* [BUG/PATCH v3] net: mvpp2: always set port pcs ops
@ 2022-02-14 23:18 Jeremy Linton
2022-02-15 8:38 ` Marcin Wojtas
2022-02-18 23:49 ` Marcin Wojtas
0 siblings, 2 replies; 8+ messages in thread
From: Jeremy Linton @ 2022-02-14 23:18 UTC (permalink / raw)
To: netdev; +Cc: mw, linux, davem, kuba, linux-kernel, Jeremy Linton, Russell King
Booting a MACCHIATObin with 5.17, the system OOPs with
a null pointer deref when the network is started. This
is caused by the pcs->ops structure being null in
mcpp2_acpi_start() when it tries to call pcs_config().
Hoisting the code which sets pcs_gmac.ops and pcs_xlg.ops,
assuring they are always set, fixes the problem.
The OOPs looks like:
[ 18.687760] Unable to handle kernel access to user memory outside uaccess routines at virtual address 0000000000000010
[ 18.698561] Mem abort info:
[ 18.698564] ESR = 0x96000004
[ 18.698567] EC = 0x25: DABT (current EL), IL = 32 bits
[ 18.709821] SET = 0, FnV = 0
[ 18.714292] EA = 0, S1PTW = 0
[ 18.718833] FSC = 0x04: level 0 translation fault
[ 18.725126] Data abort info:
[ 18.729408] ISV = 0, ISS = 0x00000004
[ 18.734655] CM = 0, WnR = 0
[ 18.738933] user pgtable: 4k pages, 48-bit VAs, pgdp=0000000111bbf000
[ 18.745409] [0000000000000010] pgd=0000000000000000, p4d=0000000000000000
[ 18.752235] Internal error: Oops: 96000004 [#1] SMP
[ 18.757134] Modules linked in: rfkill ip_set nf_tables nfnetlink qrtr sunrpc vfat fat omap_rng fuse zram xfs crct10dif_ce mvpp2 ghash_ce sbsa_gwdt phylink xhci_plat_hcd ahci_plam
[ 18.773481] CPU: 0 PID: 681 Comm: NetworkManager Not tainted 5.17.0-0.rc3.89.fc36.aarch64 #1
[ 18.781954] Hardware name: Marvell Armada 7k/8k Family Board /Armada 7k/8k Family Board , BIOS EDK II Jun 4 2019
[ 18.795222] pstate: 60400005 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[ 18.802213] pc : mvpp2_start_dev+0x2b0/0x300 [mvpp2]
[ 18.807208] lr : mvpp2_start_dev+0x298/0x300 [mvpp2]
[ 18.812197] sp : ffff80000b4732c0
[ 18.815522] x29: ffff80000b4732c0 x28: 0000000000000000 x27: ffffccab38ae57f8
[ 18.822689] x26: ffff6eeb03065a10 x25: ffff80000b473a30 x24: ffff80000b4735b8
[ 18.829855] x23: 0000000000000000 x22: 00000000000001e0 x21: ffff6eeb07b6ab68
[ 18.837021] x20: ffff6eeb07b6ab30 x19: ffff6eeb07b6a9c0 x18: 0000000000000014
[ 18.844187] x17: 00000000f6232bfe x16: ffffccab899b1dc0 x15: 000000006a30f9fa
[ 18.851353] x14: 000000003b77bd50 x13: 000006dc896f0e8e x12: 001bbbfccfd0d3a2
[ 18.858519] x11: 0000000000001528 x10: 0000000000001548 x9 : ffffccab38ad0fb0
[ 18.865685] x8 : ffff80000b473330 x7 : 0000000000000000 x6 : 0000000000000000
[ 18.872851] x5 : 0000000000000000 x4 : 0000000000000000 x3 : ffff80000b4732f8
[ 18.880017] x2 : 000000000000001a x1 : 0000000000000002 x0 : ffff6eeb07b6ab68
[ 18.887183] Call trace:
[ 18.889637] mvpp2_start_dev+0x2b0/0x300 [mvpp2]
[ 18.894279] mvpp2_open+0x134/0x2b4 [mvpp2]
[ 18.898483] __dev_open+0x128/0x1e4
[ 18.901988] __dev_change_flags+0x17c/0x1d0
[ 18.906187] dev_change_flags+0x30/0x70
[ 18.910038] do_setlink+0x278/0xa7c
[ 18.913540] __rtnl_newlink+0x44c/0x7d0
[ 18.917391] rtnl_newlink+0x5c/0x8c
[ 18.920892] rtnetlink_rcv_msg+0x254/0x314
[ 18.925006] netlink_rcv_skb+0x48/0x10c
[ 18.928858] rtnetlink_rcv+0x24/0x30
[ 18.932449] netlink_unicast+0x290/0x2f4
[ 18.936386] netlink_sendmsg+0x1d0/0x41c
[ 18.940323] sock_sendmsg+0x60/0x70
[ 18.943825] ____sys_sendmsg+0x248/0x260
[ 18.947762] ___sys_sendmsg+0x74/0xa0
[ 18.951438] __sys_sendmsg+0x64/0xcc
[ 18.955027] __arm64_sys_sendmsg+0x30/0x40
[ 18.959140] invoke_syscall+0x50/0x120
[ 18.962906] el0_svc_common.constprop.0+0x4c/0xf4
[ 18.967629] do_el0_svc+0x30/0x9c
[ 18.970958] el0_svc+0x28/0xb0
[ 18.974025] el0t_64_sync_handler+0x10c/0x140
[ 18.978400] el0t_64_sync+0x1a4/0x1a8
[ 18.982078] Code: 52800004 b9416262 aa1503e0 52800041 (f94008a5)
[ 18.988196] ---[ end trace 0000000000000000 ]---
Fixes: cff056322372 ("net: mvpp2: use .mac_select_pcs() interface")
Suggested-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Signed-off-by: Jeremy Linton <jeremy.linton@arm.com>
---
v1->v2: Apply Russell's fix
v2->v3: Fix Russell's name
drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
index 7cdbf8b8bbf6..1a835b48791b 100644
--- a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
+++ b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
@@ -6870,6 +6870,9 @@ static int mvpp2_port_probe(struct platform_device *pdev,
dev->max_mtu = MVPP2_BM_JUMBO_PKT_SIZE;
dev->dev.of_node = port_node;
+ port->pcs_gmac.ops = &mvpp2_phylink_gmac_pcs_ops;
+ port->pcs_xlg.ops = &mvpp2_phylink_xlg_pcs_ops;
+
if (!mvpp2_use_acpi_compat_mode(port_fwnode)) {
port->phylink_config.dev = &dev->dev;
port->phylink_config.type = PHYLINK_NETDEV;
@@ -6940,9 +6943,6 @@ static int mvpp2_port_probe(struct platform_device *pdev,
port->phylink_config.supported_interfaces);
}
- port->pcs_gmac.ops = &mvpp2_phylink_gmac_pcs_ops;
- port->pcs_xlg.ops = &mvpp2_phylink_xlg_pcs_ops;
-
phylink = phylink_create(&port->phylink_config, port_fwnode,
phy_mode, &mvpp2_phylink_ops);
if (IS_ERR(phylink)) {
--
2.34.1
^ permalink raw reply related [flat|nested] 8+ messages in thread
* Re: [BUG/PATCH v3] net: mvpp2: always set port pcs ops
2022-02-14 23:18 [BUG/PATCH v3] net: mvpp2: always set port pcs ops Jeremy Linton
@ 2022-02-15 8:38 ` Marcin Wojtas
2022-02-15 14:36 ` Jeremy Linton
2022-02-18 23:49 ` Marcin Wojtas
1 sibling, 1 reply; 8+ messages in thread
From: Marcin Wojtas @ 2022-02-15 8:38 UTC (permalink / raw)
To: Jeremy Linton
Cc: netdev, Russell King - ARM Linux, David S. Miller,
Jakub Kicinski, Linux Kernel Mailing List, Russell King
Hi Jeremy,
wt., 15 lut 2022 o 00:18 Jeremy Linton <jeremy.linton@arm.com> napisał(a):
>
> Booting a MACCHIATObin with 5.17, the system OOPs with
> a null pointer deref when the network is started. This
> is caused by the pcs->ops structure being null in
> mcpp2_acpi_start() when it tries to call pcs_config().
>
> Hoisting the code which sets pcs_gmac.ops and pcs_xlg.ops,
> assuring they are always set, fixes the problem.
>
> The OOPs looks like:
> [ 18.687760] Unable to handle kernel access to user memory outside uaccess routines at virtual address 0000000000000010
> [ 18.698561] Mem abort info:
> [ 18.698564] ESR = 0x96000004
> [ 18.698567] EC = 0x25: DABT (current EL), IL = 32 bits
> [ 18.709821] SET = 0, FnV = 0
> [ 18.714292] EA = 0, S1PTW = 0
> [ 18.718833] FSC = 0x04: level 0 translation fault
> [ 18.725126] Data abort info:
> [ 18.729408] ISV = 0, ISS = 0x00000004
> [ 18.734655] CM = 0, WnR = 0
> [ 18.738933] user pgtable: 4k pages, 48-bit VAs, pgdp=0000000111bbf000
> [ 18.745409] [0000000000000010] pgd=0000000000000000, p4d=0000000000000000
> [ 18.752235] Internal error: Oops: 96000004 [#1] SMP
> [ 18.757134] Modules linked in: rfkill ip_set nf_tables nfnetlink qrtr sunrpc vfat fat omap_rng fuse zram xfs crct10dif_ce mvpp2 ghash_ce sbsa_gwdt phylink xhci_plat_hcd ahci_plam
> [ 18.773481] CPU: 0 PID: 681 Comm: NetworkManager Not tainted 5.17.0-0.rc3.89.fc36.aarch64 #1
> [ 18.781954] Hardware name: Marvell Armada 7k/8k Family Board /Armada 7k/8k Family Board , BIOS EDK II Jun 4 2019
> [ 18.795222] pstate: 60400005 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
> [ 18.802213] pc : mvpp2_start_dev+0x2b0/0x300 [mvpp2]
> [ 18.807208] lr : mvpp2_start_dev+0x298/0x300 [mvpp2]
> [ 18.812197] sp : ffff80000b4732c0
> [ 18.815522] x29: ffff80000b4732c0 x28: 0000000000000000 x27: ffffccab38ae57f8
> [ 18.822689] x26: ffff6eeb03065a10 x25: ffff80000b473a30 x24: ffff80000b4735b8
> [ 18.829855] x23: 0000000000000000 x22: 00000000000001e0 x21: ffff6eeb07b6ab68
> [ 18.837021] x20: ffff6eeb07b6ab30 x19: ffff6eeb07b6a9c0 x18: 0000000000000014
> [ 18.844187] x17: 00000000f6232bfe x16: ffffccab899b1dc0 x15: 000000006a30f9fa
> [ 18.851353] x14: 000000003b77bd50 x13: 000006dc896f0e8e x12: 001bbbfccfd0d3a2
> [ 18.858519] x11: 0000000000001528 x10: 0000000000001548 x9 : ffffccab38ad0fb0
> [ 18.865685] x8 : ffff80000b473330 x7 : 0000000000000000 x6 : 0000000000000000
> [ 18.872851] x5 : 0000000000000000 x4 : 0000000000000000 x3 : ffff80000b4732f8
> [ 18.880017] x2 : 000000000000001a x1 : 0000000000000002 x0 : ffff6eeb07b6ab68
> [ 18.887183] Call trace:
> [ 18.889637] mvpp2_start_dev+0x2b0/0x300 [mvpp2]
> [ 18.894279] mvpp2_open+0x134/0x2b4 [mvpp2]
> [ 18.898483] __dev_open+0x128/0x1e4
> [ 18.901988] __dev_change_flags+0x17c/0x1d0
> [ 18.906187] dev_change_flags+0x30/0x70
> [ 18.910038] do_setlink+0x278/0xa7c
> [ 18.913540] __rtnl_newlink+0x44c/0x7d0
> [ 18.917391] rtnl_newlink+0x5c/0x8c
> [ 18.920892] rtnetlink_rcv_msg+0x254/0x314
> [ 18.925006] netlink_rcv_skb+0x48/0x10c
> [ 18.928858] rtnetlink_rcv+0x24/0x30
> [ 18.932449] netlink_unicast+0x290/0x2f4
> [ 18.936386] netlink_sendmsg+0x1d0/0x41c
> [ 18.940323] sock_sendmsg+0x60/0x70
> [ 18.943825] ____sys_sendmsg+0x248/0x260
> [ 18.947762] ___sys_sendmsg+0x74/0xa0
> [ 18.951438] __sys_sendmsg+0x64/0xcc
> [ 18.955027] __arm64_sys_sendmsg+0x30/0x40
> [ 18.959140] invoke_syscall+0x50/0x120
> [ 18.962906] el0_svc_common.constprop.0+0x4c/0xf4
> [ 18.967629] do_el0_svc+0x30/0x9c
> [ 18.970958] el0_svc+0x28/0xb0
> [ 18.974025] el0t_64_sync_handler+0x10c/0x140
> [ 18.978400] el0t_64_sync+0x1a4/0x1a8
> [ 18.982078] Code: 52800004 b9416262 aa1503e0 52800041 (f94008a5)
> [ 18.988196] ---[ end trace 0000000000000000 ]---
>
> Fixes: cff056322372 ("net: mvpp2: use .mac_select_pcs() interface")
> Suggested-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
> Signed-off-by: Jeremy Linton <jeremy.linton@arm.com>
> ---
> v1->v2: Apply Russell's fix
> v2->v3: Fix Russell's name
>
> drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c | 6 +++---
> 1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
> index 7cdbf8b8bbf6..1a835b48791b 100644
> --- a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
> +++ b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
> @@ -6870,6 +6870,9 @@ static int mvpp2_port_probe(struct platform_device *pdev,
> dev->max_mtu = MVPP2_BM_JUMBO_PKT_SIZE;
> dev->dev.of_node = port_node;
>
> + port->pcs_gmac.ops = &mvpp2_phylink_gmac_pcs_ops;
> + port->pcs_xlg.ops = &mvpp2_phylink_xlg_pcs_ops;
> +
> if (!mvpp2_use_acpi_compat_mode(port_fwnode)) {
> port->phylink_config.dev = &dev->dev;
> port->phylink_config.type = PHYLINK_NETDEV;
> @@ -6940,9 +6943,6 @@ static int mvpp2_port_probe(struct platform_device *pdev,
> port->phylink_config.supported_interfaces);
> }
>
> - port->pcs_gmac.ops = &mvpp2_phylink_gmac_pcs_ops;
> - port->pcs_xlg.ops = &mvpp2_phylink_xlg_pcs_ops;
> -
> phylink = phylink_create(&port->phylink_config, port_fwnode,
> phy_mode, &mvpp2_phylink_ops);
> if (IS_ERR(phylink)) {
> --
> 2.34.1
>
I'd like to test the patch - what EDK2 version are you using?
Best regards,
Marcin
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [BUG/PATCH v3] net: mvpp2: always set port pcs ops
2022-02-15 8:38 ` Marcin Wojtas
@ 2022-02-15 14:36 ` Jeremy Linton
2022-02-15 16:44 ` Marcin Wojtas
0 siblings, 1 reply; 8+ messages in thread
From: Jeremy Linton @ 2022-02-15 14:36 UTC (permalink / raw)
To: Marcin Wojtas
Cc: netdev, Russell King - ARM Linux, David S. Miller,
Jakub Kicinski, Linux Kernel Mailing List, Russell King
Hi,
On 2/15/22 02:38, Marcin Wojtas wrote:
> Hi Jeremy,
>
>
> wt., 15 lut 2022 o 00:18 Jeremy Linton <jeremy.linton@arm.com> napisał(a):
>>
>> Booting a MACCHIATObin with 5.17, the system OOPs with
>> a null pointer deref when the network is started. This
>> is caused by the pcs->ops structure being null in
>> mcpp2_acpi_start() when it tries to call pcs_config().
>>
>> Hoisting the code which sets pcs_gmac.ops and pcs_xlg.ops,
>> assuring they are always set, fixes the problem.
>>
>> The OOPs looks like:
>> [ 18.687760] Unable to handle kernel access to user memory outside uaccess routines at virtual address 0000000000000010
>> [ 18.698561] Mem abort info:
>> [ 18.698564] ESR = 0x96000004
>> [ 18.698567] EC = 0x25: DABT (current EL), IL = 32 bits
>> [ 18.709821] SET = 0, FnV = 0
>> [ 18.714292] EA = 0, S1PTW = 0
>> [ 18.718833] FSC = 0x04: level 0 translation fault
>> [ 18.725126] Data abort info:
>> [ 18.729408] ISV = 0, ISS = 0x00000004
>> [ 18.734655] CM = 0, WnR = 0
>> [ 18.738933] user pgtable: 4k pages, 48-bit VAs, pgdp=0000000111bbf000
>> [ 18.745409] [0000000000000010] pgd=0000000000000000, p4d=0000000000000000
>> [ 18.752235] Internal error: Oops: 96000004 [#1] SMP
>> [ 18.757134] Modules linked in: rfkill ip_set nf_tables nfnetlink qrtr sunrpc vfat fat omap_rng fuse zram xfs crct10dif_ce mvpp2 ghash_ce sbsa_gwdt phylink xhci_plat_hcd ahci_plam
>> [ 18.773481] CPU: 0 PID: 681 Comm: NetworkManager Not tainted 5.17.0-0.rc3.89.fc36.aarch64 #1
>> [ 18.781954] Hardware name: Marvell Armada 7k/8k Family Board /Armada 7k/8k Family Board , BIOS EDK II Jun 4 2019
>> [ 18.795222] pstate: 60400005 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
>> [ 18.802213] pc : mvpp2_start_dev+0x2b0/0x300 [mvpp2]
>> [ 18.807208] lr : mvpp2_start_dev+0x298/0x300 [mvpp2]
>> [ 18.812197] sp : ffff80000b4732c0
>> [ 18.815522] x29: ffff80000b4732c0 x28: 0000000000000000 x27: ffffccab38ae57f8
>> [ 18.822689] x26: ffff6eeb03065a10 x25: ffff80000b473a30 x24: ffff80000b4735b8
>> [ 18.829855] x23: 0000000000000000 x22: 00000000000001e0 x21: ffff6eeb07b6ab68
>> [ 18.837021] x20: ffff6eeb07b6ab30 x19: ffff6eeb07b6a9c0 x18: 0000000000000014
>> [ 18.844187] x17: 00000000f6232bfe x16: ffffccab899b1dc0 x15: 000000006a30f9fa
>> [ 18.851353] x14: 000000003b77bd50 x13: 000006dc896f0e8e x12: 001bbbfccfd0d3a2
>> [ 18.858519] x11: 0000000000001528 x10: 0000000000001548 x9 : ffffccab38ad0fb0
>> [ 18.865685] x8 : ffff80000b473330 x7 : 0000000000000000 x6 : 0000000000000000
>> [ 18.872851] x5 : 0000000000000000 x4 : 0000000000000000 x3 : ffff80000b4732f8
>> [ 18.880017] x2 : 000000000000001a x1 : 0000000000000002 x0 : ffff6eeb07b6ab68
>> [ 18.887183] Call trace:
>> [ 18.889637] mvpp2_start_dev+0x2b0/0x300 [mvpp2]
>> [ 18.894279] mvpp2_open+0x134/0x2b4 [mvpp2]
>> [ 18.898483] __dev_open+0x128/0x1e4
>> [ 18.901988] __dev_change_flags+0x17c/0x1d0
>> [ 18.906187] dev_change_flags+0x30/0x70
>> [ 18.910038] do_setlink+0x278/0xa7c
>> [ 18.913540] __rtnl_newlink+0x44c/0x7d0
>> [ 18.917391] rtnl_newlink+0x5c/0x8c
>> [ 18.920892] rtnetlink_rcv_msg+0x254/0x314
>> [ 18.925006] netlink_rcv_skb+0x48/0x10c
>> [ 18.928858] rtnetlink_rcv+0x24/0x30
>> [ 18.932449] netlink_unicast+0x290/0x2f4
>> [ 18.936386] netlink_sendmsg+0x1d0/0x41c
>> [ 18.940323] sock_sendmsg+0x60/0x70
>> [ 18.943825] ____sys_sendmsg+0x248/0x260
>> [ 18.947762] ___sys_sendmsg+0x74/0xa0
>> [ 18.951438] __sys_sendmsg+0x64/0xcc
>> [ 18.955027] __arm64_sys_sendmsg+0x30/0x40
>> [ 18.959140] invoke_syscall+0x50/0x120
>> [ 18.962906] el0_svc_common.constprop.0+0x4c/0xf4
>> [ 18.967629] do_el0_svc+0x30/0x9c
>> [ 18.970958] el0_svc+0x28/0xb0
>> [ 18.974025] el0t_64_sync_handler+0x10c/0x140
>> [ 18.978400] el0t_64_sync+0x1a4/0x1a8
>> [ 18.982078] Code: 52800004 b9416262 aa1503e0 52800041 (f94008a5)
>> [ 18.988196] ---[ end trace 0000000000000000 ]---
>>
>> Fixes: cff056322372 ("net: mvpp2: use .mac_select_pcs() interface")
>> Suggested-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
>> Signed-off-by: Jeremy Linton <jeremy.linton@arm.com>
>> ---
>> v1->v2: Apply Russell's fix
>> v2->v3: Fix Russell's name
>>
>> drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c | 6 +++---
>> 1 file changed, 3 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
>> index 7cdbf8b8bbf6..1a835b48791b 100644
>> --- a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
>> +++ b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
>> @@ -6870,6 +6870,9 @@ static int mvpp2_port_probe(struct platform_device *pdev,
>> dev->max_mtu = MVPP2_BM_JUMBO_PKT_SIZE;
>> dev->dev.of_node = port_node;
>>
>> + port->pcs_gmac.ops = &mvpp2_phylink_gmac_pcs_ops;
>> + port->pcs_xlg.ops = &mvpp2_phylink_xlg_pcs_ops;
>> +
>> if (!mvpp2_use_acpi_compat_mode(port_fwnode)) {
>> port->phylink_config.dev = &dev->dev;
>> port->phylink_config.type = PHYLINK_NETDEV;
>> @@ -6940,9 +6943,6 @@ static int mvpp2_port_probe(struct platform_device *pdev,
>> port->phylink_config.supported_interfaces);
>> }
>>
>> - port->pcs_gmac.ops = &mvpp2_phylink_gmac_pcs_ops;
>> - port->pcs_xlg.ops = &mvpp2_phylink_xlg_pcs_ops;
>> -
>> phylink = phylink_create(&port->phylink_config, port_fwnode,
>> phy_mode, &mvpp2_phylink_ops);
>> if (IS_ERR(phylink)) {
>> --
>> 2.34.1
>>
>
> I'd like to test the patch - what EDK2 version are you using?
I don't have access to the machine at the moment (maybe in a couple days
again) but it was running a build from late 2019 IIRC. So, definitely
not a bleeding edge version for sure.
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [BUG/PATCH v3] net: mvpp2: always set port pcs ops
2022-02-15 14:36 ` Jeremy Linton
@ 2022-02-15 16:44 ` Marcin Wojtas
2022-02-17 23:57 ` Jakub Kicinski
0 siblings, 1 reply; 8+ messages in thread
From: Marcin Wojtas @ 2022-02-15 16:44 UTC (permalink / raw)
To: Jeremy Linton
Cc: netdev, Russell King - ARM Linux, David S. Miller,
Jakub Kicinski, Linux Kernel Mailing List, Russell King
Hi,
wt., 15 lut 2022 o 15:37 Jeremy Linton <jeremy.linton@arm.com> napisał(a):
>
> Hi,
>
> On 2/15/22 02:38, Marcin Wojtas wrote:
> > Hi Jeremy,
> >
> >
> > wt., 15 lut 2022 o 00:18 Jeremy Linton <jeremy.linton@arm.com> napisał(a):
> >>
> >> Booting a MACCHIATObin with 5.17, the system OOPs with
> >> a null pointer deref when the network is started. This
> >> is caused by the pcs->ops structure being null in
> >> mcpp2_acpi_start() when it tries to call pcs_config().
> >>
> >> Hoisting the code which sets pcs_gmac.ops and pcs_xlg.ops,
> >> assuring they are always set, fixes the problem.
> >>
> >> The OOPs looks like:
> >> [ 18.687760] Unable to handle kernel access to user memory outside uaccess routines at virtual address 0000000000000010
> >> [ 18.698561] Mem abort info:
> >> [ 18.698564] ESR = 0x96000004
> >> [ 18.698567] EC = 0x25: DABT (current EL), IL = 32 bits
> >> [ 18.709821] SET = 0, FnV = 0
> >> [ 18.714292] EA = 0, S1PTW = 0
> >> [ 18.718833] FSC = 0x04: level 0 translation fault
> >> [ 18.725126] Data abort info:
> >> [ 18.729408] ISV = 0, ISS = 0x00000004
> >> [ 18.734655] CM = 0, WnR = 0
> >> [ 18.738933] user pgtable: 4k pages, 48-bit VAs, pgdp=0000000111bbf000
> >> [ 18.745409] [0000000000000010] pgd=0000000000000000, p4d=0000000000000000
> >> [ 18.752235] Internal error: Oops: 96000004 [#1] SMP
> >> [ 18.757134] Modules linked in: rfkill ip_set nf_tables nfnetlink qrtr sunrpc vfat fat omap_rng fuse zram xfs crct10dif_ce mvpp2 ghash_ce sbsa_gwdt phylink xhci_plat_hcd ahci_plam
> >> [ 18.773481] CPU: 0 PID: 681 Comm: NetworkManager Not tainted 5.17.0-0.rc3.89.fc36.aarch64 #1
> >> [ 18.781954] Hardware name: Marvell Armada 7k/8k Family Board /Armada 7k/8k Family Board , BIOS EDK II Jun 4 2019
> >> [ 18.795222] pstate: 60400005 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
> >> [ 18.802213] pc : mvpp2_start_dev+0x2b0/0x300 [mvpp2]
> >> [ 18.807208] lr : mvpp2_start_dev+0x298/0x300 [mvpp2]
> >> [ 18.812197] sp : ffff80000b4732c0
> >> [ 18.815522] x29: ffff80000b4732c0 x28: 0000000000000000 x27: ffffccab38ae57f8
> >> [ 18.822689] x26: ffff6eeb03065a10 x25: ffff80000b473a30 x24: ffff80000b4735b8
> >> [ 18.829855] x23: 0000000000000000 x22: 00000000000001e0 x21: ffff6eeb07b6ab68
> >> [ 18.837021] x20: ffff6eeb07b6ab30 x19: ffff6eeb07b6a9c0 x18: 0000000000000014
> >> [ 18.844187] x17: 00000000f6232bfe x16: ffffccab899b1dc0 x15: 000000006a30f9fa
> >> [ 18.851353] x14: 000000003b77bd50 x13: 000006dc896f0e8e x12: 001bbbfccfd0d3a2
> >> [ 18.858519] x11: 0000000000001528 x10: 0000000000001548 x9 : ffffccab38ad0fb0
> >> [ 18.865685] x8 : ffff80000b473330 x7 : 0000000000000000 x6 : 0000000000000000
> >> [ 18.872851] x5 : 0000000000000000 x4 : 0000000000000000 x3 : ffff80000b4732f8
> >> [ 18.880017] x2 : 000000000000001a x1 : 0000000000000002 x0 : ffff6eeb07b6ab68
> >> [ 18.887183] Call trace:
> >> [ 18.889637] mvpp2_start_dev+0x2b0/0x300 [mvpp2]
> >> [ 18.894279] mvpp2_open+0x134/0x2b4 [mvpp2]
> >> [ 18.898483] __dev_open+0x128/0x1e4
> >> [ 18.901988] __dev_change_flags+0x17c/0x1d0
> >> [ 18.906187] dev_change_flags+0x30/0x70
> >> [ 18.910038] do_setlink+0x278/0xa7c
> >> [ 18.913540] __rtnl_newlink+0x44c/0x7d0
> >> [ 18.917391] rtnl_newlink+0x5c/0x8c
> >> [ 18.920892] rtnetlink_rcv_msg+0x254/0x314
> >> [ 18.925006] netlink_rcv_skb+0x48/0x10c
> >> [ 18.928858] rtnetlink_rcv+0x24/0x30
> >> [ 18.932449] netlink_unicast+0x290/0x2f4
> >> [ 18.936386] netlink_sendmsg+0x1d0/0x41c
> >> [ 18.940323] sock_sendmsg+0x60/0x70
> >> [ 18.943825] ____sys_sendmsg+0x248/0x260
> >> [ 18.947762] ___sys_sendmsg+0x74/0xa0
> >> [ 18.951438] __sys_sendmsg+0x64/0xcc
> >> [ 18.955027] __arm64_sys_sendmsg+0x30/0x40
> >> [ 18.959140] invoke_syscall+0x50/0x120
> >> [ 18.962906] el0_svc_common.constprop.0+0x4c/0xf4
> >> [ 18.967629] do_el0_svc+0x30/0x9c
> >> [ 18.970958] el0_svc+0x28/0xb0
> >> [ 18.974025] el0t_64_sync_handler+0x10c/0x140
> >> [ 18.978400] el0t_64_sync+0x1a4/0x1a8
> >> [ 18.982078] Code: 52800004 b9416262 aa1503e0 52800041 (f94008a5)
> >> [ 18.988196] ---[ end trace 0000000000000000 ]---
> >>
> >> Fixes: cff056322372 ("net: mvpp2: use .mac_select_pcs() interface")
> >> Suggested-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
> >> Signed-off-by: Jeremy Linton <jeremy.linton@arm.com>
> >> ---
> >> v1->v2: Apply Russell's fix
> >> v2->v3: Fix Russell's name
> >>
> >> drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c | 6 +++---
> >> 1 file changed, 3 insertions(+), 3 deletions(-)
> >>
> >> diff --git a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
> >> index 7cdbf8b8bbf6..1a835b48791b 100644
> >> --- a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
> >> +++ b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
> >> @@ -6870,6 +6870,9 @@ static int mvpp2_port_probe(struct platform_device *pdev,
> >> dev->max_mtu = MVPP2_BM_JUMBO_PKT_SIZE;
> >> dev->dev.of_node = port_node;
> >>
> >> + port->pcs_gmac.ops = &mvpp2_phylink_gmac_pcs_ops;
> >> + port->pcs_xlg.ops = &mvpp2_phylink_xlg_pcs_ops;
> >> +
> >> if (!mvpp2_use_acpi_compat_mode(port_fwnode)) {
> >> port->phylink_config.dev = &dev->dev;
> >> port->phylink_config.type = PHYLINK_NETDEV;
> >> @@ -6940,9 +6943,6 @@ static int mvpp2_port_probe(struct platform_device *pdev,
> >> port->phylink_config.supported_interfaces);
> >> }
> >>
> >> - port->pcs_gmac.ops = &mvpp2_phylink_gmac_pcs_ops;
> >> - port->pcs_xlg.ops = &mvpp2_phylink_xlg_pcs_ops;
> >> -
> >> phylink = phylink_create(&port->phylink_config, port_fwnode,
> >> phy_mode, &mvpp2_phylink_ops);
> >> if (IS_ERR(phylink)) {
> >> --
> >> 2.34.1
> >>
> >
> > I'd like to test the patch - what EDK2 version are you using?
>
> I don't have access to the machine at the moment (maybe in a couple days
> again) but it was running a build from late 2019 IIRC. So, definitely
> not a bleeding edge version for sure.
>
2019 is enough indicator that most likely there is no MDIO description
in ACPI. Let me test the patch, thanks.
Best regards,
Marcin
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [BUG/PATCH v3] net: mvpp2: always set port pcs ops
2022-02-15 16:44 ` Marcin Wojtas
@ 2022-02-17 23:57 ` Jakub Kicinski
2022-02-18 16:00 ` Marcin Wojtas
0 siblings, 1 reply; 8+ messages in thread
From: Jakub Kicinski @ 2022-02-17 23:57 UTC (permalink / raw)
To: Marcin Wojtas
Cc: Jeremy Linton, netdev, Russell King - ARM Linux, David S. Miller,
Russell King
On Tue, 15 Feb 2022 17:44:53 +0100 Marcin Wojtas wrote:
> > I don't have access to the machine at the moment (maybe in a couple days
> > again) but it was running a build from late 2019 IIRC. So, definitely
> > not a bleeding edge version for sure.
> >
>
> 2019 is enough indicator that most likely there is no MDIO description
> in ACPI. Let me test the patch, thanks.
Any luck with the testing? I wonder if we should keep the patch in or
drop it for now and ask for repost once tested..
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [BUG/PATCH v3] net: mvpp2: always set port pcs ops
2022-02-17 23:57 ` Jakub Kicinski
@ 2022-02-18 16:00 ` Marcin Wojtas
0 siblings, 0 replies; 8+ messages in thread
From: Marcin Wojtas @ 2022-02-18 16:00 UTC (permalink / raw)
To: Jakub Kicinski
Cc: Jeremy Linton, netdev, Russell King - ARM Linux, David S. Miller,
Russell King
Hi,
pt., 18 lut 2022 o 00:57 Jakub Kicinski <kuba@kernel.org> napisał(a):
>
> On Tue, 15 Feb 2022 17:44:53 +0100 Marcin Wojtas wrote:
> > > I don't have access to the machine at the moment (maybe in a couple days
> > > again) but it was running a build from late 2019 IIRC. So, definitely
> > > not a bleeding edge version for sure.
> > >
> >
> > 2019 is enough indicator that most likely there is no MDIO description
> > in ACPI. Let me test the patch, thanks.
>
> Any luck with the testing? I wonder if we should keep the patch in or
> drop it for now and ask for repost once tested..
I'll update today, sorry for the delay.
Marcin
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [BUG/PATCH v3] net: mvpp2: always set port pcs ops
2022-02-14 23:18 [BUG/PATCH v3] net: mvpp2: always set port pcs ops Jeremy Linton
2022-02-15 8:38 ` Marcin Wojtas
@ 2022-02-18 23:49 ` Marcin Wojtas
2022-02-19 4:44 ` Jakub Kicinski
1 sibling, 1 reply; 8+ messages in thread
From: Marcin Wojtas @ 2022-02-18 23:49 UTC (permalink / raw)
To: Jeremy Linton
Cc: netdev, Russell King - ARM Linux, David S. Miller,
Jakub Kicinski, Linux Kernel Mailing List, Russell King
wt., 15 lut 2022 o 00:18 Jeremy Linton <jeremy.linton@arm.com> napisał(a):
>
> Booting a MACCHIATObin with 5.17, the system OOPs with
> a null pointer deref when the network is started. This
> is caused by the pcs->ops structure being null in
> mcpp2_acpi_start() when it tries to call pcs_config().
>
> Hoisting the code which sets pcs_gmac.ops and pcs_xlg.ops,
> assuring they are always set, fixes the problem.
>
> The OOPs looks like:
> [ 18.687760] Unable to handle kernel access to user memory outside uaccess routines at virtual address 0000000000000010
> [ 18.698561] Mem abort info:
> [ 18.698564] ESR = 0x96000004
> [ 18.698567] EC = 0x25: DABT (current EL), IL = 32 bits
> [ 18.709821] SET = 0, FnV = 0
> [ 18.714292] EA = 0, S1PTW = 0
> [ 18.718833] FSC = 0x04: level 0 translation fault
> [ 18.725126] Data abort info:
> [ 18.729408] ISV = 0, ISS = 0x00000004
> [ 18.734655] CM = 0, WnR = 0
> [ 18.738933] user pgtable: 4k pages, 48-bit VAs, pgdp=0000000111bbf000
> [ 18.745409] [0000000000000010] pgd=0000000000000000, p4d=0000000000000000
> [ 18.752235] Internal error: Oops: 96000004 [#1] SMP
> [ 18.757134] Modules linked in: rfkill ip_set nf_tables nfnetlink qrtr sunrpc vfat fat omap_rng fuse zram xfs crct10dif_ce mvpp2 ghash_ce sbsa_gwdt phylink xhci_plat_hcd ahci_plam
> [ 18.773481] CPU: 0 PID: 681 Comm: NetworkManager Not tainted 5.17.0-0.rc3.89.fc36.aarch64 #1
> [ 18.781954] Hardware name: Marvell Armada 7k/8k Family Board /Armada 7k/8k Family Board , BIOS EDK II Jun 4 2019
> [ 18.795222] pstate: 60400005 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
> [ 18.802213] pc : mvpp2_start_dev+0x2b0/0x300 [mvpp2]
> [ 18.807208] lr : mvpp2_start_dev+0x298/0x300 [mvpp2]
> [ 18.812197] sp : ffff80000b4732c0
> [ 18.815522] x29: ffff80000b4732c0 x28: 0000000000000000 x27: ffffccab38ae57f8
> [ 18.822689] x26: ffff6eeb03065a10 x25: ffff80000b473a30 x24: ffff80000b4735b8
> [ 18.829855] x23: 0000000000000000 x22: 00000000000001e0 x21: ffff6eeb07b6ab68
> [ 18.837021] x20: ffff6eeb07b6ab30 x19: ffff6eeb07b6a9c0 x18: 0000000000000014
> [ 18.844187] x17: 00000000f6232bfe x16: ffffccab899b1dc0 x15: 000000006a30f9fa
> [ 18.851353] x14: 000000003b77bd50 x13: 000006dc896f0e8e x12: 001bbbfccfd0d3a2
> [ 18.858519] x11: 0000000000001528 x10: 0000000000001548 x9 : ffffccab38ad0fb0
> [ 18.865685] x8 : ffff80000b473330 x7 : 0000000000000000 x6 : 0000000000000000
> [ 18.872851] x5 : 0000000000000000 x4 : 0000000000000000 x3 : ffff80000b4732f8
> [ 18.880017] x2 : 000000000000001a x1 : 0000000000000002 x0 : ffff6eeb07b6ab68
> [ 18.887183] Call trace:
> [ 18.889637] mvpp2_start_dev+0x2b0/0x300 [mvpp2]
> [ 18.894279] mvpp2_open+0x134/0x2b4 [mvpp2]
> [ 18.898483] __dev_open+0x128/0x1e4
> [ 18.901988] __dev_change_flags+0x17c/0x1d0
> [ 18.906187] dev_change_flags+0x30/0x70
> [ 18.910038] do_setlink+0x278/0xa7c
> [ 18.913540] __rtnl_newlink+0x44c/0x7d0
> [ 18.917391] rtnl_newlink+0x5c/0x8c
> [ 18.920892] rtnetlink_rcv_msg+0x254/0x314
> [ 18.925006] netlink_rcv_skb+0x48/0x10c
> [ 18.928858] rtnetlink_rcv+0x24/0x30
> [ 18.932449] netlink_unicast+0x290/0x2f4
> [ 18.936386] netlink_sendmsg+0x1d0/0x41c
> [ 18.940323] sock_sendmsg+0x60/0x70
> [ 18.943825] ____sys_sendmsg+0x248/0x260
> [ 18.947762] ___sys_sendmsg+0x74/0xa0
> [ 18.951438] __sys_sendmsg+0x64/0xcc
> [ 18.955027] __arm64_sys_sendmsg+0x30/0x40
> [ 18.959140] invoke_syscall+0x50/0x120
> [ 18.962906] el0_svc_common.constprop.0+0x4c/0xf4
> [ 18.967629] do_el0_svc+0x30/0x9c
> [ 18.970958] el0_svc+0x28/0xb0
> [ 18.974025] el0t_64_sync_handler+0x10c/0x140
> [ 18.978400] el0t_64_sync+0x1a4/0x1a8
> [ 18.982078] Code: 52800004 b9416262 aa1503e0 52800041 (f94008a5)
> [ 18.988196] ---[ end trace 0000000000000000 ]---
>
> Fixes: cff056322372 ("net: mvpp2: use .mac_select_pcs() interface")
> Suggested-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
> Signed-off-by: Jeremy Linton <jeremy.linton@arm.com>
> ---
> v1->v2: Apply Russell's fix
> v2->v3: Fix Russell's name
>
> drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c | 6 +++---
> 1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
> index 7cdbf8b8bbf6..1a835b48791b 100644
> --- a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
> +++ b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
> @@ -6870,6 +6870,9 @@ static int mvpp2_port_probe(struct platform_device *pdev,
> dev->max_mtu = MVPP2_BM_JUMBO_PKT_SIZE;
> dev->dev.of_node = port_node;
>
> + port->pcs_gmac.ops = &mvpp2_phylink_gmac_pcs_ops;
> + port->pcs_xlg.ops = &mvpp2_phylink_xlg_pcs_ops;
> +
> if (!mvpp2_use_acpi_compat_mode(port_fwnode)) {
> port->phylink_config.dev = &dev->dev;
> port->phylink_config.type = PHYLINK_NETDEV;
> @@ -6940,9 +6943,6 @@ static int mvpp2_port_probe(struct platform_device *pdev,
> port->phylink_config.supported_interfaces);
> }
>
> - port->pcs_gmac.ops = &mvpp2_phylink_gmac_pcs_ops;
> - port->pcs_xlg.ops = &mvpp2_phylink_xlg_pcs_ops;
> -
> phylink = phylink_create(&port->phylink_config, port_fwnode,
> phy_mode, &mvpp2_phylink_ops);
> if (IS_ERR(phylink)) {
>
All works fine in my setup, thanks!
You can add my:
Reviewed-by: Marcin Wojtas <mw@semihalf.com>
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [BUG/PATCH v3] net: mvpp2: always set port pcs ops
2022-02-18 23:49 ` Marcin Wojtas
@ 2022-02-19 4:44 ` Jakub Kicinski
0 siblings, 0 replies; 8+ messages in thread
From: Jakub Kicinski @ 2022-02-19 4:44 UTC (permalink / raw)
To: Marcin Wojtas, Jeremy Linton
Cc: netdev, David S. Miller, Linux Kernel Mailing List, Russell King
On Sat, 19 Feb 2022 00:49:27 +0100 Marcin Wojtas wrote:
> wt., 15 lut 2022 o 00:18 Jeremy Linton <jeremy.linton@arm.com> napisał(a):
> > Booting a MACCHIATObin with 5.17, the system OOPs with
> > a null pointer deref when the network is started. This
> > is caused by the pcs->ops structure being null in
> > mcpp2_acpi_start() when it tries to call pcs_config().
> >
> > Hoisting the code which sets pcs_gmac.ops and pcs_xlg.ops,
> > assuring they are always set, fixes the problem.
> > Fixes: cff056322372 ("net: mvpp2: use .mac_select_pcs() interface")
> > Suggested-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
> > Signed-off-by: Jeremy Linton <jeremy.linton@arm.com>
> You can add my:
> Reviewed-by: Marcin Wojtas <mw@semihalf.com>
Perfect, thanks everyone!
It's now 5a2aba71cd26 ("net: mvpp2: always set port pcs ops") in net.
^ permalink raw reply [flat|nested] 8+ messages in thread
end of thread, other threads:[~2022-02-19 4:44 UTC | newest]
Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-02-14 23:18 [BUG/PATCH v3] net: mvpp2: always set port pcs ops Jeremy Linton
2022-02-15 8:38 ` Marcin Wojtas
2022-02-15 14:36 ` Jeremy Linton
2022-02-15 16:44 ` Marcin Wojtas
2022-02-17 23:57 ` Jakub Kicinski
2022-02-18 16:00 ` Marcin Wojtas
2022-02-18 23:49 ` Marcin Wojtas
2022-02-19 4:44 ` Jakub Kicinski
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).