw_ops qca988x_ops;
@@ -606,7 +608,7 @@ ath10k_rx_desc_get_l3_pad_bytes(struct ath10k_hw_params *hw,#define WLAN_SI_BASE_ADDRESS 0x00010000#define WLAN_GPIO_BASE_ADDRESS 0x00014000#define WLAN_ANALOG_INTF_BASE_ADDRESS 0x0001c000 -#define WLAN_MAC_BASE_ADDRESS 0x00020000 +#define WLAN_MAC_BASE_ADDRESS ar->regs->wlan_mac_base_address#define EFUSE_BASE_ADDRESS 0x00030000#define FPGA_REG_BASE_ADDRESS 0x00039000#define WLAN_UART2_BASE_ADDRESS 0x00054c00
@@ -806,4 +808,28 @@ ath10k_rx_desc_get_l3_pad_bytes(struct ath10k_hw_params *hw,#define RTC_STATE_V_GET (x) (((x) & RTC_STATE_V_MASK) >> RTC_STATE_V_LSB) +/* Register definitions for first generation ath10k cards. These cards include +* a mac thich has a register allocation similar to ath9k and at least some +* registers including the ones relevant for modifying the coverage class are +* identical to the ath9k definitions. +* These registers are usually managed by the ath10k firmware. However by +* overriding them it is possible to support coverage class modifications. +*/ +#define WAVE1_PCU_ACK_CTS_TIMEOUT 0x8014 +#define WAVE1_PCU_ACK_CTS_TIMEOUT_MAX 0x00003FFF +#define WAVE1_PCU_ACK_CTS_TIMEOUT_ACK_MASK 0x00003FFF +#define WAVE1_PCU_ACK_CTS_TIMEOUT_ACK_LSB 0 +#define WAVE1_PCU_ACK_CTS_TIMEOUT_CTS_MASK 0x3FFF0000 +#define WAVE1_PCU_ACK_CTS_TIMEOUT_CTS_LSB 16 + +#define WAVE1_PCU_GBL_IFS_SLOT 0x1070 +#define WAVE1_PCU_GBL_IFS_SLOT_MASK 0x0000FFFF +#define WAVE1_PCU_GBL_IFS_SLOT_MAX 0x0000FFFF +#define WAVE1_PCU_GBL_IFS_SLOT_LSB 0 +#define WAVE1_PCU_GBL_IFS_SLOT_RESV0 0xFFFF0000 + +#define WAVE1_PHYCLK 0x801C +#define WAVE1_PHYCLK_USEC_MASK 0x0000007F +#define WAVE1_PHYCLK_USEC_LSB 0 +#endif /* _HW_H_ */
drivers/net/wireless/ath/ath10k/mac.c +19 -0
@@ -5372,6 +5372,20 @@ static void ath10k_bss_info_changed(struct ieee80211_hw *hw,mutex_unlock (&ar->conf_mutex); } +static void ath10k_mac_op_set_coverage_class (struct ieee80211_hw *hw,s16 value) +{ +struct ath10k *ar = hw->priv; + +/* This function should never be called if setting the coverage class +* is not supported on this hardware. +*/ +if (!ar->hw_params.hw_ops->set_coverage_class) { +WARN_ON_ONCE (1 ); +return ; + } + ar->hw_params.hw_ops->set_coverage_class (ar, value); +} +static int ath10k_hw_scan (struct ieee80211_hw *hw,struct ieee80211_vif *vif,struct ieee80211_scan_request *hw_req)
@@ -7397,6 +7411,7 @@ static const struct ieee80211_ops ath10k_ops = { .remove_interface = ath10k_remove_interface, .configure_filter = ath10k_configure_filter, .bss_info_changed = ath10k_bss_info_changed, + .set_coverage_class = ath10k_mac_op_set_coverage_class, .hw_scan = ath10k_hw_scan, .cancel_hw_scan = ath10k_cancel_hw_scan, .set_key = ath10k_set_key,
@@ -7974,6 +7989,10 @@ int ath10k_mac_register(struct ath10k *ar) ar->running_fw->fw_file.fw_features)) ar->ops->wake_tx_queue =NULL ; +/* Disable set_coverage_class for chipsets that do not support it. */ +if (!ar->hw_params.hw_ops->set_coverage_class) + ar->ops->set_coverage_class =NULL ; + ret =ath_regd_init (&ar->ath_common.regulatory, ar->hw->wiphy, ath10k_reg_notifier);if (ret) {
drivers/net/wireless/ath/ath10k/wmi.c +48 -0
222 lines not on this page · raw message
Signature · 7 lines
2.9.3 _______________________________________________ ath10k mailing list [email protected] http://lists.infradead.org/mailman/listinfo/ath10k
Re: [PATCHv3 5/5] ath10k: Allow setting coverage class
Quoted 14 lines · Benjamin Berg
Unfortunately ath10k does not generally allow modifying the coverage class with the stock firmware and Qualcomm has so far refused to implement this feature so that it can be properly supported in ath10k. If we however know the registers that need to be modified for proper operation with a higher coverage class, then we can do these modifications from the driver.
This patch implements this hack for first generation cards which are based on a core that is similar to ath9k. The registers are modified in place and need to be re-written every time the firmware sets them. To achieve this the register status is verified after certain WMI events from the firmware.
The coverage class may not be modified temporarily right after the card re-initializes the registers. This is for example the case during scanning.
Quoted 7 lines
Thanks to Sebastian Gottschall <[email protected]> for initially working on a userspace support for this. This patch wouldn't have been possible without this documentation.
Signed-off-by Benjamin Berg <[email protected]>Signed-off-by Simon Wunderlich <[email protected]>Signed-off-by Mathias Kretschmer <[email protected]>
To be honest this is an ugly hack and very fragile. But at the same time this is a quite requested feature, and the firmware doesn't support it, so I guess this is better than nothing. And it shouldn't affect in anyway if coverage class is not set from user space so hopefully it's also safe for others.
I need to review this patch more, but I noticed that it crashes on my setup with latest master-pending:
[ 89.768265] BUG: sleeping function called from invalid context at drivers/net/wireless/ath/ath10k/wmi.c:1802
[ 89.768324] in_atomic(): 1, irqs_disabled(): 0, pid: 1809, name: iw
[ 89.768374] 5 locks held by iw/1809:
[ 89.768422] #0: (cb_lock){++++++}, at: [<cc7d5465>] genl_rcv+0x15/0x30
[ 89.768727] #1: (genl_mutex){+.+.+.}, at: [<cc7d5964>] genl_lock+0x14/0x20
[ 89.769041] #2: (rtnl_mutex){+.+.+.}, at: [<cc7b0744>] rtnl_lock+0x14/0x20
[ 89.769363] #3: (&ar->conf_mutex){+.+.+.}, at: [<f8e0d214>] ath10k_hw_qca988x_set_coverage_class+0x24/0x3c0 [ath10k_core]
[ 89.769882] #4: (&(&ar->data_lock)->rlock){+.-...}, at: [<f8e0d222>] ath10k_hw_qca988x_set_coverage_class+0x32/0x3c0 [ath10k_core]
[ 89.770188] CPU: 0 PID: 1809 Comm: iw Tainted: G E 4.8.0-rc6-wt-ath+ #1468Signature · 5 lines
Kalle Valo _______________________________________________ ath10k mailing list [email protected] http://lists.infradead.org/mailman/listinfo/ath10k
PATCH 1/1 ath10k: Fix spinlock use in coverage class hack
ath10k_hw_qca988x_set_coverage_class needs to hold both conf_mutex and the data_lock spin lock for parts of the function. However, data_lock is only needed while storing the coverage_class to store the value that the card is configured to.
Fix the locking issue by only holding data_lock for the required duration.
---
And yes, I fully agree with your points of it being rather fragile. But as you said, it should be entirely safe if not used. Obviously a firmware implementation would be preferential.
This locking issue was pretty unnecessary. Lets see if any more issues show up in a closer review.
drivers/net/wireless/ath/ath10k/core.h | 2 +- drivers/net/wireless/ath/ath10k/hw.c | 5 +++-- 2 files changed, 4 insertions(+), 3 deletions(-)
drivers/net/wireless/ath/ath10k/core.h +1 -1
@@ -915,7 +915,7 @@ struct ath10k {struct work_struct set_coverage_class_work;/* protected by conf_mutex */ struct { -/* protected by data_lock */ +/* writing also protected by data_lock */ s16 coverage_class;u32 reg_phyclk;
drivers/net/wireless/ath/ath10k/hw.c +3 -2
@@ -243,7 +243,6 @@ static void ath10k_hw_qca988x_set_coverage_class(struct ath10k *ar,u32 fw_dbglog_level;mutex_lock (&ar->conf_mutex); -spin_lock_bh (&ar->data_lock);/* Only modify registers if the core is started. */ if ((ar->state != ATH10K_STATE_ON) &&
@@ -356,12 +355,14 @@ static void ath10k_hw_qca988x_set_coverage_class(struct ath10k *ar, store_regs:/* After an error we will not retry setting the coverage class. */ +spin_lock_bh (&ar->data_lock); ar->fw_coverage.coverage_class = value; +spin_unlock_bh (&ar->data_lock); + ar->fw_coverage.reg_slottime_conf = slottime_reg; ar->fw_coverage.reg_ack_cts_timeout_conf = timeout_reg; unlock: -spin_unlock_bh (&ar->data_lock);mutex_unlock (&ar->conf_mutex); }
Signature · 1 line
2.9.3
Re: [PATCH] ath10k: Fix spinlock use in coverage class hack
Quoted 6 lines · Benjamin Berg
ath10k_hw_qca988x_set_coverage_classneeds to hold bothconf_mutexand thedata_lockspin lock for parts of the function. However,data_lockis only needed while storing thecoverage_classto store the value that the card is configured to.Fix the locking issue by only holding
data_lockfor the required duration=
.
Quoted 2 lines
Signed-off-by Benjamin Berg <[email protected]>
Thanks, I also folded this with the patch in the pending branch.
Quoted 1 line
And yes, I fully agree with your points of it being rather fragile. But a=
s
Quoted 1 line
you said, it should be entirely safe if not used.
That's good.
Quoted 1 line
Obviously a firmware implementation would be preferential.
That's a shame as this feature is quite often requested. But if the firmware ever starts supporting the featrue we can then remove this hack from ath10k.
Quoted 1 line
This locking issue was pretty unnecessary. Lets see if any more issues sh=
ow
Quoted 1 line
up in a closer review.
I can't see the locking problem anymore so it seems to be fixed. I'll fix some minor things and send v2. I'll also CC linux-wireless.
--=20
Kalle Valo=
Re: [PATCHv3,5/5] ath10k: Allow setting coverage class
In reply to Benjamin BergQuoted 14 lines · Benjamin Berg
Unfortunately ath10k does not generally allow modifying the coverage class with the stock firmware and Qualcomm has so far refused to implement this feature so that it can be properly supported in ath10k. If we however know the registers that need to be modified for proper operation with a higher coverage class, then we can do these modifications from the driver.
This patch implements this hack for first generation cards which are based on a core that is similar to ath9k. The registers are modified in place and need to be re-written every time the firmware sets them. To achieve this the register status is verified after certain WMI events from the firmware.
The coverage class may not be modified temporarily right after the card re-initializes the registers. This is for example the case during scanning.
Quoted 7 lines
Thanks to Sebastian Gottschall <[email protected]> for initially working on a userspace support for this. This patch wouldn't have been possible without this documentation.
Signed-off-by Benjamin Berg <[email protected]>Signed-off-by Simon Wunderlich <[email protected]>Signed-off-by Mathias Kretschmer <[email protected]>
Patch applied to ath-next branch of ath.git, thanks.
ebee76f7fa46 ath10k: allow setting coverage class
Signature · 12 lines
https://patchwork.kernel.org/patch/9304067/ Documentation about submitting wireless patches and checking status from patchwork: https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches _______________________________________________ ath10k mailing list [email protected] http://lists.infradead.org/mailman/listinfo/ath10k
Quoted 1 line · Benjamin Berg
OK, another revision adding the fast path as suggested by Ben Greear. Onl=
y
Quoted 1 line
other change is that I am submitting it together with Vasanthakumar's cle=
anup
Quoted 6 lines
patches to allow merging both into ath-next in one go.
Changes in v4:
* Only queue register setting work if coverage class is non-zero
* Add cleanup patches from Vasanthakumar into series
Quoted 8 lines
Changes in v3:
* Only update register values if hardware is initialized
* Use the SM/MS macros instead of manual bitshifts and masks
* Use the hardware phyclk register instead of assuming a clock speed
* Forcedbglog_maskof ~0 anddbglog_levelof at least WARN if active
* Only call code for selected WMI eventsThe last two items are a bit of a guess for me. In particular it looks li=
ke
Quoted 1 line
the default log level is already WARN, but I suppose that this should be =
good
Quoted 1 line
enough to get reports for internal chip resets. I have no way of verifyin=
g
Quoted 11 lines
this myself though.
Benjamin Berg (1):
ath10k: Allow setting coverage classVasanthakumar Thiagarajan (4):
ath10k: Moveath10k_hw_paramsdefinition tohw.h
ath10k: Add provision for Rx descriptor abstraction
ath10k: Properly remove padding from the start of rx payload
ath10k: Remove 4-addr padding relatedhw_paramconfiguration
There were few conflicts, though easy. Please double check the pending branch that I didn't make any mistakes, I didn't test these yet.
https://git.kernel.org/cgit/linux/kernel/git/kvalo/ath.git/log/?h=3Dpending
Also please CC linux-wireless when submitting ath10k patches, we get wider review that way. Added it to CC now.
My log of the conflicts:
Applying: ath10k: Move ath10k_hw_params definition to hw.h
Using index info to reconstruct a base tree...
Falling back to patching base and 3-way merge...
Auto-merging drivers/net/wireless/ath/ath10k/core.h
CONFLICT (content): Merge conflict in drivers/net/wireless/ath/ath10k/core.= h
Failed to merge in the changes.
Patch failed at 0001 ath10k: Move ath10k_hw_params definition to hw.h
moved: bool sw_decrypt_mcast_mgmt;
Applying: ath10k: Add provision for Rx descriptor abstraction Using index info to reconstruct a base tree...
Falling back to patching base and 3-way merge...
Auto-merging drivers/net/wireless/ath/ath10k/hw.h
CONFLICT (content): Merge conflict in drivers/net/wireless/ath/ath10k/hw.h Auto-merging drivers/net/wireless/ath/ath10k/core.c
CONFLICT (content): Merge conflict in drivers/net/wireless/ath/ath10k/core.= c
Failed to merge in the changes.
Patch failed at 0002 ath10k: Add provision for Rx descriptor abstraction
Applying: ath10k: Allow setting coverage class
fatal: sha1 information is lacking or useless (drivers/net/wireless/ath/ath= 10k/core.c).
Repository lacks necessary blobs to fall back on 3-way merge. Cannot fall back to three-way merge.
Patch failed at 0001 ath10k: Allow setting coverage class
CONFLICT (content): Merge conflict in drivers/net/wireless/ath/ath10k/mac=
.c
patching file drivers/net/wireless/ath/ath10k/core.c Hunk #1 succeeded at 1548 (offset 5 lines). Hunk #2 succeeded at 2333 with fuzz 2 (offset 73 lines). patching file drivers/net/wireless/ath/ath10k/core.h Hunk #1 succeeded at 911 with fuzz 2 (offset 21 lines). patching file drivers/net/wireless/ath/ath10k/hw.c patching file drivers/net/wireless/ath/ath10k/hw.h Hunk #2 succeeded at 416 (offset 5 lines).
Hunk #3 succeeded at 613 (offset 5 lines).
Hunk #4 succeeded at 813 (offset 5 lines). patching file drivers/net/wireless/ath/ath10k/mac.c Hunk #1 succeeded at 5410 (offset 38 lines). Hunk #2 succeeded at 7449 (offset 38 lines). Hunk #3 succeeded at 8027 (offset 38 lines). patching file drivers/net/wireless/ath/ath10k/wmi.c Hunk #1 succeeded at 4925 (offset 46 lines). Hunk #2 succeeded at 4962 (offset 46 lines). Hunk #3 succeeded at 4972 (offset 46 lines). Hunk #4 succeeded at 4999 (offset 46 lines). Hunk #5 succeeded at 5055 (offset 46 lines). Hunk #6 succeeded at 5099 (offset 46 lines). Hunk #7 succeeded at 5109 (offset 46 lines). Hunk #8 succeeded at 5136 (offset 46 lines). Hunk #9 succeeded at 5183 (offset 46 lines). Hunk #10 succeeded at 5230 (offset 59 lines). Hunk #11 succeeded at 5240 (offset 59 lines). Hunk #12 succeeded at 5267 (offset 59 lines). Hunk #13 succeeded at 5305 (offset 59 lines). Hunk #14 succeeded at 5379 (offset 72 lines). Hunk #15 succeeded at 5396 (offset 72 lines). Hunk #16 succeeded at 5413 (offset 72 lines). Hunk #17 succeeded at 6137 (offset 73 lines).
--=20
Kalle Valo=
PATCH 1/1 ath10k: Add missing CONFIG_ATH10K_DEBUGFS check
The patch "ath10k: allow setting coverage class" was missing a check for CONFIG_ATH10K_DEBUGFS so it would try to use non-existing struct elements in some configurations. Fix this by adding the appropriate ifdef.
---
Sorry, so turns out the kbuild test robot is correct and I forgot the ifdef to check for CONFIG_ATH10K_DEBUGFS, so here a fixup commit for it.
Other than that everything looks good to me in the pending branch.
Benjamin
drivers/net/wireless/ath/ath10k/hw.c | 5 +++++ 1 file changed, 5 insertions(+)
drivers/net/wireless/ath/ath10k/hw.c +5 -0
@@ -338,8 +338,13 @@ static void ath10k_hw_qca988x_set_coverage_class(struct ath10k *ar,* set the registers again if the firmware does an internal reset and * this way we will be notified of the event. */ +#ifdef CONFIG_ATH10K_DEBUGFS fw_dbglog_mask = ar->debug.fw_dbglog_mask; fw_dbglog_level = ar->debug.fw_dbglog_level; +#else + fw_dbglog_mask =0 ; + fw_dbglog_level =0 ; +#endif if (value >0 ) {if (fw_dbglog_level >ATH10K_DBGLOG_LEVEL_WARN )
Signature · 1 line
2.9.3
Re: [PATCH] ath10k: Add missing CONFIG_ATH10K_DEBUGFS check
Quoted 8 lines · Benjamin Berg
The patch "ath10k: allow setting coverage class" was missing a check for
CONFIG_ATH10K_DEBUGFSso it would try to use non-existing struct elements in some configurations. Fix this by adding the appropriate ifdef.Signed-off-by Benjamin Berg <[email protected]>---
Sorry, so turns out the kbuild test robot is correct and I forgot the ifd=
ef
Quoted 3 lines
to check for
CONFIG_ATH10K_DEBUGFS, so here a fixup commit for it.Other than that everything looks good to me in the pending branch.
Thanks, I folded this to the original patch in the pending branch.
--=20
Kalle Valo=