From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from mail-lf0-x241.google.com ([2a00:1450:4010:c07::241]) by bombadil.infradead.org with esmtps (Exim 4.87 #1 (Red Hat Linux)) id 1dzm05-0004pH-OC for ath10k@lists.infradead.org; Wed, 04 Oct 2017 15:54:23 +0000 Received: by mail-lf0-x241.google.com with SMTP id l196so4920347lfl.3 for ; Wed, 04 Oct 2017 08:53:59 -0700 (PDT) Subject: Re: [PATCH 00/11] SDIO support for ath10k References: <1506793068-27445-1-git-send-email-alagusankar@silex-india.com> <59D47E1B.5000009@silex-india.com> From: Erik Stromdahl Message-ID: <045fef88-855e-0590-eec6-488762b3cfd6@gmail.com> Date: Wed, 4 Oct 2017 17:53:56 +0200 MIME-Version: 1.0 In-Reply-To: <59D47E1B.5000009@silex-india.com> Content-Language: en-US List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="windows-1252"; Format="flowed" Sender: "ath10k" Errors-To: ath10k-bounces+kvalo=adurom.com@lists.infradead.org To: Alagu Sankar , silexcommon@gmail.com, ath10k@lists.infradead.org Cc: linux-wireless@vger.kernel.org On 2017-10-04 08:22, Alagu Sankar wrote: > Hi Erik, > = > We will work to have this support mainlined as soon as possible. Would ap= preciate your support > in making sure that the patches do not affect the USB high latency path. > = I have added the patches in my own ath-repo and I have tested with the WUSB6100M without any problems. > On 02-10-2017 14:32, Erik Stromdahl wrote: >> Hi Alagu, >> >> It is great to see that we are finally about have fully working >> mainline support for QCA9377 SDIO chipsets! >> >> Great job! >> >> On 2017-09-30 19:37, silexcommon@gmail.com wrote: >>> From: Alagu Sankar >>> >>> This patchset, generated against master-pending branch, enables a fully >>> functional SDIO interface driver for ath10k.=A0 Patches have been verif= ied on >>> QCA9377-3 WB396 and Silex's SX-SDCAC reference cards with Station, Acce= ss Point >>> and P2P modes. >>> >>> The driver is verified with the firmware WLAN.TF.1.1.1-00061-QCATFSWPZ-1 >>> with the board data from respective SDIO card vendors. Receive performa= nce >>> matches the QCA reference driver when used with SDIO3.0 enabled platfor= ms. >>> iperf tests indicate a downlink UDP of 275Mbit/s and TCP of 150Mbit/s >>> >> Can you share any scripts etc. (wrapping hostapd and wpa_supplicant stuf= f) >> or provide some more info about you test setup? >> > I am not using any specific scripts for this. The standard ones from the = ath10k configuration page > https://wireless.wiki.kernel.org/en/users/drivers/ath10k/configuration wo= rks just fine with the > NL80211 path. >> I made a quick socat based test on an old laptop (I don't think it has S= DIO >> 3.0 support) and I did unfortunately not get the same figures as you did= :( >> > If it is SDIO v1.x, then the max bus speed is only 100Mbit/s.=A0 With v2.= x it is 200Mbit/s and 3.x it is > 832 Mbit/s.=A0 Throughput primarily depends on it. We used i.MX6 SoloX Sa= bre SDB platform > which supports SDIO3.x and could see the maximum performance. >>> This patchset differs from the previous high latency patches, specific = to SDIO. >>> HI_ACS_FLAGS_SDIO_REDUCE_TX_COMPL_SET is enabled for HI_ACS. This instr= ucts the >>> firmware to use HTT_T2H_MSG_TYPE_TX_COMPL_IND for outgoing packets. Wit= hout >>> this flag, the management frames are not sent out by the firmware. Poss= ibility >>> of management frames being sent via WMI and data frames through the red= uced Tx >>> completion needs to be probed further. >>> >> Ah, so that explains why I couldn't see any messages in the air. >> >>> Further improvements can be done on the transmit path by implementing p= acket >>> bundle. Scatter Gather is another area of improvement for both Transmit= and >>> Receive, but may not work on all platforms >>> >>> Known issues: Surprise removal of the card, when the device is in conne= cted >>> state, delays sdio function remove due to delayed WMI command failures. >>> Existing ath10k framework can not differentiate between a kernel module >>> removal and the surprise removal of teh card. >>> >>> Alagu Sankar (11): >>> =A0=A0 ath10k_sdio: sdio htt data transfer fixes >>> =A0=A0 ath10k_sdio: wb396 reference card fix >>> =A0=A0 ath10k_sdio: DMA bounce buffers for read write >>> =A0=A0 ath10k_sdio: reduce transmit msdu count >>> =A0=A0 ath10k_sdio: use clean packet headers >>> =A0=A0 ath10k_sdio: high latency fixes for beacon buffer >>> =A0=A0 ath10k_sdio: fix rssi indication >>> =A0=A0 ath10k_sdio: common read write >>> =A0=A0 ath10k_sdio: virtual scatter gather for receive >>> =A0=A0 ath10k_sdio: enable firmware crash dump >>> =A0=A0 ath10k_sdio: hif start once addition >>> >>> =A0 drivers/net/wireless/ath/ath10k/core.c=A0=A0=A0 |=A0 35 ++- >>> =A0 drivers/net/wireless/ath/ath10k/debug.c=A0=A0 |=A0=A0 3 + >>> =A0 drivers/net/wireless/ath/ath10k/htc.c=A0=A0=A0=A0 |=A0=A0 4 +- >>> =A0 drivers/net/wireless/ath/ath10k/htc.h=A0=A0=A0=A0 |=A0=A0 1 + >>> =A0 drivers/net/wireless/ath/ath10k/htt_rx.c=A0 |=A0 19 +- >>> =A0 drivers/net/wireless/ath/ath10k/htt_tx.c=A0 |=A0 24 +- >>> =A0 drivers/net/wireless/ath/ath10k/hw.c=A0=A0=A0=A0=A0 |=A0=A0 2 + >>> =A0 drivers/net/wireless/ath/ath10k/hw.h=A0=A0=A0=A0=A0 |=A0=A0 1 + >>> =A0 drivers/net/wireless/ath/ath10k/mac.c=A0=A0=A0=A0 |=A0 31 ++- >>> =A0 drivers/net/wireless/ath/ath10k/sdio.c=A0=A0=A0 | 398 +++++++++++++= +++++++++-------- >>> =A0 drivers/net/wireless/ath/ath10k/sdio.h=A0=A0=A0 |=A0 10 +- >>> =A0 drivers/net/wireless/ath/ath10k/wmi-tlv.c |=A0=A0 2 +- >>> =A0 12 files changed, 403 insertions(+), 127 deletions(-) >>> >> > = _______________________________________________ ath10k mailing list ath10k@lists.infradead.org http://lists.infradead.org/mailman/listinfo/ath10k From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from mail-lf0-f68.google.com ([209.85.215.68]:37101 "EHLO mail-lf0-f68.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750866AbdJDPyA (ORCPT ); Wed, 4 Oct 2017 11:54:00 -0400 Received: by mail-lf0-f68.google.com with SMTP id q132so4910749lfe.4 for ; Wed, 04 Oct 2017 08:53:59 -0700 (PDT) Subject: Re: [PATCH 00/11] SDIO support for ath10k To: Alagu Sankar , silexcommon@gmail.com, ath10k@lists.infradead.org Cc: linux-wireless@vger.kernel.org References: <1506793068-27445-1-git-send-email-alagusankar@silex-india.com> <59D47E1B.5000009@silex-india.com> From: Erik Stromdahl Message-ID: <045fef88-855e-0590-eec6-488762b3cfd6@gmail.com> (sfid-20171004_175405_525679_61B4ACFC) Date: Wed, 4 Oct 2017 17:53:56 +0200 MIME-Version: 1.0 In-Reply-To: <59D47E1B.5000009@silex-india.com> Content-Type: text/plain; charset=windows-1252; format=flowed Sender: linux-wireless-owner@vger.kernel.org List-ID: On 2017-10-04 08:22, Alagu Sankar wrote: > Hi Erik, > > We will work to have this support mainlined as soon as possible. Would appreciate your support > in making sure that the patches do not affect the USB high latency path. > I have added the patches in my own ath-repo and I have tested with the WUSB6100M without any problems. > On 02-10-2017 14:32, Erik Stromdahl wrote: >> Hi Alagu, >> >> It is great to see that we are finally about have fully working >> mainline support for QCA9377 SDIO chipsets! >> >> Great job! >> >> On 2017-09-30 19:37, silexcommon@gmail.com wrote: >>> From: Alagu Sankar >>> >>> This patchset, generated against master-pending branch, enables a fully >>> functional SDIO interface driver for ath10k.  Patches have been verified on >>> QCA9377-3 WB396 and Silex's SX-SDCAC reference cards with Station, Access Point >>> and P2P modes. >>> >>> The driver is verified with the firmware WLAN.TF.1.1.1-00061-QCATFSWPZ-1 >>> with the board data from respective SDIO card vendors. Receive performance >>> matches the QCA reference driver when used with SDIO3.0 enabled platforms. >>> iperf tests indicate a downlink UDP of 275Mbit/s and TCP of 150Mbit/s >>> >> Can you share any scripts etc. (wrapping hostapd and wpa_supplicant stuff) >> or provide some more info about you test setup? >> > I am not using any specific scripts for this. The standard ones from the ath10k configuration page > https://wireless.wiki.kernel.org/en/users/drivers/ath10k/configuration works just fine with the > NL80211 path. >> I made a quick socat based test on an old laptop (I don't think it has SDIO >> 3.0 support) and I did unfortunately not get the same figures as you did :( >> > If it is SDIO v1.x, then the max bus speed is only 100Mbit/s.  With v2.x it is 200Mbit/s and 3.x it is > 832 Mbit/s.  Throughput primarily depends on it. We used i.MX6 SoloX Sabre SDB platform > which supports SDIO3.x and could see the maximum performance. >>> This patchset differs from the previous high latency patches, specific to SDIO. >>> HI_ACS_FLAGS_SDIO_REDUCE_TX_COMPL_SET is enabled for HI_ACS. This instructs the >>> firmware to use HTT_T2H_MSG_TYPE_TX_COMPL_IND for outgoing packets. Without >>> this flag, the management frames are not sent out by the firmware. Possibility >>> of management frames being sent via WMI and data frames through the reduced Tx >>> completion needs to be probed further. >>> >> Ah, so that explains why I couldn't see any messages in the air. >> >>> Further improvements can be done on the transmit path by implementing packet >>> bundle. Scatter Gather is another area of improvement for both Transmit and >>> Receive, but may not work on all platforms >>> >>> Known issues: Surprise removal of the card, when the device is in connected >>> state, delays sdio function remove due to delayed WMI command failures. >>> Existing ath10k framework can not differentiate between a kernel module >>> removal and the surprise removal of teh card. >>> >>> Alagu Sankar (11): >>>    ath10k_sdio: sdio htt data transfer fixes >>>    ath10k_sdio: wb396 reference card fix >>>    ath10k_sdio: DMA bounce buffers for read write >>>    ath10k_sdio: reduce transmit msdu count >>>    ath10k_sdio: use clean packet headers >>>    ath10k_sdio: high latency fixes for beacon buffer >>>    ath10k_sdio: fix rssi indication >>>    ath10k_sdio: common read write >>>    ath10k_sdio: virtual scatter gather for receive >>>    ath10k_sdio: enable firmware crash dump >>>    ath10k_sdio: hif start once addition >>> >>>   drivers/net/wireless/ath/ath10k/core.c    |  35 ++- >>>   drivers/net/wireless/ath/ath10k/debug.c   |   3 + >>>   drivers/net/wireless/ath/ath10k/htc.c     |   4 +- >>>   drivers/net/wireless/ath/ath10k/htc.h     |   1 + >>>   drivers/net/wireless/ath/ath10k/htt_rx.c  |  19 +- >>>   drivers/net/wireless/ath/ath10k/htt_tx.c  |  24 +- >>>   drivers/net/wireless/ath/ath10k/hw.c      |   2 + >>>   drivers/net/wireless/ath/ath10k/hw.h      |   1 + >>>   drivers/net/wireless/ath/ath10k/mac.c     |  31 ++- >>>   drivers/net/wireless/ath/ath10k/sdio.c    | 398 ++++++++++++++++++++++-------- >>>   drivers/net/wireless/ath/ath10k/sdio.h    |  10 +- >>>   drivers/net/wireless/ath/ath10k/wmi-tlv.c |   2 +- >>>   12 files changed, 403 insertions(+), 127 deletions(-) >>> >> >