]> git.proxmox.com Git - mirror_ubuntu-eoan-kernel.git/commit
bpf: pass a struct with offload callbacks to bpf_offload_dev_create()
authorQuentin Monnet <quentin.monnet@netronome.com>
Fri, 9 Nov 2018 13:03:25 +0000 (13:03 +0000)
committerAlexei Starovoitov <ast@kernel.org>
Sat, 10 Nov 2018 23:39:53 +0000 (15:39 -0800)
commit1385d755cfb42f596ef1cf9f5c761010ff3b34e7
tree685f8d826e45d2e889feed5142d5c361847e5f5d
parent1da6f5733853fb230265fe1a7d1b5373f13bf5ca
bpf: pass a struct with offload callbacks to bpf_offload_dev_create()

For passing device functions for offloaded eBPF programs, there used to
be no place where to store the pointer without making the non-offloaded
programs pay a memory price.

As a consequence, three functions were called with ndo_bpf() through
specific commands. Now that we have struct bpf_offload_dev, and since
none of those operations rely on RTNL, we can turn these three commands
into hooks inside the struct bpf_prog_offload_ops, and pass them as part
of bpf_offload_dev_create().

This commit effectively passes a pointer to the struct to
bpf_offload_dev_create(). We temporarily have two struct
bpf_prog_offload_ops instances, one under offdev->ops and one under
offload->dev_ops. The next patches will make the transition towards the
former, so that offload->dev_ops can be removed, and callbacks relying
on ndo_bpf() added to offdev->ops as well.

While at it, rename "nfp_bpf_analyzer_ops" as "nfp_bpf_dev_ops" (and
similarly for netdevsim).

Suggested-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
drivers/net/ethernet/netronome/nfp/bpf/main.c
drivers/net/ethernet/netronome/nfp/bpf/main.h
drivers/net/ethernet/netronome/nfp/bpf/offload.c
drivers/net/netdevsim/bpf.c
include/linux/bpf.h
kernel/bpf/offload.c