[dpdk-dev] [PATCH v5] net/memif: zero-copy slave
Ferruh Yigit
ferruh.yigit at intel.com
Tue Oct 15 18:59:12 CEST 2019
On 10/4/2019 2:23 PM, Ferruh Yigit wrote:
> On 8/22/2019 9:18 AM, Jakub Grajciar wrote:
>> Zero-copy slave support for memif PMD.
>> Slave interface exposes DPDK memory to
>> master interface. Only single file segments
>> are supported (EAL option --single-file-segments).
>>
>> Signed-off-by: Jakub Grajciar <jgrajcia at cisco.com>
>> ---
>> doc/guides/nics/memif.rst | 42 +-
>> drivers/net/memif/Makefile | 1 +
>> drivers/net/memif/memif_socket.c | 64 +--
>> drivers/net/memif/meson.build | 1 +
>> drivers/net/memif/rte_eth_memif.c | 449 +++++++++++++++++-
>> drivers/net/memif/rte_eth_memif.h | 11 +-
>> lib/librte_eal/common/eal_common_mcfg.c | 7 +
>> .../common/include/rte_eal_memconfig.h | 10 +
>> lib/librte_eal/rte_eal_version.map | 1 +
>> 9 files changed, 513 insertions(+), 73 deletions(-)
>>
>> V2:
>> - fix coding style
>>
>> V3:
>> - fix compilation issues
>>
>> V4:
>> - don't move existing code
>> - add new EAL API rte_mcfg_get_single_file_segments,
>> mem_config is now private, this api returns
>> single_file_segments parameter value
>>
>> V5:
>> - explain single file segments limitation
>> - add zero-copy slave example
>
> Overall looks good, but I had to test this by manually modifying the PMD for the
> bind() error.
>
> I am for first fixing the PMD bind() issue before getting this patch, fyi.
>
Hi Jakub,
Just to double check if anyone is looking into the bind() error issue which is
since following commit, I am waiting for more input on it.
Commit b923866c6974 ("net/memif: allow for full key size in socket name")
Cc: stephen at networkplumber.org
More information about the dev
mailing list