Date: Tue, 6 Feb 2024 09:24:11 -0500 From: Cheng Cui <cc@freebsd.org> To: wireless@freebsd.org Subject: Weekly Development report(Jan 29 ~ Feb 02) on the iwlwifi WiFi project Message-ID: <CAGaXuiLFqBD%2Br2_4k1tuL1B_9E7TwJWcgeUQL4Csp6md8NzgvA@mail.gmail.com>
index | next in thread | raw e-mail
[-- Attachment #1 --] I was still focusing on the problem in PR 276083 <https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=276083>; last week, which caused a timeout in iwlwifi firmware during 802.11n (11ng, 11na) association, thus failing 11n association. But this problem may need a second help after I have to move it to a lower priority. 1. Given the problem could be in an earlier command or a missing command, or a flag, (a needle in a haystack), etc, I am narrowing down the problem/learning scope. - In discussion, Bjoern reached out in his contact and found some feedback about the problem which may be caused by a short hcmd size. - With the help of trace-cmd and some hex raw data dump in Linux/FreeBSD, in comparison we found the raw data in the specific hcmd (grp=5, cmd=22) was intact. So the hcmd size problem is highly likely ruled-out. - Further investigation from Bjoern showed the hcmd had reached the firmware and at least processed to some extent, but the answer didn't come back. - We decided to reach out to the contact again and hope there is some help/feedback. 2. Meanwhile, I am lowering down the priority of PR 276083 <https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=276083>, and focusing on remaining tasks. After discussion, I will work on some other TX/data path after disabling the ampdu_rx* case. 3. Finished reviewing D43646, D43635, D43512. -- Best Regards, Cheng Cui [-- Attachment #2 --] <div dir="ltr"><div dir="ltr">I was still focusing on the problem in <a href="https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=276083" target="_blank">PR 276083</a> last week, which caused a timeout in iwlwifi firmware during 802.11n (11ng, 11na) association, thus failing 11n association. But this problem may need a second help after I have to move it to a lower priority.<br></div><div><ol><li>Given the problem could be in an earlier command or a missing command, or a flag, (a needle in a haystack), etc, I am narrowing down the problem/learning scope.<br></li><ul><li>In discussion, Bjoern reached out in his contact and found some feedback about the problem which may be caused by a short hcmd size.</li><li>With the help of trace-cmd and some hex raw data dump in Linux/FreeBSD, in comparison we found the raw data in the specific hcmd (grp=5, cmd=22) was intact. So the hcmd size problem is highly likely ruled-out.</li><li>Further investigation from Bjoern showed the hcmd had reached the firmware and at least processed to some extent, but the answer didn't come back.</li><li>We decided to reach out to the contact again and hope there is some help/feedback.</li></ul><li>Meanwhile, I am lowering down the priority of <a href="https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=276083" target="_blank">PR 276083</a>, and focusing on remaining tasks. After discussion, I will work on some other TX/data path after disabling the ampdu_rx* case.</li><li>Finished reviewing D43646, D43635, D43512.<br></li></ol></div><div><br></div><span class="gmail_signature_prefix">-- </span><br><div dir="ltr" class="gmail_signature" data-smartmail="gmail_signature"><div dir="ltr"><div></div>Best Regards,<div>Cheng Cui</div></div></div></div>home | help
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAGaXuiLFqBD%2Br2_4k1tuL1B_9E7TwJWcgeUQL4Csp6md8NzgvA>