LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Horatiu Vultur <horatiu.vultur@microchip.com>
To: Andrew Lunn <andrew@lunn.ch>
Cc: <davem@davemloft.net>, <kuba@kernel.org>, <robh+dt@kernel.org>,
	<UNGLinuxDriver@microchip.com>, <p.zabel@pengutronix.de>,
	<linux@armlinux.org.uk>, <netdev@vger.kernel.org>,
	<devicetree@vger.kernel.org>, <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH net-next v2 3/6] net: lan966x: add port module support
Date: Thu, 25 Nov 2021 10:26:38 +0100	[thread overview]
Message-ID: <20211125092638.7b2u75zdv2ulekmo@soft-dev3-1.localhost> (raw)
In-Reply-To: <YZ59hpDWjNjvx5kP@lunn.ch>

The 11/24/2021 18:59, Andrew Lunn wrote:

Hi Andrew,

> 
> > +static void lan966x_ifh_inject(u32 *ifh, size_t val, size_t pos, size_t length)
> > +{
> > +     int i;
> > +
> > +     for (i = pos; i < pos + length; ++i) {
> > +             if (val & BIT(i - pos))
> > +                     ifh[IFH_LEN - i / 32 - 1] |= BIT(i % 32);
> > +             else
> > +                     ifh[IFH_LEN - i / 32 - 1] &= ~(BIT(i % 32));
> > +     }
> > +}
> > +
> > +static void lan966x_gen_ifh(u32 *ifh, struct lan966x_frame_info *info,
> > +                         struct lan966x *lan966x)
> > +{
> > +     lan966x_ifh_inject(ifh, 1, IFH_POS_BYPASS, 1);
> > +     lan966x_ifh_inject(ifh, info->port, IFH_POS_DSTS, IFH_WID_DSTS);
> > +     lan966x_ifh_inject(ifh, info->qos_class, IFH_POS_QOS_CLASS,
> > +                        IFH_WID_QOS_CLASS);
> > +     lan966x_ifh_inject(ifh, info->ipv, IFH_POS_IPV, IFH_WID_IPV);
> > +}
> > +
> 
> > +     /* Write IFH header */
> > +     for (i = 0; i < IFH_LEN; ++i) {
> > +             /* Wait until the fifo is ready */
> > +             while (!(QS_INJ_STATUS_FIFO_RDY_GET(lan_rd(lan966x, QS_INJ_STATUS)) &
> > +                      BIT(grp)))
> > +                     ;
> > +
> > +             lan_wr((__force u32)cpu_to_be32(ifh[i]), lan966x,
> > +                    QS_INJ_WR(grp));
> 
> There is a lot of magic going on here constructing the IFH. Is it
> possible to define the structure using bit fields and __be32. You
> should then be able to skip this cpu_to_be32 and the ugly cast. And
> the actual structure should be a lot clearer.

If I undestood you correctly I have tried to do the following:

struct lan966x_ifh {
    __be32 timestamp;
    __be32 bypass : 1;
    __be32 port : 3;
    ...
};

But then I start to get errors from sparse:

error: invalid bitfield specifier for type restricted __be32.

On thing that I can do is to use packing() instead of
lan966x_ifh_inject() and declare ifh as __be32.
I think this will also work and simplify a little bit the code.

> 
> > +static int lan966x_rx_frame_word(struct lan966x *lan966x, u8 grp, bool ifh,
> > +                              u32 *rval)
> > +{
> > +     u32 bytes_valid;
> > +     u32 val;
> > +
> > +     val = lan_rd(lan966x, QS_XTR_RD(grp));
> > +     if (val == XTR_NOT_READY) {
> > +             if (ifh)
> > +                     return -EIO;
> > +
> > +             do {
> > +                     val = lan_rd(lan966x, QS_XTR_RD(grp));
> > +             } while (val == XTR_NOT_READY);
> 
> I would add some sort of timeout here, just in case the hardware
> breaks. You have quite a few such loops, it would be better to make
> use of the helpers in linux/iopoll.h.

Yes, I will do that, I will also add sime timeout also for injection.

> 

-- 
/Horatiu

  reply	other threads:[~2021-11-25  9:26 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-11-23 13:55 [PATCH net-next v2 0/6] net: lan966x: Add lan966x switch driver Horatiu Vultur
2021-11-23 13:55 ` [PATCH net-next v2 1/6] dt-bindings: net: lan966x: Add lan966x-switch bindings Horatiu Vultur
2021-11-23 13:55 ` [PATCH net-next v2 2/6] net: lan966x: add the basic lan966x driver Horatiu Vultur
2021-11-23 15:31   ` Denis Kirjanov
2021-11-24  8:13     ` Horatiu Vultur
2021-11-24  3:30   ` Jakub Kicinski
2021-11-24  8:14     ` Horatiu Vultur
2021-11-23 13:55 ` [PATCH net-next v2 3/6] net: lan966x: add port module support Horatiu Vultur
2021-11-24 17:59   ` Andrew Lunn
2021-11-25  9:26     ` Horatiu Vultur [this message]
2021-11-25 14:59       ` Andrew Lunn
2021-11-26  9:12         ` Horatiu Vultur
2021-11-23 13:55 ` [PATCH net-next v2 4/6] net: lan966x: add mactable support Horatiu Vultur
2021-11-23 13:55 ` [PATCH net-next v2 5/6] net: lan966x: add ethtool configuration and statistics Horatiu Vultur
2021-11-23 13:55 ` [PATCH net-next v2 6/6] net: lan966x: Update MAINTAINERS to include lan966x driver Horatiu Vultur

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=20211125092638.7b2u75zdv2ulekmo@soft-dev3-1.localhost \
    --to=horatiu.vultur@microchip.com \
    --cc=UNGLinuxDriver@microchip.com \
    --cc=andrew@lunn.ch \
    --cc=davem@davemloft.net \
    --cc=devicetree@vger.kernel.org \
    --cc=kuba@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux@armlinux.org.uk \
    --cc=netdev@vger.kernel.org \
    --cc=p.zabel@pengutronix.de \
    --cc=robh+dt@kernel.org \
    --subject='Re: [PATCH net-next v2 3/6] net: lan966x: add port module support' \
    /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

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).