| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <MWHPR18MB142117DBE8659D68794E18D3DEB39@MWHPR18MB1421.namprd18.prod.outlook.com>
Date: Thu, 4 Feb 2021 17:37:41 +0000
From: Hariprasad Kelam <hkelam@...vell.com>
To: Jakub Kicinski <kuba@...nel.org>
CC: "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"davem@...emloft.net" <davem@...emloft.net>,
"willemdebruijn.kernel@...il.com" <willemdebruijn.kernel@...il.com>,
"andrew@...n.ch" <andrew@...n.ch>,
Sunil Kovvuri Goutham <sgoutham@...vell.com>,
Linu Cherian <lcherian@...vell.com>,
"Geethasowjanya Akula" <gakula@...vell.com>,
Jerin Jacob Kollanukkaran <jerinj@...vell.com>,
Subbaraya Sundeep Bhatta <sbhatta@...vell.com>
Subject: Re: [Patch v3 net-next 7/7] octeontx2-pf: ethtool physical link
configuration
Hi Jakub,
> -----Original Message-----
> From: Jakub Kicinski <kuba@...nel.org>
> Sent: Wednesday, February 3, 2021 6:59 AM
> To: Hariprasad Kelam <hkelam@...vell.com>
> Cc: netdev@...r.kernel.org; linux-kernel@...r.kernel.org;
> davem@...emloft.net; willemdebruijn.kernel@...il.com;
> andrew@...n.ch; Sunil Kovvuri Goutham <sgoutham@...vell.com>; Linu
> Cherian <lcherian@...vell.com>; Geethasowjanya Akula
> <gakula@...vell.com>; Jerin Jacob Kollanukkaran <jerinj@...vell.com>;
> Subbaraya Sundeep Bhatta <sbhatta@...vell.com>
> Subject: [EXT] Re: [Patch v3 net-next 7/7] octeontx2-pf: ethtool physical link
> configuration
>
> On Sun, 31 Jan 2021 18:41:05 +0530 Hariprasad Kelam wrote:
> > From: Christina Jacob <cjacob@...vell.com>
> >
> > Register set_link_ksetting callback with driver such that link
> > configurations parameters like advertised mode,speed, duplex and
> > autoneg can be configured.
> >
> > below command
> > ethtool -s eth0 advertise 0x1 speed 10 duplex full autoneg on
> >
> > Signed-off-by: Christina Jacob <cjacob@...vell.com>
> > Signed-off-by: Sunil Goutham <sgoutham@...vell.com>
> > Signed-off-by: Hariprasad Kelam <hkelam@...vell.com>
> > ---
> > .../ethernet/marvell/octeontx2/nic/otx2_ethtool.c | 67
> > ++++++++++++++++++++++
> > 1 file changed, 67 insertions(+)
> >
> > diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_ethtool.c
> > b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_ethtool.c
> > index d637815..74a62de 100644
> > --- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_ethtool.c
> > +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_ethtool.c
> > @@ -1170,6 +1170,72 @@ static int otx2_get_link_ksettings(struct
> net_device *netdev,
> > return 0;
> > }
> >
> > +static void otx2_get_advertised_mode(const struct ethtool_link_ksettings
> *cmd,
> > + u64 *mode)
> > +{
> > + u32 bit_pos;
> > +
> > + /* Firmware does not support requesting multiple advertised modes
> > + * return first set bit
> > + */
> > + bit_pos = find_first_bit(cmd->link_modes.advertising,
> > + __ETHTOOL_LINK_MODE_MASK_NBITS);
> > + if (bit_pos != __ETHTOOL_LINK_MODE_MASK_NBITS)
> > + *mode = bit_pos;
> > +}
> > +
> > +static int otx2_set_link_ksettings(struct net_device *netdev,
> > + const struct ethtool_link_ksettings *cmd) {
> > + struct otx2_nic *pf = netdev_priv(netdev);
> > + struct ethtool_link_ksettings req_ks;
> > + struct ethtool_link_ksettings cur_ks;
> > + struct cgx_set_link_mode_req *req;
> > + struct mbox *mbox = &pf->mbox;
> > + int err = 0;
> > +
> > + /* save requested link settings */
> > + memcpy(&req_ks, cmd, sizeof(struct ethtool_link_ksettings));
>
> Why do you make this copy? The comment above does not help at all.
>
Agreed copy is not necessary . Added this copy for comparing advertised modes with supported modes.
Will fix this in next version.
> > + memset(&cur_ks, 0, sizeof(struct ethtool_link_ksettings));
> > +
> > + if (!ethtool_validate_speed(cmd->base.speed) ||
> > + !ethtool_validate_duplex(cmd->base.duplex))
> > + return -EINVAL;
> > +
> > + if (cmd->base.autoneg != AUTONEG_ENABLE &&
> > + cmd->base.autoneg != AUTONEG_DISABLE)
> > + return -EINVAL;
> > +
> > + otx2_get_link_ksettings(netdev, &cur_ks);
> > +
> > + /* Check requested modes against supported modes by hardware */
> > + if (!bitmap_subset(req_ks.link_modes.advertising,
> > + cur_ks.link_modes.supported,
> > + __ETHTOOL_LINK_MODE_MASK_NBITS))
> > + return -EINVAL;
> > +
> > + mutex_lock(&mbox->lock);
> > + req = otx2_mbox_alloc_msg_cgx_set_link_mode(&pf->mbox);
> > + if (!req) {
> > + err = -ENOMEM;
> > + goto end;
> > + }
> > +
> > + req->args.speed = req_ks.base.speed;
> > + /* firmware expects 1 for half duplex and 0 for full duplex
> > + * hence inverting
> > + */
> > + req->args.duplex = req_ks.base.duplex ^ 0x1;
> > + req->args.an = req_ks.base.autoneg;
> > + otx2_get_advertised_mode(&req_ks, &req->args.mode);
>
> But that only returns the first bit set. What does the device actually do? What
> if the user cleared a middle bit?
>
This is initial patch series to support advertised modes. Current firmware design is such that
It can handle only one advertised mode. Due to this limitation we are always checking
The first set bit in advertised modes and passing it to firmware.
Will add multi advertised mode support in near future.
Thanks,
Hariprasad k
> > + err = otx2_sync_mbox_msg(&pf->mbox);
> > +end:
> > + mutex_unlock(&mbox->lock);
> > + return err;
> > +}
> > +
> > static const struct ethtool_ops otx2_ethtool_ops = {
> > .supported_coalesce_params = ETHTOOL_COALESCE_USECS |
> > ETHTOOL_COALESCE_MAX_FRAMES, @@
> -1200,6 +1266,7 @@ static
> > const struct ethtool_ops otx2_ethtool_ops = {
> > .get_fecparam = otx2_get_fecparam,
> > .set_fecparam = otx2_set_fecparam,
> > .get_link_ksettings = otx2_get_link_ksettings,
> > + .set_link_ksettings = otx2_set_link_ksettings,
> > };
> >
> > void otx2_set_ethtool_ops(struct net_device *netdev)
Powered by blists - more mailing lists