devicetree.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Prasanna Vengateshan <prasanna.vengateshan@microchip.com>
To: Vladimir Oltean <olteanv@gmail.com>
Cc: <andrew@lunn.ch>, <netdev@vger.kernel.org>, <robh+dt@kernel.org>,
	<UNGLinuxDriver@microchip.com>, <hkallweit1@gmail.com>,
	<linux@armlinux.org.uk>, <davem@davemloft.net>, <kuba@kernel.org>,
	<linux-kernel@vger.kernel.org>, <vivien.didelot@gmail.com>,
	<f.fainelli@gmail.com>, <devicetree@vger.kernel.org>
Subject: Re: [PATCH v2 net-next 9/9] net: dsa: microchip: add support for vlan operations
Date: Thu, 6 May 2021 20:21:35 +0530	[thread overview]
Message-ID: <ab9da4b759dec9ab69bac791300a6a8977ee7cc2.camel@microchip.com> (raw)
In-Reply-To: <20210422190351.qdv2xlnxghmfpjqj@skbuf>

Hi Vladimir,

On Thu, 2021-04-22 at 22:03 +0300, Vladimir Oltean wrote:
> EXTERNAL EMAIL: Do not click links or open attachments unless you know the
> content is safe
> 
> On Thu, Apr 22, 2021 at 03:12:57PM +0530, Prasanna Vengateshan wrote:
> > Support for VLAN add, del, prepare and filtering operations.
> > 
> > It aligns with latest update of removing switchdev
> > transactional logic from VLAN objects
> 
> Maybe more in the commit message about what the patch does, as opposed
> to mentioning that you had to rebase it, would be helpful.
Sure.

> 
> > 
> > Signed-off-by: Prasanna Vengateshan <prasanna.vengateshan@microchip.com>
> > ---
> >  drivers/net/dsa/microchip/lan937x_main.c | 214 +++++++++++++++++++++++
> >  1 file changed, 214 insertions(+)
> > 
> > diff --git a/drivers/net/dsa/microchip/lan937x_main.c
> > b/drivers/net/dsa/microchip/lan937x_main.c
> > index 7f6183dc0e31..35f3456c3506 100644
> 
> > +
> > +             rc = lan937x_cfg(dev, REG_SW_LUE_CTRL_0, SW_VLAN_ENABLE,
> > true);
> 
> How about this bit?

> 
> I see one bit is per port and the other is global.
> Just FYI, you can have this configuration:
> 
> ip link add br0 type bridge vlan_filtering 0
> ip link add br1 type bridge vlan_filtering 1
> ip link set swp0 master br0
> ip link set swp1 master br0
> ip link set swp2 master br1
> ip link set swp3 master br1
> 
> Do the swp0 and swp1 ports remain VLAN-unaware after you touch this
> REG_SW_LUE_CTRL_0 bit?
vlan aware is global, so ds->vlan_filtering_is_global needs to be true if VLAN
aware is global, will fix this in the next version

> 
> > +     } else {
> > +             rc = lan937x_cfg(dev, REG_SW_LUE_CTRL_0, SW_VLAN_ENABLE,
> > false);
> > +             if (rc < 0)
> > +                     return rc;
> > +
> > +             rc = lan937x_port_cfg(dev, port, REG_PORT_LUE_CTRL,
> > +                                   PORT_VLAN_LOOKUP_VID_0, false);
> > +     }
> > +
> > +     return rc;
> > +}
> > +
> > +static int lan937x_port_vlan_add(struct dsa_switch *ds, int port,
> > +                              const struct switchdev_obj_port_vlan *vlan,
> > +                              struct netlink_ext_ack *extack)
> > +{
> > +     bool untagged = vlan->flags & BRIDGE_VLAN_INFO_UNTAGGED;
> > +     struct ksz_device *dev = ds->priv;
> > +     u32 vlan_table[3];
> 
> Maybe a structure would be nicer to read than an u32 array?
Okay, will make a structure.

> 
> > +     int rc;
> > +
> > +     rc = lan937x_get_vlan_table(dev, vlan->vid, vlan_table);
> > +     if (rc < 0) {
> > +             dev_err(dev->dev, "Failed to get vlan table\n");
> 
> One of the reasons for which the extack exists is so that you can report
> errors to user space and not to the console.
Sure, will add it for port_vlan_del() as well

> 
>                 NL_SET_ERR_MSG_MOD(extack, "Failed to get vlan table");
> 
> > +             return rc;
> > +     }
> > +
> > +     vlan_table[0] = VLAN_VALID | (vlan->vid & VLAN_FID_M);
> > +
> > +     /* set/clear switch port when updating vlan table registers */
> > +     if (untagged)
> > +             vlan_table[1] |= BIT(port);
> > +     else
> > +             vlan_table[1] &= ~BIT(port);
> > +     vlan_table[1] &= ~(BIT(dev->cpu_port));
> > +
> > +     vlan_table[2] |= BIT(port) | BIT(dev->cpu_port);
> 
> What's the business with the CPU port here? Does DSA not call
> .port_vlan_add for the CPU port separately?
Calls for CPU port as well. This is to be removed.

> 
> > +
> > +     rc = lan937x_set_vlan_table(dev, vlan->vid, vlan_table);
> > +     if (rc < 0) {
> > +             dev_err(dev->dev, "Failed to set vlan table\n");
> > +             return rc;
> > +     }
> > +
> > +     /* change PVID */
> > +     if (vlan->flags & BRIDGE_VLAN_INFO_PVID) {
> > +             rc = lan937x_pwrite16(dev, port, REG_PORT_DEFAULT_VID, vlan-
> > >vid);
> > +
> > +             if (rc < 0) {
> > +                     dev_err(dev->dev, "Failed to set pvid\n");
> > +                     return rc;
> > +             }
> > +     }
> > +
> > +     return 0;
> > +}
> > +
> > +static int lan937x_port_vlan_del(struct dsa_switch *ds, int port,
> > +                              const struct switchdev_obj_port_vlan *vlan)
> > +{
> > +     bool untagged = vlan->flags & BRIDGE_VLAN_INFO_UNTAGGED;
> > +     struct ksz_device *dev = ds->priv;
> > +     u32 vlan_table[3];
> > +     u16 pvid;
> > +     int rc;
> > +
> > +     lan937x_pread16(dev, port, REG_PORT_DEFAULT_VID, &pvid);
> > +     pvid &= 0xFFF;
> > +
> > +     rc = lan937x_get_vlan_table(dev, vlan->vid, vlan_table);
> > +
> > +     if (rc < 0) {
> > +             dev_err(dev->dev, "Failed to get vlan table\n");
> > +             return rc;
> > +     }
> > +     /* clear switch port number */
> > +     vlan_table[2] &= ~BIT(port);
> > +
> > +     if (pvid == vlan->vid)
> > +             pvid = 1;
> 
> According to Documentation/networking/switchdev.rst:
> 
> When the bridge has VLAN filtering enabled and a PVID is not configured on the
> ingress port, untagged and 802.1p tagged packets must be dropped. When the
> bridge
> has VLAN filtering enabled and a PVID exists on the ingress port, untagged and
> priority-tagged packets must be accepted and forwarded according to the
> bridge's port membership of the PVID VLAN. When the bridge has VLAN filtering
> disabled, the presence/lack of a PVID should not influence the packet
> forwarding decision.
> 
> So please don't reset the pvid.
Will remove it in the next rev.

> 
> > +
> > +     if (untagged)
> > +             vlan_table[1] &= ~BIT(port);
> > +
> > +     rc = lan937x_set_vlan_table(dev, vlan->vid, vlan_table);
> > +     if (rc < 0) {
> > +             dev_err(dev->dev, "Failed to set vlan table\n");
> > +             return rc;
> > +     }
> > +
> > +     rc = lan937x_pwrite16(dev, port, REG_PORT_DEFAULT_VID, pvid);
> > +
> > +     if (rc < 0) {
> > +             dev_err(dev->dev, "Failed to set pvid\n");
> > +             return rc;
> > +     }
> > +
> > +     return 0;
> > +}
> > +
> >  static u8 lan937x_get_fid(u16 vid)
> >  {
> >       if (vid > ALU_FID_SIZE)
> > @@ -955,6 +1166,9 @@ const struct dsa_switch_ops lan937x_switch_ops = {
> >       .port_bridge_flags      = lan937x_port_bridge_flags,
> >       .port_stp_state_set     = lan937x_port_stp_state_set,
> >       .port_fast_age          = ksz_port_fast_age,
> > +     .port_vlan_filtering    = lan937x_port_vlan_filtering,
> > +     .port_vlan_add          = lan937x_port_vlan_add,
> > +     .port_vlan_del          = lan937x_port_vlan_del,
> >       .port_fdb_dump          = lan937x_port_fdb_dump,
> >       .port_fdb_add           = lan937x_port_fdb_add,
> >       .port_fdb_del           = lan937x_port_fdb_del,
> > --
> > 2.27.0
> > 



      reply	other threads:[~2021-05-06 14:51 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-04-22  9:42 [PATCH v2 net-next 0/9] net: dsa: microchip: DSA driver support for LAN937x switch Prasanna Vengateshan
2021-04-22  9:42 ` [PATCH v2 net-next 1/9] dt-bindings: net: dsa: dt bindings for microchip lan937x Prasanna Vengateshan
2021-04-22 15:30   ` Rob Herring
2021-04-22 17:38   ` Rob Herring
2021-04-26  4:05     ` Prasanna Vengateshan
2021-04-26 16:04       ` Florian Fainelli
2021-04-22  9:42 ` [PATCH v2 net-next 2/9] net: phy: Add support for LAN937x T1 phy driver Prasanna Vengateshan
2021-04-22 12:45   ` Andrew Lunn
2021-04-26  4:06     ` Prasanna Vengateshan
2021-04-22  9:42 ` [PATCH v2 net-next 3/9] net: dsa: tag_ksz: add tag handling for Microchip LAN937x Prasanna Vengateshan
2021-04-22 19:18   ` Vladimir Oltean
2021-04-26  4:33     ` Prasanna Vengateshan
2021-04-26 12:27       ` Andrew Lunn
2021-04-22  9:42 ` [PATCH v2 net-next 4/9] net: dsa: microchip: add DSA support for microchip lan937x Prasanna Vengateshan
2021-04-22 19:59   ` Vladimir Oltean
2021-04-22 23:28     ` Andrew Lunn
2021-04-26  8:54       ` Prasanna Vengateshan
2021-04-26 12:34         ` Andrew Lunn
2021-04-27  7:43     ` Prasanna Vengateshan
2021-04-22 23:32   ` Andrew Lunn
2021-04-22 23:38   ` Andrew Lunn
2021-04-22 23:43   ` Andrew Lunn
2021-04-22  9:42 ` [PATCH v2 net-next 5/9] net: dsa: microchip: add support for phylink management Prasanna Vengateshan
2021-04-22 20:05   ` Vladimir Oltean
2021-04-22  9:42 ` [PATCH v2 net-next 6/9] net: dsa: microchip: add support for ethtool port counters Prasanna Vengateshan
2021-04-22  9:42 ` [PATCH v2 net-next 7/9] net: dsa: microchip: add support for port mirror operations Prasanna Vengateshan
2021-04-22 20:11   ` Vladimir Oltean
2021-04-22  9:42 ` [PATCH v2 net-next 8/9] net: dsa: microchip: add support for fdb and mdb management Prasanna Vengateshan
2021-04-22  9:42 ` [PATCH v2 net-next 9/9] net: dsa: microchip: add support for vlan operations Prasanna Vengateshan
2021-04-22 19:03   ` Vladimir Oltean
2021-05-06 14:51     ` Prasanna Vengateshan [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=ab9da4b759dec9ab69bac791300a6a8977ee7cc2.camel@microchip.com \
    --to=prasanna.vengateshan@microchip.com \
    --cc=UNGLinuxDriver@microchip.com \
    --cc=andrew@lunn.ch \
    --cc=davem@davemloft.net \
    --cc=devicetree@vger.kernel.org \
    --cc=f.fainelli@gmail.com \
    --cc=hkallweit1@gmail.com \
    --cc=kuba@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux@armlinux.org.uk \
    --cc=netdev@vger.kernel.org \
    --cc=olteanv@gmail.com \
    --cc=robh+dt@kernel.org \
    --cc=vivien.didelot@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).