linux/net/mac80211/debugfs_sta.c

1079 lines
32 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0-only
/*
* Copyright 2003-2005 Devicescape Software, Inc.
* Copyright (c) 2006 Jiri Benc <jbenc@suse.cz>
* Copyright 2007 Johannes Berg <johannes@sipsolutions.net>
* Copyright 2013-2014 Intel Mobile Communications GmbH
* Copyright(c) 2016 Intel Deutschland GmbH
* Copyright (C) 2018 - 2021 Intel Corporation
*/
#include <linux/debugfs.h>
#include <linux/ieee80211.h>
#include "ieee80211_i.h"
#include "debugfs.h"
#include "debugfs_sta.h"
#include "sta_info.h"
#include "driver-ops.h"
/* sta attributtes */
#define STA_READ(name, field, format_string) \
static ssize_t sta_ ##name## _read(struct file *file, \
char __user *userbuf, \
size_t count, loff_t *ppos) \
{ \
struct sta_info *sta = file->private_data; \
return mac80211_format_buffer(userbuf, count, ppos, \
format_string, sta->field); \
}
#define STA_READ_D(name, field) STA_READ(name, field, "%d\n")
#define STA_OPS(name) \
static const struct file_operations sta_ ##name## _ops = { \
.read = sta_##name##_read, \
.open = simple_open, \
.llseek = generic_file_llseek, \
}
#define STA_OPS_RW(name) \
static const struct file_operations sta_ ##name## _ops = { \
.read = sta_##name##_read, \
.write = sta_##name##_write, \
.open = simple_open, \
.llseek = generic_file_llseek, \
}
#define STA_FILE(name, field, format) \
STA_READ_##format(name, field) \
STA_OPS(name)
STA_FILE(aid, sta.aid, D);
static const char * const sta_flag_names[] = {
#define FLAG(F) [WLAN_STA_##F] = #F
FLAG(AUTH),
FLAG(ASSOC),
FLAG(PS_STA),
FLAG(AUTHORIZED),
FLAG(SHORT_PREAMBLE),
FLAG(WDS),
FLAG(CLEAR_PS_FILT),
FLAG(MFP),
FLAG(BLOCK_BA),
FLAG(PS_DRIVER),
FLAG(PSPOLL),
FLAG(TDLS_PEER),
FLAG(TDLS_PEER_AUTH),
FLAG(TDLS_INITIATOR),
FLAG(TDLS_CHAN_SWITCH),
FLAG(TDLS_OFF_CHANNEL),
FLAG(TDLS_WIDER_BW),
FLAG(UAPSD),
FLAG(SP),
FLAG(4ADDR_EVENT),
FLAG(INSERTED),
FLAG(RATE_CONTROL),
FLAG(TOFFSET_KNOWN),
FLAG(MPSP_OWNER),
FLAG(MPSP_RECIPIENT),
FLAG(PS_DELIVER),
FLAG(USES_ENCRYPTION),
FLAG(DECAP_OFFLOAD),
#undef FLAG
};
static ssize_t sta_flags_read(struct file *file, char __user *userbuf,
size_t count, loff_t *ppos)
{
char buf[16 * NUM_WLAN_STA_FLAGS], *pos = buf;
char *end = buf + sizeof(buf) - 1;
struct sta_info *sta = file->private_data;
unsigned int flg;
BUILD_BUG_ON(ARRAY_SIZE(sta_flag_names) != NUM_WLAN_STA_FLAGS);
for (flg = 0; flg < NUM_WLAN_STA_FLAGS; flg++) {
if (test_sta_flag(sta, flg))
pos += scnprintf(pos, end - pos, "%s\n",
sta_flag_names[flg]);
}
return simple_read_from_buffer(userbuf, count, ppos, buf, strlen(buf));
}
STA_OPS(flags);
static ssize_t sta_num_ps_buf_frames_read(struct file *file,
char __user *userbuf,
size_t count, loff_t *ppos)
{
struct sta_info *sta = file->private_data;
char buf[17*IEEE80211_NUM_ACS], *p = buf;
int ac;
for (ac = 0; ac < IEEE80211_NUM_ACS; ac++)
p += scnprintf(p, sizeof(buf)+buf-p, "AC%d: %d\n", ac,
skb_queue_len(&sta->ps_tx_buf[ac]) +
skb_queue_len(&sta->tx_filtered[ac]));
return simple_read_from_buffer(userbuf, count, ppos, buf, p - buf);
}
STA_OPS(num_ps_buf_frames);
static ssize_t sta_last_seq_ctrl_read(struct file *file, char __user *userbuf,
size_t count, loff_t *ppos)
{
char buf[15*IEEE80211_NUM_TIDS], *p = buf;
int i;
struct sta_info *sta = file->private_data;
for (i = 0; i < IEEE80211_NUM_TIDS; i++)
p += scnprintf(p, sizeof(buf)+buf-p, "%x ",
le16_to_cpu(sta->last_seq_ctrl[i]));
p += scnprintf(p, sizeof(buf)+buf-p, "\n");
return simple_read_from_buffer(userbuf, count, ppos, buf, p - buf);
}
STA_OPS(last_seq_ctrl);
#define AQM_TXQ_ENTRY_LEN 130
static ssize_t sta_aqm_read(struct file *file, char __user *userbuf,
size_t count, loff_t *ppos)
{
struct sta_info *sta = file->private_data;
struct ieee80211_local *local = sta->local;
size_t bufsz = AQM_TXQ_ENTRY_LEN * (IEEE80211_NUM_TIDS + 2);
char *buf = kzalloc(bufsz, GFP_KERNEL), *p = buf;
struct txq_info *txqi;
ssize_t rv;
int i;
if (!buf)
return -ENOMEM;
spin_lock_bh(&local->fq.lock);
rcu_read_lock();
p += scnprintf(p,
bufsz + buf - p,
"target %uus interval %uus ecn %s\n",
codel_time_to_us(sta->cparams.target),
codel_time_to_us(sta->cparams.interval),
sta->cparams.ecn ? "yes" : "no");
p += scnprintf(p,
bufsz + buf - p,
"tid ac backlog-bytes backlog-packets new-flows drops marks overlimit collisions tx-bytes tx-packets flags\n");
for (i = 0; i < ARRAY_SIZE(sta->sta.txq); i++) {
if (!sta->sta.txq[i])
continue;
txqi = to_txq_info(sta->sta.txq[i]);
p += scnprintf(p, bufsz + buf - p,
"%d %d %u %u %u %u %u %u %u %u %u 0x%lx(%s%s%s)\n",
txqi->txq.tid,
txqi->txq.ac,
txqi->tin.backlog_bytes,
txqi->tin.backlog_packets,
txqi->tin.flows,
txqi->cstats.drop_count,
txqi->cstats.ecn_mark,
txqi->tin.overlimit,
txqi->tin.collisions,
txqi->tin.tx_bytes,
txqi->tin.tx_packets,
txqi->flags,
mac80211: Add airtime accounting and scheduling to TXQs This adds airtime accounting and scheduling to the mac80211 TXQ scheduler. A new callback, ieee80211_sta_register_airtime(), is added that drivers can call to report airtime usage for stations. When airtime information is present, mac80211 will schedule TXQs (through ieee80211_next_txq()) in a way that enforces airtime fairness between active stations. This scheduling works the same way as the ath9k in-driver airtime fairness scheduling. If no airtime usage is reported by the driver, the scheduler will default to round-robin scheduling. For drivers that don't control TXQ scheduling in software, a new API function, ieee80211_txq_may_transmit(), is added which the driver can use to check if the TXQ is eligible for transmission, or should be throttled to enforce fairness. Calls to this function must also be enclosed in ieee80211_txq_schedule_{start,end}() calls to ensure proper locking. The API ieee80211_txq_may_transmit() also ensures that TXQ list will be aligned aginst driver's own round-robin scheduler list. i.e it rotates the TXQ list till it makes the requested node becomes the first entry in TXQ list. Thus both the TXQ list and driver's list are in sync. Co-developed-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Louie Lu <git@louie.lu> [added debugfs write op to reset airtime counter] Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk> Signed-off-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2018-12-19 09:02:08 +08:00
test_bit(IEEE80211_TXQ_STOP, &txqi->flags) ? "STOP" : "RUN",
test_bit(IEEE80211_TXQ_AMPDU, &txqi->flags) ? " AMPDU" : "",
test_bit(IEEE80211_TXQ_NO_AMSDU, &txqi->flags) ? " NO-AMSDU" : "");
}
rcu_read_unlock();
spin_unlock_bh(&local->fq.lock);
rv = simple_read_from_buffer(userbuf, count, ppos, buf, p - buf);
kfree(buf);
return rv;
}
STA_OPS(aqm);
mac80211: Add airtime accounting and scheduling to TXQs This adds airtime accounting and scheduling to the mac80211 TXQ scheduler. A new callback, ieee80211_sta_register_airtime(), is added that drivers can call to report airtime usage for stations. When airtime information is present, mac80211 will schedule TXQs (through ieee80211_next_txq()) in a way that enforces airtime fairness between active stations. This scheduling works the same way as the ath9k in-driver airtime fairness scheduling. If no airtime usage is reported by the driver, the scheduler will default to round-robin scheduling. For drivers that don't control TXQ scheduling in software, a new API function, ieee80211_txq_may_transmit(), is added which the driver can use to check if the TXQ is eligible for transmission, or should be throttled to enforce fairness. Calls to this function must also be enclosed in ieee80211_txq_schedule_{start,end}() calls to ensure proper locking. The API ieee80211_txq_may_transmit() also ensures that TXQ list will be aligned aginst driver's own round-robin scheduler list. i.e it rotates the TXQ list till it makes the requested node becomes the first entry in TXQ list. Thus both the TXQ list and driver's list are in sync. Co-developed-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Louie Lu <git@louie.lu> [added debugfs write op to reset airtime counter] Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk> Signed-off-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2018-12-19 09:02:08 +08:00
static ssize_t sta_airtime_read(struct file *file, char __user *userbuf,
size_t count, loff_t *ppos)
{
struct sta_info *sta = file->private_data;
struct ieee80211_local *local = sta->sdata->local;
mac80211: Implement Airtime-based Queue Limit (AQL) In order for the Fq_CoDel algorithm integrated in mac80211 layer to operate effectively to control excessive queueing latency, the CoDel algorithm requires an accurate measure of how long packets stays in the queue, AKA sojourn time. The sojourn time measured at the mac80211 layer doesn't include queueing latency in the lower layer (firmware/hardware) and CoDel expects lower layer to have a short queue. However, most 802.11ac chipsets offload tasks such TX aggregation to firmware or hardware, thus have a deep lower layer queue. Without a mechanism to control the lower layer queue size, packets only stay in mac80211 layer transiently before being sent to firmware queue. As a result, the sojourn time measured by CoDel in the mac80211 layer is almost always lower than the CoDel latency target, hence CoDel does little to control the latency, even when the lower layer queue causes excessive latency. The Byte Queue Limits (BQL) mechanism is commonly used to address the similar issue with wired network interface. However, this method cannot be applied directly to the wireless network interface. "Bytes" is not a suitable measure of queue depth in the wireless network, as the data rate can vary dramatically from station to station in the same network, from a few Mbps to over Gbps. This patch implements an Airtime-based Queue Limit (AQL) to make CoDel work effectively with wireless drivers that utilized firmware/hardware offloading. AQL allows each txq to release just enough packets to the lower layer to form 1-2 large aggregations to keep hardware fully utilized and retains the rest of the frames in mac80211 layer to be controlled by the CoDel algorithm. Signed-off-by: Kan Yan <kyan@google.com> [ Toke: Keep API to set pending airtime internal, fix nits in commit msg ] Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/r/20191119060610.76681-4-kyan@google.com Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2019-11-19 14:06:09 +08:00
size_t bufsz = 400;
mac80211: Add airtime accounting and scheduling to TXQs This adds airtime accounting and scheduling to the mac80211 TXQ scheduler. A new callback, ieee80211_sta_register_airtime(), is added that drivers can call to report airtime usage for stations. When airtime information is present, mac80211 will schedule TXQs (through ieee80211_next_txq()) in a way that enforces airtime fairness between active stations. This scheduling works the same way as the ath9k in-driver airtime fairness scheduling. If no airtime usage is reported by the driver, the scheduler will default to round-robin scheduling. For drivers that don't control TXQ scheduling in software, a new API function, ieee80211_txq_may_transmit(), is added which the driver can use to check if the TXQ is eligible for transmission, or should be throttled to enforce fairness. Calls to this function must also be enclosed in ieee80211_txq_schedule_{start,end}() calls to ensure proper locking. The API ieee80211_txq_may_transmit() also ensures that TXQ list will be aligned aginst driver's own round-robin scheduler list. i.e it rotates the TXQ list till it makes the requested node becomes the first entry in TXQ list. Thus both the TXQ list and driver's list are in sync. Co-developed-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Louie Lu <git@louie.lu> [added debugfs write op to reset airtime counter] Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk> Signed-off-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2018-12-19 09:02:08 +08:00
char *buf = kzalloc(bufsz, GFP_KERNEL), *p = buf;
u64 rx_airtime = 0, tx_airtime = 0;
mac80211: Switch to a virtual time-based airtime scheduler This switches the airtime scheduler in mac80211 to use a virtual time-based scheduler instead of the round-robin scheduler used before. This has a couple of advantages: - No need to sync up the round-robin scheduler in firmware/hardware with the round-robin airtime scheduler. - If several stations are eligible for transmission we can schedule both of them; no need to hard-block the scheduling rotation until the head of the queue has used up its quantum. - The check of whether a station is eligible for transmission becomes simpler (in ieee80211_txq_may_transmit()). The drawback is that scheduling becomes slightly more expensive, as we need to maintain an rbtree of TXQs sorted by virtual time. This means that ieee80211_register_airtime() becomes O(logN) in the number of currently scheduled TXQs because it can change the order of the scheduled stations. We mitigate this overhead by only resorting when a station changes position in the tree, and hopefully N rarely grows too big (it's only TXQs currently backlogged, not all associated stations), so it shouldn't be too big of an issue. To prevent divisions in the fast path, we maintain both station sums and pre-computed reciprocals of the sums. This turns the fast-path operation into a multiplication, with divisions only happening as the number of active stations change (to re-compute the current sum of all active station weights). To prevent this re-computation of the reciprocal from happening too frequently, we use a time-based notion of station activity, instead of updating the weight every time a station gets scheduled or de-scheduled. As queues can oscillate between empty and occupied quite frequently, this can significantly cut down on the number of re-computations. It also has the added benefit of making the station airtime calculation independent on whether the queue happened to have drained at the time an airtime value was accounted. Co-developed-by: Yibo Zhao <yiboz@codeaurora.org> Signed-off-by: Yibo Zhao <yiboz@codeaurora.org> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/r/20210623134755.235545-1-toke@redhat.com Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2021-06-23 21:47:55 +08:00
u64 v_t[IEEE80211_NUM_ACS];
mac80211: Add airtime accounting and scheduling to TXQs This adds airtime accounting and scheduling to the mac80211 TXQ scheduler. A new callback, ieee80211_sta_register_airtime(), is added that drivers can call to report airtime usage for stations. When airtime information is present, mac80211 will schedule TXQs (through ieee80211_next_txq()) in a way that enforces airtime fairness between active stations. This scheduling works the same way as the ath9k in-driver airtime fairness scheduling. If no airtime usage is reported by the driver, the scheduler will default to round-robin scheduling. For drivers that don't control TXQ scheduling in software, a new API function, ieee80211_txq_may_transmit(), is added which the driver can use to check if the TXQ is eligible for transmission, or should be throttled to enforce fairness. Calls to this function must also be enclosed in ieee80211_txq_schedule_{start,end}() calls to ensure proper locking. The API ieee80211_txq_may_transmit() also ensures that TXQ list will be aligned aginst driver's own round-robin scheduler list. i.e it rotates the TXQ list till it makes the requested node becomes the first entry in TXQ list. Thus both the TXQ list and driver's list are in sync. Co-developed-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Louie Lu <git@louie.lu> [added debugfs write op to reset airtime counter] Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk> Signed-off-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2018-12-19 09:02:08 +08:00
ssize_t rv;
int ac;
if (!buf)
return -ENOMEM;
for (ac = 0; ac < IEEE80211_NUM_ACS; ac++) {
mac80211: Switch to a virtual time-based airtime scheduler This switches the airtime scheduler in mac80211 to use a virtual time-based scheduler instead of the round-robin scheduler used before. This has a couple of advantages: - No need to sync up the round-robin scheduler in firmware/hardware with the round-robin airtime scheduler. - If several stations are eligible for transmission we can schedule both of them; no need to hard-block the scheduling rotation until the head of the queue has used up its quantum. - The check of whether a station is eligible for transmission becomes simpler (in ieee80211_txq_may_transmit()). The drawback is that scheduling becomes slightly more expensive, as we need to maintain an rbtree of TXQs sorted by virtual time. This means that ieee80211_register_airtime() becomes O(logN) in the number of currently scheduled TXQs because it can change the order of the scheduled stations. We mitigate this overhead by only resorting when a station changes position in the tree, and hopefully N rarely grows too big (it's only TXQs currently backlogged, not all associated stations), so it shouldn't be too big of an issue. To prevent divisions in the fast path, we maintain both station sums and pre-computed reciprocals of the sums. This turns the fast-path operation into a multiplication, with divisions only happening as the number of active stations change (to re-compute the current sum of all active station weights). To prevent this re-computation of the reciprocal from happening too frequently, we use a time-based notion of station activity, instead of updating the weight every time a station gets scheduled or de-scheduled. As queues can oscillate between empty and occupied quite frequently, this can significantly cut down on the number of re-computations. It also has the added benefit of making the station airtime calculation independent on whether the queue happened to have drained at the time an airtime value was accounted. Co-developed-by: Yibo Zhao <yiboz@codeaurora.org> Signed-off-by: Yibo Zhao <yiboz@codeaurora.org> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/r/20210623134755.235545-1-toke@redhat.com Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2021-06-23 21:47:55 +08:00
spin_lock_bh(&local->airtime[ac].lock);
mac80211: Add airtime accounting and scheduling to TXQs This adds airtime accounting and scheduling to the mac80211 TXQ scheduler. A new callback, ieee80211_sta_register_airtime(), is added that drivers can call to report airtime usage for stations. When airtime information is present, mac80211 will schedule TXQs (through ieee80211_next_txq()) in a way that enforces airtime fairness between active stations. This scheduling works the same way as the ath9k in-driver airtime fairness scheduling. If no airtime usage is reported by the driver, the scheduler will default to round-robin scheduling. For drivers that don't control TXQ scheduling in software, a new API function, ieee80211_txq_may_transmit(), is added which the driver can use to check if the TXQ is eligible for transmission, or should be throttled to enforce fairness. Calls to this function must also be enclosed in ieee80211_txq_schedule_{start,end}() calls to ensure proper locking. The API ieee80211_txq_may_transmit() also ensures that TXQ list will be aligned aginst driver's own round-robin scheduler list. i.e it rotates the TXQ list till it makes the requested node becomes the first entry in TXQ list. Thus both the TXQ list and driver's list are in sync. Co-developed-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Louie Lu <git@louie.lu> [added debugfs write op to reset airtime counter] Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk> Signed-off-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2018-12-19 09:02:08 +08:00
rx_airtime += sta->airtime[ac].rx_airtime;
tx_airtime += sta->airtime[ac].tx_airtime;
mac80211: Switch to a virtual time-based airtime scheduler This switches the airtime scheduler in mac80211 to use a virtual time-based scheduler instead of the round-robin scheduler used before. This has a couple of advantages: - No need to sync up the round-robin scheduler in firmware/hardware with the round-robin airtime scheduler. - If several stations are eligible for transmission we can schedule both of them; no need to hard-block the scheduling rotation until the head of the queue has used up its quantum. - The check of whether a station is eligible for transmission becomes simpler (in ieee80211_txq_may_transmit()). The drawback is that scheduling becomes slightly more expensive, as we need to maintain an rbtree of TXQs sorted by virtual time. This means that ieee80211_register_airtime() becomes O(logN) in the number of currently scheduled TXQs because it can change the order of the scheduled stations. We mitigate this overhead by only resorting when a station changes position in the tree, and hopefully N rarely grows too big (it's only TXQs currently backlogged, not all associated stations), so it shouldn't be too big of an issue. To prevent divisions in the fast path, we maintain both station sums and pre-computed reciprocals of the sums. This turns the fast-path operation into a multiplication, with divisions only happening as the number of active stations change (to re-compute the current sum of all active station weights). To prevent this re-computation of the reciprocal from happening too frequently, we use a time-based notion of station activity, instead of updating the weight every time a station gets scheduled or de-scheduled. As queues can oscillate between empty and occupied quite frequently, this can significantly cut down on the number of re-computations. It also has the added benefit of making the station airtime calculation independent on whether the queue happened to have drained at the time an airtime value was accounted. Co-developed-by: Yibo Zhao <yiboz@codeaurora.org> Signed-off-by: Yibo Zhao <yiboz@codeaurora.org> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/r/20210623134755.235545-1-toke@redhat.com Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2021-06-23 21:47:55 +08:00
v_t[ac] = sta->airtime[ac].v_t;
spin_unlock_bh(&local->airtime[ac].lock);
}
p += scnprintf(p, bufsz + buf - p,
"RX: %llu us\nTX: %llu us\nWeight: %u\n"
mac80211: Switch to a virtual time-based airtime scheduler This switches the airtime scheduler in mac80211 to use a virtual time-based scheduler instead of the round-robin scheduler used before. This has a couple of advantages: - No need to sync up the round-robin scheduler in firmware/hardware with the round-robin airtime scheduler. - If several stations are eligible for transmission we can schedule both of them; no need to hard-block the scheduling rotation until the head of the queue has used up its quantum. - The check of whether a station is eligible for transmission becomes simpler (in ieee80211_txq_may_transmit()). The drawback is that scheduling becomes slightly more expensive, as we need to maintain an rbtree of TXQs sorted by virtual time. This means that ieee80211_register_airtime() becomes O(logN) in the number of currently scheduled TXQs because it can change the order of the scheduled stations. We mitigate this overhead by only resorting when a station changes position in the tree, and hopefully N rarely grows too big (it's only TXQs currently backlogged, not all associated stations), so it shouldn't be too big of an issue. To prevent divisions in the fast path, we maintain both station sums and pre-computed reciprocals of the sums. This turns the fast-path operation into a multiplication, with divisions only happening as the number of active stations change (to re-compute the current sum of all active station weights). To prevent this re-computation of the reciprocal from happening too frequently, we use a time-based notion of station activity, instead of updating the weight every time a station gets scheduled or de-scheduled. As queues can oscillate between empty and occupied quite frequently, this can significantly cut down on the number of re-computations. It also has the added benefit of making the station airtime calculation independent on whether the queue happened to have drained at the time an airtime value was accounted. Co-developed-by: Yibo Zhao <yiboz@codeaurora.org> Signed-off-by: Yibo Zhao <yiboz@codeaurora.org> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/r/20210623134755.235545-1-toke@redhat.com Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2021-06-23 21:47:55 +08:00
"Virt-T: VO: %lld us VI: %lld us BE: %lld us BK: %lld us\n",
rx_airtime, tx_airtime, sta->airtime[0].weight,
v_t[0], v_t[1], v_t[2], v_t[3]);
rv = simple_read_from_buffer(userbuf, count, ppos, buf, p - buf);
kfree(buf);
return rv;
}
static ssize_t sta_airtime_write(struct file *file, const char __user *userbuf,
size_t count, loff_t *ppos)
{
struct sta_info *sta = file->private_data;
struct ieee80211_local *local = sta->sdata->local;
int ac;
for (ac = 0; ac < IEEE80211_NUM_ACS; ac++) {
mac80211: Switch to a virtual time-based airtime scheduler This switches the airtime scheduler in mac80211 to use a virtual time-based scheduler instead of the round-robin scheduler used before. This has a couple of advantages: - No need to sync up the round-robin scheduler in firmware/hardware with the round-robin airtime scheduler. - If several stations are eligible for transmission we can schedule both of them; no need to hard-block the scheduling rotation until the head of the queue has used up its quantum. - The check of whether a station is eligible for transmission becomes simpler (in ieee80211_txq_may_transmit()). The drawback is that scheduling becomes slightly more expensive, as we need to maintain an rbtree of TXQs sorted by virtual time. This means that ieee80211_register_airtime() becomes O(logN) in the number of currently scheduled TXQs because it can change the order of the scheduled stations. We mitigate this overhead by only resorting when a station changes position in the tree, and hopefully N rarely grows too big (it's only TXQs currently backlogged, not all associated stations), so it shouldn't be too big of an issue. To prevent divisions in the fast path, we maintain both station sums and pre-computed reciprocals of the sums. This turns the fast-path operation into a multiplication, with divisions only happening as the number of active stations change (to re-compute the current sum of all active station weights). To prevent this re-computation of the reciprocal from happening too frequently, we use a time-based notion of station activity, instead of updating the weight every time a station gets scheduled or de-scheduled. As queues can oscillate between empty and occupied quite frequently, this can significantly cut down on the number of re-computations. It also has the added benefit of making the station airtime calculation independent on whether the queue happened to have drained at the time an airtime value was accounted. Co-developed-by: Yibo Zhao <yiboz@codeaurora.org> Signed-off-by: Yibo Zhao <yiboz@codeaurora.org> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/r/20210623134755.235545-1-toke@redhat.com Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2021-06-23 21:47:55 +08:00
spin_lock_bh(&local->airtime[ac].lock);
sta->airtime[ac].rx_airtime = 0;
sta->airtime[ac].tx_airtime = 0;
mac80211: Switch to a virtual time-based airtime scheduler This switches the airtime scheduler in mac80211 to use a virtual time-based scheduler instead of the round-robin scheduler used before. This has a couple of advantages: - No need to sync up the round-robin scheduler in firmware/hardware with the round-robin airtime scheduler. - If several stations are eligible for transmission we can schedule both of them; no need to hard-block the scheduling rotation until the head of the queue has used up its quantum. - The check of whether a station is eligible for transmission becomes simpler (in ieee80211_txq_may_transmit()). The drawback is that scheduling becomes slightly more expensive, as we need to maintain an rbtree of TXQs sorted by virtual time. This means that ieee80211_register_airtime() becomes O(logN) in the number of currently scheduled TXQs because it can change the order of the scheduled stations. We mitigate this overhead by only resorting when a station changes position in the tree, and hopefully N rarely grows too big (it's only TXQs currently backlogged, not all associated stations), so it shouldn't be too big of an issue. To prevent divisions in the fast path, we maintain both station sums and pre-computed reciprocals of the sums. This turns the fast-path operation into a multiplication, with divisions only happening as the number of active stations change (to re-compute the current sum of all active station weights). To prevent this re-computation of the reciprocal from happening too frequently, we use a time-based notion of station activity, instead of updating the weight every time a station gets scheduled or de-scheduled. As queues can oscillate between empty and occupied quite frequently, this can significantly cut down on the number of re-computations. It also has the added benefit of making the station airtime calculation independent on whether the queue happened to have drained at the time an airtime value was accounted. Co-developed-by: Yibo Zhao <yiboz@codeaurora.org> Signed-off-by: Yibo Zhao <yiboz@codeaurora.org> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/r/20210623134755.235545-1-toke@redhat.com Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2021-06-23 21:47:55 +08:00
sta->airtime[ac].v_t = 0;
spin_unlock_bh(&local->airtime[ac].lock);
}
return count;
}
STA_OPS_RW(airtime);
static ssize_t sta_aql_read(struct file *file, char __user *userbuf,
size_t count, loff_t *ppos)
{
struct sta_info *sta = file->private_data;
struct ieee80211_local *local = sta->sdata->local;
size_t bufsz = 400;
char *buf = kzalloc(bufsz, GFP_KERNEL), *p = buf;
u32 q_depth[IEEE80211_NUM_ACS];
u32 q_limit_l[IEEE80211_NUM_ACS], q_limit_h[IEEE80211_NUM_ACS];
ssize_t rv;
int ac;
if (!buf)
return -ENOMEM;
for (ac = 0; ac < IEEE80211_NUM_ACS; ac++) {
mac80211: Switch to a virtual time-based airtime scheduler This switches the airtime scheduler in mac80211 to use a virtual time-based scheduler instead of the round-robin scheduler used before. This has a couple of advantages: - No need to sync up the round-robin scheduler in firmware/hardware with the round-robin airtime scheduler. - If several stations are eligible for transmission we can schedule both of them; no need to hard-block the scheduling rotation until the head of the queue has used up its quantum. - The check of whether a station is eligible for transmission becomes simpler (in ieee80211_txq_may_transmit()). The drawback is that scheduling becomes slightly more expensive, as we need to maintain an rbtree of TXQs sorted by virtual time. This means that ieee80211_register_airtime() becomes O(logN) in the number of currently scheduled TXQs because it can change the order of the scheduled stations. We mitigate this overhead by only resorting when a station changes position in the tree, and hopefully N rarely grows too big (it's only TXQs currently backlogged, not all associated stations), so it shouldn't be too big of an issue. To prevent divisions in the fast path, we maintain both station sums and pre-computed reciprocals of the sums. This turns the fast-path operation into a multiplication, with divisions only happening as the number of active stations change (to re-compute the current sum of all active station weights). To prevent this re-computation of the reciprocal from happening too frequently, we use a time-based notion of station activity, instead of updating the weight every time a station gets scheduled or de-scheduled. As queues can oscillate between empty and occupied quite frequently, this can significantly cut down on the number of re-computations. It also has the added benefit of making the station airtime calculation independent on whether the queue happened to have drained at the time an airtime value was accounted. Co-developed-by: Yibo Zhao <yiboz@codeaurora.org> Signed-off-by: Yibo Zhao <yiboz@codeaurora.org> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/r/20210623134755.235545-1-toke@redhat.com Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2021-06-23 21:47:55 +08:00
spin_lock_bh(&local->airtime[ac].lock);
mac80211: Implement Airtime-based Queue Limit (AQL) In order for the Fq_CoDel algorithm integrated in mac80211 layer to operate effectively to control excessive queueing latency, the CoDel algorithm requires an accurate measure of how long packets stays in the queue, AKA sojourn time. The sojourn time measured at the mac80211 layer doesn't include queueing latency in the lower layer (firmware/hardware) and CoDel expects lower layer to have a short queue. However, most 802.11ac chipsets offload tasks such TX aggregation to firmware or hardware, thus have a deep lower layer queue. Without a mechanism to control the lower layer queue size, packets only stay in mac80211 layer transiently before being sent to firmware queue. As a result, the sojourn time measured by CoDel in the mac80211 layer is almost always lower than the CoDel latency target, hence CoDel does little to control the latency, even when the lower layer queue causes excessive latency. The Byte Queue Limits (BQL) mechanism is commonly used to address the similar issue with wired network interface. However, this method cannot be applied directly to the wireless network interface. "Bytes" is not a suitable measure of queue depth in the wireless network, as the data rate can vary dramatically from station to station in the same network, from a few Mbps to over Gbps. This patch implements an Airtime-based Queue Limit (AQL) to make CoDel work effectively with wireless drivers that utilized firmware/hardware offloading. AQL allows each txq to release just enough packets to the lower layer to form 1-2 large aggregations to keep hardware fully utilized and retains the rest of the frames in mac80211 layer to be controlled by the CoDel algorithm. Signed-off-by: Kan Yan <kyan@google.com> [ Toke: Keep API to set pending airtime internal, fix nits in commit msg ] Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/r/20191119060610.76681-4-kyan@google.com Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2019-11-19 14:06:09 +08:00
q_limit_l[ac] = sta->airtime[ac].aql_limit_low;
q_limit_h[ac] = sta->airtime[ac].aql_limit_high;
mac80211: Switch to a virtual time-based airtime scheduler This switches the airtime scheduler in mac80211 to use a virtual time-based scheduler instead of the round-robin scheduler used before. This has a couple of advantages: - No need to sync up the round-robin scheduler in firmware/hardware with the round-robin airtime scheduler. - If several stations are eligible for transmission we can schedule both of them; no need to hard-block the scheduling rotation until the head of the queue has used up its quantum. - The check of whether a station is eligible for transmission becomes simpler (in ieee80211_txq_may_transmit()). The drawback is that scheduling becomes slightly more expensive, as we need to maintain an rbtree of TXQs sorted by virtual time. This means that ieee80211_register_airtime() becomes O(logN) in the number of currently scheduled TXQs because it can change the order of the scheduled stations. We mitigate this overhead by only resorting when a station changes position in the tree, and hopefully N rarely grows too big (it's only TXQs currently backlogged, not all associated stations), so it shouldn't be too big of an issue. To prevent divisions in the fast path, we maintain both station sums and pre-computed reciprocals of the sums. This turns the fast-path operation into a multiplication, with divisions only happening as the number of active stations change (to re-compute the current sum of all active station weights). To prevent this re-computation of the reciprocal from happening too frequently, we use a time-based notion of station activity, instead of updating the weight every time a station gets scheduled or de-scheduled. As queues can oscillate between empty and occupied quite frequently, this can significantly cut down on the number of re-computations. It also has the added benefit of making the station airtime calculation independent on whether the queue happened to have drained at the time an airtime value was accounted. Co-developed-by: Yibo Zhao <yiboz@codeaurora.org> Signed-off-by: Yibo Zhao <yiboz@codeaurora.org> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/r/20210623134755.235545-1-toke@redhat.com Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2021-06-23 21:47:55 +08:00
spin_unlock_bh(&local->airtime[ac].lock);
mac80211: Implement Airtime-based Queue Limit (AQL) In order for the Fq_CoDel algorithm integrated in mac80211 layer to operate effectively to control excessive queueing latency, the CoDel algorithm requires an accurate measure of how long packets stays in the queue, AKA sojourn time. The sojourn time measured at the mac80211 layer doesn't include queueing latency in the lower layer (firmware/hardware) and CoDel expects lower layer to have a short queue. However, most 802.11ac chipsets offload tasks such TX aggregation to firmware or hardware, thus have a deep lower layer queue. Without a mechanism to control the lower layer queue size, packets only stay in mac80211 layer transiently before being sent to firmware queue. As a result, the sojourn time measured by CoDel in the mac80211 layer is almost always lower than the CoDel latency target, hence CoDel does little to control the latency, even when the lower layer queue causes excessive latency. The Byte Queue Limits (BQL) mechanism is commonly used to address the similar issue with wired network interface. However, this method cannot be applied directly to the wireless network interface. "Bytes" is not a suitable measure of queue depth in the wireless network, as the data rate can vary dramatically from station to station in the same network, from a few Mbps to over Gbps. This patch implements an Airtime-based Queue Limit (AQL) to make CoDel work effectively with wireless drivers that utilized firmware/hardware offloading. AQL allows each txq to release just enough packets to the lower layer to form 1-2 large aggregations to keep hardware fully utilized and retains the rest of the frames in mac80211 layer to be controlled by the CoDel algorithm. Signed-off-by: Kan Yan <kyan@google.com> [ Toke: Keep API to set pending airtime internal, fix nits in commit msg ] Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/r/20191119060610.76681-4-kyan@google.com Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2019-11-19 14:06:09 +08:00
q_depth[ac] = atomic_read(&sta->airtime[ac].aql_tx_pending);
mac80211: Add airtime accounting and scheduling to TXQs This adds airtime accounting and scheduling to the mac80211 TXQ scheduler. A new callback, ieee80211_sta_register_airtime(), is added that drivers can call to report airtime usage for stations. When airtime information is present, mac80211 will schedule TXQs (through ieee80211_next_txq()) in a way that enforces airtime fairness between active stations. This scheduling works the same way as the ath9k in-driver airtime fairness scheduling. If no airtime usage is reported by the driver, the scheduler will default to round-robin scheduling. For drivers that don't control TXQ scheduling in software, a new API function, ieee80211_txq_may_transmit(), is added which the driver can use to check if the TXQ is eligible for transmission, or should be throttled to enforce fairness. Calls to this function must also be enclosed in ieee80211_txq_schedule_{start,end}() calls to ensure proper locking. The API ieee80211_txq_may_transmit() also ensures that TXQ list will be aligned aginst driver's own round-robin scheduler list. i.e it rotates the TXQ list till it makes the requested node becomes the first entry in TXQ list. Thus both the TXQ list and driver's list are in sync. Co-developed-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Louie Lu <git@louie.lu> [added debugfs write op to reset airtime counter] Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk> Signed-off-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2018-12-19 09:02:08 +08:00
}
p += scnprintf(p, bufsz + buf - p,
mac80211: Implement Airtime-based Queue Limit (AQL) In order for the Fq_CoDel algorithm integrated in mac80211 layer to operate effectively to control excessive queueing latency, the CoDel algorithm requires an accurate measure of how long packets stays in the queue, AKA sojourn time. The sojourn time measured at the mac80211 layer doesn't include queueing latency in the lower layer (firmware/hardware) and CoDel expects lower layer to have a short queue. However, most 802.11ac chipsets offload tasks such TX aggregation to firmware or hardware, thus have a deep lower layer queue. Without a mechanism to control the lower layer queue size, packets only stay in mac80211 layer transiently before being sent to firmware queue. As a result, the sojourn time measured by CoDel in the mac80211 layer is almost always lower than the CoDel latency target, hence CoDel does little to control the latency, even when the lower layer queue causes excessive latency. The Byte Queue Limits (BQL) mechanism is commonly used to address the similar issue with wired network interface. However, this method cannot be applied directly to the wireless network interface. "Bytes" is not a suitable measure of queue depth in the wireless network, as the data rate can vary dramatically from station to station in the same network, from a few Mbps to over Gbps. This patch implements an Airtime-based Queue Limit (AQL) to make CoDel work effectively with wireless drivers that utilized firmware/hardware offloading. AQL allows each txq to release just enough packets to the lower layer to form 1-2 large aggregations to keep hardware fully utilized and retains the rest of the frames in mac80211 layer to be controlled by the CoDel algorithm. Signed-off-by: Kan Yan <kyan@google.com> [ Toke: Keep API to set pending airtime internal, fix nits in commit msg ] Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/r/20191119060610.76681-4-kyan@google.com Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2019-11-19 14:06:09 +08:00
"Q depth: VO: %u us VI: %u us BE: %u us BK: %u us\n"
"Q limit[low/high]: VO: %u/%u VI: %u/%u BE: %u/%u BK: %u/%u\n",
q_depth[0], q_depth[1], q_depth[2], q_depth[3],
q_limit_l[0], q_limit_h[0], q_limit_l[1], q_limit_h[1],
q_limit_l[2], q_limit_h[2], q_limit_l[3], q_limit_h[3]);
mac80211: Add airtime accounting and scheduling to TXQs This adds airtime accounting and scheduling to the mac80211 TXQ scheduler. A new callback, ieee80211_sta_register_airtime(), is added that drivers can call to report airtime usage for stations. When airtime information is present, mac80211 will schedule TXQs (through ieee80211_next_txq()) in a way that enforces airtime fairness between active stations. This scheduling works the same way as the ath9k in-driver airtime fairness scheduling. If no airtime usage is reported by the driver, the scheduler will default to round-robin scheduling. For drivers that don't control TXQ scheduling in software, a new API function, ieee80211_txq_may_transmit(), is added which the driver can use to check if the TXQ is eligible for transmission, or should be throttled to enforce fairness. Calls to this function must also be enclosed in ieee80211_txq_schedule_{start,end}() calls to ensure proper locking. The API ieee80211_txq_may_transmit() also ensures that TXQ list will be aligned aginst driver's own round-robin scheduler list. i.e it rotates the TXQ list till it makes the requested node becomes the first entry in TXQ list. Thus both the TXQ list and driver's list are in sync. Co-developed-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Louie Lu <git@louie.lu> [added debugfs write op to reset airtime counter] Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk> Signed-off-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2018-12-19 09:02:08 +08:00
rv = simple_read_from_buffer(userbuf, count, ppos, buf, p - buf);
kfree(buf);
return rv;
}
static ssize_t sta_aql_write(struct file *file, const char __user *userbuf,
mac80211: Add airtime accounting and scheduling to TXQs This adds airtime accounting and scheduling to the mac80211 TXQ scheduler. A new callback, ieee80211_sta_register_airtime(), is added that drivers can call to report airtime usage for stations. When airtime information is present, mac80211 will schedule TXQs (through ieee80211_next_txq()) in a way that enforces airtime fairness between active stations. This scheduling works the same way as the ath9k in-driver airtime fairness scheduling. If no airtime usage is reported by the driver, the scheduler will default to round-robin scheduling. For drivers that don't control TXQ scheduling in software, a new API function, ieee80211_txq_may_transmit(), is added which the driver can use to check if the TXQ is eligible for transmission, or should be throttled to enforce fairness. Calls to this function must also be enclosed in ieee80211_txq_schedule_{start,end}() calls to ensure proper locking. The API ieee80211_txq_may_transmit() also ensures that TXQ list will be aligned aginst driver's own round-robin scheduler list. i.e it rotates the TXQ list till it makes the requested node becomes the first entry in TXQ list. Thus both the TXQ list and driver's list are in sync. Co-developed-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Louie Lu <git@louie.lu> [added debugfs write op to reset airtime counter] Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk> Signed-off-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2018-12-19 09:02:08 +08:00
size_t count, loff_t *ppos)
{
struct sta_info *sta = file->private_data;
mac80211: Implement Airtime-based Queue Limit (AQL) In order for the Fq_CoDel algorithm integrated in mac80211 layer to operate effectively to control excessive queueing latency, the CoDel algorithm requires an accurate measure of how long packets stays in the queue, AKA sojourn time. The sojourn time measured at the mac80211 layer doesn't include queueing latency in the lower layer (firmware/hardware) and CoDel expects lower layer to have a short queue. However, most 802.11ac chipsets offload tasks such TX aggregation to firmware or hardware, thus have a deep lower layer queue. Without a mechanism to control the lower layer queue size, packets only stay in mac80211 layer transiently before being sent to firmware queue. As a result, the sojourn time measured by CoDel in the mac80211 layer is almost always lower than the CoDel latency target, hence CoDel does little to control the latency, even when the lower layer queue causes excessive latency. The Byte Queue Limits (BQL) mechanism is commonly used to address the similar issue with wired network interface. However, this method cannot be applied directly to the wireless network interface. "Bytes" is not a suitable measure of queue depth in the wireless network, as the data rate can vary dramatically from station to station in the same network, from a few Mbps to over Gbps. This patch implements an Airtime-based Queue Limit (AQL) to make CoDel work effectively with wireless drivers that utilized firmware/hardware offloading. AQL allows each txq to release just enough packets to the lower layer to form 1-2 large aggregations to keep hardware fully utilized and retains the rest of the frames in mac80211 layer to be controlled by the CoDel algorithm. Signed-off-by: Kan Yan <kyan@google.com> [ Toke: Keep API to set pending airtime internal, fix nits in commit msg ] Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/r/20191119060610.76681-4-kyan@google.com Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2019-11-19 14:06:09 +08:00
u32 ac, q_limit_l, q_limit_h;
char _buf[100] = {}, *buf = _buf;
if (count > sizeof(_buf))
return -EINVAL;
if (copy_from_user(buf, userbuf, count))
return -EFAULT;
buf[sizeof(_buf) - 1] = '\0';
if (sscanf(buf, "limit %u %u %u", &ac, &q_limit_l, &q_limit_h)
mac80211: Implement Airtime-based Queue Limit (AQL) In order for the Fq_CoDel algorithm integrated in mac80211 layer to operate effectively to control excessive queueing latency, the CoDel algorithm requires an accurate measure of how long packets stays in the queue, AKA sojourn time. The sojourn time measured at the mac80211 layer doesn't include queueing latency in the lower layer (firmware/hardware) and CoDel expects lower layer to have a short queue. However, most 802.11ac chipsets offload tasks such TX aggregation to firmware or hardware, thus have a deep lower layer queue. Without a mechanism to control the lower layer queue size, packets only stay in mac80211 layer transiently before being sent to firmware queue. As a result, the sojourn time measured by CoDel in the mac80211 layer is almost always lower than the CoDel latency target, hence CoDel does little to control the latency, even when the lower layer queue causes excessive latency. The Byte Queue Limits (BQL) mechanism is commonly used to address the similar issue with wired network interface. However, this method cannot be applied directly to the wireless network interface. "Bytes" is not a suitable measure of queue depth in the wireless network, as the data rate can vary dramatically from station to station in the same network, from a few Mbps to over Gbps. This patch implements an Airtime-based Queue Limit (AQL) to make CoDel work effectively with wireless drivers that utilized firmware/hardware offloading. AQL allows each txq to release just enough packets to the lower layer to form 1-2 large aggregations to keep hardware fully utilized and retains the rest of the frames in mac80211 layer to be controlled by the CoDel algorithm. Signed-off-by: Kan Yan <kyan@google.com> [ Toke: Keep API to set pending airtime internal, fix nits in commit msg ] Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/r/20191119060610.76681-4-kyan@google.com Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2019-11-19 14:06:09 +08:00
!= 3)
return -EINVAL;
if (ac >= IEEE80211_NUM_ACS)
return -EINVAL;
sta->airtime[ac].aql_limit_low = q_limit_l;
sta->airtime[ac].aql_limit_high = q_limit_h;
mac80211: Add airtime accounting and scheduling to TXQs This adds airtime accounting and scheduling to the mac80211 TXQ scheduler. A new callback, ieee80211_sta_register_airtime(), is added that drivers can call to report airtime usage for stations. When airtime information is present, mac80211 will schedule TXQs (through ieee80211_next_txq()) in a way that enforces airtime fairness between active stations. This scheduling works the same way as the ath9k in-driver airtime fairness scheduling. If no airtime usage is reported by the driver, the scheduler will default to round-robin scheduling. For drivers that don't control TXQ scheduling in software, a new API function, ieee80211_txq_may_transmit(), is added which the driver can use to check if the TXQ is eligible for transmission, or should be throttled to enforce fairness. Calls to this function must also be enclosed in ieee80211_txq_schedule_{start,end}() calls to ensure proper locking. The API ieee80211_txq_may_transmit() also ensures that TXQ list will be aligned aginst driver's own round-robin scheduler list. i.e it rotates the TXQ list till it makes the requested node becomes the first entry in TXQ list. Thus both the TXQ list and driver's list are in sync. Co-developed-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Louie Lu <git@louie.lu> [added debugfs write op to reset airtime counter] Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk> Signed-off-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2018-12-19 09:02:08 +08:00
return count;
}
STA_OPS_RW(aql);
mac80211: Add airtime accounting and scheduling to TXQs This adds airtime accounting and scheduling to the mac80211 TXQ scheduler. A new callback, ieee80211_sta_register_airtime(), is added that drivers can call to report airtime usage for stations. When airtime information is present, mac80211 will schedule TXQs (through ieee80211_next_txq()) in a way that enforces airtime fairness between active stations. This scheduling works the same way as the ath9k in-driver airtime fairness scheduling. If no airtime usage is reported by the driver, the scheduler will default to round-robin scheduling. For drivers that don't control TXQ scheduling in software, a new API function, ieee80211_txq_may_transmit(), is added which the driver can use to check if the TXQ is eligible for transmission, or should be throttled to enforce fairness. Calls to this function must also be enclosed in ieee80211_txq_schedule_{start,end}() calls to ensure proper locking. The API ieee80211_txq_may_transmit() also ensures that TXQ list will be aligned aginst driver's own round-robin scheduler list. i.e it rotates the TXQ list till it makes the requested node becomes the first entry in TXQ list. Thus both the TXQ list and driver's list are in sync. Co-developed-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Louie Lu <git@louie.lu> [added debugfs write op to reset airtime counter] Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk> Signed-off-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2018-12-19 09:02:08 +08:00
static ssize_t sta_agg_status_read(struct file *file, char __user *userbuf,
size_t count, loff_t *ppos)
{
char *buf, *p;
ssize_t bufsz = 71 + IEEE80211_NUM_TIDS * 40;
int i;
struct sta_info *sta = file->private_data;
struct tid_ampdu_rx *tid_rx;
struct tid_ampdu_tx *tid_tx;
ssize_t ret;
buf = kzalloc(bufsz, GFP_KERNEL);
if (!buf)
return -ENOMEM;
p = buf;
rcu_read_lock();
p += scnprintf(p, bufsz + buf - p, "next dialog_token: %#02x\n",
sta->ampdu_mlme.dialog_token_allocator + 1);
p += scnprintf(p, bufsz + buf - p,
"TID\t\tRX\tDTKN\tSSN\t\tTX\tDTKN\tpending\n");
for (i = 0; i < IEEE80211_NUM_TIDS; i++) {
bool tid_rx_valid;
tid_rx = rcu_dereference(sta->ampdu_mlme.tid_rx[i]);
tid_tx = rcu_dereference(sta->ampdu_mlme.tid_tx[i]);
tid_rx_valid = test_bit(i, sta->ampdu_mlme.agg_session_valid);
p += scnprintf(p, bufsz + buf - p, "%02d", i);
p += scnprintf(p, bufsz + buf - p, "\t\t%x",
tid_rx_valid);
p += scnprintf(p, bufsz + buf - p, "\t%#.2x",
tid_rx_valid ?
sta->ampdu_mlme.tid_rx_token[i] : 0);
p += scnprintf(p, bufsz + buf - p, "\t%#.3x",
tid_rx ? tid_rx->ssn : 0);
p += scnprintf(p, bufsz + buf - p, "\t\t%x", !!tid_tx);
p += scnprintf(p, bufsz + buf - p, "\t%#.2x",
tid_tx ? tid_tx->dialog_token : 0);
p += scnprintf(p, bufsz + buf - p, "\t%03d",
tid_tx ? skb_queue_len(&tid_tx->pending) : 0);
p += scnprintf(p, bufsz + buf - p, "\n");
}
rcu_read_unlock();
ret = simple_read_from_buffer(userbuf, count, ppos, buf, p - buf);
kfree(buf);
return ret;
}
static ssize_t sta_agg_status_write(struct file *file, const char __user *userbuf,
size_t count, loff_t *ppos)
{
char _buf[25] = {}, *buf = _buf;
struct sta_info *sta = file->private_data;
bool start, tx;
unsigned long tid;
char *pos;
int ret, timeout = 5000;
if (count > sizeof(_buf))
return -EINVAL;
if (copy_from_user(buf, userbuf, count))
return -EFAULT;
buf[sizeof(_buf) - 1] = '\0';
pos = buf;
buf = strsep(&pos, " ");
if (!buf)
return -EINVAL;
if (!strcmp(buf, "tx"))
tx = true;
else if (!strcmp(buf, "rx"))
tx = false;
else
return -EINVAL;
buf = strsep(&pos, " ");
if (!buf)
return -EINVAL;
if (!strcmp(buf, "start")) {
start = true;
if (!tx)
return -EINVAL;
} else if (!strcmp(buf, "stop")) {
start = false;
} else {
return -EINVAL;
}
buf = strsep(&pos, " ");
if (!buf)
return -EINVAL;
if (sscanf(buf, "timeout=%d", &timeout) == 1) {
buf = strsep(&pos, " ");
if (!buf || !tx || !start)
return -EINVAL;
}
ret = kstrtoul(buf, 0, &tid);
if (ret || tid >= IEEE80211_NUM_TIDS)
return -EINVAL;
if (tx) {
if (start)
ret = ieee80211_start_tx_ba_session(&sta->sta, tid,
timeout);
else
ret = ieee80211_stop_tx_ba_session(&sta->sta, tid);
} else {
__ieee80211_stop_rx_ba_session(sta, tid, WLAN_BACK_RECIPIENT,
3, true);
ret = 0;
}
return ret ?: count;
}
STA_OPS_RW(agg_status);
static ssize_t sta_ht_capa_read(struct file *file, char __user *userbuf,
size_t count, loff_t *ppos)
{
#define PRINT_HT_CAP(_cond, _str) \
do { \
if (_cond) \
p += scnprintf(p, sizeof(buf)+buf-p, "\t" _str "\n"); \
} while (0)
char *buf, *p;
int i;
ssize_t bufsz = 512;
struct sta_info *sta = file->private_data;
struct ieee80211_sta_ht_cap *htc = &sta->sta.ht_cap;
ssize_t ret;
buf = kzalloc(bufsz, GFP_KERNEL);
if (!buf)
return -ENOMEM;
p = buf;
p += scnprintf(p, bufsz + buf - p, "ht %ssupported\n",
htc->ht_supported ? "" : "not ");
if (htc->ht_supported) {
p += scnprintf(p, bufsz + buf - p, "cap: %#.4x\n", htc->cap);
PRINT_HT_CAP((htc->cap & BIT(0)), "RX LDPC");
PRINT_HT_CAP((htc->cap & BIT(1)), "HT20/HT40");
PRINT_HT_CAP(!(htc->cap & BIT(1)), "HT20");
PRINT_HT_CAP(((htc->cap >> 2) & 0x3) == 0, "Static SM Power Save");
PRINT_HT_CAP(((htc->cap >> 2) & 0x3) == 1, "Dynamic SM Power Save");
PRINT_HT_CAP(((htc->cap >> 2) & 0x3) == 3, "SM Power Save disabled");
PRINT_HT_CAP((htc->cap & BIT(4)), "RX Greenfield");
PRINT_HT_CAP((htc->cap & BIT(5)), "RX HT20 SGI");
PRINT_HT_CAP((htc->cap & BIT(6)), "RX HT40 SGI");
PRINT_HT_CAP((htc->cap & BIT(7)), "TX STBC");
PRINT_HT_CAP(((htc->cap >> 8) & 0x3) == 0, "No RX STBC");
PRINT_HT_CAP(((htc->cap >> 8) & 0x3) == 1, "RX STBC 1-stream");
PRINT_HT_CAP(((htc->cap >> 8) & 0x3) == 2, "RX STBC 2-streams");
PRINT_HT_CAP(((htc->cap >> 8) & 0x3) == 3, "RX STBC 3-streams");
PRINT_HT_CAP((htc->cap & BIT(10)), "HT Delayed Block Ack");
PRINT_HT_CAP(!(htc->cap & BIT(11)), "Max AMSDU length: "
"3839 bytes");
PRINT_HT_CAP((htc->cap & BIT(11)), "Max AMSDU length: "
"7935 bytes");
/*
* For beacons and probe response this would mean the BSS
* does or does not allow the usage of DSSS/CCK HT40.
* Otherwise it means the STA does or does not use
* DSSS/CCK HT40.
*/
PRINT_HT_CAP((htc->cap & BIT(12)), "DSSS/CCK HT40");
PRINT_HT_CAP(!(htc->cap & BIT(12)), "No DSSS/CCK HT40");
/* BIT(13) is reserved */
PRINT_HT_CAP((htc->cap & BIT(14)), "40 MHz Intolerant");
PRINT_HT_CAP((htc->cap & BIT(15)), "L-SIG TXOP protection");
p += scnprintf(p, bufsz + buf - p, "ampdu factor/density: %d/%d\n",
htc->ampdu_factor, htc->ampdu_density);
p += scnprintf(p, bufsz + buf - p, "MCS mask:");
for (i = 0; i < IEEE80211_HT_MCS_MASK_LEN; i++)
p += scnprintf(p, bufsz + buf - p, " %.2x",
htc->mcs.rx_mask[i]);
p += scnprintf(p, bufsz + buf - p, "\n");
/* If not set this is meaningless */
if (le16_to_cpu(htc->mcs.rx_highest)) {
p += scnprintf(p, bufsz + buf - p,
"MCS rx highest: %d Mbps\n",
le16_to_cpu(htc->mcs.rx_highest));
}
p += scnprintf(p, bufsz + buf - p, "MCS tx params: %x\n",
htc->mcs.tx_params);
}
ret = simple_read_from_buffer(userbuf, count, ppos, buf, p - buf);
kfree(buf);
return ret;
}
STA_OPS(ht_capa);
static ssize_t sta_vht_capa_read(struct file *file, char __user *userbuf,
size_t count, loff_t *ppos)
{
char *buf, *p;
struct sta_info *sta = file->private_data;
struct ieee80211_sta_vht_cap *vhtc = &sta->sta.vht_cap;
ssize_t ret;
ssize_t bufsz = 512;
buf = kzalloc(bufsz, GFP_KERNEL);
if (!buf)
return -ENOMEM;
p = buf;
p += scnprintf(p, bufsz + buf - p, "VHT %ssupported\n",
vhtc->vht_supported ? "" : "not ");
if (vhtc->vht_supported) {
p += scnprintf(p, bufsz + buf - p, "cap: %#.8x\n",
vhtc->cap);
#define PFLAG(a, b) \
do { \
if (vhtc->cap & IEEE80211_VHT_CAP_ ## a) \
p += scnprintf(p, bufsz + buf - p, \
"\t\t%s\n", b); \
} while (0)
switch (vhtc->cap & 0x3) {
case IEEE80211_VHT_CAP_MAX_MPDU_LENGTH_3895:
p += scnprintf(p, bufsz + buf - p,
"\t\tMAX-MPDU-3895\n");
break;
case IEEE80211_VHT_CAP_MAX_MPDU_LENGTH_7991:
p += scnprintf(p, bufsz + buf - p,
"\t\tMAX-MPDU-7991\n");
break;
case IEEE80211_VHT_CAP_MAX_MPDU_LENGTH_11454:
p += scnprintf(p, bufsz + buf - p,
"\t\tMAX-MPDU-11454\n");
break;
default:
p += scnprintf(p, bufsz + buf - p,
"\t\tMAX-MPDU-UNKNOWN\n");
}
switch (vhtc->cap & IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_MASK) {
case 0:
p += scnprintf(p, bufsz + buf - p,
"\t\t80Mhz\n");
break;
case IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_160MHZ:
p += scnprintf(p, bufsz + buf - p,
"\t\t160Mhz\n");
break;
case IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_160_80PLUS80MHZ:
p += scnprintf(p, bufsz + buf - p,
"\t\t80+80Mhz\n");
break;
default:
p += scnprintf(p, bufsz + buf - p,
"\t\tUNKNOWN-MHZ: 0x%x\n",
(vhtc->cap >> 2) & 0x3);
}
PFLAG(RXLDPC, "RXLDPC");
PFLAG(SHORT_GI_80, "SHORT-GI-80");
PFLAG(SHORT_GI_160, "SHORT-GI-160");
PFLAG(TXSTBC, "TXSTBC");
p += scnprintf(p, bufsz + buf - p,
"\t\tRXSTBC_%d\n", (vhtc->cap >> 8) & 0x7);
PFLAG(SU_BEAMFORMER_CAPABLE, "SU-BEAMFORMER-CAPABLE");
PFLAG(SU_BEAMFORMEE_CAPABLE, "SU-BEAMFORMEE-CAPABLE");
p += scnprintf(p, bufsz + buf - p,
"\t\tBEAMFORMEE-STS: 0x%x\n",
(vhtc->cap & IEEE80211_VHT_CAP_BEAMFORMEE_STS_MASK) >>
IEEE80211_VHT_CAP_BEAMFORMEE_STS_SHIFT);
p += scnprintf(p, bufsz + buf - p,
"\t\tSOUNDING-DIMENSIONS: 0x%x\n",
(vhtc->cap & IEEE80211_VHT_CAP_SOUNDING_DIMENSIONS_MASK)
>> IEEE80211_VHT_CAP_SOUNDING_DIMENSIONS_SHIFT);
PFLAG(MU_BEAMFORMER_CAPABLE, "MU-BEAMFORMER-CAPABLE");
PFLAG(MU_BEAMFORMEE_CAPABLE, "MU-BEAMFORMEE-CAPABLE");
PFLAG(VHT_TXOP_PS, "TXOP-PS");
PFLAG(HTC_VHT, "HTC-VHT");
p += scnprintf(p, bufsz + buf - p,
"\t\tMPDU-LENGTH-EXPONENT: 0x%x\n",
(vhtc->cap & IEEE80211_VHT_CAP_MAX_A_MPDU_LENGTH_EXPONENT_MASK) >>
IEEE80211_VHT_CAP_MAX_A_MPDU_LENGTH_EXPONENT_SHIFT);
PFLAG(VHT_LINK_ADAPTATION_VHT_UNSOL_MFB,
"LINK-ADAPTATION-VHT-UNSOL-MFB");
p += scnprintf(p, bufsz + buf - p,
"\t\tLINK-ADAPTATION-VHT-MRQ-MFB: 0x%x\n",
(vhtc->cap & IEEE80211_VHT_CAP_VHT_LINK_ADAPTATION_VHT_MRQ_MFB) >> 26);
PFLAG(RX_ANTENNA_PATTERN, "RX-ANTENNA-PATTERN");
PFLAG(TX_ANTENNA_PATTERN, "TX-ANTENNA-PATTERN");
p += scnprintf(p, bufsz + buf - p, "RX MCS: %.4x\n",
le16_to_cpu(vhtc->vht_mcs.rx_mcs_map));
if (vhtc->vht_mcs.rx_highest)
p += scnprintf(p, bufsz + buf - p,
"MCS RX highest: %d Mbps\n",
le16_to_cpu(vhtc->vht_mcs.rx_highest));
p += scnprintf(p, bufsz + buf - p, "TX MCS: %.4x\n",
le16_to_cpu(vhtc->vht_mcs.tx_mcs_map));
if (vhtc->vht_mcs.tx_highest)
p += scnprintf(p, bufsz + buf - p,
"MCS TX highest: %d Mbps\n",
le16_to_cpu(vhtc->vht_mcs.tx_highest));
#undef PFLAG
}
ret = simple_read_from_buffer(userbuf, count, ppos, buf, p - buf);
kfree(buf);
return ret;
}
STA_OPS(vht_capa);
static ssize_t sta_he_capa_read(struct file *file, char __user *userbuf,
size_t count, loff_t *ppos)
{
char *buf, *p;
size_t buf_sz = PAGE_SIZE;
struct sta_info *sta = file->private_data;
struct ieee80211_sta_he_cap *hec = &sta->sta.he_cap;
struct ieee80211_he_mcs_nss_supp *nss = &hec->he_mcs_nss_supp;
u8 ppe_size;
u8 *cap;
int i;
ssize_t ret;
buf = kmalloc(buf_sz, GFP_KERNEL);
if (!buf)
return -ENOMEM;
p = buf;
p += scnprintf(p, buf_sz + buf - p, "HE %ssupported\n",
hec->has_he ? "" : "not ");
if (!hec->has_he)
goto out;
cap = hec->he_cap_elem.mac_cap_info;
p += scnprintf(p, buf_sz + buf - p,
"MAC-CAP: %#.2x %#.2x %#.2x %#.2x %#.2x %#.2x\n",
cap[0], cap[1], cap[2], cap[3], cap[4], cap[5]);
#define PRINT(fmt, ...) \
p += scnprintf(p, buf_sz + buf - p, "\t\t" fmt "\n", \
##__VA_ARGS__)
#define PFLAG(t, n, a, b) \
do { \
if (cap[n] & IEEE80211_HE_##t##_CAP##n##_##a) \
PRINT("%s", b); \
} while (0)
#define PFLAG_RANGE(t, i, n, s, m, off, fmt) \
do { \
u8 msk = IEEE80211_HE_##t##_CAP##i##_##n##_MASK; \
u8 idx = ((cap[i] & msk) >> (ffs(msk) - 1)) + off; \
PRINT(fmt, (s << idx) + (m * idx)); \
} while (0)
#define PFLAG_RANGE_DEFAULT(t, i, n, s, m, off, fmt, a, b) \
do { \
if (cap[i] == IEEE80211_HE_##t ##_CAP##i##_##n##_##a) { \
PRINT("%s", b); \
break; \
} \
PFLAG_RANGE(t, i, n, s, m, off, fmt); \
} while (0)
PFLAG(MAC, 0, HTC_HE, "HTC-HE");
PFLAG(MAC, 0, TWT_REQ, "TWT-REQ");
PFLAG(MAC, 0, TWT_RES, "TWT-RES");
PFLAG_RANGE_DEFAULT(MAC, 0, DYNAMIC_FRAG, 0, 1, 0,
"DYNAMIC-FRAG-LEVEL-%d", NOT_SUPP, "NOT-SUPP");
PFLAG_RANGE_DEFAULT(MAC, 0, MAX_NUM_FRAG_MSDU, 1, 0, 0,
"MAX-NUM-FRAG-MSDU-%d", UNLIMITED, "UNLIMITED");
PFLAG_RANGE_DEFAULT(MAC, 1, MIN_FRAG_SIZE, 128, 0, -1,
"MIN-FRAG-SIZE-%d", UNLIMITED, "UNLIMITED");
PFLAG_RANGE_DEFAULT(MAC, 1, TF_MAC_PAD_DUR, 0, 8, 0,
"TF-MAC-PAD-DUR-%dUS", MASK, "UNKNOWN");
PFLAG_RANGE(MAC, 1, MULTI_TID_AGG_RX_QOS, 0, 1, 1,
"MULTI-TID-AGG-RX-QOS-%d");
if (cap[0] & IEEE80211_HE_MAC_CAP0_HTC_HE) {
switch (((cap[2] << 1) | (cap[1] >> 7)) & 0x3) {
case 0:
PRINT("LINK-ADAPTATION-NO-FEEDBACK");
break;
case 1:
PRINT("LINK-ADAPTATION-RESERVED");
break;
case 2:
PRINT("LINK-ADAPTATION-UNSOLICITED-FEEDBACK");
break;
case 3:
PRINT("LINK-ADAPTATION-BOTH");
break;
}
}
PFLAG(MAC, 2, ALL_ACK, "ALL-ACK");
PFLAG(MAC, 2, TRS, "TRS");
PFLAG(MAC, 2, BSR, "BSR");
PFLAG(MAC, 2, BCAST_TWT, "BCAST-TWT");
PFLAG(MAC, 2, 32BIT_BA_BITMAP, "32BIT-BA-BITMAP");
PFLAG(MAC, 2, MU_CASCADING, "MU-CASCADING");
PFLAG(MAC, 2, ACK_EN, "ACK-EN");
PFLAG(MAC, 3, OMI_CONTROL, "OMI-CONTROL");
PFLAG(MAC, 3, OFDMA_RA, "OFDMA-RA");
switch (cap[3] & IEEE80211_HE_MAC_CAP3_MAX_AMPDU_LEN_EXP_MASK) {
case IEEE80211_HE_MAC_CAP3_MAX_AMPDU_LEN_EXP_EXT_0:
PRINT("MAX-AMPDU-LEN-EXP-USE-EXT-0");
break;
case IEEE80211_HE_MAC_CAP3_MAX_AMPDU_LEN_EXP_EXT_1:
PRINT("MAX-AMPDU-LEN-EXP-VHT-EXT-1");
break;
case IEEE80211_HE_MAC_CAP3_MAX_AMPDU_LEN_EXP_EXT_2:
PRINT("MAX-AMPDU-LEN-EXP-VHT-EXT-2");
break;
case IEEE80211_HE_MAC_CAP3_MAX_AMPDU_LEN_EXP_EXT_3:
PRINT("MAX-AMPDU-LEN-EXP-VHT-EXT-3");
break;
}
PFLAG(MAC, 3, AMSDU_FRAG, "AMSDU-FRAG");
PFLAG(MAC, 3, FLEX_TWT_SCHED, "FLEX-TWT-SCHED");
PFLAG(MAC, 3, RX_CTRL_FRAME_TO_MULTIBSS, "RX-CTRL-FRAME-TO-MULTIBSS");
PFLAG(MAC, 4, BSRP_BQRP_A_MPDU_AGG, "BSRP-BQRP-A-MPDU-AGG");
PFLAG(MAC, 4, QTP, "QTP");
PFLAG(MAC, 4, BQR, "BQR");
PFLAG(MAC, 4, PSR_RESP, "PSR-RESP");
PFLAG(MAC, 4, NDP_FB_REP, "NDP-FB-REP");
PFLAG(MAC, 4, OPS, "OPS");
PFLAG(MAC, 4, AMSDU_IN_AMPDU, "AMSDU-IN-AMPDU");
PRINT("MULTI-TID-AGG-TX-QOS-%d", ((cap[5] << 1) | (cap[4] >> 7)) & 0x7);
PFLAG(MAC, 5, SUBCHAN_SELECTIVE_TRANSMISSION,
"SUBCHAN-SELECTIVE-TRANSMISSION");
PFLAG(MAC, 5, UL_2x996_TONE_RU, "UL-2x996-TONE-RU");
PFLAG(MAC, 5, OM_CTRL_UL_MU_DATA_DIS_RX, "OM-CTRL-UL-MU-DATA-DIS-RX");
PFLAG(MAC, 5, HE_DYNAMIC_SM_PS, "HE-DYNAMIC-SM-PS");
PFLAG(MAC, 5, PUNCTURED_SOUNDING, "PUNCTURED-SOUNDING");
PFLAG(MAC, 5, HT_VHT_TRIG_FRAME_RX, "HT-VHT-TRIG-FRAME-RX");
cap = hec->he_cap_elem.phy_cap_info;
p += scnprintf(p, buf_sz + buf - p,
"PHY CAP: %#.2x %#.2x %#.2x %#.2x %#.2x %#.2x %#.2x %#.2x %#.2x %#.2x %#.2x\n",
cap[0], cap[1], cap[2], cap[3], cap[4], cap[5], cap[6],
cap[7], cap[8], cap[9], cap[10]);
PFLAG(PHY, 0, CHANNEL_WIDTH_SET_40MHZ_IN_2G,
"CHANNEL-WIDTH-SET-40MHZ-IN-2G");
PFLAG(PHY, 0, CHANNEL_WIDTH_SET_40MHZ_80MHZ_IN_5G,
"CHANNEL-WIDTH-SET-40MHZ-80MHZ-IN-5G");
PFLAG(PHY, 0, CHANNEL_WIDTH_SET_160MHZ_IN_5G,
"CHANNEL-WIDTH-SET-160MHZ-IN-5G");
PFLAG(PHY, 0, CHANNEL_WIDTH_SET_80PLUS80_MHZ_IN_5G,
"CHANNEL-WIDTH-SET-80PLUS80-MHZ-IN-5G");
PFLAG(PHY, 0, CHANNEL_WIDTH_SET_RU_MAPPING_IN_2G,
"CHANNEL-WIDTH-SET-RU-MAPPING-IN-2G");
PFLAG(PHY, 0, CHANNEL_WIDTH_SET_RU_MAPPING_IN_5G,
"CHANNEL-WIDTH-SET-RU-MAPPING-IN-5G");
switch (cap[1] & IEEE80211_HE_PHY_CAP1_PREAMBLE_PUNC_RX_MASK) {
case IEEE80211_HE_PHY_CAP1_PREAMBLE_PUNC_RX_80MHZ_ONLY_SECOND_20MHZ:
PRINT("PREAMBLE-PUNC-RX-80MHZ-ONLY-SECOND-20MHZ");
break;
case IEEE80211_HE_PHY_CAP1_PREAMBLE_PUNC_RX_80MHZ_ONLY_SECOND_40MHZ:
PRINT("PREAMBLE-PUNC-RX-80MHZ-ONLY-SECOND-40MHZ");
break;
case IEEE80211_HE_PHY_CAP1_PREAMBLE_PUNC_RX_160MHZ_ONLY_SECOND_20MHZ:
PRINT("PREAMBLE-PUNC-RX-160MHZ-ONLY-SECOND-20MHZ");
break;
case IEEE80211_HE_PHY_CAP1_PREAMBLE_PUNC_RX_160MHZ_ONLY_SECOND_40MHZ:
PRINT("PREAMBLE-PUNC-RX-160MHZ-ONLY-SECOND-40MHZ");
break;
}
PFLAG(PHY, 1, DEVICE_CLASS_A,
"IEEE80211-HE-PHY-CAP1-DEVICE-CLASS-A");
PFLAG(PHY, 1, LDPC_CODING_IN_PAYLOAD,
"LDPC-CODING-IN-PAYLOAD");
PFLAG(PHY, 1, HE_LTF_AND_GI_FOR_HE_PPDUS_0_8US,
"HY-CAP1-HE-LTF-AND-GI-FOR-HE-PPDUS-0-8US");
PRINT("MIDAMBLE-RX-MAX-NSTS-%d", ((cap[2] << 1) | (cap[1] >> 7)) & 0x3);
PFLAG(PHY, 2, NDP_4x_LTF_AND_3_2US, "NDP-4X-LTF-AND-3-2US");
PFLAG(PHY, 2, STBC_TX_UNDER_80MHZ, "STBC-TX-UNDER-80MHZ");
PFLAG(PHY, 2, STBC_RX_UNDER_80MHZ, "STBC-RX-UNDER-80MHZ");
PFLAG(PHY, 2, DOPPLER_TX, "DOPPLER-TX");
PFLAG(PHY, 2, DOPPLER_RX, "DOPPLER-RX");
PFLAG(PHY, 2, UL_MU_FULL_MU_MIMO, "UL-MU-FULL-MU-MIMO");
PFLAG(PHY, 2, UL_MU_PARTIAL_MU_MIMO, "UL-MU-PARTIAL-MU-MIMO");
switch (cap[3] & IEEE80211_HE_PHY_CAP3_DCM_MAX_CONST_TX_MASK) {
case IEEE80211_HE_PHY_CAP3_DCM_MAX_CONST_TX_NO_DCM:
PRINT("DCM-MAX-CONST-TX-NO-DCM");
break;
case IEEE80211_HE_PHY_CAP3_DCM_MAX_CONST_TX_BPSK:
PRINT("DCM-MAX-CONST-TX-BPSK");
break;
case IEEE80211_HE_PHY_CAP3_DCM_MAX_CONST_TX_QPSK:
PRINT("DCM-MAX-CONST-TX-QPSK");
break;
case IEEE80211_HE_PHY_CAP3_DCM_MAX_CONST_TX_16_QAM:
PRINT("DCM-MAX-CONST-TX-16-QAM");
break;
}
PFLAG(PHY, 3, DCM_MAX_TX_NSS_1, "DCM-MAX-TX-NSS-1");
PFLAG(PHY, 3, DCM_MAX_TX_NSS_2, "DCM-MAX-TX-NSS-2");
switch (cap[3] & IEEE80211_HE_PHY_CAP3_DCM_MAX_CONST_RX_MASK) {
case IEEE80211_HE_PHY_CAP3_DCM_MAX_CONST_RX_NO_DCM:
PRINT("DCM-MAX-CONST-RX-NO-DCM");
break;
case IEEE80211_HE_PHY_CAP3_DCM_MAX_CONST_RX_BPSK:
PRINT("DCM-MAX-CONST-RX-BPSK");
break;
case IEEE80211_HE_PHY_CAP3_DCM_MAX_CONST_RX_QPSK:
PRINT("DCM-MAX-CONST-RX-QPSK");
break;
case IEEE80211_HE_PHY_CAP3_DCM_MAX_CONST_RX_16_QAM:
PRINT("DCM-MAX-CONST-RX-16-QAM");
break;
}
PFLAG(PHY, 3, DCM_MAX_RX_NSS_1, "DCM-MAX-RX-NSS-1");
PFLAG(PHY, 3, DCM_MAX_RX_NSS_2, "DCM-MAX-RX-NSS-2");
PFLAG(PHY, 3, RX_PARTIAL_BW_SU_IN_20MHZ_MU,
"RX-PARTIAL-BW-SU-IN-20MHZ-MU");
PFLAG(PHY, 3, SU_BEAMFORMER, "SU-BEAMFORMER");
PFLAG(PHY, 4, SU_BEAMFORMEE, "SU-BEAMFORMEE");
PFLAG(PHY, 4, MU_BEAMFORMER, "MU-BEAMFORMER");
PFLAG_RANGE(PHY, 4, BEAMFORMEE_MAX_STS_UNDER_80MHZ, 0, 1, 4,
"BEAMFORMEE-MAX-STS-UNDER-%d");
PFLAG_RANGE(PHY, 4, BEAMFORMEE_MAX_STS_ABOVE_80MHZ, 0, 1, 4,
"BEAMFORMEE-MAX-STS-ABOVE-%d");
PFLAG_RANGE(PHY, 5, BEAMFORMEE_NUM_SND_DIM_UNDER_80MHZ, 0, 1, 1,
"NUM-SND-DIM-UNDER-80MHZ-%d");
PFLAG_RANGE(PHY, 5, BEAMFORMEE_NUM_SND_DIM_ABOVE_80MHZ, 0, 1, 1,
"NUM-SND-DIM-ABOVE-80MHZ-%d");
PFLAG(PHY, 5, NG16_SU_FEEDBACK, "NG16-SU-FEEDBACK");
PFLAG(PHY, 5, NG16_MU_FEEDBACK, "NG16-MU-FEEDBACK");
PFLAG(PHY, 6, CODEBOOK_SIZE_42_SU, "CODEBOOK-SIZE-42-SU");
PFLAG(PHY, 6, CODEBOOK_SIZE_75_MU, "CODEBOOK-SIZE-75-MU");
PFLAG(PHY, 6, TRIG_SU_BEAMFORMING_FB, "TRIG-SU-BEAMFORMING-FB");
PFLAG(PHY, 6, TRIG_MU_BEAMFORMING_PARTIAL_BW_FB,
"MU-BEAMFORMING-PARTIAL-BW-FB");
PFLAG(PHY, 6, TRIG_CQI_FB, "TRIG-CQI-FB");
PFLAG(PHY, 6, PARTIAL_BW_EXT_RANGE, "PARTIAL-BW-EXT-RANGE");
PFLAG(PHY, 6, PARTIAL_BANDWIDTH_DL_MUMIMO,
"PARTIAL-BANDWIDTH-DL-MUMIMO");
PFLAG(PHY, 6, PPE_THRESHOLD_PRESENT, "PPE-THRESHOLD-PRESENT");
PFLAG(PHY, 7, PSR_BASED_SR, "PSR-BASED-SR");
PFLAG(PHY, 7, POWER_BOOST_FACTOR_SUPP, "POWER-BOOST-FACTOR-SUPP");
PFLAG(PHY, 7, HE_SU_MU_PPDU_4XLTF_AND_08_US_GI,
"HE-SU-MU-PPDU-4XLTF-AND-08-US-GI");
PFLAG_RANGE(PHY, 7, MAX_NC, 0, 1, 1, "MAX-NC-%d");
PFLAG(PHY, 7, STBC_TX_ABOVE_80MHZ, "STBC-TX-ABOVE-80MHZ");
PFLAG(PHY, 7, STBC_RX_ABOVE_80MHZ, "STBC-RX-ABOVE-80MHZ");
PFLAG(PHY, 8, HE_ER_SU_PPDU_4XLTF_AND_08_US_GI,
"HE-ER-SU-PPDU-4XLTF-AND-08-US-GI");
PFLAG(PHY, 8, 20MHZ_IN_40MHZ_HE_PPDU_IN_2G,
"20MHZ-IN-40MHZ-HE-PPDU-IN-2G");
PFLAG(PHY, 8, 20MHZ_IN_160MHZ_HE_PPDU, "20MHZ-IN-160MHZ-HE-PPDU");
PFLAG(PHY, 8, 80MHZ_IN_160MHZ_HE_PPDU, "80MHZ-IN-160MHZ-HE-PPDU");
PFLAG(PHY, 8, HE_ER_SU_1XLTF_AND_08_US_GI,
"HE-ER-SU-1XLTF-AND-08-US-GI");
PFLAG(PHY, 8, MIDAMBLE_RX_TX_2X_AND_1XLTF,
"MIDAMBLE-RX-TX-2X-AND-1XLTF");
switch (cap[8] & IEEE80211_HE_PHY_CAP8_DCM_MAX_RU_MASK) {
case IEEE80211_HE_PHY_CAP8_DCM_MAX_RU_242:
PRINT("DCM-MAX-RU-242");
break;
case IEEE80211_HE_PHY_CAP8_DCM_MAX_RU_484:
PRINT("DCM-MAX-RU-484");
break;
case IEEE80211_HE_PHY_CAP8_DCM_MAX_RU_996:
PRINT("DCM-MAX-RU-996");
break;
case IEEE80211_HE_PHY_CAP8_DCM_MAX_RU_2x996:
PRINT("DCM-MAX-RU-2x996");
break;
}
PFLAG(PHY, 9, LONGER_THAN_16_SIGB_OFDM_SYM,
"LONGER-THAN-16-SIGB-OFDM-SYM");
PFLAG(PHY, 9, NON_TRIGGERED_CQI_FEEDBACK,
"NON-TRIGGERED-CQI-FEEDBACK");
PFLAG(PHY, 9, TX_1024_QAM_LESS_THAN_242_TONE_RU,
"TX-1024-QAM-LESS-THAN-242-TONE-RU");
PFLAG(PHY, 9, RX_1024_QAM_LESS_THAN_242_TONE_RU,
"RX-1024-QAM-LESS-THAN-242-TONE-RU");
PFLAG(PHY, 9, RX_FULL_BW_SU_USING_MU_WITH_COMP_SIGB,
"RX-FULL-BW-SU-USING-MU-WITH-COMP-SIGB");
PFLAG(PHY, 9, RX_FULL_BW_SU_USING_MU_WITH_NON_COMP_SIGB,
"RX-FULL-BW-SU-USING-MU-WITH-NON-COMP-SIGB");
switch (u8_get_bits(cap[9],
IEEE80211_HE_PHY_CAP9_NOMINAL_PKT_PADDING_MASK)) {
case IEEE80211_HE_PHY_CAP9_NOMINAL_PKT_PADDING_0US:
PRINT("NOMINAL-PACKET-PADDING-0US");
break;
case IEEE80211_HE_PHY_CAP9_NOMINAL_PKT_PADDING_8US:
PRINT("NOMINAL-PACKET-PADDING-8US");
break;
case IEEE80211_HE_PHY_CAP9_NOMINAL_PKT_PADDING_16US:
PRINT("NOMINAL-PACKET-PADDING-16US");
break;
}
#undef PFLAG_RANGE_DEFAULT
#undef PFLAG_RANGE
#undef PFLAG
#define PRINT_NSS_SUPP(f, n) \
do { \
int _i; \
u16 v = le16_to_cpu(nss->f); \
p += scnprintf(p, buf_sz + buf - p, n ": %#.4x\n", v); \
for (_i = 0; _i < 8; _i += 2) { \
switch ((v >> _i) & 0x3) { \
case 0: \
PRINT(n "-%d-SUPPORT-0-7", _i / 2); \
break; \
case 1: \
PRINT(n "-%d-SUPPORT-0-9", _i / 2); \
break; \
case 2: \
PRINT(n "-%d-SUPPORT-0-11", _i / 2); \
break; \
case 3: \
PRINT(n "-%d-NOT-SUPPORTED", _i / 2); \
break; \
} \
} \
} while (0)
PRINT_NSS_SUPP(rx_mcs_80, "RX-MCS-80");
PRINT_NSS_SUPP(tx_mcs_80, "TX-MCS-80");
if (cap[0] & IEEE80211_HE_PHY_CAP0_CHANNEL_WIDTH_SET_160MHZ_IN_5G) {
PRINT_NSS_SUPP(rx_mcs_160, "RX-MCS-160");
PRINT_NSS_SUPP(tx_mcs_160, "TX-MCS-160");
}
if (cap[0] &
IEEE80211_HE_PHY_CAP0_CHANNEL_WIDTH_SET_80PLUS80_MHZ_IN_5G) {
PRINT_NSS_SUPP(rx_mcs_80p80, "RX-MCS-80P80");
PRINT_NSS_SUPP(tx_mcs_80p80, "TX-MCS-80P80");
}
#undef PRINT_NSS_SUPP
#undef PRINT
if (!(cap[6] & IEEE80211_HE_PHY_CAP6_PPE_THRESHOLD_PRESENT))
goto out;
p += scnprintf(p, buf_sz + buf - p, "PPE-THRESHOLDS: %#.2x",
hec->ppe_thres[0]);
ppe_size = ieee80211_he_ppe_size(hec->ppe_thres[0], cap);
for (i = 1; i < ppe_size; i++) {
p += scnprintf(p, buf_sz + buf - p, " %#.2x",
hec->ppe_thres[i]);
}
p += scnprintf(p, buf_sz + buf - p, "\n");
out:
ret = simple_read_from_buffer(userbuf, count, ppos, buf, p - buf);
kfree(buf);
return ret;
}
STA_OPS(he_capa);
#define DEBUGFS_ADD(name) \
debugfs_create_file(#name, 0400, \
sta->debugfs_dir, sta, &sta_ ##name## _ops)
#define DEBUGFS_ADD_COUNTER(name, field) \
debugfs_create_ulong(#name, 0400, sta->debugfs_dir, &sta->field);
void ieee80211_sta_debugfs_add(struct sta_info *sta)
{
struct ieee80211_local *local = sta->local;
struct ieee80211_sub_if_data *sdata = sta->sdata;
struct dentry *stations_dir = sta->sdata->debugfs.subdir_stations;
u8 mac[3*ETH_ALEN];
if (!stations_dir)
return;
snprintf(mac, sizeof(mac), "%pM", sta->sta.addr);
/*
* This might fail due to a race condition:
* When mac80211 unlinks a station, the debugfs entries
* remain, but it is already possible to link a new
* station with the same address which triggers adding
* it to debugfs; therefore, if the old station isn't
* destroyed quickly enough the old station's debugfs
* dir might still be around.
*/
sta->debugfs_dir = debugfs_create_dir(mac, stations_dir);
DEBUGFS_ADD(flags);
DEBUGFS_ADD(aid);
DEBUGFS_ADD(num_ps_buf_frames);
DEBUGFS_ADD(last_seq_ctrl);
DEBUGFS_ADD(agg_status);
DEBUGFS_ADD(ht_capa);
DEBUGFS_ADD(vht_capa);
DEBUGFS_ADD(he_capa);
DEBUGFS_ADD_COUNTER(rx_duplicates, rx_stats.num_duplicates);
DEBUGFS_ADD_COUNTER(rx_fragments, rx_stats.fragments);
DEBUGFS_ADD_COUNTER(tx_filtered, status_stats.filtered);
if (local->ops->wake_tx_queue) {
DEBUGFS_ADD(aqm);
mac80211: Add airtime accounting and scheduling to TXQs This adds airtime accounting and scheduling to the mac80211 TXQ scheduler. A new callback, ieee80211_sta_register_airtime(), is added that drivers can call to report airtime usage for stations. When airtime information is present, mac80211 will schedule TXQs (through ieee80211_next_txq()) in a way that enforces airtime fairness between active stations. This scheduling works the same way as the ath9k in-driver airtime fairness scheduling. If no airtime usage is reported by the driver, the scheduler will default to round-robin scheduling. For drivers that don't control TXQ scheduling in software, a new API function, ieee80211_txq_may_transmit(), is added which the driver can use to check if the TXQ is eligible for transmission, or should be throttled to enforce fairness. Calls to this function must also be enclosed in ieee80211_txq_schedule_{start,end}() calls to ensure proper locking. The API ieee80211_txq_may_transmit() also ensures that TXQ list will be aligned aginst driver's own round-robin scheduler list. i.e it rotates the TXQ list till it makes the requested node becomes the first entry in TXQ list. Thus both the TXQ list and driver's list are in sync. Co-developed-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Louie Lu <git@louie.lu> [added debugfs write op to reset airtime counter] Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk> Signed-off-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2018-12-19 09:02:08 +08:00
DEBUGFS_ADD(airtime);
}
mac80211: Add airtime accounting and scheduling to TXQs This adds airtime accounting and scheduling to the mac80211 TXQ scheduler. A new callback, ieee80211_sta_register_airtime(), is added that drivers can call to report airtime usage for stations. When airtime information is present, mac80211 will schedule TXQs (through ieee80211_next_txq()) in a way that enforces airtime fairness between active stations. This scheduling works the same way as the ath9k in-driver airtime fairness scheduling. If no airtime usage is reported by the driver, the scheduler will default to round-robin scheduling. For drivers that don't control TXQ scheduling in software, a new API function, ieee80211_txq_may_transmit(), is added which the driver can use to check if the TXQ is eligible for transmission, or should be throttled to enforce fairness. Calls to this function must also be enclosed in ieee80211_txq_schedule_{start,end}() calls to ensure proper locking. The API ieee80211_txq_may_transmit() also ensures that TXQ list will be aligned aginst driver's own round-robin scheduler list. i.e it rotates the TXQ list till it makes the requested node becomes the first entry in TXQ list. Thus both the TXQ list and driver's list are in sync. Co-developed-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Louie Lu <git@louie.lu> [added debugfs write op to reset airtime counter] Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk> Signed-off-by: Rajkumar Manoharan <rmanohar@codeaurora.org> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2018-12-19 09:02:08 +08:00
if (wiphy_ext_feature_isset(local->hw.wiphy,
NL80211_EXT_FEATURE_AQL))
DEBUGFS_ADD(aql);
debugfs_create_xul("driver_buffered_tids", 0400, sta->debugfs_dir,
&sta->driver_buffered_tids);
drv_sta_add_debugfs(local, sdata, &sta->sta, sta->debugfs_dir);
}
void ieee80211_sta_debugfs_remove(struct sta_info *sta)
{
debugfs_remove_recursive(sta->debugfs_dir);
sta->debugfs_dir = NULL;
}