Revert "mwifiex: fix system hang problem after resume"
authorBrian Norris <briannorris@chromium.org>
Mon, 5 Aug 2019 17:15:04 +0000 (10:15 -0700)
committerKalle Valo <kvalo@codeaurora.org>
Tue, 6 Aug 2019 12:43:22 +0000 (15:43 +0300)
commit654026df2635863fd695b2ca833e5c62454bd5ee
tree8ada3dae6a93e9d8b032cab63494d3f01f23f764
parentc85a6376673805a68633e5d7d7d1fccf4b207685
Revert "mwifiex: fix system hang problem after resume"

This reverts commit 437322ea2a36d112e20aa7282c869bf924b3a836.

This above-mentioned "fix" does not actually do anything to prevent a
race condition. It simply papers over it so that the issue doesn't
appear.

If this is a real problem, it should be explained better than the above
commit does, and an alternative, non-racy solution should be found.

For further reason to revert this: there's no reason we can't try
resetting the card when it's *actually* stuck in host-sleep mode. So
instead, this is unnecessarily creating scenarios where we can't recover
Wifi (and in fact, I'm fielding reports of Chromebooks that can't
recover after the aforementioned commit).

Note that this was proposed in 2017 and Ack'ed then, but due to my
marking as RFC, it never went anywhere:

https://patchwork.kernel.org/patch/9657277/
[RFC] Revert "mwifiex: fix system hang problem after resume"

Cc: Amitkumar Karwar <akarwar@marvell.com>
Signed-off-by: Brian Norris <briannorris@chromium.org>
Reviewed-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Acked-by: Amitkumar Karwar <amitkarwar@gmail.com>
Tested-by: Matthias Kaehlcke <mka@chromium.org>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
drivers/net/wireless/marvell/mwifiex/init.c