[PATCH 0/3] vhost_user: configure memory slots

[Maxime Coquelin]

On 8/29/25 12:26 PM, Bathija, Pravin wrote:
> Answers inline.
> 
> 
> Internal Use - Confidential
>> -----Original Message-----
>> From: Maxime Coquelin <maxime.coquelin at redhat.com>
>> Sent: Friday, August 29, 2025 2:18 AM
>> To: Bathija, Pravin <Pravin.Bathija at dell.com>; Bruce Richardson
>> <bruce.richardson at intel.com>; David Marchand
>> <david.marchand at redhat.com>; Chenbo Xia <chenbox at nvidia.com>
>> Cc: dev at dpdk.org; pravin.m.bathija.dev at gmail.com; Thomas Monjalon
>> <thomas at monjalon.net>
>> Subject: Re: [PATCH 0/3] vhost_user: configure memory slots
>>
>>
>> [EXTERNAL EMAIL]
>>
>> Hi Pravin,
>>
>> On 8/29/25 11:11 AM, Bathija, Pravin wrote:
>>> Dear Maxime, Bruce, David, Chenbo,
>>>
>>> Could you please help review the code ? The goal is to merge the changes
>> before September 18 when the merge window closes for SPDK v25.09.
>>
>> I will help review the code, but there are no guarantee it will be merged in main
>> branch before September 18th.
> Thank you for your consideration.
> 
>>
>> Can you elaborate about the tests you have run?
> - Run I/O tests using fio and libblkio engine to SPDK vhost user block targets with multiple targets and different block sizes
> - Custom system tests defined within SPDK framework
> https://review.spdk.io/c/spdk/spdk/+/26395
> - QEMU bring up with SPDK Virtio block device
> 
>> Have you ensured it is still working properly with Vhsot-user/Virtio-net?
> Yes tests have been run to ensure this as listed above

You have only run storage tests, not networking test.
So the DPDK Vhost datapath has not been exercised, neither the IOMMU AFAICT.

Maxime

> 
>>
>> Maxime
>>
>>> Regards,
>>> Pravin
>>>
>>>
>>> Internal Use - Confidential
>>>> -----Original Message-----
>>>> From: Bathija, Pravin
>>>> Sent: Tuesday, August 26, 2025 11:47 AM
>>>> To: Maxime Coquelin <maxime.coquelin at redhat.com>; Bruce Richardson
>>>> <bruce.richardson at intel.com>; David Marchand
>>>> <david.marchand at redhat.com>; Chenbo Xia <chenbox at nvidia.com>
>>>> Cc: dev at dpdk.org; pravin.m.bathija.dev at gmail.com; Thomas Monjalon
>>>> <thomas at monjalon.net>
>>>> Subject: RE: [PATCH 0/3] vhost_user: configure memory slots
>>>>
>>>> Answers inline.
>>>>
>>>>> -----Original Message-----
>>>>> From: Maxime Coquelin <maxime.coquelin at redhat.com>
>>>>> Sent: Monday, August 25, 2025 1:49 AM
>>>>> To: Bathija, Pravin <Pravin.Bathija at dell.com>; Bruce Richardson
>>>>> <bruce.richardson at intel.com>; David Marchand
>>>>> <david.marchand at redhat.com>; Chenbo Xia <chenbox at nvidia.com>
>>>>> Cc: dev at dpdk.org; pravin.m.bathija.dev at gmail.com; Thomas Monjalon
>>>>> <thomas at monjalon.net>
>>>>> Subject: Re: [PATCH 0/3] vhost_user: configure memory slots
>>>>>
>>>>>
>>>>> [EXTERNAL EMAIL]
>>>>>
>>>>> Hi,
>>>>>
>>>>> Please try to fix your email client so that we can follow the replies inlined.
>>>>
>>>> Thank you for the input. I have made the necessary adjustments.
>>>> Please let me know if there is anything missing.
>>>>
>>>>>
>>>>> On 8/22/25 9:33 AM, Bathija, Pravin wrote:
>>>>>> Dear Maxime, Bruce, David, Chenbo,
>>>>>>
>>>>>> I would like to follow up on the status of the patches and when we
>>>>>> should
>>>>> expect these to be merged into the mainline.
>>>>>>
>>>>>> Please let me know if I can help answer any questions you might have.
>>>>>>
>>>>>> Regards,
>>>>>> Pravin
>>>>>>
>>>>>>
>>>>>> Internal Use - Confidential
>>>>>> -----Original Message-----
>>>>>> From: Bathija, Pravin
>>>>>> Sent: Thursday, August 21, 2025 7:48 PM
>>>>>> To: Thomas Monjalon <thomas at monjalon.net>
>>>>>> Cc: dev at dpdk.org; pravin.m.bathija.dev at gmail.com; Maxime Coquelin
>>>>>> <maxime.coquelin at redhat.com>; Chenbo Xia <chenbox at nvidia.com>;
>>>> David
>>>>>> Marchand <david.marchand at redhat.com>; Bruce Richardson
>>>>>> <bruce.richardson at intel.com>
>>>>>> Subject: RE: [PATCH 0/3] vhost_user: configure memory slots
>>>>>>
>>>>>> Comments inline:
>>>>>>
>>>>>> -----Original Message-----
>>>>>> From: Thomas Monjalon <thomas at monjalon.net>
>>>>>> Sent: Tuesday, August 19, 2025 4:36 AM
>>>>>> To: Bathija, Pravin <Pravin.Bathija at dell.com>
>>>>>> Cc: dev at dpdk.org; pravin.m.bathija.dev at gmail.com; Maxime Coquelin
>>>>>> <maxime.coquelin at redhat.com>; Chenbo Xia <chenbox at nvidia.com>;
>>>> David
>>>>>> Marchand <david.marchand at redhat.com>; Bruce Richardson
>>>>>> <bruce.richardson at intel.com>
>>>>>> Subject: Re: [PATCH 0/3] vhost_user: configure memory slots
>>>>>>
>>>>>>
>>>>>> [EXTERNAL EMAIL]
>>>>>>
>>>>>> Comments inline
>>>>>>
>>>>>> 12/08/2025 04:33, Pravin M Bathija:
>>>>>>> - add protocol feature support add and removal of memory
>>>>>>> regions/slots dynamically.
>>>>>>> - this patch set has been tested with multiple instances of fio
>>>>>>
>>>>>> What is "fio"?
>>>>>> Fio is an open-source tool to test I/O workloads on disks and file systems.
>>>>>>
>>>>>>> connected via libblkio to spdk target backed unix domain socket.
>>>>>>> - please apply this this patch-set to dpdk mainline as well as
>>>>>>> version 25.03.0.
>>>>>>
>>>>>> As mentionned previously, we don't plan to maintain 25.03.
>>>>>> I understand you need it for SPDK,
>>>>>> but I think SPDK should rely only on LTS branches.
>>>>>> Is there a specific reason why 25.03 is required in SPDK?
>>>>>>
>>>>>> When we do sub-module init on spdk master, it uses dpdk version
>>>>>> 25.03 from
>>>>> it's own sub-repo.
>>>>>> Our changes for vhost have already been pushed to spdk master. We
>>>>>> have also pushed the same dpdk changes in the spdk sub-repo "dpdk"
>>>>>> which is version 25.03. I was just ensuring that there is
>>>>>> uniformity in the 25.03
>>>>> codebase, whether it is on the spdk sub-repo or 25.03 within dpdk repo.
>>>>>
>>>>> Ok, still, it would be better for SPDK to uses LTS as we have no
>>>>> plans to add extra stable branches.
>>>>>
>>>>> Also, your series may be considered as a feature and not a bug fix,
>>>>> so even if
>>>>> 25.03 would be a stable branch (which it is not), it would not be
>>>>> backported against it.
>>>>
>>>> Understood. I would request that we merge my changes into DPDK
>>>> mainline before September 18 when the merge window closes for SPDK
>> v25.09.
>>>>
>>>>>
>>>>>>
>>>>>>
>>>>>>> v2:
>>>>>>> - fixed warnings with e-mail address
>>>>>>> - fixed warnings with typos
>>>>>>> - there is a warning about __rte_packed_begin not being in pair
>>>>>>> with __rte_packed_end which on reading the code turns out to be
>>>>>>> false. It is an anomaly of how git-format generates the patch and
>>>>>>> how check-patch checks it.
>>>>>>
>>>>>> When sending a v2, please use -v2 so it is reflected in the email subject.
>>>>>> I will remember this for next time. Thank you for pointing that out.
>>>>>>
>>>>>>> Pravin M Bathija (3):
>>>>>>>      mailmap: add user
>>>>>>>      vhost_user: configure memory slots
>>>>>>>      vhost_user: support for memory regions
>>>>>>>
>>>>>>>     .mailmap               |   1 +
>>>>>>>     lib/vhost/rte_vhost.h  |   4 +
>>>>>>>     lib/vhost/vhost_user.c | 325
>>>>>>> +++++++++++++++++++++++++++++++++++------
>>>>>>>     lib/vhost/vhost_user.h |  10 ++
>>>>>>>     4 files changed, 296 insertions(+), 44 deletions(-)
>>>>>>
>>>>>>
>>>>>>
>>>
>