d/freebsd-nq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
1fe39413c4
freebsd-nq/sys/dev/iwm/if_iwm.c

6415 lines
173 KiB
C
Raw Normal View History

Add support for Intel 8265 WiFi Obtained from: OpenBSD MFC after: 1 month
2017-10-09 15:48:56 +00:00
/* $OpenBSD: if_iwm.c,v 1.167 2017/04/04 00:40:52 claudio Exp $ */
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/*
* Copyright (c) 2014 genua mbh <info@genua.de>
* Copyright (c) 2014 Fixup Software Ltd.
*
* Permission to use, copy, modify, and distribute this software for any
* purpose with or without fee is hereby granted, provided that the above
* copyright notice and this permission notice appear in all copies.
*
* THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
* WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
* ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
* WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
* ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
* OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
*/
/*-
* Based on BSD-licensed source modules in the Linux iwlwifi driver,
* which were used as the reference documentation for this implementation.
*
* Driver version we are currently based off of is
* Linux 3.14.3 (tag id a2df521e42b1d9a23f620ac79dbfe8655a8391dd)
*
***********************************************************************
*
* This file is provided under a dual BSD/GPLv2 license. When using or
* redistributing this file, you may do so under either license.
*
* GPL LICENSE SUMMARY
*
* Copyright(c) 2007 - 2013 Intel Corporation. All rights reserved.
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of version 2 of the GNU General Public License as
* published by the Free Software Foundation.
*
* This program is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110,
* USA
*
* The full GNU General Public License is included in this distribution
* in the file called COPYING.
*
* Contact Information:
* Intel Linux Wireless <ilw@linux.intel.com>
* Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
*
*
* BSD LICENSE
*
* Copyright(c) 2005 - 2013 Intel Corporation. All rights reserved.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* * Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* * Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in
* the documentation and/or other materials provided with the
* distribution.
* * Neither the name Intel Corporation nor the names of its
* contributors may be used to endorse or promote products derived
* from this software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
* A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
* OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
* LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
/*-
* Copyright (c) 2007-2010 Damien Bergamini <damien.bergamini@free.fr>
*
* Permission to use, copy, modify, and distribute this software for any
* purpose with or without fee is hereby granted, provided that the above
* copyright notice and this permission notice appear in all copies.
*
* THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
* WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
* ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
* WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
* ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
* OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
iwm: add 'opt_wlan.h' include into source files. Include net80211 options file, so IEEE80211_DEBUG_REFCNT option will be handled correctly.
2016-05-19 23:31:00 +00:00
#include "opt_wlan.h"
[iwm] include opt_iwm.h and opt_wlan.h consistently in all files.
2017-05-09 02:38:23 +00:00
#include "opt_iwm.h"
iwm: add 'opt_wlan.h' include into source files. Include net80211 options file, so IEEE80211_DEBUG_REFCNT option will be handled correctly.
2016-05-19 23:31:00 +00:00
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
#include <sys/param.h>
#include <sys/bus.h>
#include <sys/conf.h>
#include <sys/endian.h>
#include <sys/firmware.h>
#include <sys/kernel.h>
#include <sys/malloc.h>
#include <sys/mbuf.h>
#include <sys/mutex.h>
#include <sys/module.h>
#include <sys/proc.h>
#include <sys/rman.h>
#include <sys/socket.h>
#include <sys/sockio.h>
#include <sys/sysctl.h>
#include <sys/linker.h>
#include <machine/bus.h>
#include <machine/endian.h>
#include <machine/resource.h>
#include <dev/pci/pcivar.h>
#include <dev/pci/pcireg.h>
#include <net/bpf.h>
#include <net/if.h>
#include <net/if_var.h>
#include <net/if_arp.h>
#include <net/if_dl.h>
#include <net/if_media.h>
#include <net/if_types.h>
#include <netinet/in.h>
#include <netinet/in_systm.h>
#include <netinet/if_ether.h>
#include <netinet/ip.h>
#include <net80211/ieee80211_var.h>
#include <net80211/ieee80211_regdomain.h>
#include <net80211/ieee80211_ratectl.h>
#include <net80211/ieee80211_radiotap.h>
iwm: use the proper include path for iwm headers.
2015-08-08 21:08:35 +00:00
#include <dev/iwm/if_iwmreg.h>
#include <dev/iwm/if_iwmvar.h>
[iwm] Move struct iwm_cfg chipset configs to if_iwm_7000.c and _8000.c * This is more similar to how code/definitions are distributed in Linux's iwlwifi. * This should make recognizing new chipset variants, and adding additional flags from the Linux iwlwifi code easier, without blowing up if_iwm.c Obtained from: dragonflybsd.git 27d11320e707d2c41424efc1983762f6799941d6
2017-02-22 04:35:53 +00:00
#include <dev/iwm/if_iwm_config.h>
iwm: use the proper include path for iwm headers.
2015-08-08 21:08:35 +00:00
#include <dev/iwm/if_iwm_debug.h>
[iwm] Add implementation of the notification wait api from iwlwifi. Obtained from: Linux iwlwifi Obtained from: DragonflyBSD commit 94dc1dadceb57b688036211262d678bc6bbdde37
2017-02-06 05:27:07 +00:00
#include <dev/iwm/if_iwm_notif_wait.h>
iwm: use the proper include path for iwm headers.
2015-08-08 21:08:35 +00:00
#include <dev/iwm/if_iwm_util.h>
#include <dev/iwm/if_iwm_binding.h>
#include <dev/iwm/if_iwm_phy_db.h>
#include <dev/iwm/if_iwm_mac_ctxt.h>
#include <dev/iwm/if_iwm_phy_ctxt.h>
#include <dev/iwm/if_iwm_time_event.h>
#include <dev/iwm/if_iwm_power.h>
#include <dev/iwm/if_iwm_scan.h>
[iwm] Move Smart Fifo handling into if_iwm_sf.c, sync with Linux iwlwifi. * This change also fixes a possible issue in the existing smart-fifo code, which set the IWM_SF_CFG_DUMMY_NOTIF_OFF bit on AC8260 chipsets, although that's only used in iwlwifi for Family 8000 chipsets connected via SDIO interface. Obtained from: Dragonflybsd.git cb650b01526b0aeef3c4307d926e7f1428997d50
2017-06-04 21:05:58 +00:00
#include <dev/iwm/if_iwm_sf.h>
[iwm] Factor out firmware station handling into if_iwm_sta.c. * This adds iwm_mvm_rm_sta(), which will be used to tear down firmware state for better/cleaner iwm_newstate() handling. * Makes iwm_enable_txq() and iwm_mvm_flush_tx_path() non-static, add the declarations to if_iwm_util.h for now. Obtained from: dragonflybsd.git 85d1c6190c4c3564b1a347f253e823aa95c202b2
2017-05-12 06:03:23 +00:00
#include <dev/iwm/if_iwm_sta.h>
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm: use the proper include path for iwm headers.
2015-08-08 21:08:35 +00:00
#include <dev/iwm/if_iwm_pcie_trans.h>
[iwm] add LED blinking for iwm hardware that supports it. Submitted by: Imre Vadasz <imre@vdsz.com> Obtained from: DragonflyBSD, Linux iwlwifi/mvm
2016-06-02 04:42:28 +00:00
#include <dev/iwm/if_iwm_led.h>
[iwm] Add support for Firmware paging, needed for newer 8000C firmware. * Uses the IWM_FW_PAGING_BLOCK_CMD firmware command to tell the firmware what memory ranges to use for paging. Obtained from: dragonflybsd.git 8a5b199964f8e7bdb00039f0b48817a01b402f18
2017-02-24 07:07:58 +00:00
#include <dev/iwm/if_iwm_fw.h>
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
/* From DragonflyBSD */
#define mtodoff(m, t, off) ((t)((m)->m_data + (off)))
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
const uint8_t iwm_nvm_channels[] = {
/* 2.4 GHz */
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14,
/* 5 GHz */
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
36, 40, 44, 48, 52, 56, 60, 64,
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
100, 104, 108, 112, 116, 120, 124, 128, 132, 136, 140, 144,
149, 153, 157, 161, 165
};
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
_Static_assert(nitems(iwm_nvm_channels) <= IWM_NUM_CHANNELS,
"IWM_NUM_CHANNELS is too small");
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
const uint8_t iwm_nvm_channels_8000[] = {
/* 2.4 GHz */
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14,
/* 5 GHz */
36, 40, 44, 48, 52, 56, 60, 64, 68, 72, 76, 80, 84, 88, 92,
96, 100, 104, 108, 112, 116, 120, 124, 128, 132, 136, 140, 144,
149, 153, 157, 161, 165, 169, 173, 177, 181
};
_Static_assert(nitems(iwm_nvm_channels_8000) <= IWM_NUM_CHANNELS_8000,
"IWM_NUM_CHANNELS_8000 is too small");
#define IWM_NUM_2GHZ_CHANNELS 14
#define IWM_N_HW_ADDR_MASK 0xF
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/*
* XXX For now, there's simply a fixed set of rate table entries
* that are populated.
*/
const struct iwm_rate {
uint8_t rate;
uint8_t plcp;
} iwm_rates[] = {
{ 2, IWM_RATE_1M_PLCP },
{ 4, IWM_RATE_2M_PLCP },
{ 11, IWM_RATE_5M_PLCP },
{ 22, IWM_RATE_11M_PLCP },
{ 12, IWM_RATE_6M_PLCP },
{ 18, IWM_RATE_9M_PLCP },
{ 24, IWM_RATE_12M_PLCP },
{ 36, IWM_RATE_18M_PLCP },
{ 48, IWM_RATE_24M_PLCP },
{ 72, IWM_RATE_36M_PLCP },
{ 96, IWM_RATE_48M_PLCP },
{ 108, IWM_RATE_54M_PLCP },
};
#define IWM_RIDX_CCK 0
#define IWM_RIDX_OFDM 4
#define IWM_RIDX_MAX (nitems(iwm_rates)-1)
#define IWM_RIDX_IS_CCK(_i_) ((_i_) < IWM_RIDX_OFDM)
#define IWM_RIDX_IS_OFDM(_i_) ((_i_) >= IWM_RIDX_OFDM)
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
struct iwm_nvm_section {
uint16_t length;
uint8_t *data;
};
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
#define IWM_MVM_UCODE_ALIVE_TIMEOUT hz
[iwm] Use notification wait API to wait for calibration to complete. Tested: * 7260, STA mode (2g, 5g) Obtained from: DragonflyBSD commit 1e0cf8ec6fcd77978f5336297ece61a415790f84
2017-02-06 05:35:11 +00:00
#define IWM_MVM_UCODE_CALIB_TIMEOUT (2*hz)
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
struct iwm_mvm_alive_data {
int valid;
uint32_t scd_base_addr;
};
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static int iwm_store_cscheme(struct iwm_softc *, const uint8_t *, size_t);
static int iwm_firmware_store_section(struct iwm_softc *,
enum iwm_ucode_type,
const uint8_t *, size_t);
static int iwm_set_default_calib(struct iwm_softc *, const void *);
static void iwm_fw_info_free(struct iwm_fw_info *);
if_iwm - Move iwm_read_firmware() call into iwm_attach(). * We should load the firmware exactly once before the driver really initializes the hardware the first time, and unload it at detach time. There is no need to retrieve the firmware during execution of iwm_mvm_load_ucode_wait_alive(), we should make sure we already have the firmware data at hand before that. * The existing sc_preinit_hook code fails to deal with the case where if_iwm is loaded by the loader (or is statically linked) and the firmware needs to be loaded from disk. So we can just call iwm_read_firmware() from iwm_attach() directly. * A separate solution will have to be added to properly defer the firmware loading during bootup, until the necessary filesystem is mounted. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (0104ee1f4cb6a2313c00c2526c6ae98d42e5041d)
2019-01-24 03:42:23 +00:00
static int iwm_read_firmware(struct iwm_softc *);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static int iwm_alloc_fwmem(struct iwm_softc *);
static int iwm_alloc_sched(struct iwm_softc *);
static int iwm_alloc_kw(struct iwm_softc *);
static int iwm_alloc_ict(struct iwm_softc *);
static int iwm_alloc_rx_ring(struct iwm_softc *, struct iwm_rx_ring *);
static void iwm_reset_rx_ring(struct iwm_softc *, struct iwm_rx_ring *);
static void iwm_free_rx_ring(struct iwm_softc *, struct iwm_rx_ring *);
static int iwm_alloc_tx_ring(struct iwm_softc *, struct iwm_tx_ring *,
int);
static void iwm_reset_tx_ring(struct iwm_softc *, struct iwm_tx_ring *);
static void iwm_free_tx_ring(struct iwm_softc *, struct iwm_tx_ring *);
static void iwm_enable_interrupts(struct iwm_softc *);
static void iwm_restore_interrupts(struct iwm_softc *);
static void iwm_disable_interrupts(struct iwm_softc *);
static void iwm_ict_reset(struct iwm_softc *);
static int iwm_allow_mcast(struct ieee80211vap *, struct iwm_softc *);
static void iwm_stop_device(struct iwm_softc *);
static void iwm_mvm_nic_config(struct iwm_softc *);
static int iwm_nic_rx_init(struct iwm_softc *);
static int iwm_nic_tx_init(struct iwm_softc *);
static int iwm_nic_init(struct iwm_softc *);
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
static int iwm_trans_pcie_fw_alive(struct iwm_softc *, uint32_t);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static int iwm_nvm_read_chunk(struct iwm_softc *, uint16_t, uint16_t,
uint16_t, uint8_t *, uint16_t *);
static int iwm_nvm_read_section(struct iwm_softc *, uint16_t, uint8_t *,
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
uint16_t *, uint32_t);
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
static uint32_t iwm_eeprom_channel_flags(uint16_t);
static void iwm_add_channel_band(struct iwm_softc *,
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
struct ieee80211_channel[], int, int *, int, size_t,
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
const uint8_t[]);
static void iwm_init_channel_map(struct ieee80211com *, int, int *,
struct ieee80211_channel[]);
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
static struct iwm_nvm_data *
iwm_parse_nvm_data(struct iwm_softc *, const uint16_t *,
const uint16_t *, const uint16_t *,
const uint16_t *, const uint16_t *,
const uint16_t *);
static void iwm_free_nvm_data(struct iwm_nvm_data *);
static void iwm_set_hw_address_family_8000(struct iwm_softc *,
struct iwm_nvm_data *,
const uint16_t *,
const uint16_t *);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
static int iwm_get_sku(const struct iwm_softc *, const uint16_t *,
const uint16_t *);
static int iwm_get_nvm_version(const struct iwm_softc *, const uint16_t *);
static int iwm_get_radio_cfg(const struct iwm_softc *, const uint16_t *,
const uint16_t *);
static int iwm_get_n_hw_addrs(const struct iwm_softc *,
const uint16_t *);
static void iwm_set_radio_cfg(const struct iwm_softc *,
struct iwm_nvm_data *, uint32_t);
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
static struct iwm_nvm_data *
iwm_parse_nvm_sections(struct iwm_softc *, struct iwm_nvm_section *);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static int iwm_nvm_init(struct iwm_softc *);
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
static int iwm_pcie_load_section(struct iwm_softc *, uint8_t,
const struct iwm_fw_desc *);
static int iwm_pcie_load_firmware_chunk(struct iwm_softc *, uint32_t,
bus_addr_t, uint32_t);
static int iwm_pcie_load_cpu_sections_8000(struct iwm_softc *sc,
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
const struct iwm_fw_img *,
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
int, int *);
static int iwm_pcie_load_cpu_sections(struct iwm_softc *,
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
const struct iwm_fw_img *,
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
int, int *);
static int iwm_pcie_load_given_ucode_8000(struct iwm_softc *,
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
const struct iwm_fw_img *);
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
static int iwm_pcie_load_given_ucode(struct iwm_softc *,
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
const struct iwm_fw_img *);
static int iwm_start_fw(struct iwm_softc *, const struct iwm_fw_img *);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static int iwm_send_tx_ant_cfg(struct iwm_softc *, uint8_t);
static int iwm_send_phy_cfg_cmd(struct iwm_softc *);
static int iwm_mvm_load_ucode_wait_alive(struct iwm_softc *,
enum iwm_ucode_type);
static int iwm_run_init_mvm_ucode(struct iwm_softc *, int);
if_iwm - Configure the PCIe LTR, fix PCI express capability accesses. Taken-From: Linux iwlwifi Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (08a7ad5a5ff65aaaf2df6a609be7a4e1df43efc3)
2019-01-24 03:44:48 +00:00
static int iwm_mvm_config_ltr(struct iwm_softc *sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static int iwm_rx_addbuf(struct iwm_softc *, int, int);
static int iwm_mvm_get_signal_strength(struct iwm_softc *,
struct iwm_rx_phy_info *);
static void iwm_mvm_rx_rx_phy_cmd(struct iwm_softc *,
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
struct iwm_rx_packet *);
[iwm] Sync statistics API definitions with Linux iwlwifi. Obtained from: dragonflybsd.git 75895a53a9c1ba60d75be9b4bf6e49a37f91a7cf
2017-07-26 05:40:52 +00:00
static int iwm_get_noise(struct iwm_softc *,
[iwm] fix up RSSI calculations for both scan results and normal RX operations. * hard code a noise floor of -96 for now. The noise floor update code returns some "interesting" values that I can't map to anything useful right now. * Ensure a default noise floor is set - otherwise the initial scan results have a noise floor of '0'. * Fix up the RSSI calculation to be correctly relative to the noise floor. The RSSI routines return an absolute value in dBm - so fix this up. * Cap RSSI values appropriately. * Ensure we pass in a 1/2 dB unit value in to net80211. Tested: * Intel 7260, STA mode iwm0: <Intel Dual Band Wireless AC 7260> mem 0xf1400000-0xf1401fff irq 17 at device 0.0 on pci2 iwm0: hw rev 0x140, fw ver 16.242414.0, address xx:xx:xx:xx:xx:xx
2016-09-18 05:07:18 +00:00
const struct iwm_mvm_statistics_rx_non_phy *);
[iwm] Sync statistics API definitions with Linux iwlwifi. Obtained from: dragonflybsd.git 75895a53a9c1ba60d75be9b4bf6e49a37f91a7cf
2017-07-26 05:40:52 +00:00
static void iwm_mvm_handle_rx_statistics(struct iwm_softc *,
struct iwm_rx_packet *);
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
static boolean_t iwm_mvm_rx_rx_mpdu(struct iwm_softc *, struct mbuf *,
uint32_t, boolean_t);
iwm: convert to ieee80211_tx_complete() Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D4755
2016-01-03 10:10:11 +00:00
static int iwm_mvm_rx_tx_cmd_single(struct iwm_softc *,
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
struct iwm_rx_packet *,
struct iwm_node *);
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
static void iwm_mvm_rx_tx_cmd(struct iwm_softc *, struct iwm_rx_packet *);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static void iwm_cmd_done(struct iwm_softc *, struct iwm_rx_packet *);
#if 0
static void iwm_update_sched(struct iwm_softc *, int, int, uint8_t,
uint16_t);
#endif
static const struct iwm_rate *
iwm_tx_fill_cmd(struct iwm_softc *, struct iwm_node *,
[iwm] use rate control info from the node txrates; use mgmtrate for EAPOL frames This changes the transmit rate control code to do a few things: * use fixed rates (mcast, ucast, mgmt) where required. * Don't use a hard-coded 11a or 11bg rate for non-data frames - use what net80211 says we should use. * use mgmtrate for EAPOL frames. Reviewed by: avos Differential Revision: https://reviews.freebsd.org/D7994
2016-09-21 20:56:10 +00:00
struct mbuf *, struct iwm_tx_cmd *);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static int iwm_tx(struct iwm_softc *, struct mbuf *,
struct ieee80211_node *, int);
static int iwm_raw_xmit(struct ieee80211_node *, struct mbuf *,
const struct ieee80211_bpf_params *);
[iwm] Move in_phyctxt from struct iwm_node to phy_ctxt in struct iwm_vap. * This better matches how things are organized in Linux's iwlwifi. Obtained from: dragonflybsd.git 0cf16dd2e0e09a3e5140e50222ac2e69bcdb19a2
2017-05-09 05:32:35 +00:00
static int iwm_mvm_update_quotas(struct iwm_softc *, struct iwm_vap *);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static int iwm_auth(struct ieee80211vap *, struct iwm_softc *);
static struct ieee80211_node *
iwm_node_alloc(struct ieee80211vap *,
const uint8_t[IEEE80211_ADDR_LEN]);
if_iwm - Update firmware rs table, instead of indexing the table in tx cmds. * Rather than providing a non-zero index into the firmware RS table, we should always use index 0 and update the firmware RS table whenever our chosen tx rate for data-frames changes. * Send IWM_LQ_CMD updates when the tx rate gets updated by the net80211 rate control (which is after we tell the tx status to the net80211 rate-control in iwm_mvm_rx_tx_cmd_single()). * Disregard frames transferred with a different tx rate than the currently selected rate for the rate-control calculations. This way we avoid counting management frames (which are sent at a slow, and fixed rate), as well as frames we added to the tx queue just before a new IWM_LQ_CMD update took effect. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (5d6b465e288ac5b52d7115688d4e6516acbbea1c)
2019-01-24 03:41:09 +00:00
static uint8_t iwm_rate_from_ucode_rate(uint32_t);
static int iwm_rate2ridx(struct iwm_softc *, uint8_t);
static void iwm_setrates(struct iwm_softc *, struct iwm_node *, int);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static int iwm_media_change(struct ifnet *);
static int iwm_newstate(struct ieee80211vap *, enum ieee80211_state, int);
static void iwm_endscan_cb(void *, int);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
static int iwm_send_bt_init_conf(struct iwm_softc *);
[iwm] Check for lar_disable tunable, and lar_enabled flag from NVM. * LAR can be disabled with the hw.iwm.lar.disable tunable now. * On Family 8000 devices we need to check the lar_enabled flag from nvm_data in addition to the TLV_CAPA_LAR_SUPPORT flag from the firmware. * Add a separate IWM_DEBUG_LAR debugging flag. Obtained from: dragonflybsd.git 0593e39cb295aa996ecf789ed4990c3b255f1770
2017-06-04 21:10:14 +00:00
static boolean_t iwm_mvm_is_lar_supported(struct iwm_softc *);
static boolean_t iwm_mvm_is_wifi_mcc_supported(struct iwm_softc *);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
static int iwm_send_update_mcc_cmd(struct iwm_softc *, const char *);
static void iwm_mvm_tt_tx_backoff(struct iwm_softc *, uint32_t);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static int iwm_init_hw(struct iwm_softc *);
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
static void iwm_init(struct iwm_softc *);
static void iwm_start(struct iwm_softc *);
static void iwm_stop(struct iwm_softc *);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static void iwm_watchdog(void *);
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
static void iwm_parent(struct ieee80211com *);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
#ifdef IWM_DEBUG
static const char *
iwm_desc_lookup(uint32_t);
static void iwm_nic_error(struct iwm_softc *);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
static void iwm_nic_umac_error(struct iwm_softc *);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
#endif
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
static void iwm_handle_rxb(struct iwm_softc *, struct mbuf *);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static void iwm_notif_intr(struct iwm_softc *);
static void iwm_intr(void *);
static int iwm_attach(device_t);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
static int iwm_is_valid_ether_addr(uint8_t *);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static void iwm_preinit(void *);
static int iwm_detach_local(struct iwm_softc *sc, int);
static void iwm_init_task(void *);
static void iwm_radiotap_attach(struct iwm_softc *);
static struct ieee80211vap *
iwm_vap_create(struct ieee80211com *,
const char [IFNAMSIZ], int,
enum ieee80211_opmode, int,
const uint8_t [IEEE80211_ADDR_LEN],
const uint8_t [IEEE80211_ADDR_LEN]);
static void iwm_vap_delete(struct ieee80211vap *);
[iwm] Cleanup mbufq draining. Add iwm_xmit_queue_drain() like in iwn(4). * iwm_xmit_queue_drain() calls ieee80211_free_node(), removing a possible memory leak, compared to using just mbufq_drain(). * Remove duplicate mbufq_drain() from iwm_mvm_rm_sta(), this should be handled in the caller. Obtained from: dragonflybsd.git 339d45fda40072e0aca5ece639173204716f11fe
2017-07-26 05:29:08 +00:00
static void iwm_xmit_queue_drain(struct iwm_softc *);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static void iwm_scan_start(struct ieee80211com *);
static void iwm_scan_end(struct ieee80211com *);
static void iwm_update_mcast(struct ieee80211com *);
static void iwm_set_channel(struct ieee80211com *);
static void iwm_scan_curchan(struct ieee80211_scan_state *, unsigned long);
static void iwm_scan_mindwell(struct ieee80211_scan_state *);
static int iwm_detach(device_t);
[iwm] Check for lar_disable tunable, and lar_enabled flag from NVM. * LAR can be disabled with the hw.iwm.lar.disable tunable now. * On Family 8000 devices we need to check the lar_enabled flag from nvm_data in addition to the TLV_CAPA_LAR_SUPPORT flag from the firmware. * Add a separate IWM_DEBUG_LAR debugging flag. Obtained from: dragonflybsd.git 0593e39cb295aa996ecf789ed4990c3b255f1770
2017-06-04 21:10:14 +00:00
static int iwm_lar_disable = 0;
TUNABLE_INT("hw.iwm.lar.disable", &iwm_lar_disable);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/*
* Firmware parser.
*/
static int
iwm_store_cscheme(struct iwm_softc *sc, const uint8_t *data, size_t dlen)
{
const struct iwm_fw_cscheme_list *l = (const void *)data;
if (dlen < sizeof(*l) ||
dlen < sizeof(l->size) + l->size * sizeof(*l->cs))
return EINVAL;
/* we don't actually store anything for now, always use s/w crypto */
return 0;
}
static int
iwm_firmware_store_section(struct iwm_softc *sc,
enum iwm_ucode_type type, const uint8_t *data, size_t dlen)
{
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
struct iwm_fw_img *fws;
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
struct iwm_fw_desc *fwone;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (type >= IWM_UCODE_TYPE_MAX)
return EINVAL;
if (dlen < sizeof(uint32_t))
return EINVAL;
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
fws = &sc->sc_fw.img[type];
[iwm] Get rid of some gratuitous constant renaming wrt. Linux iwlwifi. * IWM_UCODE_SECT_MAX -> IWM_UCODE_SECTION_MAX * IWM_UCODE_TYPE_* -> IWM_UCODE_* (except for IWM_UCODE_TYPE_MAX which stays). Obtained from: DragonflyBSD commit ff4d1fc3ed002c9fb362423da6c45d711b65658a
2017-02-06 05:07:15 +00:00
if (fws->fw_count >= IWM_UCODE_SECTION_MAX)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
return EINVAL;
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
fwone = &fws->sec[fws->fw_count];
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* first 32bit are device load offset */
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
memcpy(&fwone->offset, data, sizeof(uint32_t));
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* rest is data */
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
fwone->data = data + sizeof(uint32_t);
fwone->len = dlen - sizeof(uint32_t);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
fws->fw_count++;
return 0;
}
[iwm] Use IWM_DEFAULT_SCAN_CHANNELS define as default for sc_capa_n_scan_channels. Approved by: adrian (mentor) Obtained from: Linux iwlwifi Differential Revision: https://reviews.freebsd.org/D7938
2016-09-19 22:23:46 +00:00
#define IWM_DEFAULT_SCAN_CHANNELS 40
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* iwlwifi: iwl-drv.c */
struct iwm_tlv_calib_data {
uint32_t ucode_type;
struct iwm_tlv_calib_ctrl calib;
} __packed;
static int
iwm_set_default_calib(struct iwm_softc *sc, const void *data)
{
const struct iwm_tlv_calib_data *def_calib = data;
uint32_t ucode_type = le32toh(def_calib->ucode_type);
if (ucode_type >= IWM_UCODE_TYPE_MAX) {
device_printf(sc->sc_dev,
"Wrong ucode_type %u for default "
"calibration.\n", ucode_type);
return EINVAL;
}
sc->sc_default_calib[ucode_type].flow_trigger =
def_calib->calib.flow_trigger;
sc->sc_default_calib[ucode_type].event_trigger =
def_calib->calib.event_trigger;
return 0;
}
[iwm] Make ucode capabilities and api flags handling more like iwlwifi. Obtained from: dragonflybsd.git 757eecf0e6c92745aa2eee95811e573c8300850e
2017-03-23 04:50:38 +00:00
static int
iwm_set_ucode_api_flags(struct iwm_softc *sc, const uint8_t *data,
struct iwm_ucode_capabilities *capa)
{
const struct iwm_ucode_api *ucode_api = (const void *)data;
uint32_t api_index = le32toh(ucode_api->api_index);
uint32_t api_flags = le32toh(ucode_api->api_flags);
int i;
if (api_index >= howmany(IWM_NUM_UCODE_TLV_API, 32)) {
device_printf(sc->sc_dev,
"api flags index %d larger than supported by driver\n",
api_index);
/* don't return an error so we can load FW that has more bits */
return 0;
}
for (i = 0; i < 32; i++) {
if (api_flags & (1U << i))
setbit(capa->enabled_api, i + 32 * api_index);
}
return 0;
}
static int
iwm_set_ucode_capabilities(struct iwm_softc *sc, const uint8_t *data,
struct iwm_ucode_capabilities *capa)
{
const struct iwm_ucode_capa *ucode_capa = (const void *)data;
uint32_t api_index = le32toh(ucode_capa->api_index);
uint32_t api_flags = le32toh(ucode_capa->api_capa);
int i;
if (api_index >= howmany(IWM_NUM_UCODE_TLV_CAPA, 32)) {
device_printf(sc->sc_dev,
"capa flags index %d larger than supported by driver\n",
api_index);
/* don't return an error so we can load FW that has more bits */
return 0;
}
for (i = 0; i < 32; i++) {
if (api_flags & (1U << i))
setbit(capa->enabled_capa, i + 32 * api_index);
}
return 0;
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static void
iwm_fw_info_free(struct iwm_fw_info *fw)
{
iwm: store pointer for 'struct firmware' instead of 'size_t' and 'void *' pair. Approved by: adrian (mentor) Obtained from: DragonFlyBSD Differential Revision: https://reviews.freebsd.org/D4765
2016-01-04 21:03:01 +00:00
firmware_put(fw->fw_fp, FIRMWARE_UNLOAD);
fw->fw_fp = NULL;
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
memset(fw->img, 0, sizeof(fw->img));
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
static int
if_iwm - Move iwm_read_firmware() call into iwm_attach(). * We should load the firmware exactly once before the driver really initializes the hardware the first time, and unload it at detach time. There is no need to retrieve the firmware during execution of iwm_mvm_load_ucode_wait_alive(), we should make sure we already have the firmware data at hand before that. * The existing sc_preinit_hook code fails to deal with the case where if_iwm is loaded by the loader (or is statically linked) and the firmware needs to be loaded from disk. So we can just call iwm_read_firmware() from iwm_attach() directly. * A separate solution will have to be added to properly defer the firmware loading during bootup, until the necessary filesystem is mounted. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (0104ee1f4cb6a2313c00c2526c6ae98d42e5041d)
2019-01-24 03:42:23 +00:00
iwm_read_firmware(struct iwm_softc *sc)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
struct iwm_fw_info *fw = &sc->sc_fw;
const struct iwm_tlv_ucode_header *uhdr;
[iwm] Sync iwm_read_firmware()'s loop to iwlwifi's code. Obtained from: dragonflybsd.git d1c10ccfcf2d6d2a664f17197add0b4f93333181
2017-05-12 05:28:49 +00:00
const struct iwm_ucode_tlv *tlv;
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
struct iwm_ucode_capabilities *capa = &sc->sc_fw.ucode_capa;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
enum iwm_ucode_tlv_type tlv_type;
const struct firmware *fwp;
const uint8_t *data;
[iwm] Sync iwm_read_firmware()'s loop to iwlwifi's code. Obtained from: dragonflybsd.git d1c10ccfcf2d6d2a664f17197add0b4f93333181
2017-05-12 05:28:49 +00:00
uint32_t tlv_len;
[iwm] Store paging_mem_size field in firmware image information struct. Obtained from: DragonflyBSD commit a8524cc6c440e5ce9490ba2b0507c99ff6777c6d
2017-02-06 05:08:21 +00:00
uint32_t usniffer_img;
[iwm] Sync iwm_read_firmware()'s loop to iwlwifi's code. Obtained from: dragonflybsd.git d1c10ccfcf2d6d2a664f17197add0b4f93333181
2017-05-12 05:28:49 +00:00
const uint8_t *tlv_data;
[iwm] Store paging_mem_size field in firmware image information struct. Obtained from: DragonflyBSD commit a8524cc6c440e5ce9490ba2b0507c99ff6777c6d
2017-02-06 05:08:21 +00:00
uint32_t paging_mem_size;
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
int num_of_cpus;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
int error = 0;
size_t len;
/*
* Load firmware into driver memory.
iwm: store pointer for 'struct firmware' instead of 'size_t' and 'void *' pair. Approved by: adrian (mentor) Obtained from: DragonFlyBSD Differential Revision: https://reviews.freebsd.org/D4765
2016-01-04 21:03:01 +00:00
* fw_fp will be set.
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
*/
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
fwp = firmware_get(sc->cfg->fw_name);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (fwp == NULL) {
device_printf(sc->sc_dev,
"could not read firmware %s (error %d)\n",
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
sc->cfg->fw_name, error);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
goto out;
}
iwm: store pointer for 'struct firmware' instead of 'size_t' and 'void *' pair. Approved by: adrian (mentor) Obtained from: DragonFlyBSD Differential Revision: https://reviews.freebsd.org/D4765
2016-01-04 21:03:01 +00:00
fw->fw_fp = fwp;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
/* (Re-)Initialize default values. */
[iwm] Make ucode capabilities and api flags handling more like iwlwifi. Obtained from: dragonflybsd.git 757eecf0e6c92745aa2eee95811e573c8300850e
2017-03-23 04:50:38 +00:00
capa->flags = 0;
capa->max_probe_length = IWM_DEFAULT_MAX_PROBE_LENGTH;
capa->n_scan_channels = IWM_DEFAULT_SCAN_CHANNELS;
memset(capa->enabled_capa, 0, sizeof(capa->enabled_capa));
memset(capa->enabled_api, 0, sizeof(capa->enabled_api));
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
memset(sc->sc_fw_mcc, 0, sizeof(sc->sc_fw_mcc));
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/*
* Parse firmware contents
*/
iwm: store pointer for 'struct firmware' instead of 'size_t' and 'void *' pair. Approved by: adrian (mentor) Obtained from: DragonFlyBSD Differential Revision: https://reviews.freebsd.org/D4765
2016-01-04 21:03:01 +00:00
uhdr = (const void *)fw->fw_fp->data;
if (*(const uint32_t *)fw->fw_fp->data != 0
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
|| le32toh(uhdr->magic) != IWM_TLV_UCODE_MAGIC) {
device_printf(sc->sc_dev, "invalid firmware %s\n",
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
sc->cfg->fw_name);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
error = EINVAL;
goto out;
}
[iwm] Sync iwm_read_firmware()'s loop to iwlwifi's code. Obtained from: dragonflybsd.git d1c10ccfcf2d6d2a664f17197add0b4f93333181
2017-05-12 05:28:49 +00:00
snprintf(sc->sc_fwver, sizeof(sc->sc_fwver), "%u.%u (API ver %u)",
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
IWM_UCODE_MAJOR(le32toh(uhdr->ver)),
IWM_UCODE_MINOR(le32toh(uhdr->ver)),
IWM_UCODE_API(le32toh(uhdr->ver)));
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
data = uhdr->data;
iwm: store pointer for 'struct firmware' instead of 'size_t' and 'void *' pair. Approved by: adrian (mentor) Obtained from: DragonFlyBSD Differential Revision: https://reviews.freebsd.org/D4765
2016-01-04 21:03:01 +00:00
len = fw->fw_fp->datasize - sizeof(*uhdr);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Sync iwm_read_firmware()'s loop to iwlwifi's code. Obtained from: dragonflybsd.git d1c10ccfcf2d6d2a664f17197add0b4f93333181
2017-05-12 05:28:49 +00:00
while (len >= sizeof(*tlv)) {
len -= sizeof(*tlv);
tlv = (const void *)data;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Sync iwm_read_firmware()'s loop to iwlwifi's code. Obtained from: dragonflybsd.git d1c10ccfcf2d6d2a664f17197add0b4f93333181
2017-05-12 05:28:49 +00:00
tlv_len = le32toh(tlv->length);
tlv_type = le32toh(tlv->type);
tlv_data = tlv->data;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (len < tlv_len) {
device_printf(sc->sc_dev,
"firmware too short: %zu bytes\n",
len);
error = EINVAL;
goto parse_out;
}
[iwm] Sync iwm_read_firmware()'s loop to iwlwifi's code. Obtained from: dragonflybsd.git d1c10ccfcf2d6d2a664f17197add0b4f93333181
2017-05-12 05:28:49 +00:00
len -= roundup2(tlv_len, 4);
With the sync from Dragonfly BSD in r318216 a bug slipped in (also still present upstream it seems). The tlv variable was changed to a pointer but the advancement of the data pointer was left as sizeof(tlv). While the sizeof the (now) pointer equals the sizeof 2 x uint32_t (size of the struct) on 64bit platforms, on 32bit platforms the size of the advancement of the data pointer was wrong leading to firmware load issues. Correctly advance the data pointer by the size of the structure and not by the size of a pointer. PR: 219683 Submitted by: waddlesplash gamil.com (Haiku) on irc MFC after: 1 week
2019-01-15 22:31:54 +00:00
data += sizeof(*tlv) + roundup2(tlv_len, 4);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
switch ((int)tlv_type) {
case IWM_UCODE_TLV_PROBE_MAX_LEN:
[iwm] Sync iwm_read_firmware()'s loop to iwlwifi's code. Obtained from: dragonflybsd.git d1c10ccfcf2d6d2a664f17197add0b4f93333181
2017-05-12 05:28:49 +00:00
if (tlv_len != sizeof(uint32_t)) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev,
if_iwm - Move iwm_read_firmware() call into iwm_attach(). * We should load the firmware exactly once before the driver really initializes the hardware the first time, and unload it at detach time. There is no need to retrieve the firmware during execution of iwm_mvm_load_ucode_wait_alive(), we should make sure we already have the firmware data at hand before that. * The existing sc_preinit_hook code fails to deal with the case where if_iwm is loaded by the loader (or is statically linked) and the firmware needs to be loaded from disk. So we can just call iwm_read_firmware() from iwm_attach() directly. * A separate solution will have to be added to properly defer the firmware loading during bootup, until the necessary filesystem is mounted. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (0104ee1f4cb6a2313c00c2526c6ae98d42e5041d)
2019-01-24 03:42:23 +00:00
"%s: PROBE_MAX_LEN (%u) != sizeof(uint32_t)\n",
__func__, tlv_len);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
error = EINVAL;
goto parse_out;
}
[iwm] Make ucode capabilities and api flags handling more like iwlwifi. Obtained from: dragonflybsd.git 757eecf0e6c92745aa2eee95811e573c8300850e
2017-03-23 04:50:38 +00:00
capa->max_probe_length =
[iwm] Sync iwm_read_firmware()'s loop to iwlwifi's code. Obtained from: dragonflybsd.git d1c10ccfcf2d6d2a664f17197add0b4f93333181
2017-05-12 05:28:49 +00:00
le32_to_cpup((const uint32_t *)tlv_data);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* limit it to something sensible */
[iwm] Make ucode capabilities and api flags handling more like iwlwifi. Obtained from: dragonflybsd.git 757eecf0e6c92745aa2eee95811e573c8300850e
2017-03-23 04:50:38 +00:00
if (capa->max_probe_length >
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
IWM_SCAN_OFFLOAD_PROBE_REQ_SIZE) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_FIRMWARE_TLV,
"%s: IWM_UCODE_TLV_PROBE_MAX_LEN "
"ridiculous\n", __func__);
error = EINVAL;
goto parse_out;
}
break;
case IWM_UCODE_TLV_PAN:
if (tlv_len) {
device_printf(sc->sc_dev,
if_iwm - Move iwm_read_firmware() call into iwm_attach(). * We should load the firmware exactly once before the driver really initializes the hardware the first time, and unload it at detach time. There is no need to retrieve the firmware during execution of iwm_mvm_load_ucode_wait_alive(), we should make sure we already have the firmware data at hand before that. * The existing sc_preinit_hook code fails to deal with the case where if_iwm is loaded by the loader (or is statically linked) and the firmware needs to be loaded from disk. So we can just call iwm_read_firmware() from iwm_attach() directly. * A separate solution will have to be added to properly defer the firmware loading during bootup, until the necessary filesystem is mounted. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (0104ee1f4cb6a2313c00c2526c6ae98d42e5041d)
2019-01-24 03:42:23 +00:00
"%s: IWM_UCODE_TLV_PAN: tlv_len (%u) > 0\n",
__func__, tlv_len);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
error = EINVAL;
goto parse_out;
}
[iwm] Make ucode capabilities and api flags handling more like iwlwifi. Obtained from: dragonflybsd.git 757eecf0e6c92745aa2eee95811e573c8300850e
2017-03-23 04:50:38 +00:00
capa->flags |= IWM_UCODE_TLV_FLAGS_PAN;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
break;
case IWM_UCODE_TLV_FLAGS:
if (tlv_len < sizeof(uint32_t)) {
device_printf(sc->sc_dev,
if_iwm - Move iwm_read_firmware() call into iwm_attach(). * We should load the firmware exactly once before the driver really initializes the hardware the first time, and unload it at detach time. There is no need to retrieve the firmware during execution of iwm_mvm_load_ucode_wait_alive(), we should make sure we already have the firmware data at hand before that. * The existing sc_preinit_hook code fails to deal with the case where if_iwm is loaded by the loader (or is statically linked) and the firmware needs to be loaded from disk. So we can just call iwm_read_firmware() from iwm_attach() directly. * A separate solution will have to be added to properly defer the firmware loading during bootup, until the necessary filesystem is mounted. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (0104ee1f4cb6a2313c00c2526c6ae98d42e5041d)
2019-01-24 03:42:23 +00:00
"%s: IWM_UCODE_TLV_FLAGS: tlv_len (%u) < sizeof(uint32_t)\n",
__func__, tlv_len);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
error = EINVAL;
goto parse_out;
}
[iwm] Sync iwm_read_firmware()'s loop to iwlwifi's code. Obtained from: dragonflybsd.git d1c10ccfcf2d6d2a664f17197add0b4f93333181
2017-05-12 05:28:49 +00:00
if (tlv_len % sizeof(uint32_t)) {
device_printf(sc->sc_dev,
if_iwm - Move iwm_read_firmware() call into iwm_attach(). * We should load the firmware exactly once before the driver really initializes the hardware the first time, and unload it at detach time. There is no need to retrieve the firmware during execution of iwm_mvm_load_ucode_wait_alive(), we should make sure we already have the firmware data at hand before that. * The existing sc_preinit_hook code fails to deal with the case where if_iwm is loaded by the loader (or is statically linked) and the firmware needs to be loaded from disk. So we can just call iwm_read_firmware() from iwm_attach() directly. * A separate solution will have to be added to properly defer the firmware loading during bootup, until the necessary filesystem is mounted. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (0104ee1f4cb6a2313c00c2526c6ae98d42e5041d)
2019-01-24 03:42:23 +00:00
"%s: IWM_UCODE_TLV_FLAGS: tlv_len (%u) %% sizeof(uint32_t)\n",
__func__, tlv_len);
[iwm] Sync iwm_read_firmware()'s loop to iwlwifi's code. Obtained from: dragonflybsd.git d1c10ccfcf2d6d2a664f17197add0b4f93333181
2017-05-12 05:28:49 +00:00
error = EINVAL;
goto parse_out;
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/*
* Apparently there can be many flags, but Linux driver
* parses only the first one, and so do we.
*
* XXX: why does this override IWM_UCODE_TLV_PAN?
* Intentional or a bug? Observations from
* current firmware file:
* 1) TLV_PAN is parsed first
* 2) TLV_FLAGS contains TLV_FLAGS_PAN
* ==> this resets TLV_PAN to itself... hnnnk
*/
[iwm] Sync iwm_read_firmware()'s loop to iwlwifi's code. Obtained from: dragonflybsd.git d1c10ccfcf2d6d2a664f17197add0b4f93333181
2017-05-12 05:28:49 +00:00
capa->flags = le32_to_cpup((const uint32_t *)tlv_data);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
break;
case IWM_UCODE_TLV_CSCHEME:
if ((error = iwm_store_cscheme(sc,
tlv_data, tlv_len)) != 0) {
device_printf(sc->sc_dev,
"%s: iwm_store_cscheme(): returned %d\n",
if_iwm - Move iwm_read_firmware() call into iwm_attach(). * We should load the firmware exactly once before the driver really initializes the hardware the first time, and unload it at detach time. There is no need to retrieve the firmware during execution of iwm_mvm_load_ucode_wait_alive(), we should make sure we already have the firmware data at hand before that. * The existing sc_preinit_hook code fails to deal with the case where if_iwm is loaded by the loader (or is statically linked) and the firmware needs to be loaded from disk. So we can just call iwm_read_firmware() from iwm_attach() directly. * A separate solution will have to be added to properly defer the firmware loading during bootup, until the necessary filesystem is mounted. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (0104ee1f4cb6a2313c00c2526c6ae98d42e5041d)
2019-01-24 03:42:23 +00:00
__func__, error);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
goto parse_out;
}
break;
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
case IWM_UCODE_TLV_NUM_OF_CPU:
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (tlv_len != sizeof(uint32_t)) {
device_printf(sc->sc_dev,
if_iwm - Move iwm_read_firmware() call into iwm_attach(). * We should load the firmware exactly once before the driver really initializes the hardware the first time, and unload it at detach time. There is no need to retrieve the firmware during execution of iwm_mvm_load_ucode_wait_alive(), we should make sure we already have the firmware data at hand before that. * The existing sc_preinit_hook code fails to deal with the case where if_iwm is loaded by the loader (or is statically linked) and the firmware needs to be loaded from disk. So we can just call iwm_read_firmware() from iwm_attach() directly. * A separate solution will have to be added to properly defer the firmware loading during bootup, until the necessary filesystem is mounted. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (0104ee1f4cb6a2313c00c2526c6ae98d42e5041d)
2019-01-24 03:42:23 +00:00
"%s: IWM_UCODE_TLV_NUM_OF_CPU: tlv_len (%u) != sizeof(uint32_t)\n",
__func__, tlv_len);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
error = EINVAL;
goto parse_out;
}
[iwm] Sync iwm_read_firmware()'s loop to iwlwifi's code. Obtained from: dragonflybsd.git d1c10ccfcf2d6d2a664f17197add0b4f93333181
2017-05-12 05:28:49 +00:00
num_of_cpus = le32_to_cpup((const uint32_t *)tlv_data);
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
if (num_of_cpus == 2) {
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
fw->img[IWM_UCODE_REGULAR].is_dual_cpus =
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
TRUE;
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
fw->img[IWM_UCODE_INIT].is_dual_cpus =
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
TRUE;
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
fw->img[IWM_UCODE_WOWLAN].is_dual_cpus =
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
TRUE;
} else if ((num_of_cpus > 2) || (num_of_cpus < 1)) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev,
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
"%s: Driver supports only 1 or 2 CPUs\n",
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
__func__);
error = EINVAL;
goto parse_out;
}
break;
case IWM_UCODE_TLV_SEC_RT:
if ((error = iwm_firmware_store_section(sc,
[iwm] Get rid of some gratuitous constant renaming wrt. Linux iwlwifi. * IWM_UCODE_SECT_MAX -> IWM_UCODE_SECTION_MAX * IWM_UCODE_TYPE_* -> IWM_UCODE_* (except for IWM_UCODE_TYPE_MAX which stays). Obtained from: DragonflyBSD commit ff4d1fc3ed002c9fb362423da6c45d711b65658a
2017-02-06 05:07:15 +00:00
IWM_UCODE_REGULAR, tlv_data, tlv_len)) != 0) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev,
[iwm] Get rid of some gratuitous constant renaming wrt. Linux iwlwifi. * IWM_UCODE_SECT_MAX -> IWM_UCODE_SECTION_MAX * IWM_UCODE_TYPE_* -> IWM_UCODE_* (except for IWM_UCODE_TYPE_MAX which stays). Obtained from: DragonflyBSD commit ff4d1fc3ed002c9fb362423da6c45d711b65658a
2017-02-06 05:07:15 +00:00
"%s: IWM_UCODE_REGULAR: iwm_firmware_store_section() failed; %d\n",
if_iwm - Move iwm_read_firmware() call into iwm_attach(). * We should load the firmware exactly once before the driver really initializes the hardware the first time, and unload it at detach time. There is no need to retrieve the firmware during execution of iwm_mvm_load_ucode_wait_alive(), we should make sure we already have the firmware data at hand before that. * The existing sc_preinit_hook code fails to deal with the case where if_iwm is loaded by the loader (or is statically linked) and the firmware needs to be loaded from disk. So we can just call iwm_read_firmware() from iwm_attach() directly. * A separate solution will have to be added to properly defer the firmware loading during bootup, until the necessary filesystem is mounted. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (0104ee1f4cb6a2313c00c2526c6ae98d42e5041d)
2019-01-24 03:42:23 +00:00
__func__, error);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
goto parse_out;
}
break;
case IWM_UCODE_TLV_SEC_INIT:
if ((error = iwm_firmware_store_section(sc,
[iwm] Get rid of some gratuitous constant renaming wrt. Linux iwlwifi. * IWM_UCODE_SECT_MAX -> IWM_UCODE_SECTION_MAX * IWM_UCODE_TYPE_* -> IWM_UCODE_* (except for IWM_UCODE_TYPE_MAX which stays). Obtained from: DragonflyBSD commit ff4d1fc3ed002c9fb362423da6c45d711b65658a
2017-02-06 05:07:15 +00:00
IWM_UCODE_INIT, tlv_data, tlv_len)) != 0) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev,
[iwm] Get rid of some gratuitous constant renaming wrt. Linux iwlwifi. * IWM_UCODE_SECT_MAX -> IWM_UCODE_SECTION_MAX * IWM_UCODE_TYPE_* -> IWM_UCODE_* (except for IWM_UCODE_TYPE_MAX which stays). Obtained from: DragonflyBSD commit ff4d1fc3ed002c9fb362423da6c45d711b65658a
2017-02-06 05:07:15 +00:00
"%s: IWM_UCODE_INIT: iwm_firmware_store_section() failed; %d\n",
if_iwm - Move iwm_read_firmware() call into iwm_attach(). * We should load the firmware exactly once before the driver really initializes the hardware the first time, and unload it at detach time. There is no need to retrieve the firmware during execution of iwm_mvm_load_ucode_wait_alive(), we should make sure we already have the firmware data at hand before that. * The existing sc_preinit_hook code fails to deal with the case where if_iwm is loaded by the loader (or is statically linked) and the firmware needs to be loaded from disk. So we can just call iwm_read_firmware() from iwm_attach() directly. * A separate solution will have to be added to properly defer the firmware loading during bootup, until the necessary filesystem is mounted. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (0104ee1f4cb6a2313c00c2526c6ae98d42e5041d)
2019-01-24 03:42:23 +00:00
__func__, error);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
goto parse_out;
}
break;
case IWM_UCODE_TLV_SEC_WOWLAN:
if ((error = iwm_firmware_store_section(sc,
[iwm] Get rid of some gratuitous constant renaming wrt. Linux iwlwifi. * IWM_UCODE_SECT_MAX -> IWM_UCODE_SECTION_MAX * IWM_UCODE_TYPE_* -> IWM_UCODE_* (except for IWM_UCODE_TYPE_MAX which stays). Obtained from: DragonflyBSD commit ff4d1fc3ed002c9fb362423da6c45d711b65658a
2017-02-06 05:07:15 +00:00
IWM_UCODE_WOWLAN, tlv_data, tlv_len)) != 0) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev,
[iwm] Get rid of some gratuitous constant renaming wrt. Linux iwlwifi. * IWM_UCODE_SECT_MAX -> IWM_UCODE_SECTION_MAX * IWM_UCODE_TYPE_* -> IWM_UCODE_* (except for IWM_UCODE_TYPE_MAX which stays). Obtained from: DragonflyBSD commit ff4d1fc3ed002c9fb362423da6c45d711b65658a
2017-02-06 05:07:15 +00:00
"%s: IWM_UCODE_WOWLAN: iwm_firmware_store_section() failed; %d\n",
if_iwm - Move iwm_read_firmware() call into iwm_attach(). * We should load the firmware exactly once before the driver really initializes the hardware the first time, and unload it at detach time. There is no need to retrieve the firmware during execution of iwm_mvm_load_ucode_wait_alive(), we should make sure we already have the firmware data at hand before that. * The existing sc_preinit_hook code fails to deal with the case where if_iwm is loaded by the loader (or is statically linked) and the firmware needs to be loaded from disk. So we can just call iwm_read_firmware() from iwm_attach() directly. * A separate solution will have to be added to properly defer the firmware loading during bootup, until the necessary filesystem is mounted. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (0104ee1f4cb6a2313c00c2526c6ae98d42e5041d)
2019-01-24 03:42:23 +00:00
__func__, error);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
goto parse_out;
}
break;
case IWM_UCODE_TLV_DEF_CALIB:
if (tlv_len != sizeof(struct iwm_tlv_calib_data)) {
device_printf(sc->sc_dev,
if_iwm - Move iwm_read_firmware() call into iwm_attach(). * We should load the firmware exactly once before the driver really initializes the hardware the first time, and unload it at detach time. There is no need to retrieve the firmware during execution of iwm_mvm_load_ucode_wait_alive(), we should make sure we already have the firmware data at hand before that. * The existing sc_preinit_hook code fails to deal with the case where if_iwm is loaded by the loader (or is statically linked) and the firmware needs to be loaded from disk. So we can just call iwm_read_firmware() from iwm_attach() directly. * A separate solution will have to be added to properly defer the firmware loading during bootup, until the necessary filesystem is mounted. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (0104ee1f4cb6a2313c00c2526c6ae98d42e5041d)
2019-01-24 03:42:23 +00:00
"%s: IWM_UCODE_TLV_DEV_CALIB: tlv_len (%u) < sizeof(iwm_tlv_calib_data) (%zu)\n",
__func__, tlv_len,
sizeof(struct iwm_tlv_calib_data));
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
error = EINVAL;
goto parse_out;
}
if ((error = iwm_set_default_calib(sc, tlv_data)) != 0) {
device_printf(sc->sc_dev,
"%s: iwm_set_default_calib() failed: %d\n",
if_iwm - Move iwm_read_firmware() call into iwm_attach(). * We should load the firmware exactly once before the driver really initializes the hardware the first time, and unload it at detach time. There is no need to retrieve the firmware during execution of iwm_mvm_load_ucode_wait_alive(), we should make sure we already have the firmware data at hand before that. * The existing sc_preinit_hook code fails to deal with the case where if_iwm is loaded by the loader (or is statically linked) and the firmware needs to be loaded from disk. So we can just call iwm_read_firmware() from iwm_attach() directly. * A separate solution will have to be added to properly defer the firmware loading during bootup, until the necessary filesystem is mounted. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (0104ee1f4cb6a2313c00c2526c6ae98d42e5041d)
2019-01-24 03:42:23 +00:00
__func__, error);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
goto parse_out;
}
break;
case IWM_UCODE_TLV_PHY_SKU:
if (tlv_len != sizeof(uint32_t)) {
error = EINVAL;
device_printf(sc->sc_dev,
if_iwm - Move iwm_read_firmware() call into iwm_attach(). * We should load the firmware exactly once before the driver really initializes the hardware the first time, and unload it at detach time. There is no need to retrieve the firmware during execution of iwm_mvm_load_ucode_wait_alive(), we should make sure we already have the firmware data at hand before that. * The existing sc_preinit_hook code fails to deal with the case where if_iwm is loaded by the loader (or is statically linked) and the firmware needs to be loaded from disk. So we can just call iwm_read_firmware() from iwm_attach() directly. * A separate solution will have to be added to properly defer the firmware loading during bootup, until the necessary filesystem is mounted. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (0104ee1f4cb6a2313c00c2526c6ae98d42e5041d)
2019-01-24 03:42:23 +00:00
"%s: IWM_UCODE_TLV_PHY_SKU: tlv_len (%u) < sizeof(uint32_t)\n",
__func__, tlv_len);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
goto parse_out;
}
[iwm] Sync valid_tx_ant and valid_rx_ant mask handling with iwlwifi. * This fixes the phy_cfg field sent in the iwm_send_phy_cfg_cmd() command, which wasn't taking into account the valid_rx_ant and valid_tx_ant masks from nvm_data before. Tested: * 7260, STA mode, 2G and 5G Obtained from: DragonflyBSD commit cbb82693c18fd71b4eb86855b82d03995f352d65
2017-02-06 05:03:41 +00:00
sc->sc_fw.phy_config =
[iwm] Sync iwm_read_firmware()'s loop to iwlwifi's code. Obtained from: dragonflybsd.git d1c10ccfcf2d6d2a664f17197add0b4f93333181
2017-05-12 05:28:49 +00:00
le32_to_cpup((const uint32_t *)tlv_data);
[iwm] Sync valid_tx_ant and valid_rx_ant mask handling with iwlwifi. * This fixes the phy_cfg field sent in the iwm_send_phy_cfg_cmd() command, which wasn't taking into account the valid_rx_ant and valid_tx_ant masks from nvm_data before. Tested: * 7260, STA mode, 2G and 5G Obtained from: DragonflyBSD commit cbb82693c18fd71b4eb86855b82d03995f352d65
2017-02-06 05:03:41 +00:00
sc->sc_fw.valid_tx_ant = (sc->sc_fw.phy_config &
IWM_FW_PHY_CFG_TX_CHAIN) >>
IWM_FW_PHY_CFG_TX_CHAIN_POS;
sc->sc_fw.valid_rx_ant = (sc->sc_fw.phy_config &
IWM_FW_PHY_CFG_RX_CHAIN) >>
IWM_FW_PHY_CFG_RX_CHAIN_POS;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
break;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
case IWM_UCODE_TLV_API_CHANGES_SET: {
[iwm] Make ucode capabilities and api flags handling more like iwlwifi. Obtained from: dragonflybsd.git 757eecf0e6c92745aa2eee95811e573c8300850e
2017-03-23 04:50:38 +00:00
if (tlv_len != sizeof(struct iwm_ucode_api)) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
error = EINVAL;
goto parse_out;
}
[iwm] Make ucode capabilities and api flags handling more like iwlwifi. Obtained from: dragonflybsd.git 757eecf0e6c92745aa2eee95811e573c8300850e
2017-03-23 04:50:38 +00:00
if (iwm_set_ucode_api_flags(sc, tlv_data, capa)) {
error = EINVAL;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
goto parse_out;
}
break;
}
case IWM_UCODE_TLV_ENABLED_CAPABILITIES: {
[iwm] Make ucode capabilities and api flags handling more like iwlwifi. Obtained from: dragonflybsd.git 757eecf0e6c92745aa2eee95811e573c8300850e
2017-03-23 04:50:38 +00:00
if (tlv_len != sizeof(struct iwm_ucode_capa)) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
error = EINVAL;
goto parse_out;
}
[iwm] Make ucode capabilities and api flags handling more like iwlwifi. Obtained from: dragonflybsd.git 757eecf0e6c92745aa2eee95811e573c8300850e
2017-03-23 04:50:38 +00:00
if (iwm_set_ucode_capabilities(sc, tlv_data, capa)) {
error = EINVAL;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
goto parse_out;
}
break;
}
case 48: /* undocumented TLV */
case IWM_UCODE_TLV_SDIO_ADMA_ADDR:
case IWM_UCODE_TLV_FW_GSCAN_CAPA:
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* ignore, not used by current driver */
break;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
case IWM_UCODE_TLV_SEC_RT_USNIFFER:
if ((error = iwm_firmware_store_section(sc,
[iwm] Get rid of some gratuitous constant renaming wrt. Linux iwlwifi. * IWM_UCODE_SECT_MAX -> IWM_UCODE_SECTION_MAX * IWM_UCODE_TYPE_* -> IWM_UCODE_* (except for IWM_UCODE_TYPE_MAX which stays). Obtained from: DragonflyBSD commit ff4d1fc3ed002c9fb362423da6c45d711b65658a
2017-02-06 05:07:15 +00:00
IWM_UCODE_REGULAR_USNIFFER, tlv_data,
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
tlv_len)) != 0)
goto parse_out;
break;
[iwm] Store paging_mem_size field in firmware image information struct. Obtained from: DragonflyBSD commit a8524cc6c440e5ce9490ba2b0507c99ff6777c6d
2017-02-06 05:08:21 +00:00
case IWM_UCODE_TLV_PAGING:
if (tlv_len != sizeof(uint32_t)) {
error = EINVAL;
goto parse_out;
}
[iwm] Sync iwm_read_firmware()'s loop to iwlwifi's code. Obtained from: dragonflybsd.git d1c10ccfcf2d6d2a664f17197add0b4f93333181
2017-05-12 05:28:49 +00:00
paging_mem_size = le32_to_cpup((const uint32_t *)tlv_data);
[iwm] Store paging_mem_size field in firmware image information struct. Obtained from: DragonflyBSD commit a8524cc6c440e5ce9490ba2b0507c99ff6777c6d
2017-02-06 05:08:21 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_FIRMWARE_TLV,
"%s: Paging: paging enabled (size = %u bytes)\n",
__func__, paging_mem_size);
if (paging_mem_size > IWM_MAX_PAGING_IMAGE_SIZE) {
device_printf(sc->sc_dev,
"%s: Paging: driver supports up to %u bytes for paging image\n",
__func__, IWM_MAX_PAGING_IMAGE_SIZE);
error = EINVAL;
goto out;
}
if (paging_mem_size & (IWM_FW_PAGING_SIZE - 1)) {
device_printf(sc->sc_dev,
"%s: Paging: image isn't multiple %u\n",
__func__, IWM_FW_PAGING_SIZE);
error = EINVAL;
goto out;
}
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
sc->sc_fw.img[IWM_UCODE_REGULAR].paging_mem_size =
[iwm] Store paging_mem_size field in firmware image information struct. Obtained from: DragonflyBSD commit a8524cc6c440e5ce9490ba2b0507c99ff6777c6d
2017-02-06 05:08:21 +00:00
paging_mem_size;
usniffer_img = IWM_UCODE_REGULAR_USNIFFER;
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
sc->sc_fw.img[usniffer_img].paging_mem_size =
[iwm] Store paging_mem_size field in firmware image information struct. Obtained from: DragonflyBSD commit a8524cc6c440e5ce9490ba2b0507c99ff6777c6d
2017-02-06 05:08:21 +00:00
paging_mem_size;
break;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
case IWM_UCODE_TLV_N_SCAN_CHANNELS:
if (tlv_len != sizeof(uint32_t)) {
error = EINVAL;
goto parse_out;
}
[iwm] Make ucode capabilities and api flags handling more like iwlwifi. Obtained from: dragonflybsd.git 757eecf0e6c92745aa2eee95811e573c8300850e
2017-03-23 04:50:38 +00:00
capa->n_scan_channels =
[iwm] Sync iwm_read_firmware()'s loop to iwlwifi's code. Obtained from: dragonflybsd.git d1c10ccfcf2d6d2a664f17197add0b4f93333181
2017-05-12 05:28:49 +00:00
le32_to_cpup((const uint32_t *)tlv_data);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
break;
case IWM_UCODE_TLV_FW_VERSION:
if (tlv_len != sizeof(uint32_t) * 3) {
error = EINVAL;
goto parse_out;
}
snprintf(sc->sc_fwver, sizeof(sc->sc_fwver),
"%d.%d.%d",
le32toh(((const uint32_t *)tlv_data)[0]),
le32toh(((const uint32_t *)tlv_data)[1]),
le32toh(((const uint32_t *)tlv_data)[2]));
break;
[iwm] Recognize the IWM_UCODE_TLV_FW_MEM_SEG firmware section type. * Will be needed for loading version 22 of 7265D firmware. Obtained from: DragonflyBSD commit 1d244c8133cf15d00d46836bc48958188cf9f510
2017-02-08 07:05:56 +00:00
case IWM_UCODE_TLV_FW_MEM_SEG:
break;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
default:
device_printf(sc->sc_dev,
"%s: unknown firmware section %d, abort\n",
__func__, tlv_type);
error = EINVAL;
goto parse_out;
}
}
KASSERT(error == 0, ("unhandled error"));
parse_out:
if (error) {
device_printf(sc->sc_dev, "firmware parse error %d, "
"section type %d\n", error, tlv_type);
}
out:
if (error) {
iwm: store pointer for 'struct firmware' instead of 'size_t' and 'void *' pair. Approved by: adrian (mentor) Obtained from: DragonFlyBSD Differential Revision: https://reviews.freebsd.org/D4765
2016-01-04 21:03:01 +00:00
if (fw->fw_fp != NULL)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm_fw_info_free(fw);
if_iwm - Move iwm_read_firmware() call into iwm_attach(). * We should load the firmware exactly once before the driver really initializes the hardware the first time, and unload it at detach time. There is no need to retrieve the firmware during execution of iwm_mvm_load_ucode_wait_alive(), we should make sure we already have the firmware data at hand before that. * The existing sc_preinit_hook code fails to deal with the case where if_iwm is loaded by the loader (or is statically linked) and the firmware needs to be loaded from disk. So we can just call iwm_read_firmware() from iwm_attach() directly. * A separate solution will have to be added to properly defer the firmware loading during bootup, until the necessary filesystem is mounted. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (0104ee1f4cb6a2313c00c2526c6ae98d42e5041d)
2019-01-24 03:42:23 +00:00
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
return error;
}
/*
* DMA resource routines
*/
/* fwmem is used to load firmware onto the card */
static int
iwm_alloc_fwmem(struct iwm_softc *sc)
{
/* Must be aligned on a 16-byte boundary. */
return iwm_dma_contig_alloc(sc->sc_dmat, &sc->fw_dma,
[iwm] Get rid of unneeded IWM_FWDMASEGSZ and IWM_FWMAXSIZE constants * We only need IWM_FH_MEM_TB_MAX_LENGTH bytes for the fw_dma memory. Obtained from: dragonflybsd.git f414b8da30816705e32573cc58097cacac34f161
2017-02-22 02:49:57 +00:00
IWM_FH_MEM_TB_MAX_LENGTH, 16);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
/* tx scheduler rings. not used? */
static int
iwm_alloc_sched(struct iwm_softc *sc)
{
/* TX scheduler rings must be aligned on a 1KB boundary. */
[iwm] When stopping TX DMA, wait for all channels at once. * Makes the TX DMA stopping more similar to Linux code, and potentially a bit faster. Also, output an error message when TX DMA idling fails. Taken-From: Linux iwlwifi Tested: * AC3165, STA mode Approved by: adrian (mentor) Obtained from: DragonFlyBSD git 2ee486ddff973ac552ff787c17e8d83e8ae0f24c Differential Revision: https://reviews.freebsd.org/D7325
2016-07-27 20:51:31 +00:00
return iwm_dma_contig_alloc(sc->sc_dmat, &sc->sched_dma,
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
nitems(sc->txq) * sizeof(struct iwm_agn_scd_bc_tbl), 1024);
}
/* keep-warm page is used internally by the card. see iwl-fh.h for more info */
static int
iwm_alloc_kw(struct iwm_softc *sc)
{
return iwm_dma_contig_alloc(sc->sc_dmat, &sc->kw_dma, 4096, 4096);
}
/* interrupt cause table */
static int
iwm_alloc_ict(struct iwm_softc *sc)
{
return iwm_dma_contig_alloc(sc->sc_dmat, &sc->ict_dma,
IWM_ICT_SIZE, 1<<IWM_ICT_PADDR_SHIFT);
}
static int
iwm_alloc_rx_ring(struct iwm_softc *sc, struct iwm_rx_ring *ring)
{
bus_size_t size;
int i, error;
ring->cur = 0;
/* Allocate RX descriptors (256-byte aligned). */
size = IWM_RX_RING_COUNT * sizeof(uint32_t);
error = iwm_dma_contig_alloc(sc->sc_dmat, &ring->desc_dma, size, 256);
if (error != 0) {
device_printf(sc->sc_dev,
"could not allocate RX ring DMA memory\n");
goto fail;
}
ring->desc = ring->desc_dma.vaddr;
/* Allocate RX status area (16-byte aligned). */
error = iwm_dma_contig_alloc(sc->sc_dmat, &ring->stat_dma,
sizeof(*ring->stat), 16);
if (error != 0) {
device_printf(sc->sc_dev,
"could not allocate RX status DMA memory\n");
goto fail;
}
ring->stat = ring->stat_dma.vaddr;
/* Create RX buffer DMA tag. */
error = bus_dma_tag_create(sc->sc_dmat, 1, 0,
BUS_SPACE_MAXADDR_32BIT, BUS_SPACE_MAXADDR, NULL, NULL,
Remove BUS_DMA_NOWAIT from bus_dma_tag_create() invocations as it's no valid flag there.
2015-10-21 06:23:57 +00:00
IWM_RBUF_SIZE, 1, IWM_RBUF_SIZE, 0, NULL, NULL, &ring->data_dmat);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (error != 0) {
device_printf(sc->sc_dev,
"%s: could not create RX buf DMA tag, error %d\n",
__func__, error);
goto fail;
}
[iwm] Fix up busdma use in the RX path When allocating a new mbuf or bus_dmamap_load()-ing it fails, we can just keep the old mbuf since we are dropping that packet anyway. Instead of doing bus_dmamap_create() and bus_dmamap_destroy() all the time, create an extra bus_dmamap_t which we can use to safely try bus_dmamap_load()-ing the new mbuf. On success we just swap the spare bus_dmamap_t with the data->map of that ring entry. Tested: Tested with Intel AC7260, verified with vmstat -m that new kernel no longer visibly leaks memory from the M_DEVBUF malloc type. Before, leakage was 1KB every few seconds while ping(8)-ing over the wlan connection. Submitted by: Imre Vadasz <imre@vdsz.com> Approved by: re@ Obtained from: DragonflyBSD.git cc440b26818b5dfdd9af504d71c1b0e6522b53ef Differential Revision: https://reviews.freebsd.org/D6742
2016-06-13 00:13:20 +00:00
/* Allocate spare bus_dmamap_t for iwm_rx_addbuf() */
error = bus_dmamap_create(ring->data_dmat, 0, &ring->spare_map);
if (error != 0) {
device_printf(sc->sc_dev,
"%s: could not create RX buf DMA map, error %d\n",
__func__, error);
goto fail;
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/*
* Allocate and map RX buffers.
*/
for (i = 0; i < IWM_RX_RING_COUNT; i++) {
[iwm] Fix up busdma use in the RX path When allocating a new mbuf or bus_dmamap_load()-ing it fails, we can just keep the old mbuf since we are dropping that packet anyway. Instead of doing bus_dmamap_create() and bus_dmamap_destroy() all the time, create an extra bus_dmamap_t which we can use to safely try bus_dmamap_load()-ing the new mbuf. On success we just swap the spare bus_dmamap_t with the data->map of that ring entry. Tested: Tested with Intel AC7260, verified with vmstat -m that new kernel no longer visibly leaks memory from the M_DEVBUF malloc type. Before, leakage was 1KB every few seconds while ping(8)-ing over the wlan connection. Submitted by: Imre Vadasz <imre@vdsz.com> Approved by: re@ Obtained from: DragonflyBSD.git cc440b26818b5dfdd9af504d71c1b0e6522b53ef Differential Revision: https://reviews.freebsd.org/D6742
2016-06-13 00:13:20 +00:00
struct iwm_rx_data *data = &ring->data[i];
error = bus_dmamap_create(ring->data_dmat, 0, &data->map);
if (error != 0) {
device_printf(sc->sc_dev,
"%s: could not create RX buf DMA map, error %d\n",
__func__, error);
goto fail;
}
data->m = NULL;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if ((error = iwm_rx_addbuf(sc, IWM_RBUF_SIZE, i)) != 0) {
goto fail;
}
}
return 0;
fail: iwm_free_rx_ring(sc, ring);
return error;
}
[iwm] Add bit-polling in Rx-DMA init code path. Taken-From: OpenBSD (if_iwm.c r1.80) Submitted by: Imre Vadasz <imre@vdsz.com> Obtained from: DragonflyBSD ed35558754288911048cb607e57c688273ebd8d4
2016-06-02 05:00:52 +00:00
static void
iwm_reset_rx_ring(struct iwm_softc *sc, struct iwm_rx_ring *ring)
{
Ensure the ring state is also blanked upon reset, otherwise duplicate rx events get handled during reset paths. Submitted by: Matthew Dillion <dillon@apollo.backplane.com> Obtained from: DragonflyBSD
2015-09-18 17:39:31 +00:00
/* Reset the ring state */
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
ring->cur = 0;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
/*
* The hw rx ring index in shared memory must also be cleared,
* otherwise the discrepancy can cause reprocessing chaos.
*/
iwm(4): Fix possible null pointer indirection Obtained from: DragonFlyBSD (obtained from Haiku) MFC after: 1 week
2019-01-21 01:45:35 +00:00
if (sc->rxq.stat)
memset(sc->rxq.stat, 0, sizeof(*sc->rxq.stat));
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
static void
iwm_free_rx_ring(struct iwm_softc *sc, struct iwm_rx_ring *ring)
{
int i;
iwm_dma_contig_free(&ring->desc_dma);
iwm_dma_contig_free(&ring->stat_dma);
for (i = 0; i < IWM_RX_RING_COUNT; i++) {
struct iwm_rx_data *data = &ring->data[i];
if (data->m != NULL) {
bus_dmamap_sync(ring->data_dmat, data->map,
BUS_DMASYNC_POSTREAD);
bus_dmamap_unload(ring->data_dmat, data->map);
m_freem(data->m);
data->m = NULL;
}
if (data->map != NULL) {
bus_dmamap_destroy(ring->data_dmat, data->map);
data->map = NULL;
}
}
[iwm] Fix up busdma use in the RX path When allocating a new mbuf or bus_dmamap_load()-ing it fails, we can just keep the old mbuf since we are dropping that packet anyway. Instead of doing bus_dmamap_create() and bus_dmamap_destroy() all the time, create an extra bus_dmamap_t which we can use to safely try bus_dmamap_load()-ing the new mbuf. On success we just swap the spare bus_dmamap_t with the data->map of that ring entry. Tested: Tested with Intel AC7260, verified with vmstat -m that new kernel no longer visibly leaks memory from the M_DEVBUF malloc type. Before, leakage was 1KB every few seconds while ping(8)-ing over the wlan connection. Submitted by: Imre Vadasz <imre@vdsz.com> Approved by: re@ Obtained from: DragonflyBSD.git cc440b26818b5dfdd9af504d71c1b0e6522b53ef Differential Revision: https://reviews.freebsd.org/D6742
2016-06-13 00:13:20 +00:00
if (ring->spare_map != NULL) {
bus_dmamap_destroy(ring->data_dmat, ring->spare_map);
ring->spare_map = NULL;
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (ring->data_dmat != NULL) {
bus_dma_tag_destroy(ring->data_dmat);
ring->data_dmat = NULL;
}
}
static int
iwm_alloc_tx_ring(struct iwm_softc *sc, struct iwm_tx_ring *ring, int qid)
{
bus_addr_t paddr;
bus_size_t size;
[iwm] Use mbuf for large firmware commands, like OpenBSD does. We also need to consider the size of large firmware commands in iwm_alloc_tx_ring(), in the dma tag creation, when qid == IWM_MVM_CMD_QUEUE. The old code apparently only allocated a 2KB (MCLBYTES) sized buffer when it actually expected 4KB. Submitted by: Imre Vadasz <imre@vdsz.com> Approved by: re (gjb) Differential Revision: https://reviews.freebsd.org/D6824
2016-06-23 01:13:30 +00:00
size_t maxsize;
int nsegments;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
int i, error;
ring->qid = qid;
ring->queued = 0;
ring->cur = 0;
/* Allocate TX descriptors (256-byte aligned). */
size = IWM_TX_RING_COUNT * sizeof (struct iwm_tfd);
error = iwm_dma_contig_alloc(sc->sc_dmat, &ring->desc_dma, size, 256);
if (error != 0) {
device_printf(sc->sc_dev,
"could not allocate TX ring DMA memory\n");
goto fail;
}
ring->desc = ring->desc_dma.vaddr;
/*
* We only use rings 0 through 9 (4 EDCA + cmd) so there is no need
* to allocate commands space for other rings.
*/
if (qid > IWM_MVM_CMD_QUEUE)
return 0;
size = IWM_TX_RING_COUNT * sizeof(struct iwm_device_cmd);
error = iwm_dma_contig_alloc(sc->sc_dmat, &ring->cmd_dma, size, 4);
if (error != 0) {
device_printf(sc->sc_dev,
"could not allocate TX cmd DMA memory\n");
goto fail;
}
ring->cmd = ring->cmd_dma.vaddr;
[iwm] Use mbuf for large firmware commands, like OpenBSD does. We also need to consider the size of large firmware commands in iwm_alloc_tx_ring(), in the dma tag creation, when qid == IWM_MVM_CMD_QUEUE. The old code apparently only allocated a 2KB (MCLBYTES) sized buffer when it actually expected 4KB. Submitted by: Imre Vadasz <imre@vdsz.com> Approved by: re (gjb) Differential Revision: https://reviews.freebsd.org/D6824
2016-06-23 01:13:30 +00:00
/* FW commands may require more mapped space than packets. */
if (qid == IWM_MVM_CMD_QUEUE) {
maxsize = IWM_RBUF_SIZE;
nsegments = 1;
} else {
maxsize = MCLBYTES;
nsegments = IWM_MAX_SCATTER - 2;
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
error = bus_dma_tag_create(sc->sc_dmat, 1, 0,
[iwm] Use mbuf for large firmware commands, like OpenBSD does. We also need to consider the size of large firmware commands in iwm_alloc_tx_ring(), in the dma tag creation, when qid == IWM_MVM_CMD_QUEUE. The old code apparently only allocated a 2KB (MCLBYTES) sized buffer when it actually expected 4KB. Submitted by: Imre Vadasz <imre@vdsz.com> Approved by: re (gjb) Differential Revision: https://reviews.freebsd.org/D6824
2016-06-23 01:13:30 +00:00
BUS_SPACE_MAXADDR_32BIT, BUS_SPACE_MAXADDR, NULL, NULL, maxsize,
nsegments, maxsize, 0, NULL, NULL, &ring->data_dmat);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (error != 0) {
device_printf(sc->sc_dev, "could not create TX buf DMA tag\n");
goto fail;
}
paddr = ring->cmd_dma.paddr;
for (i = 0; i < IWM_TX_RING_COUNT; i++) {
struct iwm_tx_data *data = &ring->data[i];
data->cmd_paddr = paddr;
data->scratch_paddr = paddr + sizeof(struct iwm_cmd_header)
+ offsetof(struct iwm_tx_cmd, scratch);
paddr += sizeof(struct iwm_device_cmd);
error = bus_dmamap_create(ring->data_dmat, 0, &data->map);
if (error != 0) {
device_printf(sc->sc_dev,
"could not create TX buf DMA map\n");
goto fail;
}
}
KASSERT(paddr == ring->cmd_dma.paddr + size,
("invalid physical address"));
return 0;
fail: iwm_free_tx_ring(sc, ring);
return error;
}
static void
iwm_reset_tx_ring(struct iwm_softc *sc, struct iwm_tx_ring *ring)
{
int i;
for (i = 0; i < IWM_TX_RING_COUNT; i++) {
struct iwm_tx_data *data = &ring->data[i];
if (data->m != NULL) {
bus_dmamap_sync(ring->data_dmat, data->map,
BUS_DMASYNC_POSTWRITE);
bus_dmamap_unload(ring->data_dmat, data->map);
m_freem(data->m);
data->m = NULL;
}
}
/* Clear TX descriptors. */
memset(ring->desc, 0, ring->desc_dma.size);
bus_dmamap_sync(ring->desc_dma.tag, ring->desc_dma.map,
BUS_DMASYNC_PREWRITE);
sc->qfullmsk &= ~(1 << ring->qid);
ring->queued = 0;
ring->cur = 0;
[iwm] Implement apmg_wake_up_wa workaround properly for 7000 family. * Add iwm_pcie_set_cmd_in_flight() and iwm_pcie_clear_cmd_in_flight() helper methods. * Use ring->queued tracking in the command queue to set/clear the cmd_hold_nic_awake bit at the right points. Taken-From: Linux iwlwifi Obtained from: DragonflyBSD commit ce43f57f5308b579ea21e8a5a29969114ba2247d
2017-02-08 06:53:23 +00:00
if (ring->qid == IWM_MVM_CMD_QUEUE && sc->cmd_hold_nic_awake)
iwm_pcie_clear_cmd_in_flight(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
static void
iwm_free_tx_ring(struct iwm_softc *sc, struct iwm_tx_ring *ring)
{
int i;
iwm_dma_contig_free(&ring->desc_dma);
iwm_dma_contig_free(&ring->cmd_dma);
for (i = 0; i < IWM_TX_RING_COUNT; i++) {
struct iwm_tx_data *data = &ring->data[i];
if (data->m != NULL) {
bus_dmamap_sync(ring->data_dmat, data->map,
BUS_DMASYNC_POSTWRITE);
bus_dmamap_unload(ring->data_dmat, data->map);
m_freem(data->m);
data->m = NULL;
}
if (data->map != NULL) {
bus_dmamap_destroy(ring->data_dmat, data->map);
data->map = NULL;
}
}
if (ring->data_dmat != NULL) {
bus_dma_tag_destroy(ring->data_dmat);
ring->data_dmat = NULL;
}
}
/*
* High-level hardware frobbing routines
*/
static void
iwm_enable_interrupts(struct iwm_softc *sc)
{
sc->sc_intmask = IWM_CSR_INI_SET_MASK;
IWM_WRITE(sc, IWM_CSR_INT_MASK, sc->sc_intmask);
}
static void
iwm_restore_interrupts(struct iwm_softc *sc)
{
IWM_WRITE(sc, IWM_CSR_INT_MASK, sc->sc_intmask);
}
static void
iwm_disable_interrupts(struct iwm_softc *sc)
{
/* disable interrupts */
IWM_WRITE(sc, IWM_CSR_INT_MASK, 0);
/* acknowledge all interrupts */
IWM_WRITE(sc, IWM_CSR_INT, ~0);
IWM_WRITE(sc, IWM_CSR_FH_INT_STATUS, ~0);
}
static void
iwm_ict_reset(struct iwm_softc *sc)
{
iwm_disable_interrupts(sc);
/* Reset ICT table. */
memset(sc->ict_dma.vaddr, 0, IWM_ICT_SIZE);
sc->ict_cur = 0;
/* Set physical address of ICT table (4KB aligned). */
IWM_WRITE(sc, IWM_CSR_DRAM_INT_TBL_REG,
IWM_CSR_DRAM_INT_TBL_ENABLE
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
| IWM_CSR_DRAM_INIT_TBL_WRITE_POINTER
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
| IWM_CSR_DRAM_INIT_TBL_WRAP_CHECK
| sc->ict_dma.paddr >> IWM_ICT_PADDR_SHIFT);
/* Switch to ICT interrupt mode in driver. */
sc->sc_flags |= IWM_FLAG_USE_ICT;
/* Re-enable interrupts. */
IWM_WRITE(sc, IWM_CSR_INT, ~0);
iwm_enable_interrupts(sc);
}
/* iwlwifi pcie/trans.c */
/*
* Since this .. hard-resets things, it's time to actually
* mark the first vap (if any) as having no mac context.
* It's annoying, but since the driver is potentially being
* stop/start'ed whilst active (thanks openbsd port!) we
* have to correctly track this.
*/
static void
iwm_stop_device(struct iwm_softc *sc)
{
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
struct ieee80211com *ic = &sc->sc_ic;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
struct ieee80211vap *vap = TAILQ_FIRST(&ic->ic_vaps);
[iwm] When stopping TX DMA, wait for all channels at once. * Makes the TX DMA stopping more similar to Linux code, and potentially a bit faster. Also, output an error message when TX DMA idling fails. Taken-From: Linux iwlwifi Tested: * AC3165, STA mode Approved by: adrian (mentor) Obtained from: DragonFlyBSD git 2ee486ddff973ac552ff787c17e8d83e8ae0f24c Differential Revision: https://reviews.freebsd.org/D7325
2016-07-27 20:51:31 +00:00
int chnl, qid;
uint32_t mask = 0;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* tell the device to stop sending interrupts */
iwm_disable_interrupts(sc);
/*
* FreeBSD-local: mark the first vap as not-uploaded,
* so the next transition through auth/assoc
* will correctly populate the MAC context.
*/
if (vap) {
struct iwm_vap *iv = IWM_VAP(vap);
[iwm] Move in_phyctxt from struct iwm_node to phy_ctxt in struct iwm_vap. * This better matches how things are organized in Linux's iwlwifi. Obtained from: dragonflybsd.git 0cf16dd2e0e09a3e5140e50222ac2e69bcdb19a2
2017-05-09 05:32:35 +00:00
iv->phy_ctxt = NULL;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iv->is_uploaded = 0;
}
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
sc->sc_firmware_state = 0;
iwm - Improve firmware Time Event handling. * This is a mix of the OpenBSD Git 7fd9664469d1b717a307eebd74aeececbd3c41cc change, and syncing with the Linux iwlwifi code. Taken-From: Linux iwlwifi, and OpenBSD Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (706a3044afd27c3fecfdf57bec1695310e53e228)
2019-01-24 03:47:04 +00:00
sc->sc_flags &= ~IWM_FLAG_TE_ACTIVE;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* device going down, Stop using ICT table */
sc->sc_flags &= ~IWM_FLAG_USE_ICT;
/* stop tx and rx. tx and rx bits, as usual, are from if_iwn */
if (iwm_nic_lock(sc)) {
[iwm] iwm_{read,write}_prph() don't grab the nic lock in iwm themselves. * Fix a couple of cases where the nic lock ended up not being grabbed during an iwm_read_prph() or iwm_write_prph(). Obtained from: dragonflybsd.git 6c5470f2db219c61e362c981fea969d97e1b8293
2017-05-09 02:41:33 +00:00
iwm_write_prph(sc, IWM_SCD_TXFACT, 0);
[iwm] When stopping TX DMA, wait for all channels at once. * Makes the TX DMA stopping more similar to Linux code, and potentially a bit faster. Also, output an error message when TX DMA idling fails. Taken-From: Linux iwlwifi Tested: * AC3165, STA mode Approved by: adrian (mentor) Obtained from: DragonFlyBSD git 2ee486ddff973ac552ff787c17e8d83e8ae0f24c Differential Revision: https://reviews.freebsd.org/D7325
2016-07-27 20:51:31 +00:00
/* Stop each Tx DMA channel */
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
for (chnl = 0; chnl < IWM_FH_TCSR_CHNL_NUM; chnl++) {
IWM_WRITE(sc,
IWM_FH_TCSR_CHNL_TX_CONFIG_REG(chnl), 0);
[iwm] When stopping TX DMA, wait for all channels at once. * Makes the TX DMA stopping more similar to Linux code, and potentially a bit faster. Also, output an error message when TX DMA idling fails. Taken-From: Linux iwlwifi Tested: * AC3165, STA mode Approved by: adrian (mentor) Obtained from: DragonFlyBSD git 2ee486ddff973ac552ff787c17e8d83e8ae0f24c Differential Revision: https://reviews.freebsd.org/D7325
2016-07-27 20:51:31 +00:00
mask |= IWM_FH_TSSR_TX_STATUS_REG_MSK_CHNL_IDLE(chnl);
}
/* Wait for DMA channels to be idle */
[iwm] Fix iwm_poll_bit() usage in iwm_stop_device(), fixup r303418. * iwm_poll_bit() returns 1 on success and 0 on failure, whereas iwl_poll_bit() in Linux's iwlwifi returns >= 0 on success and < 0 on failure. * Because of the wrong iwm_poll_bit return code check, no warning was printed if tx DMA stopping failed. Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D7371
2016-07-30 19:03:32 +00:00
if (!iwm_poll_bit(sc, IWM_FH_TSSR_TX_STATUS_REG, mask, mask,
5000)) {
[iwm] When stopping TX DMA, wait for all channels at once. * Makes the TX DMA stopping more similar to Linux code, and potentially a bit faster. Also, output an error message when TX DMA idling fails. Taken-From: Linux iwlwifi Tested: * AC3165, STA mode Approved by: adrian (mentor) Obtained from: DragonFlyBSD git 2ee486ddff973ac552ff787c17e8d83e8ae0f24c Differential Revision: https://reviews.freebsd.org/D7325
2016-07-27 20:51:31 +00:00
device_printf(sc->sc_dev,
"Failing on timeout while stopping DMA channel: [0x%08x]\n",
IWM_READ(sc, IWM_FH_TSSR_TX_STATUS_REG));
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
iwm_nic_unlock(sc);
}
[iwm] Get rid of iwm_disable_rx_dma, just use iwm_pcie_rx_stop directly. * This also fixes one of many small nic lock handling bugs, and matches iwlwifi's code. Obtained from: DragonflyBSD git 50787d03cd0a0366c9cc4a055bb6977e5f65c85d
2017-02-08 06:43:02 +00:00
iwm_pcie_rx_stop(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Stop RX ring. */
iwm_reset_rx_ring(sc, &sc->rxq);
/* Reset all TX rings. */
for (qid = 0; qid < nitems(sc->txq); qid++)
iwm_reset_tx_ring(sc, &sc->txq[qid]);
[iwm] Only for family 7000 power-down busmaster DMA clocks when stopping. Taken-From: Linux iwlwifi Obtained from: DragonflyBSD commit 4c45994fcc77373ae2fb0901db15368c9731f641
2017-02-08 06:54:08 +00:00
if (sc->cfg->device_family == IWM_DEVICE_FAMILY_7000) {
/* Power-down device's busmaster DMA clocks */
[iwm] iwm_{read,write}_prph() don't grab the nic lock in iwm themselves. * Fix a couple of cases where the nic lock ended up not being grabbed during an iwm_read_prph() or iwm_write_prph(). Obtained from: dragonflybsd.git 6c5470f2db219c61e362c981fea969d97e1b8293
2017-05-09 02:41:33 +00:00
if (iwm_nic_lock(sc)) {
iwm_write_prph(sc, IWM_APMG_CLK_DIS_REG,
IWM_APMG_CLK_VAL_DMA_CLK_RQT);
[iwm] Add iwm_nic_unlock() calls missing from previous commit. Obtained from: dragonflybsd.git f88ab372284e63c4c13da93e9026a203b9b4cdc5
2017-05-09 02:42:53 +00:00
iwm_nic_unlock(sc);
[iwm] iwm_{read,write}_prph() don't grab the nic lock in iwm themselves. * Fix a couple of cases where the nic lock ended up not being grabbed during an iwm_read_prph() or iwm_write_prph(). Obtained from: dragonflybsd.git 6c5470f2db219c61e362c981fea969d97e1b8293
2017-05-09 02:41:33 +00:00
}
[iwm] Only for family 7000 power-down busmaster DMA clocks when stopping. Taken-From: Linux iwlwifi Obtained from: DragonflyBSD commit 4c45994fcc77373ae2fb0901db15368c9731f641
2017-02-08 06:54:08 +00:00
DELAY(5);
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Make sure (redundant) we've released our request to stay awake */
IWM_CLRBITS(sc, IWM_CSR_GP_CNTRL,
IWM_CSR_GP_CNTRL_REG_FLAG_MAC_ACCESS_REQ);
/* Stop the device, and put it in low power state */
iwm_apm_stop(sc);
/* Upon stop, the APM issues an interrupt if HW RF kill is set.
* Clean again the interrupt here
*/
iwm_disable_interrupts(sc);
/* stop and reset the on-board processor */
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
IWM_WRITE(sc, IWM_CSR_RESET, IWM_CSR_RESET_REG_FLAG_SW_RESET);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/*
* Even if we stop the HW, we still want the RF kill
* interrupt
*/
iwm_enable_rfkill_int(sc);
iwm_check_rfkill(sc);
}
/* iwlwifi: mvm/ops.c */
static void
iwm_mvm_nic_config(struct iwm_softc *sc)
{
uint8_t radio_cfg_type, radio_cfg_step, radio_cfg_dash;
uint32_t reg_val = 0;
[iwm] Sync valid_tx_ant and valid_rx_ant mask handling with iwlwifi. * This fixes the phy_cfg field sent in the iwm_send_phy_cfg_cmd() command, which wasn't taking into account the valid_rx_ant and valid_tx_ant masks from nvm_data before. Tested: * 7260, STA mode, 2G and 5G Obtained from: DragonflyBSD commit cbb82693c18fd71b4eb86855b82d03995f352d65
2017-02-06 05:03:41 +00:00
uint32_t phy_config = iwm_mvm_get_phy_config(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Sync valid_tx_ant and valid_rx_ant mask handling with iwlwifi. * This fixes the phy_cfg field sent in the iwm_send_phy_cfg_cmd() command, which wasn't taking into account the valid_rx_ant and valid_tx_ant masks from nvm_data before. Tested: * 7260, STA mode, 2G and 5G Obtained from: DragonflyBSD commit cbb82693c18fd71b4eb86855b82d03995f352d65
2017-02-06 05:03:41 +00:00
radio_cfg_type = (phy_config & IWM_FW_PHY_CFG_RADIO_TYPE) >>
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_FW_PHY_CFG_RADIO_TYPE_POS;
[iwm] Sync valid_tx_ant and valid_rx_ant mask handling with iwlwifi. * This fixes the phy_cfg field sent in the iwm_send_phy_cfg_cmd() command, which wasn't taking into account the valid_rx_ant and valid_tx_ant masks from nvm_data before. Tested: * 7260, STA mode, 2G and 5G Obtained from: DragonflyBSD commit cbb82693c18fd71b4eb86855b82d03995f352d65
2017-02-06 05:03:41 +00:00
radio_cfg_step = (phy_config & IWM_FW_PHY_CFG_RADIO_STEP) >>
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_FW_PHY_CFG_RADIO_STEP_POS;
[iwm] Sync valid_tx_ant and valid_rx_ant mask handling with iwlwifi. * This fixes the phy_cfg field sent in the iwm_send_phy_cfg_cmd() command, which wasn't taking into account the valid_rx_ant and valid_tx_ant masks from nvm_data before. Tested: * 7260, STA mode, 2G and 5G Obtained from: DragonflyBSD commit cbb82693c18fd71b4eb86855b82d03995f352d65
2017-02-06 05:03:41 +00:00
radio_cfg_dash = (phy_config & IWM_FW_PHY_CFG_RADIO_DASH) >>
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_FW_PHY_CFG_RADIO_DASH_POS;
/* SKU control */
reg_val |= IWM_CSR_HW_REV_STEP(sc->sc_hw_rev) <<
IWM_CSR_HW_IF_CONFIG_REG_POS_MAC_STEP;
reg_val |= IWM_CSR_HW_REV_DASH(sc->sc_hw_rev) <<
IWM_CSR_HW_IF_CONFIG_REG_POS_MAC_DASH;
/* radio configuration */
reg_val |= radio_cfg_type << IWM_CSR_HW_IF_CONFIG_REG_POS_PHY_TYPE;
reg_val |= radio_cfg_step << IWM_CSR_HW_IF_CONFIG_REG_POS_PHY_STEP;
reg_val |= radio_cfg_dash << IWM_CSR_HW_IF_CONFIG_REG_POS_PHY_DASH;
IWM_WRITE(sc, IWM_CSR_HW_IF_CONFIG_REG, reg_val);
IWM_DPRINTF(sc, IWM_DEBUG_RESET,
"Radio type=0x%x-0x%x-0x%x\n", radio_cfg_type,
radio_cfg_step, radio_cfg_dash);
/*
* W/A : NIC is stuck in a reset state after Early PCIe power off
* (PCIe power is lost before PERST# is asserted), causing ME FW
* to lose ownership and not being able to obtain it back.
*/
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->device_family == IWM_DEVICE_FAMILY_7000) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
iwm_set_bits_mask_prph(sc, IWM_APMG_PS_CTRL_REG,
IWM_APMG_PS_CTRL_EARLY_PWR_OFF_RESET_DIS,
~IWM_APMG_PS_CTRL_EARLY_PWR_OFF_RESET_DIS);
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
static int
iwm_nic_rx_init(struct iwm_softc *sc)
{
/*
* Initialize RX ring. This is from the iwn driver.
*/
memset(sc->rxq.stat, 0, sizeof(*sc->rxq.stat));
[iwm] Get rid of iwm_disable_rx_dma, just use iwm_pcie_rx_stop directly. * This also fixes one of many small nic lock handling bugs, and matches iwlwifi's code. Obtained from: DragonflyBSD git 50787d03cd0a0366c9cc4a055bb6977e5f65c85d
2017-02-08 06:43:02 +00:00
/* Stop Rx DMA */
iwm_pcie_rx_stop(sc);
if (!iwm_nic_lock(sc))
return EBUSY;
/* reset and flush pointers */
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_WRITE(sc, IWM_FH_MEM_RCSR_CHNL0_RBDCB_WPTR, 0);
IWM_WRITE(sc, IWM_FH_MEM_RCSR_CHNL0_FLUSH_RB_REQ, 0);
IWM_WRITE(sc, IWM_FH_RSCSR_CHNL0_RDPTR, 0);
IWM_WRITE(sc, IWM_FH_RSCSR_CHNL0_RBDCB_WPTR_REG, 0);
/* Set physical address of RX ring (256-byte aligned). */
IWM_WRITE(sc,
IWM_FH_RSCSR_CHNL0_RBDCB_BASE_REG, sc->rxq.desc_dma.paddr >> 8);
/* Set physical address of RX status (16-byte aligned). */
IWM_WRITE(sc,
IWM_FH_RSCSR_CHNL0_STTS_WPTR_REG, sc->rxq.stat_dma.paddr >> 4);
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
/* Enable Rx DMA
* XXX 5000 HW isn't supported by the iwm(4) driver.
* IWM_FH_RCSR_CHNL0_RX_IGNORE_RXF_EMPTY is set because of HW bug in
* the credit mechanism in 5000 HW RX FIFO
* Direct rx interrupts to hosts
* Rx buffer size 4 or 8k or 12k
* RB timeout 0x10
* 256 RBDs
*/
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_WRITE(sc, IWM_FH_MEM_RCSR_CHNL0_CONFIG_REG,
IWM_FH_RCSR_RX_CONFIG_CHNL_EN_ENABLE_VAL |
IWM_FH_RCSR_CHNL0_RX_IGNORE_RXF_EMPTY | /* HW bug */
IWM_FH_RCSR_CHNL0_RX_CONFIG_IRQ_DEST_INT_HOST_VAL |
IWM_FH_RCSR_RX_CONFIG_REG_VAL_RB_SIZE_4K |
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
(IWM_RX_RB_TIMEOUT << IWM_FH_RCSR_RX_CONFIG_REG_IRQ_RBTH_POS) |
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_RX_QUEUE_SIZE_LOG << IWM_FH_RCSR_RX_CONFIG_RBDCB_SIZE_POS);
IWM_WRITE_1(sc, IWM_CSR_INT_COALESCING, IWM_HOST_INT_TIMEOUT_DEF);
/* W/A for interrupt coalescing bug in 7260 and 3160 */
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->host_interrupt_operation_mode)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_SETBITS(sc, IWM_CSR_INT_COALESCING, IWM_HOST_INT_OPER_MODE);
/*
* Thus sayeth el jefe (iwlwifi) via a comment:
*
* This value should initially be 0 (before preparing any
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
* RBs), should be 8 after preparing the first 8 RBs (for example)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
*/
IWM_WRITE(sc, IWM_FH_RSCSR_CHNL0_WPTR, 8);
iwm_nic_unlock(sc);
return 0;
}
static int
iwm_nic_tx_init(struct iwm_softc *sc)
{
int qid;
if (!iwm_nic_lock(sc))
return EBUSY;
/* Deactivate TX scheduler. */
iwm_write_prph(sc, IWM_SCD_TXFACT, 0);
/* Set physical address of "keep warm" page (16-byte aligned). */
IWM_WRITE(sc, IWM_FH_KW_MEM_ADDR_REG, sc->kw_dma.paddr >> 4);
/* Initialize TX rings. */
for (qid = 0; qid < nitems(sc->txq); qid++) {
struct iwm_tx_ring *txq = &sc->txq[qid];
/* Set physical address of TX ring (256-byte aligned). */
IWM_WRITE(sc, IWM_FH_MEM_CBBC_QUEUE(qid),
txq->desc_dma.paddr >> 8);
IWM_DPRINTF(sc, IWM_DEBUG_XMIT,
"%s: loading ring %d descriptors (%p) at %lx\n",
__func__,
qid, txq->desc,
(unsigned long) (txq->desc_dma.paddr >> 8));
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
iwm_write_prph(sc, IWM_SCD_GP_CTRL, IWM_SCD_GP_CTRL_AUTO_ACTIVE_MODE);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm_nic_unlock(sc);
return 0;
}
static int
iwm_nic_init(struct iwm_softc *sc)
{
int error;
iwm_apm_init(sc);
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->device_family == IWM_DEVICE_FAMILY_7000)
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
iwm_set_pwr(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm_mvm_nic_config(sc);
if ((error = iwm_nic_rx_init(sc)) != 0)
return error;
/*
* Ditto for TX, from iwn
*/
if ((error = iwm_nic_tx_init(sc)) != 0)
return error;
IWM_DPRINTF(sc, IWM_DEBUG_RESET,
"%s: shadow registers enabled\n", __func__);
IWM_SETBITS(sc, IWM_CSR_MAC_SHADOW_REG_CTRL, 0x800fffff);
return 0;
}
[iwm] Factor out firmware station handling into if_iwm_sta.c. * This adds iwm_mvm_rm_sta(), which will be used to tear down firmware state for better/cleaner iwm_newstate() handling. * Makes iwm_enable_txq() and iwm_mvm_flush_tx_path() non-static, add the declarations to if_iwm_util.h for now. Obtained from: dragonflybsd.git 85d1c6190c4c3564b1a347f253e823aa95c202b2
2017-05-12 06:03:23 +00:00
int
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
iwm_enable_txq(struct iwm_softc *sc, int sta_id, int qid, int fifo)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
if (!iwm_nic_lock(sc)) {
device_printf(sc->sc_dev,
"%s: cannot enable txq %d\n",
__func__,
qid);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
return EBUSY;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
IWM_WRITE(sc, IWM_HBUS_TARG_WRPTR, qid << 8 | 0);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
if (qid == IWM_MVM_CMD_QUEUE) {
/* unactivate before configuration */
iwm_write_prph(sc, IWM_SCD_QUEUE_STATUS_BITS(qid),
(0 << IWM_SCD_QUEUE_STTS_REG_POS_ACTIVE)
| (1 << IWM_SCD_QUEUE_STTS_REG_POS_SCD_ACT_EN));
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
iwm_nic_unlock(sc);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
iwm_clear_bits_prph(sc, IWM_SCD_AGGR_SEL, (1 << qid));
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
if (!iwm_nic_lock(sc)) {
device_printf(sc->sc_dev,
"%s: cannot enable txq %d\n", __func__, qid);
return EBUSY;
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
iwm_write_prph(sc, IWM_SCD_QUEUE_RDPTR(qid), 0);
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
iwm_nic_unlock(sc);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
iwm_write_mem32(sc, sc->scd_base_addr + IWM_SCD_CONTEXT_QUEUE_OFFSET(qid), 0);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
/* Set scheduler window size and frame limit. */
iwm_write_mem32(sc,
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
sc->scd_base_addr + IWM_SCD_CONTEXT_QUEUE_OFFSET(qid) +
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
sizeof(uint32_t),
((IWM_FRAME_LIMIT << IWM_SCD_QUEUE_CTX_REG2_WIN_SIZE_POS) &
IWM_SCD_QUEUE_CTX_REG2_WIN_SIZE_MSK) |
((IWM_FRAME_LIMIT << IWM_SCD_QUEUE_CTX_REG2_FRAME_LIMIT_POS) &
IWM_SCD_QUEUE_CTX_REG2_FRAME_LIMIT_MSK));
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
if (!iwm_nic_lock(sc)) {
device_printf(sc->sc_dev,
"%s: cannot enable txq %d\n", __func__, qid);
return EBUSY;
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
iwm_write_prph(sc, IWM_SCD_QUEUE_STATUS_BITS(qid),
(1 << IWM_SCD_QUEUE_STTS_REG_POS_ACTIVE) |
(fifo << IWM_SCD_QUEUE_STTS_REG_POS_TXF) |
(1 << IWM_SCD_QUEUE_STTS_REG_POS_WSL) |
IWM_SCD_QUEUE_STTS_REG_MSK);
} else {
struct iwm_scd_txq_cfg_cmd cmd;
int error;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
iwm_nic_unlock(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
memset(&cmd, 0, sizeof(cmd));
cmd.scd_queue = qid;
cmd.enable = 1;
cmd.sta_id = sta_id;
cmd.tx_fifo = fifo;
cmd.aggregate = 0;
cmd.window = IWM_FRAME_LIMIT;
error = iwm_mvm_send_cmd_pdu(sc, IWM_SCD_QUEUE_CFG, IWM_CMD_SYNC,
sizeof(cmd), &cmd);
if (error) {
device_printf(sc->sc_dev,
"cannot enable txq %d\n", qid);
return error;
}
if (!iwm_nic_lock(sc))
return EBUSY;
}
iwm_write_prph(sc, IWM_SCD_EN_CTRL,
iwm_read_prph(sc, IWM_SCD_EN_CTRL) | qid);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm_nic_unlock(sc);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_XMIT, "%s: enabled txq %d FIFO %d\n",
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
__func__, qid, fifo);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
return 0;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
static int
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
iwm_trans_pcie_fw_alive(struct iwm_softc *sc, uint32_t scd_base_addr)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
int error, chnl;
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
int clear_dwords = (IWM_SCD_TRANS_TBL_MEM_UPPER_BOUND -
IWM_SCD_CONTEXT_MEM_LOWER_BOUND) / sizeof(uint32_t);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (!iwm_nic_lock(sc))
return EBUSY;
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
iwm_ict_reset(sc);
sc->scd_base_addr = iwm_read_prph(sc, IWM_SCD_SRAM_BASE_ADDR);
if (scd_base_addr != 0 &&
scd_base_addr != sc->scd_base_addr) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev,
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
"%s: sched addr mismatch: alive: 0x%x prph: 0x%x\n",
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
__func__, sc->scd_base_addr, scd_base_addr);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] iwm_{read,write}_prph() don't grab the nic lock in iwm themselves. * Fix a couple of cases where the nic lock ended up not being grabbed during an iwm_read_prph() or iwm_write_prph(). Obtained from: dragonflybsd.git 6c5470f2db219c61e362c981fea969d97e1b8293
2017-05-09 02:41:33 +00:00
iwm_nic_unlock(sc);
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
/* reset context data, TX status and translation data */
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
error = iwm_write_mem(sc,
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
sc->scd_base_addr + IWM_SCD_CONTEXT_MEM_LOWER_BOUND,
NULL, clear_dwords);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (error)
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
return EBUSY;
if (!iwm_nic_lock(sc))
return EBUSY;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Set physical address of TX scheduler rings (1KB aligned). */
iwm_write_prph(sc, IWM_SCD_DRAM_BASE_ADDR, sc->sched_dma.paddr >> 10);
iwm_write_prph(sc, IWM_SCD_CHAINEXT_EN, 0);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
iwm_nic_unlock(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* enable command channel */
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
error = iwm_enable_txq(sc, 0 /* unused */, IWM_MVM_CMD_QUEUE, 7);
if (error)
return error;
if (!iwm_nic_lock(sc))
return EBUSY;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm_write_prph(sc, IWM_SCD_TXFACT, 0xff);
/* Enable DMA channels. */
for (chnl = 0; chnl < IWM_FH_TCSR_CHNL_NUM; chnl++) {
IWM_WRITE(sc, IWM_FH_TCSR_CHNL_TX_CONFIG_REG(chnl),
IWM_FH_TCSR_TX_CONFIG_REG_VAL_DMA_CHNL_ENABLE |
IWM_FH_TCSR_TX_CONFIG_REG_VAL_DMA_CREDIT_ENABLE);
}
IWM_SETBITS(sc, IWM_FH_TX_CHICKEN_BITS_REG,
IWM_FH_TX_CHICKEN_BITS_SCD_AUTO_RETRY_EN);
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
iwm_nic_unlock(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Enable L1-Active */
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->device_family != IWM_DEVICE_FAMILY_8000) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
iwm_clear_bits_prph(sc, IWM_APMG_PCIDEV_STT_REG,
IWM_APMG_PCIDEV_STT_VAL_L1_ACT_DIS);
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
return error;
}
/*
* NVM read access and content parsing. We do not support
* external NVM or writing NVM.
* iwlwifi/mvm/nvm.c
*/
/* Default NVM size to read */
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
#define IWM_NVM_DEFAULT_CHUNK_SIZE (2*1024)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
#define IWM_NVM_WRITE_OPCODE 1
#define IWM_NVM_READ_OPCODE 0
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
/* load nvm chunk response */
[iwm] Sync iwm_nvm_read_chunk() function with Linux iwlwifi. This fixes an error handling detail in iwm_nvm_read_chunk(), where an error response from the firmware for an NVM read shouldn't be fatal if the offset was non-zero. Approved by: adrian (mentor) Obtained from: DragonFlyBSD git 250a1c33fca1725121fe499f9cebc90267d209f9 Differential Revision: https://reviews.freebsd.org/D8542
2016-11-17 20:00:20 +00:00
enum {
IWM_READ_NVM_CHUNK_SUCCEED = 0,
IWM_READ_NVM_CHUNK_NOT_VALID_ADDRESS = 1
};
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static int
iwm_nvm_read_chunk(struct iwm_softc *sc, uint16_t section,
uint16_t offset, uint16_t length, uint8_t *data, uint16_t *len)
{
struct iwm_nvm_access_cmd nvm_access_cmd = {
.offset = htole16(offset),
.length = htole16(length),
.type = htole16(section),
.op_code = IWM_NVM_READ_OPCODE,
};
struct iwm_nvm_access_resp *nvm_resp;
struct iwm_rx_packet *pkt;
struct iwm_host_cmd cmd = {
.id = IWM_NVM_ACCESS_CMD,
[iwm] Sync iwm_nvm_read_chunk() function with Linux iwlwifi. This fixes an error handling detail in iwm_nvm_read_chunk(), where an error response from the firmware for an NVM read shouldn't be fatal if the offset was non-zero. Approved by: adrian (mentor) Obtained from: DragonFlyBSD git 250a1c33fca1725121fe499f9cebc90267d209f9 Differential Revision: https://reviews.freebsd.org/D8542
2016-11-17 20:00:20 +00:00
.flags = IWM_CMD_WANT_SKB | IWM_CMD_SEND_IN_RFKILL,
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
.data = { &nvm_access_cmd, },
};
[iwm] Sync iwm_nvm_read_chunk() function with Linux iwlwifi. This fixes an error handling detail in iwm_nvm_read_chunk(), where an error response from the firmware for an NVM read shouldn't be fatal if the offset was non-zero. Approved by: adrian (mentor) Obtained from: DragonFlyBSD git 250a1c33fca1725121fe499f9cebc90267d209f9 Differential Revision: https://reviews.freebsd.org/D8542
2016-11-17 20:00:20 +00:00
int ret, bytes_read, offset_read;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
uint8_t *resp_data;
cmd.len[0] = sizeof(struct iwm_nvm_access_cmd);
ret = iwm_send_cmd(sc, &cmd);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
if (ret) {
device_printf(sc->sc_dev,
"Could not send NVM_ACCESS command (error=%d)\n", ret);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
return ret;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
pkt = cmd.resp_pkt;
/* Extract NVM response */
nvm_resp = (void *)pkt->data;
ret = le16toh(nvm_resp->status);
bytes_read = le16toh(nvm_resp->length);
offset_read = le16toh(nvm_resp->offset);
resp_data = nvm_resp->data;
if (ret) {
[iwm] Sync iwm_nvm_read_chunk() function with Linux iwlwifi. This fixes an error handling detail in iwm_nvm_read_chunk(), where an error response from the firmware for an NVM read shouldn't be fatal if the offset was non-zero. Approved by: adrian (mentor) Obtained from: DragonFlyBSD git 250a1c33fca1725121fe499f9cebc90267d209f9 Differential Revision: https://reviews.freebsd.org/D8542
2016-11-17 20:00:20 +00:00
if ((offset != 0) &&
(ret == IWM_READ_NVM_CHUNK_NOT_VALID_ADDRESS)) {
/*
* meaning of NOT_VALID_ADDRESS:
* driver try to read chunk from address that is
* multiple of 2K and got an error since addr is empty.
* meaning of (offset != 0): driver already
* read valid data from another chunk so this case
* is not an error.
*/
IWM_DPRINTF(sc, IWM_DEBUG_EEPROM | IWM_DEBUG_RESET,
"NVM access command failed on offset 0x%x since that section size is multiple 2K\n",
offset);
*len = 0;
ret = 0;
} else {
IWM_DPRINTF(sc, IWM_DEBUG_EEPROM | IWM_DEBUG_RESET,
"NVM access command failed with status %d\n", ret);
ret = EIO;
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
goto exit;
}
if (offset_read != offset) {
device_printf(sc->sc_dev,
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
"NVM ACCESS response with invalid offset %d\n",
offset_read);
ret = EINVAL;
goto exit;
}
if (bytes_read > length) {
device_printf(sc->sc_dev,
"NVM ACCESS response with too much data "
[iwm] Sync iwm_nvm_read_chunk() function with Linux iwlwifi. This fixes an error handling detail in iwm_nvm_read_chunk(), where an error response from the firmware for an NVM read shouldn't be fatal if the offset was non-zero. Approved by: adrian (mentor) Obtained from: DragonFlyBSD git 250a1c33fca1725121fe499f9cebc90267d209f9 Differential Revision: https://reviews.freebsd.org/D8542
2016-11-17 20:00:20 +00:00
"(%d bytes requested, %d bytes received)\n",
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
length, bytes_read);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
ret = EINVAL;
goto exit;
}
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
/* Write data to NVM */
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
memcpy(data + offset, resp_data, bytes_read);
*len = bytes_read;
exit:
iwm_free_resp(sc, &cmd);
return ret;
}
/*
* Reads an NVM section completely.
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
* NICs prior to 7000 family don't have a real NVM, but just read
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
* section 0 which is the EEPROM. Because the EEPROM reading is unlimited
* by uCode, we need to manually check in this case that we don't
* overflow and try to read more than the EEPROM size.
* For 7000 family NICs, we supply the maximal size we can read, and
* the uCode fills the response with as much data as we can,
* without overflowing, so no check is needed.
*/
static int
iwm_nvm_read_section(struct iwm_softc *sc,
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
uint16_t section, uint8_t *data, uint16_t *len, uint32_t size_read)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
uint16_t seglen, length, offset = 0;
int ret;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
/* Set nvm section read length */
length = IWM_NVM_DEFAULT_CHUNK_SIZE;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
seglen = length;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
/* Read the NVM until exhausted (reading less than requested) */
while (seglen == length) {
/* Check no memory assumptions fail and cause an overflow */
if ((size_read + offset + length) >
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
sc->cfg->eeprom_size) {
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
device_printf(sc->sc_dev,
"EEPROM size is too small for NVM\n");
return ENOBUFS;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
ret = iwm_nvm_read_chunk(sc, section, offset, length, data, &seglen);
if (ret) {
IWM_DPRINTF(sc, IWM_DEBUG_EEPROM | IWM_DEBUG_RESET,
"Cannot read NVM from section %d offset %d, length %d\n",
section, offset, length);
return ret;
}
offset += seglen;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_EEPROM | IWM_DEBUG_RESET,
"NVM section %d read completed\n", section);
*len = offset;
return 0;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
/*
* BEGIN IWM_NVM_PARSE
*/
/* iwlwifi/iwl-nvm-parse.c */
/* NVM offsets (in words) definitions */
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
enum iwm_nvm_offsets {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* NVM HW-Section offset (in words) definitions */
IWM_HW_ADDR = 0x15,
/* NVM SW-Section offset (in words) definitions */
IWM_NVM_SW_SECTION = 0x1C0,
IWM_NVM_VERSION = 0,
IWM_RADIO_CFG = 1,
IWM_SKU = 2,
IWM_N_HW_ADDRS = 3,
IWM_NVM_CHANNELS = 0x1E0 - IWM_NVM_SW_SECTION,
/* NVM calibration section offset (in words) definitions */
IWM_NVM_CALIB_SECTION = 0x2B8,
IWM_XTAL_CALIB = 0x316 - IWM_NVM_CALIB_SECTION
};
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
enum iwm_8000_nvm_offsets {
/* NVM HW-Section offset (in words) definitions */
IWM_HW_ADDR0_WFPM_8000 = 0x12,
IWM_HW_ADDR1_WFPM_8000 = 0x16,
IWM_HW_ADDR0_PCIE_8000 = 0x8A,
IWM_HW_ADDR1_PCIE_8000 = 0x8E,
IWM_MAC_ADDRESS_OVERRIDE_8000 = 1,
/* NVM SW-Section offset (in words) definitions */
IWM_NVM_SW_SECTION_8000 = 0x1C0,
IWM_NVM_VERSION_8000 = 0,
IWM_RADIO_CFG_8000 = 0,
IWM_SKU_8000 = 2,
IWM_N_HW_ADDRS_8000 = 3,
/* NVM REGULATORY -Section offset (in words) definitions */
IWM_NVM_CHANNELS_8000 = 0,
IWM_NVM_LAR_OFFSET_8000_OLD = 0x4C7,
IWM_NVM_LAR_OFFSET_8000 = 0x507,
IWM_NVM_LAR_ENABLED_8000 = 0x7,
/* NVM calibration section offset (in words) definitions */
IWM_NVM_CALIB_SECTION_8000 = 0x2B8,
IWM_XTAL_CALIB_8000 = 0x316 - IWM_NVM_CALIB_SECTION_8000
};
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* SKU Capabilities (actual values from NVM definition) */
enum nvm_sku_bits {
IWM_NVM_SKU_CAP_BAND_24GHZ = (1 << 0),
IWM_NVM_SKU_CAP_BAND_52GHZ = (1 << 1),
IWM_NVM_SKU_CAP_11N_ENABLE = (1 << 2),
IWM_NVM_SKU_CAP_11AC_ENABLE = (1 << 3),
};
/* radio config bits (actual values from NVM definition) */
#define IWM_NVM_RF_CFG_DASH_MSK(x) (x & 0x3) /* bits 0-1 */
#define IWM_NVM_RF_CFG_STEP_MSK(x) ((x >> 2) & 0x3) /* bits 2-3 */
#define IWM_NVM_RF_CFG_TYPE_MSK(x) ((x >> 4) & 0x3) /* bits 4-5 */
#define IWM_NVM_RF_CFG_PNUM_MSK(x) ((x >> 6) & 0x3) /* bits 6-7 */
#define IWM_NVM_RF_CFG_TX_ANT_MSK(x) ((x >> 8) & 0xF) /* bits 8-11 */
#define IWM_NVM_RF_CFG_RX_ANT_MSK(x) ((x >> 12) & 0xF) /* bits 12-15 */
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
#define IWM_NVM_RF_CFG_FLAVOR_MSK_8000(x) (x & 0xF)
#define IWM_NVM_RF_CFG_DASH_MSK_8000(x) ((x >> 4) & 0xF)
#define IWM_NVM_RF_CFG_STEP_MSK_8000(x) ((x >> 8) & 0xF)
#define IWM_NVM_RF_CFG_TYPE_MSK_8000(x) ((x >> 12) & 0xFFF)
#define IWM_NVM_RF_CFG_TX_ANT_MSK_8000(x) ((x >> 24) & 0xF)
#define IWM_NVM_RF_CFG_RX_ANT_MSK_8000(x) ((x >> 28) & 0xF)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/**
* enum iwm_nvm_channel_flags - channel flags in NVM
* @IWM_NVM_CHANNEL_VALID: channel is usable for this SKU/geo
* @IWM_NVM_CHANNEL_IBSS: usable as an IBSS channel
* @IWM_NVM_CHANNEL_ACTIVE: active scanning allowed
* @IWM_NVM_CHANNEL_RADAR: radar detection required
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
* XXX cannot find this (DFS) flag in iwm-nvm-parse.c
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
* @IWM_NVM_CHANNEL_DFS: dynamic freq selection candidate
* @IWM_NVM_CHANNEL_WIDE: 20 MHz channel okay (?)
* @IWM_NVM_CHANNEL_40MHZ: 40 MHz channel okay (?)
* @IWM_NVM_CHANNEL_80MHZ: 80 MHz channel okay (?)
* @IWM_NVM_CHANNEL_160MHZ: 160 MHz channel okay (?)
*/
enum iwm_nvm_channel_flags {
IWM_NVM_CHANNEL_VALID = (1 << 0),
IWM_NVM_CHANNEL_IBSS = (1 << 1),
IWM_NVM_CHANNEL_ACTIVE = (1 << 3),
IWM_NVM_CHANNEL_RADAR = (1 << 4),
IWM_NVM_CHANNEL_DFS = (1 << 7),
IWM_NVM_CHANNEL_WIDE = (1 << 8),
IWM_NVM_CHANNEL_40MHZ = (1 << 9),
IWM_NVM_CHANNEL_80MHZ = (1 << 10),
IWM_NVM_CHANNEL_160MHZ = (1 << 11),
};
/*
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
* Translate EEPROM flags to net80211.
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
*/
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
static uint32_t
iwm_eeprom_channel_flags(uint16_t ch_flags)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
uint32_t nflags;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
nflags = 0;
if ((ch_flags & IWM_NVM_CHANNEL_ACTIVE) == 0)
nflags |= IEEE80211_CHAN_PASSIVE;
if ((ch_flags & IWM_NVM_CHANNEL_IBSS) == 0)
nflags |= IEEE80211_CHAN_NOADHOC;
if (ch_flags & IWM_NVM_CHANNEL_RADAR) {
nflags |= IEEE80211_CHAN_DFS;
/* Just in case. */
nflags |= IEEE80211_CHAN_NOADHOC;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
return (nflags);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
static void
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
iwm_add_channel_band(struct iwm_softc *sc, struct ieee80211_channel chans[],
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
int maxchans, int *nchans, int ch_idx, size_t ch_num,
const uint8_t bands[])
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
const uint16_t * const nvm_ch_flags = sc->nvm_data->nvm_ch_flags;
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
uint32_t nflags;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
uint16_t ch_flags;
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
uint8_t ieee;
int error;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
for (; ch_idx < ch_num; ch_idx++) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
ch_flags = le16_to_cpup(nvm_ch_flags + ch_idx);
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->device_family == IWM_DEVICE_FAMILY_7000)
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
ieee = iwm_nvm_channels[ch_idx];
else
ieee = iwm_nvm_channels_8000[ch_idx];
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (!(ch_flags & IWM_NVM_CHANNEL_VALID)) {
IWM_DPRINTF(sc, IWM_DEBUG_EEPROM,
"Ch. %d Flags %x [%sGHz] - No traffic\n",
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
ieee, ch_flags,
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
(ch_idx >= IWM_NUM_2GHZ_CHANNELS) ?
"5.2" : "2.4");
continue;
}
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
nflags = iwm_eeprom_channel_flags(ch_flags);
error = ieee80211_add_channel(chans, maxchans, nchans,
ieee, 0, 0, nflags, bands);
if (error != 0)
break;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_EEPROM,
"Ch. %d Flags %x [%sGHz] - Added\n",
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
ieee, ch_flags,
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
(ch_idx >= IWM_NUM_2GHZ_CHANNELS) ?
"5.2" : "2.4");
}
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
}
static void
iwm_init_channel_map(struct ieee80211com *ic, int maxchans, int *nchans,
struct ieee80211_channel chans[])
{
struct iwm_softc *sc = ic->ic_softc;
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
struct iwm_nvm_data *data = sc->nvm_data;
Follow-up r298818: hide size of 'bands' array behind a macro.
2016-05-16 02:27:28 +00:00
uint8_t bands[IEEE80211_MODE_BYTES];
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
size_t ch_num;
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
memset(bands, 0, sizeof(bands));
/* 1-13: 11b/g channels. */
setbit(bands, IEEE80211_MODE_11B);
setbit(bands, IEEE80211_MODE_11G);
iwm_add_channel_band(sc, chans, maxchans, nchans, 0,
IWM_NUM_2GHZ_CHANNELS - 1, bands);
/* 14: 11b channel only. */
clrbit(bands, IEEE80211_MODE_11G);
iwm_add_channel_band(sc, chans, maxchans, nchans,
iwm: fix parameters for iwm_add_channel_band() It accepts <first index, max index + 1>, not <first index, number of entries>. Reported by: adrian
2016-05-02 19:07:44 +00:00
IWM_NUM_2GHZ_CHANNELS - 1, IWM_NUM_2GHZ_CHANNELS, bands);
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
if (data->sku_cap_band_52GHz_enable) {
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->device_family == IWM_DEVICE_FAMILY_7000)
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
ch_num = nitems(iwm_nvm_channels);
else
ch_num = nitems(iwm_nvm_channels_8000);
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
memset(bands, 0, sizeof(bands));
setbit(bands, IEEE80211_MODE_11A);
iwm_add_channel_band(sc, chans, maxchans, nchans,
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
IWM_NUM_2GHZ_CHANNELS, ch_num, bands);
}
}
static void
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
iwm_set_hw_address_family_8000(struct iwm_softc *sc, struct iwm_nvm_data *data,
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
const uint16_t *mac_override, const uint16_t *nvm_hw)
{
const uint8_t *hw_addr;
if (mac_override) {
static const uint8_t reserved_mac[] = {
0x02, 0xcc, 0xaa, 0xff, 0xee, 0x00
};
hw_addr = (const uint8_t *)(mac_override +
IWM_MAC_ADDRESS_OVERRIDE_8000);
/*
* Store the MAC address from MAO section.
* No byte swapping is required in MAO section
*/
IEEE80211_ADDR_COPY(data->hw_addr, hw_addr);
/*
* Force the use of the OTP MAC address in case of reserved MAC
* address in the NVM, or if address is given but invalid.
*/
if (!IEEE80211_ADDR_EQ(reserved_mac, hw_addr) &&
!IEEE80211_ADDR_EQ(ieee80211broadcastaddr, data->hw_addr) &&
iwm_is_valid_ether_addr(data->hw_addr) &&
!IEEE80211_IS_MULTICAST(data->hw_addr))
return;
IWM_DPRINTF(sc, IWM_DEBUG_RESET,
"%s: mac address from nvm override section invalid\n",
__func__);
}
if (nvm_hw) {
/* read the mac address from WFMP registers */
uint32_t mac_addr0 =
htole32(iwm_read_prph(sc, IWM_WFMP_MAC_ADDR_0));
uint32_t mac_addr1 =
htole32(iwm_read_prph(sc, IWM_WFMP_MAC_ADDR_1));
hw_addr = (const uint8_t *)&mac_addr0;
data->hw_addr[0] = hw_addr[3];
data->hw_addr[1] = hw_addr[2];
data->hw_addr[2] = hw_addr[1];
data->hw_addr[3] = hw_addr[0];
hw_addr = (const uint8_t *)&mac_addr1;
data->hw_addr[4] = hw_addr[1];
data->hw_addr[5] = hw_addr[0];
return;
}
device_printf(sc->sc_dev, "%s: mac address not found\n", __func__);
memset(data->hw_addr, 0, sizeof(data->hw_addr));
}
static int
iwm_get_sku(const struct iwm_softc *sc, const uint16_t *nvm_sw,
const uint16_t *phy_sku)
{
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->device_family != IWM_DEVICE_FAMILY_8000)
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
return le16_to_cpup(nvm_sw + IWM_SKU);
return le32_to_cpup((const uint32_t *)(phy_sku + IWM_SKU_8000));
}
static int
iwm_get_nvm_version(const struct iwm_softc *sc, const uint16_t *nvm_sw)
{
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->device_family != IWM_DEVICE_FAMILY_8000)
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
return le16_to_cpup(nvm_sw + IWM_NVM_VERSION);
else
return le32_to_cpup((const uint32_t *)(nvm_sw +
IWM_NVM_VERSION_8000));
}
static int
iwm_get_radio_cfg(const struct iwm_softc *sc, const uint16_t *nvm_sw,
const uint16_t *phy_sku)
{
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->device_family != IWM_DEVICE_FAMILY_8000)
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
return le16_to_cpup(nvm_sw + IWM_RADIO_CFG);
return le32_to_cpup((const uint32_t *)(phy_sku + IWM_RADIO_CFG_8000));
}
static int
iwm_get_n_hw_addrs(const struct iwm_softc *sc, const uint16_t *nvm_sw)
{
int n_hw_addr;
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->device_family != IWM_DEVICE_FAMILY_8000)
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
return le16_to_cpup(nvm_sw + IWM_N_HW_ADDRS);
n_hw_addr = le32_to_cpup((const uint32_t *)(nvm_sw + IWM_N_HW_ADDRS_8000));
return n_hw_addr & IWM_N_HW_ADDR_MASK;
}
static void
iwm_set_radio_cfg(const struct iwm_softc *sc, struct iwm_nvm_data *data,
uint32_t radio_cfg)
{
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->device_family != IWM_DEVICE_FAMILY_8000) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
data->radio_cfg_type = IWM_NVM_RF_CFG_TYPE_MSK(radio_cfg);
data->radio_cfg_step = IWM_NVM_RF_CFG_STEP_MSK(radio_cfg);
data->radio_cfg_dash = IWM_NVM_RF_CFG_DASH_MSK(radio_cfg);
data->radio_cfg_pnum = IWM_NVM_RF_CFG_PNUM_MSK(radio_cfg);
return;
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
/* set the radio configuration for family 8000 */
data->radio_cfg_type = IWM_NVM_RF_CFG_TYPE_MSK_8000(radio_cfg);
data->radio_cfg_step = IWM_NVM_RF_CFG_STEP_MSK_8000(radio_cfg);
data->radio_cfg_dash = IWM_NVM_RF_CFG_DASH_MSK_8000(radio_cfg);
data->radio_cfg_pnum = IWM_NVM_RF_CFG_FLAVOR_MSK_8000(radio_cfg);
data->valid_tx_ant = IWM_NVM_RF_CFG_TX_ANT_MSK_8000(radio_cfg);
data->valid_rx_ant = IWM_NVM_RF_CFG_RX_ANT_MSK_8000(radio_cfg);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
static int
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
iwm_set_hw_address(struct iwm_softc *sc, struct iwm_nvm_data *data,
const uint16_t *nvm_hw, const uint16_t *mac_override)
{
#ifdef notyet /* for FAMILY 9000 */
if (cfg->mac_addr_from_csr) {
iwm_set_hw_address_from_csr(sc, data);
} else
#endif
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->device_family != IWM_DEVICE_FAMILY_8000) {
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
const uint8_t *hw_addr = (const uint8_t *)(nvm_hw + IWM_HW_ADDR);
/* The byte order is little endian 16 bit, meaning 214365 */
data->hw_addr[0] = hw_addr[1];
data->hw_addr[1] = hw_addr[0];
data->hw_addr[2] = hw_addr[3];
data->hw_addr[3] = hw_addr[2];
data->hw_addr[4] = hw_addr[5];
data->hw_addr[5] = hw_addr[4];
} else {
iwm_set_hw_address_family_8000(sc, data, mac_override, nvm_hw);
}
if (!iwm_is_valid_ether_addr(data->hw_addr)) {
device_printf(sc->sc_dev, "no valid mac address was found\n");
return EINVAL;
}
return 0;
}
static struct iwm_nvm_data *
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm_parse_nvm_data(struct iwm_softc *sc,
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
const uint16_t *nvm_hw, const uint16_t *nvm_sw,
const uint16_t *nvm_calib, const uint16_t *mac_override,
const uint16_t *phy_sku, const uint16_t *regulatory)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
struct iwm_nvm_data *data;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
uint32_t sku, radio_cfg;
[iwm] Check for lar_disable tunable, and lar_enabled flag from NVM. * LAR can be disabled with the hw.iwm.lar.disable tunable now. * On Family 8000 devices we need to check the lar_enabled flag from nvm_data in addition to the TLV_CAPA_LAR_SUPPORT flag from the firmware. * Add a separate IWM_DEBUG_LAR debugging flag. Obtained from: dragonflybsd.git 0593e39cb295aa996ecf789ed4990c3b255f1770
2017-06-04 21:10:14 +00:00
uint16_t lar_config;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->device_family != IWM_DEVICE_FAMILY_8000) {
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
data = malloc(sizeof(*data) +
IWM_NUM_CHANNELS * sizeof(uint16_t),
M_DEVBUF, M_NOWAIT | M_ZERO);
} else {
data = malloc(sizeof(*data) +
IWM_NUM_CHANNELS_8000 * sizeof(uint16_t),
M_DEVBUF, M_NOWAIT | M_ZERO);
}
if (!data)
return NULL;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
data->nvm_version = iwm_get_nvm_version(sc, nvm_sw);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
radio_cfg = iwm_get_radio_cfg(sc, nvm_sw, phy_sku);
iwm_set_radio_cfg(sc, data, radio_cfg);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
sku = iwm_get_sku(sc, nvm_sw, phy_sku);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
data->sku_cap_band_24GHz_enable = sku & IWM_NVM_SKU_CAP_BAND_24GHZ;
data->sku_cap_band_52GHz_enable = sku & IWM_NVM_SKU_CAP_BAND_52GHZ;
data->sku_cap_11n_enable = 0;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
data->n_hw_addrs = iwm_get_n_hw_addrs(sc, nvm_sw);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Check for lar_disable tunable, and lar_enabled flag from NVM. * LAR can be disabled with the hw.iwm.lar.disable tunable now. * On Family 8000 devices we need to check the lar_enabled flag from nvm_data in addition to the TLV_CAPA_LAR_SUPPORT flag from the firmware. * Add a separate IWM_DEBUG_LAR debugging flag. Obtained from: dragonflybsd.git 0593e39cb295aa996ecf789ed4990c3b255f1770
2017-06-04 21:10:14 +00:00
if (sc->cfg->device_family == IWM_DEVICE_FAMILY_8000) {
uint16_t lar_offset = data->nvm_version < 0xE39 ?
IWM_NVM_LAR_OFFSET_8000_OLD :
IWM_NVM_LAR_OFFSET_8000;
lar_config = le16_to_cpup(regulatory + lar_offset);
data->lar_enabled = !!(lar_config &
IWM_NVM_LAR_ENABLED_8000);
}
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
/* If no valid mac address was found - bail out */
if (iwm_set_hw_address(sc, data, nvm_hw, mac_override)) {
free(data, M_DEVBUF);
return NULL;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->device_family == IWM_DEVICE_FAMILY_7000) {
iwm(4): use correct channel list source for Intel 3168 Intel 3168 uses another EEPROM section to store channel flags; port missing bits from iwlwifi to make it work. PR: 230750, 236235 Tested by: Bert JW Regeer <xistence@0x58.com> MFC after: 3 days
2019-03-11 08:30:29 +00:00
memcpy(data->nvm_ch_flags, sc->cfg->nvm_type == IWM_NVM_SDP ?
&regulatory[0] : &nvm_sw[IWM_NVM_CHANNELS],
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
IWM_NUM_CHANNELS * sizeof(uint16_t));
} else {
memcpy(data->nvm_ch_flags, &regulatory[IWM_NVM_CHANNELS_8000],
IWM_NUM_CHANNELS_8000 * sizeof(uint16_t));
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
return data;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
static void
iwm_free_nvm_data(struct iwm_nvm_data *data)
{
if (data != NULL)
free(data, M_DEVBUF);
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
static struct iwm_nvm_data *
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm_parse_nvm_sections(struct iwm_softc *sc, struct iwm_nvm_section *sections)
{
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
const uint16_t *hw, *sw, *calib, *regulatory, *mac_override, *phy_sku;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Checking for required sections */
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->device_family == IWM_DEVICE_FAMILY_7000) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
if (!sections[IWM_NVM_SECTION_TYPE_SW].data ||
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
!sections[sc->cfg->nvm_hw_section_num].data) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
device_printf(sc->sc_dev,
"Can't parse empty OTP/NVM sections\n");
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
return NULL;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
} else if (sc->cfg->device_family == IWM_DEVICE_FAMILY_8000) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
/* SW and REGULATORY sections are mandatory */
if (!sections[IWM_NVM_SECTION_TYPE_SW].data ||
!sections[IWM_NVM_SECTION_TYPE_REGULATORY].data) {
device_printf(sc->sc_dev,
"Can't parse empty OTP/NVM sections\n");
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
return NULL;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
/* MAC_OVERRIDE or at least HW section must exist */
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (!sections[sc->cfg->nvm_hw_section_num].data &&
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
!sections[IWM_NVM_SECTION_TYPE_MAC_OVERRIDE].data) {
device_printf(sc->sc_dev,
"Can't parse mac_address, empty sections\n");
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
return NULL;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
/* PHY_SKU section is mandatory in B0 */
if (!sections[IWM_NVM_SECTION_TYPE_PHY_SKU].data) {
device_printf(sc->sc_dev,
"Can't parse phy_sku in B0, empty sections\n");
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
return NULL;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
} else {
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
panic("unknown device family %d\n", sc->cfg->device_family);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
hw = (const uint16_t *) sections[sc->cfg->nvm_hw_section_num].data;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
sw = (const uint16_t *)sections[IWM_NVM_SECTION_TYPE_SW].data;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
calib = (const uint16_t *)
sections[IWM_NVM_SECTION_TYPE_CALIBRATION].data;
iwm(4): use correct channel list source for Intel 3168 Intel 3168 uses another EEPROM section to store channel flags; port missing bits from iwlwifi to make it work. PR: 230750, 236235 Tested by: Bert JW Regeer <xistence@0x58.com> MFC after: 3 days
2019-03-11 08:30:29 +00:00
regulatory = sc->cfg->nvm_type == IWM_NVM_SDP ?
(const uint16_t *)sections[IWM_NVM_SECTION_TYPE_REGULATORY_SDP].data :
(const uint16_t *)sections[IWM_NVM_SECTION_TYPE_REGULATORY].data;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
mac_override = (const uint16_t *)
sections[IWM_NVM_SECTION_TYPE_MAC_OVERRIDE].data;
phy_sku = (const uint16_t *)sections[IWM_NVM_SECTION_TYPE_PHY_SKU].data;
return iwm_parse_nvm_data(sc, hw, sw, calib, mac_override,
phy_sku, regulatory);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
static int
iwm_nvm_init(struct iwm_softc *sc)
{
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
struct iwm_nvm_section nvm_sections[IWM_NVM_MAX_NUM_SECTIONS];
int i, ret, section;
uint32_t size_read = 0;
uint8_t *nvm_buffer, *temp;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
uint16_t len;
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
memset(nvm_sections, 0, sizeof(nvm_sections));
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->nvm_hw_section_num >= IWM_NVM_MAX_NUM_SECTIONS)
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
return EINVAL;
[iwm] free RX ring / NVM memory after they're used. * Free RX ring during detach * Free NVM memory after parsing Tested: * 7260, STA mode Submitted by: Imre Vadasz <imre@vdsz.com> Approved by: re (gjb) Obtained from: dragonflybsd Differential Revision: https://reviews.freebsd.org/D6817
2016-06-16 17:59:15 +00:00
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
/* load NVM values from nic */
/* Read From FW NVM */
IWM_DPRINTF(sc, IWM_DEBUG_EEPROM, "Read from NVM\n");
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
nvm_buffer = malloc(sc->cfg->eeprom_size, M_DEVBUF, M_NOWAIT | M_ZERO);
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
if (!nvm_buffer)
return ENOMEM;
for (section = 0; section < IWM_NVM_MAX_NUM_SECTIONS; section++) {
/* we override the constness for initial read */
ret = iwm_nvm_read_section(sc, section, nvm_buffer,
&len, size_read);
if (ret)
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
continue;
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
size_read += len;
temp = malloc(len, M_DEVBUF, M_NOWAIT);
if (!temp) {
ret = ENOMEM;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
break;
}
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
memcpy(temp, nvm_buffer, len);
nvm_sections[section].data = temp;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
nvm_sections[section].length = len;
}
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
if (!size_read)
device_printf(sc->sc_dev, "OTP is blank\n");
free(nvm_buffer, M_DEVBUF);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
sc->nvm_data = iwm_parse_nvm_sections(sc, nvm_sections);
if (!sc->nvm_data)
return EINVAL;
IWM_DPRINTF(sc, IWM_DEBUG_EEPROM | IWM_DEBUG_RESET,
"nvm version = %x\n", sc->nvm_data->nvm_version);
for (i = 0; i < IWM_NVM_MAX_NUM_SECTIONS; i++) {
[iwm] free RX ring / NVM memory after they're used. * Free RX ring during detach * Free NVM memory after parsing Tested: * 7260, STA mode Submitted by: Imre Vadasz <imre@vdsz.com> Approved by: re (gjb) Obtained from: dragonflybsd Differential Revision: https://reviews.freebsd.org/D6817
2016-06-16 17:59:15 +00:00
if (nvm_sections[i].data != NULL)
free(nvm_sections[i].data, M_DEVBUF);
}
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
return 0;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
static int
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
iwm_pcie_load_section(struct iwm_softc *sc, uint8_t section_num,
const struct iwm_fw_desc *section)
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
{
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
struct iwm_dma_info *dma = &sc->fw_dma;
uint8_t *v_addr;
bus_addr_t p_addr;
uint32_t offset, chunk_sz = MIN(IWM_FH_MEM_TB_MAX_LENGTH, section->len);
int ret = 0;
IWM_DPRINTF(sc, IWM_DEBUG_RESET,
"%s: [%d] uCode section being loaded...\n",
__func__, section_num);
v_addr = dma->vaddr;
p_addr = dma->paddr;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
for (offset = 0; offset < section->len; offset += chunk_sz) {
uint32_t copy_size, dst_addr;
int extended_addr = FALSE;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
copy_size = MIN(chunk_sz, section->len - offset);
dst_addr = section->offset + offset;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
if (dst_addr >= IWM_FW_MEM_EXTENDED_START &&
dst_addr <= IWM_FW_MEM_EXTENDED_END)
extended_addr = TRUE;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
if (extended_addr)
iwm_set_bits_prph(sc, IWM_LMPM_CHICK,
IWM_LMPM_CHICK_EXTENDED_ADDR_SPACE);
memcpy(v_addr, (const uint8_t *)section->data + offset,
copy_size);
bus_dmamap_sync(dma->tag, dma->map, BUS_DMASYNC_PREWRITE);
ret = iwm_pcie_load_firmware_chunk(sc, dst_addr, p_addr,
copy_size);
if (extended_addr)
iwm_clear_bits_prph(sc, IWM_LMPM_CHICK,
IWM_LMPM_CHICK_EXTENDED_ADDR_SPACE);
if (ret) {
device_printf(sc->sc_dev,
"%s: Could not load the [%d] uCode section\n",
__func__, section_num);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
break;
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
return ret;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
/*
* ucode
*/
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
static int
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
iwm_pcie_load_firmware_chunk(struct iwm_softc *sc, uint32_t dst_addr,
bus_addr_t phy_addr, uint32_t byte_cnt)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
sc->sc_fw_chunk_done = 0;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
if (!iwm_nic_lock(sc))
return EBUSY;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_WRITE(sc, IWM_FH_TCSR_CHNL_TX_CONFIG_REG(IWM_FH_SRVC_CHNL),
IWM_FH_TCSR_TX_CONFIG_REG_VAL_DMA_CHNL_PAUSE);
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_WRITE(sc, IWM_FH_SRVC_CHNL_SRAM_ADDR_REG(IWM_FH_SRVC_CHNL),
dst_addr);
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_WRITE(sc, IWM_FH_TFDIB_CTRL0_REG(IWM_FH_SRVC_CHNL),
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
phy_addr & IWM_FH_MEM_TFDIB_DRAM_ADDR_LSB_MSK);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_WRITE(sc, IWM_FH_TFDIB_CTRL1_REG(IWM_FH_SRVC_CHNL),
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
(iwm_get_dma_hi_addr(phy_addr)
<< IWM_FH_MEM_TFDIB_REG1_ADDR_BITSHIFT) | byte_cnt);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_WRITE(sc, IWM_FH_TCSR_CHNL_TX_BUF_STS_REG(IWM_FH_SRVC_CHNL),
1 << IWM_FH_TCSR_CHNL_TX_BUF_STS_REG_POS_TB_NUM |
1 << IWM_FH_TCSR_CHNL_TX_BUF_STS_REG_POS_TB_IDX |
IWM_FH_TCSR_CHNL_TX_BUF_STS_REG_VAL_TFDB_VALID);
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_WRITE(sc, IWM_FH_TCSR_CHNL_TX_CONFIG_REG(IWM_FH_SRVC_CHNL),
IWM_FH_TCSR_TX_CONFIG_REG_VAL_DMA_CHNL_ENABLE |
IWM_FH_TCSR_TX_CONFIG_REG_VAL_DMA_CREDIT_DISABLE |
IWM_FH_TCSR_TX_CONFIG_REG_VAL_CIRQ_HOST_ENDTFD);
iwm_nic_unlock(sc);
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
/* wait up to 5s for this segment to load */
Fix logic errors in iwm_pcie_load_firmware_chunk introduced in r314065. * There's no reason to have a while() loop here, because: - if msleep returns 0, that means we were woken up by the interrupt handler, and we are going to exit immediately as sc_fw_chunk_done will now be 1 (there is nothing else that sleeps on sc_fw.) - if msleep doesn't return 0 (i.e. it returned ETIMEDOUT) then we will exit immediately because of the if-test. So, just use a single msleep() and then check sc_fw_chunk_done as before. * The comment said we were sleeping for 5 seconds, but the msleep was only for 1. Before r314065, this was 1 second and so was the comment, and in that commit the comment was changed and the function call wasn't. Possibly fixes failures to initialize uCode on certain devices. Submitted by: Augustin Cavalier (waddlesplash gmail.com) Obtained from: Haiku 132990ecdcb072f2ce597b5d497ff3e5b1f09c20 MFC after: 10 days
2019-01-26 17:52:12 +00:00
msleep(&sc->sc_fw, &sc->sc_mtx, 0, "iwmfw", hz * 5);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
Fix logic errors in iwm_pcie_load_firmware_chunk introduced in r314065. * There's no reason to have a while() loop here, because: - if msleep returns 0, that means we were woken up by the interrupt handler, and we are going to exit immediately as sc_fw_chunk_done will now be 1 (there is nothing else that sleeps on sc_fw.) - if msleep doesn't return 0 (i.e. it returned ETIMEDOUT) then we will exit immediately because of the if-test. So, just use a single msleep() and then check sc_fw_chunk_done as before. * The comment said we were sleeping for 5 seconds, but the msleep was only for 1. Before r314065, this was 1 second and so was the comment, and in that commit the comment was changed and the function call wasn't. Possibly fixes failures to initialize uCode on certain devices. Submitted by: Augustin Cavalier (waddlesplash gmail.com) Obtained from: Haiku 132990ecdcb072f2ce597b5d497ff3e5b1f09c20 MFC after: 10 days
2019-01-26 17:52:12 +00:00
if (!sc->sc_fw_chunk_done) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
device_printf(sc->sc_dev,
"fw chunk addr 0x%x len %d failed to load\n",
dst_addr, byte_cnt);
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
return ETIMEDOUT;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
return 0;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
static int
iwm_pcie_load_cpu_sections_8000(struct iwm_softc *sc,
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
const struct iwm_fw_img *image, int cpu, int *first_ucode_section)
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
{
int shift_param;
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
int i, ret = 0, sec_num = 0x1;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
uint32_t val, last_read_idx = 0;
if (cpu == 1) {
shift_param = 0;
*first_ucode_section = 0;
} else {
shift_param = 16;
(*first_ucode_section)++;
}
[iwm] Get rid of some gratuitous constant renaming wrt. Linux iwlwifi. * IWM_UCODE_SECT_MAX -> IWM_UCODE_SECTION_MAX * IWM_UCODE_TYPE_* -> IWM_UCODE_* (except for IWM_UCODE_TYPE_MAX which stays). Obtained from: DragonflyBSD commit ff4d1fc3ed002c9fb362423da6c45d711b65658a
2017-02-06 05:07:15 +00:00
for (i = *first_ucode_section; i < IWM_UCODE_SECTION_MAX; i++) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
last_read_idx = i;
/*
* CPU1_CPU2_SEPARATOR_SECTION delimiter - separate between
* CPU1 to CPU2.
* PAGING_SEPARATOR_SECTION delimiter - separate between
* CPU2 non paged to CPU2 paging sec.
*/
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
if (!image->sec[i].data ||
image->sec[i].offset == IWM_CPU1_CPU2_SEPARATOR_SECTION ||
image->sec[i].offset == IWM_PAGING_SEPARATOR_SECTION) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_RESET,
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
"Break since Data not valid or Empty section, sec = %d\n",
i);
break;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
ret = iwm_pcie_load_section(sc, i, &image->sec[i]);
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
if (ret)
return ret;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
/* Notify the ucode of the loaded section number and status */
if (iwm_nic_lock(sc)) {
val = IWM_READ(sc, IWM_FH_UCODE_LOAD_STATUS);
val = val | (sec_num << shift_param);
IWM_WRITE(sc, IWM_FH_UCODE_LOAD_STATUS, val);
sec_num = (sec_num << 1) | 0x1;
iwm_nic_unlock(sc);
}
}
*first_ucode_section = last_read_idx;
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
iwm_enable_interrupts(sc);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
if (iwm_nic_lock(sc)) {
if (cpu == 1)
IWM_WRITE(sc, IWM_FH_UCODE_LOAD_STATUS, 0xFFFF);
else
IWM_WRITE(sc, IWM_FH_UCODE_LOAD_STATUS, 0xFFFFFFFF);
iwm_nic_unlock(sc);
}
return 0;
}
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
static int
iwm_pcie_load_cpu_sections(struct iwm_softc *sc,
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
const struct iwm_fw_img *image, int cpu, int *first_ucode_section)
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
{
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
int shift_param;
int i, ret = 0;
uint32_t last_read_idx = 0;
if (cpu == 1) {
shift_param = 0;
*first_ucode_section = 0;
} else {
shift_param = 16;
(*first_ucode_section)++;
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
for (i = *first_ucode_section; i < IWM_UCODE_SECTION_MAX; i++) {
last_read_idx = i;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
/*
* CPU1_CPU2_SEPARATOR_SECTION delimiter - separate between
* CPU1 to CPU2.
* PAGING_SEPARATOR_SECTION delimiter - separate between
* CPU2 non paged to CPU2 paging sec.
*/
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
if (!image->sec[i].data ||
image->sec[i].offset == IWM_CPU1_CPU2_SEPARATOR_SECTION ||
image->sec[i].offset == IWM_PAGING_SEPARATOR_SECTION) {
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_RESET,
"Break since Data not valid or Empty section, sec = %d\n",
i);
break;
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
ret = iwm_pcie_load_section(sc, i, &image->sec[i]);
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
if (ret)
return ret;
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
*first_ucode_section = last_read_idx;
return 0;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static int
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
iwm_pcie_load_given_ucode(struct iwm_softc *sc, const struct iwm_fw_img *image)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
int ret = 0;
int first_ucode_section;
IWM_DPRINTF(sc, IWM_DEBUG_RESET, "working with %s CPU\n",
image->is_dual_cpus ? "Dual" : "Single");
/* load to FW the binary non secured sections of CPU1 */
ret = iwm_pcie_load_cpu_sections(sc, image, 1, &first_ucode_section);
if (ret)
return ret;
if (image->is_dual_cpus) {
/* set CPU2 header address */
[iwm] iwm_{read,write}_prph() don't grab the nic lock in iwm themselves. * Fix a couple of cases where the nic lock ended up not being grabbed during an iwm_read_prph() or iwm_write_prph(). Obtained from: dragonflybsd.git 6c5470f2db219c61e362c981fea969d97e1b8293
2017-05-09 02:41:33 +00:00
if (iwm_nic_lock(sc)) {
iwm_write_prph(sc,
IWM_LMPM_SECURE_UCODE_LOAD_CPU2_HDR_ADDR,
IWM_LMPM_SECURE_CPU2_HDR_MEM_SPACE);
[iwm] Add iwm_nic_unlock() calls missing from previous commit. Obtained from: dragonflybsd.git f88ab372284e63c4c13da93e9026a203b9b4cdc5
2017-05-09 02:42:53 +00:00
iwm_nic_unlock(sc);
[iwm] iwm_{read,write}_prph() don't grab the nic lock in iwm themselves. * Fix a couple of cases where the nic lock ended up not being grabbed during an iwm_read_prph() or iwm_write_prph(). Obtained from: dragonflybsd.git 6c5470f2db219c61e362c981fea969d97e1b8293
2017-05-09 02:41:33 +00:00
}
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
/* load to FW the binary sections of CPU2 */
ret = iwm_pcie_load_cpu_sections(sc, image, 2,
&first_ucode_section);
if (ret)
return ret;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
iwm_enable_interrupts(sc);
/* release CPU reset */
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_WRITE(sc, IWM_CSR_RESET, 0);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
return 0;
}
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
int
iwm_pcie_load_given_ucode_8000(struct iwm_softc *sc,
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
const struct iwm_fw_img *image)
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
{
int ret = 0;
int first_ucode_section;
IWM_DPRINTF(sc, IWM_DEBUG_RESET, "working with %s CPU\n",
image->is_dual_cpus ? "Dual" : "Single");
/* configure the ucode to be ready to get the secured image */
/* release CPU reset */
[iwm] iwm_{read,write}_prph() don't grab the nic lock in iwm themselves. * Fix a couple of cases where the nic lock ended up not being grabbed during an iwm_read_prph() or iwm_write_prph(). Obtained from: dragonflybsd.git 6c5470f2db219c61e362c981fea969d97e1b8293
2017-05-09 02:41:33 +00:00
if (iwm_nic_lock(sc)) {
iwm_write_prph(sc, IWM_RELEASE_CPU_RESET,
IWM_RELEASE_CPU_RESET_BIT);
[iwm] Add iwm_nic_unlock() calls missing from previous commit. Obtained from: dragonflybsd.git f88ab372284e63c4c13da93e9026a203b9b4cdc5
2017-05-09 02:42:53 +00:00
iwm_nic_unlock(sc);
[iwm] iwm_{read,write}_prph() don't grab the nic lock in iwm themselves. * Fix a couple of cases where the nic lock ended up not being grabbed during an iwm_read_prph() or iwm_write_prph(). Obtained from: dragonflybsd.git 6c5470f2db219c61e362c981fea969d97e1b8293
2017-05-09 02:41:33 +00:00
}
[iwm] Synchronize firmware loading code with Linux iwlwifi. * While there, rename some functions to match the names and functionality of the similarly named functions in Linux iwlwifi. Obtained from: dragonflybsd.git e98ee77a816bfd8b4912047b93dfb2c560788f24
2017-02-22 02:18:04 +00:00
/* load to FW the binary Secured sections of CPU1 */
ret = iwm_pcie_load_cpu_sections_8000(sc, image, 1,
&first_ucode_section);
if (ret)
return ret;
/* load to FW the binary sections of CPU2 */
return iwm_pcie_load_cpu_sections_8000(sc, image, 2,
&first_ucode_section);
}
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
/* XXX Get rid of this definition */
static inline void
iwm_enable_fw_load_int(struct iwm_softc *sc)
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
{
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_INTR, "Enabling FW load interrupt\n");
sc->sc_intmask = IWM_CSR_INT_BIT_FH_TX;
IWM_WRITE(sc, IWM_CSR_INT_MASK, sc->sc_intmask);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
/* XXX Add proper rfkill support code */
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static int
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
iwm_start_fw(struct iwm_softc *sc, const struct iwm_fw_img *fw)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
int ret;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
/* This may fail if AMT took ownership of the device */
if (iwm_prepare_card_hw(sc)) {
device_printf(sc->sc_dev,
"%s: Exit HW not ready\n", __func__);
ret = EIO;
goto out;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
IWM_WRITE(sc, IWM_CSR_INT, 0xFFFFFFFF);
iwm_disable_interrupts(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* make sure rfkill handshake bits are cleared */
IWM_WRITE(sc, IWM_CSR_UCODE_DRV_GP1_CLR, IWM_CSR_UCODE_SW_BIT_RFKILL);
IWM_WRITE(sc, IWM_CSR_UCODE_DRV_GP1_CLR,
IWM_CSR_UCODE_DRV_GP1_BIT_CMD_BLOCKED);
/* clear (again), then enable host interrupts */
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
IWM_WRITE(sc, IWM_CSR_INT, 0xFFFFFFFF);
ret = iwm_nic_init(sc);
if (ret) {
device_printf(sc->sc_dev, "%s: Unable to init nic\n", __func__);
goto out;
}
/*
* Now, we load the firmware and don't want to be interrupted, even
* by the RF-Kill interrupt (hence mask all the interrupt besides the
* FH_TX interrupt which is needed to load the firmware). If the
* RF-Kill switch is toggled, we will find out after having loaded
* the firmware and return the proper value to the caller.
*/
iwm_enable_fw_load_int(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* really make sure rfkill handshake bits are cleared */
/* maybe we should write a few times more? just to make sure */
IWM_WRITE(sc, IWM_CSR_UCODE_DRV_GP1_CLR, IWM_CSR_UCODE_SW_BIT_RFKILL);
IWM_WRITE(sc, IWM_CSR_UCODE_DRV_GP1_CLR, IWM_CSR_UCODE_SW_BIT_RFKILL);
/* Load the given image to the HW */
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
if (sc->cfg->device_family == IWM_DEVICE_FAMILY_8000)
ret = iwm_pcie_load_given_ucode_8000(sc, fw);
else
ret = iwm_pcie_load_given_ucode(sc, fw);
/* XXX re-check RF-Kill state */
out:
return ret;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
static int
iwm_send_tx_ant_cfg(struct iwm_softc *sc, uint8_t valid_tx_ant)
{
struct iwm_tx_ant_cfg_cmd tx_ant_cmd = {
.valid = htole32(valid_tx_ant),
};
return iwm_mvm_send_cmd_pdu(sc, IWM_TX_ANT_CONFIGURATION_CMD,
IWM_CMD_SYNC, sizeof(tx_ant_cmd), &tx_ant_cmd);
}
/* iwlwifi: mvm/fw.c */
static int
iwm_send_phy_cfg_cmd(struct iwm_softc *sc)
{
struct iwm_phy_cfg_cmd phy_cfg_cmd;
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
enum iwm_ucode_type ucode_type = sc->cur_ucode;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Set parameters */
[iwm] Sync valid_tx_ant and valid_rx_ant mask handling with iwlwifi. * This fixes the phy_cfg field sent in the iwm_send_phy_cfg_cmd() command, which wasn't taking into account the valid_rx_ant and valid_tx_ant masks from nvm_data before. Tested: * 7260, STA mode, 2G and 5G Obtained from: DragonflyBSD commit cbb82693c18fd71b4eb86855b82d03995f352d65
2017-02-06 05:03:41 +00:00
phy_cfg_cmd.phy_cfg = htole32(iwm_mvm_get_phy_config(sc));
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
phy_cfg_cmd.calib_control.event_trigger =
sc->sc_default_calib[ucode_type].event_trigger;
phy_cfg_cmd.calib_control.flow_trigger =
sc->sc_default_calib[ucode_type].flow_trigger;
IWM_DPRINTF(sc, IWM_DEBUG_CMD | IWM_DEBUG_RESET,
"Sending Phy CFG command: 0x%x\n", phy_cfg_cmd.phy_cfg);
return iwm_mvm_send_cmd_pdu(sc, IWM_PHY_CONFIGURATION_CMD, IWM_CMD_SYNC,
sizeof(phy_cfg_cmd), &phy_cfg_cmd);
}
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
static int
iwm_alive_fn(struct iwm_softc *sc, struct iwm_rx_packet *pkt, void *data)
{
struct iwm_mvm_alive_data *alive_data = data;
iwm - Update alive response handling, add v4 and remove old versions. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (3820e2bf3331ced3541d1811a38c5a5136dfab93)
2019-01-24 03:49:35 +00:00
struct iwm_mvm_alive_resp_v3 *palive3;
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
struct iwm_mvm_alive_resp *palive;
iwm - Update alive response handling, add v4 and remove old versions. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (3820e2bf3331ced3541d1811a38c5a5136dfab93)
2019-01-24 03:49:35 +00:00
struct iwm_umac_alive *umac;
struct iwm_lmac_alive *lmac1;
struct iwm_lmac_alive *lmac2 = NULL;
uint16_t status;
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
iwm - Update alive response handling, add v4 and remove old versions. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (3820e2bf3331ced3541d1811a38c5a5136dfab93)
2019-01-24 03:49:35 +00:00
if (iwm_rx_packet_payload_len(pkt) == sizeof(*palive)) {
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
palive = (void *)pkt->data;
iwm - Update alive response handling, add v4 and remove old versions. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (3820e2bf3331ced3541d1811a38c5a5136dfab93)
2019-01-24 03:49:35 +00:00
umac = &palive->umac_data;
lmac1 = &palive->lmac_data[0];
lmac2 = &palive->lmac_data[1];
status = le16toh(palive->status);
} else {
palive3 = (void *)pkt->data;
umac = &palive3->umac_data;
lmac1 = &palive3->lmac_data;
status = le16toh(palive3->status);
}
sc->error_event_table[0] = le32toh(lmac1->error_event_table_ptr);
if (lmac2)
sc->error_event_table[1] =
le32toh(lmac2->error_event_table_ptr);
sc->log_event_table = le32toh(lmac1->log_event_table_ptr);
sc->umac_error_event_table = le32toh(umac->error_info_addr);
alive_data->scd_base_addr = le32toh(lmac1->scd_base_ptr);
alive_data->valid = status == IWM_ALIVE_STATUS_OK;
if (sc->umac_error_event_table)
sc->support_umac_log = TRUE;
IWM_DPRINTF(sc, IWM_DEBUG_FW,
"Alive ucode status 0x%04x revision 0x%01X 0x%01X\n",
status, lmac1->ver_type, lmac1->ver_subtype);
if (lmac2)
IWM_DPRINTF(sc, IWM_DEBUG_FW, "Alive ucode CDB\n");
IWM_DPRINTF(sc, IWM_DEBUG_FW,
"UMAC version: Major - 0x%x, Minor - 0x%x\n",
le32toh(umac->umac_major),
le32toh(umac->umac_minor));
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
return TRUE;
}
[iwm] Use notification wait API to wait for calibration to complete. Tested: * 7260, STA mode (2g, 5g) Obtained from: DragonflyBSD commit 1e0cf8ec6fcd77978f5336297ece61a415790f84
2017-02-06 05:35:11 +00:00
static int
iwm_wait_phy_db_entry(struct iwm_softc *sc,
struct iwm_rx_packet *pkt, void *data)
{
struct iwm_phy_db *phy_db = data;
if (pkt->hdr.code != IWM_CALIB_RES_NOTIF_PHY_DB) {
if(pkt->hdr.code != IWM_INIT_COMPLETE_NOTIF) {
device_printf(sc->sc_dev, "%s: Unexpected cmd: %d\n",
__func__, pkt->hdr.code);
}
return TRUE;
}
if (iwm_phy_db_set_section(phy_db, pkt)) {
device_printf(sc->sc_dev,
"%s: iwm_phy_db_set_section failed\n", __func__);
}
return FALSE;
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static int
iwm_mvm_load_ucode_wait_alive(struct iwm_softc *sc,
enum iwm_ucode_type ucode_type)
{
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
struct iwm_notification_wait alive_wait;
struct iwm_mvm_alive_data alive_data;
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
const struct iwm_fw_img *fw;
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
enum iwm_ucode_type old_type = sc->cur_ucode;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
int error;
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
static const uint16_t alive_cmd[] = { IWM_MVM_ALIVE };
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
fw = &sc->sc_fw.img[ucode_type];
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
sc->cur_ucode = ucode_type;
sc->ucode_loaded = FALSE;
memset(&alive_data, 0, sizeof(alive_data));
iwm_init_notification_wait(sc->sc_notif_wait, &alive_wait,
alive_cmd, nitems(alive_cmd),
iwm_alive_fn, &alive_data);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
error = iwm_start_fw(sc, fw);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (error) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
device_printf(sc->sc_dev, "iwm_start_fw: failed %d\n", error);
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
sc->cur_ucode = old_type;
iwm_remove_notification(sc->sc_notif_wait, &alive_wait);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
return error;
}
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
/*
* Some things may run in the background now, but we
* just wait for the ALIVE notification here.
*/
IWM_UNLOCK(sc);
error = iwm_wait_notification(sc->sc_notif_wait, &alive_wait,
IWM_MVM_UCODE_ALIVE_TIMEOUT);
IWM_LOCK(sc);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
if (error) {
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
if (sc->cfg->device_family == IWM_DEVICE_FAMILY_8000) {
[iwm] iwm_{read,write}_prph() don't grab the nic lock in iwm themselves. * Fix a couple of cases where the nic lock ended up not being grabbed during an iwm_read_prph() or iwm_write_prph(). Obtained from: dragonflybsd.git 6c5470f2db219c61e362c981fea969d97e1b8293
2017-05-09 02:41:33 +00:00
uint32_t a = 0x5a5a5a5a, b = 0x5a5a5a5a;
if (iwm_nic_lock(sc)) {
a = iwm_read_prph(sc, IWM_SB_CPU_1_STATUS);
b = iwm_read_prph(sc, IWM_SB_CPU_2_STATUS);
[iwm] Add iwm_nic_unlock() calls missing from previous commit. Obtained from: dragonflybsd.git f88ab372284e63c4c13da93e9026a203b9b4cdc5
2017-05-09 02:42:53 +00:00
iwm_nic_unlock(sc);
[iwm] iwm_{read,write}_prph() don't grab the nic lock in iwm themselves. * Fix a couple of cases where the nic lock ended up not being grabbed during an iwm_read_prph() or iwm_write_prph(). Obtained from: dragonflybsd.git 6c5470f2db219c61e362c981fea969d97e1b8293
2017-05-09 02:41:33 +00:00
}
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
device_printf(sc->sc_dev,
"SecBoot CPU1 Status: 0x%x, CPU2 Status: 0x%x\n",
[iwm] iwm_{read,write}_prph() don't grab the nic lock in iwm themselves. * Fix a couple of cases where the nic lock ended up not being grabbed during an iwm_read_prph() or iwm_write_prph(). Obtained from: dragonflybsd.git 6c5470f2db219c61e362c981fea969d97e1b8293
2017-05-09 02:41:33 +00:00
a, b);
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
}
sc->cur_ucode = old_type;
return error;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
if (!alive_data.valid) {
device_printf(sc->sc_dev, "%s: Loaded ucode is not valid\n",
__func__);
sc->cur_ucode = old_type;
return EIO;
}
iwm_trans_pcie_fw_alive(sc, alive_data.scd_base_addr);
/*
* configure and operate fw paging mechanism.
* driver configures the paging flow only once, CPU2 paging image
* included in the IWM_UCODE_INIT image.
*/
if (fw->paging_mem_size) {
[iwm] Add support for Firmware paging, needed for newer 8000C firmware. * Uses the IWM_FW_PAGING_BLOCK_CMD firmware command to tell the firmware what memory ranges to use for paging. Obtained from: dragonflybsd.git 8a5b199964f8e7bdb00039f0b48817a01b402f18
2017-02-24 07:07:58 +00:00
error = iwm_save_fw_paging(sc, fw);
if (error) {
device_printf(sc->sc_dev,
"%s: failed to save the FW paging image\n",
__func__);
return error;
}
error = iwm_send_paging_cmd(sc, fw);
if (error) {
device_printf(sc->sc_dev,
"%s: failed to send the paging cmd\n", __func__);
iwm_free_fw_paging(sc);
return error;
}
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
}
if (!error)
sc->ucode_loaded = TRUE;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
return error;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
/*
* mvm misc bits
*/
/*
* follows iwlwifi/fw.c
*/
static int
iwm_run_init_mvm_ucode(struct iwm_softc *sc, int justnvm)
{
[iwm] Use notification wait API to wait for calibration to complete. Tested: * 7260, STA mode (2g, 5g) Obtained from: DragonflyBSD commit 1e0cf8ec6fcd77978f5336297ece61a415790f84
2017-02-06 05:35:11 +00:00
struct iwm_notification_wait calib_wait;
static const uint16_t init_complete[] = {
IWM_INIT_COMPLETE_NOTIF,
IWM_CALIB_RES_NOTIF_PHY_DB
};
int ret;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* do not operate with rfkill switch turned on */
if ((sc->sc_flags & IWM_FLAG_RFKILL) && !justnvm) {
device_printf(sc->sc_dev,
"radio is disabled by hardware switch\n");
return EPERM;
}
[iwm] Use notification wait API to wait for calibration to complete. Tested: * 7260, STA mode (2g, 5g) Obtained from: DragonflyBSD commit 1e0cf8ec6fcd77978f5336297ece61a415790f84
2017-02-06 05:35:11 +00:00
iwm_init_notification_wait(sc->sc_notif_wait,
&calib_wait,
init_complete,
nitems(init_complete),
iwm_wait_phy_db_entry,
sc->sc_phy_db);
/* Will also start the device */
ret = iwm_mvm_load_ucode_wait_alive(sc, IWM_UCODE_INIT);
if (ret) {
device_printf(sc->sc_dev, "Failed to start INIT ucode: %d\n",
ret);
goto error;
[iwm] Fix several nitpicks in iwm(4). Move some declarations to if_iwmreg.h. Remove iwm_fw_alive(); just call iwm_post_alive() directly. Simplify iwm_mvm_add_sta(). Return timeout error from iwm_apm_init(). Print a message when init (i.e. boot) firmware fails to load. Remove some commented-out code which wouldn't compile anyway. Move iwm_mvm_tx_fifo to if_iwmreg.h to match better where Linux puts it. Taken-From: OpenBSD (if_iwm.c r1.80 and if_iwmreg.h r1.11) Submitted by: Imre Vadasz <imre@vdsz.com> Obtained from: DragonflyBSD 29fcb331e5620ae145a6ab9cdda830e22fff626a
2016-06-02 05:43:16 +00:00
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (justnvm) {
[iwm] Use notification wait API to wait for calibration to complete. Tested: * 7260, STA mode (2g, 5g) Obtained from: DragonflyBSD commit 1e0cf8ec6fcd77978f5336297ece61a415790f84
2017-02-06 05:35:11 +00:00
/* Read nvm */
ret = iwm_nvm_init(sc);
if (ret) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev, "failed to read nvm\n");
[iwm] Use notification wait API to wait for calibration to complete. Tested: * 7260, STA mode (2g, 5g) Obtained from: DragonflyBSD commit 1e0cf8ec6fcd77978f5336297ece61a415790f84
2017-02-06 05:35:11 +00:00
goto error;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
IEEE80211_ADDR_COPY(sc->sc_ic.ic_macaddr, sc->nvm_data->hw_addr);
[iwm] Use notification wait API to wait for calibration to complete. Tested: * 7260, STA mode (2g, 5g) Obtained from: DragonflyBSD commit 1e0cf8ec6fcd77978f5336297ece61a415790f84
2017-02-06 05:35:11 +00:00
goto error;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Use notification wait API to wait for calibration to complete. Tested: * 7260, STA mode (2g, 5g) Obtained from: DragonflyBSD commit 1e0cf8ec6fcd77978f5336297ece61a415790f84
2017-02-06 05:35:11 +00:00
ret = iwm_send_bt_init_conf(sc);
if (ret) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
device_printf(sc->sc_dev,
[iwm] Use notification wait API to wait for calibration to complete. Tested: * 7260, STA mode (2g, 5g) Obtained from: DragonflyBSD commit 1e0cf8ec6fcd77978f5336297ece61a415790f84
2017-02-06 05:35:11 +00:00
"failed to send bt coex configuration: %d\n", ret);
goto error;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Send TX valid antennas before triggering calibrations */
[iwm] Use notification wait API to wait for calibration to complete. Tested: * 7260, STA mode (2g, 5g) Obtained from: DragonflyBSD commit 1e0cf8ec6fcd77978f5336297ece61a415790f84
2017-02-06 05:35:11 +00:00
ret = iwm_send_tx_ant_cfg(sc, iwm_mvm_get_valid_tx_ant(sc));
if (ret) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
device_printf(sc->sc_dev,
[iwm] Use notification wait API to wait for calibration to complete. Tested: * 7260, STA mode (2g, 5g) Obtained from: DragonflyBSD commit 1e0cf8ec6fcd77978f5336297ece61a415790f84
2017-02-06 05:35:11 +00:00
"failed to send antennas before calibration: %d\n", ret);
goto error;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/*
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
* Send phy configurations command to init uCode
* to start the 16.0 uCode init image internal calibrations.
*/
[iwm] Use notification wait API to wait for calibration to complete. Tested: * 7260, STA mode (2g, 5g) Obtained from: DragonflyBSD commit 1e0cf8ec6fcd77978f5336297ece61a415790f84
2017-02-06 05:35:11 +00:00
ret = iwm_send_phy_cfg_cmd(sc);
if (ret) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev,
[iwm] Use notification wait API to wait for calibration to complete. Tested: * 7260, STA mode (2g, 5g) Obtained from: DragonflyBSD commit 1e0cf8ec6fcd77978f5336297ece61a415790f84
2017-02-06 05:35:11 +00:00
"%s: Failed to run INIT calibrations: %d\n",
__func__, ret);
goto error;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
/*
* Nothing to do but wait for the init complete notification
[iwm] Use notification wait API to wait for calibration to complete. Tested: * 7260, STA mode (2g, 5g) Obtained from: DragonflyBSD commit 1e0cf8ec6fcd77978f5336297ece61a415790f84
2017-02-06 05:35:11 +00:00
* from the firmware.
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
*/
[iwm] Use notification wait API to wait for calibration to complete. Tested: * 7260, STA mode (2g, 5g) Obtained from: DragonflyBSD commit 1e0cf8ec6fcd77978f5336297ece61a415790f84
2017-02-06 05:35:11 +00:00
IWM_UNLOCK(sc);
ret = iwm_wait_notification(sc->sc_notif_wait, &calib_wait,
IWM_MVM_UCODE_CALIB_TIMEOUT);
IWM_LOCK(sc);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Use notification wait API to wait for calibration to complete. Tested: * 7260, STA mode (2g, 5g) Obtained from: DragonflyBSD commit 1e0cf8ec6fcd77978f5336297ece61a415790f84
2017-02-06 05:35:11 +00:00
goto out;
error:
iwm_remove_notification(sc->sc_notif_wait, &calib_wait);
out:
return ret;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
if_iwm - Configure the PCIe LTR, fix PCI express capability accesses. Taken-From: Linux iwlwifi Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (08a7ad5a5ff65aaaf2df6a609be7a4e1df43efc3)
2019-01-24 03:44:48 +00:00
static int
iwm_mvm_config_ltr(struct iwm_softc *sc)
{
struct iwm_ltr_config_cmd cmd = {
.flags = htole32(IWM_LTR_CFG_FLAG_FEATURE_ENABLE),
};
if (!sc->sc_ltr_enabled)
return 0;
return iwm_mvm_send_cmd_pdu(sc, IWM_LTR_CONFIG, 0, sizeof(cmd), &cmd);
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/*
* receive side
*/
/* (re)stock rx ring, called at init-time and at runtime */
static int
iwm_rx_addbuf(struct iwm_softc *sc, int size, int idx)
{
struct iwm_rx_ring *ring = &sc->rxq;
struct iwm_rx_data *data = &ring->data[idx];
struct mbuf *m;
[iwm] Tiny cleanup in iwm_rx_addbuf(). Obtained from: dragonflybsd.git 3370bc5504ebb3c1b9bb960a185cd0c8052a2845
2017-03-25 02:42:52 +00:00
bus_dmamap_t dmamap;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
bus_dma_segment_t seg;
int nsegs, error;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
m = m_getjcl(M_NOWAIT, MT_DATA, M_PKTHDR, IWM_RBUF_SIZE);
if (m == NULL)
return ENOBUFS;
m->m_len = m->m_pkthdr.len = m->m_ext.ext_size;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
error = bus_dmamap_load_mbuf_sg(ring->data_dmat, ring->spare_map, m,
&seg, &nsegs, BUS_DMA_NOWAIT);
if (error != 0) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev,
[iwm] Fix up busdma use in the RX path When allocating a new mbuf or bus_dmamap_load()-ing it fails, we can just keep the old mbuf since we are dropping that packet anyway. Instead of doing bus_dmamap_create() and bus_dmamap_destroy() all the time, create an extra bus_dmamap_t which we can use to safely try bus_dmamap_load()-ing the new mbuf. On success we just swap the spare bus_dmamap_t with the data->map of that ring entry. Tested: Tested with Intel AC7260, verified with vmstat -m that new kernel no longer visibly leaks memory from the M_DEVBUF malloc type. Before, leakage was 1KB every few seconds while ping(8)-ing over the wlan connection. Submitted by: Imre Vadasz <imre@vdsz.com> Approved by: re@ Obtained from: DragonflyBSD.git cc440b26818b5dfdd9af504d71c1b0e6522b53ef Differential Revision: https://reviews.freebsd.org/D6742
2016-06-13 00:13:20 +00:00
"%s: can't map mbuf, error %d\n", __func__, error);
[iwm] Tiny cleanup in iwm_rx_addbuf(). Obtained from: dragonflybsd.git 3370bc5504ebb3c1b9bb960a185cd0c8052a2845
2017-03-25 02:42:52 +00:00
m_freem(m);
return error;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Fix up busdma use in the RX path When allocating a new mbuf or bus_dmamap_load()-ing it fails, we can just keep the old mbuf since we are dropping that packet anyway. Instead of doing bus_dmamap_create() and bus_dmamap_destroy() all the time, create an extra bus_dmamap_t which we can use to safely try bus_dmamap_load()-ing the new mbuf. On success we just swap the spare bus_dmamap_t with the data->map of that ring entry. Tested: Tested with Intel AC7260, verified with vmstat -m that new kernel no longer visibly leaks memory from the M_DEVBUF malloc type. Before, leakage was 1KB every few seconds while ping(8)-ing over the wlan connection. Submitted by: Imre Vadasz <imre@vdsz.com> Approved by: re@ Obtained from: DragonflyBSD.git cc440b26818b5dfdd9af504d71c1b0e6522b53ef Differential Revision: https://reviews.freebsd.org/D6742
2016-06-13 00:13:20 +00:00
if (data->m != NULL)
bus_dmamap_unload(ring->data_dmat, data->map);
/* Swap ring->spare_map with data->map */
dmamap = data->map;
data->map = ring->spare_map;
ring->spare_map = dmamap;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
bus_dmamap_sync(ring->data_dmat, data->map, BUS_DMASYNC_PREREAD);
[iwm] Fix up busdma use in the RX path When allocating a new mbuf or bus_dmamap_load()-ing it fails, we can just keep the old mbuf since we are dropping that packet anyway. Instead of doing bus_dmamap_create() and bus_dmamap_destroy() all the time, create an extra bus_dmamap_t which we can use to safely try bus_dmamap_load()-ing the new mbuf. On success we just swap the spare bus_dmamap_t with the data->map of that ring entry. Tested: Tested with Intel AC7260, verified with vmstat -m that new kernel no longer visibly leaks memory from the M_DEVBUF malloc type. Before, leakage was 1KB every few seconds while ping(8)-ing over the wlan connection. Submitted by: Imre Vadasz <imre@vdsz.com> Approved by: re@ Obtained from: DragonflyBSD.git cc440b26818b5dfdd9af504d71c1b0e6522b53ef Differential Revision: https://reviews.freebsd.org/D6742
2016-06-13 00:13:20 +00:00
data->m = m;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Update RX descriptor. */
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
KASSERT((seg.ds_addr & 255) == 0, ("seg.ds_addr not aligned"));
ring->desc[idx] = htole32(seg.ds_addr >> 8);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
bus_dmamap_sync(ring->desc_dma.tag, ring->desc_dma.map,
BUS_DMASYNC_PREWRITE);
return 0;
}
/* iwlwifi: mvm/rx.c */
/*
* iwm_mvm_get_signal_strength - use new rx PHY INFO API
* values are reported by the fw as positive values - need to negate
* to obtain their dBM. Account for missing antennas by replacing 0
* values by -256dBm: practically 0 power and a non-feasible 8 bit value.
*/
static int
iwm_mvm_get_signal_strength(struct iwm_softc *sc, struct iwm_rx_phy_info *phy_info)
{
int energy_a, energy_b, energy_c, max_energy;
uint32_t val;
val = le32toh(phy_info->non_cfg_phy[IWM_RX_INFO_ENERGY_ANT_ABC_IDX]);
energy_a = (val & IWM_RX_INFO_ENERGY_ANT_A_MSK) >>
IWM_RX_INFO_ENERGY_ANT_A_POS;
energy_a = energy_a ? -energy_a : -256;
energy_b = (val & IWM_RX_INFO_ENERGY_ANT_B_MSK) >>
IWM_RX_INFO_ENERGY_ANT_B_POS;
energy_b = energy_b ? -energy_b : -256;
energy_c = (val & IWM_RX_INFO_ENERGY_ANT_C_MSK) >>
IWM_RX_INFO_ENERGY_ANT_C_POS;
energy_c = energy_c ? -energy_c : -256;
max_energy = MAX(energy_a, energy_b);
max_energy = MAX(max_energy, energy_c);
IWM_DPRINTF(sc, IWM_DEBUG_RECV,
"energy In A %d B %d C %d , and max %d\n",
energy_a, energy_b, energy_c, max_energy);
return max_energy;
}
static void
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
iwm_mvm_rx_rx_phy_cmd(struct iwm_softc *sc, struct iwm_rx_packet *pkt)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
struct iwm_rx_phy_info *phy_info = (void *)pkt->data;
IWM_DPRINTF(sc, IWM_DEBUG_RECV, "received PHY stats\n");
memcpy(&sc->sc_last_phy_info, phy_info, sizeof(sc->sc_last_phy_info));
}
/*
* Retrieve the average noise (in dBm) among receivers.
*/
static int
[iwm] fix up RSSI calculations for both scan results and normal RX operations. * hard code a noise floor of -96 for now. The noise floor update code returns some "interesting" values that I can't map to anything useful right now. * Ensure a default noise floor is set - otherwise the initial scan results have a noise floor of '0'. * Fix up the RSSI calculation to be correctly relative to the noise floor. The RSSI routines return an absolute value in dBm - so fix this up. * Cap RSSI values appropriately. * Ensure we pass in a 1/2 dB unit value in to net80211. Tested: * Intel 7260, STA mode iwm0: <Intel Dual Band Wireless AC 7260> mem 0xf1400000-0xf1401fff irq 17 at device 0.0 on pci2 iwm0: hw rev 0x140, fw ver 16.242414.0, address xx:xx:xx:xx:xx:xx
2016-09-18 05:07:18 +00:00
iwm_get_noise(struct iwm_softc *sc,
const struct iwm_mvm_statistics_rx_non_phy *stats)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
int i, total, nbant, noise;
total = nbant = noise = 0;
for (i = 0; i < 3; i++) {
noise = le32toh(stats->beacon_silence_rssi[i]) & 0xff;
[iwm] fix up RSSI calculations for both scan results and normal RX operations. * hard code a noise floor of -96 for now. The noise floor update code returns some "interesting" values that I can't map to anything useful right now. * Ensure a default noise floor is set - otherwise the initial scan results have a noise floor of '0'. * Fix up the RSSI calculation to be correctly relative to the noise floor. The RSSI routines return an absolute value in dBm - so fix this up. * Cap RSSI values appropriately. * Ensure we pass in a 1/2 dB unit value in to net80211. Tested: * Intel 7260, STA mode iwm0: <Intel Dual Band Wireless AC 7260> mem 0xf1400000-0xf1401fff irq 17 at device 0.0 on pci2 iwm0: hw rev 0x140, fw ver 16.242414.0, address xx:xx:xx:xx:xx:xx
2016-09-18 05:07:18 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_RECV, "%s: i=%d, noise=%d\n",
__func__,
i,
noise);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (noise) {
total += noise;
nbant++;
}
}
[iwm] fix up RSSI calculations for both scan results and normal RX operations. * hard code a noise floor of -96 for now. The noise floor update code returns some "interesting" values that I can't map to anything useful right now. * Ensure a default noise floor is set - otherwise the initial scan results have a noise floor of '0'. * Fix up the RSSI calculation to be correctly relative to the noise floor. The RSSI routines return an absolute value in dBm - so fix this up. * Cap RSSI values appropriately. * Ensure we pass in a 1/2 dB unit value in to net80211. Tested: * Intel 7260, STA mode iwm0: <Intel Dual Band Wireless AC 7260> mem 0xf1400000-0xf1401fff irq 17 at device 0.0 on pci2 iwm0: hw rev 0x140, fw ver 16.242414.0, address xx:xx:xx:xx:xx:xx
2016-09-18 05:07:18 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_RECV, "%s: nbant=%d, total=%d\n",
__func__, nbant, total);
#if 0
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* There should be at least one antenna but check anyway. */
return (nbant == 0) ? -127 : (total / nbant) - 107;
[iwm] fix up RSSI calculations for both scan results and normal RX operations. * hard code a noise floor of -96 for now. The noise floor update code returns some "interesting" values that I can't map to anything useful right now. * Ensure a default noise floor is set - otherwise the initial scan results have a noise floor of '0'. * Fix up the RSSI calculation to be correctly relative to the noise floor. The RSSI routines return an absolute value in dBm - so fix this up. * Cap RSSI values appropriately. * Ensure we pass in a 1/2 dB unit value in to net80211. Tested: * Intel 7260, STA mode iwm0: <Intel Dual Band Wireless AC 7260> mem 0xf1400000-0xf1401fff irq 17 at device 0.0 on pci2 iwm0: hw rev 0x140, fw ver 16.242414.0, address xx:xx:xx:xx:xx:xx
2016-09-18 05:07:18 +00:00
#else
/* For now, just hard-code it to -96 to be safe */
return (-96);
#endif
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Sync statistics API definitions with Linux iwlwifi. Obtained from: dragonflybsd.git 75895a53a9c1ba60d75be9b4bf6e49a37f91a7cf
2017-07-26 05:40:52 +00:00
static void
iwm_mvm_handle_rx_statistics(struct iwm_softc *sc, struct iwm_rx_packet *pkt)
{
struct iwm_notif_statistics_v10 *stats = (void *)&pkt->data;
memcpy(&sc->sc_stats, stats, sizeof(sc->sc_stats));
sc->sc_noise = iwm_get_noise(sc, &stats->rx.general);
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/*
* iwm_mvm_rx_rx_mpdu - IWM_REPLY_RX_MPDU_CMD handler
*
* Handles the actual data of the Rx packet from the fw
*/
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
static boolean_t
iwm_mvm_rx_rx_mpdu(struct iwm_softc *sc, struct mbuf *m, uint32_t offset,
boolean_t stolen)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
struct ieee80211com *ic = &sc->sc_ic;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
struct ieee80211vap *vap = TAILQ_FIRST(&ic->ic_vaps);
struct ieee80211_frame *wh;
struct ieee80211_node *ni;
struct ieee80211_rx_stats rxs;
struct iwm_rx_phy_info *phy_info;
struct iwm_rx_mpdu_res_start *rx_res;
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
struct iwm_rx_packet *pkt = mtodoff(m, struct iwm_rx_packet *, offset);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
uint32_t len;
uint32_t rx_pkt_status;
int rssi;
phy_info = &sc->sc_last_phy_info;
rx_res = (struct iwm_rx_mpdu_res_start *)pkt->data;
wh = (struct ieee80211_frame *)(pkt->data + sizeof(*rx_res));
len = le16toh(rx_res->byte_count);
rx_pkt_status = le32toh(*(uint32_t *)(pkt->data + sizeof(*rx_res) + len));
if (__predict_false(phy_info->cfg_phy_cnt > 20)) {
device_printf(sc->sc_dev,
"dsp size out of range [0,20]: %d\n",
phy_info->cfg_phy_cnt);
iwm: add missing 'Rx errors' counter incrementation.
2016-10-09 10:21:28 +00:00
goto fail;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
if (!(rx_pkt_status & IWM_RX_MPDU_RES_STATUS_CRC_OK) ||
!(rx_pkt_status & IWM_RX_MPDU_RES_STATUS_OVERRUN_OK)) {
IWM_DPRINTF(sc, IWM_DEBUG_RECV,
"Bad CRC or FIFO: 0x%08X.\n", rx_pkt_status);
iwm: add missing 'Rx errors' counter incrementation.
2016-10-09 10:21:28 +00:00
goto fail;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Remove a couple of unneeded IWM_UCODE_TLV_FLAGS_* flags. * All the supported firmwares have these flags set. * This removes the following flags: IWM_UCODE_TLV_FLAGS_PM_CMD_SUPPORT, IWM_UCODE_TLV_FLAGS_NEWBT_COEX, IWM_UCODE_TLV_FLAGS_BF_UPDATED, IWM_UCODE_TLV_FLAGS_D3_CONTINUITY_API, IWM_UCODE_TLV_FLAGS_STA_KEY_CMD, IWM_UCODE_TLV_FLAGS_DEVICE_PS_CMD, IWM_UCODE_TLV_FLAGS_SCHED_SCAN, IWM_UCODE_TLV_FLAGS_RX_ENERGY_API, IWM_UCODE_TLV_FLAGS_TIME_EVENT_API_V2 * Also remove definitions and code for dealing with the v1 time-event api. * Remove unneeded calc_rssi() function. Obtained from: dragonflybsd.git d078c812418d0e2c3392e99fa25fc776d07bdfad
2017-03-23 04:43:04 +00:00
rssi = iwm_mvm_get_signal_strength(sc, phy_info);
[iwm] fix up RSSI calculations for both scan results and normal RX operations. * hard code a noise floor of -96 for now. The noise floor update code returns some "interesting" values that I can't map to anything useful right now. * Ensure a default noise floor is set - otherwise the initial scan results have a noise floor of '0'. * Fix up the RSSI calculation to be correctly relative to the noise floor. The RSSI routines return an absolute value in dBm - so fix this up. * Cap RSSI values appropriately. * Ensure we pass in a 1/2 dB unit value in to net80211. Tested: * Intel 7260, STA mode iwm0: <Intel Dual Band Wireless AC 7260> mem 0xf1400000-0xf1401fff irq 17 at device 0.0 on pci2 iwm0: hw rev 0x140, fw ver 16.242414.0, address xx:xx:xx:xx:xx:xx
2016-09-18 05:07:18 +00:00
/* Map it to relative value */
rssi = rssi - sc->sc_noise;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* replenish ring for the buffer we're going to feed to the sharks */
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
if (!stolen && iwm_rx_addbuf(sc, IWM_RBUF_SIZE, sc->rxq.cur) != 0) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev, "%s: unable to add more buffers\n",
__func__);
iwm: add missing 'Rx errors' counter incrementation.
2016-10-09 10:21:28 +00:00
goto fail;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Move mbuf hacks after sanity checks in iwm_mvm_rx_rx_mpdu(). * This avoids leaving the mbuf in a weird state, when dropping a packet. Obtained from: dragonflybsd.git 96eaecf93d9f731459a0df8efc72cfad034320bd
2017-03-23 04:34:25 +00:00
m->m_data = pkt->data + sizeof(*rx_res);
m->m_pkthdr.len = m->m_len = len;
[iwm] fix up RSSI calculations for both scan results and normal RX operations. * hard code a noise floor of -96 for now. The noise floor update code returns some "interesting" values that I can't map to anything useful right now. * Ensure a default noise floor is set - otherwise the initial scan results have a noise floor of '0'. * Fix up the RSSI calculation to be correctly relative to the noise floor. The RSSI routines return an absolute value in dBm - so fix this up. * Cap RSSI values appropriately. * Ensure we pass in a 1/2 dB unit value in to net80211. Tested: * Intel 7260, STA mode iwm0: <Intel Dual Band Wireless AC 7260> mem 0xf1400000-0xf1401fff irq 17 at device 0.0 on pci2 iwm0: hw rev 0x140, fw ver 16.242414.0, address xx:xx:xx:xx:xx:xx
2016-09-18 05:07:18 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_RECV,
"%s: rssi=%d, noise=%d\n", __func__, rssi, sc->sc_noise);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
ni = ieee80211_find_rxnode(ic, (struct ieee80211_frame_min *)wh);
IWM_DPRINTF(sc, IWM_DEBUG_RECV,
"%s: phy_info: channel=%d, flags=0x%08x\n",
__func__,
le16toh(phy_info->channel),
le16toh(phy_info->phy_flags));
/*
* Populate an RX state struct with the provided information.
*/
bzero(&rxs, sizeof(rxs));
rxs.r_flags |= IEEE80211_R_IEEE | IEEE80211_R_FREQ;
rxs.r_flags |= IEEE80211_R_NF | IEEE80211_R_RSSI;
rxs.c_ieee = le16toh(phy_info->channel);
if (le16toh(phy_info->phy_flags & IWM_RX_RES_PHY_FLAGS_BAND_24)) {
rxs.c_freq = ieee80211_ieee2mhz(rxs.c_ieee, IEEE80211_CHAN_2GHZ);
} else {
rxs.c_freq = ieee80211_ieee2mhz(rxs.c_ieee, IEEE80211_CHAN_5GHZ);
}
[iwm] fix up RSSI calculations for both scan results and normal RX operations. * hard code a noise floor of -96 for now. The noise floor update code returns some "interesting" values that I can't map to anything useful right now. * Ensure a default noise floor is set - otherwise the initial scan results have a noise floor of '0'. * Fix up the RSSI calculation to be correctly relative to the noise floor. The RSSI routines return an absolute value in dBm - so fix this up. * Cap RSSI values appropriately. * Ensure we pass in a 1/2 dB unit value in to net80211. Tested: * Intel 7260, STA mode iwm0: <Intel Dual Band Wireless AC 7260> mem 0xf1400000-0xf1401fff irq 17 at device 0.0 on pci2 iwm0: hw rev 0x140, fw ver 16.242414.0, address xx:xx:xx:xx:xx:xx
2016-09-18 05:07:18 +00:00
/* rssi is in 1/2db units */
[net80211] extend the ieee80211_rx_stats struct to include more information. There are a variety of more interesting RX statistics that we should keep track of but we don't. This is a starting point for adding more information. Specifically: * now the RX rate information and some of the packet status is passed up; * The 32 bit or 64 bit TSF is passed up; * the PHY mode is passed up; * the "I'm decap'ed AMSDU!" state is passed up; * number of RX chains is bumped to 4. This is all mostly a placeholder for getting the data into the RX status before we pass it up to net80211 - unfortunately we don't yet enforce that drivers provide it, nor do we pass the provided info back up the stack so anyone can use the data. We're going to need to use some of this data moving forward. Notably, now that some hardware can do AMSDU decap for us (the intel iwm driver can do it when we flip it on; the ath10k port I'm doing does it for us) then we need to pass it up through the stack so the duplicate RX sequence numbers and crypto/IV details don't cause the packet to be dropped and/or counted against a replay counter. It's also the beginning of being able to do more interesting node accounting in net80211. Specifically, once drivers start populating per-packet rate information, AMPDU information, timestamps, etc, we can start providing histograms of rate-versus-RSSI, account for receive time spent per node and other such interesting things. (Note: I'm also hoping to include ranging and RTT information for future chipset support; and it's likely going to include it in this kind of fashion.)
2016-10-08 01:12:29 +00:00
rxs.c_rssi = rssi * 2;
rxs.c_nf = sc->sc_noise;
[iwm] free node reference if rxparams addition fails.
2017-02-06 01:57:41 +00:00
if (ieee80211_add_rx_params(m, &rxs) == 0) {
if (ni)
ieee80211_free_node(ni);
net80211: convert all ieee80211_input_mimo*() consumers to ieee80211_add_rx_params() + drop last (ieee80211_rx_stats) parameter Note: there is an additional check for ieee80211_get_rx_params() return value (which does not exist in the original diff). Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D8207
2016-10-12 20:50:13 +00:00
goto fail;
[iwm] free node reference if rxparams addition fails.
2017-02-06 01:57:41 +00:00
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (ieee80211_radiotap_active_vap(vap)) {
struct iwm_rx_radiotap_header *tap = &sc->sc_rxtap;
tap->wr_flags = 0;
if (phy_info->phy_flags & htole16(IWM_PHY_INFO_FLAG_SHPREAMBLE))
tap->wr_flags |= IEEE80211_RADIOTAP_F_SHORTPRE;
tap->wr_chan_freq = htole16(rxs.c_freq);
/* XXX only if ic->ic_curchan->ic_ieee == rxs.c_ieee */
tap->wr_chan_flags = htole16(ic->ic_curchan->ic_flags);
tap->wr_dbm_antsignal = (int8_t)rssi;
tap->wr_dbm_antnoise = (int8_t)sc->sc_noise;
tap->wr_tsft = phy_info->system_timestamp;
switch (phy_info->rate) {
/* CCK rates. */
case 10: tap->wr_rate = 2; break;
case 20: tap->wr_rate = 4; break;
case 55: tap->wr_rate = 11; break;
case 110: tap->wr_rate = 22; break;
/* OFDM rates. */
case 0xd: tap->wr_rate = 12; break;
case 0xf: tap->wr_rate = 18; break;
case 0x5: tap->wr_rate = 24; break;
case 0x7: tap->wr_rate = 36; break;
case 0x9: tap->wr_rate = 48; break;
case 0xb: tap->wr_rate = 72; break;
case 0x1: tap->wr_rate = 96; break;
case 0x3: tap->wr_rate = 108; break;
/* Unknown rate: should not happen. */
default: tap->wr_rate = 0;
}
}
IWM_UNLOCK(sc);
if (ni != NULL) {
IWM_DPRINTF(sc, IWM_DEBUG_RECV, "input m %p\n", m);
net80211: convert all ieee80211_input_mimo*() consumers to ieee80211_add_rx_params() + drop last (ieee80211_rx_stats) parameter Note: there is an additional check for ieee80211_get_rx_params() return value (which does not exist in the original diff). Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D8207
2016-10-12 20:50:13 +00:00
ieee80211_input_mimo(ni, m);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
ieee80211_free_node(ni);
} else {
IWM_DPRINTF(sc, IWM_DEBUG_RECV, "inputall m %p\n", m);
net80211: convert all ieee80211_input_mimo*() consumers to ieee80211_add_rx_params() + drop last (ieee80211_rx_stats) parameter Note: there is an additional check for ieee80211_get_rx_params() return value (which does not exist in the original diff). Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D8207
2016-10-12 20:50:13 +00:00
ieee80211_input_mimo_all(ic, m);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
IWM_LOCK(sc);
iwm: add missing 'Rx errors' counter incrementation.
2016-10-09 10:21:28 +00:00
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
return TRUE;
iwm: add missing 'Rx errors' counter incrementation.
2016-10-09 10:21:28 +00:00
[iwm] free node reference if rxparams addition fails.
2017-02-06 01:57:41 +00:00
fail:
counter_u64_add(ic->ic_ierrors, 1);
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
return FALSE;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
iwm: convert to ieee80211_tx_complete() Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D4755
2016-01-03 10:10:11 +00:00
static int
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm_mvm_rx_tx_cmd_single(struct iwm_softc *sc, struct iwm_rx_packet *pkt,
struct iwm_node *in)
{
struct iwm_mvm_tx_resp *tx_resp = (void *)pkt->data;
net80211: ieee80211_ratectl*: switch to reusable KPI Replace various void * / int argument combinations with common structures: - ieee80211_ratectl_tx_status for *_tx_complete(); - ieee80211_ratectl_tx_stats for *_tx_update(); While here, improve amrr_tx_update() for a bit: 1. In case, if receiver is not known (typical for Ralink USB drivers), refresh Tx rate for all nodes on the interface. 2. There was a misuse: - otus(4) sends non-decreasing counters (as originally intended); - but ural(4), rum(4) and run(4) are using 'read & clear' registers to obtain statistics for some period of time (and those 'last period' values are used as arguments for tx_update()). If arguments are not big enough, they are just discarded after the next call. Fix: move counting into *_tx_update() (now otus(4) will zero out all node counters after every tx_update() call) Tested with: - Intel 3945BG (wpi(4)), STA mode. - WUSB54GC (rum(4)), STA / HOSTAP mode. - RTL8188EU (urtwn(4)), STA mode. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D8037
2016-10-02 20:35:55 +00:00
struct ieee80211_ratectl_tx_status *txs = &sc->sc_txs;
iwm: convert to ieee80211_tx_complete() Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D4755
2016-01-03 10:10:11 +00:00
struct ieee80211_node *ni = &in->in_ni;
if_iwm - Update firmware rs table, instead of indexing the table in tx cmds. * Rather than providing a non-zero index into the firmware RS table, we should always use index 0 and update the firmware RS table whenever our chosen tx rate for data-frames changes. * Send IWM_LQ_CMD updates when the tx rate gets updated by the net80211 rate control (which is after we tell the tx status to the net80211 rate-control in iwm_mvm_rx_tx_cmd_single()). * Disregard frames transferred with a different tx rate than the currently selected rate for the rate-control calculations. This way we avoid counting management frames (which are sent at a slow, and fixed rate), as well as frames we added to the tx queue just before a new IWM_LQ_CMD update took effect. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (5d6b465e288ac5b52d7115688d4e6516acbbea1c)
2019-01-24 03:41:09 +00:00
struct ieee80211vap *vap = ni->ni_vap;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
int status = le16toh(tx_resp->status.status) & IWM_TX_STATUS_MSK;
if_iwm - Update firmware rs table, instead of indexing the table in tx cmds. * Rather than providing a non-zero index into the firmware RS table, we should always use index 0 and update the firmware RS table whenever our chosen tx rate for data-frames changes. * Send IWM_LQ_CMD updates when the tx rate gets updated by the net80211 rate control (which is after we tell the tx status to the net80211 rate-control in iwm_mvm_rx_tx_cmd_single()). * Disregard frames transferred with a different tx rate than the currently selected rate for the rate-control calculations. This way we avoid counting management frames (which are sent at a slow, and fixed rate), as well as frames we added to the tx queue just before a new IWM_LQ_CMD update took effect. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (5d6b465e288ac5b52d7115688d4e6516acbbea1c)
2019-01-24 03:41:09 +00:00
int new_rate, cur_rate = vap->iv_bss->ni_txrate;
boolean_t rate_matched;
uint8_t tx_resp_rate;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
KASSERT(tx_resp->frame_count == 1, ("too many frames"));
/* Update rate control statistics. */
[iwm] add extra transmit setup/completion logging so I can see what's going on. I'm seeing 5GHz association work but data not work until the rate drops, so I need way more information about what's being programmed into the transmit descriptors. Tested: * 7260AC, STA mode
2016-04-26 03:24:28 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_XMIT, "%s: status=0x%04x, seq=%d, fc=%d, btc=%d, frts=%d, ff=%d, irate=%08x, wmt=%d\n",
__func__,
(int) le16toh(tx_resp->status.status),
(int) le16toh(tx_resp->status.sequence),
tx_resp->frame_count,
tx_resp->bt_kill_count,
tx_resp->failure_rts,
tx_resp->failure_frame,
le32toh(tx_resp->initial_rate),
(int) le16toh(tx_resp->wireless_media_time));
if_iwm - Update firmware rs table, instead of indexing the table in tx cmds. * Rather than providing a non-zero index into the firmware RS table, we should always use index 0 and update the firmware RS table whenever our chosen tx rate for data-frames changes. * Send IWM_LQ_CMD updates when the tx rate gets updated by the net80211 rate control (which is after we tell the tx status to the net80211 rate-control in iwm_mvm_rx_tx_cmd_single()). * Disregard frames transferred with a different tx rate than the currently selected rate for the rate-control calculations. This way we avoid counting management frames (which are sent at a slow, and fixed rate), as well as frames we added to the tx queue just before a new IWM_LQ_CMD update took effect. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (5d6b465e288ac5b52d7115688d4e6516acbbea1c)
2019-01-24 03:41:09 +00:00
tx_resp_rate = iwm_rate_from_ucode_rate(le32toh(tx_resp->initial_rate));
/* For rate control, ignore frames sent at different initial rate */
rate_matched = (tx_resp_rate != 0 && tx_resp_rate == cur_rate);
if (tx_resp_rate != 0 && cur_rate != 0 && !rate_matched) {
IWM_DPRINTF(sc, IWM_DEBUG_TXRATE,
"tx_resp_rate doesn't match ni_txrate (tx_resp_rate=%u "
"ni_txrate=%d)\n", tx_resp_rate, cur_rate);
}
net80211: ieee80211_ratectl*: switch to reusable KPI Replace various void * / int argument combinations with common structures: - ieee80211_ratectl_tx_status for *_tx_complete(); - ieee80211_ratectl_tx_stats for *_tx_update(); While here, improve amrr_tx_update() for a bit: 1. In case, if receiver is not known (typical for Ralink USB drivers), refresh Tx rate for all nodes on the interface. 2. There was a misuse: - otus(4) sends non-decreasing counters (as originally intended); - but ural(4), rum(4) and run(4) are using 'read & clear' registers to obtain statistics for some period of time (and those 'last period' values are used as arguments for tx_update()). If arguments are not big enough, they are just discarded after the next call. Fix: move counting into *_tx_update() (now otus(4) will zero out all node counters after every tx_update() call) Tested with: - Intel 3945BG (wpi(4)), STA mode. - WUSB54GC (rum(4)), STA / HOSTAP mode. - RTL8188EU (urtwn(4)), STA mode. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D8037
2016-10-02 20:35:55 +00:00
txs->flags = IEEE80211_RATECTL_STATUS_SHORT_RETRY |
IEEE80211_RATECTL_STATUS_LONG_RETRY;
txs->short_retries = tx_resp->failure_rts;
txs->long_retries = tx_resp->failure_frame;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (status != IWM_TX_STATUS_SUCCESS &&
status != IWM_TX_STATUS_DIRECT_DONE) {
net80211: ieee80211_ratectl*: switch to reusable KPI Replace various void * / int argument combinations with common structures: - ieee80211_ratectl_tx_status for *_tx_complete(); - ieee80211_ratectl_tx_stats for *_tx_update(); While here, improve amrr_tx_update() for a bit: 1. In case, if receiver is not known (typical for Ralink USB drivers), refresh Tx rate for all nodes on the interface. 2. There was a misuse: - otus(4) sends non-decreasing counters (as originally intended); - but ural(4), rum(4) and run(4) are using 'read & clear' registers to obtain statistics for some period of time (and those 'last period' values are used as arguments for tx_update()). If arguments are not big enough, they are just discarded after the next call. Fix: move counting into *_tx_update() (now otus(4) will zero out all node counters after every tx_update() call) Tested with: - Intel 3945BG (wpi(4)), STA mode. - WUSB54GC (rum(4)), STA / HOSTAP mode. - RTL8188EU (urtwn(4)), STA mode. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D8037
2016-10-02 20:35:55 +00:00
switch (status) {
case IWM_TX_STATUS_FAIL_SHORT_LIMIT:
txs->status = IEEE80211_RATECTL_TX_FAIL_SHORT;
break;
case IWM_TX_STATUS_FAIL_LONG_LIMIT:
txs->status = IEEE80211_RATECTL_TX_FAIL_LONG;
break;
case IWM_TX_STATUS_FAIL_LIFE_EXPIRE:
txs->status = IEEE80211_RATECTL_TX_FAIL_EXPIRED;
break;
default:
txs->status = IEEE80211_RATECTL_TX_FAIL_UNSPECIFIED;
break;
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
} else {
net80211: ieee80211_ratectl*: switch to reusable KPI Replace various void * / int argument combinations with common structures: - ieee80211_ratectl_tx_status for *_tx_complete(); - ieee80211_ratectl_tx_stats for *_tx_update(); While here, improve amrr_tx_update() for a bit: 1. In case, if receiver is not known (typical for Ralink USB drivers), refresh Tx rate for all nodes on the interface. 2. There was a misuse: - otus(4) sends non-decreasing counters (as originally intended); - but ural(4), rum(4) and run(4) are using 'read & clear' registers to obtain statistics for some period of time (and those 'last period' values are used as arguments for tx_update()). If arguments are not big enough, they are just discarded after the next call. Fix: move counting into *_tx_update() (now otus(4) will zero out all node counters after every tx_update() call) Tested with: - Intel 3945BG (wpi(4)), STA mode. - WUSB54GC (rum(4)), STA / HOSTAP mode. - RTL8188EU (urtwn(4)), STA mode. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D8037
2016-10-02 20:35:55 +00:00
txs->status = IEEE80211_RATECTL_TX_SUCCESS;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
if_iwm - Update firmware rs table, instead of indexing the table in tx cmds. * Rather than providing a non-zero index into the firmware RS table, we should always use index 0 and update the firmware RS table whenever our chosen tx rate for data-frames changes. * Send IWM_LQ_CMD updates when the tx rate gets updated by the net80211 rate control (which is after we tell the tx status to the net80211 rate-control in iwm_mvm_rx_tx_cmd_single()). * Disregard frames transferred with a different tx rate than the currently selected rate for the rate-control calculations. This way we avoid counting management frames (which are sent at a slow, and fixed rate), as well as frames we added to the tx queue just before a new IWM_LQ_CMD update took effect. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (5d6b465e288ac5b52d7115688d4e6516acbbea1c)
2019-01-24 03:41:09 +00:00
if (rate_matched) {
ieee80211_ratectl_tx_complete(ni, txs);
int rix = ieee80211_ratectl_rate(vap->iv_bss, NULL, 0);
new_rate = vap->iv_bss->ni_txrate;
if (new_rate != 0 && new_rate != cur_rate) {
struct iwm_node *in = IWM_NODE(vap->iv_bss);
iwm_setrates(sc, in, rix);
iwm_mvm_send_lq_cmd(sc, &in->in_lq, FALSE);
}
}
net80211: ieee80211_ratectl*: switch to reusable KPI Replace various void * / int argument combinations with common structures: - ieee80211_ratectl_tx_status for *_tx_complete(); - ieee80211_ratectl_tx_stats for *_tx_update(); While here, improve amrr_tx_update() for a bit: 1. In case, if receiver is not known (typical for Ralink USB drivers), refresh Tx rate for all nodes on the interface. 2. There was a misuse: - otus(4) sends non-decreasing counters (as originally intended); - but ural(4), rum(4) and run(4) are using 'read & clear' registers to obtain statistics for some period of time (and those 'last period' values are used as arguments for tx_update()). If arguments are not big enough, they are just discarded after the next call. Fix: move counting into *_tx_update() (now otus(4) will zero out all node counters after every tx_update() call) Tested with: - Intel 3945BG (wpi(4)), STA mode. - WUSB54GC (rum(4)), STA / HOSTAP mode. - RTL8188EU (urtwn(4)), STA mode. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D8037
2016-10-02 20:35:55 +00:00
return (txs->status != IEEE80211_RATECTL_TX_SUCCESS);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
static void
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
iwm_mvm_rx_tx_cmd(struct iwm_softc *sc, struct iwm_rx_packet *pkt)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
struct iwm_cmd_header *cmd_hdr = &pkt->hdr;
int idx = cmd_hdr->idx;
int qid = cmd_hdr->qid;
struct iwm_tx_ring *ring = &sc->txq[qid];
struct iwm_tx_data *txd = &ring->data[idx];
struct iwm_node *in = txd->in;
iwm: convert to ieee80211_tx_complete() Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D4755
2016-01-03 10:10:11 +00:00
struct mbuf *m = txd->m;
int status;
KASSERT(txd->done == 0, ("txd not done"));
KASSERT(txd->in != NULL, ("txd without node"));
KASSERT(txd->m != NULL, ("txd without mbuf"));
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
sc->sc_tx_timer = 0;
iwm: convert to ieee80211_tx_complete() Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D4755
2016-01-03 10:10:11 +00:00
status = iwm_mvm_rx_tx_cmd_single(sc, pkt, in);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Unmap and free mbuf. */
bus_dmamap_sync(ring->data_dmat, txd->map, BUS_DMASYNC_POSTWRITE);
bus_dmamap_unload(ring->data_dmat, txd->map);
IWM_DPRINTF(sc, IWM_DEBUG_XMIT,
"free txd %p, in %p\n", txd, txd->in);
txd->done = 1;
txd->m = NULL;
txd->in = NULL;
iwm: convert to ieee80211_tx_complete() Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D4755
2016-01-03 10:10:11 +00:00
ieee80211_tx_complete(&in->in_ni, m, status);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (--ring->queued < IWM_TX_RING_LOMARK) {
sc->qfullmsk &= ~(1 << ring->qid);
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
if (sc->qfullmsk == 0) {
iwm_start(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
}
}
/*
* transmit side
*/
/*
* Process a "command done" firmware notification. This is where we wakeup
* processes waiting for a synchronous command completion.
* from if_iwn
*/
static void
iwm_cmd_done(struct iwm_softc *sc, struct iwm_rx_packet *pkt)
{
struct iwm_tx_ring *ring = &sc->txq[IWM_MVM_CMD_QUEUE];
struct iwm_tx_data *data;
if (pkt->hdr.qid != IWM_MVM_CMD_QUEUE) {
return; /* Not a command ack. */
}
iwm: add 'command accepted' debug notification (copied from wpi(4)). Now it should be easier to find out which command causes firmware panics when few commands are sent in a short period of time.
2016-08-27 08:34:20 +00:00
/* XXX wide commands? */
IWM_DPRINTF(sc, IWM_DEBUG_CMD,
"cmd notification type 0x%x qid %d idx %d\n",
pkt->hdr.code, pkt->hdr.qid, pkt->hdr.idx);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
data = &ring->data[pkt->hdr.idx];
/* If the command was mapped in an mbuf, free it. */
if (data->m != NULL) {
bus_dmamap_sync(ring->data_dmat, data->map,
BUS_DMASYNC_POSTWRITE);
bus_dmamap_unload(ring->data_dmat, data->map);
m_freem(data->m);
data->m = NULL;
}
wakeup(&ring->desc[pkt->hdr.idx]);
[iwm] Implement apmg_wake_up_wa workaround properly for 7000 family. * Add iwm_pcie_set_cmd_in_flight() and iwm_pcie_clear_cmd_in_flight() helper methods. * Use ring->queued tracking in the command queue to set/clear the cmd_hold_nic_awake bit at the right points. Taken-From: Linux iwlwifi Obtained from: DragonflyBSD commit ce43f57f5308b579ea21e8a5a29969114ba2247d
2017-02-08 06:53:23 +00:00
if (((pkt->hdr.idx + ring->queued) % IWM_TX_RING_COUNT) != ring->cur) {
device_printf(sc->sc_dev,
"%s: Some HCMDs skipped?: idx=%d queued=%d cur=%d\n",
__func__, pkt->hdr.idx, ring->queued, ring->cur);
/* XXX call iwm_force_nmi() */
}
KASSERT(ring->queued > 0, ("ring->queued is empty?"));
ring->queued--;
if (ring->queued == 0)
iwm_pcie_clear_cmd_in_flight(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
#if 0
/*
* necessary only for block ack mode
*/
void
iwm_update_sched(struct iwm_softc *sc, int qid, int idx, uint8_t sta_id,
uint16_t len)
{
struct iwm_agn_scd_bc_tbl *scd_bc_tbl;
uint16_t w_val;
scd_bc_tbl = sc->sched_dma.vaddr;
len += 8; /* magic numbers came naturally from paris */
[iwm] Remove support for fw older than -17 and -22 * iwm(4) didn't use any of these definitions yet, anyway. Obtained from: dragonflybsd.git f95003b8f1f7382c8396a6d408e3072632afdd3d
2017-06-04 21:28:52 +00:00
len = roundup(len, 4) / 4;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
w_val = htole16(sta_id << 12 | len);
/* Update TX scheduler. */
scd_bc_tbl[qid].tfd_offset[idx] = w_val;
bus_dmamap_sync(sc->sched_dma.tag, sc->sched_dma.map,
BUS_DMASYNC_PREWRITE);
/* I really wonder what this is ?!? */
if (idx < IWM_TFD_QUEUE_SIZE_BC_DUP) {
scd_bc_tbl[qid].tfd_offset[IWM_TFD_QUEUE_SIZE_MAX + idx] = w_val;
bus_dmamap_sync(sc->sched_dma.tag, sc->sched_dma.map,
BUS_DMASYNC_PREWRITE);
}
}
#endif
[iwm] fix default antenna selection a bit; fix 5GHz rate control handling. * Don't do the antenna switching when setting up the rate table - we don't take into account whether it's an active antenna or not (eg shared with BT.) I'll look into this a bit more later. * The default antenna is still 1, I'll look into that a bit more later. (So no, this doesn't fix it for Larry who needs ANT_B to be active, not ANT_A.) * My changes to the rate control setup used the wrong method for finding a suitable rate, which led to 1M CCK frames being queued for 11a operation. This is .. sub-optimal. Change the rate control lookup path to use the global table instead of the per-node table, which won't be setup until the node rate table is setup (which happens way too late in this game.) Tested: * Intel 7260, 2G and 5G operation.
2016-09-24 01:21:42 +00:00
static int
iwm_tx_rateidx_global_lookup(struct iwm_softc *sc, uint8_t rate)
{
int i;
for (i = 0; i < nitems(iwm_rates); i++) {
if (iwm_rates[i].rate == rate)
return (i);
}
/* XXX error? */
IWM_DPRINTF(sc, IWM_DEBUG_XMIT | IWM_DEBUG_TXRATE,
"%s: couldn't find an entry for rate=%d\n",
__func__,
rate);
return (0);
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/*
[iwn] fill in the TX command rate info flags consistently. The Linux driver sets the rate_n_flags regardless of whether it's being sent using firmware rate control or local rate control. This includes the antenna configuration. Thanks to Kyle Evans <kevans91@ksu.edu> for pointing this out to me and doing some investigation/testing on his end. Tested: * Intel 7260 STA, 2G and 5G networks
2016-05-01 05:01:10 +00:00
* Fill in the rate related information for a transmit command.
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
*/
static const struct iwm_rate *
iwm_tx_fill_cmd(struct iwm_softc *sc, struct iwm_node *in,
[iwm] use rate control info from the node txrates; use mgmtrate for EAPOL frames This changes the transmit rate control code to do a few things: * use fixed rates (mcast, ucast, mgmt) where required. * Don't use a hard-coded 11a or 11bg rate for non-data frames - use what net80211 says we should use. * use mgmtrate for EAPOL frames. Reviewed by: avos Differential Revision: https://reviews.freebsd.org/D7994
2016-09-21 20:56:10 +00:00
struct mbuf *m, struct iwm_tx_cmd *tx)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
struct ieee80211_node *ni = &in->in_ni;
[iwm] use rate control info from the node txrates; use mgmtrate for EAPOL frames This changes the transmit rate control code to do a few things: * use fixed rates (mcast, ucast, mgmt) where required. * Don't use a hard-coded 11a or 11bg rate for non-data frames - use what net80211 says we should use. * use mgmtrate for EAPOL frames. Reviewed by: avos Differential Revision: https://reviews.freebsd.org/D7994
2016-09-21 20:56:10 +00:00
struct ieee80211_frame *wh;
const struct ieee80211_txparam *tp = ni->ni_txparms;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
const struct iwm_rate *rinfo;
[iwm] use rate control info from the node txrates; use mgmtrate for EAPOL frames This changes the transmit rate control code to do a few things: * use fixed rates (mcast, ucast, mgmt) where required. * Don't use a hard-coded 11a or 11bg rate for non-data frames - use what net80211 says we should use. * use mgmtrate for EAPOL frames. Reviewed by: avos Differential Revision: https://reviews.freebsd.org/D7994
2016-09-21 20:56:10 +00:00
int type;
[iwm] fix default antenna selection a bit; fix 5GHz rate control handling. * Don't do the antenna switching when setting up the rate table - we don't take into account whether it's an active antenna or not (eg shared with BT.) I'll look into this a bit more later. * The default antenna is still 1, I'll look into that a bit more later. (So no, this doesn't fix it for Larry who needs ANT_B to be active, not ANT_A.) * My changes to the rate control setup used the wrong method for finding a suitable rate, which led to 1M CCK frames being queued for 11a operation. This is .. sub-optimal. Change the rate control lookup path to use the global table instead of the per-node table, which won't be setup until the node rate table is setup (which happens way too late in this game.) Tested: * Intel 7260, 2G and 5G operation.
2016-09-24 01:21:42 +00:00
int ridx, rate_flags;
[iwm] use rate control info from the node txrates; use mgmtrate for EAPOL frames This changes the transmit rate control code to do a few things: * use fixed rates (mcast, ucast, mgmt) where required. * Don't use a hard-coded 11a or 11bg rate for non-data frames - use what net80211 says we should use. * use mgmtrate for EAPOL frames. Reviewed by: avos Differential Revision: https://reviews.freebsd.org/D7994
2016-09-21 20:56:10 +00:00
wh = mtod(m, struct ieee80211_frame *);
type = wh->i_fc[0] & IEEE80211_FC0_TYPE_MASK;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
tx->rts_retry_limit = IWM_RTS_DFAULT_RETRY_LIMIT;
tx->data_retry_limit = IWM_DEFAULT_TX_RETRY;
net80211 drivers: fix rate setup for EAPOL frames, obtain Tx parameters directly from the node. - Use ni_txparms directly instead of calculating them manually every time - Move M_EAPOL flag check upper; otherwise it may be skipped due to 'ucastrate' / 'mcastrate' check - Use 'mgtrate' for control frames too (see ifconfig(8), mgtrate parameter) - Add few more M_EAPOL checks where it was missing (zyd(4), ural(4), urtw(4)) - Few unrelated cleanups Tested with: - Intel 6205 (iwn(4)), STA mode; - WUSB54GC (rum(4)), HOSTAP mode + RTL8188EU (rtwn(4)), STA mode. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D9811
2017-02-26 20:49:35 +00:00
if (type == IEEE80211_FC0_TYPE_MGT ||
type == IEEE80211_FC0_TYPE_CTL ||
(m->m_flags & M_EAPOL) != 0) {
[iwm] fix default antenna selection a bit; fix 5GHz rate control handling. * Don't do the antenna switching when setting up the rate table - we don't take into account whether it's an active antenna or not (eg shared with BT.) I'll look into this a bit more later. * The default antenna is still 1, I'll look into that a bit more later. (So no, this doesn't fix it for Larry who needs ANT_B to be active, not ANT_A.) * My changes to the rate control setup used the wrong method for finding a suitable rate, which led to 1M CCK frames being queued for 11a operation. This is .. sub-optimal. Change the rate control lookup path to use the global table instead of the per-node table, which won't be setup until the node rate table is setup (which happens way too late in this game.) Tested: * Intel 7260, 2G and 5G operation.
2016-09-24 01:21:42 +00:00
ridx = iwm_tx_rateidx_global_lookup(sc, tp->mgmtrate);
IWM_DPRINTF(sc, IWM_DEBUG_TXRATE,
"%s: MGT (%d)\n", __func__, tp->mgmtrate);
[iwm] use rate control info from the node txrates; use mgmtrate for EAPOL frames This changes the transmit rate control code to do a few things: * use fixed rates (mcast, ucast, mgmt) where required. * Don't use a hard-coded 11a or 11bg rate for non-data frames - use what net80211 says we should use. * use mgmtrate for EAPOL frames. Reviewed by: avos Differential Revision: https://reviews.freebsd.org/D7994
2016-09-21 20:56:10 +00:00
} else if (IEEE80211_IS_MULTICAST(wh->i_addr1)) {
[iwm] fix default antenna selection a bit; fix 5GHz rate control handling. * Don't do the antenna switching when setting up the rate table - we don't take into account whether it's an active antenna or not (eg shared with BT.) I'll look into this a bit more later. * The default antenna is still 1, I'll look into that a bit more later. (So no, this doesn't fix it for Larry who needs ANT_B to be active, not ANT_A.) * My changes to the rate control setup used the wrong method for finding a suitable rate, which led to 1M CCK frames being queued for 11a operation. This is .. sub-optimal. Change the rate control lookup path to use the global table instead of the per-node table, which won't be setup until the node rate table is setup (which happens way too late in this game.) Tested: * Intel 7260, 2G and 5G operation.
2016-09-24 01:21:42 +00:00
ridx = iwm_tx_rateidx_global_lookup(sc, tp->mcastrate);
IWM_DPRINTF(sc, IWM_DEBUG_TXRATE,
"%s: MCAST (%d)\n", __func__, tp->mcastrate);
[iwm] use rate control info from the node txrates; use mgmtrate for EAPOL frames This changes the transmit rate control code to do a few things: * use fixed rates (mcast, ucast, mgmt) where required. * Don't use a hard-coded 11a or 11bg rate for non-data frames - use what net80211 says we should use. * use mgmtrate for EAPOL frames. Reviewed by: avos Differential Revision: https://reviews.freebsd.org/D7994
2016-09-21 20:56:10 +00:00
} else if (tp->ucastrate != IEEE80211_FIXED_RATE_NONE) {
[iwm] fix default antenna selection a bit; fix 5GHz rate control handling. * Don't do the antenna switching when setting up the rate table - we don't take into account whether it's an active antenna or not (eg shared with BT.) I'll look into this a bit more later. * The default antenna is still 1, I'll look into that a bit more later. (So no, this doesn't fix it for Larry who needs ANT_B to be active, not ANT_A.) * My changes to the rate control setup used the wrong method for finding a suitable rate, which led to 1M CCK frames being queued for 11a operation. This is .. sub-optimal. Change the rate control lookup path to use the global table instead of the per-node table, which won't be setup until the node rate table is setup (which happens way too late in this game.) Tested: * Intel 7260, 2G and 5G operation.
2016-09-24 01:21:42 +00:00
ridx = iwm_tx_rateidx_global_lookup(sc, tp->ucastrate);
IWM_DPRINTF(sc, IWM_DEBUG_TXRATE,
"%s: FIXED_RATE (%d)\n", __func__, tp->ucastrate);
net80211 drivers: fix rate setup for EAPOL frames, obtain Tx parameters directly from the node. - Use ni_txparms directly instead of calculating them manually every time - Move M_EAPOL flag check upper; otherwise it may be skipped due to 'ucastrate' / 'mcastrate' check - Use 'mgtrate' for control frames too (see ifconfig(8), mgtrate parameter) - Add few more M_EAPOL checks where it was missing (zyd(4), ural(4), urtw(4)) - Few unrelated cleanups Tested with: - Intel 6205 (iwn(4)), STA mode; - WUSB54GC (rum(4)), HOSTAP mode + RTL8188EU (rtwn(4)), STA mode. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D9811
2017-02-26 20:49:35 +00:00
} else {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* for data frames, use RS table */
[iwm] fix default antenna selection a bit; fix 5GHz rate control handling. * Don't do the antenna switching when setting up the rate table - we don't take into account whether it's an active antenna or not (eg shared with BT.) I'll look into this a bit more later. * The default antenna is still 1, I'll look into that a bit more later. (So no, this doesn't fix it for Larry who needs ANT_B to be active, not ANT_A.) * My changes to the rate control setup used the wrong method for finding a suitable rate, which led to 1M CCK frames being queued for 11a operation. This is .. sub-optimal. Change the rate control lookup path to use the global table instead of the per-node table, which won't be setup until the node rate table is setup (which happens way too late in this game.) Tested: * Intel 7260, 2G and 5G operation.
2016-09-24 01:21:42 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_TXRATE, "%s: DATA\n", __func__);
if_iwm - Update firmware rs table, instead of indexing the table in tx cmds. * Rather than providing a non-zero index into the firmware RS table, we should always use index 0 and update the firmware RS table whenever our chosen tx rate for data-frames changes. * Send IWM_LQ_CMD updates when the tx rate gets updated by the net80211 rate control (which is after we tell the tx status to the net80211 rate-control in iwm_mvm_rx_tx_cmd_single()). * Disregard frames transferred with a different tx rate than the currently selected rate for the rate-control calculations. This way we avoid counting management frames (which are sent at a slow, and fixed rate), as well as frames we added to the tx queue just before a new IWM_LQ_CMD update took effect. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (5d6b465e288ac5b52d7115688d4e6516acbbea1c)
2019-01-24 03:41:09 +00:00
ridx = iwm_rate2ridx(sc, ni->ni_txrate);
if (ridx == -1)
ridx = 0;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* This is the index into the programmed table */
if_iwm - Update firmware rs table, instead of indexing the table in tx cmds. * Rather than providing a non-zero index into the firmware RS table, we should always use index 0 and update the firmware RS table whenever our chosen tx rate for data-frames changes. * Send IWM_LQ_CMD updates when the tx rate gets updated by the net80211 rate control (which is after we tell the tx status to the net80211 rate-control in iwm_mvm_rx_tx_cmd_single()). * Disregard frames transferred with a different tx rate than the currently selected rate for the rate-control calculations. This way we avoid counting management frames (which are sent at a slow, and fixed rate), as well as frames we added to the tx queue just before a new IWM_LQ_CMD update took effect. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (5d6b465e288ac5b52d7115688d4e6516acbbea1c)
2019-01-24 03:41:09 +00:00
tx->initial_rate_index = 0;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
tx->tx_flags |= htole32(IWM_TX_CMD_FLG_STA_RATE);
}
[iwm] fix default antenna selection a bit; fix 5GHz rate control handling. * Don't do the antenna switching when setting up the rate table - we don't take into account whether it's an active antenna or not (eg shared with BT.) I'll look into this a bit more later. * The default antenna is still 1, I'll look into that a bit more later. (So no, this doesn't fix it for Larry who needs ANT_B to be active, not ANT_A.) * My changes to the rate control setup used the wrong method for finding a suitable rate, which led to 1M CCK frames being queued for 11a operation. This is .. sub-optimal. Change the rate control lookup path to use the global table instead of the per-node table, which won't be setup until the node rate table is setup (which happens way too late in this game.) Tested: * Intel 7260, 2G and 5G operation.
2016-09-24 01:21:42 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_XMIT | IWM_DEBUG_TXRATE,
"%s: frame type=%d txrate %d\n",
__func__, type, iwm_rates[ridx].rate);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
rinfo = &iwm_rates[ridx];
IWM_DPRINTF(sc, IWM_DEBUG_TXRATE, "%s: ridx=%d; rate=%d, CCK=%d\n",
__func__, ridx,
rinfo->rate,
!! (IWM_RIDX_IS_CCK(ridx))
);
/* XXX TODO: hard-coded TX antenna? */
rate_flags = 1 << IWM_RATE_MCS_ANT_POS;
if (IWM_RIDX_IS_CCK(ridx))
rate_flags |= IWM_RATE_MCS_CCK_MSK;
tx->rate_n_flags = htole32(rate_flags | rinfo->plcp);
return rinfo;
}
#define TB0_SIZE 16
static int
iwm_tx(struct iwm_softc *sc, struct mbuf *m, struct ieee80211_node *ni, int ac)
{
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
struct ieee80211com *ic = &sc->sc_ic;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
struct ieee80211vap *vap = TAILQ_FIRST(&ic->ic_vaps);
iwm: reorganize if_iwmvar.h - Change order of data in if_iwmvar.h (like it is in other drivers: defines, data structures, vap/node structures, softc struct and locks); use indentation. - Fix IWM_LOCK(_sc) / IWM_UNLOCK(_sc) macro. - Add IWM_LOCK_INIT / DESTROY(sc) + fix mtx_init() usage. - Wrap iwm_node casts into IWM_NODE() macro. - Drop some fields: * wt_hwqueue from Tx radiotap header; * macaddr[6] from iwm_vap; Approved by: adrian Differential Revision: https://reviews.freebsd.org/D4753
2016-01-03 10:06:10 +00:00
struct iwm_node *in = IWM_NODE(ni);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
struct iwm_tx_ring *ring;
struct iwm_tx_data *data;
struct iwm_tfd *desc;
struct iwm_device_cmd *cmd;
struct iwm_tx_cmd *tx;
struct ieee80211_frame *wh;
struct ieee80211_key *k = NULL;
struct mbuf *m1;
const struct iwm_rate *rinfo;
uint32_t flags;
u_int hdrlen;
bus_dma_segment_t *seg, segs[IWM_MAX_SCATTER];
int nsegs;
uint8_t tid, type;
int i, totlen, error, pad;
wh = mtod(m, struct ieee80211_frame *);
hdrlen = ieee80211_anyhdrsize(wh);
type = wh->i_fc[0] & IEEE80211_FC0_TYPE_MASK;
tid = 0;
ring = &sc->txq[ac];
desc = &ring->desc[ring->cur];
memset(desc, 0, sizeof(*desc));
data = &ring->data[ring->cur];
/* Fill out iwm_tx_cmd to send to the firmware */
cmd = &ring->cmd[ring->cur];
cmd->hdr.code = IWM_TX_CMD;
cmd->hdr.flags = 0;
cmd->hdr.qid = ring->qid;
cmd->hdr.idx = ring->cur;
tx = (void *)cmd->data;
memset(tx, 0, sizeof(*tx));
[iwm] use rate control info from the node txrates; use mgmtrate for EAPOL frames This changes the transmit rate control code to do a few things: * use fixed rates (mcast, ucast, mgmt) where required. * Don't use a hard-coded 11a or 11bg rate for non-data frames - use what net80211 says we should use. * use mgmtrate for EAPOL frames. Reviewed by: avos Differential Revision: https://reviews.freebsd.org/D7994
2016-09-21 20:56:10 +00:00
rinfo = iwm_tx_fill_cmd(sc, in, m, tx);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Encrypt the frame if need be. */
if (wh->i_fc[1] & IEEE80211_FC1_PROTECTED) {
/* Retrieve key for TX && do software encryption. */
k = ieee80211_crypto_encap(ni, m);
if (k == NULL) {
m_freem(m);
return (ENOBUFS);
}
/* 802.11 header may have moved. */
wh = mtod(m, struct ieee80211_frame *);
}
if (ieee80211_radiotap_active_vap(vap)) {
struct iwm_tx_radiotap_header *tap = &sc->sc_txtap;
tap->wt_flags = 0;
tap->wt_chan_freq = htole16(ni->ni_chan->ic_freq);
tap->wt_chan_flags = htole16(ni->ni_chan->ic_flags);
tap->wt_rate = rinfo->rate;
if (k != NULL)
tap->wt_flags |= IEEE80211_RADIOTAP_F_WEP;
ieee80211_radiotap_tx(vap, m);
}
totlen = m->m_pkthdr.len;
flags = 0;
if (!IEEE80211_IS_MULTICAST(wh->i_addr1)) {
flags |= IWM_TX_CMD_FLG_ACK;
}
[iwm] Fix inverted logic in iwm_tx(). The PROT_REQUIRE flag in should be set for data frames above a certain length, but we were setting it for !data frames above a certain length, which makes no sense at all. Taken-From: OpenBSD, Linux iwlwifi Approved by: adrian (mentor) Obtained from: DragonFlyBSD git 8cc03924a36c572c2908e659e624f44636dc2b33 Differential Revision: https://reviews.freebsd.org/D7323
2016-07-27 20:43:08 +00:00
if (type == IEEE80211_FC0_TYPE_DATA
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
&& (totlen + IEEE80211_CRC_LEN > vap->iv_rtsthreshold)
&& !IEEE80211_IS_MULTICAST(wh->i_addr1)) {
flags |= IWM_TX_CMD_FLG_PROT_REQUIRE;
}
if (IEEE80211_IS_MULTICAST(wh->i_addr1) ||
type != IEEE80211_FC0_TYPE_DATA)
tx->sta_id = sc->sc_aux_sta.sta_id;
else
tx->sta_id = IWM_STATION_ID;
if (type == IEEE80211_FC0_TYPE_MGT) {
uint8_t subtype = wh->i_fc[0] & IEEE80211_FC0_SUBTYPE_MASK;
if (subtype == IEEE80211_FC0_SUBTYPE_ASSOC_REQ ||
[iwm] Set different pm_timeout for action frames. When building a Tx Command for management frames, we are lacking a check for action frames, for which we should set a different pm_timeout. This cause the fw to stay awake for 100TU after each such frame is transmitted, resulting an excessive power consumption. Taken-From: Linux iwlwifi (git b084a35663c3f1f7) Approved by: adrian (mentor) Obtained from: Linux git b084a35663c3f1f7de1c45c4ae3006864c940fe7 Obtained from: DragonFlyBSD git ba00f0e3ae873d6f0d5743e22c3ebc49c44dfdac Differential Revision: https://reviews.freebsd.org/D7324
2016-07-27 20:46:51 +00:00
subtype == IEEE80211_FC0_SUBTYPE_REASSOC_REQ) {
tx->pm_frame_timeout = htole16(IWM_PM_FRAME_ASSOC);
} else if (subtype == IEEE80211_FC0_SUBTYPE_ACTION) {
tx->pm_frame_timeout = htole16(IWM_PM_FRAME_NONE);
} else {
tx->pm_frame_timeout = htole16(IWM_PM_FRAME_MGMT);
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
} else {
[iwm] Set different pm_timeout for action frames. When building a Tx Command for management frames, we are lacking a check for action frames, for which we should set a different pm_timeout. This cause the fw to stay awake for 100TU after each such frame is transmitted, resulting an excessive power consumption. Taken-From: Linux iwlwifi (git b084a35663c3f1f7) Approved by: adrian (mentor) Obtained from: Linux git b084a35663c3f1f7de1c45c4ae3006864c940fe7 Obtained from: DragonFlyBSD git ba00f0e3ae873d6f0d5743e22c3ebc49c44dfdac Differential Revision: https://reviews.freebsd.org/D7324
2016-07-27 20:46:51 +00:00
tx->pm_frame_timeout = htole16(IWM_PM_FRAME_NONE);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
if (hdrlen & 3) {
/* First segment length must be a multiple of 4. */
flags |= IWM_TX_CMD_FLG_MH_PAD;
pad = 4 - (hdrlen & 3);
} else
pad = 0;
tx->driver_txop = 0;
tx->next_frame_len = 0;
tx->len = htole16(totlen);
tx->tid_tspec = tid;
tx->life_time = htole32(IWM_TX_CMD_LIFE_TIME_INFINITE);
/* Set physical address of "scratch area". */
tx->dram_lsb_ptr = htole32(data->scratch_paddr);
tx->dram_msb_ptr = iwm_get_dma_hi_addr(data->scratch_paddr);
/* Copy 802.11 header in TX command. */
memcpy(((uint8_t *)tx) + sizeof(*tx), wh, hdrlen);
flags |= IWM_TX_CMD_FLG_BT_DIS | IWM_TX_CMD_FLG_SEQ_CTL;
tx->sec_ctl = 0;
tx->tx_flags |= htole32(flags);
/* Trim 802.11 header. */
m_adj(m, hdrlen);
error = bus_dmamap_load_mbuf_sg(ring->data_dmat, data->map, m,
segs, &nsegs, BUS_DMA_NOWAIT);
if (error != 0) {
if (error != EFBIG) {
device_printf(sc->sc_dev, "can't map mbuf (error %d)\n",
error);
m_freem(m);
return error;
}
/* Too many DMA segments, linearize mbuf. */
iwm: use m_collapse() to defragment a mbuf chain - Simplify defragmentation code. - Use proper number of dma segments for data. Approved by: adrian (mentor) Obtained from: DragonFlyBSD (mostly) Differential Revision: https://reviews.freebsd.org/D4754
2016-01-03 21:32:47 +00:00
m1 = m_collapse(m, M_NOWAIT, IWM_MAX_SCATTER - 2);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (m1 == NULL) {
iwm: use m_collapse() to defragment a mbuf chain - Simplify defragmentation code. - Use proper number of dma segments for data. Approved by: adrian (mentor) Obtained from: DragonFlyBSD (mostly) Differential Revision: https://reviews.freebsd.org/D4754
2016-01-03 21:32:47 +00:00
device_printf(sc->sc_dev,
"%s: could not defrag mbuf\n", __func__);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
m_freem(m);
iwm: use m_collapse() to defragment a mbuf chain - Simplify defragmentation code. - Use proper number of dma segments for data. Approved by: adrian (mentor) Obtained from: DragonFlyBSD (mostly) Differential Revision: https://reviews.freebsd.org/D4754
2016-01-03 21:32:47 +00:00
return (ENOBUFS);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
m = m1;
iwm: use m_collapse() to defragment a mbuf chain - Simplify defragmentation code. - Use proper number of dma segments for data. Approved by: adrian (mentor) Obtained from: DragonFlyBSD (mostly) Differential Revision: https://reviews.freebsd.org/D4754
2016-01-03 21:32:47 +00:00
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
error = bus_dmamap_load_mbuf_sg(ring->data_dmat, data->map, m,
segs, &nsegs, BUS_DMA_NOWAIT);
if (error != 0) {
device_printf(sc->sc_dev, "can't map mbuf (error %d)\n",
error);
m_freem(m);
return error;
}
}
data->m = m;
data->in = in;
data->done = 0;
IWM_DPRINTF(sc, IWM_DEBUG_XMIT,
"sending txd %p, in %p\n", data, data->in);
KASSERT(data->in != NULL, ("node is NULL"));
IWM_DPRINTF(sc, IWM_DEBUG_XMIT,
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
"sending data: qid=%d idx=%d len=%d nsegs=%d txflags=0x%08x rate_n_flags=0x%08x rateidx=%u\n",
[iwm] add extra transmit setup/completion logging so I can see what's going on. I'm seeing 5GHz association work but data not work until the rate drops, so I need way more information about what's being programmed into the transmit descriptors. Tested: * 7260AC, STA mode
2016-04-26 03:24:28 +00:00
ring->qid, ring->cur, totlen, nsegs,
le32toh(tx->tx_flags),
le32toh(tx->rate_n_flags),
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
tx->initial_rate_index
[iwm] add extra transmit setup/completion logging so I can see what's going on. I'm seeing 5GHz association work but data not work until the rate drops, so I need way more information about what's being programmed into the transmit descriptors. Tested: * 7260AC, STA mode
2016-04-26 03:24:28 +00:00
);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Fill TX descriptor. */
desc->num_tbs = 2 + nsegs;
desc->tbs[0].lo = htole32(data->cmd_paddr);
desc->tbs[0].hi_n_len = htole16(iwm_get_dma_hi_addr(data->cmd_paddr)) |
(TB0_SIZE << 4);
desc->tbs[1].lo = htole32(data->cmd_paddr + TB0_SIZE);
desc->tbs[1].hi_n_len = htole16(iwm_get_dma_hi_addr(data->cmd_paddr)) |
((sizeof(struct iwm_cmd_header) + sizeof(*tx)
+ hdrlen + pad - TB0_SIZE) << 4);
/* Other DMA segments are for data payload. */
for (i = 0; i < nsegs; i++) {
seg = &segs[i];
desc->tbs[i+2].lo = htole32(seg->ds_addr);
desc->tbs[i+2].hi_n_len = \
htole16(iwm_get_dma_hi_addr(seg->ds_addr))
| ((seg->ds_len) << 4);
}
bus_dmamap_sync(ring->data_dmat, data->map,
BUS_DMASYNC_PREWRITE);
bus_dmamap_sync(ring->cmd_dma.tag, ring->cmd_dma.map,
BUS_DMASYNC_PREWRITE);
bus_dmamap_sync(ring->desc_dma.tag, ring->desc_dma.map,
BUS_DMASYNC_PREWRITE);
#if 0
iwm_update_sched(sc, ring->qid, ring->cur, tx->sta_id, le16toh(tx->len));
#endif
/* Kick TX ring. */
ring->cur = (ring->cur + 1) % IWM_TX_RING_COUNT;
IWM_WRITE(sc, IWM_HBUS_TARG_WRPTR, ring->qid << 8 | ring->cur);
/* Mark TX ring as full if we reach a certain threshold. */
if (++ring->queued > IWM_TX_RING_HIMARK) {
sc->qfullmsk |= 1 << ring->qid;
}
return 0;
}
static int
iwm_raw_xmit(struct ieee80211_node *ni, struct mbuf *m,
const struct ieee80211_bpf_params *params)
{
struct ieee80211com *ic = ni->ni_ic;
Convert more ifp->if_softc -> ic->ic_softc. These should be a big no-op. Tested: * make universe
2015-08-17 23:44:38 +00:00
struct iwm_softc *sc = ic->ic_softc;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
int error = 0;
IWM_DPRINTF(sc, IWM_DEBUG_XMIT,
"->%s begin\n", __func__);
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
if ((sc->sc_flags & IWM_FLAG_HW_INITED) == 0) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
m_freem(m);
IWM_DPRINTF(sc, IWM_DEBUG_XMIT,
"<-%s not RUNNING\n", __func__);
return (ENETDOWN);
}
IWM_LOCK(sc);
/* XXX fix this */
if (params == NULL) {
error = iwm_tx(sc, m, ni, 0);
} else {
error = iwm_tx(sc, m, ni, 0);
}
if_iwm - Stop iwm_watchdog callout when idle. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (6a8683b0e9d734f23bd9647e117da198c2b9a74e)
2019-01-24 03:48:50 +00:00
if (sc->sc_tx_timer == 0)
callout_reset(&sc->sc_watchdog_to, hz, iwm_watchdog, sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
sc->sc_tx_timer = 5;
IWM_UNLOCK(sc);
return (error);
}
/*
* mvm/tx.c
*/
/*
* Note that there are transports that buffer frames before they reach
* the firmware. This means that after flush_tx_path is called, the
* queue might not be empty. The race-free way to handle this is to:
* 1) set the station as draining
* 2) flush the Tx path
* 3) wait for the transport queues to be empty
*/
int
[iwm] Improve reliability of iwm_release on disassociation a bit. * We need to first call ivp->iv_newstate(), to enqueue the deauth/deassoc mgmt frame, then flush the tx queue, before actually calling iwm_release(). * cycling a wlan connection via wpa_cli frontend to wpa_supplicant, by issuing disconnect and reconnect commandos works quite well. (There is still an issue when disconnecting/reconnecting too quickly) * Reassociating or roaming via wpa_supplicant is still broken. Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D7943
2016-09-19 22:36:30 +00:00
iwm_mvm_flush_tx_path(struct iwm_softc *sc, uint32_t tfd_msk, uint32_t flags)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
[iwm] Improve reliability of iwm_release on disassociation a bit. * We need to first call ivp->iv_newstate(), to enqueue the deauth/deassoc mgmt frame, then flush the tx queue, before actually calling iwm_release(). * cycling a wlan connection via wpa_cli frontend to wpa_supplicant, by issuing disconnect and reconnect commandos works quite well. (There is still an issue when disconnecting/reconnecting too quickly) * Reassociating or roaming via wpa_supplicant is still broken. Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D7943
2016-09-19 22:36:30 +00:00
int ret;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
struct iwm_tx_path_flush_cmd flush_cmd = {
.queues_ctl = htole32(tfd_msk),
.flush_ctl = htole16(IWM_DUMP_TX_FIFO_FLUSH),
};
[iwm] Improve reliability of iwm_release on disassociation a bit. * We need to first call ivp->iv_newstate(), to enqueue the deauth/deassoc mgmt frame, then flush the tx queue, before actually calling iwm_release(). * cycling a wlan connection via wpa_cli frontend to wpa_supplicant, by issuing disconnect and reconnect commandos works quite well. (There is still an issue when disconnecting/reconnecting too quickly) * Reassociating or roaming via wpa_supplicant is still broken. Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D7943
2016-09-19 22:36:30 +00:00
ret = iwm_mvm_send_cmd_pdu(sc, IWM_TXPATH_FLUSH, flags,
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
sizeof(flush_cmd), &flush_cmd);
if (ret)
device_printf(sc->sc_dev,
"Flushing tx queue failed: %d\n", ret);
return ret;
}
/*
* BEGIN mvm/quota.c
*/
static int
[iwm] Move in_phyctxt from struct iwm_node to phy_ctxt in struct iwm_vap. * This better matches how things are organized in Linux's iwlwifi. Obtained from: dragonflybsd.git 0cf16dd2e0e09a3e5140e50222ac2e69bcdb19a2
2017-05-09 05:32:35 +00:00
iwm_mvm_update_quotas(struct iwm_softc *sc, struct iwm_vap *ivp)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
struct iwm_time_quota_cmd cmd;
int i, idx, ret, num_active_macs, quota, quota_rem;
int colors[IWM_MAX_BINDINGS] = { -1, -1, -1, -1, };
int n_ifs[IWM_MAX_BINDINGS] = {0, };
uint16_t id;
memset(&cmd, 0, sizeof(cmd));
/* currently, PHY ID == binding ID */
[iwm] Move in_phyctxt from struct iwm_node to phy_ctxt in struct iwm_vap. * This better matches how things are organized in Linux's iwlwifi. Obtained from: dragonflybsd.git 0cf16dd2e0e09a3e5140e50222ac2e69bcdb19a2
2017-05-09 05:32:35 +00:00
if (ivp) {
id = ivp->phy_ctxt->id;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
KASSERT(id < IWM_MAX_BINDINGS, ("invalid id"));
[iwm] Move in_phyctxt from struct iwm_node to phy_ctxt in struct iwm_vap. * This better matches how things are organized in Linux's iwlwifi. Obtained from: dragonflybsd.git 0cf16dd2e0e09a3e5140e50222ac2e69bcdb19a2
2017-05-09 05:32:35 +00:00
colors[id] = ivp->phy_ctxt->color;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (1)
n_ifs[id] = 1;
}
/*
* The FW's scheduling session consists of
* IWM_MVM_MAX_QUOTA fragments. Divide these fragments
* equally between all the bindings that require quota
*/
num_active_macs = 0;
for (i = 0; i < IWM_MAX_BINDINGS; i++) {
cmd.quotas[i].id_and_color = htole32(IWM_FW_CTXT_INVALID);
num_active_macs += n_ifs[i];
}
quota = 0;
quota_rem = 0;
if (num_active_macs) {
quota = IWM_MVM_MAX_QUOTA / num_active_macs;
quota_rem = IWM_MVM_MAX_QUOTA % num_active_macs;
}
for (idx = 0, i = 0; i < IWM_MAX_BINDINGS; i++) {
if (colors[i] < 0)
continue;
cmd.quotas[idx].id_and_color =
htole32(IWM_FW_CMD_ID_AND_COLOR(i, colors[i]));
if (n_ifs[i] <= 0) {
cmd.quotas[idx].quota = htole32(0);
cmd.quotas[idx].max_duration = htole32(0);
} else {
cmd.quotas[idx].quota = htole32(quota * n_ifs[i]);
cmd.quotas[idx].max_duration = htole32(0);
}
idx++;
}
/* Give the remainder of the session to the first binding */
cmd.quotas[0].quota = htole32(le32toh(cmd.quotas[0].quota) + quota_rem);
ret = iwm_mvm_send_cmd_pdu(sc, IWM_TIME_QUOTA_CMD, IWM_CMD_SYNC,
sizeof(cmd), &cmd);
if (ret)
device_printf(sc->sc_dev,
"%s: Failed to send quota: %d\n", __func__, ret);
return ret;
}
/*
* END mvm/quota.c
*/
/*
* ieee80211 routines
*/
/*
* Change to AUTH state in 80211 state machine. Roughly matches what
* Linux does in bss_info_changed().
*/
static int
iwm_auth(struct ieee80211vap *vap, struct iwm_softc *sc)
{
struct ieee80211_node *ni;
struct iwm_node *in;
struct iwm_vap *iv = IWM_VAP(vap);
uint32_t duration;
int error;
/*
* XXX i have a feeling that the vap node is being
* freed from underneath us. Grr.
*/
ni = ieee80211_ref_node(vap->iv_bss);
iwm: reorganize if_iwmvar.h - Change order of data in if_iwmvar.h (like it is in other drivers: defines, data structures, vap/node structures, softc struct and locks); use indentation. - Fix IWM_LOCK(_sc) / IWM_UNLOCK(_sc) macro. - Add IWM_LOCK_INIT / DESTROY(sc) + fix mtx_init() usage. - Wrap iwm_node casts into IWM_NODE() macro. - Drop some fields: * wt_hwqueue from Tx radiotap header; * macaddr[6] from iwm_vap; Approved by: adrian Differential Revision: https://reviews.freebsd.org/D4753
2016-01-03 10:06:10 +00:00
in = IWM_NODE(ni);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_RESET | IWM_DEBUG_STATE,
"%s: called; vap=%p, bss ni=%p\n",
__func__,
vap,
ni);
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_STATE, "%s: Current node bssid: %s\n",
__func__, ether_sprintf(ni->ni_bssid));
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
in->in_assoc = 0;
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
iv->iv_auth = 1;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Refuse connection to APs with beacon interval < 16. Obtained from: dragonflybsd.git aba448de727e9b122adadeb36fd00a8ad6018d4f
2017-05-12 06:16:11 +00:00
/*
* Firmware bug - it'll crash if the beacon interval is less
* than 16. We can't avoid connecting at all, so refuse the
* station state change, this will cause net80211 to abandon
* attempts to connect to this AP, and eventually wpa_s will
* blacklist the AP...
*/
if (ni->ni_intval < 16) {
device_printf(sc->sc_dev,
"AP %s beacon interval is %d, refusing due to firmware bug!\n",
ether_sprintf(ni->ni_bssid), ni->ni_intval);
error = EINVAL;
goto out;
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
error = iwm_allow_mcast(vap, sc);
if (error) {
device_printf(sc->sc_dev,
"%s: failed to set multicast\n", __func__);
goto out;
}
/*
* This is where it deviates from what Linux does.
*
* Linux iwlwifi doesn't reset the nic each time, nor does it
* call ctxt_add() here. Instead, it adds it during vap creation,
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
* and always does a mac_ctx_changed().
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
*
* The openbsd port doesn't attempt to do that - it reset things
* at odd states and does the add here.
*
* So, until the state handling is fixed (ie, we never reset
* the NIC except for a firmware failure, which should drag
* the NIC back to IDLE, re-setup and re-add all the mac/phy
* contexts that are required), let's do a dirty hack here.
*/
if (iv->is_uploaded) {
if ((error = iwm_mvm_mac_ctxt_changed(sc, vap)) != 0) {
device_printf(sc->sc_dev,
[iwn] fix firmware command use in iwm_auth(). The iwm firmware has separate commands for add, modify and delete for various things (mac, phy context, etc.) The openbsd driver has a habit of just completely resetting the NIC each time, which is technically mostly okay (as long as the reset doesn't actually fail!) but it means a lot of the code is doing ADD when it should do MODIFY. The firmware responds in kind - it just asserts. This fixes auth attempts that occur after the NIC has been already configured. (I'm sure there are more instances of this!) Tested: iwm0: <Intel Dual Band Wireless AC 7260> mem 0xf1400000-0xf1401fff irq 17 at device 0.0 on pci2 iwm0: revision: 0x140, firmware 25.228 (API ver. 9) .. STA mode. Submitted by: Masachika ISHIZUKA <ish@amail.plala.or.jp>
2016-04-25 16:13:04 +00:00
"%s: failed to update MAC\n", __func__);
goto out;
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
} else {
if ((error = iwm_mvm_mac_ctxt_add(sc, vap)) != 0) {
device_printf(sc->sc_dev,
"%s: failed to add MAC\n", __func__);
goto out;
}
[iwm] Deduplicate code in iwm_auth() from an if condition. Obtained from: dragonflybsd.git 03c6e6970115727c9d39f9358e0500ab4f4634cd
2017-05-12 05:53:28 +00:00
}
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
sc->sc_firmware_state = 1;
[iwn] fix firmware command use in iwm_auth(). The iwm firmware has separate commands for add, modify and delete for various things (mac, phy context, etc.) The openbsd driver has a habit of just completely resetting the NIC each time, which is technically mostly okay (as long as the reset doesn't actually fail!) but it means a lot of the code is doing ADD when it should do MODIFY. The firmware responds in kind - it just asserts. This fixes auth attempts that occur after the NIC has been already configured. (I'm sure there are more instances of this!) Tested: iwm0: <Intel Dual Band Wireless AC 7260> mem 0xf1400000-0xf1401fff irq 17 at device 0.0 on pci2 iwm0: revision: 0x140, firmware 25.228 (API ver. 9) .. STA mode. Submitted by: Masachika ISHIZUKA <ish@amail.plala.or.jp>
2016-04-25 16:13:04 +00:00
[iwm] Deduplicate code in iwm_auth() from an if condition. Obtained from: dragonflybsd.git 03c6e6970115727c9d39f9358e0500ab4f4634cd
2017-05-12 05:53:28 +00:00
if ((error = iwm_mvm_phy_ctxt_changed(sc, &sc->sc_phyctxt[0],
in->in_ni.ni_chan, 1, 1)) != 0) {
device_printf(sc->sc_dev,
"%s: failed update phy ctxt\n", __func__);
goto out;
}
iv->phy_ctxt = &sc->sc_phyctxt[0];
if ((error = iwm_mvm_binding_add_vif(sc, iv)) != 0) {
device_printf(sc->sc_dev,
"%s: binding update cmd\n", __func__);
goto out;
}
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
sc->sc_firmware_state = 2;
[iwm] Make powersaving more similar to Linux iwlwifi behaviour. * Add a per-vap ps_disabled flag, and use it for a workaround which fixes an association issue when powersaving is enabled. * Compute flag that should correpsond to the mvmif->bss_conf.ps flag in Linux's iwlwifi (e.g. this disallows powersaving when not associated yet). Inspired-By: Linux iwlwifi Obtained from: dragonflybsd.git dc2e69bdfe8c9d7049c8a28da0adffbfbc6de5c0
2017-05-12 06:31:57 +00:00
/*
* Authentication becomes unreliable when powersaving is left enabled
* here. Powersaving will be activated again when association has
* finished or is aborted.
*/
iv->ps_disabled = TRUE;
error = iwm_mvm_power_update_mac(sc);
iv->ps_disabled = FALSE;
if (error != 0) {
[iwm] Already call iwm_mvm_power_update_mac() during SCAN<->AUTH paths. * Otherwise we would never update powersaving settings until we complete an association, after the first authentication attempt. * This corresponds to what Linux iwlwifi seems to do. Obtained from: dragonflybsd.git aa128dc02a17c2e616232ef0fa997121e969c995
2017-05-12 06:21:03 +00:00
device_printf(sc->sc_dev,
"%s: failed to update power management\n",
__func__);
goto out;
}
[iwm] Deduplicate code in iwm_auth() from an if condition. Obtained from: dragonflybsd.git 03c6e6970115727c9d39f9358e0500ab4f4634cd
2017-05-12 05:53:28 +00:00
if ((error = iwm_mvm_add_sta(sc, in)) != 0) {
device_printf(sc->sc_dev,
"%s: failed to add sta\n", __func__);
goto out;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
sc->sc_firmware_state = 3;
[iwn] fix firmware command use in iwm_auth(). The iwm firmware has separate commands for add, modify and delete for various things (mac, phy context, etc.) The openbsd driver has a habit of just completely resetting the NIC each time, which is technically mostly okay (as long as the reset doesn't actually fail!) but it means a lot of the code is doing ADD when it should do MODIFY. The firmware responds in kind - it just asserts. This fixes auth attempts that occur after the NIC has been already configured. (I'm sure there are more instances of this!) Tested: iwm0: <Intel Dual Band Wireless AC 7260> mem 0xf1400000-0xf1401fff irq 17 at device 0.0 on pci2 iwm0: revision: 0x140, firmware 25.228 (API ver. 9) .. STA mode. Submitted by: Masachika ISHIZUKA <ish@amail.plala.or.jp>
2016-04-25 16:13:04 +00:00
/*
* Prevent the FW from wandering off channel during association
* by "protecting" the session with a time event.
*/
/* XXX duration is in units of TU, not MS */
duration = IWM_MVM_TE_SESSION_PROTECTION_MAX_TIME_MS;
iwm - Improve firmware Time Event handling. * This is a mix of the OpenBSD Git 7fd9664469d1b717a307eebd74aeececbd3c41cc change, and syncing with the Linux iwlwifi code. Taken-From: Linux iwlwifi, and OpenBSD Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (706a3044afd27c3fecfdf57bec1695310e53e228)
2019-01-24 03:47:04 +00:00
iwm_mvm_protect_session(sc, iv, duration, 500 /* XXX magic number */, TRUE);
[iwn] fix firmware command use in iwm_auth(). The iwm firmware has separate commands for add, modify and delete for various things (mac, phy context, etc.) The openbsd driver has a habit of just completely resetting the NIC each time, which is technically mostly okay (as long as the reset doesn't actually fail!) but it means a lot of the code is doing ADD when it should do MODIFY. The firmware responds in kind - it just asserts. This fixes auth attempts that occur after the NIC has been already configured. (I'm sure there are more instances of this!) Tested: iwm0: <Intel Dual Band Wireless AC 7260> mem 0xf1400000-0xf1401fff irq 17 at device 0.0 on pci2 iwm0: revision: 0x140, firmware 25.228 (API ver. 9) .. STA mode. Submitted by: Masachika ISHIZUKA <ish@amail.plala.or.jp>
2016-04-25 16:13:04 +00:00
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
error = 0;
out:
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
if (error != 0)
iv->iv_auth = 0;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
ieee80211_free_node(ni);
return (error);
}
static struct ieee80211_node *
iwm_node_alloc(struct ieee80211vap *vap, const uint8_t mac[IEEE80211_ADDR_LEN])
{
return malloc(sizeof (struct iwm_node), M_80211_NODE,
M_NOWAIT | M_ZERO);
}
if_iwm - Update firmware rs table, instead of indexing the table in tx cmds. * Rather than providing a non-zero index into the firmware RS table, we should always use index 0 and update the firmware RS table whenever our chosen tx rate for data-frames changes. * Send IWM_LQ_CMD updates when the tx rate gets updated by the net80211 rate control (which is after we tell the tx status to the net80211 rate-control in iwm_mvm_rx_tx_cmd_single()). * Disregard frames transferred with a different tx rate than the currently selected rate for the rate-control calculations. This way we avoid counting management frames (which are sent at a slow, and fixed rate), as well as frames we added to the tx queue just before a new IWM_LQ_CMD update took effect. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (5d6b465e288ac5b52d7115688d4e6516acbbea1c)
2019-01-24 03:41:09 +00:00
static uint8_t
iwm_rate_from_ucode_rate(uint32_t rate_n_flags)
{
uint8_t plcp = rate_n_flags & 0xff;
int i;
for (i = 0; i <= IWM_RIDX_MAX; i++) {
if (iwm_rates[i].plcp == plcp)
return iwm_rates[i].rate;
}
return 0;
}
[iwm] Add the BSS's basic rates to iwm's LQ command, not all the rates. Makes the firmware use appropriate Tx rates for ACKs. Obtained from: dragonflybsd.git ab1d3efc208e797c1e09759cd506c95c0aeaa06e
2017-03-25 02:55:13 +00:00
uint8_t
iwm_ridx2rate(struct ieee80211_rateset *rs, int ridx)
{
int i;
uint8_t rval;
for (i = 0; i < rs->rs_nrates; i++) {
rval = (rs->rs_rates[i] & IEEE80211_RATE_VAL);
if (rval == iwm_rates[ridx].rate)
return rs->rs_rates[i];
}
return 0;
}
if_iwm - Update firmware rs table, instead of indexing the table in tx cmds. * Rather than providing a non-zero index into the firmware RS table, we should always use index 0 and update the firmware RS table whenever our chosen tx rate for data-frames changes. * Send IWM_LQ_CMD updates when the tx rate gets updated by the net80211 rate control (which is after we tell the tx status to the net80211 rate-control in iwm_mvm_rx_tx_cmd_single()). * Disregard frames transferred with a different tx rate than the currently selected rate for the rate-control calculations. This way we avoid counting management frames (which are sent at a slow, and fixed rate), as well as frames we added to the tx queue just before a new IWM_LQ_CMD update took effect. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (5d6b465e288ac5b52d7115688d4e6516acbbea1c)
2019-01-24 03:41:09 +00:00
static int
iwm_rate2ridx(struct iwm_softc *sc, uint8_t rate)
{
int i;
for (i = 0; i <= IWM_RIDX_MAX; i++) {
if (iwm_rates[i].rate == rate)
return i;
}
device_printf(sc->sc_dev,
"%s: WARNING: device rate for %u not found!\n",
__func__, rate);
return -1;
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static void
if_iwm - Update firmware rs table, instead of indexing the table in tx cmds. * Rather than providing a non-zero index into the firmware RS table, we should always use index 0 and update the firmware RS table whenever our chosen tx rate for data-frames changes. * Send IWM_LQ_CMD updates when the tx rate gets updated by the net80211 rate control (which is after we tell the tx status to the net80211 rate-control in iwm_mvm_rx_tx_cmd_single()). * Disregard frames transferred with a different tx rate than the currently selected rate for the rate-control calculations. This way we avoid counting management frames (which are sent at a slow, and fixed rate), as well as frames we added to the tx queue just before a new IWM_LQ_CMD update took effect. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (5d6b465e288ac5b52d7115688d4e6516acbbea1c)
2019-01-24 03:41:09 +00:00
iwm_setrates(struct iwm_softc *sc, struct iwm_node *in, int rix)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
struct ieee80211_node *ni = &in->in_ni;
struct iwm_lq_cmd *lq = &in->in_lq;
if_iwm - Update firmware rs table, instead of indexing the table in tx cmds. * Rather than providing a non-zero index into the firmware RS table, we should always use index 0 and update the firmware RS table whenever our chosen tx rate for data-frames changes. * Send IWM_LQ_CMD updates when the tx rate gets updated by the net80211 rate control (which is after we tell the tx status to the net80211 rate-control in iwm_mvm_rx_tx_cmd_single()). * Disregard frames transferred with a different tx rate than the currently selected rate for the rate-control calculations. This way we avoid counting management frames (which are sent at a slow, and fixed rate), as well as frames we added to the tx queue just before a new IWM_LQ_CMD update took effect. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (5d6b465e288ac5b52d7115688d4e6516acbbea1c)
2019-01-24 03:41:09 +00:00
struct ieee80211_rateset *rs = &ni->ni_rates;
int nrates = rs->rs_nrates;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
int i, ridx, tab = 0;
[iwm] fix default antenna selection a bit; fix 5GHz rate control handling. * Don't do the antenna switching when setting up the rate table - we don't take into account whether it's an active antenna or not (eg shared with BT.) I'll look into this a bit more later. * The default antenna is still 1, I'll look into that a bit more later. (So no, this doesn't fix it for Larry who needs ANT_B to be active, not ANT_A.) * My changes to the rate control setup used the wrong method for finding a suitable rate, which led to 1M CCK frames being queued for 11a operation. This is .. sub-optimal. Change the rate control lookup path to use the global table instead of the per-node table, which won't be setup until the node rate table is setup (which happens way too late in this game.) Tested: * Intel 7260, 2G and 5G operation.
2016-09-24 01:21:42 +00:00
// int txant = 0;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if_iwm - Update firmware rs table, instead of indexing the table in tx cmds. * Rather than providing a non-zero index into the firmware RS table, we should always use index 0 and update the firmware RS table whenever our chosen tx rate for data-frames changes. * Send IWM_LQ_CMD updates when the tx rate gets updated by the net80211 rate control (which is after we tell the tx status to the net80211 rate-control in iwm_mvm_rx_tx_cmd_single()). * Disregard frames transferred with a different tx rate than the currently selected rate for the rate-control calculations. This way we avoid counting management frames (which are sent at a slow, and fixed rate), as well as frames we added to the tx queue just before a new IWM_LQ_CMD update took effect. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (5d6b465e288ac5b52d7115688d4e6516acbbea1c)
2019-01-24 03:41:09 +00:00
KASSERT(rix >= 0 && rix < nrates, ("invalid rix"));
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (nrates > nitems(lq->rs_table)) {
device_printf(sc->sc_dev,
"%s: node supports %d rates, driver handles "
"only %zu\n", __func__, nrates, nitems(lq->rs_table));
return;
}
[iwm] fix up the rate control setup code to initialise rates in the order we want to use it. The rate table was being initialised in low->high, but the link quality table was being initialised high->low. So, when we did a lookup, we would get the indexes wrong. This started by a patch from dragonflybsd which reversed how the ni->in_ridx[] array is being used; I'd rather it all be consistent. So, this is consistent. Inspired by: what I did to iwn(4) a while ago Inspired by: DragonflyBSD; <imre@vdsz.com>
2016-01-18 05:43:34 +00:00
if (nrates == 0) {
device_printf(sc->sc_dev,
"%s: node supports 0 rates, odd!\n", __func__);
return;
}
if_iwm - Update firmware rs table, instead of indexing the table in tx cmds. * Rather than providing a non-zero index into the firmware RS table, we should always use index 0 and update the firmware RS table whenever our chosen tx rate for data-frames changes. * Send IWM_LQ_CMD updates when the tx rate gets updated by the net80211 rate control (which is after we tell the tx status to the net80211 rate-control in iwm_mvm_rx_tx_cmd_single()). * Disregard frames transferred with a different tx rate than the currently selected rate for the rate-control calculations. This way we avoid counting management frames (which are sent at a slow, and fixed rate), as well as frames we added to the tx queue just before a new IWM_LQ_CMD update took effect. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (5d6b465e288ac5b52d7115688d4e6516acbbea1c)
2019-01-24 03:41:09 +00:00
nrates = imin(rix + 1, nrates);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_TXRATE,
"%s: nrates=%d\n", __func__, nrates);
[iwm] fix up the rate control setup code to initialise rates in the order we want to use it. The rate table was being initialised in low->high, but the link quality table was being initialised high->low. So, when we did a lookup, we would get the indexes wrong. This started by a patch from dragonflybsd which reversed how the ni->in_ridx[] array is being used; I'd rather it all be consistent. So, this is consistent. Inspired by: what I did to iwn(4) a while ago Inspired by: DragonflyBSD; <imre@vdsz.com>
2016-01-18 05:43:34 +00:00
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* then construct a lq_cmd based on those */
memset(lq, 0, sizeof(*lq));
lq->sta_id = IWM_STATION_ID;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
/* For HT, always enable RTS/CTS to avoid excessive retries. */
if (ni->ni_flags & IEEE80211_NODE_HT)
lq->flags |= IWM_LQ_FLAG_USE_RTS_MSK;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/*
* are these used? (we don't do SISO or MIMO)
* need to set them to non-zero, though, or we get an error.
*/
lq->single_stream_ant_msk = 1;
lq->dual_stream_ant_msk = 1;
/*
* Build the actual rate selection table.
* The lowest bits are the rates. Additionally,
* CCK needs bit 9 to be set. The rest of the bits
* we add to the table select the tx antenna
* Note that we add the rates in the highest rate first
* (opposite of ni_rates).
*/
for (i = 0; i < nrates; i++) {
if_iwm - Update firmware rs table, instead of indexing the table in tx cmds. * Rather than providing a non-zero index into the firmware RS table, we should always use index 0 and update the firmware RS table whenever our chosen tx rate for data-frames changes. * Send IWM_LQ_CMD updates when the tx rate gets updated by the net80211 rate control (which is after we tell the tx status to the net80211 rate-control in iwm_mvm_rx_tx_cmd_single()). * Disregard frames transferred with a different tx rate than the currently selected rate for the rate-control calculations. This way we avoid counting management frames (which are sent at a slow, and fixed rate), as well as frames we added to the tx queue just before a new IWM_LQ_CMD update took effect. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (5d6b465e288ac5b52d7115688d4e6516acbbea1c)
2019-01-24 03:41:09 +00:00
int rate = rs->rs_rates[rix - i] & IEEE80211_RATE_VAL;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
int nextant;
if_iwm - Update firmware rs table, instead of indexing the table in tx cmds. * Rather than providing a non-zero index into the firmware RS table, we should always use index 0 and update the firmware RS table whenever our chosen tx rate for data-frames changes. * Send IWM_LQ_CMD updates when the tx rate gets updated by the net80211 rate control (which is after we tell the tx status to the net80211 rate-control in iwm_mvm_rx_tx_cmd_single()). * Disregard frames transferred with a different tx rate than the currently selected rate for the rate-control calculations. This way we avoid counting management frames (which are sent at a slow, and fixed rate), as well as frames we added to the tx queue just before a new IWM_LQ_CMD update took effect. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (5d6b465e288ac5b52d7115688d4e6516acbbea1c)
2019-01-24 03:41:09 +00:00
/* Map 802.11 rate to HW rate index. */
ridx = iwm_rate2ridx(sc, rate);
if (ridx == -1)
continue;
[iwm] fix default antenna selection a bit; fix 5GHz rate control handling. * Don't do the antenna switching when setting up the rate table - we don't take into account whether it's an active antenna or not (eg shared with BT.) I'll look into this a bit more later. * The default antenna is still 1, I'll look into that a bit more later. (So no, this doesn't fix it for Larry who needs ANT_B to be active, not ANT_A.) * My changes to the rate control setup used the wrong method for finding a suitable rate, which led to 1M CCK frames being queued for 11a operation. This is .. sub-optimal. Change the rate control lookup path to use the global table instead of the per-node table, which won't be setup until the node rate table is setup (which happens way too late in this game.) Tested: * Intel 7260, 2G and 5G operation.
2016-09-24 01:21:42 +00:00
#if 0
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (txant == 0)
[iwm] Sync valid_tx_ant and valid_rx_ant mask handling with iwlwifi. * This fixes the phy_cfg field sent in the iwm_send_phy_cfg_cmd() command, which wasn't taking into account the valid_rx_ant and valid_tx_ant masks from nvm_data before. Tested: * 7260, STA mode, 2G and 5G Obtained from: DragonflyBSD commit cbb82693c18fd71b4eb86855b82d03995f352d65
2017-02-06 05:03:41 +00:00
txant = iwm_mvm_get_valid_tx_ant(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
nextant = 1<<(ffs(txant)-1);
txant &= ~nextant;
[iwm] fix default antenna selection a bit; fix 5GHz rate control handling. * Don't do the antenna switching when setting up the rate table - we don't take into account whether it's an active antenna or not (eg shared with BT.) I'll look into this a bit more later. * The default antenna is still 1, I'll look into that a bit more later. (So no, this doesn't fix it for Larry who needs ANT_B to be active, not ANT_A.) * My changes to the rate control setup used the wrong method for finding a suitable rate, which led to 1M CCK frames being queued for 11a operation. This is .. sub-optimal. Change the rate control lookup path to use the global table instead of the per-node table, which won't be setup until the node rate table is setup (which happens way too late in this game.) Tested: * Intel 7260, 2G and 5G operation.
2016-09-24 01:21:42 +00:00
#else
[iwm] Sync valid_tx_ant and valid_rx_ant mask handling with iwlwifi. * This fixes the phy_cfg field sent in the iwm_send_phy_cfg_cmd() command, which wasn't taking into account the valid_rx_ant and valid_tx_ant masks from nvm_data before. Tested: * 7260, STA mode, 2G and 5G Obtained from: DragonflyBSD commit cbb82693c18fd71b4eb86855b82d03995f352d65
2017-02-06 05:03:41 +00:00
nextant = iwm_mvm_get_valid_tx_ant(sc);
[iwm] fix default antenna selection a bit; fix 5GHz rate control handling. * Don't do the antenna switching when setting up the rate table - we don't take into account whether it's an active antenna or not (eg shared with BT.) I'll look into this a bit more later. * The default antenna is still 1, I'll look into that a bit more later. (So no, this doesn't fix it for Larry who needs ANT_B to be active, not ANT_A.) * My changes to the rate control setup used the wrong method for finding a suitable rate, which led to 1M CCK frames being queued for 11a operation. This is .. sub-optimal. Change the rate control lookup path to use the global table instead of the per-node table, which won't be setup until the node rate table is setup (which happens way too late in this game.) Tested: * Intel 7260, 2G and 5G operation.
2016-09-24 01:21:42 +00:00
#endif
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
tab = iwm_rates[ridx].plcp;
tab |= nextant << IWM_RATE_MCS_ANT_POS;
if (IWM_RIDX_IS_CCK(ridx))
tab |= IWM_RATE_MCS_CCK_MSK;
IWM_DPRINTF(sc, IWM_DEBUG_TXRATE,
"station rate i=%d, rate=%d, hw=%x\n",
i, iwm_rates[ridx].rate, tab);
lq->rs_table[i] = htole32(tab);
}
/* then fill the rest with the lowest possible rate */
for (i = nrates; i < nitems(lq->rs_table); i++) {
KASSERT(tab != 0, ("invalid tab"));
lq->rs_table[i] = htole32(tab);
}
}
static int
iwm_media_change(struct ifnet *ifp)
{
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
struct ieee80211vap *vap = ifp->if_softc;
struct ieee80211com *ic = vap->iv_ic;
struct iwm_softc *sc = ic->ic_softc;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
int error;
error = ieee80211_media_change(ifp);
if (error != ENETRESET)
return error;
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
IWM_LOCK(sc);
if (ic->ic_nrunning > 0) {
iwm_stop(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm_init(sc);
}
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
IWM_UNLOCK(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
return error;
}
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
static void
iwm_bring_down_firmware(struct iwm_softc *sc, struct ieee80211vap *vap)
{
struct iwm_vap *ivp = IWM_VAP(vap);
int error;
iwm - Always clear watchdog timer, when bringing down firmware state. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8abdc2b36a45c4e9c95fc8263ca532ea26633dcb)
2019-01-24 03:48:27 +00:00
/* Avoid Tx watchdog triggering, when transfers get dropped here. */
sc->sc_tx_timer = 0;
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
ivp->iv_auth = 0;
if (sc->sc_firmware_state == 3) {
iwm_xmit_queue_drain(sc);
// iwm_mvm_flush_tx_path(sc, 0xf, IWM_CMD_SYNC);
error = iwm_mvm_rm_sta(sc, vap, TRUE);
if (error) {
device_printf(sc->sc_dev,
"%s: Failed to remove station: %d\n",
__func__, error);
}
}
if (sc->sc_firmware_state == 3) {
error = iwm_mvm_mac_ctxt_changed(sc, vap);
if (error) {
device_printf(sc->sc_dev,
"%s: Failed to change mac context: %d\n",
__func__, error);
}
}
if (sc->sc_firmware_state == 3) {
error = iwm_mvm_sf_update(sc, vap, FALSE);
if (error) {
device_printf(sc->sc_dev,
"%s: Failed to update smart FIFO: %d\n",
__func__, error);
}
}
if (sc->sc_firmware_state == 3) {
error = iwm_mvm_rm_sta_id(sc, vap);
if (error) {
device_printf(sc->sc_dev,
"%s: Failed to remove station id: %d\n",
__func__, error);
}
}
if (sc->sc_firmware_state == 3) {
error = iwm_mvm_update_quotas(sc, NULL);
if (error) {
device_printf(sc->sc_dev,
"%s: Failed to update PHY quota: %d\n",
__func__, error);
}
}
if (sc->sc_firmware_state == 3) {
/* XXX Might need to specify bssid correctly. */
error = iwm_mvm_mac_ctxt_changed(sc, vap);
if (error) {
device_printf(sc->sc_dev,
"%s: Failed to change mac context: %d\n",
__func__, error);
}
}
if (sc->sc_firmware_state == 3) {
sc->sc_firmware_state = 2;
}
if (sc->sc_firmware_state > 1) {
error = iwm_mvm_binding_remove_vif(sc, ivp);
if (error) {
device_printf(sc->sc_dev,
"%s: Failed to remove channel ctx: %d\n",
__func__, error);
}
}
if (sc->sc_firmware_state > 1) {
sc->sc_firmware_state = 1;
}
ivp->phy_ctxt = NULL;
if (sc->sc_firmware_state > 0) {
error = iwm_mvm_mac_ctxt_changed(sc, vap);
if (error) {
device_printf(sc->sc_dev,
"%s: Failed to change mac context: %d\n",
__func__, error);
}
}
if (sc->sc_firmware_state > 0) {
error = iwm_mvm_power_update_mac(sc);
if (error != 0) {
device_printf(sc->sc_dev,
"%s: failed to update power management\n",
__func__);
}
}
sc->sc_firmware_state = 0;
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static int
iwm_newstate(struct ieee80211vap *vap, enum ieee80211_state nstate, int arg)
{
struct iwm_vap *ivp = IWM_VAP(vap);
struct ieee80211com *ic = vap->iv_ic;
Convert more ifp->if_softc -> ic->ic_softc. These should be a big no-op. Tested: * make universe
2015-08-17 23:44:38 +00:00
struct iwm_softc *sc = ic->ic_softc;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
struct iwm_node *in;
int error;
IWM_DPRINTF(sc, IWM_DEBUG_STATE,
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
"switching state %s -> %s arg=0x%x\n",
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
ieee80211_state_name[vap->iv_state],
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
ieee80211_state_name[nstate],
arg);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IEEE80211_UNLOCK(ic);
IWM_LOCK(sc);
[iwm] add LED blinking for iwm hardware that supports it. Submitted by: Imre Vadasz <imre@vdsz.com> Obtained from: DragonflyBSD, Linux iwlwifi/mvm
2016-06-02 04:42:28 +00:00
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
if ((sc->sc_flags & IWM_FLAG_SCAN_RUNNING) &&
(nstate == IEEE80211_S_AUTH ||
nstate == IEEE80211_S_ASSOC ||
nstate == IEEE80211_S_RUN)) {
/* Stop blinking for a scan, when authenticating. */
[iwm] add LED blinking for iwm hardware that supports it. Submitted by: Imre Vadasz <imre@vdsz.com> Obtained from: DragonflyBSD, Linux iwlwifi/mvm
2016-06-02 04:42:28 +00:00
iwm_led_blink_stop(sc);
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
}
[iwm] add LED blinking for iwm hardware that supports it. Submitted by: Imre Vadasz <imre@vdsz.com> Obtained from: DragonflyBSD, Linux iwlwifi/mvm
2016-06-02 04:42:28 +00:00
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
if (vap->iv_state == IEEE80211_S_RUN && nstate != IEEE80211_S_RUN) {
iwm_mvm_led_disable(sc);
/* disable beacon filtering if we're hopping out of RUN */
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm_mvm_disable_beacon_filter(sc);
iwm: reorganize if_iwmvar.h - Change order of data in if_iwmvar.h (like it is in other drivers: defines, data structures, vap/node structures, softc struct and locks); use indentation. - Fix IWM_LOCK(_sc) / IWM_UNLOCK(_sc) macro. - Add IWM_LOCK_INIT / DESTROY(sc) + fix mtx_init() usage. - Wrap iwm_node casts into IWM_NODE() macro. - Drop some fields: * wt_hwqueue from Tx radiotap header; * macaddr[6] from iwm_vap; Approved by: adrian Differential Revision: https://reviews.freebsd.org/D4753
2016-01-03 10:06:10 +00:00
if (((in = IWM_NODE(vap->iv_bss)) != NULL))
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
in->in_assoc = 0;
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm - Improve firmware Time Event handling. * This is a mix of the OpenBSD Git 7fd9664469d1b717a307eebd74aeececbd3c41cc change, and syncing with the Linux iwlwifi code. Taken-From: Linux iwlwifi, and OpenBSD Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (706a3044afd27c3fecfdf57bec1695310e53e228)
2019-01-24 03:47:04 +00:00
if ((vap->iv_state == IEEE80211_S_AUTH ||
vap->iv_state == IEEE80211_S_ASSOC ||
vap->iv_state == IEEE80211_S_RUN) &&
(nstate == IEEE80211_S_INIT ||
nstate == IEEE80211_S_SCAN ||
nstate == IEEE80211_S_AUTH)) {
iwm_mvm_stop_session_protection(sc, ivp);
}
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
if ((vap->iv_state == IEEE80211_S_RUN ||
vap->iv_state == IEEE80211_S_ASSOC) &&
nstate == IEEE80211_S_INIT) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/*
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
* In this case, iv_newstate() wants to send an 80211 frame on
* the network that we are leaving. So we need to call it,
* before tearing down all the firmware state.
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
*/
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
IWM_UNLOCK(sc);
IEEE80211_LOCK(ic);
ivp->iv_newstate(vap, nstate, arg);
IEEE80211_UNLOCK(ic);
IWM_LOCK(sc);
iwm_bring_down_firmware(sc, vap);
IWM_UNLOCK(sc);
IEEE80211_LOCK(ic);
return 0;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
switch (nstate) {
case IEEE80211_S_INIT:
[iwm] Handle AUTH->SCAN/INIT and ASSOC->SCAN/INIT better * Tear down the relevant firmware state (i.e. the station, the vif binding) in these transition cases. * Before this case would leave the firmware state lying around, resulting in errors and firmware panics in the subsequent association attempts. Obtained from: dragonflybsd.git 94b501399fde6368ae388a669c95b099a6e66e93
2017-05-12 06:05:34 +00:00
case IEEE80211_S_SCAN:
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
break;
case IEEE80211_S_AUTH:
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
iwm_bring_down_firmware(sc, vap);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if ((error = iwm_auth(vap, sc)) != 0) {
device_printf(sc->sc_dev,
"%s: could not move to auth state: %d\n",
__func__, error);
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
iwm_bring_down_firmware(sc, vap);
IWM_UNLOCK(sc);
IEEE80211_LOCK(ic);
return 1;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
break;
case IEEE80211_S_ASSOC:
[iwm] Enable Energy Based Scan (EBS). This can significantly reduce scan duration thus saving time and power. EBS failure reported by FW disables EBS for current connection. It is re-enabled upon new connection attempt on any WLAN interface. Obtained from: dragonflybsd.git 89f579e9823a5c446ca172cf82bbc210d6a054a4
2017-03-25 02:49:20 +00:00
/*
* EBS may be disabled due to previous failures reported by FW.
* Reset EBS status here assuming environment has been changed.
*/
[iwm] No need for iwm_assoc() in AUTH->ASSOC transition. * Hence no need to keep stuff in separate iwm_assoc() function, just inline the stuff into iwm_newstate(). Obtained from: dragonflybsd.git e8f7d88e0d030f138f95ecdb7c1a729d9fb0d6ab
2017-05-12 05:51:50 +00:00
sc->last_ebs_successful = TRUE;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
break;
case IEEE80211_S_RUN:
[iwm] No need for iwm_assoc() in AUTH->ASSOC transition. * Hence no need to keep stuff in separate iwm_assoc() function, just inline the stuff into iwm_newstate(). Obtained from: dragonflybsd.git e8f7d88e0d030f138f95ecdb7c1a729d9fb0d6ab
2017-05-12 05:51:50 +00:00
in = IWM_NODE(vap->iv_bss);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Update the association state, now we have it all */
/* (eg associd comes in at this point */
[iwm] No need for iwm_assoc() in AUTH->ASSOC transition. * Hence no need to keep stuff in separate iwm_assoc() function, just inline the stuff into iwm_newstate(). Obtained from: dragonflybsd.git e8f7d88e0d030f138f95ecdb7c1a729d9fb0d6ab
2017-05-12 05:51:50 +00:00
error = iwm_mvm_update_sta(sc, in);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (error != 0) {
device_printf(sc->sc_dev,
[iwm] No need for iwm_assoc() in AUTH->ASSOC transition. * Hence no need to keep stuff in separate iwm_assoc() function, just inline the stuff into iwm_newstate(). Obtained from: dragonflybsd.git e8f7d88e0d030f138f95ecdb7c1a729d9fb0d6ab
2017-05-12 05:51:50 +00:00
"%s: failed to update STA\n", __func__);
IWM_UNLOCK(sc);
IEEE80211_LOCK(ic);
return error;
}
in->in_assoc = 1;
error = iwm_mvm_mac_ctxt_changed(sc, vap);
if (error != 0) {
device_printf(sc->sc_dev,
"%s: failed to update MAC: %d\n", __func__, error);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Move Smart Fifo handling into if_iwm_sf.c, sync with Linux iwlwifi. * This change also fixes a possible issue in the existing smart-fifo code, which set the IWM_SF_CFG_DUMMY_NOTIF_OFF bit on AC8260 chipsets, although that's only used in iwlwifi for Family 8000 chipsets connected via SDIO interface. Obtained from: Dragonflybsd.git cb650b01526b0aeef3c4307d926e7f1428997d50
2017-06-04 21:05:58 +00:00
iwm_mvm_sf_update(sc, vap, FALSE);
[iwm] Switch arguments from iwm_node* to iwm_vap* in if_iwm_power.c. * Power management handling is per-vap, not per-node, so we should pass the iwm_vap in these arguments. Obtained from: dragonflybsd.git 62a4e7957a736b4de38938b02fa7eb9b45bc5d0d
2017-05-12 06:30:06 +00:00
iwm_mvm_enable_beacon_filter(sc, ivp);
[iwm] Add basic powermanagement support via ifconfig wlan0 powersave. * The DEVICE_POWER_FLAGS_CAM_MSK flag was removed in the upstream iwlwifi in Linux commit ceef91c89480dd18bb3ac51e91280a233d0ca41f. * Add sc_ps_disabled flag to struct iwm_softc, which corresponds to mvm->ps_disabled in struct iwl_mvm in Linux iwlwifi. * Adds a hw.iwm.power_scheme tunable which corresponds to the power_scheme module parameter in Linux iwlwifi. Set this to 1 for completely disabling power management, 2 (default) for balanced powermanagement, and 3 for lowerpower mode (which does dtim period skipping). * Imports the constants.h file from iwlwifi as if_iwm_constants.h. * This doesn't allow changing the powermanagement setting while connected, also one can only choose between enabled and disabled powersaving with ifconfig (so switching between balanced and low-power mode requires rebooting to change the tunable). * After any changes to powermanagement (i.e. "ifconfig wlan0 powersave" to enable powermanagement, or "ifconfig wlan0 -powersave" for disabling powermanagement), one has to disconnect and reconnect to the accespoint for the change to take effect. Obtained from: dragonflybsd.git d7002a7990d077c92585978ea998474af50f91e0
2017-05-09 04:15:07 +00:00
iwm_mvm_power_update_mac(sc);
[iwm] Move in_phyctxt from struct iwm_node to phy_ctxt in struct iwm_vap. * This better matches how things are organized in Linux's iwlwifi. Obtained from: dragonflybsd.git 0cf16dd2e0e09a3e5140e50222ac2e69bcdb19a2
2017-05-09 05:32:35 +00:00
iwm_mvm_update_quotas(sc, ivp);
if_iwm - Update firmware rs table, instead of indexing the table in tx cmds. * Rather than providing a non-zero index into the firmware RS table, we should always use index 0 and update the firmware RS table whenever our chosen tx rate for data-frames changes. * Send IWM_LQ_CMD updates when the tx rate gets updated by the net80211 rate control (which is after we tell the tx status to the net80211 rate-control in iwm_mvm_rx_tx_cmd_single()). * Disregard frames transferred with a different tx rate than the currently selected rate for the rate-control calculations. This way we avoid counting management frames (which are sent at a slow, and fixed rate), as well as frames we added to the tx queue just before a new IWM_LQ_CMD update took effect. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (5d6b465e288ac5b52d7115688d4e6516acbbea1c)
2019-01-24 03:41:09 +00:00
int rix = ieee80211_ratectl_rate(&in->in_ni, NULL, 0);
iwm_setrates(sc, in, rix);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Add iwm_mvm_send_lq_cmd() from Linux iwlwifi to if_iwm_util.c. Obtained from: dragonflybsd.git 8a5dd7783e407856754093f5b1c9c757c64534b7
2017-07-26 05:51:31 +00:00
if ((error = iwm_mvm_send_lq_cmd(sc, &in->in_lq, TRUE)) != 0) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev,
[iwm] Add iwm_mvm_send_lq_cmd() from Linux iwlwifi to if_iwm_util.c. Obtained from: dragonflybsd.git 8a5dd7783e407856754093f5b1c9c757c64534b7
2017-07-26 05:51:31 +00:00
"%s: IWM_LQ_CMD failed: %d\n", __func__, error);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
iwm_mvm_led_enable(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
break;
default:
break;
}
IWM_UNLOCK(sc);
IEEE80211_LOCK(ic);
return (ivp->iv_newstate(vap, nstate, arg));
}
void
iwm_endscan_cb(void *arg, int pending)
{
struct iwm_softc *sc = arg;
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
struct ieee80211com *ic = &sc->sc_ic;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_SCAN | IWM_DEBUG_TRACE,
"%s: scan ended\n",
__func__);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
ieee80211_scan_done(TAILQ_FIRST(&ic->ic_vaps));
}
static int
iwm_send_bt_init_conf(struct iwm_softc *sc)
{
struct iwm_bt_coex_cmd bt_cmd;
bt_cmd.mode = htole32(IWM_BT_COEX_WIFI);
bt_cmd.enabled_modules = htole32(IWM_BT_COEX_HIGH_BAND_RET);
return iwm_mvm_send_cmd_pdu(sc, IWM_BT_CONFIG, 0, sizeof(bt_cmd),
&bt_cmd);
}
[iwm] Check for lar_disable tunable, and lar_enabled flag from NVM. * LAR can be disabled with the hw.iwm.lar.disable tunable now. * On Family 8000 devices we need to check the lar_enabled flag from nvm_data in addition to the TLV_CAPA_LAR_SUPPORT flag from the firmware. * Add a separate IWM_DEBUG_LAR debugging flag. Obtained from: dragonflybsd.git 0593e39cb295aa996ecf789ed4990c3b255f1770
2017-06-04 21:10:14 +00:00
static boolean_t
iwm_mvm_is_lar_supported(struct iwm_softc *sc)
{
boolean_t nvm_lar = sc->nvm_data->lar_enabled;
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
boolean_t tlv_lar = fw_has_capa(&sc->sc_fw.ucode_capa,
[iwm] Check for lar_disable tunable, and lar_enabled flag from NVM. * LAR can be disabled with the hw.iwm.lar.disable tunable now. * On Family 8000 devices we need to check the lar_enabled flag from nvm_data in addition to the TLV_CAPA_LAR_SUPPORT flag from the firmware. * Add a separate IWM_DEBUG_LAR debugging flag. Obtained from: dragonflybsd.git 0593e39cb295aa996ecf789ed4990c3b255f1770
2017-06-04 21:10:14 +00:00
IWM_UCODE_TLV_CAPA_LAR_SUPPORT);
if (iwm_lar_disable)
return FALSE;
/*
* Enable LAR only if it is supported by the FW (TLV) &&
* enabled in the NVM
*/
if (sc->cfg->device_family == IWM_DEVICE_FAMILY_8000)
return nvm_lar && tlv_lar;
else
return tlv_lar;
}
static boolean_t
iwm_mvm_is_wifi_mcc_supported(struct iwm_softc *sc)
{
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
return fw_has_api(&sc->sc_fw.ucode_capa,
[iwm] Check for lar_disable tunable, and lar_enabled flag from NVM. * LAR can be disabled with the hw.iwm.lar.disable tunable now. * On Family 8000 devices we need to check the lar_enabled flag from nvm_data in addition to the TLV_CAPA_LAR_SUPPORT flag from the firmware. * Add a separate IWM_DEBUG_LAR debugging flag. Obtained from: dragonflybsd.git 0593e39cb295aa996ecf789ed4990c3b255f1770
2017-06-04 21:10:14 +00:00
IWM_UCODE_TLV_API_WIFI_MCC_UPDATE) ||
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
fw_has_capa(&sc->sc_fw.ucode_capa,
[iwm] Check for lar_disable tunable, and lar_enabled flag from NVM. * LAR can be disabled with the hw.iwm.lar.disable tunable now. * On Family 8000 devices we need to check the lar_enabled flag from nvm_data in addition to the TLV_CAPA_LAR_SUPPORT flag from the firmware. * Add a separate IWM_DEBUG_LAR debugging flag. Obtained from: dragonflybsd.git 0593e39cb295aa996ecf789ed4990c3b255f1770
2017-06-04 21:10:14 +00:00
IWM_UCODE_TLV_CAPA_LAR_MULTI_MCC);
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
static int
iwm_send_update_mcc_cmd(struct iwm_softc *sc, const char *alpha2)
{
struct iwm_mcc_update_cmd mcc_cmd;
struct iwm_host_cmd hcmd = {
.id = IWM_MCC_UPDATE_CMD,
.flags = (IWM_CMD_SYNC | IWM_CMD_WANT_SKB),
.data = { &mcc_cmd },
};
int ret;
#ifdef IWM_DEBUG
struct iwm_rx_packet *pkt;
struct iwm_mcc_update_resp_v1 *mcc_resp_v1 = NULL;
struct iwm_mcc_update_resp *mcc_resp;
int n_channels;
uint16_t mcc;
#endif
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
int resp_v2 = fw_has_capa(&sc->sc_fw.ucode_capa,
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
IWM_UCODE_TLV_CAPA_LAR_SUPPORT_V2);
[iwm] Check for lar_disable tunable, and lar_enabled flag from NVM. * LAR can be disabled with the hw.iwm.lar.disable tunable now. * On Family 8000 devices we need to check the lar_enabled flag from nvm_data in addition to the TLV_CAPA_LAR_SUPPORT flag from the firmware. * Add a separate IWM_DEBUG_LAR debugging flag. Obtained from: dragonflybsd.git 0593e39cb295aa996ecf789ed4990c3b255f1770
2017-06-04 21:10:14 +00:00
if (!iwm_mvm_is_lar_supported(sc)) {
IWM_DPRINTF(sc, IWM_DEBUG_LAR, "%s: no LAR support\n",
__func__);
return 0;
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
memset(&mcc_cmd, 0, sizeof(mcc_cmd));
mcc_cmd.mcc = htole16(alpha2[0] << 8 | alpha2[1]);
[iwm] Check for lar_disable tunable, and lar_enabled flag from NVM. * LAR can be disabled with the hw.iwm.lar.disable tunable now. * On Family 8000 devices we need to check the lar_enabled flag from nvm_data in addition to the TLV_CAPA_LAR_SUPPORT flag from the firmware. * Add a separate IWM_DEBUG_LAR debugging flag. Obtained from: dragonflybsd.git 0593e39cb295aa996ecf789ed4990c3b255f1770
2017-06-04 21:10:14 +00:00
if (iwm_mvm_is_wifi_mcc_supported(sc))
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
mcc_cmd.source_id = IWM_MCC_SOURCE_GET_CURRENT;
else
mcc_cmd.source_id = IWM_MCC_SOURCE_OLD_FW;
if (resp_v2)
hcmd.len[0] = sizeof(struct iwm_mcc_update_cmd);
else
hcmd.len[0] = sizeof(struct iwm_mcc_update_cmd_v1);
[iwm] Check for lar_disable tunable, and lar_enabled flag from NVM. * LAR can be disabled with the hw.iwm.lar.disable tunable now. * On Family 8000 devices we need to check the lar_enabled flag from nvm_data in addition to the TLV_CAPA_LAR_SUPPORT flag from the firmware. * Add a separate IWM_DEBUG_LAR debugging flag. Obtained from: dragonflybsd.git 0593e39cb295aa996ecf789ed4990c3b255f1770
2017-06-04 21:10:14 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_LAR,
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
"send MCC update to FW with '%c%c' src = %d\n",
alpha2[0], alpha2[1], mcc_cmd.source_id);
ret = iwm_send_cmd(sc, &hcmd);
if (ret)
return ret;
#ifdef IWM_DEBUG
pkt = hcmd.resp_pkt;
/* Extract MCC response */
if (resp_v2) {
mcc_resp = (void *)pkt->data;
mcc = mcc_resp->mcc;
n_channels = le32toh(mcc_resp->n_channels);
} else {
mcc_resp_v1 = (void *)pkt->data;
mcc = mcc_resp_v1->mcc;
n_channels = le32toh(mcc_resp_v1->n_channels);
}
/* W/A for a FW/NVM issue - returns 0x00 for the world domain */
if (mcc == 0)
mcc = 0x3030; /* "00" - world */
[iwm] Check for lar_disable tunable, and lar_enabled flag from NVM. * LAR can be disabled with the hw.iwm.lar.disable tunable now. * On Family 8000 devices we need to check the lar_enabled flag from nvm_data in addition to the TLV_CAPA_LAR_SUPPORT flag from the firmware. * Add a separate IWM_DEBUG_LAR debugging flag. Obtained from: dragonflybsd.git 0593e39cb295aa996ecf789ed4990c3b255f1770
2017-06-04 21:10:14 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_LAR,
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
"regulatory domain '%c%c' (%d channels available)\n",
mcc >> 8, mcc & 0xff, n_channels);
#endif
iwm_free_resp(sc, &hcmd);
return 0;
}
static void
iwm_mvm_tt_tx_backoff(struct iwm_softc *sc, uint32_t backoff)
{
struct iwm_host_cmd cmd = {
.id = IWM_REPLY_THERMAL_MNG_BACKOFF,
.len = { sizeof(uint32_t), },
.data = { &backoff, },
};
if (iwm_send_cmd(sc, &cmd) != 0) {
device_printf(sc->sc_dev,
"failed to change thermal tx backoff\n");
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
}
static int
iwm_init_hw(struct iwm_softc *sc)
{
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
struct ieee80211com *ic = &sc->sc_ic;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
int error, i, ac;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Move Smart Fifo handling into if_iwm_sf.c, sync with Linux iwlwifi. * This change also fixes a possible issue in the existing smart-fifo code, which set the IWM_SF_CFG_DUMMY_NOTIF_OFF bit on AC8260 chipsets, although that's only used in iwlwifi for Family 8000 chipsets connected via SDIO interface. Obtained from: Dragonflybsd.git cb650b01526b0aeef3c4307d926e7f1428997d50
2017-06-04 21:05:58 +00:00
sc->sf_state = IWM_SF_UNINIT;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
if ((error = iwm_start_hw(sc)) != 0) {
printf("iwm_start_hw: failed %d\n", error);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
return error;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if ((error = iwm_run_init_mvm_ucode(sc, 0)) != 0) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
printf("iwm_run_init_mvm_ucode: failed %d\n", error);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
return error;
}
/*
* should stop and start HW since that INIT
* image just loaded
*/
iwm_stop_device(sc);
[iwm] Add basic powermanagement support via ifconfig wlan0 powersave. * The DEVICE_POWER_FLAGS_CAM_MSK flag was removed in the upstream iwlwifi in Linux commit ceef91c89480dd18bb3ac51e91280a233d0ca41f. * Add sc_ps_disabled flag to struct iwm_softc, which corresponds to mvm->ps_disabled in struct iwl_mvm in Linux iwlwifi. * Adds a hw.iwm.power_scheme tunable which corresponds to the power_scheme module parameter in Linux iwlwifi. Set this to 1 for completely disabling power management, 2 (default) for balanced powermanagement, and 3 for lowerpower mode (which does dtim period skipping). * Imports the constants.h file from iwlwifi as if_iwm_constants.h. * This doesn't allow changing the powermanagement setting while connected, also one can only choose between enabled and disabled powersaving with ifconfig (so switching between balanced and low-power mode requires rebooting to change the tunable). * After any changes to powermanagement (i.e. "ifconfig wlan0 powersave" to enable powermanagement, or "ifconfig wlan0 -powersave" for disabling powermanagement), one has to disconnect and reconnect to the accespoint for the change to take effect. Obtained from: dragonflybsd.git d7002a7990d077c92585978ea998474af50f91e0
2017-05-09 04:15:07 +00:00
sc->sc_ps_disabled = FALSE;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if ((error = iwm_start_hw(sc)) != 0) {
device_printf(sc->sc_dev, "could not initialize hardware\n");
return error;
}
/* omstart, this time with the regular firmware */
[iwm] Get rid of some gratuitous constant renaming wrt. Linux iwlwifi. * IWM_UCODE_SECT_MAX -> IWM_UCODE_SECTION_MAX * IWM_UCODE_TYPE_* -> IWM_UCODE_* (except for IWM_UCODE_TYPE_MAX which stays). Obtained from: DragonflyBSD commit ff4d1fc3ed002c9fb362423da6c45d711b65658a
2017-02-06 05:07:15 +00:00
error = iwm_mvm_load_ucode_wait_alive(sc, IWM_UCODE_REGULAR);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (error) {
device_printf(sc->sc_dev, "could not load firmware\n");
goto error;
}
[iwm] Move Smart Fifo handling into if_iwm_sf.c, sync with Linux iwlwifi. * This change also fixes a possible issue in the existing smart-fifo code, which set the IWM_SF_CFG_DUMMY_NOTIF_OFF bit on AC8260 chipsets, although that's only used in iwlwifi for Family 8000 chipsets connected via SDIO interface. Obtained from: Dragonflybsd.git cb650b01526b0aeef3c4307d926e7f1428997d50
2017-06-04 21:05:58 +00:00
error = iwm_mvm_sf_update(sc, NULL, FALSE);
if (error)
device_printf(sc->sc_dev, "Failed to initialize Smart Fifo\n");
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
if ((error = iwm_send_bt_init_conf(sc)) != 0) {
device_printf(sc->sc_dev, "bt init conf failed\n");
goto error;
}
[iwm] Sync valid_tx_ant and valid_rx_ant mask handling with iwlwifi. * This fixes the phy_cfg field sent in the iwm_send_phy_cfg_cmd() command, which wasn't taking into account the valid_rx_ant and valid_tx_ant masks from nvm_data before. Tested: * 7260, STA mode, 2G and 5G Obtained from: DragonflyBSD commit cbb82693c18fd71b4eb86855b82d03995f352d65
2017-02-06 05:03:41 +00:00
error = iwm_send_tx_ant_cfg(sc, iwm_mvm_get_valid_tx_ant(sc));
if (error != 0) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
device_printf(sc->sc_dev, "antenna config failed\n");
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
goto error;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Sync if_iwm_phy_db code with Linux iwlwifi. Obtained from: Dragonflybsd commit c1019b6bfff36c856f7b4fccbdf3bb13ac27750c
2017-02-06 02:14:34 +00:00
/* Send phy db control command and then phy db calibration */
if ((error = iwm_send_phy_db_data(sc->sc_phy_db)) != 0)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
goto error;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
if ((error = iwm_send_phy_cfg_cmd(sc)) != 0) {
device_printf(sc->sc_dev, "phy_cfg_cmd failed\n");
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
goto error;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Add auxiliary station for scanning */
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
if ((error = iwm_mvm_add_aux_sta(sc)) != 0) {
device_printf(sc->sc_dev, "add_aux_sta failed\n");
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
goto error;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
for (i = 0; i < IWM_NUM_PHY_CTX; i++) {
/*
* The channel used here isn't relevant as it's
* going to be overwritten in the other flows.
* For now use the first channel we have.
*/
if ((error = iwm_mvm_phy_ctxt_add(sc,
&sc->sc_phyctxt[i], &ic->ic_channels[1], 1, 1)) != 0)
goto error;
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
/* Initialize tx backoffs to the minimum. */
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->device_family == IWM_DEVICE_FAMILY_7000)
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
iwm_mvm_tt_tx_backoff(sc, 0);
if_iwm - Configure the PCIe LTR, fix PCI express capability accesses. Taken-From: Linux iwlwifi Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (08a7ad5a5ff65aaaf2df6a609be7a4e1df43efc3)
2019-01-24 03:44:48 +00:00
if (iwm_mvm_config_ltr(sc) != 0)
device_printf(sc->sc_dev, "PCIe LTR configuration failed\n");
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
error = iwm_mvm_power_update_device(sc);
if (error)
goto error;
[iwm] Check for lar_disable tunable, and lar_enabled flag from NVM. * LAR can be disabled with the hw.iwm.lar.disable tunable now. * On Family 8000 devices we need to check the lar_enabled flag from nvm_data in addition to the TLV_CAPA_LAR_SUPPORT flag from the firmware. * Add a separate IWM_DEBUG_LAR debugging flag. Obtained from: dragonflybsd.git 0593e39cb295aa996ecf789ed4990c3b255f1770
2017-06-04 21:10:14 +00:00
if ((error = iwm_send_update_mcc_cmd(sc, "ZZ")) != 0)
goto error;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
if (fw_has_capa(&sc->sc_fw.ucode_capa, IWM_UCODE_TLV_CAPA_UMAC_SCAN)) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
if ((error = iwm_mvm_config_umac_scan(sc)) != 0)
goto error;
}
/* Enable Tx queues. */
for (ac = 0; ac < WME_NUM_AC; ac++) {
error = iwm_enable_txq(sc, IWM_STATION_ID, ac,
iwm_mvm_ac_to_tx_fifo[ac]);
if (error)
goto error;
}
if ((error = iwm_mvm_disable_beacon_filter(sc)) != 0) {
device_printf(sc->sc_dev, "failed to disable beacon filter\n");
goto error;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
return 0;
error:
iwm_stop_device(sc);
return error;
}
/* Allow multicast from our BSSID. */
static int
iwm_allow_mcast(struct ieee80211vap *vap, struct iwm_softc *sc)
{
struct ieee80211_node *ni = vap->iv_bss;
struct iwm_mcast_filter_cmd *cmd;
size_t size;
int error;
size = roundup(sizeof(*cmd), 4);
cmd = malloc(size, M_DEVBUF, M_NOWAIT | M_ZERO);
if (cmd == NULL)
return ENOMEM;
cmd->filter_own = 1;
cmd->port_id = 0;
cmd->count = 0;
cmd->pass_all = 1;
IEEE80211_ADDR_COPY(cmd->bssid, ni->ni_bssid);
error = iwm_mvm_send_cmd_pdu(sc, IWM_MCAST_FILTER_CMD,
IWM_CMD_SYNC, size, cmd);
free(cmd, M_DEVBUF);
return (error);
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
/*
* ifnet interfaces
*/
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static void
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
iwm_init(struct iwm_softc *sc)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
int error;
if (sc->sc_flags & IWM_FLAG_HW_INITED) {
return;
}
sc->sc_generation++;
sc->sc_flags &= ~IWM_FLAG_STOPPED;
if ((error = iwm_init_hw(sc)) != 0) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
printf("iwm_init_hw failed %d\n", error);
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
iwm_stop(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
return;
}
/*
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
* Ok, firmware loaded and we are jogging
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
*/
sc->sc_flags |= IWM_FLAG_HW_INITED;
}
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
static int
iwm_transmit(struct ieee80211com *ic, struct mbuf *m)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
struct iwm_softc *sc;
int error;
sc = ic->ic_softc;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_LOCK(sc);
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
if ((sc->sc_flags & IWM_FLAG_HW_INITED) == 0) {
IWM_UNLOCK(sc);
return (ENXIO);
}
error = mbufq_enqueue(&sc->sc_snd, m);
if (error) {
IWM_UNLOCK(sc);
return (error);
}
iwm_start(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_UNLOCK(sc);
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
return (0);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
/*
* Dequeue packets from sendq and call send.
*/
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static void
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
iwm_start(struct iwm_softc *sc)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
struct ieee80211_node *ni;
struct mbuf *m;
int ac = 0;
IWM_DPRINTF(sc, IWM_DEBUG_XMIT | IWM_DEBUG_TRACE, "->%s\n", __func__);
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
while (sc->qfullmsk == 0 &&
(m = mbufq_dequeue(&sc->sc_snd)) != NULL) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
ni = (struct ieee80211_node *)m->m_pkthdr.rcvif;
if (iwm_tx(sc, m, ni, ac) != 0) {
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
if_inc_counter(ni->ni_vap->iv_ifp,
IFCOUNTER_OERRORS, 1);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
ieee80211_free_node(ni);
continue;
}
if_iwm - Stop iwm_watchdog callout when idle. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (6a8683b0e9d734f23bd9647e117da198c2b9a74e)
2019-01-24 03:48:50 +00:00
if (sc->sc_tx_timer == 0) {
callout_reset(&sc->sc_watchdog_to, hz, iwm_watchdog,
sc);
}
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
sc->sc_tx_timer = 15;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
IWM_DPRINTF(sc, IWM_DEBUG_XMIT | IWM_DEBUG_TRACE, "<-%s\n", __func__);
}
static void
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
iwm_stop(struct iwm_softc *sc)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
sc->sc_flags &= ~IWM_FLAG_HW_INITED;
sc->sc_flags |= IWM_FLAG_STOPPED;
sc->sc_generation++;
[iwm] add LED blinking for iwm hardware that supports it. Submitted by: Imre Vadasz <imre@vdsz.com> Obtained from: DragonflyBSD, Linux iwlwifi/mvm
2016-06-02 04:42:28 +00:00
iwm_led_blink_stop(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
sc->sc_tx_timer = 0;
iwm_stop_device(sc);
[iwm] Use iwm_mvm_scan_stop_wait to properly abort scans. * Add IWM_FLAG_SCAN_RUNNING to sc->sc_flags to track whether the firmware is currently running a scan, in order to decide wheter iwm_scan_end needs to abort a running scan. * In iwm_scan_end, if the scan is still running, we now abort it, in order to keep the firmware scanning state in sync. * Try to make things a bit simpler, by reacting on the IWM_SCAN_OFFLOAD_COMPLETE and IWM_SCAN_COMPLETE_UMAC notifications, instead of IWM_SCAN_ITERATION_COMPLETE and IWM_SCAN_ITERATION_COMPLETE_UMAC. This should be fine since we always only tell the firmware to do a single scan iteration anyway. Obtained from: DragonflyBSD commit 1f249c981c4e89e7cde1836a75b61cac36dc7ac5
2017-02-08 06:50:59 +00:00
sc->sc_flags &= ~IWM_FLAG_SCAN_RUNNING;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
static void
iwm_watchdog(void *arg)
{
struct iwm_softc *sc = arg;
iwm: restart device after watchdog timeouts.
2016-05-19 22:43:21 +00:00
struct ieee80211com *ic = &sc->sc_ic;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm - Fix race during detach, where a callout is left after driver is gone. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (ba3b4ff9a1fc04a349df05d6d3449f4d9b15c4be)
2019-01-24 03:49:11 +00:00
if (sc->sc_attached == 0)
return;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (sc->sc_tx_timer > 0) {
if (--sc->sc_tx_timer == 0) {
device_printf(sc->sc_dev, "device timeout\n");
#ifdef IWM_DEBUG
iwm_nic_error(sc);
#endif
iwm: restart device after watchdog timeouts.
2016-05-19 22:43:21 +00:00
ieee80211_restart_all(ic);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
counter_u64_add(sc->sc_ic.ic_oerrors, 1);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
return;
}
if_iwm - Stop iwm_watchdog callout when idle. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (6a8683b0e9d734f23bd9647e117da198c2b9a74e)
2019-01-24 03:48:50 +00:00
callout_reset(&sc->sc_watchdog_to, hz, iwm_watchdog, sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
}
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
static void
iwm_parent(struct ieee80211com *ic)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
struct iwm_softc *sc = ic->ic_softc;
int startall = 0;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
IWM_LOCK(sc);
if (ic->ic_nrunning > 0) {
if (!(sc->sc_flags & IWM_FLAG_HW_INITED)) {
iwm_init(sc);
startall = 1;
}
} else if (sc->sc_flags & IWM_FLAG_HW_INITED)
iwm_stop(sc);
IWM_UNLOCK(sc);
if (startall)
ieee80211_start_all(ic);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
/*
* The interrupt side of things
*/
/*
* error dumping routines are from iwlwifi/mvm/utils.c
*/
/*
* Note: This structure is read from the device with IO accesses,
* and the reading already does the endian conversion. As it is
* read with uint32_t-sized accesses, any members with a different size
* need to be ordered correctly though!
*/
struct iwm_error_event_table {
uint32_t valid; /* (nonzero) valid, (0) log is empty */
uint32_t error_id; /* type of error */
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
uint32_t trm_hw_status0; /* TRM HW status */
uint32_t trm_hw_status1; /* TRM HW status */
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
uint32_t blink2; /* branch link */
uint32_t ilink1; /* interrupt link */
uint32_t ilink2; /* interrupt link */
uint32_t data1; /* error-specific data */
uint32_t data2; /* error-specific data */
uint32_t data3; /* error-specific data */
uint32_t bcon_time; /* beacon timer */
uint32_t tsf_low; /* network timestamp function timer */
uint32_t tsf_hi; /* network timestamp function timer */
uint32_t gp1; /* GP1 timer register */
uint32_t gp2; /* GP2 timer register */
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
uint32_t fw_rev_type; /* firmware revision type */
uint32_t major; /* uCode version major */
uint32_t minor; /* uCode version minor */
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
uint32_t hw_ver; /* HW Silicon version */
uint32_t brd_ver; /* HW board version */
uint32_t log_pc; /* log program counter */
uint32_t frame_ptr; /* frame pointer */
uint32_t stack_ptr; /* stack pointer */
uint32_t hcmd; /* last host command header */
uint32_t isr0; /* isr status register LMPM_NIC_ISR0:
* rxtx_flag */
uint32_t isr1; /* isr status register LMPM_NIC_ISR1:
* host_flag */
uint32_t isr2; /* isr status register LMPM_NIC_ISR2:
* enc_flag */
uint32_t isr3; /* isr status register LMPM_NIC_ISR3:
* time_flag */
uint32_t isr4; /* isr status register LMPM_NIC_ISR4:
* wico interrupt */
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
uint32_t last_cmd_id; /* last HCMD id handled by the firmware */
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
uint32_t wait_event; /* wait event() caller address */
uint32_t l2p_control; /* L2pControlField */
uint32_t l2p_duration; /* L2pDurationField */
uint32_t l2p_mhvalid; /* L2pMhValidBits */
uint32_t l2p_addr_match; /* L2pAddrMatchStat */
uint32_t lmpm_pmg_sel; /* indicate which clocks are turned on
* (LMPM_PMG_SEL) */
uint32_t u_timestamp; /* indicate when the date and time of the
* compilation */
uint32_t flow_handler; /* FH read/write pointers, RX credit */
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
} __packed /* LOG_ERROR_TABLE_API_S_VER_3 */;
/*
* UMAC error struct - relevant starting from family 8000 chip.
* Note: This structure is read from the device with IO accesses,
* and the reading already does the endian conversion. As it is
* read with u32-sized accesses, any members with a different size
* need to be ordered correctly though!
*/
struct iwm_umac_error_event_table {
uint32_t valid; /* (nonzero) valid, (0) log is empty */
uint32_t error_id; /* type of error */
uint32_t blink1; /* branch link */
uint32_t blink2; /* branch link */
uint32_t ilink1; /* interrupt link */
uint32_t ilink2; /* interrupt link */
uint32_t data1; /* error-specific data */
uint32_t data2; /* error-specific data */
uint32_t data3; /* error-specific data */
uint32_t umac_major;
uint32_t umac_minor;
uint32_t frame_pointer; /* core register 27*/
uint32_t stack_pointer; /* core register 28 */
uint32_t cmd_header; /* latest host cmd sent to UMAC */
uint32_t nic_isr_pref; /* ISR status register */
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
} __packed;
#define ERROR_START_OFFSET (1 * sizeof(uint32_t))
#define ERROR_ELEM_SIZE (7 * sizeof(uint32_t))
#ifdef IWM_DEBUG
struct {
const char *name;
uint8_t num;
} advanced_lookup[] = {
{ "NMI_INTERRUPT_WDG", 0x34 },
{ "SYSASSERT", 0x35 },
{ "UCODE_VERSION_MISMATCH", 0x37 },
{ "BAD_COMMAND", 0x38 },
{ "NMI_INTERRUPT_DATA_ACTION_PT", 0x3C },
{ "FATAL_ERROR", 0x3D },
{ "NMI_TRM_HW_ERR", 0x46 },
{ "NMI_INTERRUPT_TRM", 0x4C },
{ "NMI_INTERRUPT_BREAK_POINT", 0x54 },
{ "NMI_INTERRUPT_WDG_RXF_FULL", 0x5C },
{ "NMI_INTERRUPT_WDG_NO_RBD_RXF_FULL", 0x64 },
{ "NMI_INTERRUPT_HOST", 0x66 },
{ "NMI_INTERRUPT_ACTION_PT", 0x7C },
{ "NMI_INTERRUPT_UNKNOWN", 0x84 },
{ "NMI_INTERRUPT_INST_ACTION_PT", 0x86 },
{ "ADVANCED_SYSASSERT", 0 },
};
static const char *
iwm_desc_lookup(uint32_t num)
{
int i;
for (i = 0; i < nitems(advanced_lookup) - 1; i++)
if (advanced_lookup[i].num == num)
return advanced_lookup[i].name;
/* No entry matches 'num', so it is the last: ADVANCED_SYSASSERT */
return advanced_lookup[i].name;
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
static void
iwm_nic_umac_error(struct iwm_softc *sc)
{
struct iwm_umac_error_event_table table;
uint32_t base;
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
base = sc->umac_error_event_table;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
if (base < 0x800000) {
device_printf(sc->sc_dev, "Invalid error log pointer 0x%08x\n",
base);
return;
}
if (iwm_read_mem(sc, base, &table, sizeof(table)/sizeof(uint32_t))) {
device_printf(sc->sc_dev, "reading errlog failed\n");
return;
}
if (ERROR_START_OFFSET <= table.valid * ERROR_ELEM_SIZE) {
device_printf(sc->sc_dev, "Start UMAC Error Log Dump:\n");
device_printf(sc->sc_dev, "Status: 0x%x, count: %d\n",
sc->sc_flags, table.valid);
}
device_printf(sc->sc_dev, "0x%08X | %s\n", table.error_id,
iwm_desc_lookup(table.error_id));
device_printf(sc->sc_dev, "0x%08X | umac branchlink1\n", table.blink1);
device_printf(sc->sc_dev, "0x%08X | umac branchlink2\n", table.blink2);
device_printf(sc->sc_dev, "0x%08X | umac interruptlink1\n",
table.ilink1);
device_printf(sc->sc_dev, "0x%08X | umac interruptlink2\n",
table.ilink2);
device_printf(sc->sc_dev, "0x%08X | umac data1\n", table.data1);
device_printf(sc->sc_dev, "0x%08X | umac data2\n", table.data2);
device_printf(sc->sc_dev, "0x%08X | umac data3\n", table.data3);
device_printf(sc->sc_dev, "0x%08X | umac major\n", table.umac_major);
device_printf(sc->sc_dev, "0x%08X | umac minor\n", table.umac_minor);
device_printf(sc->sc_dev, "0x%08X | frame pointer\n",
table.frame_pointer);
device_printf(sc->sc_dev, "0x%08X | stack pointer\n",
table.stack_pointer);
device_printf(sc->sc_dev, "0x%08X | last host cmd\n", table.cmd_header);
device_printf(sc->sc_dev, "0x%08X | isr status reg\n",
table.nic_isr_pref);
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/*
* Support for dumping the error log seemed like a good idea ...
* but it's mostly hex junk and the only sensible thing is the
* hw/ucode revision (which we know anyway). Since it's here,
* I'll just leave it in, just in case e.g. the Intel guys want to
* help us decipher some "ADVANCED_SYSASSERT" later.
*/
static void
iwm_nic_error(struct iwm_softc *sc)
{
struct iwm_error_event_table table;
uint32_t base;
device_printf(sc->sc_dev, "dumping device error log\n");
iwm - Update alive response handling, add v4 and remove old versions. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (3820e2bf3331ced3541d1811a38c5a5136dfab93)
2019-01-24 03:49:35 +00:00
base = sc->error_event_table[0];
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
if (base < 0x800000) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev,
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
"Invalid error log pointer 0x%08x\n", base);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
return;
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
if (iwm_read_mem(sc, base, &table, sizeof(table)/sizeof(uint32_t))) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev, "reading errlog failed\n");
return;
}
if (!table.valid) {
device_printf(sc->sc_dev, "errlog not found, skipping\n");
return;
}
if (ERROR_START_OFFSET <= table.valid * ERROR_ELEM_SIZE) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
device_printf(sc->sc_dev, "Start Error Log Dump:\n");
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev, "Status: 0x%x, count: %d\n",
sc->sc_flags, table.valid);
}
device_printf(sc->sc_dev, "0x%08X | %-28s\n", table.error_id,
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
iwm_desc_lookup(table.error_id));
device_printf(sc->sc_dev, "%08X | trm_hw_status0\n",
table.trm_hw_status0);
device_printf(sc->sc_dev, "%08X | trm_hw_status1\n",
table.trm_hw_status1);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev, "%08X | branchlink2\n", table.blink2);
device_printf(sc->sc_dev, "%08X | interruptlink1\n", table.ilink1);
device_printf(sc->sc_dev, "%08X | interruptlink2\n", table.ilink2);
device_printf(sc->sc_dev, "%08X | data1\n", table.data1);
device_printf(sc->sc_dev, "%08X | data2\n", table.data2);
device_printf(sc->sc_dev, "%08X | data3\n", table.data3);
device_printf(sc->sc_dev, "%08X | beacon time\n", table.bcon_time);
device_printf(sc->sc_dev, "%08X | tsf low\n", table.tsf_low);
device_printf(sc->sc_dev, "%08X | tsf hi\n", table.tsf_hi);
device_printf(sc->sc_dev, "%08X | time gp1\n", table.gp1);
device_printf(sc->sc_dev, "%08X | time gp2\n", table.gp2);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
device_printf(sc->sc_dev, "%08X | uCode revision type\n",
table.fw_rev_type);
device_printf(sc->sc_dev, "%08X | uCode version major\n", table.major);
device_printf(sc->sc_dev, "%08X | uCode version minor\n", table.minor);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev, "%08X | hw version\n", table.hw_ver);
device_printf(sc->sc_dev, "%08X | board version\n", table.brd_ver);
device_printf(sc->sc_dev, "%08X | hcmd\n", table.hcmd);
device_printf(sc->sc_dev, "%08X | isr0\n", table.isr0);
device_printf(sc->sc_dev, "%08X | isr1\n", table.isr1);
device_printf(sc->sc_dev, "%08X | isr2\n", table.isr2);
device_printf(sc->sc_dev, "%08X | isr3\n", table.isr3);
device_printf(sc->sc_dev, "%08X | isr4\n", table.isr4);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
device_printf(sc->sc_dev, "%08X | last cmd Id\n", table.last_cmd_id);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev, "%08X | wait_event\n", table.wait_event);
device_printf(sc->sc_dev, "%08X | l2p_control\n", table.l2p_control);
device_printf(sc->sc_dev, "%08X | l2p_duration\n", table.l2p_duration);
device_printf(sc->sc_dev, "%08X | l2p_mhvalid\n", table.l2p_mhvalid);
device_printf(sc->sc_dev, "%08X | l2p_addr_match\n", table.l2p_addr_match);
device_printf(sc->sc_dev, "%08X | lmpm_pmg_sel\n", table.lmpm_pmg_sel);
device_printf(sc->sc_dev, "%08X | timestamp\n", table.u_timestamp);
device_printf(sc->sc_dev, "%08X | flow_handler\n", table.flow_handler);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
if (sc->umac_error_event_table)
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
iwm_nic_umac_error(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
#endif
static void
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
iwm_handle_rxb(struct iwm_softc *sc, struct mbuf *m)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
iwm(4): switch to ieee80211_runtask() Submitted by: Andiry Voskoboinyk (s3erios@gmail.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D5054
2016-07-25 23:44:44 +00:00
struct ieee80211com *ic = &sc->sc_ic;
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
struct iwm_cmd_response *cresp;
struct mbuf *m1;
uint32_t offset = 0;
uint32_t maxoff = IWM_RBUF_SIZE;
uint32_t nextoff;
boolean_t stolen = FALSE;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
#define HAVEROOM(a) \
((a) + sizeof(uint32_t) + sizeof(struct iwm_cmd_header) < maxoff)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
while (HAVEROOM(offset)) {
struct iwm_rx_packet *pkt = mtodoff(m, struct iwm_rx_packet *,
offset);
int qid, idx, code, len;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
qid = pkt->hdr.qid;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
idx = pkt->hdr.idx;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
code = IWM_WIDE_ID(pkt->hdr.flags, pkt->hdr.code);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/*
* randomly get these from the firmware, no idea why.
* they at least seem harmless, so just ignore them for now
*/
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
if ((pkt->hdr.code == 0 && (qid & ~0x80) == 0 && idx == 0) ||
pkt->len_n_flags == htole32(IWM_FH_RSCSR_FRAME_INVALID)) {
break;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_INTR,
"rx packet qid=%d idx=%d type=%x\n",
qid & ~0x80, pkt->hdr.idx, code);
[iwm] Ignore IWM_DEBUG_LOG_MSG notifications. * Firmware versions 21 and 22 generate some IWM_DEBUG_LOG_MSG notifications, which seem to be harmless. Avoid spamming the system log with "frame ... UNHANDLED (this should not happen)" messages. Obtained from: dragonflybsd.git dda889ac57d8e5b46bb1b1ecf53c17a18481c7c8
2017-06-04 21:14:23 +00:00
len = iwm_rx_packet_len(pkt);
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
len += sizeof(uint32_t); /* account for status word */
nextoff = offset + roundup2(len, IWM_FH_RSCSR_FRAME_ALIGN);
[iwm] Add implementation of the notification wait api from iwlwifi. Obtained from: Linux iwlwifi Obtained from: DragonflyBSD commit 94dc1dadceb57b688036211262d678bc6bbdde37
2017-02-06 05:27:07 +00:00
iwm_notification_wait_notify(sc->sc_notif_wait, code, pkt);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
switch (code) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
case IWM_REPLY_RX_PHY_CMD:
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
iwm_mvm_rx_rx_phy_cmd(sc, pkt);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
break;
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
case IWM_REPLY_RX_MPDU_CMD: {
/*
* If this is the last frame in the RX buffer, we
* can directly feed the mbuf to the sharks here.
*/
struct iwm_rx_packet *nextpkt = mtodoff(m,
struct iwm_rx_packet *, nextoff);
if (!HAVEROOM(nextoff) ||
(nextpkt->hdr.code == 0 &&
(nextpkt->hdr.qid & ~0x80) == 0 &&
nextpkt->hdr.idx == 0) ||
(nextpkt->len_n_flags ==
htole32(IWM_FH_RSCSR_FRAME_INVALID))) {
if (iwm_mvm_rx_rx_mpdu(sc, m, offset, stolen)) {
stolen = FALSE;
/* Make sure we abort the loop */
nextoff = maxoff;
}
break;
}
/*
* Use m_copym instead of m_split, because that
* makes it easier to keep a valid rx buffer in
* the ring, when iwm_mvm_rx_rx_mpdu() fails.
*
* We need to start m_copym() at offset 0, to get the
* M_PKTHDR flag preserved.
*/
m1 = m_copym(m, 0, M_COPYALL, M_NOWAIT);
if (m1) {
if (iwm_mvm_rx_rx_mpdu(sc, m1, offset, stolen))
stolen = TRUE;
else
m_freem(m1);
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
break;
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
case IWM_TX_CMD:
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
iwm_mvm_rx_tx_cmd(sc, pkt);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
break;
case IWM_MISSED_BEACONS_NOTIFICATION: {
struct iwm_missed_beacons_notif *resp;
int missed;
/* XXX look at mac_id to determine interface ID */
struct ieee80211vap *vap = TAILQ_FIRST(&ic->ic_vaps);
[iwm] Get rid of SYNC_RESP_STRUCT and SYNC_RESP_PTR macros. * SYNC_RESP_STRUCT and SYNC_RESP_PTR originate from the OpenBSD version of iwm, and they weren't serving any real purpose in the FreeBSD port. * We just do a single bus_dmamap_sync for syncing the complete received frame, instead of explicitly bus_dmamap_sync-ing subranges of the frame like in the OpenBSD iwm code. Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D7939
2016-11-01 22:38:26 +00:00
resp = (void *)pkt->data;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
missed = le32toh(resp->consec_missed_beacons);
IWM_DPRINTF(sc, IWM_DEBUG_BEACON | IWM_DEBUG_STATE,
"%s: MISSED_BEACON: mac_id=%d, "
"consec_since_last_rx=%d, consec=%d, num_expect=%d "
"num_rx=%d\n",
__func__,
le32toh(resp->mac_id),
le32toh(resp->consec_missed_beacons_since_last_rx),
le32toh(resp->consec_missed_beacons),
le32toh(resp->num_expected_beacons),
le32toh(resp->num_recvd_beacons));
/* Be paranoid */
if (vap == NULL)
break;
/* XXX no net80211 locking? */
if (vap->iv_state == IEEE80211_S_RUN &&
(ic->ic_flags & IEEE80211_F_SCAN) == 0) {
if (missed > vap->iv_bmissthreshold) {
/* XXX bad locking; turn into task */
IWM_UNLOCK(sc);
ieee80211_beacon_miss(ic);
IWM_LOCK(sc);
}
}
[iwm] Get rid of struct iwm_rx_data argument for iwm_mvm_rx_rx_mpdu. Obtained from: dragonflybsd.git b5cdd8067951dc90271ab104ef555b3b5a4d5d5a
2017-03-23 04:33:15 +00:00
break;
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
case IWM_MFUART_LOAD_NOTIFICATION:
break;
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
case IWM_MVM_ALIVE:
break;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Change 2nd arg of iwm_phy_db_set_section() to struct iwm_rx_packet. * This matches the function declaration in Linux's iwlwifi. Obtained from: DragonflyBSD commit de7995a5e0ebf2d0016a87a0142a98c75db58fb4
2017-02-06 05:09:42 +00:00
case IWM_CALIB_RES_NOTIF_PHY_DB:
break;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Sync statistics API definitions with Linux iwlwifi. Obtained from: dragonflybsd.git 75895a53a9c1ba60d75be9b4bf6e49a37f91a7cf
2017-07-26 05:40:52 +00:00
case IWM_STATISTICS_NOTIFICATION:
iwm_mvm_handle_rx_statistics(sc, pkt);
[iwm] Get rid of struct iwm_rx_data argument for iwm_mvm_rx_rx_mpdu. Obtained from: dragonflybsd.git b5cdd8067951dc90271ab104ef555b3b5a4d5d5a
2017-03-23 04:33:15 +00:00
break;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
case IWM_NVM_ACCESS_CMD:
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
case IWM_MCC_UPDATE_CMD:
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
if (sc->sc_wantresp == (((qid & ~0x80) << 16) | idx)) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
memcpy(sc->sc_cmd_resp,
pkt, sizeof(sc->sc_cmd_resp));
}
break;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
case IWM_MCC_CHUB_UPDATE_CMD: {
struct iwm_mcc_chub_notif *notif;
[iwm] Get rid of SYNC_RESP_STRUCT and SYNC_RESP_PTR macros. * SYNC_RESP_STRUCT and SYNC_RESP_PTR originate from the OpenBSD version of iwm, and they weren't serving any real purpose in the FreeBSD port. * We just do a single bus_dmamap_sync for syncing the complete received frame, instead of explicitly bus_dmamap_sync-ing subranges of the frame like in the OpenBSD iwm code. Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D7939
2016-11-01 22:38:26 +00:00
notif = (void *)pkt->data;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
sc->sc_fw_mcc[0] = (notif->mcc & 0xff00) >> 8;
sc->sc_fw_mcc[1] = notif->mcc & 0xff;
sc->sc_fw_mcc[2] = '\0';
[iwm] Check for lar_disable tunable, and lar_enabled flag from NVM. * LAR can be disabled with the hw.iwm.lar.disable tunable now. * On Family 8000 devices we need to check the lar_enabled flag from nvm_data in addition to the TLV_CAPA_LAR_SUPPORT flag from the firmware. * Add a separate IWM_DEBUG_LAR debugging flag. Obtained from: dragonflybsd.git 0593e39cb295aa996ecf789ed4990c3b255f1770
2017-06-04 21:10:14 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_LAR,
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
"fw source %d sent CC '%s'\n",
notif->source_id, sc->sc_fw_mcc);
[iwm] Get rid of struct iwm_rx_data argument for iwm_mvm_rx_rx_mpdu. Obtained from: dragonflybsd.git b5cdd8067951dc90271ab104ef555b3b5a4d5d5a
2017-03-23 04:33:15 +00:00
break;
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
[iwm] Recognize IWM_DTS_MEASUREMENT_NOTIF_WIDE notification. * Add the command groups enum, and the iwm_phy_ops_subcmd_ids enum to if_iwmreg.h definitions. * The IWM_DTS_MEASUREMENT_NOTIF_WIDE notification will be generated by version 17 firmware. Taken-From: Linux iwlwifi Obtained from: DragonflyBSD commit 4d8d6f9def2ffb60aaf2d88f72f069a96c0b4e3f
2017-02-08 06:57:21 +00:00
case IWM_DTS_MEASUREMENT_NOTIFICATION:
case IWM_WIDE_ID(IWM_PHY_OPS_GROUP,
IWM_DTS_MEASUREMENT_NOTIF_WIDE): {
[iwm] Very basic DTS thermal sensor support (prints temp as debug msg). * Adds IWM_DEBUG_TEMP debug message type, for printing messages related to temperature sensors and thermal/TDP infos. * The firmware regularly sends us DTS measurement notifications, so just print the temperature value as a debugging message. (Adrian's addition): * Eventually this can be used by the driver to limit transmit rate / power to try and do some thermal throttling. Obtained from: DragonflyBSD commit efb7d4eb5c3140889a8880e12fd83c7bbfd0059d
2017-02-08 06:56:28 +00:00
struct iwm_dts_measurement_notif_v1 *notif;
if (iwm_rx_packet_payload_len(pkt) < sizeof(*notif)) {
device_printf(sc->sc_dev,
"Invalid DTS_MEASUREMENT_NOTIFICATION\n");
break;
}
notif = (void *)pkt->data;
IWM_DPRINTF(sc, IWM_DEBUG_TEMP,
"IWM_DTS_MEASUREMENT_NOTIFICATION - %d\n",
notif->temp);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
break;
[iwm] Very basic DTS thermal sensor support (prints temp as debug msg). * Adds IWM_DEBUG_TEMP debug message type, for printing messages related to temperature sensors and thermal/TDP infos. * The firmware regularly sends us DTS measurement notifications, so just print the temperature value as a debugging message. (Adrian's addition): * Eventually this can be used by the driver to limit transmit rate / power to try and do some thermal throttling. Obtained from: DragonflyBSD commit efb7d4eb5c3140889a8880e12fd83c7bbfd0059d
2017-02-08 06:56:28 +00:00
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
case IWM_PHY_CONFIGURATION_CMD:
case IWM_TX_ANT_CONFIGURATION_CMD:
case IWM_ADD_STA:
case IWM_MAC_CONTEXT_CMD:
case IWM_REPLY_SF_CFG_CMD:
case IWM_POWER_TABLE_CMD:
if_iwm - Configure the PCIe LTR, fix PCI express capability accesses. Taken-From: Linux iwlwifi Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (08a7ad5a5ff65aaaf2df6a609be7a4e1df43efc3)
2019-01-24 03:44:48 +00:00
case IWM_LTR_CONFIG:
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
case IWM_PHY_CONTEXT_CMD:
case IWM_BINDING_CONTEXT_CMD:
case IWM_TIME_EVENT_CMD:
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
case IWM_WIDE_ID(IWM_ALWAYS_LONG_GROUP, IWM_SCAN_CFG_CMD):
case IWM_WIDE_ID(IWM_ALWAYS_LONG_GROUP, IWM_SCAN_REQ_UMAC):
[iwm] Revert "if_iwm - SCAN_ABORT_UMAC response doesn't use a wide id" This reverts commit cef47a9cbb0a3ce5f18369fed9403d2764884bc2. Obtained from: dragonflybsd.git f62d325820ee7f7c2bcf721ada9cef8b70f74471
2017-05-12 06:34:43 +00:00
case IWM_WIDE_ID(IWM_ALWAYS_LONG_GROUP, IWM_SCAN_ABORT_UMAC):
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
case IWM_SCAN_OFFLOAD_REQUEST_CMD:
[iwm] Add scan abort functions, to properly cancel a running scan. * Uses the notification wait api to wait for the corresponding scan complete notification after sending the abort command. Taken-From: Linux iwlwifi Obtained from: DragonflyBSD commit b484d09d54301740f036ddf02008117f563960c2
2017-02-08 06:44:50 +00:00
case IWM_SCAN_OFFLOAD_ABORT_CMD:
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
case IWM_REPLY_BEACON_FILTERING_CMD:
case IWM_MAC_PM_POWER_TABLE:
case IWM_TIME_QUOTA_CMD:
case IWM_REMOVE_STA:
case IWM_TXPATH_FLUSH:
case IWM_LQ_CMD:
[iwm] Recognize IWM_FW_PAGING_BLOCK_CMD wide cmd response correctly. Obtained from: dragonflybsd.git ef688cebb9b29b67f7a011846589971987949e0d
2017-05-12 06:35:20 +00:00
case IWM_WIDE_ID(IWM_ALWAYS_LONG_GROUP,
IWM_FW_PAGING_BLOCK_CMD):
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
case IWM_BT_CONFIG:
case IWM_REPLY_THERMAL_MNG_BACKOFF:
[iwm] Get rid of SYNC_RESP_STRUCT and SYNC_RESP_PTR macros. * SYNC_RESP_STRUCT and SYNC_RESP_PTR originate from the OpenBSD version of iwm, and they weren't serving any real purpose in the FreeBSD port. * We just do a single bus_dmamap_sync for syncing the complete received frame, instead of explicitly bus_dmamap_sync-ing subranges of the frame like in the OpenBSD iwm code. Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D7939
2016-11-01 22:38:26 +00:00
cresp = (void *)pkt->data;
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
if (sc->sc_wantresp == (((qid & ~0x80) << 16) | idx)) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
memcpy(sc->sc_cmd_resp,
pkt, sizeof(*pkt)+sizeof(*cresp));
}
break;
/* ignore */
[iwm] Set command code for PHY_DB as well. Obtained from: dragonflybsd.git 58318c956a74382d1286ccabaf767012fdcfe1a2
2017-06-04 21:13:13 +00:00
case IWM_PHY_DB_CMD:
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
break;
case IWM_INIT_COMPLETE_NOTIF:
break;
[iwm] Enable Energy Based Scan (EBS). This can significantly reduce scan duration thus saving time and power. EBS failure reported by FW disables EBS for current connection. It is re-enabled upon new connection attempt on any WLAN interface. Obtained from: dragonflybsd.git 89f579e9823a5c446ca172cf82bbc210d6a054a4
2017-03-25 02:49:20 +00:00
case IWM_SCAN_OFFLOAD_COMPLETE:
iwm_mvm_rx_lmac_scan_complete_notif(sc, pkt);
[iwm] Use iwm_mvm_scan_stop_wait to properly abort scans. * Add IWM_FLAG_SCAN_RUNNING to sc->sc_flags to track whether the firmware is currently running a scan, in order to decide wheter iwm_scan_end needs to abort a running scan. * In iwm_scan_end, if the scan is still running, we now abort it, in order to keep the firmware scanning state in sync. * Try to make things a bit simpler, by reacting on the IWM_SCAN_OFFLOAD_COMPLETE and IWM_SCAN_COMPLETE_UMAC notifications, instead of IWM_SCAN_ITERATION_COMPLETE and IWM_SCAN_ITERATION_COMPLETE_UMAC. This should be fine since we always only tell the firmware to do a single scan iteration anyway. Obtained from: DragonflyBSD commit 1f249c981c4e89e7cde1836a75b61cac36dc7ac5
2017-02-08 06:50:59 +00:00
if (sc->sc_flags & IWM_FLAG_SCAN_RUNNING) {
sc->sc_flags &= ~IWM_FLAG_SCAN_RUNNING;
ieee80211_runtask(ic, &sc->sc_es_task);
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
break;
case IWM_SCAN_ITERATION_COMPLETE: {
struct iwm_lmac_scan_complete_notif *notif;
[iwm] Get rid of SYNC_RESP_STRUCT and SYNC_RESP_PTR macros. * SYNC_RESP_STRUCT and SYNC_RESP_PTR originate from the OpenBSD version of iwm, and they weren't serving any real purpose in the FreeBSD port. * We just do a single bus_dmamap_sync for syncing the complete received frame, instead of explicitly bus_dmamap_sync-ing subranges of the frame like in the OpenBSD iwm code. Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D7939
2016-11-01 22:38:26 +00:00
notif = (void *)pkt->data;
[iwm] Enable Energy Based Scan (EBS). This can significantly reduce scan duration thus saving time and power. EBS failure reported by FW disables EBS for current connection. It is re-enabled upon new connection attempt on any WLAN interface. Obtained from: dragonflybsd.git 89f579e9823a5c446ca172cf82bbc210d6a054a4
2017-03-25 02:49:20 +00:00
break;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
[iwm] Enable Energy Based Scan (EBS). This can significantly reduce scan duration thus saving time and power. EBS failure reported by FW disables EBS for current connection. It is re-enabled upon new connection attempt on any WLAN interface. Obtained from: dragonflybsd.git 89f579e9823a5c446ca172cf82bbc210d6a054a4
2017-03-25 02:49:20 +00:00
case IWM_SCAN_COMPLETE_UMAC:
iwm_mvm_rx_umac_scan_complete_notif(sc, pkt);
[iwm] Use iwm_mvm_scan_stop_wait to properly abort scans. * Add IWM_FLAG_SCAN_RUNNING to sc->sc_flags to track whether the firmware is currently running a scan, in order to decide wheter iwm_scan_end needs to abort a running scan. * In iwm_scan_end, if the scan is still running, we now abort it, in order to keep the firmware scanning state in sync. * Try to make things a bit simpler, by reacting on the IWM_SCAN_OFFLOAD_COMPLETE and IWM_SCAN_COMPLETE_UMAC notifications, instead of IWM_SCAN_ITERATION_COMPLETE and IWM_SCAN_ITERATION_COMPLETE_UMAC. This should be fine since we always only tell the firmware to do a single scan iteration anyway. Obtained from: DragonflyBSD commit 1f249c981c4e89e7cde1836a75b61cac36dc7ac5
2017-02-08 06:50:59 +00:00
if (sc->sc_flags & IWM_FLAG_SCAN_RUNNING) {
sc->sc_flags &= ~IWM_FLAG_SCAN_RUNNING;
ieee80211_runtask(ic, &sc->sc_es_task);
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
break;
case IWM_SCAN_ITERATION_COMPLETE_UMAC: {
struct iwm_umac_scan_iter_complete_notif *notif;
[iwm] Get rid of SYNC_RESP_STRUCT and SYNC_RESP_PTR macros. * SYNC_RESP_STRUCT and SYNC_RESP_PTR originate from the OpenBSD version of iwm, and they weren't serving any real purpose in the FreeBSD port. * We just do a single bus_dmamap_sync for syncing the complete received frame, instead of explicitly bus_dmamap_sync-ing subranges of the frame like in the OpenBSD iwm code. Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D7939
2016-11-01 22:38:26 +00:00
notif = (void *)pkt->data;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_SCAN, "UMAC scan iteration "
"complete, status=0x%x, %d channels scanned\n",
notif->status, notif->scanned_channels);
break;
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
case IWM_REPLY_ERROR: {
struct iwm_error_resp *resp;
[iwm] Get rid of SYNC_RESP_STRUCT and SYNC_RESP_PTR macros. * SYNC_RESP_STRUCT and SYNC_RESP_PTR originate from the OpenBSD version of iwm, and they weren't serving any real purpose in the FreeBSD port. * We just do a single bus_dmamap_sync for syncing the complete received frame, instead of explicitly bus_dmamap_sync-ing subranges of the frame like in the OpenBSD iwm code. Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D7939
2016-11-01 22:38:26 +00:00
resp = (void *)pkt->data;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev,
"firmware error 0x%x, cmd 0x%x\n",
le32toh(resp->error_type),
resp->cmd_id);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
break;
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm - Clear Time Event active state, when receiving End Notification. * This hopefully avoids some firmware panics, I was occasionally seeing, when iwm disconnects upon losing signal to an access point at some point. * This is synchronizing the if_iwm_time_event.c file a bit more from the corresponding Linux iwlwifi/mvm/time-event.c. Taken-From: Linux iwlwifi Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (e8cb71584a6a72232c13151d60e57f7f229220eb)
2019-01-24 03:47:47 +00:00
case IWM_TIME_EVENT_NOTIFICATION:
iwm_mvm_rx_time_event_notif(sc, pkt);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
break;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Ignore IWM_DEBUG_LOG_MSG notifications. * Firmware versions 21 and 22 generate some IWM_DEBUG_LOG_MSG notifications, which seem to be harmless. Avoid spamming the system log with "frame ... UNHANDLED (this should not happen)" messages. Obtained from: dragonflybsd.git dda889ac57d8e5b46bb1b1ecf53c17a18481c7c8
2017-06-04 21:14:23 +00:00
/*
* Firmware versions 21 and 22 generate some DEBUG_LOG_MSG
* messages. Just ignore them for now.
*/
case IWM_DEBUG_LOG_MSG:
break;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
case IWM_MCAST_FILTER_CMD:
break;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
case IWM_SCD_QUEUE_CFG: {
struct iwm_scd_txq_cfg_rsp *rsp;
[iwm] Get rid of SYNC_RESP_STRUCT and SYNC_RESP_PTR macros. * SYNC_RESP_STRUCT and SYNC_RESP_PTR originate from the OpenBSD version of iwm, and they weren't serving any real purpose in the FreeBSD port. * We just do a single bus_dmamap_sync for syncing the complete received frame, instead of explicitly bus_dmamap_sync-ing subranges of the frame like in the OpenBSD iwm code. Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D7939
2016-11-01 22:38:26 +00:00
rsp = (void *)pkt->data;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_CMD,
"queue cfg token=0x%x sta_id=%d "
"tid=%d scd_queue=%d\n",
rsp->token, rsp->sta_id, rsp->tid,
rsp->scd_queue);
break;
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
default:
device_printf(sc->sc_dev,
"frame %d/%d %x UNHANDLED (this should "
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
"not happen)\n", qid & ~0x80, idx,
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
pkt->len_n_flags);
break;
}
/*
* Why test bit 0x80? The Linux driver:
*
* There is one exception: uCode sets bit 15 when it
* originates the response/notification, i.e. when the
* response/notification is not a direct response to a
* command sent by the driver. For example, uCode issues
* IWM_REPLY_RX when it sends a received frame to the driver;
* it is not a direct response to any driver command.
*
* Ok, so since when is 7 == 15? Well, the Linux driver
* uses a slightly different format for pkt->hdr, and "qid"
* is actually the upper byte of a two-byte field.
*/
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
if (!(qid & (1 << 7)))
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm_cmd_done(sc, pkt);
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
offset = nextoff;
}
if (stolen)
m_freem(m);
#undef HAVEROOM
}
/*
* Process an IWM_CSR_INT_BIT_FH_RX or IWM_CSR_INT_BIT_SW_RX interrupt.
* Basic structure from if_iwn
*/
static void
iwm_notif_intr(struct iwm_softc *sc)
{
uint16_t hw;
bus_dmamap_sync(sc->rxq.stat_dma.tag, sc->rxq.stat_dma.map,
BUS_DMASYNC_POSTREAD);
hw = le16toh(sc->rxq.stat->closed_rb_num) & 0xfff;
/*
* Process responses
*/
while (sc->rxq.cur != hw) {
struct iwm_rx_ring *ring = &sc->rxq;
struct iwm_rx_data *data = &ring->data[ring->cur];
bus_dmamap_sync(ring->data_dmat, data->map,
BUS_DMASYNC_POSTREAD);
IWM_DPRINTF(sc, IWM_DEBUG_INTR,
"%s: hw = %d cur = %d\n", __func__, hw, ring->cur);
iwm_handle_rxb(sc, data->m);
ring->cur = (ring->cur + 1) % IWM_RX_RING_COUNT;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
/*
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
* Tell the firmware that it can reuse the ring entries that
* we have just processed.
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
* Seems like the hardware gets upset unless we align
* the write by 8??
*/
hw = (hw == 0) ? IWM_RX_RING_COUNT - 1 : hw - 1;
[iwm] Process multiple frames per RX buffer. * Factor out iwm_handle_rxb() function from iwm_notif_intr(). * Removing the IWM_FH_RCSR_CHNL0_RX_CONFIG_SINGLE_FRAME_MSK flag allows the device to put multiple frames (both command responses and 80211 frames) into a single RX buffer. * Uses m_copym() to split up the receive buffers when multiple 80211 frames are received in one RX buffer. The effect is basically the same as when using m_split(), but we want to keep the original mbuf around when calling iwm_mvm_rx_rx_mpdu() to make error handling a bit easier for now. * Contains a small optimization to avoid the m_copym() when only a single 80211 frame is received in one RX buffer (i.e. matching the existing behaviour). Obtained from: dragonflybsd.git b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67
2017-05-12 05:49:24 +00:00
IWM_WRITE(sc, IWM_FH_RSCSR_CHNL0_WPTR, rounddown2(hw, 8));
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
static void
iwm_intr(void *arg)
{
struct iwm_softc *sc = arg;
int handled = 0;
int r1, r2, rv = 0;
int isperiodic = 0;
IWM_LOCK(sc);
IWM_WRITE(sc, IWM_CSR_INT_MASK, 0);
if (sc->sc_flags & IWM_FLAG_USE_ICT) {
uint32_t *ict = sc->ict_dma.vaddr;
int tmp;
tmp = htole32(ict[sc->ict_cur]);
if (!tmp)
goto out_ena;
/*
* ok, there was something. keep plowing until we have all.
*/
r1 = r2 = 0;
while (tmp) {
r1 |= tmp;
ict[sc->ict_cur] = 0;
sc->ict_cur = (sc->ict_cur+1) % IWM_ICT_COUNT;
tmp = htole32(ict[sc->ict_cur]);
}
/* this is where the fun begins. don't ask */
if (r1 == 0xffffffff)
r1 = 0;
/* i am not expected to understand this */
if (r1 & 0xc0000)
r1 |= 0x8000;
r1 = (0xff & r1) | ((0xff00 & r1) << 16);
} else {
r1 = IWM_READ(sc, IWM_CSR_INT);
/* "hardware gone" (where, fishing?) */
if (r1 == 0xffffffff || (r1 & 0xfffffff0) == 0xa5a5a5a0)
goto out;
r2 = IWM_READ(sc, IWM_CSR_FH_INT_STATUS);
}
if (r1 == 0 && r2 == 0) {
goto out_ena;
}
IWM_WRITE(sc, IWM_CSR_INT, r1 | ~sc->sc_intmask);
[iwm] Sync IWM_MVM_ALIVE waiting and start_fw handling with iwlwifi. * Use the notification wait API, like it's done in the Linux iwlwifi code, to wait for the IWM_MVM_ALIVE notification. * This also should fix some firmware load interrupt issues, and errors in the nic lock using. Tested: * (adrian) Intel 7260, STA mode Obtained from: dragonflybsd.git a7697ea01c11fd493aec52260a02f31df680eb91
2017-02-22 02:23:14 +00:00
/* Safely ignore these bits for debug checks below */
r1 &= ~(IWM_CSR_INT_BIT_ALIVE | IWM_CSR_INT_BIT_SCD);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (r1 & IWM_CSR_INT_BIT_SW_ERR) {
int i;
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
struct ieee80211com *ic = &sc->sc_ic;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
struct ieee80211vap *vap = TAILQ_FIRST(&ic->ic_vaps);
The iwm_nic_error() defined if IWM_DEBUG is defined.
2016-04-29 19:10:22 +00:00
#ifdef IWM_DEBUG
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm_nic_error(sc);
The iwm_nic_error() defined if IWM_DEBUG is defined.
2016-04-29 19:10:22 +00:00
#endif
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Dump driver status (TX and RX rings) while we're here. */
device_printf(sc->sc_dev, "driver status:\n");
for (i = 0; i < IWM_MVM_MAX_QUEUES; i++) {
struct iwm_tx_ring *ring = &sc->txq[i];
device_printf(sc->sc_dev,
" tx ring %2d: qid=%-2d cur=%-3d "
"queued=%-3d\n",
i, ring->qid, ring->cur, ring->queued);
}
device_printf(sc->sc_dev,
" rx ring: cur=%d\n", sc->rxq.cur);
device_printf(sc->sc_dev,
iwm(4): Don't dereference potentially NULL pointer before NULL check Introduced in r298594. There is no path before the 'vap == NULL' check where vap is not already dereferenced. Reported by: Coverity CID: 1354979 Sponsored by: EMC / Isilon Storage Division
2016-04-26 19:06:28 +00:00
" 802.11 state %d\n", (vap == NULL) ? -1 : vap->iv_state);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm - Track firmware state better, and improve handling in iwm_newstate(). * This avoids firmware resets in all the cases in iwm_newstate(). Instead iwm_bring_down_firmware() is called, which tears down all the STA connection state, according to the sc->sc_firmware_state value. * Improve the behaviour of the LED blinking a bit, so it only blinks when there really is a wireless scan going on. * Print the newstate arg in debug output of iwm_newstate(), to help in debugging. This is inspired by the firmware state maintaining change in OpenBSD's iwm, by stsp@openbsd.org (OpenBSD Git 0ddb056fb7370664b1d4b84392697cb17d1a414a). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (8a41b10ac639d0609878696808387a6799d39b57)
2019-01-24 03:45:55 +00:00
/* Reset our firmware state tracking. */
sc->sc_firmware_state = 0;
[iwm] restart the VAP if there's a firmware panic. Always print out the firmware panic info before restarting; don't put it behind IWM_DEBUG. Submitted by: bapt Differential Revision: https://reviews.freebsd.org/D6081
2016-04-25 22:00:05 +00:00
/* Don't stop the device; just do a VAP restart */
IWM_UNLOCK(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] restart the VAP if there's a firmware panic. Always print out the firmware panic info before restarting; don't put it behind IWM_DEBUG. Submitted by: bapt Differential Revision: https://reviews.freebsd.org/D6081
2016-04-25 22:00:05 +00:00
if (vap == NULL) {
printf("%s: null vap\n", __func__);
return;
}
device_printf(sc->sc_dev, "%s: controller panicked, iv_state = %d; "
"restarting\n", __func__, vap->iv_state);
ieee80211_restart_all(ic);
return;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
if (r1 & IWM_CSR_INT_BIT_HW_ERR) {
handled |= IWM_CSR_INT_BIT_HW_ERR;
device_printf(sc->sc_dev, "hardware error, stopping device\n");
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
iwm_stop(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
rv = 1;
goto out;
}
/* firmware chunk loaded */
if (r1 & IWM_CSR_INT_BIT_FH_TX) {
IWM_WRITE(sc, IWM_CSR_FH_INT_STATUS, IWM_CSR_FH_INT_TX_MASK);
handled |= IWM_CSR_INT_BIT_FH_TX;
sc->sc_fw_chunk_done = 1;
wakeup(&sc->sc_fw);
}
if (r1 & IWM_CSR_INT_BIT_RF_KILL) {
handled |= IWM_CSR_INT_BIT_RF_KILL;
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
if (iwm_check_rfkill(sc)) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_printf(sc->sc_dev,
"%s: rfkill switch, disabling interface\n",
__func__);
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
iwm_stop(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
}
/*
* The Linux driver uses periodic interrupts to avoid races.
* We cargo-cult like it's going out of fashion.
*/
if (r1 & IWM_CSR_INT_BIT_RX_PERIODIC) {
handled |= IWM_CSR_INT_BIT_RX_PERIODIC;
IWM_WRITE(sc, IWM_CSR_INT, IWM_CSR_INT_BIT_RX_PERIODIC);
if ((r1 & (IWM_CSR_INT_BIT_FH_RX | IWM_CSR_INT_BIT_SW_RX)) == 0)
IWM_WRITE_1(sc,
IWM_CSR_INT_PERIODIC_REG, IWM_CSR_INT_PERIODIC_DIS);
isperiodic = 1;
}
if ((r1 & (IWM_CSR_INT_BIT_FH_RX | IWM_CSR_INT_BIT_SW_RX)) || isperiodic) {
handled |= (IWM_CSR_INT_BIT_FH_RX | IWM_CSR_INT_BIT_SW_RX);
IWM_WRITE(sc, IWM_CSR_FH_INT_STATUS, IWM_CSR_FH_INT_RX_MASK);
iwm_notif_intr(sc);
/* enable periodic interrupt, see above */
if (r1 & (IWM_CSR_INT_BIT_FH_RX | IWM_CSR_INT_BIT_SW_RX) && !isperiodic)
IWM_WRITE_1(sc, IWM_CSR_INT_PERIODIC_REG,
IWM_CSR_INT_PERIODIC_ENA);
}
if (__predict_false(r1 & ~handled))
IWM_DPRINTF(sc, IWM_DEBUG_INTR,
"%s: unhandled interrupts: %x\n", __func__, r1);
rv = 1;
out_ena:
iwm_restore_interrupts(sc);
out:
IWM_UNLOCK(sc);
return;
}
/*
* Autoconf glue-sniffing
*/
#define PCI_VENDOR_INTEL 0x8086
#define PCI_PRODUCT_INTEL_WL_3160_1 0x08b3
#define PCI_PRODUCT_INTEL_WL_3160_2 0x08b4
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
#define PCI_PRODUCT_INTEL_WL_3165_1 0x3165
#define PCI_PRODUCT_INTEL_WL_3165_2 0x3166
[iwm] Add support for iwm 3168 cards ``` iwm0@pci0:5:0:0: class=0x028000 card=0x21108086 chip=0x24fb8086 rev=0x10 hdr=0x00 vendor = 'Intel Corporation' device = 'Dual Band Wireless-AC 3168NGW [Stone Peak]' class = network [94829] iwm0: <Intel(R) Dual Band Wireless AC 3168> mem 0xef700000-0xef701fff at device 0.0 on pci5 [94829] iwm0: hw rev 0x220, fw ver 22.361476.0, address 28:c6:3f:15:43:c5 ``` MFC After: 2 weeks Reviewed by: ivadasz (over IRC) PR: 224886 Differential Revision: https://reviews.freebsd.org/D14865
2018-03-28 07:59:16 +00:00
#define PCI_PRODUCT_INTEL_WL_3168_1 0x24fb
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
#define PCI_PRODUCT_INTEL_WL_7260_1 0x08b1
#define PCI_PRODUCT_INTEL_WL_7260_2 0x08b2
#define PCI_PRODUCT_INTEL_WL_7265_1 0x095a
#define PCI_PRODUCT_INTEL_WL_7265_2 0x095b
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
#define PCI_PRODUCT_INTEL_WL_8260_1 0x24f3
#define PCI_PRODUCT_INTEL_WL_8260_2 0x24f4
Add support for Intel 8265 WiFi Obtained from: OpenBSD MFC after: 1 month
2017-10-09 15:48:56 +00:00
#define PCI_PRODUCT_INTEL_WL_8265_1 0x24fd
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static const struct iwm_devices {
[iwm] Simplify device id matching code a bit. * Just add the struct iwm_cfg pointers to the iwm_devices array, to get rid of the large switch clause. Obtained from: dragonflybsd.git 35f0e6c86c1654323d6b19f7a077f4ab8ac85868
2017-02-22 04:28:10 +00:00
uint16_t device;
const struct iwm_cfg *cfg;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
} iwm_devices[] = {
[iwm] Move struct iwm_cfg chipset configs to if_iwm_7000.c and _8000.c * This is more similar to how code/definitions are distributed in Linux's iwlwifi. * This should make recognizing new chipset variants, and adding additional flags from the Linux iwlwifi code easier, without blowing up if_iwm.c Obtained from: dragonflybsd.git 27d11320e707d2c41424efc1983762f6799941d6
2017-02-22 04:35:53 +00:00
{ PCI_PRODUCT_INTEL_WL_3160_1, &iwm3160_cfg },
{ PCI_PRODUCT_INTEL_WL_3160_2, &iwm3160_cfg },
{ PCI_PRODUCT_INTEL_WL_3165_1, &iwm3165_cfg },
{ PCI_PRODUCT_INTEL_WL_3165_2, &iwm3165_cfg },
[iwm] Add support for iwm 3168 cards ``` iwm0@pci0:5:0:0: class=0x028000 card=0x21108086 chip=0x24fb8086 rev=0x10 hdr=0x00 vendor = 'Intel Corporation' device = 'Dual Band Wireless-AC 3168NGW [Stone Peak]' class = network [94829] iwm0: <Intel(R) Dual Band Wireless AC 3168> mem 0xef700000-0xef701fff at device 0.0 on pci5 [94829] iwm0: hw rev 0x220, fw ver 22.361476.0, address 28:c6:3f:15:43:c5 ``` MFC After: 2 weeks Reviewed by: ivadasz (over IRC) PR: 224886 Differential Revision: https://reviews.freebsd.org/D14865
2018-03-28 07:59:16 +00:00
{ PCI_PRODUCT_INTEL_WL_3168_1, &iwm3168_cfg },
[iwm] Move struct iwm_cfg chipset configs to if_iwm_7000.c and _8000.c * This is more similar to how code/definitions are distributed in Linux's iwlwifi. * This should make recognizing new chipset variants, and adding additional flags from the Linux iwlwifi code easier, without blowing up if_iwm.c Obtained from: dragonflybsd.git 27d11320e707d2c41424efc1983762f6799941d6
2017-02-22 04:35:53 +00:00
{ PCI_PRODUCT_INTEL_WL_7260_1, &iwm7260_cfg },
{ PCI_PRODUCT_INTEL_WL_7260_2, &iwm7260_cfg },
{ PCI_PRODUCT_INTEL_WL_7265_1, &iwm7265_cfg },
{ PCI_PRODUCT_INTEL_WL_7265_2, &iwm7265_cfg },
{ PCI_PRODUCT_INTEL_WL_8260_1, &iwm8260_cfg },
{ PCI_PRODUCT_INTEL_WL_8260_2, &iwm8260_cfg },
Add support for Intel 8265 WiFi Obtained from: OpenBSD MFC after: 1 month
2017-10-09 15:48:56 +00:00
{ PCI_PRODUCT_INTEL_WL_8265_1, &iwm8265_cfg },
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
};
static int
iwm_probe(device_t dev)
{
int i;
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
for (i = 0; i < nitems(iwm_devices); i++) {
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
if (pci_get_vendor(dev) == PCI_VENDOR_INTEL &&
pci_get_device(dev) == iwm_devices[i].device) {
[iwm] Move struct iwm_cfg chipset configs to if_iwm_7000.c and _8000.c * This is more similar to how code/definitions are distributed in Linux's iwlwifi. * This should make recognizing new chipset variants, and adding additional flags from the Linux iwlwifi code easier, without blowing up if_iwm.c Obtained from: dragonflybsd.git 27d11320e707d2c41424efc1983762f6799941d6
2017-02-22 04:35:53 +00:00
device_set_desc(dev, iwm_devices[i].cfg->name);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
return (BUS_PROBE_DEFAULT);
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
return (ENXIO);
}
static int
iwm_dev_check(device_t dev)
{
struct iwm_softc *sc;
[iwm] Simplify device id matching code a bit. * Just add the struct iwm_cfg pointers to the iwm_devices array, to get rid of the large switch clause. Obtained from: dragonflybsd.git 35f0e6c86c1654323d6b19f7a077f4ab8ac85868
2017-02-22 04:28:10 +00:00
uint16_t devid;
int i;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
sc = device_get_softc(dev);
[iwm] Simplify device id matching code a bit. * Just add the struct iwm_cfg pointers to the iwm_devices array, to get rid of the large switch clause. Obtained from: dragonflybsd.git 35f0e6c86c1654323d6b19f7a077f4ab8ac85868
2017-02-22 04:28:10 +00:00
devid = pci_get_device(dev);
for (i = 0; i < nitems(iwm_devices); i++) {
if (iwm_devices[i].device == devid) {
sc->cfg = iwm_devices[i].cfg;
return (0);
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Simplify device id matching code a bit. * Just add the struct iwm_cfg pointers to the iwm_devices array, to get rid of the large switch clause. Obtained from: dragonflybsd.git 35f0e6c86c1654323d6b19f7a077f4ab8ac85868
2017-02-22 04:28:10 +00:00
device_printf(dev, "unknown adapter type\n");
return ENXIO;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
[iwm] Remove OpenBSD-specific comment. Beautify pci cfg space accesses. Obtained from: dragonflybsd.git 710fdd011c30a1bd9f53b87c843fb8907c4a6cfd
2017-02-22 04:27:07 +00:00
/* PCI registers */
#define PCI_CFG_RETRY_TIMEOUT 0x041
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static int
iwm_pci_attach(device_t dev)
{
struct iwm_softc *sc;
int count, error, rid;
uint16_t reg;
sc = device_get_softc(dev);
[iwm] Remove OpenBSD-specific comment. Beautify pci cfg space accesses. Obtained from: dragonflybsd.git 710fdd011c30a1bd9f53b87c843fb8907c4a6cfd
2017-02-22 04:27:07 +00:00
/* We disable the RETRY_TIMEOUT register (0x41) to keep
* PCI Tx retries from interfering with C3 CPU state */
pci_write_config(dev, PCI_CFG_RETRY_TIMEOUT, 0x00, 1);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Enable bus-mastering and hardware bug workaround. */
pci_enable_busmaster(dev);
reg = pci_read_config(dev, PCIR_STATUS, sizeof(reg));
/* if !MSI */
if (reg & PCIM_STATUS_INTxSTATE) {
reg &= ~PCIM_STATUS_INTxSTATE;
}
pci_write_config(dev, PCIR_STATUS, reg, sizeof(reg));
rid = PCIR_BAR(0);
sc->sc_mem = bus_alloc_resource_any(dev, SYS_RES_MEMORY, &rid,
RF_ACTIVE);
if (sc->sc_mem == NULL) {
device_printf(sc->sc_dev, "can't map mem space\n");
return (ENXIO);
}
sc->sc_st = rman_get_bustag(sc->sc_mem);
sc->sc_sh = rman_get_bushandle(sc->sc_mem);
/* Install interrupt handler. */
count = 1;
rid = 0;
if (pci_alloc_msi(dev, &count) == 0)
rid = 1;
sc->sc_irq = bus_alloc_resource_any(dev, SYS_RES_IRQ, &rid, RF_ACTIVE |
(rid != 0 ? 0 : RF_SHAREABLE));
if (sc->sc_irq == NULL) {
device_printf(dev, "can't map interrupt\n");
return (ENXIO);
}
error = bus_setup_intr(dev, sc->sc_irq, INTR_TYPE_NET | INTR_MPSAFE,
NULL, iwm_intr, sc, &sc->sc_ih);
if (sc->sc_ih == NULL) {
device_printf(dev, "can't establish interrupt");
return (ENXIO);
}
sc->sc_dmat = bus_get_dma_tag(sc->sc_dev);
return (0);
}
static void
iwm_pci_detach(device_t dev)
{
struct iwm_softc *sc = device_get_softc(dev);
if (sc->sc_irq != NULL) {
bus_teardown_intr(dev, sc->sc_irq, sc->sc_ih);
bus_release_resource(dev, SYS_RES_IRQ,
rman_get_rid(sc->sc_irq), sc->sc_irq);
pci_release_msi(dev);
}
if (sc->sc_mem != NULL)
bus_release_resource(dev, SYS_RES_MEMORY,
rman_get_rid(sc->sc_mem), sc->sc_mem);
}
static int
iwm_attach(device_t dev)
{
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
struct iwm_softc *sc = device_get_softc(dev);
struct ieee80211com *ic = &sc->sc_ic;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
int error;
int txq_i, i;
sc->sc_dev = dev;
[iwm] make sure we call iwm_detach_local() only once. Obtained from: DragonflyBSD git ebd4ceab76a6f161362029cbfd08efaedaab0519
2017-02-06 02:20:05 +00:00
sc->sc_attached = 1;
iwm: reorganize if_iwmvar.h - Change order of data in if_iwmvar.h (like it is in other drivers: defines, data structures, vap/node structures, softc struct and locks); use indentation. - Fix IWM_LOCK(_sc) / IWM_UNLOCK(_sc) macro. - Add IWM_LOCK_INIT / DESTROY(sc) + fix mtx_init() usage. - Wrap iwm_node casts into IWM_NODE() macro. - Drop some fields: * wt_hwqueue from Tx radiotap header; * macaddr[6] from iwm_vap; Approved by: adrian Differential Revision: https://reviews.freebsd.org/D4753
2016-01-03 10:06:10 +00:00
IWM_LOCK_INIT(sc);
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
mbufq_init(&sc->sc_snd, ifqmaxlen);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
callout_init_mtx(&sc->sc_watchdog_to, &sc->sc_mtx, 0);
[iwm] add LED blinking for iwm hardware that supports it. Submitted by: Imre Vadasz <imre@vdsz.com> Obtained from: DragonflyBSD, Linux iwlwifi/mvm
2016-06-02 04:42:28 +00:00
callout_init_mtx(&sc->sc_led_blink_to, &sc->sc_mtx, 0);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
TASK_INIT(&sc->sc_es_task, 0, iwm_endscan_cb, sc);
[iwm] Add implementation of the notification wait api from iwlwifi. Obtained from: Linux iwlwifi Obtained from: DragonflyBSD commit 94dc1dadceb57b688036211262d678bc6bbdde37
2017-02-06 05:27:07 +00:00
sc->sc_notif_wait = iwm_notification_wait_init(sc);
if (sc->sc_notif_wait == NULL) {
device_printf(dev, "failed to init notification wait struct\n");
goto fail;
}
[iwm] Move Smart Fifo handling into if_iwm_sf.c, sync with Linux iwlwifi. * This change also fixes a possible issue in the existing smart-fifo code, which set the IWM_SF_CFG_DUMMY_NOTIF_OFF bit on AC8260 chipsets, although that's only used in iwlwifi for Family 8000 chipsets connected via SDIO interface. Obtained from: Dragonflybsd.git cb650b01526b0aeef3c4307d926e7f1428997d50
2017-06-04 21:05:58 +00:00
sc->sf_state = IWM_SF_UNINIT;
[iwm] Sync if_iwm_phy_db code with Linux iwlwifi. Obtained from: Dragonflybsd commit c1019b6bfff36c856f7b4fccbdf3bb13ac27750c
2017-02-06 02:14:34 +00:00
/* Init phy db */
sc->sc_phy_db = iwm_phy_db_init(sc);
if (!sc->sc_phy_db) {
device_printf(dev, "Cannot init phy_db\n");
goto fail;
}
[iwm] Enable Energy Based Scan (EBS). This can significantly reduce scan duration thus saving time and power. EBS failure reported by FW disables EBS for current connection. It is re-enabled upon new connection attempt on any WLAN interface. Obtained from: dragonflybsd.git 89f579e9823a5c446ca172cf82bbc210d6a054a4
2017-03-25 02:49:20 +00:00
/* Set EBS as successful as long as not stated otherwise by the FW. */
sc->last_ebs_successful = TRUE;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* PCI attach */
error = iwm_pci_attach(dev);
if (error != 0)
goto fail;
sc->sc_wantresp = -1;
if_iwm - Move iwm_read_firmware() call into iwm_attach(). * We should load the firmware exactly once before the driver really initializes the hardware the first time, and unload it at detach time. There is no need to retrieve the firmware during execution of iwm_mvm_load_ucode_wait_alive(), we should make sure we already have the firmware data at hand before that. * The existing sc_preinit_hook code fails to deal with the case where if_iwm is loaded by the loader (or is statically linked) and the firmware needs to be loaded from disk. So we can just call iwm_read_firmware() from iwm_attach() directly. * A separate solution will have to be added to properly defer the firmware loading during bootup, until the necessary filesystem is mounted. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (0104ee1f4cb6a2313c00c2526c6ae98d42e5041d)
2019-01-24 03:42:23 +00:00
/* Match device id */
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
error = iwm_dev_check(dev);
if (error != 0)
goto fail;
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
sc->sc_hw_rev = IWM_READ(sc, IWM_CSR_HW_REV);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
/*
* In the 8000 HW family the format of the 4 bytes of CSR_HW_REV have
* changed, and now the revision step also includes bit 0-1 (no more
* "dash" value). To keep hw_rev backwards compatible - we'll store it
* in the old format.
*/
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
if (sc->cfg->device_family == IWM_DEVICE_FAMILY_8000) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
int ret;
uint32_t hw_step;
if_iwm - The iwm_prepare_card_hw() in iwm_attach() is only needed on 8K hw. * Doing the iwm_prepare_card_hw() call in iwm_attach() only on Family 8000 hardware matches the code in Linux iwlwifi. * While there remove DEFAULT_MAX_TX_POWER definition which is unused, and has a value different from IWL_DEFAULT_MAX_TX_POWER in iwlwifi. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (e8560f8dc58df12a7c79a6bb4e6ccb156e001085)
2019-01-24 03:41:44 +00:00
sc->sc_hw_rev = (sc->sc_hw_rev & 0xfff0) |
(IWM_CSR_HW_REV_STEP(sc->sc_hw_rev << 2) << 2);
if (iwm_prepare_card_hw(sc) != 0) {
device_printf(dev, "could not initialize hardware\n");
goto fail;
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
/*
* In order to recognize C step the driver should read the
* chip version id located at the AUX bus MISC address.
*/
IWM_SETBITS(sc, IWM_CSR_GP_CNTRL,
IWM_CSR_GP_CNTRL_REG_FLAG_INIT_DONE);
DELAY(2);
ret = iwm_poll_bit(sc, IWM_CSR_GP_CNTRL,
IWM_CSR_GP_CNTRL_REG_FLAG_MAC_CLOCK_READY,
IWM_CSR_GP_CNTRL_REG_FLAG_MAC_CLOCK_READY,
25000);
[iwm] Fix iwm_poll_bit() error value check in iwm_attach(). The iwm(4) iwm_poll_bit() function returns 1 on success, and 0 on failure, whereas the iwl_poll_bit() in Linux iwlwifi returns < 0 on failure. So the (ret < 0) check ended up ignoring any error returned by iwm_poll_bit(). Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D7932
2016-09-19 22:13:08 +00:00
if (!ret) {
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
device_printf(sc->sc_dev,
"Failed to wake up the nic\n");
goto fail;
}
if (iwm_nic_lock(sc)) {
hw_step = iwm_read_prph(sc, IWM_WFPM_CTRL_REG);
hw_step |= IWM_ENABLE_WFPM;
iwm_write_prph(sc, IWM_WFPM_CTRL_REG, hw_step);
hw_step = iwm_read_prph(sc, IWM_AUX_MISC_REG);
hw_step = (hw_step >> IWM_HW_STEP_LOCATION_BITS) & 0xF;
if (hw_step == 0x3)
sc->sc_hw_rev = (sc->sc_hw_rev & 0xFFFFFFF3) |
(IWM_SILICON_C_STEP << 2);
iwm_nic_unlock(sc);
} else {
device_printf(sc->sc_dev, "Failed to lock the nic\n");
goto fail;
}
}
[iwm] Use chipset configuration structs like iwlwifi does. * This makes it a bit easier to factor out common parts for e.g. the 7000 chipset family. * Add iwm7265d config, and recognize the 7265D chipset variant via the hardware revision. Tested: * 7260, STA mode (2ghz) Obtained from: Dragonflybsd commit cc8d6ccf5583fd45964f3bde9b057ee4f834c0e0
2017-02-06 03:29:50 +00:00
/* special-case 7265D, it has the same PCI IDs. */
if (sc->cfg == &iwm7265_cfg &&
(sc->sc_hw_rev & IWM_CSR_HW_REV_TYPE_MSK) == IWM_CSR_HW_REV_TYPE_7265D) {
sc->cfg = &iwm7265d_cfg;
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Allocate DMA memory for firmware transfers. */
if ((error = iwm_alloc_fwmem(sc)) != 0) {
device_printf(dev, "could not allocate memory for firmware\n");
goto fail;
}
/* Allocate "Keep Warm" page. */
if ((error = iwm_alloc_kw(sc)) != 0) {
device_printf(dev, "could not allocate keep warm page\n");
goto fail;
}
/* We use ICT interrupts */
if ((error = iwm_alloc_ict(sc)) != 0) {
device_printf(dev, "could not allocate ICT table\n");
goto fail;
}
/* Allocate TX scheduler "rings". */
if ((error = iwm_alloc_sched(sc)) != 0) {
device_printf(dev, "could not allocate TX scheduler rings\n");
goto fail;
}
/* Allocate TX rings */
for (txq_i = 0; txq_i < nitems(sc->txq); txq_i++) {
if ((error = iwm_alloc_tx_ring(sc,
&sc->txq[txq_i], txq_i)) != 0) {
device_printf(dev,
"could not allocate TX ring %d\n",
txq_i);
goto fail;
}
}
/* Allocate RX ring. */
if ((error = iwm_alloc_rx_ring(sc, &sc->rxq)) != 0) {
device_printf(dev, "could not allocate RX ring\n");
goto fail;
}
/* Clear pending interrupts. */
IWM_WRITE(sc, IWM_CSR_INT, 0xffffffff);
ic->ic_softc = sc;
ic->ic_name = device_get_nameunit(sc->sc_dev);
ic->ic_phytype = IEEE80211_T_OFDM; /* not only, but not used */
ic->ic_opmode = IEEE80211_M_STA; /* default to BSS mode */
/* Set device capabilities. */
ic->ic_caps =
IEEE80211_C_STA |
IEEE80211_C_WPA | /* WPA/RSN */
IEEE80211_C_WME |
[iwm] Add basic powermanagement support via ifconfig wlan0 powersave. * The DEVICE_POWER_FLAGS_CAM_MSK flag was removed in the upstream iwlwifi in Linux commit ceef91c89480dd18bb3ac51e91280a233d0ca41f. * Add sc_ps_disabled flag to struct iwm_softc, which corresponds to mvm->ps_disabled in struct iwl_mvm in Linux iwlwifi. * Adds a hw.iwm.power_scheme tunable which corresponds to the power_scheme module parameter in Linux iwlwifi. Set this to 1 for completely disabling power management, 2 (default) for balanced powermanagement, and 3 for lowerpower mode (which does dtim period skipping). * Imports the constants.h file from iwlwifi as if_iwm_constants.h. * This doesn't allow changing the powermanagement setting while connected, also one can only choose between enabled and disabled powersaving with ifconfig (so switching between balanced and low-power mode requires rebooting to change the tunable). * After any changes to powermanagement (i.e. "ifconfig wlan0 powersave" to enable powermanagement, or "ifconfig wlan0 -powersave" for disabling powermanagement), one has to disconnect and reconnect to the accespoint for the change to take effect. Obtained from: dragonflybsd.git d7002a7990d077c92585978ea998474af50f91e0
2017-05-09 04:15:07 +00:00
IEEE80211_C_PMGT |
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IEEE80211_C_SHSLOT | /* short slot time supported */
IEEE80211_C_SHPREAMBLE /* short preamble supported */
// IEEE80211_C_BGSCAN /* capable of bg scanning */
;
[iwm] Set full-offload scan flag. Fixes fw panic when already associated. * Starting a scan from wpa_supplicant or via ifconfig while associated, should no longer cause firmware panics or abort early. Tested: * AC7260, STA mode Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D8412
2016-11-01 22:47:58 +00:00
/* Advertise full-offload scanning */
ic->ic_flags_ext = IEEE80211_FEXT_SCAN_OFFLOAD;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
for (i = 0; i < nitems(sc->sc_phyctxt); i++) {
sc->sc_phyctxt[i].id = i;
sc->sc_phyctxt[i].color = 0;
sc->sc_phyctxt[i].ref = 0;
sc->sc_phyctxt[i].channel = NULL;
}
[iwm] fix up RSSI calculations for both scan results and normal RX operations. * hard code a noise floor of -96 for now. The noise floor update code returns some "interesting" values that I can't map to anything useful right now. * Ensure a default noise floor is set - otherwise the initial scan results have a noise floor of '0'. * Fix up the RSSI calculation to be correctly relative to the noise floor. The RSSI routines return an absolute value in dBm - so fix this up. * Cap RSSI values appropriately. * Ensure we pass in a 1/2 dB unit value in to net80211. Tested: * Intel 7260, STA mode iwm0: <Intel Dual Band Wireless AC 7260> mem 0xf1400000-0xf1401fff irq 17 at device 0.0 on pci2 iwm0: hw rev 0x140, fw ver 16.242414.0, address xx:xx:xx:xx:xx:xx
2016-09-18 05:07:18 +00:00
/* Default noise floor */
sc->sc_noise = -96;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Max RSSI */
sc->sc_max_rssi = IWM_MAX_DBM - IWM_MIN_DBM;
[iwm] fix up RSSI calculations for both scan results and normal RX operations. * hard code a noise floor of -96 for now. The noise floor update code returns some "interesting" values that I can't map to anything useful right now. * Ensure a default noise floor is set - otherwise the initial scan results have a noise floor of '0'. * Fix up the RSSI calculation to be correctly relative to the noise floor. The RSSI routines return an absolute value in dBm - so fix this up. * Cap RSSI values appropriately. * Ensure we pass in a 1/2 dB unit value in to net80211. Tested: * Intel 7260, STA mode iwm0: <Intel Dual Band Wireless AC 7260> mem 0xf1400000-0xf1401fff irq 17 at device 0.0 on pci2 iwm0: hw rev 0x140, fw ver 16.242414.0, address xx:xx:xx:xx:xx:xx
2016-09-18 05:07:18 +00:00
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
#ifdef IWM_DEBUG
SYSCTL_ADD_INT(device_get_sysctl_ctx(dev),
SYSCTL_CHILDREN(device_get_sysctl_tree(dev)), OID_AUTO, "debug",
CTLFLAG_RW, &sc->sc_debug, 0, "control debugging");
#endif
if_iwm - Move iwm_read_firmware() call into iwm_attach(). * We should load the firmware exactly once before the driver really initializes the hardware the first time, and unload it at detach time. There is no need to retrieve the firmware during execution of iwm_mvm_load_ucode_wait_alive(), we should make sure we already have the firmware data at hand before that. * The existing sc_preinit_hook code fails to deal with the case where if_iwm is loaded by the loader (or is statically linked) and the firmware needs to be loaded from disk. So we can just call iwm_read_firmware() from iwm_attach() directly. * A separate solution will have to be added to properly defer the firmware loading during bootup, until the necessary filesystem is mounted. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (0104ee1f4cb6a2313c00c2526c6ae98d42e5041d)
2019-01-24 03:42:23 +00:00
error = iwm_read_firmware(sc);
if (error) {
goto fail;
} else if (sc->sc_fw.fw_fp == NULL) {
/*
* XXX Add a solution for properly deferring firmware load
* during bootup.
*/
goto fail;
} else {
sc->sc_preinit_hook.ich_func = iwm_preinit;
sc->sc_preinit_hook.ich_arg = sc;
if (config_intrhook_establish(&sc->sc_preinit_hook) != 0) {
device_printf(dev,
"config_intrhook_establish failed\n");
goto fail;
}
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_RESET | IWM_DEBUG_TRACE,
"<-%s\n", __func__);
return 0;
/* Free allocated memory if something failed during attachment. */
fail:
iwm_detach_local(sc, 0);
return ENXIO;
}
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
static int
iwm_is_valid_ether_addr(uint8_t *addr)
{
char zero_addr[IEEE80211_ADDR_LEN] = { 0, 0, 0, 0, 0, 0 };
if ((addr[0] & 1) || IEEE80211_ADDR_EQ(zero_addr, addr))
return (FALSE);
return (TRUE);
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static int
[iwm] Properly implement iwm_wme_update callback function. * Inspired by iwn(4) and Linux iwlwifi. * Read wme parameters into a buffer within struct iwm_vap in iwm_wme_update(). * If we aren't associated yet, the new settings will soon be sent by iwm_mvm_mac_ctxt_changed() during association. * If we are already associated, explicitly call iwm_mvm_mac_ctxt_changed() from iwm_wme_update() to send the new settings to the firmware. * Change iwm_mvm_ac_to_tx_fifo mapping, to fit the freebsd net80211 WME stream class numbering, instead of Linux's enum ieee80211_ac_numbers. Obtained from: dragonflybsd.git b8bd6cd746d1f45e616ccfcbeed06dfe452a1108
2017-05-12 05:50:38 +00:00
iwm_wme_update(struct ieee80211com *ic)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
{
[iwm] Properly implement iwm_wme_update callback function. * Inspired by iwn(4) and Linux iwlwifi. * Read wme parameters into a buffer within struct iwm_vap in iwm_wme_update(). * If we aren't associated yet, the new settings will soon be sent by iwm_mvm_mac_ctxt_changed() during association. * If we are already associated, explicitly call iwm_mvm_mac_ctxt_changed() from iwm_wme_update() to send the new settings to the firmware. * Change iwm_mvm_ac_to_tx_fifo mapping, to fit the freebsd net80211 WME stream class numbering, instead of Linux's enum ieee80211_ac_numbers. Obtained from: dragonflybsd.git b8bd6cd746d1f45e616ccfcbeed06dfe452a1108
2017-05-12 05:50:38 +00:00
#define IWM_EXP2(x) ((1 << (x)) - 1) /* CWmin = 2^ECWmin - 1 */
Convert more ifp->if_softc -> ic->ic_softc. These should be a big no-op. Tested: * make universe
2015-08-17 23:44:38 +00:00
struct iwm_softc *sc = ic->ic_softc;
[net80211] convert all of the WME use over to a temporary copy of WME info. This removes the direct WME info access in the ieee80211com struct and instead provides a method of fetching the data. Right now it's a no-op but eventually it'll turn into a per-VAP method for drivers that support it (eg iwn, iwm, upcoming ath10k work) as things like p2p support require this kind of behaviour. Tested: * ath(4), STA and AP mode TODO: * yes, this is slightly stack size-y, but it is an important first step to get drivers migrated over to a sensible WME API. A lot of per-phy things need to be converted to per-VAP before P2P, 11ac firmware, etc stuff shows up.
2018-01-02 00:07:28 +00:00
struct chanAccParams chp;
[iwm] Properly implement iwm_wme_update callback function. * Inspired by iwn(4) and Linux iwlwifi. * Read wme parameters into a buffer within struct iwm_vap in iwm_wme_update(). * If we aren't associated yet, the new settings will soon be sent by iwm_mvm_mac_ctxt_changed() during association. * If we are already associated, explicitly call iwm_mvm_mac_ctxt_changed() from iwm_wme_update() to send the new settings to the firmware. * Change iwm_mvm_ac_to_tx_fifo mapping, to fit the freebsd net80211 WME stream class numbering, instead of Linux's enum ieee80211_ac_numbers. Obtained from: dragonflybsd.git b8bd6cd746d1f45e616ccfcbeed06dfe452a1108
2017-05-12 05:50:38 +00:00
struct ieee80211vap *vap = TAILQ_FIRST(&ic->ic_vaps);
struct iwm_vap *ivp = IWM_VAP(vap);
struct iwm_node *in;
struct wmeParams tmp[WME_NUM_AC];
int aci, error;
if (vap == NULL)
return (0);
[net80211] convert all of the WME use over to a temporary copy of WME info. This removes the direct WME info access in the ieee80211com struct and instead provides a method of fetching the data. Right now it's a no-op but eventually it'll turn into a per-VAP method for drivers that support it (eg iwn, iwm, upcoming ath10k work) as things like p2p support require this kind of behaviour. Tested: * ath(4), STA and AP mode TODO: * yes, this is slightly stack size-y, but it is an important first step to get drivers migrated over to a sensible WME API. A lot of per-phy things need to be converted to per-VAP before P2P, 11ac firmware, etc stuff shows up.
2018-01-02 00:07:28 +00:00
ieee80211_wme_ic_getparams(ic, &chp);
[iwm] Properly implement iwm_wme_update callback function. * Inspired by iwn(4) and Linux iwlwifi. * Read wme parameters into a buffer within struct iwm_vap in iwm_wme_update(). * If we aren't associated yet, the new settings will soon be sent by iwm_mvm_mac_ctxt_changed() during association. * If we are already associated, explicitly call iwm_mvm_mac_ctxt_changed() from iwm_wme_update() to send the new settings to the firmware. * Change iwm_mvm_ac_to_tx_fifo mapping, to fit the freebsd net80211 WME stream class numbering, instead of Linux's enum ieee80211_ac_numbers. Obtained from: dragonflybsd.git b8bd6cd746d1f45e616ccfcbeed06dfe452a1108
2017-05-12 05:50:38 +00:00
IEEE80211_LOCK(ic);
for (aci = 0; aci < WME_NUM_AC; aci++)
[net80211] convert all of the WME use over to a temporary copy of WME info. This removes the direct WME info access in the ieee80211com struct and instead provides a method of fetching the data. Right now it's a no-op but eventually it'll turn into a per-VAP method for drivers that support it (eg iwn, iwm, upcoming ath10k work) as things like p2p support require this kind of behaviour. Tested: * ath(4), STA and AP mode TODO: * yes, this is slightly stack size-y, but it is an important first step to get drivers migrated over to a sensible WME API. A lot of per-phy things need to be converted to per-VAP before P2P, 11ac firmware, etc stuff shows up.
2018-01-02 00:07:28 +00:00
tmp[aci] = chp.cap_wmeParams[aci];
[iwm] Properly implement iwm_wme_update callback function. * Inspired by iwn(4) and Linux iwlwifi. * Read wme parameters into a buffer within struct iwm_vap in iwm_wme_update(). * If we aren't associated yet, the new settings will soon be sent by iwm_mvm_mac_ctxt_changed() during association. * If we are already associated, explicitly call iwm_mvm_mac_ctxt_changed() from iwm_wme_update() to send the new settings to the firmware. * Change iwm_mvm_ac_to_tx_fifo mapping, to fit the freebsd net80211 WME stream class numbering, instead of Linux's enum ieee80211_ac_numbers. Obtained from: dragonflybsd.git b8bd6cd746d1f45e616ccfcbeed06dfe452a1108
2017-05-12 05:50:38 +00:00
IEEE80211_UNLOCK(ic);
IWM_LOCK(sc);
for (aci = 0; aci < WME_NUM_AC; aci++) {
const struct wmeParams *ac = &tmp[aci];
ivp->queue_params[aci].aifsn = ac->wmep_aifsn;
ivp->queue_params[aci].cw_min = IWM_EXP2(ac->wmep_logcwmin);
ivp->queue_params[aci].cw_max = IWM_EXP2(ac->wmep_logcwmax);
ivp->queue_params[aci].edca_txop =
IEEE80211_TXOP_TO_US(ac->wmep_txopLimit);
}
ivp->have_wme = TRUE;
if (ivp->is_uploaded && vap->iv_bss != NULL) {
in = IWM_NODE(vap->iv_bss);
if (in->in_assoc) {
if ((error = iwm_mvm_mac_ctxt_changed(sc, vap)) != 0) {
device_printf(sc->sc_dev,
"%s: failed to update MAC\n", __func__);
}
}
}
IWM_UNLOCK(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
return (0);
[iwm] Properly implement iwm_wme_update callback function. * Inspired by iwn(4) and Linux iwlwifi. * Read wme parameters into a buffer within struct iwm_vap in iwm_wme_update(). * If we aren't associated yet, the new settings will soon be sent by iwm_mvm_mac_ctxt_changed() during association. * If we are already associated, explicitly call iwm_mvm_mac_ctxt_changed() from iwm_wme_update() to send the new settings to the firmware. * Change iwm_mvm_ac_to_tx_fifo mapping, to fit the freebsd net80211 WME stream class numbering, instead of Linux's enum ieee80211_ac_numbers. Obtained from: dragonflybsd.git b8bd6cd746d1f45e616ccfcbeed06dfe452a1108
2017-05-12 05:50:38 +00:00
#undef IWM_EXP2
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
static void
iwm_preinit(void *arg)
{
struct iwm_softc *sc = arg;
device_t dev = sc->sc_dev;
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
struct ieee80211com *ic = &sc->sc_ic;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
int error;
IWM_DPRINTF(sc, IWM_DEBUG_RESET | IWM_DEBUG_TRACE,
"->%s\n", __func__);
IWM_LOCK(sc);
if ((error = iwm_start_hw(sc)) != 0) {
device_printf(dev, "could not initialize hardware\n");
IWM_UNLOCK(sc);
goto fail;
}
error = iwm_run_init_mvm_ucode(sc, 1);
iwm_stop_device(sc);
if (error) {
IWM_UNLOCK(sc);
goto fail;
}
device_printf(dev,
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
"hw rev 0x%x, fw ver %s, address %s\n",
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
sc->sc_hw_rev & IWM_CSR_HW_REV_TYPE_MSK,
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
sc->sc_fwver, ether_sprintf(sc->nvm_data->hw_addr));
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* not all hardware can do 5GHz band */
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
if (!sc->nvm_data->sku_cap_band_52GHz_enable)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
memset(&ic->ic_sup_rates[IEEE80211_MODE_11A], 0,
sizeof(ic->ic_sup_rates[IEEE80211_MODE_11A]));
IWM_UNLOCK(sc);
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
iwm_init_channel_map(ic, IEEE80211_CHAN_MAX, &ic->ic_nchans,
ic->ic_channels);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/*
* At this point we've committed - if we fail to do setup,
* we now also have to tear down the net80211 state.
*/
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
ieee80211_ifattach(ic);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
ic->ic_vap_create = iwm_vap_create;
ic->ic_vap_delete = iwm_vap_delete;
ic->ic_raw_xmit = iwm_raw_xmit;
ic->ic_node_alloc = iwm_node_alloc;
ic->ic_scan_start = iwm_scan_start;
ic->ic_scan_end = iwm_scan_end;
ic->ic_update_mcast = iwm_update_mcast;
iwm: switch to ieee80211_add_channel() - Switch to ieee80211_add_channel(). - Parse/set more channel flags (DFS, NOADHOC). - Add ic_getradiocaps() method. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D6150
2016-05-01 14:27:27 +00:00
ic->ic_getradiocaps = iwm_init_channel_map;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
ic->ic_set_channel = iwm_set_channel;
ic->ic_scan_curchan = iwm_scan_curchan;
ic->ic_scan_mindwell = iwm_scan_mindwell;
[iwm] Properly implement iwm_wme_update callback function. * Inspired by iwn(4) and Linux iwlwifi. * Read wme parameters into a buffer within struct iwm_vap in iwm_wme_update(). * If we aren't associated yet, the new settings will soon be sent by iwm_mvm_mac_ctxt_changed() during association. * If we are already associated, explicitly call iwm_mvm_mac_ctxt_changed() from iwm_wme_update() to send the new settings to the firmware. * Change iwm_mvm_ac_to_tx_fifo mapping, to fit the freebsd net80211 WME stream class numbering, instead of Linux's enum ieee80211_ac_numbers. Obtained from: dragonflybsd.git b8bd6cd746d1f45e616ccfcbeed06dfe452a1108
2017-05-12 05:50:38 +00:00
ic->ic_wme.wme_update = iwm_wme_update;
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
ic->ic_parent = iwm_parent;
ic->ic_transmit = iwm_transmit;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm_radiotap_attach(sc);
if (bootverbose)
ieee80211_announce(ic);
IWM_DPRINTF(sc, IWM_DEBUG_RESET | IWM_DEBUG_TRACE,
"<-%s\n", __func__);
config_intrhook_disestablish(&sc->sc_preinit_hook);
return;
fail:
config_intrhook_disestablish(&sc->sc_preinit_hook);
iwm_detach_local(sc, 0);
}
/*
* Attach the interface to 802.11 radiotap.
*/
static void
iwm_radiotap_attach(struct iwm_softc *sc)
{
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
struct ieee80211com *ic = &sc->sc_ic;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_DPRINTF(sc, IWM_DEBUG_RESET | IWM_DEBUG_TRACE,
"->%s begin\n", __func__);
ieee80211_radiotap_attach(ic,
&sc->sc_txtap.wt_ihdr, sizeof(sc->sc_txtap),
IWM_TX_RADIOTAP_PRESENT,
&sc->sc_rxtap.wr_ihdr, sizeof(sc->sc_rxtap),
IWM_RX_RADIOTAP_PRESENT);
IWM_DPRINTF(sc, IWM_DEBUG_RESET | IWM_DEBUG_TRACE,
"->%s end\n", __func__);
}
static struct ieee80211vap *
iwm_vap_create(struct ieee80211com *ic, const char name[IFNAMSIZ], int unit,
enum ieee80211_opmode opmode, int flags,
const uint8_t bssid[IEEE80211_ADDR_LEN],
const uint8_t mac[IEEE80211_ADDR_LEN])
{
struct iwm_vap *ivp;
struct ieee80211vap *vap;
if (!TAILQ_EMPTY(&ic->ic_vaps)) /* only one at a time */
return NULL;
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
ivp = malloc(sizeof(struct iwm_vap), M_80211_VAP, M_WAITOK | M_ZERO);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
vap = &ivp->iv_vap;
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
ieee80211_vap_setup(ic, vap, name, unit, opmode, flags, bssid);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
vap->iv_bmissthreshold = 10; /* override default */
/* Override with driver methods. */
ivp->iv_newstate = vap->iv_newstate;
vap->iv_newstate = iwm_newstate;
[if_iwm] Partly sync if_iwm_binding.c to Linux iwlwifi code. * Store macid and color values in struct iwm_vap, to avoid hardcoded constants a bit. * Add iwm_mvm_binding_remove_vif() function (will be used in disconnecting from an access point without resetting the whole device). * Not adding code from Linux iwlwifi yet, to handle one PHY context to be bound to several VAPs/virtual-interfaces, it's definitely not needed in the near future. Obtained from: dragonflybsd.git f16ef74977e51e1bfc7a625dd18b98b02158e0e5
2017-05-12 05:19:08 +00:00
ivp->id = IWM_DEFAULT_MACID;
ivp->color = IWM_DEFAULT_COLOR;
[iwm] Properly implement iwm_wme_update callback function. * Inspired by iwn(4) and Linux iwlwifi. * Read wme parameters into a buffer within struct iwm_vap in iwm_wme_update(). * If we aren't associated yet, the new settings will soon be sent by iwm_mvm_mac_ctxt_changed() during association. * If we are already associated, explicitly call iwm_mvm_mac_ctxt_changed() from iwm_wme_update() to send the new settings to the firmware. * Change iwm_mvm_ac_to_tx_fifo mapping, to fit the freebsd net80211 WME stream class numbering, instead of Linux's enum ieee80211_ac_numbers. Obtained from: dragonflybsd.git b8bd6cd746d1f45e616ccfcbeed06dfe452a1108
2017-05-12 05:50:38 +00:00
ivp->have_wme = FALSE;
[iwm] Make powersaving more similar to Linux iwlwifi behaviour. * Add a per-vap ps_disabled flag, and use it for a workaround which fixes an association issue when powersaving is enabled. * Compute flag that should correpsond to the mvmif->bss_conf.ps flag in Linux's iwlwifi (e.g. this disallows powersaving when not associated yet). Inspired-By: Linux iwlwifi Obtained from: dragonflybsd.git dc2e69bdfe8c9d7049c8a28da0adffbfbc6de5c0
2017-05-12 06:31:57 +00:00
ivp->ps_disabled = FALSE;
[iwm] Properly implement iwm_wme_update callback function. * Inspired by iwn(4) and Linux iwlwifi. * Read wme parameters into a buffer within struct iwm_vap in iwm_wme_update(). * If we aren't associated yet, the new settings will soon be sent by iwm_mvm_mac_ctxt_changed() during association. * If we are already associated, explicitly call iwm_mvm_mac_ctxt_changed() from iwm_wme_update() to send the new settings to the firmware. * Change iwm_mvm_ac_to_tx_fifo mapping, to fit the freebsd net80211 WME stream class numbering, instead of Linux's enum ieee80211_ac_numbers. Obtained from: dragonflybsd.git b8bd6cd746d1f45e616ccfcbeed06dfe452a1108
2017-05-12 05:50:38 +00:00
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
ieee80211_ratectl_init(vap);
/* Complete setup. */
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
ieee80211_vap_attach(vap, iwm_media_change, ieee80211_media_status,
mac);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
ic->ic_opmode = opmode;
return vap;
}
static void
iwm_vap_delete(struct ieee80211vap *vap)
{
struct iwm_vap *ivp = IWM_VAP(vap);
ieee80211_ratectl_deinit(vap);
ieee80211_vap_detach(vap);
free(ivp, M_80211_VAP);
}
[iwm] Cleanup mbufq draining. Add iwm_xmit_queue_drain() like in iwn(4). * iwm_xmit_queue_drain() calls ieee80211_free_node(), removing a possible memory leak, compared to using just mbufq_drain(). * Remove duplicate mbufq_drain() from iwm_mvm_rm_sta(), this should be handled in the caller. Obtained from: dragonflybsd.git 339d45fda40072e0aca5ece639173204716f11fe
2017-07-26 05:29:08 +00:00
static void
iwm_xmit_queue_drain(struct iwm_softc *sc)
{
struct mbuf *m;
struct ieee80211_node *ni;
while ((m = mbufq_dequeue(&sc->sc_snd)) != NULL) {
ni = (struct ieee80211_node *)m->m_pkthdr.rcvif;
ieee80211_free_node(ni);
m_freem(m);
}
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
static void
iwm_scan_start(struct ieee80211com *ic)
{
struct ieee80211vap *vap = TAILQ_FIRST(&ic->ic_vaps);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
struct iwm_softc *sc = ic->ic_softc;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
int error;
IWM_LOCK(sc);
[iwm] Use iwm_mvm_scan_stop_wait to properly abort scans. * Add IWM_FLAG_SCAN_RUNNING to sc->sc_flags to track whether the firmware is currently running a scan, in order to decide wheter iwm_scan_end needs to abort a running scan. * In iwm_scan_end, if the scan is still running, we now abort it, in order to keep the firmware scanning state in sync. * Try to make things a bit simpler, by reacting on the IWM_SCAN_OFFLOAD_COMPLETE and IWM_SCAN_COMPLETE_UMAC notifications, instead of IWM_SCAN_ITERATION_COMPLETE and IWM_SCAN_ITERATION_COMPLETE_UMAC. This should be fine since we always only tell the firmware to do a single scan iteration anyway. Obtained from: DragonflyBSD commit 1f249c981c4e89e7cde1836a75b61cac36dc7ac5
2017-02-08 06:50:59 +00:00
if (sc->sc_flags & IWM_FLAG_SCAN_RUNNING) {
/* This should not be possible */
device_printf(sc->sc_dev,
"%s: Previous scan not completed yet\n", __func__);
}
iwm - Reduce gratuitous differences with Linux iwlwifi in struct naming. * Rename some structs and struct members for firmware handling. Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (4b1006a6e4d0f61d48c67b46e1f791e30837db67)
2019-01-24 03:43:45 +00:00
if (fw_has_capa(&sc->sc_fw.ucode_capa, IWM_UCODE_TLV_CAPA_UMAC_SCAN))
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
error = iwm_mvm_umac_scan(sc);
else
error = iwm_mvm_lmac_scan(sc);
if (error != 0) {
[iwm] Use iwm_mvm_scan_stop_wait to properly abort scans. * Add IWM_FLAG_SCAN_RUNNING to sc->sc_flags to track whether the firmware is currently running a scan, in order to decide wheter iwm_scan_end needs to abort a running scan. * In iwm_scan_end, if the scan is still running, we now abort it, in order to keep the firmware scanning state in sync. * Try to make things a bit simpler, by reacting on the IWM_SCAN_OFFLOAD_COMPLETE and IWM_SCAN_COMPLETE_UMAC notifications, instead of IWM_SCAN_ITERATION_COMPLETE and IWM_SCAN_ITERATION_COMPLETE_UMAC. This should be fine since we always only tell the firmware to do a single scan iteration anyway. Obtained from: DragonflyBSD commit 1f249c981c4e89e7cde1836a75b61cac36dc7ac5
2017-02-08 06:50:59 +00:00
device_printf(sc->sc_dev, "could not initiate scan\n");
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_UNLOCK(sc);
ieee80211_cancel_scan(vap);
[iwm] add LED blinking for iwm hardware that supports it. Submitted by: Imre Vadasz <imre@vdsz.com> Obtained from: DragonflyBSD, Linux iwlwifi/mvm
2016-06-02 04:42:28 +00:00
} else {
[iwm] Use iwm_mvm_scan_stop_wait to properly abort scans. * Add IWM_FLAG_SCAN_RUNNING to sc->sc_flags to track whether the firmware is currently running a scan, in order to decide wheter iwm_scan_end needs to abort a running scan. * In iwm_scan_end, if the scan is still running, we now abort it, in order to keep the firmware scanning state in sync. * Try to make things a bit simpler, by reacting on the IWM_SCAN_OFFLOAD_COMPLETE and IWM_SCAN_COMPLETE_UMAC notifications, instead of IWM_SCAN_ITERATION_COMPLETE and IWM_SCAN_ITERATION_COMPLETE_UMAC. This should be fine since we always only tell the firmware to do a single scan iteration anyway. Obtained from: DragonflyBSD commit 1f249c981c4e89e7cde1836a75b61cac36dc7ac5
2017-02-08 06:50:59 +00:00
sc->sc_flags |= IWM_FLAG_SCAN_RUNNING;
[iwm] add LED blinking for iwm hardware that supports it. Submitted by: Imre Vadasz <imre@vdsz.com> Obtained from: DragonflyBSD, Linux iwlwifi/mvm
2016-06-02 04:42:28 +00:00
iwm_led_blink_start(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
IWM_UNLOCK(sc);
[iwm] add LED blinking for iwm hardware that supports it. Submitted by: Imre Vadasz <imre@vdsz.com> Obtained from: DragonflyBSD, Linux iwlwifi/mvm
2016-06-02 04:42:28 +00:00
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
static void
iwm_scan_end(struct ieee80211com *ic)
{
[iwm] add LED blinking for iwm hardware that supports it. Submitted by: Imre Vadasz <imre@vdsz.com> Obtained from: DragonflyBSD, Linux iwlwifi/mvm
2016-06-02 04:42:28 +00:00
struct ieee80211vap *vap = TAILQ_FIRST(&ic->ic_vaps);
struct iwm_softc *sc = ic->ic_softc;
IWM_LOCK(sc);
iwm_led_blink_stop(sc);
if (vap->iv_state == IEEE80211_S_RUN)
iwm_mvm_led_enable(sc);
[iwm] Use iwm_mvm_scan_stop_wait to properly abort scans. * Add IWM_FLAG_SCAN_RUNNING to sc->sc_flags to track whether the firmware is currently running a scan, in order to decide wheter iwm_scan_end needs to abort a running scan. * In iwm_scan_end, if the scan is still running, we now abort it, in order to keep the firmware scanning state in sync. * Try to make things a bit simpler, by reacting on the IWM_SCAN_OFFLOAD_COMPLETE and IWM_SCAN_COMPLETE_UMAC notifications, instead of IWM_SCAN_ITERATION_COMPLETE and IWM_SCAN_ITERATION_COMPLETE_UMAC. This should be fine since we always only tell the firmware to do a single scan iteration anyway. Obtained from: DragonflyBSD commit 1f249c981c4e89e7cde1836a75b61cac36dc7ac5
2017-02-08 06:50:59 +00:00
if (sc->sc_flags & IWM_FLAG_SCAN_RUNNING) {
/*
* Removing IWM_FLAG_SCAN_RUNNING now, is fine because
* both iwm_scan_end and iwm_scan_start run in the ic->ic_tq
* taskqueue.
*/
sc->sc_flags &= ~IWM_FLAG_SCAN_RUNNING;
iwm_mvm_scan_stop_wait(sc);
}
[iwm] add LED blinking for iwm hardware that supports it. Submitted by: Imre Vadasz <imre@vdsz.com> Obtained from: DragonflyBSD, Linux iwlwifi/mvm
2016-06-02 04:42:28 +00:00
IWM_UNLOCK(sc);
[iwm] Use iwm_mvm_scan_stop_wait to properly abort scans. * Add IWM_FLAG_SCAN_RUNNING to sc->sc_flags to track whether the firmware is currently running a scan, in order to decide wheter iwm_scan_end needs to abort a running scan. * In iwm_scan_end, if the scan is still running, we now abort it, in order to keep the firmware scanning state in sync. * Try to make things a bit simpler, by reacting on the IWM_SCAN_OFFLOAD_COMPLETE and IWM_SCAN_COMPLETE_UMAC notifications, instead of IWM_SCAN_ITERATION_COMPLETE and IWM_SCAN_ITERATION_COMPLETE_UMAC. This should be fine since we always only tell the firmware to do a single scan iteration anyway. Obtained from: DragonflyBSD commit 1f249c981c4e89e7cde1836a75b61cac36dc7ac5
2017-02-08 06:50:59 +00:00
/*
* Make sure we don't race, if sc_es_task is still enqueued here.
* This is to make sure that it won't call ieee80211_scan_done
* when we have already started the next scan.
*/
taskqueue_cancel(ic->ic_tq, &sc->sc_es_task, NULL);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
}
static void
iwm_update_mcast(struct ieee80211com *ic)
{
}
static void
iwm_set_channel(struct ieee80211com *ic)
{
}
static void
iwm_scan_curchan(struct ieee80211_scan_state *ss, unsigned long maxdwell)
{
}
static void
iwm_scan_mindwell(struct ieee80211_scan_state *ss)
{
return;
}
void
iwm_init_task(void *arg1)
{
struct iwm_softc *sc = arg1;
IWM_LOCK(sc);
while (sc->sc_flags & IWM_FLAG_BUSY)
msleep(&sc->sc_flags, &sc->sc_mtx, 0, "iwmpwr", 0);
sc->sc_flags |= IWM_FLAG_BUSY;
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
iwm_stop(sc);
if (sc->sc_ic.ic_nrunning > 0)
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm_init(sc);
sc->sc_flags &= ~IWM_FLAG_BUSY;
wakeup(&sc->sc_flags);
IWM_UNLOCK(sc);
}
static int
iwm_resume(device_t dev)
{
[iwm] implement suspend/resume through ieee80211_{suspend,resume}_all This allows wifi to associate correctly after a suspend/resume cycle. Yes, I'm using this now day to day. Tested: * Intel 7260AC, STA mode
2016-04-26 04:40:59 +00:00
struct iwm_softc *sc = device_get_softc(dev);
int do_reinit = 0;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Remove OpenBSD-specific comment. Beautify pci cfg space accesses. Obtained from: dragonflybsd.git 710fdd011c30a1bd9f53b87c843fb8907c4a6cfd
2017-02-22 04:27:07 +00:00
/*
* We disable the RETRY_TIMEOUT register (0x41) to keep
* PCI Tx retries from interfering with C3 CPU state.
*/
pci_write_config(dev, PCI_CFG_RETRY_TIMEOUT, 0x00, 1);
if_iwm - Check sc->sc_attached flag in suspend/resume callbacks. * There is (almost) nothing to do in suspend/resume if if_iwm has failed during initialization (e.g. because of firmware load failure) and was already uninitialized by iwm_detach_local(). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (67b5e090efb225654815fed91020db6cfc16bb19)
2019-01-24 03:42:59 +00:00
if (!sc->sc_attached)
return 0;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
iwm_init_task(device_get_softc(dev));
[iwm] implement suspend/resume through ieee80211_{suspend,resume}_all This allows wifi to associate correctly after a suspend/resume cycle. Yes, I'm using this now day to day. Tested: * Intel 7260AC, STA mode
2016-04-26 04:40:59 +00:00
IWM_LOCK(sc);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
if (sc->sc_flags & IWM_FLAG_SCANNING) {
sc->sc_flags &= ~IWM_FLAG_SCANNING;
[iwm] implement suspend/resume through ieee80211_{suspend,resume}_all This allows wifi to associate correctly after a suspend/resume cycle. Yes, I'm using this now day to day. Tested: * Intel 7260AC, STA mode
2016-04-26 04:40:59 +00:00
do_reinit = 1;
}
IWM_UNLOCK(sc);
if (do_reinit)
ieee80211_resume_all(&sc->sc_ic);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
return 0;
}
static int
iwm_suspend(device_t dev)
{
[iwm] implement suspend/resume through ieee80211_{suspend,resume}_all This allows wifi to associate correctly after a suspend/resume cycle. Yes, I'm using this now day to day. Tested: * Intel 7260AC, STA mode
2016-04-26 04:40:59 +00:00
int do_stop = 0;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
struct iwm_softc *sc = device_get_softc(dev);
[iwm] implement suspend/resume through ieee80211_{suspend,resume}_all This allows wifi to associate correctly after a suspend/resume cycle. Yes, I'm using this now day to day. Tested: * Intel 7260AC, STA mode
2016-04-26 04:40:59 +00:00
do_stop = !! (sc->sc_ic.ic_nrunning > 0);
if_iwm - Check sc->sc_attached flag in suspend/resume callbacks. * There is (almost) nothing to do in suspend/resume if if_iwm has failed during initialization (e.g. because of firmware load failure) and was already uninitialized by iwm_detach_local(). Submitted by: Augustin Cavalier <waddlesplash@gmail.com> (Haiku) Obtained from: DragonFlyBSD (67b5e090efb225654815fed91020db6cfc16bb19)
2019-01-24 03:42:59 +00:00
if (!sc->sc_attached)
return (0);
[iwm] implement suspend/resume through ieee80211_{suspend,resume}_all This allows wifi to associate correctly after a suspend/resume cycle. Yes, I'm using this now day to day. Tested: * Intel 7260AC, STA mode
2016-04-26 04:40:59 +00:00
ieee80211_suspend_all(&sc->sc_ic);
if (do_stop) {
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
IWM_LOCK(sc);
iwm_stop(sc);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
sc->sc_flags |= IWM_FLAG_SCANNING;
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
IWM_UNLOCK(sc);
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
return (0);
}
static int
iwm_detach_local(struct iwm_softc *sc, int do_net80211)
{
iwm: revert r293178 This optimization is not proper (and causes kernel panic), since driver checks fw_status to optimize away parsing stage if it was already done. Reported by: dchagin
2016-01-05 20:09:26 +00:00
struct iwm_fw_info *fw = &sc->sc_fw;
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
device_t dev = sc->sc_dev;
int i;
[iwm] make sure we call iwm_detach_local() only once. Obtained from: DragonflyBSD git ebd4ceab76a6f161362029cbfd08efaedaab0519
2017-02-06 02:20:05 +00:00
if (!sc->sc_attached)
return 0;
sc->sc_attached = 0;
iwm: If firmware load fails during init via iwm_preinit() we bail out before calling ieee80211_ifattach() so the taskqueue hasn't been initialized. Don't try to drain it, we'll panic. Looks like this issue was introduced in r303326. Reviewed by: avos, sbruno, adrian Differential Revision: https://reviews.freebsd.org/D8499
2016-11-12 18:33:56 +00:00
if (do_net80211)
ieee80211_draintask(&sc->sc_ic, &sc->sc_es_task);
iwm(4): switch to ieee80211_runtask() Submitted by: Andiry Voskoboinyk (s3erios@gmail.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D5054
2016-07-25 23:44:44 +00:00
[iwm] add LED blinking for iwm hardware that supports it. Submitted by: Imre Vadasz <imre@vdsz.com> Obtained from: DragonflyBSD, Linux iwlwifi/mvm
2016-06-02 04:42:28 +00:00
callout_drain(&sc->sc_led_blink_to);
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
callout_drain(&sc->sc_watchdog_to);
iwm_stop_device(sc);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
if (do_net80211) {
[iwm] Cleanup mbufq draining. Add iwm_xmit_queue_drain() like in iwn(4). * iwm_xmit_queue_drain() calls ieee80211_free_node(), removing a possible memory leak, compared to using just mbufq_drain(). * Remove duplicate mbufq_drain() from iwm_mvm_rm_sta(), this should be handled in the caller. Obtained from: dragonflybsd.git 339d45fda40072e0aca5ece639173204716f11fe
2017-07-26 05:29:08 +00:00
IWM_LOCK(sc);
iwm_xmit_queue_drain(sc);
IWM_UNLOCK(sc);
Replay r286410. Change KPI of how device drivers that provide wireless connectivity interact with the net80211 stack. Historical background: originally wireless devices created an interface, just like Ethernet devices do. Name of an interface matched the name of the driver that created. Later, wlan(4) layer was introduced, and the wlanX interfaces become the actual interface, leaving original ones as "a parent interface" of wlanX. Kernelwise, the KPI between net80211 layer and a driver became a mix of methods that pass a pointer to struct ifnet as identifier and methods that pass pointer to struct ieee80211com. From user point of view, the parent interface just hangs on in the ifconfig list, and user can't do anything useful with it. Now, the struct ifnet goes away. The struct ieee80211com is the only KPI between a device driver and net80211. Details: - The struct ieee80211com is embedded into drivers softc. - Packets are sent via new ic_transmit method, which is very much like the previous if_transmit. - Bringing parent up/down is done via new ic_parent method, which notifies driver about any changes: number of wlan(4) interfaces, number of them in promisc or allmulti state. - Device specific ioctls (if any) are received on new ic_ioctl method. - Packets/errors accounting are done by the stack. In certain cases, when driver experiences errors and can not attribute them to any specific interface, driver updates ic_oerrors or ic_ierrors counters. Details on interface configuration with new world order: - A sequence of commands needed to bring up wireless DOESN"T change. - /etc/rc.conf parameters DON'T change. - List of devices that can be used to create wlan(4) interfaces is now provided by net.wlan.devices sysctl. Most drivers in this change were converted by me, except of wpi(4), that was done by Andriy Voskoboinyk. Big thanks to Kevin Lo for testing changes to at least 8 drivers. Thanks to pluknet@, Oliver Hartmann, Olivier Cochard, gjb@, mmoll@, op@ and lev@, who also participated in testing. Reviewed by: adrian Sponsored by: Netflix Sponsored by: Nginx, Inc.
2015-08-27 08:56:39 +00:00
ieee80211_ifdetach(&sc->sc_ic);
iwm(4) synchronize driver to DragonFlyBSD version and recent f/w update. Submitted by: Kevin Bowling (kevin.bowling@kev009.com) MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6967
2016-07-26 00:02:17 +00:00
}
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Sync if_iwm_phy_db code with Linux iwlwifi. Obtained from: Dragonflybsd commit c1019b6bfff36c856f7b4fccbdf3bb13ac27750c
2017-02-06 02:14:34 +00:00
iwm_phy_db_free(sc->sc_phy_db);
sc->sc_phy_db = NULL;
[iwm] Add and use iwm_phy_db_free(), to plug phy_db memory leak. (Together with other iwm(4) memory leak fixes) Memory leakage in M_DEVBUF is now at ca. 2KB for each iwm(4) module load/unload cycle. Submitted by: Imre Vadasz <imre@vdsz.com> Approved by: re (gjb) Obtained from: DragonflyBSD git eaf551a1d464c643e98ce5781971dd32124e9af1 Differential Revision: https://reviews.freebsd.org/D6819
2016-06-23 01:11:52 +00:00
[iwm] Sync nvm parsing code with Linux iwlwifi. * sc->sc_nvm becomes sc->nvm_data and is now a pointer instead of an inlined struct. * Add sc->eeprom_size and sc->nvm_hw_section_num configuration values to struct iwm_softc. * For now continue to avoid negative error return-values, and use pointer variables for some return values, as before. * Continue to omit LAR (location aware regulatory) related code as well. Tested: * Intel 7260, STA mode (2GHz) Obtained from: dragonflybsd commit 39f8331b1a6f295291e08c377da12a8e7a5436c0
2017-02-06 03:06:11 +00:00
iwm_free_nvm_data(sc->nvm_data);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Free descriptor rings */
[iwm] free RX ring / NVM memory after they're used. * Free RX ring during detach * Free NVM memory after parsing Tested: * 7260, STA mode Submitted by: Imre Vadasz <imre@vdsz.com> Approved by: re (gjb) Obtained from: dragonflybsd Differential Revision: https://reviews.freebsd.org/D6817
2016-06-16 17:59:15 +00:00
iwm_free_rx_ring(sc, &sc->rxq);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
for (i = 0; i < nitems(sc->txq); i++)
iwm_free_tx_ring(sc, &sc->txq[i]);
iwm: revert r293178 This optimization is not proper (and causes kernel panic), since driver checks fw_status to optimize away parsing stage if it was already done. Reported by: dchagin
2016-01-05 20:09:26 +00:00
/* Free firmware */
if (fw->fw_fp != NULL)
iwm_fw_info_free(fw);
iwm: free firmware related resources after uploading it to the hardware iwn(4) / wpi(4) works in the same way (read_firmware() -> hw_init() -> firmware_put()) Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D4766
2016-01-04 21:07:08 +00:00
/* Free scheduler */
[iwm] Remove wrappers around iwm_dma_contig_free() calls. No functional changes. Inspired by: OpenBSD Approved by: adrian (mentor) Differential Revision: https://reviews.freebsd.org/D7933
2016-09-19 22:16:03 +00:00
iwm_dma_contig_free(&sc->sched_dma);
iwm_dma_contig_free(&sc->ict_dma);
iwm_dma_contig_free(&sc->kw_dma);
iwm_dma_contig_free(&sc->fw_dma);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
[iwm] Add support for Firmware paging, needed for newer 8000C firmware. * Uses the IWM_FW_PAGING_BLOCK_CMD firmware command to tell the firmware what memory ranges to use for paging. Obtained from: dragonflybsd.git 8a5b199964f8e7bdb00039f0b48817a01b402f18
2017-02-24 07:07:58 +00:00
iwm_free_fw_paging(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
/* Finished with the hardware - detach things */
iwm_pci_detach(dev);
[iwm] Add implementation of the notification wait api from iwlwifi. Obtained from: Linux iwlwifi Obtained from: DragonflyBSD commit 94dc1dadceb57b688036211262d678bc6bbdde37
2017-02-06 05:27:07 +00:00
if (sc->sc_notif_wait != NULL) {
iwm_notification_wait_free(sc->sc_notif_wait);
sc->sc_notif_wait = NULL;
}
iwm: reorganize if_iwmvar.h - Change order of data in if_iwmvar.h (like it is in other drivers: defines, data structures, vap/node structures, softc struct and locks); use indentation. - Fix IWM_LOCK(_sc) / IWM_UNLOCK(_sc) macro. - Add IWM_LOCK_INIT / DESTROY(sc) + fix mtx_init() usage. - Wrap iwm_node casts into IWM_NODE() macro. - Drop some fields: * wt_hwqueue from Tx radiotap header; * macaddr[6] from iwm_vap; Approved by: adrian Differential Revision: https://reviews.freebsd.org/D4753
2016-01-03 10:06:10 +00:00
IWM_LOCK_DESTROY(sc);
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
return (0);
}
static int
iwm_detach(device_t dev)
{
struct iwm_softc *sc = device_get_softc(dev);
return (iwm_detach_local(sc, 1));
}
static device_method_t iwm_pci_methods[] = {
/* Device interface */
DEVMETHOD(device_probe, iwm_probe),
DEVMETHOD(device_attach, iwm_attach),
DEVMETHOD(device_detach, iwm_detach),
DEVMETHOD(device_suspend, iwm_suspend),
DEVMETHOD(device_resume, iwm_resume),
DEVMETHOD_END
};
static driver_t iwm_pci_driver = {
"iwm",
iwm_pci_methods,
sizeof (struct iwm_softc)
};
static devclass_t iwm_devclass;
DRIVER_MODULE(iwm, pci, iwm_pci_driver, iwm_devclass, NULL, NULL);
Fix the MODULE_PNP_INFO() for iwm(4) where I got the bus and module arguments wrong in r339020. PR: 231625 Reported by: Yuri Pankov (yuripv yuripv.net) Reviewed by: cem, Yuri Pankov (yuripv yuripv.net) Approved by: re (kib) Pointyhat to: bz (a rather big one for this one)
2018-10-01 10:44:33 +00:00
MODULE_PNP_INFO("U16:device;P:#;T:vendor=0x8086", pci, iwm_pci_driver,
Provide MODULE_PNP_INFO() for iwm(4) so that devmatch(8) can do its job. PR: 231625 Submitted by: Yuri Pankov (yuripv yuripv.net) Approved by: re (kib)
2018-09-29 21:14:54 +00:00
iwm_devices, nitems(iwm_devices));
Import OpenBSD's iwm WiFi driver for Intel 3160/7260/7265. There are still several bugs, but I've been using it for a while now. Thanks to all the testers and to Adrian for his help with this driver. This driver isn't connected to the build yet, but it will be soon. There's no MFC planned because the driver isn't very stable yet. Reviewed by: adrian Obtained from: https://github.com/rpaulo/iwm Tested by: adrian, gjb, dumbbell (others that I forgot). Relnotes: yes
2015-08-08 06:06:48 +00:00
MODULE_DEPEND(iwm, firmware, 1, 1, 1);
MODULE_DEPEND(iwm, pci, 1, 1, 1);
MODULE_DEPEND(iwm, wlan, 1, 1, 1);
Reference in New Issue Copy Permalink