bug#26257: [PATCH] gnu: Add rdma-core.

bug#26257: [PATCH] gnu: Add rdma-core.

[Ludovic Courtès]

Marius Bakke <mbakke@fastmail.com> skribis:

> * gnu/packages/linux.scm (rdma-core): New variable.

LGTM!

> +    (home-page "https://github.com/linux-rdma/rdma-core")
> +    (synopsis "Utilities and libraries for working with RDMA devices")
> +    (description
> +     "This package provides the userspace components for the InfiniBand
> +subsystem of the Linux kernel.  Specifically it contains userspace libraries
> +for the following device nodes:
> +
> +@enumerate
> +@item @file{/dev/infiniband/uverbsX} (@code{libibverbs})
> +@item @file{/dev/infiniband/rdma_cm} (@code{librdmacm})
> +@item @file{/dev/infiniband/umadX} (@code{libibumad})
> +@end enumerate
> +
> +The following service daemons are also provided:
> +@enumerate
> +@item @code{srp_daemon} (for the @code{ib_srp} kernel module)
> +@item @code{iwpmd} (for iwarp kernel providers)
> +@item @code{ibacm} (for InfiniBand communication management assistant)
> +@end enumerate")

I was wondering whether this code was relying on proprietary InfiniBand
drivers, which would not be okay, but

  find $LINUX_MODULE_DIRECTORY -name ib\*ko

shows a bunch of drivers, including ‘ib_srp’ (mentioned above), so I
guess this is fine.

Thanks!

Ludo’.

bug#26257: [PATCH] gnu: Add rdma-core.

[Marius Bakke]

[-- Attachment #1: Type: text/plain, Size: 183 bytes --]

Ludovic Courtès <ludo@gnu.org> writes:

> Marius Bakke <mbakke@fastmail.com> skribis:
>
>> * gnu/packages/linux.scm (rdma-core): New variable.
>
> LGTM!

Closing, thanks!


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 487 bytes --]