Re: [PATCH v3 net-next 01/11] net: stmmac: prepare dma op mode config for multiple queues

2017-05-08 Thread Jan Kiszka 
On 2017-05-08 14:02, Joao Pinto wrote:
> Às 12:56 PM de 5/8/2017, Andy Shevchenko escreveu:
>> On Mon, May 8, 2017 at 2:40 PM, Joao Pinto  wrote:
>>> Às 12:34 PM de 5/8/2017, Andy Shevchenko escreveu:
 On Mon, May 8, 2017 at 1:42 PM, Joao Pinto  wrote:
> Às 11:12 AM de 5/8/2017, Andy Shevchenko escreveu:
>> On Mon, May 8, 2017 at 12:54 PM, Joao Pinto  
>> wrote:
>>> Às 10:36 AM de 5/8/2017, Andy Shevchenko escreveu:
>>

 [   44.374161] stmmac_dvr_probe <<< 0 0

>>>
>>> Ok, so this is the cause of the problem. The driver is geting 0 for real RX 
>>> and
>>> TX queues.
>>>
>>> Your setup uses standard DT parsing from stmmac_platform or a custom one?
>>>
>>> If you are using stmmac_probe_config_dt():
>>> https://urldefense.proofpoint.com/v2/url?u=https-3A__git.kernel.org_pub_scm_linux_kernel_git_torvalds_linux.git_tree_drivers_net_ethernet_stmicro_stmmac_stmmac-5Fplatform.c-23n363=DwIFaQ=DPL6_X_6JkXFx7AXWqB0tg=s2fO0hii0OGNOv9qQy_HRXy-xAJUD1NNoEcc3io_kx0=fJQj7RiT2sksJYOAZ9VSJUDnxPR7RlE6Fw_cTV0_Mqc=KhdAPUtP0twDkibE89cLYs8JjnxEvBgav5uf08WL_e8=
>>>  
>>>
>>> You will find a function named stmmac_mtl_setup() being called:
>>> https://urldefense.proofpoint.com/v2/url?u=https-3A__git.kernel.org_pub_scm_linux_kernel_git_torvalds_linux.git_tree_drivers_net_ethernet_stmicro_stmmac_stmmac-5Fplatform.c-23n492=DwIFaQ=DPL6_X_6JkXFx7AXWqB0tg=s2fO0hii0OGNOv9qQy_HRXy-xAJUD1NNoEcc3io_kx0=fJQj7RiT2sksJYOAZ9VSJUDnxPR7RlE6Fw_cTV0_Mqc=rTxn0fwdudwq9XAquH60xNHN538KBQ6_n4wODdLoyA0=
>>>  
>>>
>>> In this function, the number of RX and TX queues is being set to 1 by 
>>> default.
>>
>> Ah-ha, now I know how it's happened.
>> You forget to update all setup() hooks in PCI bus driver :-)
>>
>> I will prepare a fix.
>> Just tell me should I put Fixes tag or not? And if yes, what commit
>> should I refer to?
>>
> 
> Great, you can use this commit:
> 
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c?id=26d6851fd24ed5d88580d66b4c8384947d5ca29b
> 
> Thanks!
> 
> Joao
> 

Perfect, looking forward to try out a fix. Thanks, folks!

Jan

-- 
Siemens AG, Corporate Technology, CT RDA ITP SES-DE
Corporate Competence Center Embedded Linux

Re: [PATCH v3 net-next 01/11] net: stmmac: prepare dma op mode config for multiple queues

2017-05-08 Thread Joao Pinto 
Às 12:56 PM de 5/8/2017, Andy Shevchenko escreveu:
> On Mon, May 8, 2017 at 2:40 PM, Joao Pinto  wrote:
>> Às 12:34 PM de 5/8/2017, Andy Shevchenko escreveu:
>>> On Mon, May 8, 2017 at 1:42 PM, Joao Pinto  wrote:
 Às 11:12 AM de 5/8/2017, Andy Shevchenko escreveu:
> On Mon, May 8, 2017 at 12:54 PM, Joao Pinto  
> wrote:
>> Às 10:36 AM de 5/8/2017, Andy Shevchenko escreveu:
> 
>>>
>>> [   44.374161] stmmac_dvr_probe <<< 0 0
>>>
>>
>> Ok, so this is the cause of the problem. The driver is geting 0 for real RX 
>> and
>> TX queues.
>>
>> Your setup uses standard DT parsing from stmmac_platform or a custom one?
>>
>> If you are using stmmac_probe_config_dt():
>> https://urldefense.proofpoint.com/v2/url?u=https-3A__git.kernel.org_pub_scm_linux_kernel_git_torvalds_linux.git_tree_drivers_net_ethernet_stmicro_stmmac_stmmac-5Fplatform.c-23n363=DwIFaQ=DPL6_X_6JkXFx7AXWqB0tg=s2fO0hii0OGNOv9qQy_HRXy-xAJUD1NNoEcc3io_kx0=fJQj7RiT2sksJYOAZ9VSJUDnxPR7RlE6Fw_cTV0_Mqc=KhdAPUtP0twDkibE89cLYs8JjnxEvBgav5uf08WL_e8=
>>  
>>
>> You will find a function named stmmac_mtl_setup() being called:
>> https://urldefense.proofpoint.com/v2/url?u=https-3A__git.kernel.org_pub_scm_linux_kernel_git_torvalds_linux.git_tree_drivers_net_ethernet_stmicro_stmmac_stmmac-5Fplatform.c-23n492=DwIFaQ=DPL6_X_6JkXFx7AXWqB0tg=s2fO0hii0OGNOv9qQy_HRXy-xAJUD1NNoEcc3io_kx0=fJQj7RiT2sksJYOAZ9VSJUDnxPR7RlE6Fw_cTV0_Mqc=rTxn0fwdudwq9XAquH60xNHN538KBQ6_n4wODdLoyA0=
>>  
>>
>> In this function, the number of RX and TX queues is being set to 1 by 
>> default.
> 
> Ah-ha, now I know how it's happened.
> You forget to update all setup() hooks in PCI bus driver :-)
> 
> I will prepare a fix.
> Just tell me should I put Fixes tag or not? And if yes, what commit
> should I refer to?
> 

Great, you can use this commit:

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c?id=26d6851fd24ed5d88580d66b4c8384947d5ca29b

Thanks!

Joao

Re: [PATCH v3 net-next 01/11] net: stmmac: prepare dma op mode config for multiple queues

2017-05-08 Thread Andy Shevchenko 
On Mon, May 8, 2017 at 2:40 PM, Joao Pinto  wrote:
> Às 12:34 PM de 5/8/2017, Andy Shevchenko escreveu:
>> On Mon, May 8, 2017 at 1:42 PM, Joao Pinto  wrote:
>>> Às 11:12 AM de 5/8/2017, Andy Shevchenko escreveu:
 On Mon, May 8, 2017 at 12:54 PM, Joao Pinto  
 wrote:
> Às 10:36 AM de 5/8/2017, Andy Shevchenko escreveu:

>>
>> [   44.374161] stmmac_dvr_probe <<< 0 0
>>
>
> Ok, so this is the cause of the problem. The driver is geting 0 for real RX 
> and
> TX queues.
>
> Your setup uses standard DT parsing from stmmac_platform or a custom one?
>
> If you are using stmmac_probe_config_dt():
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c#n363
>
> You will find a function named stmmac_mtl_setup() being called:
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c#n492
>
> In this function, the number of RX and TX queues is being set to 1 by default.

Ah-ha, now I know how it's happened.
You forget to update all setup() hooks in PCI bus driver :-)

I will prepare a fix.
Just tell me should I put Fixes tag or not? And if yes, what commit
should I refer to?

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH v3 net-next 01/11] net: stmmac: prepare dma op mode config for multiple queues

2017-05-08 Thread Joao Pinto 
Às 12:34 PM de 5/8/2017, Andy Shevchenko escreveu:
> On Mon, May 8, 2017 at 1:42 PM, Joao Pinto  wrote:
>> Às 11:12 AM de 5/8/2017, Andy Shevchenko escreveu:
>>> On Mon, May 8, 2017 at 12:54 PM, Joao Pinto  wrote:
 Às 10:36 AM de 5/8/2017, Andy Shevchenko escreveu:
> 
> JFYI: With today's linux-next when _kexec:ed_ kernel boots I tried and
> got the following:
> 
 Are you using the same version of Ethernet IP, 10/100?
>>>
>>> I'm running on Intel Galileo Gen2 board (v4.11 by the way works fine
>>> with direct boot from SD card)
>>>
 Could you please verify if the crash you are experiencing is this place?
 https://urldefense.proofpoint.com/v2/url?u=https-3A__git.kernel.org_pub_scm_linux_kernel_git_torvalds_linux.git_tree_drivers_net_ethernet_stmicro_stmmac_stmmac-5Fmain.c-23n2956=DwIFaQ=DPL6_X_6JkXFx7AXWqB0tg=s2fO0hii0OGNOv9qQy_HRXy-xAJUD1NNoEcc3io_kx0=UF269QZ9ExFRw1XXpgdvO2QeTCLEp-GquRe8OqZwRf0=yZu3uME5PK-3nJlxz-H-HfHh3Shjzg0je5If_jSXVb4=

 I would say that for rather old IPs, the napi is not capable of giving a 
 valid
 queue number. Could you please print the queue index returned by this line?

 https://urldefense.proofpoint.com/v2/url?u=https-3A__git.kernel.org_pub_scm_linux_kernel_git_torvalds_linux.git_tree_drivers_net_ethernet_stmicro_stmmac_stmmac-5Fmain.c-23n2948=DwIFaQ=DPL6_X_6JkXFx7AXWqB0tg=s2fO0hii0OGNOv9qQy_HRXy-xAJUD1NNoEcc3io_kx0=UF269QZ9ExFRw1XXpgdvO2QeTCLEp-GquRe8OqZwRf0=p_TgHODJum23I2N4AldR4oIaOPffSDpk9agmbRMQgoM=
>>>
>>> --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
>>> +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
>>> @@ -2953,7 +2953,9 @@ static netdev_tx_t stmmac_xmit(struct sk_buff
>>> *skb, struct net_device *dev)
>>>unsigned int enh_desc;
>>>unsigned int des;
>>>
>>> +   pr_info("%s <<< 1: priv %p, queue: %u\n", __func__, priv, queue);
>>>tx_q = >tx_queue[queue];
>>> +   pr_info("%s <<< 2: priv %p, queue: %u tx_q: %p\n", __func__,
>>> priv, queue, tx_q);
>>>
>>>
>>> [  101.591040] stmmac_xmit <<< 1: priv cdd1c4c0, queue: 7
>>> [  101.596377] stmmac_xmit <<< 2: priv cdd1c4c0, queue: 7 tx_q: cdd1caac
> 
>> I assume that the queue index is always 7 right? By return 7, the napi 
>> interface
>> 'thinks' that your setup is using 8 TX queues which I assume it is not and 
>> thats
>> the problem causing your board to malfuntion.
>>
>> Could you please check the values of the 'real' tx and rx queues count in 
>> this line?
>> https://urldefense.proofpoint.com/v2/url?u=https-3A__git.kernel.org_pub_scm_linux_kernel_git_torvalds_linux.git_tree_drivers_net_ethernet_stmicro_stmmac_stmmac-5Fmain.c-23n4107=DwIFaQ=DPL6_X_6JkXFx7AXWqB0tg=s2fO0hii0OGNOv9qQy_HRXy-xAJUD1NNoEcc3io_kx0=6PN46fgWi1XTHkxFzV9wkYHPkKJWvkRC1OOlEhyKdcA=cyYmWeYuPwacYmVRzJbhRm3Krz6XNyHbxq8t7ZUi8Ec=
>>  
>>
>> For default they are =1, so napi should be assuming 1RX and 1TX, and so you
>> should be getting queue index =0 in reception and transmission.
>>
>> In terms of reception, could you print the queue index that stmmac_poll is 
>> using
>> here:
>>
>> https://urldefense.proofpoint.com/v2/url?u=https-3A__git.kernel.org_pub_scm_linux_kernel_git_torvalds_linux.git_tree_drivers_net_ethernet_stmicro_stmmac_stmmac-5Fmain.c-23n3468=DwIFaQ=DPL6_X_6JkXFx7AXWqB0tg=s2fO0hii0OGNOv9qQy_HRXy-xAJUD1NNoEcc3io_kx0=6PN46fgWi1XTHkxFzV9wkYHPkKJWvkRC1OOlEhyKdcA=Xli0e7Key3FA7Rve_opcwc6W7nd4khVX15wwoNpFHL4=
>>  
> 
> +   pr_info("%s <<< %u\n", __func__, rx_q->queue_index);
>work_done = stmmac_rx(priv, budget, rx_q->queue_index);
>if (work_done < budget) {
>napi_complete_done(napi, work_done);
> 
>/* Configure real RX and TX queues */
>netif_set_real_num_rx_queues(ndev, priv->plat->rx_queues_to_use);
>netif_set_real_num_tx_queues(ndev, priv->plat->tx_queues_to_use);
> +   pr_info("%s <<< %hhu %hhu\n", __func__,
> priv->plat->rx_queues_to_use, priv->plat->tx_queues_to_use);
> 
> 
> [   44.374161] stmmac_dvr_probe <<< 0 0
> 

Ok, so this is the cause of the problem. The driver is geting 0 for real RX and
TX queues.

Your setup uses standard DT parsing from stmmac_platform or a custom one?

If you are using stmmac_probe_config_dt():
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c#n363

You will find a function named stmmac_mtl_setup() being called:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c#n492

In this function, the number of RX and TX queues is being set to 1 by default.

Joao


> [  109.014763] stmmac_xmit <<< 1: priv cdcea4c0, queue: 2
> [  109.020099] stmmac_xmit <<< 2: priv cdcea4c0, queue: 2 tx_q: cdcea9e4
> 
> That's all, no poll activated.
>

Re: [PATCH v3 net-next 01/11] net: stmmac: prepare dma op mode config for multiple queues

2017-05-08 Thread Andy Shevchenko 
On Mon, May 8, 2017 at 1:42 PM, Joao Pinto  wrote:
> Às 11:12 AM de 5/8/2017, Andy Shevchenko escreveu:
>> On Mon, May 8, 2017 at 12:54 PM, Joao Pinto  wrote:
>>> Às 10:36 AM de 5/8/2017, Andy Shevchenko escreveu:

 JFYI: With today's linux-next when _kexec:ed_ kernel boots I tried and
 got the following:

>>> Are you using the same version of Ethernet IP, 10/100?
>>
>> I'm running on Intel Galileo Gen2 board (v4.11 by the way works fine
>> with direct boot from SD card)
>>
>>> Could you please verify if the crash you are experiencing is this place?
>>> https://urldefense.proofpoint.com/v2/url?u=https-3A__git.kernel.org_pub_scm_linux_kernel_git_torvalds_linux.git_tree_drivers_net_ethernet_stmicro_stmmac_stmmac-5Fmain.c-23n2956=DwIFaQ=DPL6_X_6JkXFx7AXWqB0tg=s2fO0hii0OGNOv9qQy_HRXy-xAJUD1NNoEcc3io_kx0=UF269QZ9ExFRw1XXpgdvO2QeTCLEp-GquRe8OqZwRf0=yZu3uME5PK-3nJlxz-H-HfHh3Shjzg0je5If_jSXVb4=
>>>
>>> I would say that for rather old IPs, the napi is not capable of giving a 
>>> valid
>>> queue number. Could you please print the queue index returned by this line?
>>>
>>> https://urldefense.proofpoint.com/v2/url?u=https-3A__git.kernel.org_pub_scm_linux_kernel_git_torvalds_linux.git_tree_drivers_net_ethernet_stmicro_stmmac_stmmac-5Fmain.c-23n2948=DwIFaQ=DPL6_X_6JkXFx7AXWqB0tg=s2fO0hii0OGNOv9qQy_HRXy-xAJUD1NNoEcc3io_kx0=UF269QZ9ExFRw1XXpgdvO2QeTCLEp-GquRe8OqZwRf0=p_TgHODJum23I2N4AldR4oIaOPffSDpk9agmbRMQgoM=
>>
>> --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
>> +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
>> @@ -2953,7 +2953,9 @@ static netdev_tx_t stmmac_xmit(struct sk_buff
>> *skb, struct net_device *dev)
>>unsigned int enh_desc;
>>unsigned int des;
>>
>> +   pr_info("%s <<< 1: priv %p, queue: %u\n", __func__, priv, queue);
>>tx_q = >tx_queue[queue];
>> +   pr_info("%s <<< 2: priv %p, queue: %u tx_q: %p\n", __func__,
>> priv, queue, tx_q);
>>
>>
>> [  101.591040] stmmac_xmit <<< 1: priv cdd1c4c0, queue: 7
>> [  101.596377] stmmac_xmit <<< 2: priv cdd1c4c0, queue: 7 tx_q: cdd1caac

> I assume that the queue index is always 7 right? By return 7, the napi 
> interface
> 'thinks' that your setup is using 8 TX queues which I assume it is not and 
> thats
> the problem causing your board to malfuntion.
>
> Could you please check the values of the 'real' tx and rx queues count in 
> this line?
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c#n4107
>
> For default they are =1, so napi should be assuming 1RX and 1TX, and so you
> should be getting queue index =0 in reception and transmission.
>
> In terms of reception, could you print the queue index that stmmac_poll is 
> using
> here:
>
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c#n3468

+   pr_info("%s <<< %u\n", __func__, rx_q->queue_index);
   work_done = stmmac_rx(priv, budget, rx_q->queue_index);
   if (work_done < budget) {
   napi_complete_done(napi, work_done);

   /* Configure real RX and TX queues */
   netif_set_real_num_rx_queues(ndev, priv->plat->rx_queues_to_use);
   netif_set_real_num_tx_queues(ndev, priv->plat->tx_queues_to_use);
+   pr_info("%s <<< %hhu %hhu\n", __func__,
priv->plat->rx_queues_to_use, priv->plat->tx_queues_to_use);


[   44.374161] stmmac_dvr_probe <<< 0 0

[  109.014763] stmmac_xmit <<< 1: priv cdcea4c0, queue: 2
[  109.020099] stmmac_xmit <<< 2: priv cdcea4c0, queue: 2 tx_q: cdcea9e4

That's all, no poll activated.

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH v3 net-next 01/11] net: stmmac: prepare dma op mode config for multiple queues

2017-05-08 Thread Joao Pinto 
Às 11:12 AM de 5/8/2017, Andy Shevchenko escreveu:
> On Mon, May 8, 2017 at 12:54 PM, Joao Pinto  wrote:
>> Hi Andy and Jan,
>>
>> Às 10:36 AM de 5/8/2017, Andy Shevchenko escreveu:
>>> On Mon, May 8, 2017 at 9:56 AM, Jan Kiszka  wrote:
 On 2017-03-15 12:04, Joao Pinto wrote:
> This patch prepares DMA Operation Mode configuration for multiple queues.
> The work consisted on breaking the DMA operation Mode configuration 
> function
> into RX and TX scope and adapting its mechanism in stmmac_main.
>>>
 Starting with this patch, the stmmac-based network adapters of the Intel
 Quark SoC stop working. I'm getting an IP via DHCP, I can ping, but TCP
 connections can no longer be established.
> 
>>> JFYI: With today's linux-next when _kexec:ed_ kernel boots I tried and
>>> got the following:
>>>

snip (...)

>>>
>>>
>>
>> Are you using the same version of Ethernet IP, 10/100?
> 
> I'm running on Intel Galileo Gen2 board (v4.11 by the way works fine
> with direct boot from SD card)
> 
>> Could you please verify if the crash you are experiencing is this place?
>> https://urldefense.proofpoint.com/v2/url?u=https-3A__git.kernel.org_pub_scm_linux_kernel_git_torvalds_linux.git_tree_drivers_net_ethernet_stmicro_stmmac_stmmac-5Fmain.c-23n2956=DwIFaQ=DPL6_X_6JkXFx7AXWqB0tg=s2fO0hii0OGNOv9qQy_HRXy-xAJUD1NNoEcc3io_kx0=UF269QZ9ExFRw1XXpgdvO2QeTCLEp-GquRe8OqZwRf0=yZu3uME5PK-3nJlxz-H-HfHh3Shjzg0je5If_jSXVb4=
>>  
>>
>> I would say that for rather old IPs, the napi is not capable of giving a 
>> valid
>> queue number. Could you please print the queue index returned by this line?
>>
>> https://urldefense.proofpoint.com/v2/url?u=https-3A__git.kernel.org_pub_scm_linux_kernel_git_torvalds_linux.git_tree_drivers_net_ethernet_stmicro_stmmac_stmmac-5Fmain.c-23n2948=DwIFaQ=DPL6_X_6JkXFx7AXWqB0tg=s2fO0hii0OGNOv9qQy_HRXy-xAJUD1NNoEcc3io_kx0=UF269QZ9ExFRw1XXpgdvO2QeTCLEp-GquRe8OqZwRf0=p_TgHODJum23I2N4AldR4oIaOPffSDpk9agmbRMQgoM=
>>  
> 
> --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> @@ -2953,7 +2953,9 @@ static netdev_tx_t stmmac_xmit(struct sk_buff
> *skb, struct net_device *dev)
>unsigned int enh_desc;
>unsigned int des;
> 
> +   pr_info("%s <<< 1: priv %p, queue: %u\n", __func__, priv, queue);
>tx_q = >tx_queue[queue];
> +   pr_info("%s <<< 2: priv %p, queue: %u tx_q: %p\n", __func__,
> priv, queue, tx_q);
> 
> 
> [  101.591040] stmmac_xmit <<< 1: priv cdd1c4c0, queue: 7
> [  101.596377] stmmac_xmit <<< 2: priv cdd1c4c0, queue: 7 tx_q: cdd1caac
> 

I assume that the queue index is always 7 right? By return 7, the napi interface
'thinks' that your setup is using 8 TX queues which I assume it is not and thats
the problem causing your board to malfuntion.

Could you please check the values of the 'real' tx and rx queues count in this 
line?
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c#n4107

For default they are =1, so napi should be assuming 1RX and 1TX, and so you
should be getting queue index =0 in reception and transmission.

In terms of reception, could you print the queue index that stmmac_poll is using
here:

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c#n3468

> 
> Also noticed warning that have to be addressed:
> 
> drivers/net/ethernet/stmicro/stmmac/stmmac_main.c:2504:49: warning:
> incorrect type in argument 1 (different address spaces)
> drivers/net/ethernet/stmicro/stmmac/stmmac_main.c:2504:49:expected
> void [noderef] *ioaddr
> drivers/net/ethernet/stmicro/stmmac/stmmac_main.c:2504:49:got
> struct mac_device_info *hw

This one was well caught! Although it has no influence in your setup, since you
don't have this callback implemented, eQOS (>= 4.00) and 1000 cores will have
issues if using PCS. I can make a patch for this one.

> 
> drivers/net/ethernet/stmicro/stmmac/stmmac_main.c: In function
> ‘init_dma_rx_desc_rings’:
> drivers/net/ethernet/stmicro/stmmac/stmmac_main.c:1274:15: warning:
> comparison of
> unsigned expression >= 0 is always true [-Wtype-limits]
>  while (queue >= 0) {
>   ^~

This one I have in my agenda to improve it, I also talked about it with Dan
Carpenter about it.

Re: [PATCH v3 net-next 01/11] net: stmmac: prepare dma op mode config for multiple queues

2017-05-08 Thread Andy Shevchenko 
On Mon, May 8, 2017 at 12:54 PM, Joao Pinto  wrote:
> Hi Andy and Jan,
>
> Às 10:36 AM de 5/8/2017, Andy Shevchenko escreveu:
>> On Mon, May 8, 2017 at 9:56 AM, Jan Kiszka  wrote:
>>> On 2017-03-15 12:04, Joao Pinto wrote:
 This patch prepares DMA Operation Mode configuration for multiple queues.
 The work consisted on breaking the DMA operation Mode configuration 
 function
 into RX and TX scope and adapting its mechanism in stmmac_main.
>>
>>> Starting with this patch, the stmmac-based network adapters of the Intel
>>> Quark SoC stop working. I'm getting an IP via DHCP, I can ping, but TCP
>>> connections can no longer be established.

>> JFYI: With today's linux-next when _kexec:ed_ kernel boots I tried and
>> got the following:
>>
>>
>> # ip a s
>> 1: lo:  mtu 65536 qdisc noqueue qlen 1000
>>link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
>>inet 127.0.0.1/8 scope host lo
>>   valid_lft forever preferred_lft forever
>>inet6 ::1/128 scope host
>>   valid_lft forever preferre[  130.403995] random: fast init done
>> d_lft forever
>> 2: eth0:  mtu 1500 qdisc noop qlen 1000
>>link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff
>> 3: eth1:  mtu 1500 qdisc noop qlen 1000
>>link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff
>> 4: sit0@NONE:  mtu 1480 qdisc noop qlen 1000
>>link/sit 0.0.0.0 brd 0.0.0.0
>> # udhcpc -i eth0
>> udhcpc: started, v1.26.2
>> [  140.825131] stmmaceth :00:14.6 eth0: device MAC address 
>> 98:4f:ee:05:ac:47
>> [  140.834304] Generic PHY stmmac-a6:01: attached PHY driver [Generic
>> PHY] (mii_bus:phy_addr=stmmac-a6:01, irq=-1)
>> [  140.930871] stmmaceth :00:14.6 eth0: IEEE 1588-2008 Advanced
>> Timestamp supported
>> [  140.941109] stmmaceth :00:14.6 eth0: registered PTP clock
>> [  140.953626] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
>> udhcpc: sending discover
>> [  142.979557] stmmaceth :00:14.6 eth0: Link is Up - 100Mbps/Full
>> - flow control off
>> [  142.988756] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
>> [  142.998810] BUG: unable to handle kernel NULL pointer dereference at   
>> (null)
>> [  143.006193] IP: stmmac_xmit+0xf1/0x1080
>> [  143.010168] *pde = 
>> [  143.010177]
>> [  143.014762] Oops: 0002 [#1]
>> [  143.017672] Modules linked in: at24 nvmem_core pwm_pca9685
>> [  143.023338] CPU: 0 PID: 0 Comm: swapper Not tainted 4.11.0-next-20170508+ 
>> #2
>> [  143.030539] task: c8533580 task.stack: c852c000
>> [  143.035237] EIP: stmmac_xmit+0xf1/0x1080
>> [  143.039302] EFLAGS: 00010216 CPU: 0
>> [  143.042915] EAX:  EBX: 0050 ECX:  EDX: ceb6a0c0
>> [  143.049326] ESI:  EDI: cdd16000 EBP: cdc25d70 ESP: cdc25d20
>> [  143.055735]  DS: 007b ES: 007b FS:  GS:  SS: 0068
>> [  143.061271] CR0: 80050033 CR2:  CR3: 0eb5c000 CR4: 00100010
>> [  143.067671] Call Trace:
>> [  143.070238]  
>> [  143.072763]  dev_hard_start_xmit+0x7c/0x1a0
>> [  143.077120]  sch_direct_xmit+0xf0/0x120
>> [  143.081130]  __dev_queue_xmit+0x181/0x430
>> [  143.085311]  ? eth_commit_mac_addr_change+0x20/0x20
>> [  143.090362]  dev_queue_xmit+0xa/0x10
>> [  143.094100]  neigh_resolve_output+0xdb/0x190
>> [  143.098561]  ip6_finish_output2+0x184/0x500
>> [  143.102945]  ip6_finish_output+0x91/0xe0
>> [  143.107057]  ? ip6_finish_output+0x91/0xe0
>> [  143.111338]  ip6_output+0x36/0x110
>> [  143.114924]  ? ip6_fragment+0xb00/0xb00
>> [  143.118935]  mld_sendpack+0x191/0x2b0
>> [  143.122769]  ? mld_newpack+0xda/0x180
>> [  143.126598]  ? ipv6_icmp_sysctl_init+0x30/0x30
>> [  143.131224]  mld_ifc_timer_expire+0x158/0x240
>> [  143.135756]  ? find_next_bit+0xa/0x10
>> [  143.139584]  ? mld_dad_timer_expire+0x50/0x50
>> [  143.144112]  call_timer_fn+0x2a/0xf0
>> [  143.147862]  ? mld_dad_timer_expire+0x50/0x50
>> [  143.152395]  run_timer_softirq+0x158/0x300
>> [  143.156668]  ? file_free_rcu+0x1e/0x30
>> [  143.160589]  __do_softirq+0xc4/0x200
>> [  143.164341]  ? __hrtimer_tasklet_trampoline+0x30/0x30
>> [  143.169575]  do_softirq_own_stack+0x1e/0x30
>> [  143.173902]  
>> [  143.176502]  irq_exit+0x95/0xa0
>> [  143.179812]  smp_apic_timer_interrupt+0x31/0x40
>> [  143.184530]  apic_timer_interrupt+0x32/0x40
>> [  143.19] EIP: default_idle+0xc/0x70
>> [  143.192774] EFLAGS: 0246 CPU: 0
>> [  143.196386] EAX:  EBX:  ECX: 0001 EDX: 
>> [  143.202795] ESI:  EDI: c8533580 EBP: c852df54 ESP: c852df4c
>> [  143.209205]  DS: 007b ES: 007b FS:  GS:  SS: 0068
>> [  143.214780]  arch_cpu_idle+0x9/0x10
>> [  143.218446]  default_idle_call+0x17/0x30
>> [  143.222551]  do_idle+0xed/0x130
>> [  143.225873]  cpu_startup_entry+0x15/0x20
>> [  143.229965]  rest_init+0x5c/0x60
>> [  143.233370]  start_kernel+0x313/0x318
>> [  143.237221]  i386_start_kernel+0x98/0x9c
>> [  143.241315]  startup_32_smp+0x16b/0x16d
>> [

Re: [PATCH v3 net-next 01/11] net: stmmac: prepare dma op mode config for multiple queues

2017-05-08 Thread Joao Pinto 
Hi Andy and Jan,

Às 10:36 AM de 5/8/2017, Andy Shevchenko escreveu:
> On Mon, May 8, 2017 at 9:56 AM, Jan Kiszka  wrote:
>> On 2017-03-15 12:04, Joao Pinto wrote:
>>> This patch prepares DMA Operation Mode configuration for multiple queues.
>>> The work consisted on breaking the DMA operation Mode configuration function
>>> into RX and TX scope and adapting its mechanism in stmmac_main.
> 
>> Starting with this patch, the stmmac-based network adapters of the Intel
>> Quark SoC stop working. I'm getting an IP via DHCP, I can ping, but TCP
>> connections can no longer be established.
>>
>> Moving on a few patches (didn't bisect the exact one yet), the TX
>> watchdog starts to fire, and DHCP fails completely. And if I go to
>> current master in Linus tree (reverting an unrelated boot regression), I
>> even get a crash in stmmac_xmit.
>>
>> Here are some details about the hw from dma_cap POV, if this helps:
>>
>> ==
>> DMA HW features
>> ==
>> 10/100 Mbps: Y
>> 1000 Mbps: N
>> Half duplex: Y
>> Hash Filter: Y
>> Multiple MAC address registers: N
>> PCS (TBI/SGMII/RTBI PHY interfaces): N
>> SMA (MDIO) Interface: Y
>> PMT Remote wake up: N
>> PMT Magic Frame: N
>> RMON module: Y
>> IEEE 1588-2002 Time Stamp: N
>> IEEE 1588-2008 Advanced Time Stamp: Y
>> 802.3az - Energy-Efficient Ethernet (EEE): N
>> AV features: N
>> Checksum Offload in TX: Y
>> IP Checksum Offload (type1) in RX: N
>> IP Checksum Offload (type2) in RX: Y
>> RXFIFO > 2048bytes: Y
>> Number of Additional RX channel: 0
>> Number of Additional TX channel: 0
>> Enhanced descriptors: Y
>>
>> Given the number of different failure modes, my feeling is that there
>> are multiple regressions coming with these patches...
>>
>> I've tested on the IOT2000 board, but I suspect the Galileo Gen2 will be
>> affected equally. If you don't have access to any such device, let me
>> know what I can debug for you.
> 
> JFYI: With today's linux-next when _kexec:ed_ kernel boots I tried and
> got the following:
> 
> 
> # ip a s
> 1: lo:  mtu 65536 qdisc noqueue qlen 1000
>link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
>inet 127.0.0.1/8 scope host lo
>   valid_lft forever preferred_lft forever
>inet6 ::1/128 scope host
>   valid_lft forever preferre[  130.403995] random: fast init done
> d_lft forever
> 2: eth0:  mtu 1500 qdisc noop qlen 1000
>link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff
> 3: eth1:  mtu 1500 qdisc noop qlen 1000
>link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff
> 4: sit0@NONE:  mtu 1480 qdisc noop qlen 1000
>link/sit 0.0.0.0 brd 0.0.0.0
> # udhcpc -i eth0
> udhcpc: started, v1.26.2
> [  140.825131] stmmaceth :00:14.6 eth0: device MAC address 
> 98:4f:ee:05:ac:47
> [  140.834304] Generic PHY stmmac-a6:01: attached PHY driver [Generic
> PHY] (mii_bus:phy_addr=stmmac-a6:01, irq=-1)
> [  140.930871] stmmaceth :00:14.6 eth0: IEEE 1588-2008 Advanced
> Timestamp supported
> [  140.941109] stmmaceth :00:14.6 eth0: registered PTP clock
> [  140.953626] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
> udhcpc: sending discover
> [  142.979557] stmmaceth :00:14.6 eth0: Link is Up - 100Mbps/Full
> - flow control off
> [  142.988756] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
> [  142.998810] BUG: unable to handle kernel NULL pointer dereference at   
> (null)
> [  143.006193] IP: stmmac_xmit+0xf1/0x1080
> [  143.010168] *pde = 
> [  143.010177]
> [  143.014762] Oops: 0002 [#1]
> [  143.017672] Modules linked in: at24 nvmem_core pwm_pca9685
> [  143.023338] CPU: 0 PID: 0 Comm: swapper Not tainted 4.11.0-next-20170508+ 
> #2
> [  143.030539] task: c8533580 task.stack: c852c000
> [  143.035237] EIP: stmmac_xmit+0xf1/0x1080
> [  143.039302] EFLAGS: 00010216 CPU: 0
> [  143.042915] EAX:  EBX: 0050 ECX:  EDX: ceb6a0c0
> [  143.049326] ESI:  EDI: cdd16000 EBP: cdc25d70 ESP: cdc25d20
> [  143.055735]  DS: 007b ES: 007b FS:  GS:  SS: 0068
> [  143.061271] CR0: 80050033 CR2:  CR3: 0eb5c000 CR4: 00100010
> [  143.067671] Call Trace:
> [  143.070238]  
> [  143.072763]  dev_hard_start_xmit+0x7c/0x1a0
> [  143.077120]  sch_direct_xmit+0xf0/0x120
> [  143.081130]  __dev_queue_xmit+0x181/0x430
> [  143.085311]  ? eth_commit_mac_addr_change+0x20/0x20
> [  143.090362]  dev_queue_xmit+0xa/0x10
> [  143.094100]  neigh_resolve_output+0xdb/0x190
> [  143.098561]  ip6_finish_output2+0x184/0x500
> [  143.102945]  ip6_finish_output+0x91/0xe0
> [  143.107057]  ? ip6_finish_output+0x91/0xe0
> [  143.111338]  ip6_output+0x36/0x110
> [  143.114924]  ? ip6_fragment+0xb00/0xb00
> [  143.118935]  mld_sendpack+0x191/0x2b0
> [  143.122769]  ?

Re: [PATCH v3 net-next 01/11] net: stmmac: prepare dma op mode config for multiple queues

2017-05-08 Thread Andy Shevchenko 
On Mon, May 8, 2017 at 9:56 AM, Jan Kiszka  wrote:
> On 2017-03-15 12:04, Joao Pinto wrote:
>> This patch prepares DMA Operation Mode configuration for multiple queues.
>> The work consisted on breaking the DMA operation Mode configuration function
>> into RX and TX scope and adapting its mechanism in stmmac_main.

> Starting with this patch, the stmmac-based network adapters of the Intel
> Quark SoC stop working. I'm getting an IP via DHCP, I can ping, but TCP
> connections can no longer be established.
>
> Moving on a few patches (didn't bisect the exact one yet), the TX
> watchdog starts to fire, and DHCP fails completely. And if I go to
> current master in Linus tree (reverting an unrelated boot regression), I
> even get a crash in stmmac_xmit.
>
> Here are some details about the hw from dma_cap POV, if this helps:
>
> ==
> DMA HW features
> ==
> 10/100 Mbps: Y
> 1000 Mbps: N
> Half duplex: Y
> Hash Filter: Y
> Multiple MAC address registers: N
> PCS (TBI/SGMII/RTBI PHY interfaces): N
> SMA (MDIO) Interface: Y
> PMT Remote wake up: N
> PMT Magic Frame: N
> RMON module: Y
> IEEE 1588-2002 Time Stamp: N
> IEEE 1588-2008 Advanced Time Stamp: Y
> 802.3az - Energy-Efficient Ethernet (EEE): N
> AV features: N
> Checksum Offload in TX: Y
> IP Checksum Offload (type1) in RX: N
> IP Checksum Offload (type2) in RX: Y
> RXFIFO > 2048bytes: Y
> Number of Additional RX channel: 0
> Number of Additional TX channel: 0
> Enhanced descriptors: Y
>
> Given the number of different failure modes, my feeling is that there
> are multiple regressions coming with these patches...
>
> I've tested on the IOT2000 board, but I suspect the Galileo Gen2 will be
> affected equally. If you don't have access to any such device, let me
> know what I can debug for you.

JFYI: With today's linux-next when _kexec:ed_ kernel boots I tried and
got the following:


# ip a s
1: lo:  mtu 65536 qdisc noqueue qlen 1000
   link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
   inet 127.0.0.1/8 scope host lo
  valid_lft forever preferred_lft forever
   inet6 ::1/128 scope host
  valid_lft forever preferre[  130.403995] random: fast init done
d_lft forever
2: eth0:  mtu 1500 qdisc noop qlen 1000
   link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff
3: eth1:  mtu 1500 qdisc noop qlen 1000
   link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff
4: sit0@NONE:  mtu 1480 qdisc noop qlen 1000
   link/sit 0.0.0.0 brd 0.0.0.0
# udhcpc -i eth0
udhcpc: started, v1.26.2
[  140.825131] stmmaceth :00:14.6 eth0: device MAC address 98:4f:ee:05:ac:47
[  140.834304] Generic PHY stmmac-a6:01: attached PHY driver [Generic
PHY] (mii_bus:phy_addr=stmmac-a6:01, irq=-1)
[  140.930871] stmmaceth :00:14.6 eth0: IEEE 1588-2008 Advanced
Timestamp supported
[  140.941109] stmmaceth :00:14.6 eth0: registered PTP clock
[  140.953626] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
udhcpc: sending discover
[  142.979557] stmmaceth :00:14.6 eth0: Link is Up - 100Mbps/Full
- flow control off
[  142.988756] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
[  142.998810] BUG: unable to handle kernel NULL pointer dereference at   (null)
[  143.006193] IP: stmmac_xmit+0xf1/0x1080
[  143.010168] *pde = 
[  143.010177]
[  143.014762] Oops: 0002 [#1]
[  143.017672] Modules linked in: at24 nvmem_core pwm_pca9685
[  143.023338] CPU: 0 PID: 0 Comm: swapper Not tainted 4.11.0-next-20170508+ #2
[  143.030539] task: c8533580 task.stack: c852c000
[  143.035237] EIP: stmmac_xmit+0xf1/0x1080
[  143.039302] EFLAGS: 00010216 CPU: 0
[  143.042915] EAX:  EBX: 0050 ECX:  EDX: ceb6a0c0
[  143.049326] ESI:  EDI: cdd16000 EBP: cdc25d70 ESP: cdc25d20
[  143.055735]  DS: 007b ES: 007b FS:  GS:  SS: 0068
[  143.061271] CR0: 80050033 CR2:  CR3: 0eb5c000 CR4: 00100010
[  143.067671] Call Trace:
[  143.070238]  
[  143.072763]  dev_hard_start_xmit+0x7c/0x1a0
[  143.077120]  sch_direct_xmit+0xf0/0x120
[  143.081130]  __dev_queue_xmit+0x181/0x430
[  143.085311]  ? eth_commit_mac_addr_change+0x20/0x20
[  143.090362]  dev_queue_xmit+0xa/0x10
[  143.094100]  neigh_resolve_output+0xdb/0x190
[  143.098561]  ip6_finish_output2+0x184/0x500
[  143.102945]  ip6_finish_output+0x91/0xe0
[  143.107057]  ? ip6_finish_output+0x91/0xe0
[  143.111338]  ip6_output+0x36/0x110
[  143.114924]  ? ip6_fragment+0xb00/0xb00
[  143.118935]  mld_sendpack+0x191/0x2b0
[  143.122769]  ? mld_newpack+0xda/0x180
[  143.126598]  ? ipv6_icmp_sysctl_init+0x30/0x30
[  143.131224]  mld_ifc_timer_expire+0x158/0x240
[  143.135756]  ? find_next_bit+0xa/0x10
[  143.139584]  ? mld_dad_timer_expire+0x50/0x50
[  143.144112]  call_timer_fn+0x2a/0xf0
[

Re: [PATCH v3 net-next 01/11] net: stmmac: prepare dma op mode config for multiple queues

2017-05-08 Thread Jan Kiszka 
On 2017-03-15 12:04, Joao Pinto wrote:
> This patch prepares DMA Operation Mode configuration for multiple queues.
> The work consisted on breaking the DMA operation Mode configuration function
> into RX and TX scope and adapting its mechanism in stmmac_main.
> 
> Signed-off-by: Joao Pinto 
> ---
> changes v1->v3:
> - Just to keep up the patch-set version
> 
>  drivers/net/ethernet/stmicro/stmmac/common.h  |   3 +
>  drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c  | 118 
> +++---
>  drivers/net/ethernet/stmicro/stmmac/stmmac_main.c |  82 +++
>  3 files changed, 124 insertions(+), 79 deletions(-)
> 
> diff --git a/drivers/net/ethernet/stmicro/stmmac/common.h 
> b/drivers/net/ethernet/stmicro/stmmac/common.h
> index 9f0d26d..13bd3d4 100644
> --- a/drivers/net/ethernet/stmicro/stmmac/common.h
> +++ b/drivers/net/ethernet/stmicro/stmmac/common.h
> @@ -424,6 +424,9 @@ struct stmmac_dma_ops {
>* An invalid value enables the store-and-forward mode */
>   void (*dma_mode)(void __iomem *ioaddr, int txmode, int rxmode,
>int rxfifosz);
> + void (*dma_rx_mode)(void __iomem *ioaddr, int mode, u32 channel,
> + int fifosz);
> + void (*dma_tx_mode)(void __iomem *ioaddr, int mode, u32 channel);
>   /* To track extra statistic (if supported) */
>   void (*dma_diagnostic_fr) (void *data, struct stmmac_extra_stats *x,
>  void __iomem *ioaddr);
> diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c 
> b/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c
> index 6ac6b26..6285e8a 100644
> --- a/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c
> +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c
> @@ -182,70 +182,26 @@ static void dwmac4_rx_watchdog(void __iomem *ioaddr, 
> u32 riwt)
>   writel(riwt, ioaddr + DMA_CHAN_RX_WATCHDOG(i));
>  }
>  
> -static void dwmac4_dma_chan_op_mode(void __iomem *ioaddr, int txmode,
> - int rxmode, u32 channel, int rxfifosz)
> +static void dwmac4_dma_rx_chan_op_mode(void __iomem *ioaddr, int mode,
> +u32 channel, int fifosz)
>  {
> - unsigned int rqs = rxfifosz / 256 - 1;
> - u32 mtl_tx_op, mtl_rx_op, mtl_rx_int;
> -
> - /* Following code only done for channel 0, other channels not yet
> -  * supported.
> -  */
> - mtl_tx_op = readl(ioaddr + MTL_CHAN_TX_OP_MODE(channel));
> -
> - if (txmode == SF_DMA_MODE) {
> - pr_debug("GMAC: enable TX store and forward mode\n");
> - /* Transmit COE type 2 cannot be done in cut-through mode. */
> - mtl_tx_op |= MTL_OP_MODE_TSF;
> - } else {
> - pr_debug("GMAC: disabling TX SF (threshold %d)\n", txmode);
> - mtl_tx_op &= ~MTL_OP_MODE_TSF;
> - mtl_tx_op &= MTL_OP_MODE_TTC_MASK;
> - /* Set the transmit threshold */
> - if (txmode <= 32)
> - mtl_tx_op |= MTL_OP_MODE_TTC_32;
> - else if (txmode <= 64)
> - mtl_tx_op |= MTL_OP_MODE_TTC_64;
> - else if (txmode <= 96)
> - mtl_tx_op |= MTL_OP_MODE_TTC_96;
> - else if (txmode <= 128)
> - mtl_tx_op |= MTL_OP_MODE_TTC_128;
> - else if (txmode <= 192)
> - mtl_tx_op |= MTL_OP_MODE_TTC_192;
> - else if (txmode <= 256)
> - mtl_tx_op |= MTL_OP_MODE_TTC_256;
> - else if (txmode <= 384)
> - mtl_tx_op |= MTL_OP_MODE_TTC_384;
> - else
> - mtl_tx_op |= MTL_OP_MODE_TTC_512;
> - }
> - /* For an IP with DWC_EQOS_NUM_TXQ == 1, the fields TXQEN and TQS are RO
> -  * with reset values: TXQEN on, TQS == DWC_EQOS_TXFIFO_SIZE.
> -  * For an IP with DWC_EQOS_NUM_TXQ > 1, the fields TXQEN and TQS are R/W
> -  * with reset values: TXQEN off, TQS 256 bytes.
> -  *
> -  * Write the bits in both cases, since it will have no effect when RO.
> -  * For DWC_EQOS_NUM_TXQ > 1, the top bits in MTL_OP_MODE_TQS_MASK might
> -  * be RO, however, writing the whole TQS field will result in a value
> -  * equal to DWC_EQOS_TXFIFO_SIZE, just like for DWC_EQOS_NUM_TXQ == 1.
> -  */
> - mtl_tx_op |= MTL_OP_MODE_TXQEN | MTL_OP_MODE_TQS_MASK;
> - writel(mtl_tx_op, ioaddr +  MTL_CHAN_TX_OP_MODE(channel));
> + unsigned int rqs = fifosz / 256 - 1;
> + u32 mtl_rx_op, mtl_rx_int;
>  
>   mtl_rx_op = readl(ioaddr + MTL_CHAN_RX_OP_MODE(channel));
>  
> - if (rxmode == SF_DMA_MODE) {
> + if (mode == SF_DMA_MODE) {
>   pr_debug("GMAC: enable RX store and forward mode\n");
>   mtl_rx_op |= MTL_OP_MODE_RSF;
>   } else {
> - pr_debug("GMAC: disable RX SF mode (threshold %d)\n", rxmode);
> + pr_debug("GMAC: disable RX SF mode (threshold