In ena_com_mem_addr_set(), use the above functions to split dma address
to the lower 32 bits and the higher 16 bits.
Signed-off-by: Netanel Belgazal <netanel@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Current driver tries to allocate msix vectors as the number of the
negotiated io queues. (with another msix vector for management).
If pci_alloc_irq_vectors() fails, the driver aborts the probe
and the ENA network device is never brought up.
With this patch, the driver's logic will reduce the number of IO
queues to the number of allocated msix vectors (minus one for management)
instead of failing probe().
Signed-off-by: Netanel Belgazal <netanel@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
ENA driver post Rx buffers through the Rx submission queue
for the ENA device to fill them with receive packets.
Each Rx buffer is marked with req_id in the Rx descriptor.
Newer ENA devices could consume the posted Rx buffer in out of order,
and as result the corresponding Rx completion queue will have Rx
completion descriptors with non contiguous req_id(s)
In this change the driver holds two rings.
The first ring (called free_rx_ids) is a mapping ring.
It holds all the unused request ids.
The values in this ring are from 0 to ring_size -1.
When the driver wants to allocate a new Rx buffer it uses the head of
free_rx_ids and uses it's value as the index for rx_buffer_info ring.
The req_id is also written to the Rx descriptor
Upon Rx completion,
The driver took the req_id from the completion descriptor and uses it
as index in rx_buffer_info.
The req_id is then return to the free_rx_ids ring.
This patch also adds statistics to inform when the driver receive out
of range or unused req_id.
Note:
free_rx_ids is only accessible from the napi handler, so no locking is
required
Signed-off-by: Netanel Belgazal <netanel@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For each device reset, log to the device what is the cause
the reset occur.
Signed-off-by: Netanel Belgazal <netanel@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Instead of using:
memset(ptr, 0x0, sizeof(struct ...))
use:
memset(ptr, 0x0, sizeor(*ptr))
Signed-off-by: Netanel Belgazal <netanel@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With this patch, ENA device can update the ena driver about
the desired timeout values:
These values are part of the "hardware hints" which are transmitted
to the driver as Asynchronous event through ENA async
event notification queue.
In case the ENA device does not support this capability,
the driver will use its own default values.
Signed-off-by: Netanel Belgazal <netanel@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The functions dwmac4_dma_init_rx_chan, dwmac4_dma_init_tx_chan and
dwmac4_dma_init_channel do not need to be in global scope, so them
static.
Cleans up sparse warnings:
"symbol 'dwmac4_dma_init_rx_chan' was not declared. Should it be static?"
"symbol 'dwmac4_dma_init_tx_chan' was not declared. Should it be static?"
"symbol 'dwmac4_dma_init_channel' was not declared. Should it be static?"
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Respect the XDP_FLAGS_HW_MODE. When it's set install the program
on the NIC and skip enabling XDP in the driver.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The xdp_prog member of the adapter's data path structure is used
for XDP in driver mode. In case a XDP program is loaded with in
HW-only mode, we need to store it somewhere else. Add a new XDP
prog pointer in the main structure and use that when we need to
know whether any XDP program is loaded, not only a driver mode
one. Only release our reference on adapter free instead of
immediately after netdev unregister to allow offload to be disabled
first.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
DRV_MODE means that user space wants the program to be run in
the driver. Do not try to offload. Only offload if no mode
flags have been specified.
Remember what the mode is when the program is installed and refuse
new setup requests if there is already a program loaded in a
different mode. This should leave it open for us to implement
simultaneous loading of two programs - one in the drv path and
another to the NIC later.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In preparation of XDP offload flags move the driver setup into
a function. Otherwise the number of conditions in one function
would make it slightly hard to follow. The offload handler may
now be called with NULL prog, even if no offload is currently
active, but that's fine, offload code can handle that.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When trying to TX through a monitor interface, the
conditions in iwl_mvm_tx_skb_non_sta() don't match
and the frame tries to go out from an usued TXQ.
Add a check for monitor iface, and use the AUX queue
in such a case.
In non-DQA mode the frame is sent through the
static-allocated queues anyway, so the problem is
in DQA mode only.
Signed-off-by: Liad Kaufman <liad.kaufman@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
In some platforms, having the device enabled with certain radio
frontends causes the platform to not be able to resume properly
from suspend, regardless of the wakeup cause. This was traced to
a hardware issue with the integrated 9000-series A-step variant.
Set the right hardware bit to disable the problematic state.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
When the firmware crashes, the transmit queues can't make
any progress. This is why we stop the counter that monitor
the transmit queues' activity.
The call that notifies the error to the op_mode may take
a bit of time, so stop the timer of the transmit queues
earlier.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
The driver prints "L1 Enabled - LTR Enabled" all the time as dev_info,
which is just useless noise in most cases. Convert this to
IWL_DEBUG_POWER() so we don't pollute the log unnecessarily but still
can get this info on demand.
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
It's sometimes hard to find out which HW address the iwlwifi device is
using, for instance when reading crouded sniffer logs. To make it
easier, print out an info level message with the HW address as soon as
we know it.
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
This new API allows flushing queues based on station ID and TID in A000
devices. One reason for using this is that tfd_queue_mask is only good
for 32 queues, which is not enough for A000 devices.
Signed-off-by: Sara Sharon <sara.sharon@intel.com>
Signed-off-by: Mordechai Goodstein <mordechay.goodstein@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
When we get an ASSERT, the fw_dump_desc pointer points to
iwl_mvm_dump_desc_assert which can't be freed since it is
a global. We still need to NULL'ify the pointer when we
call iwl_mvm_free_fw_dump_desc otherwise we will hit
int iwl_mvm_fw_dbg_collect_desc(struct iwl_mvm *mvm,
const struct iwl_mvm_dump_desc *desc,
const struct iwl_fw_dbg_trigger_tlv *trigger)
{
<snip>
if (WARN_ON(mvm->fw_dump_desc))
iwl_mvm_free_fw_dump_desc(mvm);
Fixes: b6eaa45aa1 ("iwlwifi: mvm: add the cause of the firmware dump in the dump")
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
When updating the mac context after association,
assoc_beacon_arrive_time is not being set, which causes the FW to
set a wrong TSF to the MAC.
Fix this by setting the assoc_beacon_arrive_time when updating the
mac context after association.
Signed-off-by: Avraham Stern <avraham.stern@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Nothing ever checks the return value of iwl_pcie_apm_stop_master(),
so there's no point in it having one - make it return void.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
In case we need to move the scheduler write pointer by
steps of 0x40, 0x80 or 0xc0, the scheduler gets stuck.
This leads to hardware error interrupts with status:
0x5A5A5A5A or alike.
In order to work around this, detect in the transport
layer that we are going to hit this case and tell iwlmvm
to increment the sequence number of the packets. This
allows to keep the requirement that the WiFi sequence
number is in sync with the index in the scheduler Tx queue
and it also allows to avoid the problematic sequence.
This means that from time to time, we will start a queue
from ssn + 1, but that shouldn't be a problem since we
don't switch to new queues for AMPDU now that we have
DQA which allows to keep the same queue while toggling
the AMPDU state.
This bug has been fixed on 9000 devices and up.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
A TID may not have traffic but still have a BA agreement
active (or being setup / torn down) since a BA agreement
can be triggered by a debugfs hook.
Just avoid to consider such a TID as inactive to make the
logic safer.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
If the hardware is stuck, we can't read any of the memory we need to
dump it, so we end up printing only 0xa5a5a5a5, which is useless.
To solve this, poke the hardware by triggering a reset and re-enabling
the clocks if we detect a HW error.
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
When we started using threaded irqs, all the opmode calls were changed
to be called with local_bh disabled. The reason for this was it was
that mac80211 needs that. When we are handling FW errors, mac80211 is
not involved, so we don't need it.
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
In BSS mode in the disconnection flow, mac80211 removes
the AP station before the vif is set to unassociated.
Our firmware wants it the other way around: first set
the vif as unassociated, and then remove the AP station.
In order to bridge between those two different behaviors,
iwlmvm doesn't remove the station from the firmware when
mac80211 removes it, but only after the vif is set to
unassociated. The implementation is in
iwl_mvm_bss_info_changed_station:
if (assoc state was modified && mvmvif->ap_sta_id is VALID
&& assoc state is now UNASSC)
remove_the_station_from_the_firmware()
During the recovery flow, mac80211 re-adds the AP station
and then reconfigures the vif. Since the vif is not
associated, and then, we enter the if above (which was
intended to be taken in the disconnection flow only) and
remove the station we just added. This defeats the
recovery flow.
Fix this by not removing the AP station in this flow if
we are in recovery flow.
Cc: stable@vger.kernel.org
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
We allocate nvm_data in iwl_mvm_nvm_get_from_fw(). If something goes
wrong after the allocation (i.e. if no valid MAC address is valid), we
should free nvm_data before returning an error.
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
This function just calls kfree(), so it only obscures the code without
bringing any benefits. Remove it.
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Properly document the transmit buffer bits using an enum and
kernel-doc documentation.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Add kernel-doc documentation for enum iwl_debug_cmds, linking the
structures used by the commands.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
To stop and start the FW monitor in the 7000 HW
family we need to use a different bit, otherwise
after stopping it for the first time - it won't
get restarted. Use the correct bitmask.
Note: This fix is only for DRAM collection mode.
For other modes, an additional fix will be
needed.
Signed-off-by: Liad Kaufman <liad.kaufman@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Fix the enum link by adding the missing & and provide the link
to the TX response documentation.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
When the command name is printed on command completion, the wrong
group is used, leading to the wrong name being printed. Fix this
by using the group ID without inappropriately mangling it through
iwl_cmd_groupid() - it's already a u8. Also, while at it, use it
from the same place as the command ID, everything else is just
confusing.
Fixes: ab02165cce ("iwlwifi: add wide firmware command infrastructure for TX")
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
When sending non-linear SKBs that should be included in the regular
TX tracing completely (and not be pushed into the tx_data tracing),
the (tracing) code didn't correctly take the fact that they were
non-linear into account and added only the skb head portion.
This probably never really triggered, since those frames we want
traced fully are most likely linear anyway, but the code gets easier
to understand and we lose an argument to the tracing function, so
overall fixing this is better.
Fixes: 206eea7833 ("iwlwifi: pcie: support frag SKBs")
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
There's no need to calculate the data_len outside of the tracepoint,
since it's always skb->len - hdr_len, which are both available inside.
Simplify the callers and move the calculation in.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Show the name of the member (scanned_channels) that provides the
length with some better markup.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
On A000 HW, the SCD rdptr has only 8 bits allocated
for it, thus when checking if a queue is full, or
when checking if the SSN is equal to the TID's
next_reclaimed, A000 HW should trim the SSN.
Fix this by "normalizing" the SSN to wrap around
0xFF when comparing to the next_reclaimed on A000
HW.
Signed-off-by: Liad Kaufman <liad.kaufman@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
We don't actually care about the value at all, just making sure
that we can successfully parse a single integer value, but that's
entirely pointless - remove it.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
When toggling the RF-kill pin quickly in succession, the driver can
get rather confused because it might be in the process of shutting
down, expecting all commands to go through quickly due to rfkill,
but the transport already thinks the device is accessible again,
even though it previously shut it down. This leads to bugs, and I
even observed a kernel panic.
Avoid this by making the PCIe code only report that the radio is
enabled again after the higher layers actually decided to shut it
off.
This also pulls out this common RF-kill checking code into a common
function called by both transport generations and also moves it to
the direct method - in the internal helper we don't really care
about the RF-kill status anymore since we won't report it up until
the stop anyway.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
If we happen to be in or get into the queue sync when RF-kill
is asserted, we return from there and warn since there are
still queue sync notifications outstanding. These can't ever
come though, because we're in RF-kill, so don't WARN then.
While at it, also move the warning to the appropriate place,
if the request is not synchronous then we shouldn't warn, but
currently always will.
To make it fast, also trigger the waitq when on rfkill assert.
Fixes: 0636b93821 ("iwlwifi: mvm: implement driver RX queues sync command")
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
In order to debug "hardware" RF-kill flows, add a low-level hook to
allow changing the "hardware" RF-kill from debugfs.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
There's no point in duplicating exactly the same code here
for legacy and MSI-X interrupts, so pull it out into a new
function to call in both places.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
