As USB_INT_ID_RETRY_FAILED can override USB_INT_ID_REGS, beacon interrupt
(CR_INTERRUPT) might be lost. Problem is that when device trigger CR_INTERRUPT
it disables HW interrupt. Now if USB_INT_ID_REGS with CR_INTERRUPT gets lost,
beacon interrupt stays disabled until beacon watchdog notices the stall. This
happen very often on heavy TX. Improve watchdog to trigger earlier, after three
missing beacon interrupts.
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
zd_mac_config_beacon() has only limited time to set up beacon when called from
beacon interrupt handler/worker. So do not let it retry acquiring beacon fifo
semaphore in interrupt handler. Beacon fifo semaphore should not be locked by
firmware anyway at this time, it's only locked when device is using/txing
beacon.
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Device losses read-reg interrupts. By looking at usbmon it appears that
USB_INT_ID_RETRY_FAILED can override USB_INT_ID_REGS. This causes read
command to timeout, usually under heavy TX.
Fix by retrying read registers again if USB_INT_ID_RETRY_FAILED is received
while waiting for USB_INT_ID_REGS.
However USB_INT_ID_REGS is not always lost but is received after
USB_INT_ID_RETRY_FAILED and is usually received by the retried read
command. USB_INT_ID_REGS of the retry is then left unhandled and might
be received by next read command. Handle this by ignoring previous
USB_INT_ID_REGS that doesn't match current read command request.
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Update HW beacon only when needed. This appears to make device work in AP-mode
(dtim_period=1) somewhat more stable.
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This message is should be debug-only as it tells almost nothing useful. It also
happens very often and just floods logs.
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Driver uses IEEE80211_HW_SIGNAL_UNSPEC and so signal values reported to
mac80211 should be in range 0..100. Sometimes device return out of range
values. These out of range values can then trigger warning in
cfg80211_inform_bss_frame.
This patch adds checks to enforce range returned from driver to
mac80211 be in 0..100 range.
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
After uploading radio values calibration goes in. In MMIO dump it is:
radio_read(0x002b) -> 0x0008
radio_write(0x002b) <- 0x0008
radio_read(0x002e) -> 0x0004
radio_write(0x002e) <- 0x0000
radio_read(0x002e) -> 0x0000
radio_write(0x002e) <- 0x0004
radio_read(0x002b) -> 0x0008
radio_write(0x002b) <- 0x0009
To find masks and sets, MMIO hacks were used to fool closed driver.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
When module is reloaded, device may fail to work, with messages:
[ 342.211926] phy40 -> rt2x00lib_rxdone_read_signal: Warning - Frame received with unrecognized signal, mode=0x0000, signal=0x0059, type=2.
[ 342.314254] phy40 -> rt2x00lib_rxdone_read_signal: Warning - Frame received with unrecognized signal, mode=0x0000, signal=0x004a, type=2.
[ 342.416458] phy40 -> rt2x00lib_rxdone: Warning - Wrong frame size 3183 max 2432.
[ 342.518605] phy40 -> rt2x00lib_rxdone_read_signal: Warning - Frame received with unrecognized signal, mode=0x0000, signal=0x00c9, type=2.
[ 342.620836] phy40 -> rt2x00lib_rxdone_read_signal: Warning - Frame received with unrecognized signal, mode=0x0000, signal=0x00ae, type=1.
[ 342.723201] phy40 -> rt2x00lib_rxdone: Warning - Wrong frame size 0 max 2432.
[ 342.825399] phy40 -> rt2x00lib_rxdone: Warning - Wrong frame size 0 max 2432.
[ 342.927624] phy40 -> rt2x00lib_rxdone: Warning - Wrong frame size 0 max 2432.
[ 343.029804] phy40 -> rt2x00lib_rxdone: Warning - Wrong frame size 2491 max 2432.
[ 343.132008] phy40 -> rt2x00lib_rxdone: Warning - Wrong frame size 2576 max 2432.
[ 343.234326] phy40 -> rt2x00lib_rxdone_read_signal: Warning - Frame received with unrecognized signal, mode=0x0000, signal=0x004c, type=1.
[ 343.438723] phy40 -> rt2x00lib_rxdone_read_signal: Warning - Frame received with unrecognized signal, mode=0x0000, signal=0x00e6, type=1.
Whereas replugging device make it functional. To solve that problem
force reset device during probe.
With patch messages are gone. Unfortunately device may sometimes
still does not operate correctly after module reload (fail to receive
data after associate), but such cases are rarer than without the patch.
Signed-off-by: Stanislaw Gruszka <stf_xl@wp.pl>
Acked-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Sometimes rxdesc descriptor provided by hardware contains invalid
(random) data. For example rxdesc.size can be bigger than actual
size of the buffer. When this happen rt2x00crypto_rx_insert_iv()
corrupt memory doing memmove outside of buffer boundaries.
Signed-off-by: Stanislaw Gruszka <stf_xl@wp.pl>
Acked-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Switching channel happens after specific SHM write to B43_SHM_SH_CHAN.
This is the way we found it in BCM4331 MMIO dumps. By comparing with
N-PHY code we noticed there is routing used for SYN and TX/RX.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
The trick was to find 0x810 PHY reg ops close to analog enabling code.
To find out proper masks and sets, MMIO hacks were used.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Closed drivers kill radio right after reading radio version and MACCTL,
so it was easy to find related PHY ops:
phy_read(0x0810) -> 0x0000
phy_write(0x0810) <- 0x0000
To find out the mask of above OP, MMIO hack was used to fake read val:
phy_read(0x0810) -> 0xffff
phy_write(0x0810) <- 0x0000
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
These are a potential source of confusion, as most C code treats all
non-zero values as true.
Signed-off-by: Mike McCormack <mikem@ring3k.org>
Acked-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
rtl_ps_set_rf_state's protect_or_not parameter is not set to
true anywhere, except for commented out code.
It enables some legacy locking code, which is no longer used,
so delete the parameter and the old locking code.
Signed-off-by: Mike McCormack <mikem@ring3k.org>
Acked-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
offchan_deley should be offchan_delay
Signed-off-by: Mike McCormack <mikem@ring3k.org>
Acked-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Nothing else is freeing it, and this mirrors the error handling path
already in ath5k_init_softc.
https://bugzilla.kernel.org/show_bug.cgi?id=37592
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Reported-by: Tsyvarev Andrey <tsyvarev@ispras.ru>
checking the status of PAPRD_AGC2_POWER(Log(ADC_power) measured after
last gain-change in dB) field suggests whether the PAPRD is completely/properly
done. This is an additional check apart from polling for PAPRD done bit being set.
Susinder suggests that the ideal power range value should be
0xf0 to 0xfe. With AR9382 we do have the values in this range. to have a
common check for all platforms we take agc2_power should be atleast greater
than 0xe0
Cc: susinder@qca.qualcomm.com
Cc: senthilb@qca.qualcomm.com
Cc: kmuthusa@qca.qualcomm.com
Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qca.qualcomm.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
when the AP is configured with HT40/2.4GHz and when PAPRD is enabled in station
we have PAPRD training frames never sent out because of the following
failure "PAPRD target power delta out of range". This consistently happens
in channel 10,11 in Over The Air testing and rarely in channel 6 under wired
setup.
the reason behind this issue is all the HT40 rates target power
are 0 when we operate in channel 10/11 at 2.4GHz
Signed-off-by: Mohammed Shafi Shajakhan <mshajakhan@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
core.h references ieee80211_ops but doesn't include mac80211.h, which issues
a warning once mac80211-compat.h is taken into use.
Acked-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: Kalle Valo <kvalo@adurom.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
zd_chip.h contains one function with a pointer to ieee80211_hw but doesn't
include mac80211.h. As a temporary solution include mac80211.h to avoid
a warning.
Signed-off-by: Kalle Valo <kvalo@adurom.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
The hacks break once mac80211-compat.h is taken into use. Fix it by adding
an ifdef check.
Signed-off-by: Kalle Valo <kvalo@adurom.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
The hacks break once mac80211-compat.h is taken into use. Fix it by adding
an ifdef check.
Signed-off-by: Kalle Valo <kvalo@adurom.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Add power_level module parameter to set the default power save level.
Power save level has range from 1 - 5, default power save level is 1.
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Add power_save module parameter to enable power management if needed
Default power management is disabled.
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Since the irq number is just an unsigned int, store it inside iwl_bus
instead of calling the get_irq ops every time it is needed.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Since there is no protection around SYNC host command mechanism, at least WARN
when collision happens between two SYNC host comamnds. I am not sure there is a
real issue (beyond the HCMD_ACTIVE flag maintenance) with having two SYNC host
commands at the same time, but at least now, we will know about it.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
All pci related stuff is in iwl-pci.c.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
In order to remove a few more dereference to priv->pdev that will be killed
[Asoon, there is now a method to get the IRQ number.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
The PCI_REVISION_ID is read and printed in iwl_pci_probe anyway using pr_info
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Continue to popule the PCI layer and the iwl_bus_ops with the power related
stuff.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Get a pointer to the struct device during probe and get the rid of all the PCI
specific DMA wrappers.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Since we have now a PCI layer, all the init and deinit code that is PCI
related should move to there.
Also move the IO functions: read8/read32/write32. They need hw_base which
is killed from priv.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Bus specific layer must know how to return the struct device* of the device.
Implement that as a callback of iwl_bus_ops and use that callback instead of
using the priv->pdev pointer which is meant to disappear soon.
Since the struct device * is needed in hot path, iwl_bus holds a pointer to it
instead of calling get_dev all the time.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
iwl_bus will represent a bus, and iwl_bus_ops all the operations that can be
done on this bus.
For the moment only set_prv_data is implemented. More to come...
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Move some PCI functionality to the new iwl_pci.[ch] files:
* the PCI_DEVICE_TABLE
* the pci_driver struct definition
* the PCI probe / remove functions
* the PCI suspend / resume functions
All these functions are now split: the trigger comes from the PCI layer which
calls to the bus generic code located in the other files.
This is the beginning only. There are still a lot of PCI related code needs
to be gathered.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
It is superfluous to disable the interrupts after we reset the NIC. The only
entity that could enable the interrupts after the NIC is reset is the driver.
So remove this pointless action.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
Jussi Kivilinna