On Thu, 30 Apr 2020, Badel, Laurent wrote: > -----Original Message----- >> From: Heiner Kallweit >> Sent: Wednesday, April 29, 2020 7:06 PM >> To: Badel, Laurent ; fugang.duan@nxp.com; >> netdev@vger.kernel.org; andrew@lunn.ch; f.fainelli@gmail.com; >> linux@armlinux.org.uk; richard.leitner@skidata.com; >> davem@davemloft.net; alexander.levin@microsoft.com; >> gregkh@linuxfoundation.org >> Cc: Quette, Arnaud >> Subject: [EXTERNAL] Re: [PATCH 2/2] Reset PHY in phy_init_hw() before >> interrupt configuration >> >> On 29.04.2020 11:03, Badel, Laurent wrote: >>> Description: this patch adds a reset of the PHY in phy_init_hw() >>> for PHY drivers bearing the PHY_RST_AFTER_CLK_EN flag. >>> >>> Rationale: due to the PHY reset reverting the interrupt mask to default, >>> it is necessary to either perform the reset before PHY configuration, >>> or re-configure the PHY after reset. This patch implements the former >>> as it is simpler and more generic. >>> >>> Fixes: 1b0a83ac04e383e3bed21332962b90710fcf2828 ("net: fec: add >> phy_reset_after_clk_enable() support") >>> Signed-off-by: Laurent Badel >>> >>> --- >>> drivers/net/phy/phy_device.c | 7 +++++-- >>> 1 file changed, 5 insertions(+), 2 deletions(-) >>> >>> diff --git a/drivers/net/phy/phy_device.c b/drivers/net/phy/phy_device.c >>> index 28e3c5c0e..2cc511364 100644 >>> --- a/drivers/net/phy/phy_device.c >>> +++ b/drivers/net/phy/phy_device.c >>> @@ -1082,8 +1082,11 @@ int phy_init_hw(struct phy_device *phydev) >>> { >>> int ret = 0; >>> >>> - /* Deassert the reset signal */ >>> - phy_device_reset(phydev, 0); >>> + /* Deassert the reset signal >>> + * If the PHY needs a reset, do it now >>> + */ >>> + if (!phy_reset_after_clk_enable(phydev)) >> >> If reset is asserted when entering phy_init_hw(), then >> phy_reset_after_clk_enable() basically becomes a no-op. >> Still it should work as expected due to the reset signal being >> deasserted. It would be worth describing in the comment >> why the code still works in this case. >> > > Thank you for the comment, this is a very good point. > I will make sure to include some description when resubmitting. > I had previously tested this and what I saw was that the first > time you bring up the interface, the reset is not asserted so > that phy_reset_after_clk_enable() is effective. > The subsequent times the interface is brought up, the reset > is already asserted when entering phy_init_hw(), so that > it becomes a no-op as you correctly pointed out. However, > that didn't cause any problem on my board, presumably because > in that case the clock is already running when the PHY comes > out of reset. > I will re-test this carefully against the 'net' tree, though, > before coming to conclusions. > I have two additional things to take into account: * phy_reset_after_clk_enable() shoulnd't be longer called that way since it is now misleading -> the phy is no longer reset after clock enable but during hw_init() * how about fec_resume()? I don't think hw_init() is called then and so phy_reset_after_clk_enable() will no longer be called. >>> + phy_device_reset(phydev, 0); >>> >>> if (!phydev->drv) >>> return 0; >>> > >