From: Sungbo Eo <mans0n@gorani.run>
To: Seth Forshee <sforshee@kernel.org>
Cc: wireless-regdb@lists.infradead.org, linux-wireless@vger.kernel.org
Subject: Re: [PATCH v2] wireless-regdb: Update regulatory rules for South Korea (KR)
Date: Sun, 30 Jan 2022 21:24:39 +0900 [thread overview]
Message-ID: <77045141-4a4a-4827-8577-c1f78060aff8@gorani.run> (raw)
In-Reply-To: <YeCgsJPwqFQiSmpa@ubuntu-x1>
On 2022-01-14 06:59, Seth Forshee wrote:
> On Mon, Dec 20, 2021 at 03:10:12PM -0600, Seth Forshee wrote:
>> On Sat, Dec 18, 2021 at 04:31:17PM +0900, Sungbo Eo wrote:
>>> On 2021-12-18 07:02, Seth Forshee wrote:
>>>> On Sat, Oct 30, 2021 at 01:12:08AM +0900, Sungbo Eo wrote:
>>>>> On 2021-10-28 03:06, Seth Forshee wrote:
>>>>>> On Sun, Oct 24, 2021 at 08:38:21PM +0900, Sungbo Eo wrote:
>>>>>>> This patch is based on MSIT Public Notification 2020-113 ("Unlicensed Radio
>>>>>>> Equipment Established Without Notice"), officially announced on 2021-01-06.
>>>>>>>
>>>>>>> The PSD must not exceed 2.5 mW/MHz if the frequency range includes all or
>>>>>>> part of 5230-5250 MHz and the bandwidth is equal to or less than 40 MHz.
>>>>>>> This leads to the following:
>>>>>>> * 5230-5250 @ 20 -> 17 dBm
>>>>>>> * 5210-5250 @ 40 -> 20 dBm
>>>>>>> Here the power limits for 80/160 MHz bandwidth are also lowered to 17 dBm,
>>>>>>> as it's not possible to set different power limits for different bandwidths
>>>>>>> at the moment.
>>>>>>>
>>>>>>> Extend the last 5 GHz frequency range to 5850 MHz.
>>>>>>>
>>>>>>> WiFi 6E is now allowed with the following restrictions:
>>>>>>> * Indoor: the full 1.2 GHz range, up to 160 MHz bandwidth and 250mW EIRP
>>>>>>> * Outdoor: the lower 500 MHz range, up to 160 MHz bandwidth and 25mW EIRP
>>>>>>> Here only the former entry is added.
>>>>>>>
>>>>>>> And also update the regulatory source links.
>>>>>>>
>>>>>>> Signed-off-by: Sungbo Eo <mans0n@gorani.run>
>>>>>>> ---
>>>>>>> v2:
>>>>>>> * split 5150-5250 MHz band rule to accommodate the PSD limit
>>>>>>> * remove AUTO-BW flag from 6 GHz band rule
>>>>>>> ---
>>>>>>> db.txt | 17 ++++++++++++-----
>>>>>>> 1 file changed, 12 insertions(+), 5 deletions(-)
>>>>>>>
>>>>>>> diff --git a/db.txt b/db.txt
>>>>>>> index 6e8dbef..387ac93 100644
>>>>>>> --- a/db.txt
>>>>>>> +++ b/db.txt
>>>>>>> @@ -862,15 +862,22 @@ country KP: DFS-JP
>>>>>>> (5490 - 5630 @ 20), (30), DFS
>>>>>>> (5735 - 5815 @ 20), (30)
>>>>>>> +# Source:
>>>>>>> +# https://www.law.go.kr/LSW//admRulLsInfoP.do?chrClsCd=&admRulSeq=2100000196972
>>>>>>> +# https://www.law.go.kr/LSW//admRulLsInfoP.do?chrClsCd=&admRulSeq=2100000196973
>>>>>>> +# https://www.law.go.kr/LSW//admRulLsInfoP.do?chrClsCd=&admRulSeq=2100000196974
>>>>>>> country KR: DFS-JP
>>>>>>> - # ref: https://www.rra.go.kr
>>>>>>> (2400 - 2483.5 @ 40), (23)
>>>>>>> - (5150 - 5250 @ 80), (23), AUTO-BW
>>>>>>> + (5150 - 5210 @ 40), (23), AUTO-BW
>>>>>>> + # max. PSD 2.5 mW/MHz in 5230-5250 MHz frequency range
>>>>>>> + (5210 - 5230 @ 20), (20), AUTO-BW
>>>>>>> + (5230 - 5250 @ 20), (17), AUTO-BW
>>>>>>
>>>>>> Even with 5210-5230 split out like this, 5210-5250 @ 40 still gets
>>>>>> limited to 17 dBm by the 5230-5250 rule. So why do we need to split out
>>>>>> 5210-5230 separate from 5150-5210?
>>>>>
>>>>> I did some tests with regdb and I found out that the rule to be applied
>>>>> depends on the main 20 MHz channel. With the rules above, if I select
>>>>> channel 48 (5230-5250) and enable VHT40, then the TX power is set to 17 dBm.
>>>>> OTOH if I select channel 44 (5210–5230) and enable VHT40, then the TX power
>>>>> is set to 20 dBm. I'm not sure whether my result is really correct, though.
>>>>
>>>> I'm sorry, when I saw the v3 patch I realized that I forgot to reply to
>>>> this email. I had wanted to take a look at the kernel source to try and
>>>> understand the behavior you described.
>>>>
>>>> I took a quick look, and from what I'm seeing the minimum of the power
>>>> limits for the two ranges should be used. This is the behavior I expect.
>>>> Can you provide more detail about how you're getting that result?
>>>>
>>>> Thanks,
>>>> Seth
>>>
>>> I've got the result on MT7623+MT7615N AP with OpenWrt installed. (kernel
>>> 5.10)
>>> I set the channel and bandwidth in /etc/config/wireless (in OpenWrt-standard
>>> way) and checked the tx power with `iw dev wlan0 info` and Android WiFi
>>> Analyzer app.
>>>
>>> # channel 48, htmode VHT40
>>> type AP
>>> channel 48 (5240 MHz), width: 40 MHz, center1: 5230 MHz
>>> txpower 17.00 dBm
>>>
>>> # channel 44, htmode VHT40
>>> type AP
>>> channel 44 (5220 MHz), width: 40 MHz, center1: 5230 MHz
>>> txpower 20.00 dBm
>>>
>>> The WiFi Analyzer app also reported higher RSSI for the channel 44 case.
>>>
>>> I also had a quick look at net/wireless/reg.c and it seems you're right.
>>> Perhaps my tests were bad, I do believe your comments more than my result.
>>> ;)
>>
>> That definitely seems odd. I'll try to have another look and see if I
>> can figure out why that's happening. I'd have suspected that what's
>> displayed by iw might just be a result of how the information is
>> reported or displayed, but the higher RSSI argues against that.
>
> I've spent a little more time looking at this. I think the issue is that
> the regulatory rules are only used to establish channel definititions
> for 20 MHz channels, and these channel definitions are what get used
> when establishing tx parameters. While the flags are checked against the
> chandefs of secondary channels, only the primary channel's power limits
> are used for tx power. So I think the behavior of the kernel is incorect
> here.
Thanks for your confirm. I also did my own research and came to the same
conclusion.
>
> However, I think the regulatory database should reflect the regulatory
> data as accurately as possible, so I'd prefer we not split out 5210-5230
> and try to get the bug fixed in the kernel.
Okay, sounds reasonable. I'll send v4 soon.
Regards,
Sungbo
next prev parent reply other threads:[~2022-01-30 12:25 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-09-29 17:27 [PATCH] wireless-regdb: Update regulatory rules for South Korea (KR) Sungbo Eo
2021-10-06 14:07 ` Seth Forshee
2021-10-06 19:18 ` Sungbo Eo
2021-10-06 21:16 ` [wireless-regdb] " Seth Forshee
2021-10-24 11:38 ` [PATCH v2] " Sungbo Eo
2021-10-27 18:06 ` Seth Forshee
2021-10-29 16:12 ` Sungbo Eo
2021-12-17 22:02 ` Seth Forshee
2021-12-18 7:31 ` Sungbo Eo
2021-12-20 21:10 ` Seth Forshee
2022-01-13 21:59 ` Seth Forshee
2022-01-30 12:24 ` Sungbo Eo [this message]
2021-12-12 15:20 ` [PATCH v3] " Sungbo Eo
2022-01-30 12:49 ` [PATCH v4] " Sungbo Eo
2022-02-12 16:26 ` Seth Forshee
2022-02-12 17:04 ` Sungbo Eo
2022-02-13 22:24 ` Seth Forshee
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=77045141-4a4a-4827-8577-c1f78060aff8@gorani.run \
--to=mans0n@gorani.run \
--cc=linux-wireless@vger.kernel.org \
--cc=sforshee@kernel.org \
--cc=wireless-regdb@lists.infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).