[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <6BEA4AB8-F99A-4D8D-847F-C5691F3EE1DB@gmail.com>
Date: Tue, 16 Feb 2016 12:12:50 +0900
From: Jaedon Shin <jaedon.shin@...il.com>
To: Florian Fainelli <f.fainelli@...il.com>
Cc: netdev@...r.kernel.org, David Miller <davem@...emloft.net>,
Petri Gynther <pgynther@...gle.com>, opendmb@...il.com
Subject: Re: [PATCH] net: bcmgenet: Add MDIO_INTR in GENETv2
Hi Florian,
> On Feb 16, 2016, at 3:18 AM, Florian Fainelli <f.fainelli@...il.com> wrote:
>
> Hi Jaedon,
>
> On 15/02/2016 00:42, Jaedon Shin wrote:
>> The GENETv2 chipsets has MDIO interrupt like the GENETv3+ chipsets.
>>
>> The previous commit d5c3d84657db ("net: phy: Avoid polling PHY with
>> PHY_IGNORE_INTERRUPTS") and commit 49f7a471e4d1 ("net: bcmgenet: Properly
>> configure PHY to ignore interrupt") cause link-down PHY always in some
>> 40nm generation chipsets.
>
> Humm, these are two different things here:
>
> - GENET_HAS_MDIO_INTR is about telling the driver whether the hardware
> supports MDIO_INTR_DONE and MDIO_INTR_ERROR
> - eliminating PHY polling is about utilizing LINK_UP and LINK_DOWN to
> avoid polling the PHY
>
> So the original problem is actually here:
>
> bcmgenet_irq_task():
>
> /* Link UP/DOWN event */
> if ((priv->hw_params->flags & GENET_HAS_MDIO_INTR) &&
> (priv->irq0_stat & UMAC_IRQ_LINK_EVENT)) {
>
> These two checks are actually orthogonal, so we should remove the first
> check on GENET_HAS_MDIO_INTR.
>
As you said, the part in bcmgenet_irq_task() is a problem.
The bcmgenet using internal PHY should use phy_mac_interrupt() cause it has not
PHY_POLL, and it depends on Ethernet MAC ISR.
UMAC_IRQ_LINK_EVENT(LINK_UP and LINK_DOWN) was working correctly in GENETv2,
but (priv->hw_params->flags & GENET_HAS_MDIO_INTR) was blocking to call
phy_mac_interrupt(). I didn't find a reason through datasheet without MDIO_INTR
in GENETv2. However, I'm not sure using MDIO_INTR.
Therefore if MDIO_INTR is not valid in GENETv2, I will send the patch again
to remove the first chicken GENET_HAS_MDIO_INTR after your confirm.
Thanks,
Jaedon
> Your patch remains valid though, just the explanation needs a bit
> tweaking, thanks!
>
>>
>> Signed-off-by: Jaedon Shin <jaedon.shin@...il.com>
>> ---
>> drivers/net/ethernet/broadcom/genet/bcmgenet.c | 4 ++--
>> drivers/net/ethernet/broadcom/genet/bcmgenet.h | 2 +-
>> 2 files changed, 3 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c
>> index b15a60d787c7..8e9aa8f6390d 100644
>> --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c
>> +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c
>> @@ -1904,7 +1904,7 @@ static int init_umac(struct bcmgenet_priv *priv)
>> bcmgenet_bp_mc_set(priv, reg);
>> }
>>
>> - /* Enable MDIO interrupts on GENET v3+ */
>> + /* Enable MDIO interrupts on GENET v2+ */
>> if (priv->hw_params->flags & GENET_HAS_MDIO_INTR)
>> int0_enable |= (UMAC_IRQ_MDIO_DONE | UMAC_IRQ_MDIO_ERROR);
>>
>> @@ -3168,7 +3168,7 @@ static struct bcmgenet_hw_params bcmgenet_hw_params[] = {
>> .rdma_offset = 0x3000,
>> .tdma_offset = 0x4000,
>> .words_per_bd = 2,
>> - .flags = GENET_HAS_EXT,
>> + .flags = GENET_HAS_EXT | GENET_HAS_MDIO_INTR,
>> },
>> [GENET_V3] = {
>> .tx_queues = 4,
>> diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.h b/drivers/net/ethernet/broadcom/genet/bcmgenet.h
>> index 967367557309..c14bfbfbe06a 100644
>> --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.h
>> +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.h
>> @@ -310,7 +310,7 @@ struct bcmgenet_mib_counters {
>> #define UMAC_IRQ_TXDMA_BDONE (1 << 18)
>> #define UMAC_IRQ_TXDMA_DONE UMAC_IRQ_TXDMA_MBDONE
>>
>> -/* Only valid for GENETv3+ */
>> +/* Only valid for GENETv2+ */
>> #define UMAC_IRQ_MDIO_DONE (1 << 23)
>> #define UMAC_IRQ_MDIO_ERROR (1 << 24)
>>
>>
Powered by blists - more mailing lists