[stable request <= 3.11] net/mlx4_en: Fix BlueFlame race

[stable request <= 3.11] net/mlx4_en: Fix BlueFlame race

[Vinson Lee]

Hi.

Please consider backporting upstream commit
2d4b646613d6b12175b017aca18113945af1faf3 "net/mlx4_en: Fix BlueFlame
race" to stable kernels <= 3.11.

commit 2d4b646613d6b12175b017aca18113945af1faf3
Author: Eugenia Emantayev <eugenia@mellanox.com>
Date:   Thu Jul 25 19:21:23 2013 +0300

    net/mlx4_en: Fix BlueFlame race

    Fix a race between BlueFlame flow and stamping in post send flow.
    Example:
        SW: Build WQE 0 on the TX buffer, except the ownership bit
        SW: Set ownership for WQE 0 on the TX buffer
        SW: Ring doorbell for WQE 0
        SW: Build WQE 1 on the TX buffer, except the ownership bit
        SW: Set ownership for WQE 1 on the TX buffer
        HW: Read WQE 0 and then WQE 1, before doorbell was rung/BF was
done for WQE 1
        HW: Produce CQEs for WQE 0 and WQE 1
        SW: Process the CQEs, and stamp WQE 0 and WQE 1 accordingly
(on the TX buffer)
        SW: Copy WQE 1 from the TX buffer to the BF register - ALREADY STAMPED!
        HW: CQE error with index 0xFFFF  - the BF WQE's control
segment is STAMPED,
                so the BF index is 0xFFFF. Error: Invalid Opcode.
    As a result QP enters the error state and no traffic can be sent.

    Solution:
    When stamping - do not stamp last completed wqe.

    Signed-off-by: Eugenia Emantayev <eugenia@mellanox.com>
    Signed-off-by: Amir Vadai <amirv@mellanox.com>
    Signed-off-by: David S. Miller <davem@davemloft.net>

This patch fixes Mellanox network card packet send failures on Linux
kernel 3.10.

Tested-by: Chris Suttles <csuttles@twitter.com>
Tested-by: David Mackey <tdmackey@twitter.com>
Tested-by: Matthew Mullins <mmullins@twopensource.com>
Tested-by: Vinson Lee <vlee@twitter.com>

Cheers,
Vinson

Re: [stable request <= 3.11] net/mlx4_en: Fix BlueFlame race

[Cong Wang]

On Sat, Oct 18, 2014 at 2:14 PM, Vinson Lee <vlee@twopensource.com> wrote:
> Hi.
>
> Please consider backporting upstream commit
> 2d4b646613d6b12175b017aca18113945af1faf3 "net/mlx4_en: Fix BlueFlame
> race" to stable kernels <= 3.11.
>

David, could you take care of it if you have time? It fixes a real
bug in production. :)

Thanks a lot!

Re: [stable request <= 3.11] net/mlx4_en: Fix BlueFlame race

[Or Gerlitz]

On Wed, Oct 22, 2014 at 2:15 AM, Cong Wang <cwang@twopensource.com> wrote:
> On Sat, Oct 18, 2014 at 2:14 PM, Vinson Lee <vlee@twopensource.com> wrote:
>> Hi.
>>
>> Please consider backporting upstream commit
>> 2d4b646613d6b12175b017aca18113945af1faf3 "net/mlx4_en: Fix BlueFlame
>> race" to stable kernels <= 3.11.
>>
>
> David, could you take care of it if you have time? It fixes a real
> bug in production. :)

Let out folks here look on that 1st.

Or.

Re: [stable request <= 3.11] net/mlx4_en: Fix BlueFlame race

[Amir Vadai]

On 10/22/2014 7:17 AM, Or Gerlitz wrote:
> On Wed, Oct 22, 2014 at 2:15 AM, Cong Wang <cwang@twopensource.com> wrote:
>> On Sat, Oct 18, 2014 at 2:14 PM, Vinson Lee <vlee@twopensource.com> wrote:
>>> Hi.
>>>
>>> Please consider backporting upstream commit
>>> 2d4b646613d6b12175b017aca18113945af1faf3 "net/mlx4_en: Fix BlueFlame
>>> race" to stable kernels <= 3.11.
>>>
>>
>> David, could you take care of it if you have time? It fixes a real
>> bug in production. :)
> 
> Let out folks here look on that 1st.
> 
> Or.
> 
Verified patch above [1] on 3.10.y (v3.10.58) as is.

Dave, please pull it to stable 3.10

I don't know about an LTS 3.11 kernel - but if there is such an option,
I verified that the patch also applies cleanly on 3.11.y.

Thanks,
Amir

[1] - 2d4b646 ("net/mlx4_en: Fix BlueFlame race")

Re: [stable request <= 3.11] net/mlx4_en: Fix BlueFlame race

[Ben Hutchings]

[-- Attachment #1: Type: text/plain, Size: 753 bytes --]

On Sat, 2014-10-18 at 14:14 -0700, Vinson Lee wrote:
> Hi.
> 
> Please consider backporting upstream commit
> 2d4b646613d6b12175b017aca18113945af1faf3 "net/mlx4_en: Fix BlueFlame
> race" to stable kernels <= 3.11.
> 
> commit 2d4b646613d6b12175b017aca18113945af1faf3
> Author: Eugenia Emantayev <eugenia@mellanox.com>
> Date:   Thu Jul 25 19:21:23 2013 +0300
> 
>     net/mlx4_en: Fix BlueFlame race
[...]

David doesn't support anything older than 3.12 now, so you should send
requests for older stable branches directly to the stable list.

I tried to apply this one to 3.2-stable, but it failed.  If it's needed
there you'll have to provide a backport.

Ben.

-- 
Ben Hutchings
The world is coming to an end.	Please log off.

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 811 bytes --]

Re: [stable request <= 3.11] net/mlx4_en: Fix BlueFlame race

[Cong Wang]

On Fri, Oct 31, 2014 at 6:29 PM, Ben Hutchings <ben@decadent.org.uk> wrote:
> On Sat, 2014-10-18 at 14:14 -0700, Vinson Lee wrote:
>> Hi.
>>
>> Please consider backporting upstream commit
>> 2d4b646613d6b12175b017aca18113945af1faf3 "net/mlx4_en: Fix BlueFlame
>> race" to stable kernels <= 3.11.
>>
>> commit 2d4b646613d6b12175b017aca18113945af1faf3
>> Author: Eugenia Emantayev <eugenia@mellanox.com>
>> Date:   Thu Jul 25 19:21:23 2013 +0300
>>
>>     net/mlx4_en: Fix BlueFlame race
> [...]
>
> David doesn't support anything older than 3.12 now, so you should send
> requests for older stable branches directly to the stable list.

Interesting, is there a document saying which stable kernels netdev
supports? and how are they chosen?

I thought netdev supports all -stable maintainers support.

Anyway, we will send the request directly to Greg then.

Thanks.

Re: [stable request <= 3.11] net/mlx4_en: Fix BlueFlame race

[David Miller]

From: Cong Wang <cwang@twopensource.com>
Date: Sat, 1 Nov 2014 10:17:43 -0700

> On Fri, Oct 31, 2014 at 6:29 PM, Ben Hutchings <ben@decadent.org.uk> wrote:
>> On Sat, 2014-10-18 at 14:14 -0700, Vinson Lee wrote:
>>> Hi.
>>>
>>> Please consider backporting upstream commit
>>> 2d4b646613d6b12175b017aca18113945af1faf3 "net/mlx4_en: Fix BlueFlame
>>> race" to stable kernels <= 3.11.
>>>
>>> commit 2d4b646613d6b12175b017aca18113945af1faf3
>>> Author: Eugenia Emantayev <eugenia@mellanox.com>
>>> Date:   Thu Jul 25 19:21:23 2013 +0300
>>>
>>>     net/mlx4_en: Fix BlueFlame race
>> [...]
>>
>> David doesn't support anything older than 3.12 now, so you should send
>> requests for older stable branches directly to the stable list.
> 
> Interesting, is there a document saying which stable kernels netdev
> supports? and how are they chosen?
> 
> I thought netdev supports all -stable maintainers support.
> 
> Anyway, we will send the request directly to Greg then.

There is no documented way nor do I wish to state anything so strictly.
I want maximum flexibility for such a time consuming task.

I tend to go back 3 or 4 releases at most, and it really depends upon
the difficulty of the backports and my own time constraints.

Re: [stable request <= 3.11] net/mlx4_en: Fix BlueFlame race

[Cong Wang]

On Sat, Nov 1, 2014 at 10:41 AM, David Miller <davem@davemloft.net> wrote:
>
> There is no documented way nor do I wish to state anything so strictly.
> I want maximum flexibility for such a time consuming task.
>
> I tend to go back 3 or 4 releases at most, and it really depends upon
> the difficulty of the backports and my own time constraints.

You should really offload to developers, otherwise too much work for you. :)

Thanks!

Re: [stable request <= 3.11] net/mlx4_en: Fix BlueFlame race

[David Miller]

From: Cong Wang <cwang@twopensource.com>
Date: Mon, 3 Nov 2014 09:22:18 -0800

> On Sat, Nov 1, 2014 at 10:41 AM, David Miller <davem@davemloft.net> wrote:
>>
>> There is no documented way nor do I wish to state anything so strictly.
>> I want maximum flexibility for such a time consuming task.
>>
>> I tend to go back 3 or 4 releases at most, and it really depends upon
>> the difficulty of the backports and my own time constraints.
> 
> You should really offload to developers, otherwise too much work for you. :)

That's exactly what I am doing by having the -stable maintainers for older
releases deal with the backports and other pains.

Re: [stable request <= 3.11] net/mlx4_en: Fix BlueFlame race

[Vinson Lee]

On Mon, Nov 3, 2014 at 12:09 PM, David Miller <davem@davemloft.net> wrote:
> From: Cong Wang <cwang@twopensource.com>
> Date: Mon, 3 Nov 2014 09:22:18 -0800
>
>> On Sat, Nov 1, 2014 at 10:41 AM, David Miller <davem@davemloft.net> wrote:
>>>
>>> There is no documented way nor do I wish to state anything so strictly.
>>> I want maximum flexibility for such a time consuming task.
>>>
>>> I tend to go back 3 or 4 releases at most, and it really depends upon
>>> the difficulty of the backports and my own time constraints.
>>
>> You should really offload to developers, otherwise too much work for you. :)
>
> That's exactly what I am doing by having the -stable maintainers for older
> releases deal with the backports and other pains.

+ stable

Please backport upstream commit
2d4b646613d6b12175b017aca18113945af1faf3 "net/mlx4_en: Fix BlueFlame
race" to stable kernels. David is no longer supporting 3.10 and older.

Please refer to http://marc.info/?l=linux-netdev&m=141366689811844 for
previous comments in thread.

Cheers,
Vinson

Re: [stable request <= 3.11] net/mlx4_en: Fix BlueFlame race

[Greg KH]

On Wed, Nov 05, 2014 at 11:38:37AM -0800, Vinson Lee wrote:
> On Mon, Nov 3, 2014 at 12:09 PM, David Miller <davem@davemloft.net> wrote:
> > From: Cong Wang <cwang@twopensource.com>
> > Date: Mon, 3 Nov 2014 09:22:18 -0800
> >
> >> On Sat, Nov 1, 2014 at 10:41 AM, David Miller <davem@davemloft.net> wrote:
> >>>
> >>> There is no documented way nor do I wish to state anything so strictly.
> >>> I want maximum flexibility for such a time consuming task.
> >>>
> >>> I tend to go back 3 or 4 releases at most, and it really depends upon
> >>> the difficulty of the backports and my own time constraints.
> >>
> >> You should really offload to developers, otherwise too much work for you. :)
> >
> > That's exactly what I am doing by having the -stable maintainers for older
> > releases deal with the backports and other pains.
> 
> + stable
> 
> Please backport upstream commit
> 2d4b646613d6b12175b017aca18113945af1faf3 "net/mlx4_en: Fix BlueFlame
> race" to stable kernels. David is no longer supporting 3.10 and older.
> 
> Please refer to http://marc.info/?l=linux-netdev&m=141366689811844 for
> previous comments in thread.

Now applied, thanks.

greg k-h