* [ath9k-devel] Ack and Block Ack processing @ 2014-02-03 15:48 Olivier Marce 2014-02-03 16:29 ` Adrian Chadd 2014-02-03 17:01 ` abhinav narain 0 siblings, 2 replies; 7+ messages in thread From: Olivier Marce @ 2014-02-03 15:48 UTC (permalink / raw) To: ath9k-devel Hi everybody, thanks to all of you who share your knowledge and expertise on ath9k. I wonder if Ack and Block Ack are processed directly by the HW on receipt and not the firmware. There are several messages that confirm that Ack and Block Ack are *sent* by HW (http://www.spinics.net/lists/linux-wireless/msg106950.html) But are they processed (on receipt) also by HW ? I analysed firmware code and I did not find any Ack/BA processing. Is my understanding correct ? Best regards -- Olivier Marc? Alcatel-Lucent Bell Labs France ^ permalink raw reply [flat|nested] 7+ messages in thread
* [ath9k-devel] Ack and Block Ack processing 2014-02-03 15:48 [ath9k-devel] Ack and Block Ack processing Olivier Marce @ 2014-02-03 16:29 ` Adrian Chadd 2014-02-10 8:57 ` Olivier Marce 2014-02-03 17:01 ` abhinav narain 1 sibling, 1 reply; 7+ messages in thread From: Adrian Chadd @ 2014-02-03 16:29 UTC (permalink / raw) To: ath9k-devel Hi, So on the Atheros NICs, the block-acks are generated by the receiver in hardware and processed by the transmitter in hardware; you then get a bitmap in the TX completion descriptor telling you which frames did succeed/fail. The TX completion path does check the two 32 bit blockack (ba) fields in the TX descriptor for completion. It doesn't hand-parse a BA frame. -a On 3 February 2014 07:48, Olivier Marce <Olivier.Marce@alcatel-lucent.com> wrote: > Hi everybody, > thanks to all of you who share your knowledge and expertise on ath9k. > > I wonder if Ack and Block Ack are processed directly by the HW on > receipt and not the firmware. > > There are several messages that confirm that Ack and Block Ack are > *sent* by HW (http://www.spinics.net/lists/linux-wireless/msg106950.html) > But are they processed (on receipt) also by HW ? I analysed firmware > code and I did not find any Ack/BA processing. > > Is my understanding correct ? > > Best regards > > > -- > Olivier Marc? > Alcatel-Lucent Bell Labs France > _______________________________________________ > ath9k-devel mailing list > ath9k-devel at lists.ath9k.org > https://lists.ath9k.org/mailman/listinfo/ath9k-devel ^ permalink raw reply [flat|nested] 7+ messages in thread
* [ath9k-devel] Ack and Block Ack processing 2014-02-03 16:29 ` Adrian Chadd @ 2014-02-10 8:57 ` Olivier Marce 2014-02-10 9:45 ` Kamran Nishat 0 siblings, 1 reply; 7+ messages in thread From: Olivier Marce @ 2014-02-10 8:57 UTC (permalink / raw) To: ath9k-devel Thanks for the quick and precise answer. Then, I wonder is there is a way to be notified of sending/receiving of frames by the HW. For example, can we be notified of ACK (or BA) received and processed by the HW, or by frame retransmited by the HW ? Thanks On 03/02/2014 17:29, Adrian Chadd wrote: > Hi, > > So on the Atheros NICs, the block-acks are generated by the receiver > in hardware and processed by the transmitter in hardware; you then get > a bitmap in the TX completion descriptor telling you which frames did > succeed/fail. > > The TX completion path does check the two 32 bit blockack (ba) fields > in the TX descriptor for completion. It doesn't hand-parse a BA frame. > > > -a > > > On 3 February 2014 07:48, Olivier Marce > <Olivier.Marce@alcatel-lucent.com> wrote: >> Hi everybody, >> thanks to all of you who share your knowledge and expertise on ath9k. >> >> I wonder if Ack and Block Ack are processed directly by the HW on >> receipt and not the firmware. >> >> There are several messages that confirm that Ack and Block Ack are >> *sent* by HW (http://www.spinics.net/lists/linux-wireless/msg106950.html) >> But are they processed (on receipt) also by HW ? I analysed firmware >> code and I did not find any Ack/BA processing. >> >> Is my understanding correct ? >> >> Best regards >> >> >> -- >> Olivier Marc? >> Alcatel-Lucent Bell Labs France >> _______________________________________________ >> ath9k-devel mailing list >> ath9k-devel at lists.ath9k.org >> https://lists.ath9k.org/mailman/listinfo/ath9k-devel > > -- Olivier Marc? Alcatel-Lucent Bell Labs France ^ permalink raw reply [flat|nested] 7+ messages in thread
* [ath9k-devel] Ack and Block Ack processing 2014-02-10 8:57 ` Olivier Marce @ 2014-02-10 9:45 ` Kamran Nishat 2014-02-10 10:11 ` Olivier Marce 0 siblings, 1 reply; 7+ messages in thread From: Kamran Nishat @ 2014-02-10 9:45 UTC (permalink / raw) To: ath9k-devel as Adrian said we got information received in a BlockACK (ath_tx_complete_aggr in xmit.c) you get a bitmap, SSN and retires done for AMPDU. On Mon, Feb 10, 2014 at 1:57 PM, Olivier Marce < Olivier.Marce@alcatel-lucent.com> wrote: > Thanks for the quick and precise answer. > > Then, I wonder is there is a way to be notified of sending/receiving of > frames by the HW. For example, can we be notified of ACK (or BA) > received and processed by the HW, or by frame retransmited by the HW ? > > Thanks > > On 03/02/2014 17:29, Adrian Chadd wrote: > > Hi, > > > > So on the Atheros NICs, the block-acks are generated by the receiver > > in hardware and processed by the transmitter in hardware; you then get > > a bitmap in the TX completion descriptor telling you which frames did > > succeed/fail. > > > > The TX completion path does check the two 32 bit blockack (ba) fields > > in the TX descriptor for completion. It doesn't hand-parse a BA frame. > > > > > > -a > > > > > > On 3 February 2014 07:48, Olivier Marce > > <Olivier.Marce@alcatel-lucent.com> wrote: > >> Hi everybody, > >> thanks to all of you who share your knowledge and expertise on ath9k. > >> > >> I wonder if Ack and Block Ack are processed directly by the HW on > >> receipt and not the firmware. > >> > >> There are several messages that confirm that Ack and Block Ack are > >> *sent* by HW ( > http://www.spinics.net/lists/linux-wireless/msg106950.html) > >> But are they processed (on receipt) also by HW ? I analysed firmware > >> code and I did not find any Ack/BA processing. > >> > >> Is my understanding correct ? > >> > >> Best regards > >> > >> > >> -- > >> Olivier Marc? > >> Alcatel-Lucent Bell Labs France > >> _______________________________________________ > >> ath9k-devel mailing list > >> ath9k-devel at lists.ath9k.org > >> https://lists.ath9k.org/mailman/listinfo/ath9k-devel > > > > > > -- > Olivier Marc? > Alcatel-Lucent Bell Labs France > _______________________________________________ > ath9k-devel mailing list > ath9k-devel at lists.ath9k.org > https://lists.ath9k.org/mailman/listinfo/ath9k-devel > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20140210/08182b15/attachment.htm ^ permalink raw reply [flat|nested] 7+ messages in thread
* [ath9k-devel] Ack and Block Ack processing 2014-02-10 9:45 ` Kamran Nishat @ 2014-02-10 10:11 ` Olivier Marce 2014-02-10 18:02 ` abhinav narain 0 siblings, 1 reply; 7+ messages in thread From: Olivier Marce @ 2014-02-10 10:11 UTC (permalink / raw) To: ath9k-devel Hi Kamram, thanks, this is a useful information indeed. But my question was about to get notified of the details of exchange between emitter and transmitter. E.g. how many retries have been needed to get the a MPDU to be acked. On 10/02/2014 10:45, Kamran Nishat wrote: > as Adrian said we got information received in a BlockACK > (ath_tx_complete_aggr in xmit.c) > you get a bitmap, SSN and retires done for AMPDU. > > > > On Mon, Feb 10, 2014 at 1:57 PM, Olivier Marce > <Olivier.Marce@alcatel-lucent.com > <mailto:Olivier.Marce@alcatel-lucent.com>> wrote: > > Thanks for the quick and precise answer. > > Then, I wonder is there is a way to be notified of sending/receiving of > frames by the HW. For example, can we be notified of ACK (or BA) > received and processed by the HW, or by frame retransmited by the HW ? > > Thanks > > On 03/02/2014 17:29, Adrian Chadd wrote: > > Hi, > > > > So on the Atheros NICs, the block-acks are generated by the receiver > > in hardware and processed by the transmitter in hardware; you > then get > > a bitmap in the TX completion descriptor telling you which frames did > > succeed/fail. > > > > The TX completion path does check the two 32 bit blockack (ba) fields > > in the TX descriptor for completion. It doesn't hand-parse a BA > frame. > > > > > > -a > > > > > > On 3 February 2014 07:48, Olivier Marce > > <Olivier.Marce@alcatel-lucent.com > <mailto:Olivier.Marce@alcatel-lucent.com>> wrote: > >> Hi everybody, > >> thanks to all of you who share your knowledge and expertise on > ath9k. > >> > >> I wonder if Ack and Block Ack are processed directly by the HW on > >> receipt and not the firmware. > >> > >> There are several messages that confirm that Ack and Block Ack are > >> *sent* by HW > (http://www.spinics.net/lists/linux-wireless/msg106950.html) > >> But are they processed (on receipt) also by HW ? I analysed firmware > >> code and I did not find any Ack/BA processing. > >> > >> Is my understanding correct ? > >> > >> Best regards > >> > >> > >> -- > >> Olivier Marc? > >> Alcatel-Lucent Bell Labs France > >> _______________________________________________ > >> ath9k-devel mailing list > >> ath9k-devel at lists.ath9k.org <mailto:ath9k-devel@lists.ath9k.org> > >> https://lists.ath9k.org/mailman/listinfo/ath9k-devel > > > > > > -- > Olivier Marc? > Alcatel-Lucent Bell Labs France > _______________________________________________ > ath9k-devel mailing list > ath9k-devel at lists.ath9k.org <mailto:ath9k-devel@lists.ath9k.org> > https://lists.ath9k.org/mailman/listinfo/ath9k-devel > > -- Olivier Marc? Alcatel-Lucent Bell Labs France ^ permalink raw reply [flat|nested] 7+ messages in thread
* [ath9k-devel] Ack and Block Ack processing 2014-02-10 10:11 ` Olivier Marce @ 2014-02-10 18:02 ` abhinav narain 0 siblings, 0 replies; 7+ messages in thread From: abhinav narain @ 2014-02-10 18:02 UTC (permalink / raw) To: ath9k-devel Hi Kamran, > But my question was about to get notified of the details of exchange > between emitter and transmitter. E.g. how many retries have been needed > to get the a MPDU to be acked. > > Read xmit.c about the retry counts for each MPDU. I am not clear what happens for AMPDU, but essentially you give a buffer (chain of mpdus or one ampdu ) to the hardware and once it's done with it, it gives a callback with details in the transmit descriptor. You can extract information from there struct ath_tx_status *ts is the descriptor, which fills details into the mac80211 struct ieee80211_tx_info * - Abhinav > > On 10/02/2014 10:45, Kamran Nishat wrote: > > as Adrian said we got information received in a BlockACK > > (ath_tx_complete_aggr in xmit.c) > > you get a bitmap, SSN and retires done for AMPDU. > > > > > > > > On Mon, Feb 10, 2014 at 1:57 PM, Olivier Marce > > <Olivier.Marce@alcatel-lucent.com > > <mailto:Olivier.Marce@alcatel-lucent.com>> wrote: > > > > Thanks for the quick and precise answer. > > > > Then, I wonder is there is a way to be notified of sending/receiving > of > > frames by the HW. For example, can we be notified of ACK (or BA) > > received and processed by the HW, or by frame retransmited by the HW > ? > > > > Thanks > > > > On 03/02/2014 17:29, Adrian Chadd wrote: > > > Hi, > > > > > > So on the Atheros NICs, the block-acks are generated by the > receiver > > > in hardware and processed by the transmitter in hardware; you > > then get > > > a bitmap in the TX completion descriptor telling you which frames > did > > > succeed/fail. > > > > > > The TX completion path does check the two 32 bit blockack (ba) > fields > > > in the TX descriptor for completion. It doesn't hand-parse a BA > > frame. > > > > > > > > > -a > > > > > > > > > On 3 February 2014 07:48, Olivier Marce > > > <Olivier.Marce@alcatel-lucent.com > > <mailto:Olivier.Marce@alcatel-lucent.com>> wrote: > > >> Hi everybody, > > >> thanks to all of you who share your knowledge and expertise on > > ath9k. > > >> > > >> I wonder if Ack and Block Ack are processed directly by the HW on > > >> receipt and not the firmware. > > >> > > >> There are several messages that confirm that Ack and Block Ack > are > > >> *sent* by HW > > (http://www.spinics.net/lists/linux-wireless/msg106950.html) > > >> But are they processed (on receipt) also by HW ? I analysed > firmware > > >> code and I did not find any Ack/BA processing. > > >> > > >> Is my understanding correct ? > > >> > > >> Best regards > > >> > > >> > > >> -- > > >> Olivier Marc? > > >> Alcatel-Lucent Bell Labs France > > >> _______________________________________________ > > >> ath9k-devel mailing list > > >> ath9k-devel at lists.ath9k.org <mailto:ath9k-devel@lists.ath9k.org> > > >> https://lists.ath9k.org/mailman/listinfo/ath9k-devel > > > > > > > > > > -- > > Olivier Marc? > > Alcatel-Lucent Bell Labs France > > _______________________________________________ > > ath9k-devel mailing list > > ath9k-devel at lists.ath9k.org <mailto:ath9k-devel@lists.ath9k.org> > > https://lists.ath9k.org/mailman/listinfo/ath9k-devel > > > > > > -- > Olivier Marc? > Alcatel-Lucent Bell Labs France > _______________________________________________ > ath9k-devel mailing list > ath9k-devel at lists.ath9k.org > https://lists.ath9k.org/mailman/listinfo/ath9k-devel > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20140210/10b30251/attachment.htm ^ permalink raw reply [flat|nested] 7+ messages in thread
* [ath9k-devel] Ack and Block Ack processing 2014-02-03 15:48 [ath9k-devel] Ack and Block Ack processing Olivier Marce 2014-02-03 16:29 ` Adrian Chadd @ 2014-02-03 17:01 ` abhinav narain 1 sibling, 0 replies; 7+ messages in thread From: abhinav narain @ 2014-02-03 17:01 UTC (permalink / raw) To: ath9k-devel Hi, > I wonder if Ack and Block Ack are processed directly by the HW on > receipt and not the firmware. > Please correct me if I am wrong, ath9k series drivers directly interact with hardware. I haven't seen any code where the driver talks to firmware. Thanks, Abhinav -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20140203/f1dc7962/attachment.htm ^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2014-02-10 18:02 UTC | newest] Thread overview: 7+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2014-02-03 15:48 [ath9k-devel] Ack and Block Ack processing Olivier Marce 2014-02-03 16:29 ` Adrian Chadd 2014-02-10 8:57 ` Olivier Marce 2014-02-10 9:45 ` Kamran Nishat 2014-02-10 10:11 ` Olivier Marce 2014-02-10 18:02 ` abhinav narain 2014-02-03 17:01 ` abhinav narain
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.