linux-stable.git/drivers/net/wireless/intel, branch v5.4.7

iwlwifi: pcie: move power gating workaround earlier in the flow

2019-12-31T15:46:35+00:00

commit 0df36b90c47d93295b7e393da2d961b2f3b6cde4 upstream.

We need to reset the NIC after setting the bits to enable power
gating and that cannot be done too late in the flow otherwise it
cleans other registers and things that were already configured,
causing initialization to fail.

In order to fix this, move the function to the common code in trans.c
so it can be called directly from there at an earlier point, just
after the reset we already do during initialization.

Fixes: 9a47cb988338 ("iwlwifi: pcie: add workaround for power gating in integrated 22000")
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=205719
Cc: stable@ver.kernel.org # 5.4+
Reported-by: Anders Kaseorg 
Signed-off-by: Luca Coelho 
Signed-off-by: Kalle Valo 
Signed-off-by: Greg Kroah-Hartman

net: wireless: intel: iwlwifi: fix GRO_NORMAL packet stalling

2019-12-31T15:45:49+00:00

[ Upstream commit b167191e2a851cb2e4c6ef8b91c83ff73ef41872 ]

Commit 6570bc79c0df ("net: core: use listified Rx for GRO_NORMAL in
napi_gro_receive()") has applied batched GRO_NORMAL packets processing
to all napi_gro_receive() users, including mac80211-based drivers.

However, this change has led to a regression in iwlwifi driver [1][2] as
it is required for NAPI users to call napi_complete_done() or
napi_complete() and the end of every polling iteration, whilst iwlwifi
doesn't use NAPI scheduling at all and just calls napi_gro_flush().
In that particular case, packets which have not been already flushed
from napi->rx_list stall in it until at least next Rx cycle.

Fix this by adding a manual flushing of the list to iwlwifi driver right
before napi_gro_flush() call to mimic napi_complete() logics.

I prefer to open-code gro_normal_list() rather than exporting it for 2
reasons:
* to prevent from using it and napi_gro_flush() in any new drivers,
  as it is the *really* bad way to use NAPI that should be avoided;
* to keep gro_normal_list() static and don't lose any CC optimizations.

I also don't add the "Fixes:" tag as the mentioned commit was only a
trigger that only exposed an improper usage of NAPI in this particular
driver.

[1] https://lore.kernel.org/netdev/PSXP216MB04388962C411CD0B17A86F47804A0@PSXP216MB0438.KORP216.PROD.OUTLOOK.COM
[2] https://bugzilla.kernel.org/show_bug.cgi?id=205647

Signed-off-by: Alexander Lobakin 
Acked-by: Luca Coelho 
Reported-by: Nicholas Johnson 
Tested-by: Nicholas Johnson 
Reviewed-by: Edward Cree 
Signed-off-by: David S. Miller 
Signed-off-by: Sasha Levin

iwlwifi: check kasprintf() return value

2019-12-31T15:45:42+00:00

[ Upstream commit 5974fbb5e10b018fdbe3c3b81cb4cc54e1105ab9 ]

kasprintf() can fail, we should check the return value.

Fixes: 5ed540aecc2a ("iwlwifi: use mac80211 throughput trigger")
Fixes: 8ca151b568b6 ("iwlwifi: add the MVM driver")
Signed-off-by: Johannes Berg 
Signed-off-by: Luca Coelho 
Signed-off-by: Sasha Levin

iwlwifi: mvm: fix unaligned read of rx_pkt_status

2019-12-31T15:45:31+00:00

[ Upstream commit c5aaa8be29b25dfe1731e9a8b19fd91b7b789ee3 ]

This is present since the introduction of iwlmvm.
Example stack trace on MIPS:

[] iwl_mvm_rx_rx_mpdu+0xa8/0xb88 [iwlmvm]
[] iwl_pcie_rx_handle+0x420/0xc48 [iwlwifi]

Tested with a Wireless AC 7265 for ~6 months, confirmed to fix the
problem. No other unaligned accesses are spotted yet.

Signed-off-by: Wang Xuerui 
Signed-off-by: Luca Coelho 
Signed-off-by: Kalle Valo 
Signed-off-by: Sasha Levin

iwlwifi: pcie: fix support for transmitting SKBs with fraglist

2019-12-17T18:56:01+00:00

commit 4f4925a7b23428d5719af5a2816586b2a0e6fd19 upstream.

When the implementation of SKBs with fraglist was sent upstream, a
merge-damage occurred and half the patch was not applied.

This causes problems in high-throughput situations with AX200 devices,
including low throughput and FW crashes.

Introduce the part that was missing from the original patch.

Fixes: 0044f1716c4d ("iwlwifi: pcie: support transmitting SKBs with fraglist")
Cc: stable@vger.kernel.org # 4.20+
Signed-off-by: Johannes Berg 
[ This patch was created by me, but the original author of this code
  is Johannes, so his s-o-b is here and he's marked as the author of
  the patch. ]
Signed-off-by: Luca Coelho 
Signed-off-by: Greg Kroah-Hartman

iwlwifi: pcie: don't consider IV len in A-MSDU

2019-11-08T19:33:37+00:00

From gen2 PN is totally offloaded to hardware (also the space for the
IV isn't part of the skb).  As you can see in mvm/mac80211.c:3545, the
MAC for cipher types CCMP/GCMP doesn't set
IEEE80211_KEY_FLAG_PUT_IV_SPACE for gen2 NICs.

This causes all the AMSDU data to be corrupted with cipher enabled.

Signed-off-by: Mordechay Goodstein 
Signed-off-by: Luca Coelho 
Signed-off-by: Kalle Valo

iwlwifi: fw api: support new API for scan config cmd

2019-10-30T15:00:26+00:00

The API was reduced to include only knowledge currently needed by the
FW scan logic, the rest is legacy.  Support the new, reduced version.

Using the old API with newer firmwares (starting from
iwlwifi-*-50.ucode, which implements and requires the new API version)
causes an assertion failure similar to this one:

[    2.854505] iwlwifi 0000:00:14.3: 0x20000038 | BAD_COMMAND

Signed-off-by: Ayala Beker 
Signed-off-by: Luca Coelho 
Signed-off-by: Kalle Valo

iwlwifi: pcie: 0x2720 is qu and 0x30DC is not

2019-10-23T10:31:34+00:00

When converting the wrong qu configurations in an earlier commit, I
accidentally swapped 0x2720 and 0x30DC.  Instead of converting 0x2720,
I converted 0x30DC.  Undo 0x30DC and convert 0x2720.

Signed-off-by: Luca Coelho 
Signed-off-by: Kalle Valo

iwlwifi: pcie: add workaround for power gating in integrated 22000

2019-10-23T10:31:33+00:00

Add a workaround that forces power gating to be enabled on integrated
22000 devices.  This improves power saving in certain situations.

Signed-off-by: Luca Coelho 
Signed-off-by: Kalle Valo

iwlwifi: mvm: handle iwl_mvm_tvqm_enable_txq() error return

2019-10-23T10:31:33+00:00

iwl_mvm_tvqm_enable_txq() can return an error, notably if unable
to allocate memory for the queue. Handle this error throughout,
avoiding storing the invalid value into a u16 which later leads
to a disable of an invalid queue ("queue 65524 not used", where
65524 is just -ENOMEM in a u16).

Signed-off-by: Johannes Berg 
Signed-off-by: Luca Coelho 
Signed-off-by: Kalle Valo