Re: [OpenWrt-Devel] [PATCH 0/2] mac80211: Update QCA9880 firmware (ath10k)
On Thu, 25 Jun 2015 20:59:19 +0200, Matti Laakso malaakso at elisanet.fi https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel wrote: / // On 22/06/15 20:19, Matti Laakso wrote: // / These patches backport the new FW API v5 to ath10k // // (upstream commits 8348db2, 53513c3, and dc3632a) // // and change mac80211 to use the new firmware. // // // // // /Could you also update // target/linux/ar71xx/base-files/lib/preinit/82_patch_ath10k? // See https://patchwork.ozlabs.org/patch/466945/ and possibly // https://patchwork.ozlabs.org/patch/481472/ // // Thanks, // Stijn // // Hi, // // That seems like a burden to maintain. Do I understand correctly that the // device has valid data in the PCIe card EEPROM (so it is not necessary to // create a file with the calibration data in 81_load_ath10k_board_bin), // but the MAC is incorrect? Is the MAC completely bogus (like same for all // devices), or valid but just not what it says on the router cover? If the // latter, I'd like to drop the whole script altogether and just live with // a different MAC. / On the dgl5500 the mac in the eeprom is the same on all devices (and iirc it's the default mac the firmware also uses). Patching it to the right one in the firmware file is what happens in 82_patch_ath10k. Imre That's... interesting. I'll make v2 with a patched 82_patch_ath10k then, unless someone can think of a better way to provide a proper MAC for the board. Matti ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] [PATCH 0/2] mac80211: Update QCA9880 firmware (ath10k)
Hi Imre, On the dgl5500 the mac in the eeprom is the same on all devices (and iirc it's the default mac the firmware also uses). Patching it to the right one in the firmware file is what happens in 82_patch_ath10k. I am not very familiar with ath10k, but why are you not patching the calibration data instead (like other devices do it in target/linux/ar71xx/base-files/lib/preinit/81_load_ath10k_board_bin)? ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] [PATCH 0/2] mac80211: Update QCA9880 firmware (ath10k)
On Thu, 25 Jun 2015 20:59:19 +0200, Matti Laakso malaa...@elisanet.fi wrote: On 22/06/15 20:19, Matti Laakso wrote: / These patches backport the new FW API v5 to ath10k // (upstream commits 8348db2, 53513c3, and dc3632a) // and change mac80211 to use the new firmware. // // /Could you also update target/linux/ar71xx/base-files/lib/preinit/82_patch_ath10k? See https://patchwork.ozlabs.org/patch/466945/ and possibly https://patchwork.ozlabs.org/patch/481472/ Thanks, Stijn Hi, That seems like a burden to maintain. Do I understand correctly that the device has valid data in the PCIe card EEPROM (so it is not necessary to create a file with the calibration data in 81_load_ath10k_board_bin), but the MAC is incorrect? Is the MAC completely bogus (like same for all devices), or valid but just not what it says on the router cover? If the latter, I'd like to drop the whole script altogether and just live with a different MAC. On the dgl5500 the mac in the eeprom is the same on all devices (and iirc it's the default mac the firmware also uses). Patching it to the right one in the firmware file is what happens in 82_patch_ath10k. Imre ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] [PATCH 0/2] mac80211: Update QCA9880 firmware (ath10k)
On Fri, 26 Jun 2015 10:11:28 +0200, Martin Blumenstingl martin.blumensti...@googlemail.com wrote: Hi Imre, On the dgl5500 the mac in the eeprom is the same on all devices (and iirc it's the default mac the firmware also uses). Patching it to the right one in the firmware file is what happens in 82_patch_ath10k. I am not very familiar with ath10k, but why are you not patching the calibration data instead (like other devices do it in target/linux/ar71xx/base-files/lib/preinit/81_load_ath10k_board_bin)? Because the calibration data is not stored on the flash like on those, but on the PCIe card's eeprom. Imre ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] [PATCH 0/2] mac80211: Update QCA9880 firmware (ath10k)
On 22/06/15 20:19, Matti Laakso wrote: / These patches backport the new FW API v5 to ath10k // (upstream commits 8348db2, 53513c3, and dc3632a) // and change mac80211 to use the new firmware. // // /Could you also update target/linux/ar71xx/base-files/lib/preinit/82_patch_ath10k? See https://patchwork.ozlabs.org/patch/466945/ and possibly https://patchwork.ozlabs.org/patch/481472/ Thanks, Stijn Hi, That seems like a burden to maintain. Do I understand correctly that the device has valid data in the PCIe card EEPROM (so it is not necessary to create a file with the calibration data in 81_load_ath10k_board_bin), but the MAC is incorrect? Is the MAC completely bogus (like same for all devices), or valid but just not what it says on the router cover? If the latter, I'd like to drop the whole script altogether and just live with a different MAC. Best regards, Matti ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] [PATCH 0/2] mac80211: Update QCA9880 firmware (ath10k)
On 22/06/15 20:19, Matti Laakso wrote: These patches backport the new FW API v5 to ath10k (upstream commits 8348db2, 53513c3, and dc3632a) and change mac80211 to use the new firmware. Could you also update target/linux/ar71xx/base-files/lib/preinit/82_patch_ath10k? See https://patchwork.ozlabs.org/patch/466945/ and possibly https://patchwork.ozlabs.org/patch/481472/ Thanks, Stijn ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
[OpenWrt-Devel] [PATCH 0/2] mac80211: Update QCA9880 firmware (ath10k)
Current firmware version for QCA9880 is known to stall for some time when a station in power saving mode (e.g. mobile phone) leaves the AP range. This stall causes ath10k timeouts during station removal and subsequently firmware thinks station is still connected, causing Spurios quick kickout for STA -messages in kernel log. A fix is to use a newer firmware where this problem has been worked around by sending management frames via normal packet transport channel instead of the problematic management interface. These patches backport the new FW API v5 to ath10k (upstream commits 8348db2, 53513c3, and dc3632a) and change mac80211 to use the new firmware. Fixes https://dev.openwrt.org/ticket/18794 (even though it is already marked as fixed). Matti Laakso (2): mac80211: Backport ath10k firmware API v5 support mac80211: Update QCA9880 firmware to 10.2.4.70-2 package/kernel/mac80211/Makefile | 6 +- ...31-ath10k_add_ath10k_fw_ie_htt_op_version.patch | 385 + .../patches/932-ath10k_bump_up_fw_api_to_5.patch | 43 +++ ...tt_op_backwards_compatibility_for_qca6174.patch | 47 +++ 4 files changed, 478 insertions(+), 3 deletions(-) create mode 100644 package/kernel/mac80211/patches/931-ath10k_add_ath10k_fw_ie_htt_op_version.patch create mode 100644 package/kernel/mac80211/patches/932-ath10k_bump_up_fw_api_to_5.patch create mode 100644 package/kernel/mac80211/patches/933-ath10k_fix_htt_op_backwards_compatibility_for_qca6174.patch -- 2.1.0 ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel