Hi Rafał, Hauke,

I got a report of another issue, which was introduced by your commit

commit aa51e598d04c6acf5477934cd6383f5a17ce9029
Author: Hauke Mehrtens <hauke@hauke-m.de>
Date:   Sat Aug 24 00:32:31 2013 +0200

     brcmsmac: use bcma PCIe up and down functions

     replace the calls to bcma_core_pci_extend_L1timer() by calls to the
     newly introduced bcma_core_pci_ip() and bcma_core_pci_down()

     Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
     Cc: Arend van Spriel <arend@broadcom.com>
     Signed-off-by: John W. Linville <linville@tuxdriver.com>

The call to bcma_core_pci_up() ends up in bcma_pcie_mdio_read() which 
does a usleep_range(). However, brcmsmac is in atomic context during 
this call so we get a warning. I am in doubt on how to resolve this. 
Should we require drivers to call this (low-level) function from 
non-atomic context or not?

Regards,
Arend

-------- Original Message --------
Subject: brcmsmac follow up
Date: Mon, 9 Sep 2013 13:06:16 -0700
From: Tod Jackson <tod.jackson@gmail.com>
To: brcm80211-dev-list <brcm80211-dev-list@broadcom.com>

This is a follow-up to my brcmsmac bug report from just a bit ago.

I tested the wireless-testing tree that has a bcma commit about half
an hour ago, and internet works, but there are a bunch of backtraces.
I've attached the full dmesg.