e1000e: log when swflag is cleared unexpectedly on ICH/PCH devices
authorBruce Allan <bruce.w.allan@intel.com>
Fri, 13 May 2011 07:19:53 +0000 (07:19 +0000)
committerJeff Kirsher <jeffrey.t.kirsher@intel.com>
Fri, 10 Jun 2011 03:32:28 +0000 (20:32 -0700)
Since EXTCNF_CTRL.SWFLAG (used in the ownership arbitration of shared
resources, e.g. the PHY shared between the s/w, f/w, and h/w clients)
can be cleared by any of those clients, log a debug message when
software attempts to clear it and it is already cleared unexpectedly.
And since the swflag is cleared by a hardware reset, the driver does
not need to do that, but the mutex acquired when the bit is set must
still be cleared.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
drivers/net/e1000e/ich8lan.c

index dcd5db5e34e5d6577667a25af28622c8146622af..dddd0b3dd21b7da0b0cabf3dffb029d8cd2fc80e 100644 (file)
@@ -889,8 +889,13 @@ static void e1000_release_swflag_ich8lan(struct e1000_hw *hw)
        u32 extcnf_ctrl;
 
        extcnf_ctrl = er32(EXTCNF_CTRL);
-       extcnf_ctrl &= ~E1000_EXTCNF_CTRL_SWFLAG;
-       ew32(EXTCNF_CTRL, extcnf_ctrl);
+
+       if (extcnf_ctrl & E1000_EXTCNF_CTRL_SWFLAG) {
+               extcnf_ctrl &= ~E1000_EXTCNF_CTRL_SWFLAG;
+               ew32(EXTCNF_CTRL, extcnf_ctrl);
+       } else {
+               e_dbg("Semaphore unexpectedly released by sw/fw/hw\n");
+       }
 
        mutex_unlock(&swflag_mutex);
 }
@@ -3066,7 +3071,7 @@ static s32 e1000_reset_hw_ich8lan(struct e1000_hw *hw)
        msleep(20);
 
        if (!ret_val)
-               e1000_release_swflag_ich8lan(hw);
+               mutex_unlock(&swflag_mutex);
 
        if (ctrl & E1000_CTRL_PHY_RST) {
                ret_val = hw->phy.ops.get_cfg_done(hw);