Re: [PATCH 1/4] net: thunderx: Fix IOMMU translation faults

From	Sunil Kovvuri <sunil.kovvuri@gmail.com>
Newsgroups	linux.kernel
Subject	Re: [PATCH 1/4] net: thunderx: Fix IOMMU translation faults
Date	2017-03-06 14:10 +0100
Message-ID	<ti0yK-9I-9@gated-at.bofh.it> (permalink)
References	<tgT6i-8tS-11@gated-at.bofh.it> <tgT6i-8tS-21@gated-at.bofh.it> <tgZEK-4Br-13@gated-at.bofh.it> <thaTv-4bC-1@gated-at.bofh.it> <ti0fn-8f0-17@gated-at.bofh.it>
Organization	linux.* mail to news gateway

Show all headers | View raw

>>
>> We are seeing a 0.75Mpps drop with IP forwarding rate due to that.
>> Hence I have restricted calling DMA interfaces to only when IOMMU is enabled.
>
> What's 0.07Mpps as a percentage of baseline? On a correctly configured
> coherent arm64 system, in the absence of an IOMMU, dma_map_*() is
> essentially just virt_to_phys() behind a function call or two, so I'd be
> interested to know where any non-trivial overhead might be coming from.

It's a 5% drop and yes device is configured as coherent.
And the drop is due to additional function calls.

Thanks,
Sunil.

Thread

[PATCH 1/4] net: thunderx: Fix IOMMU translation faults sunil.kovvuri@gmail.com - 2017-03-03 12:00 +0100
  Re: [PATCH 1/4] net: thunderx: Fix IOMMU translation faults David Miller <davem@davemloft.net> - 2017-03-03 19:00 +0100
    Re: [PATCH 1/4] net: thunderx: Fix IOMMU translation faults Sunil Kovvuri <sunil.kovvuri@gmail.com> - 2017-03-04 07:00 +0100
      Re: [PATCH 1/4] net: thunderx: Fix IOMMU translation faults Robin Murphy <robin.murphy@arm.com> - 2017-03-06 13:50 +0100
        Re: [PATCH 1/4] net: thunderx: Fix IOMMU translation faults Sunil Kovvuri <sunil.kovvuri@gmail.com> - 2017-03-06 14:10 +0100
          Re: [PATCH 1/4] net: thunderx: Fix IOMMU translation faults Robin Murphy <robin.murphy@arm.com> - 2017-03-06 17:40 +0100
            Re: [PATCH 1/4] net: thunderx: Fix IOMMU translation faults Sunil Kovvuri <sunil.kovvuri@gmail.com> - 2017-03-07 13:50 +0100

csiph-web