| Message ID | 20180604115715.17895-1-bjorn.topel@gmail.com |
|---|---|
| Headers | show |
| Series | AF_XDP: bug fixes and descriptor changes | expand |
On Mon, Jun 04, 2018 at 01:57:10PM +0200, Björn Töpel wrote: > From: Björn Töpel <bjorn.topel@intel.com> > > An issue with the current AF_XDP uapi raised by Mykyta Iziumtsev (see > https://www.spinics.net/lists/netdev/msg503664.html) is that it does > not support NICs that have a "type-writer" model in an efficient > way. In this model, a memory window is passed to the hardware and > multiple frames might be filled into that window, instead of just one > that we have in the current fixed frame-size model. > > This patch set fixes two bugs in the current implementation and then > changes the uapi so that the type-writer model can be supported > efficiently by a possible future extension of AF_XDP. > > These are the uapi changes in this patch: > > * Change the "u32 idx" in the descriptors to "u64 addr". The current > idx based format does NOT work for the type-writer model (as packets > can start anywhere within a frame) but that a relative address > pointer (the u64 addr) works well for both models in the prototype > code we have that supports both models. We increased it from u32 to > u64 to support umems larger than 4G. We have also removed the u16 > offset when having a "u64 addr" since that information is already > carried in the least significant bits of the address. > > * We want to use "u8 padding[5]" for something useful in the future > (since we are not allowed to change its name), so we now call it > just options so it can be extended for various purposes in the > future. It is an u32 as that it what is left of the 16 byte > descriptor. > > * We changed the name of frame_size in the UMEM_REG setsockopt to > chunk_size since this naming also makes sense to the type-writer > model. > > With these changes to the uapi, we believe the type-writer model can > be supported without having to resort to a new descriptor format. The > type-writer model could then be supported, from the uapi point of > view, by setting a flag at bind time and providing a new flag bit in > the options field of the descriptor that signals to user space that > all packets have been written in a chunk. Or with a new chunk > completion queue as suggested by Mykyta in his latest feedback mail on > the list. for the set: Acked-by: Alexei Starovoitov <ast@kernel.org> Thank you for these fixes. According to unofficial feedback from brcm and netronome folks the descriptor format should work for these nics too. At some point we may consider second format, but I think SW should drive HW requirements and not the other way around.
On 06/04/2018 06:24 PM, Alexei Starovoitov wrote: > On Mon, Jun 04, 2018 at 01:57:10PM +0200, Björn Töpel wrote: >> From: Björn Töpel <bjorn.topel@intel.com> >> >> An issue with the current AF_XDP uapi raised by Mykyta Iziumtsev (see >> https://www.spinics.net/lists/netdev/msg503664.html) is that it does >> not support NICs that have a "type-writer" model in an efficient >> way. In this model, a memory window is passed to the hardware and >> multiple frames might be filled into that window, instead of just one >> that we have in the current fixed frame-size model. >> >> This patch set fixes two bugs in the current implementation and then >> changes the uapi so that the type-writer model can be supported >> efficiently by a possible future extension of AF_XDP. >> >> These are the uapi changes in this patch: >> >> * Change the "u32 idx" in the descriptors to "u64 addr". The current >> idx based format does NOT work for the type-writer model (as packets >> can start anywhere within a frame) but that a relative address >> pointer (the u64 addr) works well for both models in the prototype >> code we have that supports both models. We increased it from u32 to >> u64 to support umems larger than 4G. We have also removed the u16 >> offset when having a "u64 addr" since that information is already >> carried in the least significant bits of the address. >> >> * We want to use "u8 padding[5]" for something useful in the future >> (since we are not allowed to change its name), so we now call it >> just options so it can be extended for various purposes in the >> future. It is an u32 as that it what is left of the 16 byte >> descriptor. >> >> * We changed the name of frame_size in the UMEM_REG setsockopt to >> chunk_size since this naming also makes sense to the type-writer >> model. >> >> With these changes to the uapi, we believe the type-writer model can >> be supported without having to resort to a new descriptor format. The >> type-writer model could then be supported, from the uapi point of >> view, by setting a flag at bind time and providing a new flag bit in >> the options field of the descriptor that signals to user space that >> all packets have been written in a chunk. Or with a new chunk >> completion queue as suggested by Mykyta in his latest feedback mail on >> the list. > > for the set: > Acked-by: Alexei Starovoitov <ast@kernel.org> > Thank you for these fixes. > According to unofficial feedback from brcm and netronome folks > the descriptor format should work for these nics too. > At some point we may consider second format, but I think SW > should drive HW requirements and not the other way around. LGTM as well, applied to bpf-next, thanks!
From: Björn Töpel <bjorn.topel@intel.com> An issue with the current AF_XDP uapi raised by Mykyta Iziumtsev (see https://www.spinics.net/lists/netdev/msg503664.html) is that it does not support NICs that have a "type-writer" model in an efficient way. In this model, a memory window is passed to the hardware and multiple frames might be filled into that window, instead of just one that we have in the current fixed frame-size model. This patch set fixes two bugs in the current implementation and then changes the uapi so that the type-writer model can be supported efficiently by a possible future extension of AF_XDP. These are the uapi changes in this patch: * Change the "u32 idx" in the descriptors to "u64 addr". The current idx based format does NOT work for the type-writer model (as packets can start anywhere within a frame) but that a relative address pointer (the u64 addr) works well for both models in the prototype code we have that supports both models. We increased it from u32 to u64 to support umems larger than 4G. We have also removed the u16 offset when having a "u64 addr" since that information is already carried in the least significant bits of the address. * We want to use "u8 padding[5]" for something useful in the future (since we are not allowed to change its name), so we now call it just options so it can be extended for various purposes in the future. It is an u32 as that it what is left of the 16 byte descriptor. * We changed the name of frame_size in the UMEM_REG setsockopt to chunk_size since this naming also makes sense to the type-writer model. With these changes to the uapi, we believe the type-writer model can be supported without having to resort to a new descriptor format. The type-writer model could then be supported, from the uapi point of view, by setting a flag at bind time and providing a new flag bit in the options field of the descriptor that signals to user space that all packets have been written in a chunk. Or with a new chunk completion queue as suggested by Mykyta in his latest feedback mail on the list. We based this patch set on bpf-next commit bd3a08aaa9a3 ("bpf: flowlabel in bpf_fib_lookup should be flowinfo") The structure of the patch set is as follows: Patches 1-2: Fixes two bugs in the current implementation. Patches 3-4: Prepares the uapi for a "type-writer" model and modifies the sample application so that it works with the new uapi. Patch 5: Small performance improvement patch for the sample application. Cheers: Magnus and Björn Björn Töpel (4): xsk: proper fill queue descriptor validation xsk: proper Rx drop statistics update xsk: new descriptor addressing scheme samples/bpf: adapted to new uapi Magnus Karlsson (1): samples/bpf: minor *_nb_free performance fix Documentation/networking/af_xdp.rst | 101 +++++++++++++++++++++--------------- include/uapi/linux/if_xdp.h | 12 ++--- net/xdp/xdp_umem.c | 33 ++++++------ net/xdp/xdp_umem.h | 27 +++------- net/xdp/xdp_umem_props.h | 4 +- net/xdp/xsk.c | 41 ++++++++------- net/xdp/xsk_queue.c | 2 +- net/xdp/xsk_queue.h | 63 ++++++++-------------- samples/bpf/xdpsock_user.c | 92 +++++++++++++++----------------- 9 files changed, 172 insertions(+), 203 deletions(-)