On Tue, Jan 02, 2018 at 10:58:48AM +0000, Russell King wrote:
> For paged accesses to be truely safe, we need to hold the bus lock to
> prevent anyone else gaining access to the registers while we modify
> them.
>
> The phydev->lock mutex does not do this: userspace via the MII ioctl
> can still sneak in and read or write any register while we are on a
> different page, and the suspend/resume methods can be called by a
> thread different to the thread polling the phy status.
>
> Races have been observed with mvneta on SolidRun Clearfog with phylink,
> particularly between the phylib worker reading the PHYs status, and
> the thread resuming mvneta, calling phy_start() which then calls
> through to m88e1121_config_aneg_rgmii_delays(), which tries to
> read-modify-write the MSCR register:
>
> CPU0 CPU1
> marvell_read_status_page()
> marvell_set_page(phydev, MII_MARVELL_FIBER_PAGE)
> ...
> m88e1121_config_aneg_rgmii_delays()
> set_page(MII_MARVELL_MSCR_PAGE)
> phy_read(phydev, MII_88E1121_PHY_MSCR_REG)
> marvell_set_page(phydev, MII_MARVELL_COPPER_PAGE);
> ...
> phy_write(phydev, MII_88E1121_PHY_MSCR_REG)
>
> The result of this is we end up writing the copper page register 21,
> which causes the copper PHY to be disabled, and the link partner sees
> the link immediately go down.
>
> Solve this by taking the bus lock instead of the PHY lock, thereby
> preventing other accesses to the PHY while we are accessing other PHY
> pages.
>
> Signed-off-by: Russell King <rmk+ker...@armlinux.org.uk>
Reviewed-by: Andrew Lunn <and...@lunn.ch>
Andrew
