[PATCH net] net: stmmac: Fix E2E delay mechanism

[PATCH net] net: stmmac: Fix E2E delay mechanism

[Kurt Kanzenbach]

When utilizing End to End delay mechanism, the following error messages show up:

|root@ehl1:~# ptp4l --tx_timestamp_timeout=50 -H -i eno2 -E -m
|ptp4l[950.573]: selected /dev/ptp3 as PTP clock
|ptp4l[950.586]: port 1: INITIALIZING to LISTENING on INIT_COMPLETE
|ptp4l[950.586]: port 0: INITIALIZING to LISTENING on INIT_COMPLETE
|ptp4l[952.879]: port 1: new foreign master 001395.fffe.4897b4-1
|ptp4l[956.879]: selected best master clock 001395.fffe.4897b4
|ptp4l[956.879]: port 1: assuming the grand master role
|ptp4l[956.879]: port 1: LISTENING to GRAND_MASTER on RS_GRAND_MASTER
|ptp4l[962.017]: port 1: received DELAY_REQ without timestamp
|ptp4l[962.273]: port 1: received DELAY_REQ without timestamp
|ptp4l[963.090]: port 1: received DELAY_REQ without timestamp

Commit f2fb6b6275eb ("net: stmmac: enable timestamp snapshot for required PTP
packets in dwmac v5.10a") already addresses this problem for the dwmac
v5.10. However, same holds true for all dwmacs above version v4.10. Correct the
check accordingly. Afterwards everything works as expected.

Tested on Intel Atom(R) x6414RE Processor.

Fixes: 14f347334bf2 ("net: stmmac: Correctly take timestamp for PTPv2")
Fixes: f2fb6b6275eb ("net: stmmac: enable timestamp snapshot for required PTP packets in dwmac v5.10a")
Suggested-by: Ong Boon Leong <boon.leong.ong@intel.com>
Signed-off-by: Kurt Kanzenbach <kurt@linutronix.de>
---
 drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
index eb3b7bf771d7..3d67d1fa3690 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
@@ -736,7 +736,7 @@ static int stmmac_hwtstamp_set(struct net_device *dev, struct ifreq *ifr)
 			config.rx_filter = HWTSTAMP_FILTER_PTP_V2_EVENT;
 			ptp_v2 = PTP_TCR_TSVER2ENA;
 			snap_type_sel = PTP_TCR_SNAPTYPSEL_1;
-			if (priv->synopsys_id != DWMAC_CORE_5_10)
+			if (priv->synopsys_id < DWMAC_CORE_4_10)
 				ts_event_en = PTP_TCR_TSEVNTENA;
 			ptp_over_ipv4_udp = PTP_TCR_TSIPV4ENA;
 			ptp_over_ipv6_udp = PTP_TCR_TSIPV6ENA;
-- 
2.30.2

Re: [PATCH net] net: stmmac: Fix E2E delay mechanism

[patchwork-bot+netdevbpf]

Hello:

This patch was applied to netdev/net.git (master)
by David S. Miller <davem@davemloft.net>:

On Wed, 20 Oct 2021 09:04:33 +0200 you wrote:
> When utilizing End to End delay mechanism, the following error messages show up:
> 
> |root@ehl1:~# ptp4l --tx_timestamp_timeout=50 -H -i eno2 -E -m
> |ptp4l[950.573]: selected /dev/ptp3 as PTP clock
> |ptp4l[950.586]: port 1: INITIALIZING to LISTENING on INIT_COMPLETE
> |ptp4l[950.586]: port 0: INITIALIZING to LISTENING on INIT_COMPLETE
> |ptp4l[952.879]: port 1: new foreign master 001395.fffe.4897b4-1
> |ptp4l[956.879]: selected best master clock 001395.fffe.4897b4
> |ptp4l[956.879]: port 1: assuming the grand master role
> |ptp4l[956.879]: port 1: LISTENING to GRAND_MASTER on RS_GRAND_MASTER
> |ptp4l[962.017]: port 1: received DELAY_REQ without timestamp
> |ptp4l[962.273]: port 1: received DELAY_REQ without timestamp
> |ptp4l[963.090]: port 1: received DELAY_REQ without timestamp
> 
> [...]

Here is the summary with links:
  - [net] net: stmmac: Fix E2E delay mechanism
    https://git.kernel.org/netdev/net/c/3cb958027cb8

You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html

Re: [PATCH net] net: stmmac: Fix E2E delay mechanism

[Wong Vee Khee]

On Wed, Oct 20, 2021 at 09:04:33AM +0200, Kurt Kanzenbach wrote:
> When utilizing End to End delay mechanism, the following error messages show up:
> 
> |root@ehl1:~# ptp4l --tx_timestamp_timeout=50 -H -i eno2 -E -m
> |ptp4l[950.573]: selected /dev/ptp3 as PTP clock
> |ptp4l[950.586]: port 1: INITIALIZING to LISTENING on INIT_COMPLETE
> |ptp4l[950.586]: port 0: INITIALIZING to LISTENING on INIT_COMPLETE
> |ptp4l[952.879]: port 1: new foreign master 001395.fffe.4897b4-1
> |ptp4l[956.879]: selected best master clock 001395.fffe.4897b4
> |ptp4l[956.879]: port 1: assuming the grand master role
> |ptp4l[956.879]: port 1: LISTENING to GRAND_MASTER on RS_GRAND_MASTER
> |ptp4l[962.017]: port 1: received DELAY_REQ without timestamp
> |ptp4l[962.273]: port 1: received DELAY_REQ without timestamp
> |ptp4l[963.090]: port 1: received DELAY_REQ without timestamp
> 
> Commit f2fb6b6275eb ("net: stmmac: enable timestamp snapshot for required PTP
> packets in dwmac v5.10a") already addresses this problem for the dwmac
> v5.10. However, same holds true for all dwmacs above version v4.10. Correct the
> check accordingly. Afterwards everything works as expected.
> 
> Tested on Intel Atom(R) x6414RE Processor.
> 
> Fixes: 14f347334bf2 ("net: stmmac: Correctly take timestamp for PTPv2")
> Fixes: f2fb6b6275eb ("net: stmmac: enable timestamp snapshot for required PTP packets in dwmac v5.10a")
> Suggested-by: Ong Boon Leong <boon.leong.ong@intel.com>
> Signed-off-by: Kurt Kanzenbach <kurt@linutronix.de>
> ---
>  drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> index eb3b7bf771d7..3d67d1fa3690 100644
> --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> @@ -736,7 +736,7 @@ static int stmmac_hwtstamp_set(struct net_device *dev, struct ifreq *ifr)
>  			config.rx_filter = HWTSTAMP_FILTER_PTP_V2_EVENT;
>  			ptp_v2 = PTP_TCR_TSVER2ENA;
>  			snap_type_sel = PTP_TCR_SNAPTYPSEL_1;
> -			if (priv->synopsys_id != DWMAC_CORE_5_10)
> +			if (priv->synopsys_id < DWMAC_CORE_4_10)
>  				ts_event_en = PTP_TCR_TSEVNTENA;
>  			ptp_over_ipv4_udp = PTP_TCR_TSIPV4ENA;
>  			ptp_over_ipv6_udp = PTP_TCR_TSIPV6ENA;

This fixes PTPv2 E2E issue on EHL-based boards. Can this also
be applied to 5.4 and onwards?

Thanks,
Vee Khee

Re: [PATCH net] net: stmmac: Fix E2E delay mechanism

[Kurt Kanzenbach]

[-- Attachment #1: Type: text/plain, Size: 275 bytes --]

On Thu Oct 21 2021, Wong Vee Khee wrote:
> This fixes PTPv2 E2E issue on EHL-based boards. Can this also
> be applied to 5.4 and onwards?

I've submitted the patch to net tree (instead of net-next) meaning it
will be picked up for stable kernels automatically.

Thanks,
Kurt

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 861 bytes --]

[PATCH net] net: stmmac: Fix E2E delay mechanism

[muhammad.nazim.amirul.nazle.asmade]

From: Nazim Amirul <muhammad.nazim.amirul.nazle.asmade@altera.com>

For E2E delay mechanism, "received DELAY_REQ without timestamp" error
messages show up for dwmac v3.70+ and dwxgmac IPs.

This issue affects socfpga platforms, Agilex7 (dwmac 3.70) and
Agilex5 (dwxgmac). According to the databook, to enable timestamping
for all events, the SNAPTYPSEL bits in the MAC_Timestamp_Control
register must be set to 2'b01, and the TSEVNTENA bit must be cleared
to 0'b0.

Commit 3cb958027cb8 ("net: stmmac: Fix E2E delay mechanism") already
addresses this problem for all dwmacs above version v4.10. However,
same holds true for v3.70 and above, as well as for dwxgmac. Updates
the check accordingly.

Fixes: 14f347334bf2 ("net: stmmac: Correctly take timestamp for PTPv2")
Fixes: f2fb6b6275eb ("net: stmmac: enable timestamp snapshot for required PTP packets in dwmac v5.10a")
Fixes: 3cb958027cb8 ("net: stmmac: Fix E2E delay mechanism")
Signed-off-by: Rohan G Thomas <rohan.g.thomas@altera.com>
Signed-off-by: Nazim Amirul <muhammad.nazim.amirul.nazle.asmade@altera.com>
---
This is a resend of a fix originally submitted by Rohan G Thomas. The
previous submission (v2):
https://lkml.org/lkml/2025/11/29/71

 drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
index 2a0d7eff88d3..24d8c6e9f1df 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
@@ -755,7 +755,8 @@ static int stmmac_hwtstamp_set(struct net_device *dev,
 			config->rx_filter = HWTSTAMP_FILTER_PTP_V2_EVENT;
 			ptp_v2 = PTP_TCR_TSVER2ENA;
 			snap_type_sel = PTP_TCR_SNAPTYPSEL_1;
-			if (priv->synopsys_id < DWMAC_CORE_4_10)
+			if (priv->synopsys_id < DWMAC_CORE_3_70 &&
+			    !priv->plat->has_xgmac)
 				ts_event_en = PTP_TCR_TSEVNTENA;
 			ptp_over_ipv4_udp = PTP_TCR_TSIPV4ENA;
 			ptp_over_ipv6_udp = PTP_TCR_TSIPV6ENA;
-- 
2.43.7

Re: [PATCH net] net: stmmac: Fix E2E delay mechanism

[Maxime Chevallier]

On 7/27/26 04:25, muhammad.nazim.amirul.nazle.asmade@altera.com wrote:
> From: Nazim Amirul <muhammad.nazim.amirul.nazle.asmade@altera.com>
> 
> For E2E delay mechanism, "received DELAY_REQ without timestamp" error
> messages show up for dwmac v3.70+ and dwxgmac IPs.
> 
> This issue affects socfpga platforms, Agilex7 (dwmac 3.70) and
> Agilex5 (dwxgmac). According to the databook, to enable timestamping
> for all events, the SNAPTYPSEL bits in the MAC_Timestamp_Control
> register must be set to 2'b01, and the TSEVNTENA bit must be cleared
> to 0'b0.
> 
> Commit 3cb958027cb8 ("net: stmmac: Fix E2E delay mechanism") already
> addresses this problem for all dwmacs above version v4.10. However,
> same holds true for v3.70 and above, as well as for dwxgmac. Updates
> the check accordingly.
> 
> Fixes: 14f347334bf2 ("net: stmmac: Correctly take timestamp for PTPv2")
> Fixes: f2fb6b6275eb ("net: stmmac: enable timestamp snapshot for required PTP packets in dwmac v5.10a")
> Fixes: 3cb958027cb8 ("net: stmmac: Fix E2E delay mechanism")
> Signed-off-by: Rohan G Thomas <rohan.g.thomas@altera.com>
> Signed-off-by: Nazim Amirul <muhammad.nazim.amirul.nazle.asmade@altera.com>

Given the original discussions that took place, it seems that this is
the correct fix. As Russell suggest, we need some timestamping selftest
to make it easy to find regressions with this kind of patches, but
I've experimented with stmmac selftests over the weekend and they are
in a sorry state right now...

Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com>

Maxime

Re: [PATCH net] net: stmmac: Fix E2E delay mechanism

[Nazle Asmade, Muhammad Nazim Amirul]

On 27/7/2026 3:52 pm, Maxime Chevallier wrote:
> 
> 
> On 7/27/26 04:25, muhammad.nazim.amirul.nazle.asmade@altera.com wrote:
>> From: Nazim Amirul <muhammad.nazim.amirul.nazle.asmade@altera.com>
>>
>> For E2E delay mechanism, "received DELAY_REQ without timestamp" error
>> messages show up for dwmac v3.70+ and dwxgmac IPs.
>>
>> This issue affects socfpga platforms, Agilex7 (dwmac 3.70) and
>> Agilex5 (dwxgmac). According to the databook, to enable timestamping
>> for all events, the SNAPTYPSEL bits in the MAC_Timestamp_Control
>> register must be set to 2'b01, and the TSEVNTENA bit must be cleared
>> to 0'b0.
>>
>> Commit 3cb958027cb8 ("net: stmmac: Fix E2E delay mechanism") already
>> addresses this problem for all dwmacs above version v4.10. However,
>> same holds true for v3.70 and above, as well as for dwxgmac. Updates
>> the check accordingly.
>>
>> Fixes: 14f347334bf2 ("net: stmmac: Correctly take timestamp for PTPv2")
>> Fixes: f2fb6b6275eb ("net: stmmac: enable timestamp snapshot for required PTP packets in dwmac v5.10a")
>> Fixes: 3cb958027cb8 ("net: stmmac: Fix E2E delay mechanism")
>> Signed-off-by: Rohan G Thomas <rohan.g.thomas@altera.com>
>> Signed-off-by: Nazim Amirul <muhammad.nazim.amirul.nazle.asmade@altera.com>
> 
> Given the original discussions that took place, it seems that this is
> the correct fix. As Russell suggest, we need some timestamping selftest
> to make it easy to find regressions with this kind of patches, but
> I've experimented with stmmac selftests over the weekend and they are
> in a sorry state right now...
> 
> Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com>

Thanks Maxime!>
> Maxime
> 
> 

Re: [PATCH net] net: stmmac: Fix E2E delay mechanism

[Jakub Raczynski]

[-- Attachment #1: Type: text/plain, Size: 1044 bytes --]

On Sun, Jul 26, 2026 at 07:25:24PM -0700, muhammad.nazim.amirul.nazle.asmade@altera.com wrote:
> From: Nazim Amirul <muhammad.nazim.amirul.nazle.asmade@altera.com>
> 
> diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> index 2a0d7eff88d3..24d8c6e9f1df 100644
> --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> @@ -755,7 +755,8 @@ static int stmmac_hwtstamp_set(struct net_device *dev,
>  			config->rx_filter = HWTSTAMP_FILTER_PTP_V2_EVENT;
>  			ptp_v2 = PTP_TCR_TSVER2ENA;
>  			snap_type_sel = PTP_TCR_SNAPTYPSEL_1;
> -			if (priv->synopsys_id < DWMAC_CORE_4_10)
> +			if (priv->synopsys_id < DWMAC_CORE_3_70 &&
> +			    !priv->plat->has_xgmac)

This change does not compile, as has_xgmac no longer exists
since 26ab9830beabd. Please check patchwork for more details.

But on my test kernel 6.18 from a while ago where E2E is broken,
it fixes the issue.
So Tested-by me but not really.

BR
Jakub Raczynski

[-- Attachment #2: Type: text/plain, Size: 0 bytes --]

Re: [PATCH net] net: stmmac: Fix E2E delay mechanism

[Nazle Asmade, Muhammad Nazim Amirul]

On 27/7/2026 11:29 pm, Jakub Raczynski wrote:
> [You don't often get email from j.raczynski@samsung.com. Learn why this is important at https://aka.ms/LearnAboutSenderIdentification ]
> 
> On Sun, Jul 26, 2026 at 07:25:24PM -0700, muhammad.nazim.amirul.nazle.asmade@altera.com wrote:
>> From: Nazim Amirul <muhammad.nazim.amirul.nazle.asmade@altera.com>
>>
>> diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
>> index 2a0d7eff88d3..24d8c6e9f1df 100644
>> --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
>> +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
>> @@ -755,7 +755,8 @@ static int stmmac_hwtstamp_set(struct net_device *dev,
>>                        config->rx_filter = HWTSTAMP_FILTER_PTP_V2_EVENT;
>>                        ptp_v2 = PTP_TCR_TSVER2ENA;
>>                        snap_type_sel = PTP_TCR_SNAPTYPSEL_1;
>> -                     if (priv->synopsys_id < DWMAC_CORE_4_10)
>> +                     if (priv->synopsys_id < DWMAC_CORE_3_70 &&
>> +                         !priv->plat->has_xgmac)
> 
> This change does not compile, as has_xgmac no longer exists
> since 26ab9830beabd. Please check patchwork for more details.
> 
> But on my test kernel 6.18 from a while ago where E2E is broken,
> it fixes the issue.
> So Tested-by me but not really.
> 
> BR
> Jakub Raczynski
> 
> 
Thanks Jakub!

Let me fix it and repost with v2.

BR,
Nazim