From mboxrd@z Thu Jan 1 00:00:00 1970
From: Michael Tremer <michael.tremer@ipfire.org>
To: development@lists.ipfire.org
Subject: Re: [PATCH 2/4] kernel: add patches for SFP support on NXP
Layerscape/DPAA2 (arm64)
Date: Tue, 04 Oct 2022 09:57:23 +0100
Message-ID: <FDB3BF2D-A859-45E5-BE68-A0FE72A24BC9@ipfire.org>
In-Reply-To: <20221003062019.19636-3-matt@traverse.com.au>
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="===============0235944955369385512=="
List-Id: <development.lists.ipfire.org>
--===============0235944955369385512==
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable
Reviewed-by: Michael Tremer <michael.tremer(a)ipfire.org>
> On 3 Oct 2022, at 07:20, Mathew McBride <matt(a)traverse.com.au> wrote:
>=20
> These two patches are needed to support SFP's on NXP DPAA2 platforms
> (e.g Traverse Ten64).
>=20
> The deadlock issue patch was submitted upstream a while ago and
> rejected, however I am not aware of any better solutions at present.
>=20
> The 10G mode additions are part of mainline since 5.16.
>=20
> These two .patches were sourced from our patchset over here:
> https://gitlab.com/traversetech/traverse-kernel-patches/-/tree/lts-5-15/pat=
ches
>=20
> Signed-off-by: Mathew McBride <matt(a)traverse.com.au>
> ---
> lfs/linux | 3 +
> ...rm64-dpaa2-add-support-for-10g-modes.patch | 39 +++++++++
> ...inux-5.15-arm64-dpaa2-fix-lock-issue.patch | 81 +++++++++++++++++++
> 3 files changed, 123 insertions(+)
> create mode 100644 src/patches/linux/linux-5-15-arm64-dpaa2-add-support-for=
-10g-modes.patch
> create mode 100644 src/patches/linux/linux-5.15-arm64-dpaa2-fix-lock-issue.=
patch
>=20
> diff --git a/lfs/linux b/lfs/linux
> index a1b32cc25..3a7dbc91d 100644
> --- a/lfs/linux
> +++ b/lfs/linux
> @@ -153,6 +153,9 @@ endif
> ifeq "$(BUILD_ARCH)" "aarch64"
> # Apply Arm-multiarch kernel patches.
> cd $(DIR_APP) && xzcat $(DIR_DL)/arm-multi-patches-$(ARM_PATCHES).patch.xz=
| patch -Np1
> + # Apply NXP DPAA2 specific patches
> + cd $(DIR_APP) && patch -Np1 < $(DIR_SRC)/src/patches/linux/linux-5.15-arm=
64-dpaa2-fix-lock-issue.patch
> + cd $(DIR_APP) && patch -Np1 < $(DIR_SRC)/src/patches/linux/linux-5-15-arm=
64-dpaa2-add-support-for-10g-modes.patch
> endif
> cd $(DIR_APP) && patch -Np1 < $(DIR_SRC)/src/patches/linux/linux-3.14.79-a=
mba-fix.patch
>=20
> diff --git a/src/patches/linux/linux-5-15-arm64-dpaa2-add-support-for-10g-m=
odes.patch b/src/patches/linux/linux-5-15-arm64-dpaa2-add-support-for-10g-mod=
es.patch
> new file mode 100644
> index 000000000..ef8d459b7
> --- /dev/null
> +++ b/src/patches/linux/linux-5-15-arm64-dpaa2-add-support-for-10g-modes.pa=
tch
> @@ -0,0 +1,39 @@
> +From c314138bd045e050432158ab021160de3ba51c5e Mon Sep 17 00:00:00 2001
> +From: Russell King <rmk+kernel(a)armlinux.org.uk>
> +Date: Thu, 30 Jan 2020 22:42:38 +0000
> +Subject: [PATCH 2/4] net: dpaa2-mac: add support for more 10G modes
> +
> +Phylink documentation says:
> + * Note that the PHY may be able to transform from one connection
> + * technology to another, so, eg, don't clear 1000BaseX just
> + * because the MAC is unable to BaseX mode. This is more about
> + * clearing unsupported speeds and duplex settings. The port modes
> + * should not be cleared; phylink_set_port_modes() will help with this.
> +
> +So add the missing 10G modes.
> +
> +Signed-off-by: Russell King <rmk+kernel(a)armlinux.org.uk>
> +---
> + drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c | 6 ++++++
> + 1 file changed, 6 insertions(+)
> +
> +diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c b/drivers/ne=
t/ethernet/freescale/dpaa2/dpaa2-mac.c
> +index 8fe32ed4f6dc..3be849cee47b 100644
> +--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c
> ++++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c
> +@@ -140,6 +140,12 @@ static void dpaa2_mac_validate(struct phylink_config =
*config,
> + case PHY_INTERFACE_MODE_10GBASER:
> + case PHY_INTERFACE_MODE_USXGMII:
> + phylink_set(mask, 10000baseT_Full);
> ++ phylink_set(mask, 10000baseKR_Full);
> ++ phylink_set(mask, 10000baseCR_Full);
> ++ phylink_set(mask, 10000baseSR_Full);
> ++ phylink_set(mask, 10000baseLR_Full);
> ++ phylink_set(mask, 10000baseLRM_Full);
> ++ phylink_set(mask, 10000baseER_Full);
> + if (state->interface =3D=3D PHY_INTERFACE_MODE_10GBASER)
> + break;
> + phylink_set(mask, 5000baseT_Full);
> +--=20
> +2.30.1
> +
> diff --git a/src/patches/linux/linux-5.15-arm64-dpaa2-fix-lock-issue.patch =
b/src/patches/linux/linux-5.15-arm64-dpaa2-fix-lock-issue.patch
> new file mode 100644
> index 000000000..587821bac
> --- /dev/null
> +++ b/src/patches/linux/linux-5.15-arm64-dpaa2-fix-lock-issue.patch
> @@ -0,0 +1,81 @@
> +From 3a39dbe0c0c41f8dba5246ce6e2c5c4bcd6ba661 Mon Sep 17 00:00:00 2001
> +From: Ioana Ciornei <ioana.ciornei(a)nxp.com>
> +Date: Thu, 21 Nov 2019 21:15:25 +0200
> +Subject: [PATCH 1/4] dpaa2-eth: do not hold rtnl_lock on phylink_create() =
or
> + _destroy()
> +
> +The rtnl_lock should not be held when calling phylink_create() or
> +phylink_destroy() since it leads to the deadlock listed below:
> +
> +[ 18.656576] rtnl_lock+0x18/0x20
> +[ 18.659798] sfp_bus_add_upstream+0x28/0x90
> +[ 18.663974] phylink_create+0x2cc/0x828
> +[ 18.667803] dpaa2_mac_connect+0x14c/0x2a8
> +[ 18.671890] dpaa2_eth_connect_mac+0x94/0xd8
> +
> +Fix this by moving the _lock() and _unlock() calls just outside of
> +phylink_of_phy_connect() and phylink_disconnect_phy().
> +
> +Fixes: 719479230893 ("dpaa2-eth: add MAC/PHY support through phylink")
> +Reported-by: Russell King <linux(a)armlinux.org.uk>
> +Signed-off-by: Ioana Ciornei <ioana.ciornei(a)nxp.com>
> +Signed-off-by: Russell King <rmk+kernel(a)armlinux.org.uk>
> +---
> + drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c | 4 ----
> + drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c | 4 ++++
> + 2 files changed, 4 insertions(+), 4 deletions(-)
> +
> +diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c b/drivers/ne=
t/ethernet/freescale/dpaa2/dpaa2-eth.c
> +index 8b7a29e1e221..20e65053f036 100644
> +--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
> ++++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c
> +@@ -4214,12 +4214,10 @@ static irqreturn_t dpni_irq0_handler_thread(int ir=
q_num, void *arg)
> + dpaa2_eth_set_mac_addr(netdev_priv(net_dev));
> + dpaa2_eth_update_tx_fqids(priv);
> +=20
> +- rtnl_lock();
> + if (dpaa2_eth_has_mac(priv))
> + dpaa2_eth_disconnect_mac(priv);
> + else
> + dpaa2_eth_connect_mac(priv);
> +- rtnl_unlock();
> + }
> +=20
> + return IRQ_HANDLED;
> +@@ -4513,9 +4511,7 @@ static int dpaa2_eth_remove(struct fsl_mc_device *ls=
_dev)
> + #endif
> +=20
> + unregister_netdev(net_dev);
> +- rtnl_lock();
> + dpaa2_eth_disconnect_mac(priv);
> +- rtnl_unlock();
> +=20
> + dpaa2_eth_dl_port_del(priv);
> + dpaa2_eth_dl_traps_unregister(priv);
> +diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c b/drivers/ne=
t/ethernet/freescale/dpaa2/dpaa2-mac.c
> +index ae6d382d8735..8fe32ed4f6dc 100644
> +--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c
> ++++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-mac.c
> +@@ -351,7 +351,9 @@ int dpaa2_mac_connect(struct dpaa2_mac *mac)
> + if (mac->pcs)
> + phylink_set_pcs(mac->phylink, &mac->pcs->pcs);
> +=20
> ++ rtnl_lock();
> + err =3D phylink_fwnode_phy_connect(mac->phylink, dpmac_node, 0);
> ++ rtnl_unlock();
> + if (err) {
> + netdev_err(net_dev, "phylink_fwnode_phy_connect() =3D %d\n", err);
> + goto err_phylink_destroy;
> +@@ -372,7 +374,9 @@ void dpaa2_mac_disconnect(struct dpaa2_mac *mac)
> + if (!mac->phylink)
> + return;
> +=20
> ++ rtnl_lock();
> + phylink_disconnect_phy(mac->phylink);
> ++ rtnl_unlock();
> + phylink_destroy(mac->phylink);
> + dpaa2_pcs_destroy(mac);
> + }
> +--=20
> +2.30.1
> +
> --=20
> 2.30.1
>=20
--===============0235944955369385512==--