| Message ID | 20210422030413.9738-2-pkshih@realtek.com |
|---|---|
| State | New |
| Headers | show |
| Series | [v2,1/2] rtw88: follow the AP basic rates for tx mgmt frame | expand |
On Thu, Apr 22, 2021 at 11:04:13AM +0800, Ping-Ke Shih wrote: > From: Yu-Yen Ting <steventing@realtek.com> > > The management frame with high rate e.g. 24M may not be transmitted > smoothly in long range environment. > Add a debugfs to force to use the lowest basic rate > in order to debug the reachability of transmitting management frame. > > obtain current setting > cat /sys/kernel/debug/ieee80211/phyX/rtw88/basic_rates > > force lowest rate: > echo 1 > /sys/kernel/debug/ieee80211/phyX/rtw88/basic_rates > > Signed-off-by: Yu-Yen Ting <steventing@realtek.com> > Signed-off-by: Ping-Ke Shih <pkshih@realtek.com> I believe some initial objection to this was because it was unclear if this is for "production" use (e.g., recommending distros to play with this) or for debugging. I'll admit, I requested the feature for patch 1, because I've seen that for those networks where people *do* configure odd Basic Rates, they intend for stations to follow those, and not use the lowest (and most airtime-hogging) rates. And I can say, I don't see why distributions should be turning that back off. If the Basic Rates setting is wrong, then the that's up to the network admin to fix. All that is to say: I agree that this patch is purely for debugging, as stated, and that it belongs in debugfs. I also maintain a distribution, and I don't plan on using this beyond debugging. Therefore: Reviewed-by: Brian Norris <briannorris@chromium.org> BTW, if we have clear guidelines on debugfs, module parameters, etc., maybe those should be going on the wiki? I know this came up before: https://lore.kernel.org/linux-wireless/87d09u7tyr.fsf@codeaurora.org/ At this point, I'm willing to write such guidelines, if I get an ack from the relevant folks (I guess that's just Kalle?). It probably belongs somewhere in this tree: https://wireless.wiki.kernel.org/en/developers/documentation similar to this: https://wireless.wiki.kernel.org/en/developers/documentation/nl80211#vendor-specific_api except it's not really an nl80211 thing. Suggestions welcome. Side note: it could really use some cleanup -- like this page: https://wireless.wiki.kernel.org/en/developers/process Brian
On 10/11/2021 11:23 AM, Kalle Valo wrote: > (changing subject, was "Re: [PATCH v2 2/2] rtw88: add debugfs to force lowest basic rate") > > Brian Norris <briannorris@chromium.org> writes: > >> BTW, if we have clear guidelines on debugfs, module parameters, etc., >> maybe those should be going on the wiki? I know this came up before: >> >> https://lore.kernel.org/linux-wireless/87d09u7tyr.fsf@codeaurora.org/ >> >> At this point, I'm willing to write such guidelines, if I get an ack >> from the relevant folks (I guess that's just Kalle?). It probably >> belongs somewhere in this tree: >> >> https://wireless.wiki.kernel.org/en/developers/documentation >> >> similar to this: >> https://wireless.wiki.kernel.org/en/developers/documentation/nl80211#vendor-specific_api >> except it's not really an nl80211 thing. Suggestions welcome. > > I think this is a very good idea. Having general guidelines for wireless > drivers using user space interfaces would help both people submitting > patches and also people like me reviewing the patches. > > We should try to get an ack for the guidelines at least from Johannes, > but I would prefer also involve Jakub and Dave (CCed) as they might have > some input from the network subsystem point of view. > > Just to get this started, here's a draft list I came up of different > user space interfaces upstream wireless drivers are using: > > * generic nl80211 (excluding testmode and vendor commands) > > * nl80211 testmode commands > > * nl80211 vendor commands > > * sysfs[1] > > * debugfs > > * relayfs > > * configfs[1] > > * module parameters > > * thermal subsystem > > * firmware_request() > > I'm not saying that we need to document all these in the first version, > I'm just trying to come up a comprehensive overview how wireless drivers > interact with the user space. And I'm sure I missed something. so please > do fill in. Not sure if all of the above can be considered user-space interfaces, but wireless driver developers could benefit from guidelines for them regardless. Maybe following needs to be considered as well although I think cfg80211 is taking care of it: * rfkill Regards, Arend >> Side note: it could really use some cleanup -- like this page: >> https://wireless.wiki.kernel.org/en/developers/process > > Heh, that is old information. TBH in practise I maintain only the > submittingpatches page (link in the signature), other pages I rarely > touch. And naturally I also look after ath10k and ath11k pages. > > Any volunteers to clean that up? > > [1] Actually I don't know if there are any valid use cases for sysfs and > configfs at the moment, but I'll include them in the list for > completeness. > -- This electronic communication and the information and any files transmitted with it, or attached to it, are confidential and are intended solely for the use of the individual or entity to whom it is addressed and may contain information that is confidential, legally privileged, protected by privacy laws, or otherwise restricted from disclosure to anyone else. If you are not the intended recipient or the person responsible for delivering the e-mail to the intended recipient, you are hereby notified that any use, copying, distributing, dissemination, forwarding, printing, or copying of this e-mail is strictly prohibited. If you received this e-mail in error, please return the e-mail to the sender, delete it from your computer, and destroy any printed copy of it.
Arend van Spriel <arend.vanspriel@broadcom.com> writes: > On 10/11/2021 11:23 AM, Kalle Valo wrote: >> (changing subject, was "Re: [PATCH v2 2/2] rtw88: add debugfs to >> force lowest basic rate") >> >> Brian Norris <briannorris@chromium.org> writes: >> >>> BTW, if we have clear guidelines on debugfs, module parameters, etc., >>> maybe those should be going on the wiki? I know this came up before: >>> >>> https://lore.kernel.org/linux-wireless/87d09u7tyr.fsf@codeaurora.org/ >>> >>> At this point, I'm willing to write such guidelines, if I get an ack >>> from the relevant folks (I guess that's just Kalle?). It probably >>> belongs somewhere in this tree: >>> >>> https://wireless.wiki.kernel.org/en/developers/documentation >>> >>> similar to this: >>> https://wireless.wiki.kernel.org/en/developers/documentation/nl80211#vendor-specific_api >>> except it's not really an nl80211 thing. Suggestions welcome. >> >> I think this is a very good idea. Having general guidelines for wireless >> drivers using user space interfaces would help both people submitting >> patches and also people like me reviewing the patches. >> >> We should try to get an ack for the guidelines at least from Johannes, >> but I would prefer also involve Jakub and Dave (CCed) as they might have >> some input from the network subsystem point of view. >> >> Just to get this started, here's a draft list I came up of different >> user space interfaces upstream wireless drivers are using: >> >> * generic nl80211 (excluding testmode and vendor commands) >> >> * nl80211 testmode commands >> >> * nl80211 vendor commands >> >> * sysfs[1] >> >> * debugfs >> >> * relayfs >> >> * configfs[1] >> >> * module parameters >> >> * thermal subsystem >> >> * firmware_request() >> >> I'm not saying that we need to document all these in the first version, >> I'm just trying to come up a comprehensive overview how wireless drivers >> interact with the user space. And I'm sure I missed something. so please >> do fill in. > > Not sure if all of the above can be considered user-space interfaces, > but wireless driver developers could benefit from guidelines for them > regardless. Maybe the name should be "Guidelines for wireless drivers" without being too specific? > Maybe following needs to be considered as well although I think > cfg80211 is taking care of it: > > * rfkill Yeah, it would be good to include rfkill. -- https://patchwork.kernel.org/project/linux-wireless/list/ https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches
diff --git a/drivers/net/wireless/realtek/rtw88/debug.c b/drivers/net/wireless/realtek/rtw88/debug.c index 18ab472ea46c..a99a3dc6f9f9 100644 --- a/drivers/net/wireless/realtek/rtw88/debug.c +++ b/drivers/net/wireless/realtek/rtw88/debug.c @@ -861,6 +861,39 @@ static int rtw_debugfs_get_fw_crash(struct seq_file *m, void *v) return 0; } +static ssize_t rtw_debugfs_set_basic_rates(struct file *filp, + const char __user *buffer, + size_t count, loff_t *loff) +{ + struct seq_file *seqpriv = (struct seq_file *)filp->private_data; + struct rtw_debugfs_priv *debugfs_priv = seqpriv->private; + struct rtw_dev *rtwdev = debugfs_priv->rtwdev; + bool input; + int err; + + err = kstrtobool_from_user(buffer, count, &input); + if (err) + return err; + + if (input) + set_bit(RTW_FLAG_USE_LOWEST_RATE, rtwdev->flags); + else + clear_bit(RTW_FLAG_USE_LOWEST_RATE, rtwdev->flags); + + return count; +} + +static int rtw_debugfs_get_basic_rates(struct seq_file *m, void *v) +{ + struct rtw_debugfs_priv *debugfs_priv = m->private; + struct rtw_dev *rtwdev = debugfs_priv->rtwdev; + + seq_printf(m, "use lowest: %d\n", + test_bit(RTW_FLAG_USE_LOWEST_RATE, rtwdev->flags)); + + return 0; +} + static ssize_t rtw_debugfs_set_dm_cap(struct file *filp, const char __user *buffer, size_t count, loff_t *loff) @@ -1046,6 +1079,11 @@ static struct rtw_debugfs_priv rtw_debug_priv_fw_crash = { .cb_read = rtw_debugfs_get_fw_crash, }; +static struct rtw_debugfs_priv rtw_debug_priv_basic_rates = { + .cb_write = rtw_debugfs_set_basic_rates, + .cb_read = rtw_debugfs_get_basic_rates, +}; + static struct rtw_debugfs_priv rtw_debug_priv_dm_cap = { .cb_write = rtw_debugfs_set_dm_cap, .cb_read = rtw_debugfs_get_dm_cap, @@ -1125,6 +1163,7 @@ void rtw_debugfs_init(struct rtw_dev *rtwdev) rtw_debugfs_add_r(rf_dump); rtw_debugfs_add_r(tx_pwr_tbl); rtw_debugfs_add_rw(fw_crash); + rtw_debugfs_add_rw(basic_rates); rtw_debugfs_add_rw(dm_cap); } diff --git a/drivers/net/wireless/realtek/rtw88/main.h b/drivers/net/wireless/realtek/rtw88/main.h index dc3744847ba9..0df5df3a62ab 100644 --- a/drivers/net/wireless/realtek/rtw88/main.h +++ b/drivers/net/wireless/realtek/rtw88/main.h @@ -362,6 +362,7 @@ enum rtw_flags { RTW_FLAG_BUSY_TRAFFIC, RTW_FLAG_WOWLAN, RTW_FLAG_RESTARTING, + RTW_FLAG_USE_LOWEST_RATE, NUM_OF_RTW_FLAGS, }; diff --git a/drivers/net/wireless/realtek/rtw88/tx.c b/drivers/net/wireless/realtek/rtw88/tx.c index e5949a775283..0aeed15736c8 100644 --- a/drivers/net/wireless/realtek/rtw88/tx.c +++ b/drivers/net/wireless/realtek/rtw88/tx.c @@ -238,8 +238,9 @@ static u8 rtw_get_mgmt_rate(struct rtw_dev *rtwdev, struct sk_buff *skb, { struct ieee80211_tx_info *tx_info = IEEE80211_SKB_CB(skb); struct ieee80211_vif *vif = tx_info->control.vif; + bool use_lowest = test_bit(RTW_FLAG_USE_LOWEST_RATE, rtwdev->flags); - if (!vif || !vif->bss_conf.basic_rates || ignore_rate) + if (!vif || !vif->bss_conf.basic_rates || ignore_rate || use_lowest) return lowest_rate; return __ffs(vif->bss_conf.basic_rates) + lowest_rate;