Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 2 Apr 2024 09:28:50 -0400
From:      Cheng Cui <cc@freebsd.org>
To:        wireless@freebsd.org
Subject:   Weekly Development report(Mar 25 ~ Mar 29) on the iwlwifi WiFi project
Message-ID:  <CAGaXuiJSbR-tVNfNxNrYFD_wR4MSpHmK_pyec21pJTx-WEueqQ@mail.gmail.com>

next in thread | raw e-mail | index | archive | help
--0000000000005ade7006151d185b
Content-Type: text/plain; charset="UTF-8"

As mentioned earlier there is a problem with the HW encryption per packet
once `LKPI_80211_HW_CRYPTO` is defined. This problem (tracking in PR 277996
<https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=277996>) is on top of
the patch D44463 <https://reviews.freebsd.org/D44463>, which I would
welcome more feedback. Current investigation confirms that HW key setup in
firmware/driver is successful, 80211 key mapping between the FreeBSD and
the LinuxKPI is correct, and the HW key encryption stub in skbuf for the
driver is available until the driver handles it. After sync-up with Bjoern,
I found this problem only happens in the tx path, not the rx path, which
means this is only impacting the tx encryption. My next step is to
investigate the tx path with HW encryption in the Linux kernel, and figure
out how to check these parameters/flags. Most likely I am going to dump a
Linux kernel core file on the point where I am interested to see the
internal stats.

-- 
Best Regards,
Cheng Cui

--0000000000005ade7006151d185b
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">As mentioned earlier there is a problem with the HW encryp=
tion per packet once `LKPI_80211_HW_CRYPTO`<span>=C2=A0is defined. This pro=
blem (tracking in PR <a href=3D"https://bugs.freebsd.org/bugzilla/show_bug.=
cgi?id=3D277996">277996</a>) is on top of the=C2=A0</span>patch=C2=A0<a hre=
f=3D"https://reviews.freebsd.org/D44463" target=3D"_blank">D44463</a>, whic=
h I would welcome more feedback. Current investigation confirms that HW key=
 setup in firmware/driver is successful, 80211 key mapping between the Free=
BSD and the LinuxKPI is correct, and the HW key encryption stub in skbuf=C2=
=A0for the driver=C2=A0is available until the driver handles it. After sync=
-up with Bjoern, I found this problem only happens in the tx path, not the =
rx path, which means this is only impacting the tx encryption. My next step=
 is to investigate the tx path with HW encryption in the Linux kernel, and =
figure out how to check these parameters/flags. Most likely I am going to d=
ump a Linux kernel core file on the point where I am interested=C2=A0to see=
 the internal stats.<div><div><div><br></div><span class=3D"gmail_signature=
_prefix">-- </span><br><div dir=3D"ltr" class=3D"gmail_signature" data-smar=
tmail=3D"gmail_signature"><div dir=3D"ltr"><div></div>Best Regards,<div>Che=
ng Cui</div></div></div></div></div></div>

--0000000000005ade7006151d185b--



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAGaXuiJSbR-tVNfNxNrYFD_wR4MSpHmK_pyec21pJTx-WEueqQ>