From: Björn Töpel <hidden> Date: 2021-01-19 18:29:30
This series extends bind() for XDP sockets, so that the bound socket
is added to the netdev_rx_queue _rx array in the netdevice. We call
this to register the socket. To redirect packets to the registered
socket, a new BPF helper is used: bpf_redirect_xsk().
For shared XDP sockets, only the first bound socket is
registered. Users that need more complex setup has to use XSKMAP and
bpf_redirect_map().
Now, why would one use bpf_redirect_xsk() over the regular
bpf_redirect_map() helper?
* Better performance!
* Convenience; Most user use one socket per queue. This scenario is
what registered sockets support. There is no need to create an
XSKMAP. This can also reduce complexity from containerized setups,
where users might what to use XDP sockets without CAP_SYS_ADMIN
capabilities.
The first patch restructures xdp_do_redirect() a bit, to make it
easier to add the new helper. This restructure also give us a slight
performance benefit. The following three patches extends bind() and
adds the new helper. After that, two libbpf patches that selects XDP
program based on what kernel is running. Finally, selftests for the new
functionality is added.
Note that the libbpf "auto-selection" is based on kernel version, so
it is hard coded to the "-next" version (5.12). If you would like to
try this is out, you will need to change the libbpf patch locally!
Thanks to Maciej and Magnus for the internal review/comments!
Performance (rxdrop, zero-copy)
Baseline
Two cores: 21.3 Mpps
One core: 24.5 Mpps
Patched
Two cores, bpf_redirect_map: 21.7 Mpps + 2%
One core, bpf_redirect_map: 24.9 Mpps + 2%
Two cores, bpf_redirect_xsk: 24.0 Mpps +13%
One core, bpf_redirect_xsk: 25.5 Mpps + 4%
Thanks!
Björn
v1->v2:
* Added missing XDP programs to selftests.
* Fixed checkpatch warning in selftests.
Björn Töpel (8):
xdp: restructure redirect actions
xsk: remove explicit_free parameter from __xsk_rcv()
xsk: fold xp_assign_dev and __xp_assign_dev
xsk: register XDP sockets at bind(), and add new AF_XDP BPF helper
libbpf, xsk: select AF_XDP BPF program based on kernel version
libbpf, xsk: select bpf_redirect_xsk(), if supported
selftest/bpf: add XDP socket tests for bpf_redirect_{xsk, map}()
selftest/bpf: remove a lot of ifobject casting in xdpxceiver
include/linux/filter.h | 10 +
include/linux/netdevice.h | 1 +
include/net/xdp_sock.h | 12 +
include/net/xsk_buff_pool.h | 2 +-
include/trace/events/xdp.h | 46 ++--
include/uapi/linux/bpf.h | 7 +
net/core/filter.c | 205 ++++++++++--------
net/xdp/xsk.c | 112 ++++++++--
net/xdp/xsk_buff_pool.c | 12 +-
tools/include/uapi/linux/bpf.h | 7 +
tools/lib/bpf/libbpf.c | 2 +-
tools/lib/bpf/libbpf_internal.h | 2 +
tools/lib/bpf/libbpf_probes.c | 16 --
tools/lib/bpf/xsk.c | 83 ++++++-
.../selftests/bpf/progs/xdpxceiver_ext1.c | 15 ++
.../selftests/bpf/progs/xdpxceiver_ext2.c | 9 +
tools/testing/selftests/bpf/test_xsk.sh | 48 ++++
tools/testing/selftests/bpf/xdpxceiver.c | 164 +++++++++-----
tools/testing/selftests/bpf/xdpxceiver.h | 2 +
19 files changed, 554 insertions(+), 201 deletions(-)
create mode 100644 tools/testing/selftests/bpf/progs/xdpxceiver_ext1.c
create mode 100644 tools/testing/selftests/bpf/progs/xdpxceiver_ext2.c
base-commit: 95204c9bfa48d2f4d3bab7df55c1cc823957ff81
--
2.27.0
From: Björn Töpel <hidden> Date: 2021-01-19 15:51:59
From: Björn Töpel <redacted>
Add detection for kernel version, and adapt the BPF program based on
kernel support. This way, users will get the best possible performance
from the BPF program.
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Acked-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Björn Töpel <redacted>
Signed-off-by: Marek Majtyka <redacted>
---
tools/lib/bpf/libbpf.c | 2 +-
tools/lib/bpf/libbpf_internal.h | 2 ++
tools/lib/bpf/libbpf_probes.c | 16 -------------
tools/lib/bpf/xsk.c | 41 ++++++++++++++++++++++++++++++---
4 files changed, 41 insertions(+), 20 deletions(-)
@@ -48,22 +48,6 @@ static int get_vendor_id(int ifindex)returnstrtol(buf,NULL,0);}-staticintget_kernel_version(void)-{-intversion,subversion,patchlevel;-structutsnameutsn;--/* Return 0 on failure, and attempt to probe with empty kversion */-if(uname(&utsn))-return0;--if(sscanf(utsn.release,"%d.%d.%d",-&version,&subversion,&patchlevel)!=3)-return0;--return(version<<16)+(subversion<<8)+patchlevel;-}-staticvoidprobe_load(enumbpf_prog_typeprog_type,conststructbpf_insn*insns,size_tinsns_cnt,char*buf,size_tbuf_len,__u32ifindex)
@@ -351,6 +357,13 @@ int xsk_umem__create_v0_0_2(struct xsk_umem **umem_ptr, void *umem_area,COMPAT_VERSION(xsk_umem__create_v0_0_2,xsk_umem__create,LIBBPF_0.0.2)DEFAULT_VERSION(xsk_umem__create_v0_0_4,xsk_umem__create,LIBBPF_0.0.4)+staticenumxsk_progget_xsk_prog(void)+{+__u32kver=get_kernel_version();++returnkver<KERNEL_VERSION(5,3,0)?XSK_PROG_FALLBACK:XSK_PROG_REDIRECT_FLAGS;+}+staticintxsk_load_xdp_prog(structxsk_socket*xsk){staticconstintlog_buf_size=16*1024;
@@ -358,7 +371,7 @@ static int xsk_load_xdp_prog(struct xsk_socket *xsk)charlog_buf[log_buf_size];interr,prog_fd;-/* This is the C-program:+/* This is the fallback C-program:*SEC("xdp_sock")intxdp_sock_prog(structxdp_md*ctx)*{*intret,index=ctx->rx_queue_index;
@@ -414,9 +427,31 @@ static int xsk_load_xdp_prog(struct xsk_socket *xsk)/* The jumps are to this instruction */BPF_EXIT_INSN(),};-size_tinsns_cnt=sizeof(prog)/sizeof(structbpf_insn);-prog_fd=bpf_load_program(BPF_PROG_TYPE_XDP,prog,insns_cnt,+/* This is the post-5.3 kernel C-program:+*SEC("xdp_sock")intxdp_sock_prog(structxdp_md*ctx)+*{+*returnbpf_redirect_map(&xsks_map,ctx->rx_queue_index,XDP_PASS);+*}+*/+structbpf_insnprog_redirect_flags[]={+/* r2 = *(u32 *)(r1 + 16) */+BPF_LDX_MEM(BPF_W,BPF_REG_2,BPF_REG_1,16),+/* r1 = xskmap[] */+BPF_LD_MAP_FD(BPF_REG_1,ctx->xsks_map_fd),+/* r3 = XDP_PASS */+BPF_MOV64_IMM(BPF_REG_3,2),+/* call bpf_redirect_map */+BPF_EMIT_CALL(BPF_FUNC_redirect_map),+BPF_EXIT_INSN(),+};+size_tinsns_cnt[]={sizeof(prog)/sizeof(structbpf_insn),+sizeof(prog_redirect_flags)/sizeof(structbpf_insn),+};+structbpf_insn*progs[]={prog,prog_redirect_flags};+enumxsk_progoption=get_xsk_prog();++prog_fd=bpf_load_program(BPF_PROG_TYPE_XDP,progs[option],insns_cnt[option],"LGPL-2.1 or BSD-2-Clause",0,log_buf,log_buf_size);if(prog_fd<0){
From: Björn Töpel <hidden> Date: 2021-01-19 15:53:15
From: Björn Töpel <redacted>
Instead of passing void * all over the place, let us pass the actual
type (ifobject) and remove the void-ptr-to-type-ptr casting.
Signed-off-by: Björn Töpel <redacted>
---
tools/testing/selftests/bpf/xdpxceiver.c | 87 ++++++++++++------------
1 file changed, 42 insertions(+), 45 deletions(-)
@@ -845,14 +844,14 @@ static void worker_pkt_validate(void)}}-staticvoidthread_common_ops(void*arg,void*bufs,pthread_mutex_t*mutexptr,+staticvoidthread_common_ops(structifobject*ifobject,void*bufs,pthread_mutex_t*mutexptr,atomic_int*spinningptr){intctr=0;intret;-xsk_configure_umem((structifobject*)arg,bufs,num_frames*XSK_UMEM__DEFAULT_FRAME_SIZE);-ret=xsk_configure_socket((structifobject*)arg);+xsk_configure_umem(ifobject,bufs,num_frames*XSK_UMEM__DEFAULT_FRAME_SIZE);+ret=xsk_configure_socket(ifobject);/* Retry Create Socket if it fails as xsk_socket__create()*isasynchronous
@@ -245,6 +245,54 @@ retval=$? test_status$retval"${TEST_NAME}"statusList+=($retval)+### TEST 10+TEST_NAME="SKB EXT BPF_REDIRECT_MAP"++vethXDPgeneric${VETH0}${VETH1}${NS1}++params=("-S""--ext-prog1")+execxdpxceiverparams++retval=$?+test_status$retval"${TEST_NAME}"+statusList+=($retval)++### TEST 11+TEST_NAME="DRV EXT BPF_REDIRECT_MAP"++vethXDPnative${VETH0}${VETH1}${NS1}++params=("-N""--ext-prog1")+execxdpxceiverparams++retval=$?+test_status$retval"${TEST_NAME}"+statusList+=($retval)++### TEST 12+TEST_NAME="SKB EXT BPF_REDIRECT_XSK"++vethXDPgeneric${VETH0}${VETH1}${NS1}++params=("-S""--ext-prog2")+execxdpxceiverparams++retval=$?+test_status$retval"${TEST_NAME}"+statusList+=($retval)++### TEST 13+TEST_NAME="DRV EXT BPF_REDIRECT_XSK"++vethXDPnative${VETH0}${VETH1}${NS1}++params=("-N""--ext-prog2")+execxdpxceiverparams++retval=$?+test_status$retval"${TEST_NAME}"+statusList+=($retval)+## END TESTS cleanup_exit${VETH0}${VETH1}${NS1}
From: Björn Töpel <hidden> Date: 2021-01-19 16:10:06
From: Björn Töpel <redacted>
Select bpf_redirect_xsk() as the default AF_XDP BPF program, if
supported.
The bpf_redirect_xsk() helper does not require an XSKMAP, so make sure
that no map is created/updated when using it.
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Björn Töpel <redacted>
---
tools/lib/bpf/xsk.c | 46 +++++++++++++++++++++++++++++++++++++++++----
1 file changed, 42 insertions(+), 4 deletions(-)
From: Björn Töpel <hidden> Date: 2021-01-19 18:29:29
From: Björn Töpel <redacted>
Fold xp_assign_dev and __xp_assign_dev. The former directly calls the
latter.
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Björn Töpel <redacted>
---
net/xdp/xsk_buff_pool.c | 12 +++---------
1 file changed, 3 insertions(+), 9 deletions(-)
From: Björn Töpel <hidden> Date: 2021-01-19 18:29:29
From: Björn Töpel <redacted>
Extend bind() for XDP sockets, so that the bound socket is added to
the netdev_rx_queue _rx array in the netdevice. We call this to
register an XDP socket. To redirect packets to a registered socket, a
new BPF helper is used: bpf_redirect_xsk().
For shared XDP sockets, only the first bound socket is
registered. Users that require more advanced setups, continue to the
XSKMAP and bpf_redirect_map().
Now, why would one use bpf_redirect_xsk() over the regular
bpf_redirect_map() helper? First: Slightly better performance. Second:
Convenience. Most user use one socket per queue. This scenario is what
registered sockets support. There is no need to create an XSKMAP. This
can also reduce complexity from containerized setups, where users
might what to use XDP sockets without CAP_SYS_ADMIN capabilities.
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Björn Töpel <redacted>
---
include/linux/filter.h | 1 +
include/linux/netdevice.h | 1 +
include/net/xdp_sock.h | 12 +++++
include/net/xsk_buff_pool.h | 2 +-
include/uapi/linux/bpf.h | 7 +++
net/core/filter.c | 49 ++++++++++++++++--
net/xdp/xsk.c | 93 ++++++++++++++++++++++++++++------
net/xdp/xsk_buff_pool.c | 4 +-
tools/include/uapi/linux/bpf.h | 7 +++
9 files changed, 153 insertions(+), 23 deletions(-)
@@ -3836,6 +3836,12 @@ union bpf_attr {*Return*ApointertoastructsocketonsuccessorNULLifthefileis*notasocket.+*+*longbpf_redirect_xsk(structxdp_buff*xdp_md,u64action)+*Description+*RedirecttotheregisteredAF_XDPsocket.+*Return+***XDP_REDIRECT**onsuccess,otherwisetheactionparameterisreturned.*/#define __BPF_FUNC_MAPPER(FN) \FN(unspec),\
@@ -4001,6 +4007,7 @@ union bpf_attr {FN(ktime_get_coarse_ns),\FN(ima_inode_hash),\FN(sock_from_file),\+FN(redirect_xsk),\/* *//* integer value in 'imm' field of BPF_CALL instruction selects which helper
@@ -661,6 +722,7 @@ static void xsk_unbind_dev(struct xdp_sock *xs)if(xs->state!=XSK_BOUND)return;+xsk_clear(xs);WRITE_ONCE(xs->state,XSK_UNBOUND);/* Wait for driver to stop using the xdp socket. */
@@ -892,7 +954,7 @@ static int xsk_bind(struct socket *sock, struct sockaddr *addr, int addr_len)gotoout_unlock;}-err=xp_assign_dev(xs->pool,dev,qid,flags);+err=xp_assign_dev(xs,xs->pool,dev,qid,flags);if(err){xp_destroy(xs->pool);xs->pool=NULL;
@@ -918,6 +980,7 @@ static int xsk_bind(struct socket *sock, struct sockaddr *addr, int addr_len)*/smp_wmb();WRITE_ONCE(xs->state,XSK_BOUND);+xsk_reg(xs);}out_release:mutex_unlock(&xs->mutex);
@@ -3836,6 +3836,12 @@ union bpf_attr {*Return*ApointertoastructsocketonsuccessorNULLifthefileis*notasocket.+*+*longbpf_redirect_xsk(structxdp_buff*xdp_md,u64action)+*Description+*RedirecttotheregisteredAF_XDPsocket.+*Return+***XDP_REDIRECT**onsuccess,otherwisetheactionparameterisreturned.*/#define __BPF_FUNC_MAPPER(FN) \FN(unspec),\
@@ -4001,6 +4007,7 @@ union bpf_attr {FN(ktime_get_coarse_ns),\FN(ima_inode_hash),\FN(sock_from_file),\+FN(redirect_xsk),\/* *//* integer value in 'imm' field of BPF_CALL instruction selects which helper
From: Björn Töpel <hidden> Date: 2021-01-19 18:29:29
From: Björn Töpel <redacted>
The explicit_free parameter of the __xsk_rcv() function was used to
mark whether the call was via the generic XDP or the native XDP
path. Instead of clutter the code with if-statements and "true/false"
parameters which are hard to understand, simply move the explicit free
to the __xsk_map_redirect() which is always called from the native XDP
path.
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Björn Töpel <redacted>
---
net/xdp/xsk.c | 47 +++++++++++++++++++++++++++++++----------------
1 file changed, 31 insertions(+), 16 deletions(-)
From: Björn Töpel <hidden> Date: 2021-01-19 18:29:29
From: Björn Töpel <redacted>
The XDP_REDIRECT implementations for maps and non-maps are fairly
similar, but obviously need to take different code paths depending on
if the target is using a map or not. Today, the redirect targets for
XDP either uses a map, or is based on ifindex.
Future commits will introduce yet another redirect target via the a
new helper, bpf_redirect_xsk(). To pave the way for that, we introduce
an explicit redirect type to bpf_redirect_info. This makes the code
easier to follow, and makes it easier to add new redirect targets.
Further, using an explicit type in bpf_redirect_info has a slight
positive performance impact by avoiding a pointer indirection for the
map type lookup, and instead use the hot cacheline for
bpf_redirect_info.
The bpf_redirect_info flags member is not used by XDP, and not
read/written any more. The map member is only written to when
required/used, and not unconditionally.
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Björn Töpel <redacted>
---
include/linux/filter.h | 9 ++
include/trace/events/xdp.h | 46 +++++++----
net/core/filter.c | 164 ++++++++++++++++++-------------------
3 files changed, 117 insertions(+), 102 deletions(-)
@@ -133,45 +141,49 @@ DEFINE_EVENT(xdp_redirect_template, xdp_redirect,TP_PROTO(conststructnet_device*dev,conststructbpf_prog*xdp,constvoid*tgt,interr,-conststructbpf_map*map,u32index),-TP_ARGS(dev,xdp,tgt,err,map,index)+enumxdp_redirect_typetype,+conststructbpf_redirect_info*ri),+TP_ARGS(dev,xdp,tgt,err,type,ri));DEFINE_EVENT(xdp_redirect_template,xdp_redirect_err,TP_PROTO(conststructnet_device*dev,conststructbpf_prog*xdp,constvoid*tgt,interr,-conststructbpf_map*map,u32index),-TP_ARGS(dev,xdp,tgt,err,map,index)+enumxdp_redirect_typetype,+conststructbpf_redirect_info*ri),+TP_ARGS(dev,xdp,tgt,err,type,ri));#define _trace_xdp_redirect(dev, xdp, to) \-trace_xdp_redirect(dev,xdp,NULL,0,NULL,to)+trace_xdp_redirect(dev,xdp,NULL,0,XDP_REDIR_DEV_IFINDEX,NULL)#define _trace_xdp_redirect_err(dev, xdp, to, err) \-trace_xdp_redirect_err(dev,xdp,NULL,err,NULL,to)+trace_xdp_redirect_err(dev,xdp,NULL,err,XDP_REDIR_DEV_IFINDEX,NULL)-#define _trace_xdp_redirect_map(dev, xdp, to, map, index) \-trace_xdp_redirect(dev,xdp,to,0,map,index)+#define _trace_xdp_redirect_map(dev, xdp, to, type, ri) \+trace_xdp_redirect(dev,xdp,to,0,type,ri)-#define _trace_xdp_redirect_map_err(dev, xdp, to, map, index, err) \-trace_xdp_redirect_err(dev,xdp,to,err,map,index)+#define _trace_xdp_redirect_map_err(dev, xdp, to, type, ri, err) \+trace_xdp_redirect_err(dev,xdp,to,err,type,ri)/* not used anymore, but kept around so as not to break old programs */DEFINE_EVENT(xdp_redirect_template,xdp_redirect_map,TP_PROTO(conststructnet_device*dev,conststructbpf_prog*xdp,constvoid*tgt,interr,-conststructbpf_map*map,u32index),-TP_ARGS(dev,xdp,tgt,err,map,index)+enumxdp_redirect_typetype,+conststructbpf_redirect_info*ri),+TP_ARGS(dev,xdp,tgt,err,type,ri));DEFINE_EVENT(xdp_redirect_template,xdp_redirect_map_err,TP_PROTO(conststructnet_device*dev,conststructbpf_prog*xdp,constvoid*tgt,interr,-conststructbpf_map*map,u32index),-TP_ARGS(dev,xdp,tgt,err,map,index)+enumxdp_redirect_typetype,+conststructbpf_redirect_info*ri),+TP_ARGS(dev,xdp,tgt,err,type,ri));TRACE_EVENT(xdp_cpumap_kthread,
@@ -4123,18 +4098,37 @@ BPF_CALL_3(bpf_xdp_redirect_map, struct bpf_map *, map, u32, ifindex,if(unlikely(flags>XDP_TX))returnXDP_ABORTED;-ri->tgt_value=__xdp_map_lookup_elem(map,ifindex);+switch(map->map_type){+caseBPF_MAP_TYPE_DEVMAP:+ri->tgt_value=__dev_map_lookup_elem(map,ifindex);+ri->tgt_type=XDP_REDIR_DEV_MAP;+break;+caseBPF_MAP_TYPE_DEVMAP_HASH:+ri->tgt_value=__dev_map_hash_lookup_elem(map,ifindex);+ri->tgt_type=XDP_REDIR_DEV_MAP;+break;+caseBPF_MAP_TYPE_CPUMAP:+ri->tgt_value=__cpu_map_lookup_elem(map,ifindex);+ri->tgt_type=XDP_REDIR_CPU_MAP;+break;+caseBPF_MAP_TYPE_XSKMAP:+ri->tgt_value=__xsk_map_lookup_elem(map,ifindex);+ri->tgt_type=XDP_REDIR_XSK_MAP;+break;+default:+ri->tgt_value=NULL;+}+if(unlikely(!ri->tgt_value)){/* If the lookup fails we want to clear out the state in the*redirect_infostructcompletely,sothatifaneBPFprogram*performsmultiplelookups,thelastonealwaystakes*precedence.*/-WRITE_ONCE(ri->map,NULL);+ri->tgt_type=XDP_REDIR_UNSET;returnflags;}-ri->flags=flags;ri->tgt_index=ifindex;WRITE_ONCE(ri->map,map);
From: kernel test robot <hidden> Date: 2021-01-20 08:28:30
Hi "Björn,
I love your patch! Yet something to improve:
[auto build test ERROR on 95204c9bfa48d2f4d3bab7df55c1cc823957ff81]
url: https://github.com/0day-ci/linux/commits/Bj-rn-T-pel/Introduce-bpf_redirect_xsk-helper/20210120-150357
base: 95204c9bfa48d2f4d3bab7df55c1cc823957ff81
config: x86_64-randconfig-m031-20210120 (attached as .config)
compiler: gcc-9 (Debian 9.3.0-15) 9.3.0
reproduce (this is a W=1 build):
# https://github.com/0day-ci/linux/commit/419b1341d7980ee57fb10f4306719eef3c1a15f8
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Bj-rn-T-pel/Introduce-bpf_redirect_xsk-helper/20210120-150357
git checkout 419b1341d7980ee57fb10f4306719eef3c1a15f8
# save the attached .config to linux build tree
make W=1 ARCH=x86_64
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <redacted>
All errors (new ones prefixed by >>):
In file included from <command-line>:
net/core/filter.c: In function '____bpf_xdp_redirect_xsk':
quoted
net/core/filter.c:4165:35: error: 'struct netdev_rx_queue' has no member named 'xsk'
4165 | xs = READ_ONCE(dev->_rx[queue_id].xsk);
| ^
include/linux/compiler_types.h:300:9: note: in definition of macro '__compiletime_assert'
300 | if (!(condition)) \
| ^~~~~~~~~
include/linux/compiler_types.h:320:2: note: in expansion of macro '_compiletime_assert'
320 | _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
| ^~~~~~~~~~~~~~~~~~~
include/asm-generic/rwonce.h:36:2: note: in expansion of macro 'compiletime_assert'
36 | compiletime_assert(__native_word(t) || sizeof(t) == sizeof(long long), \
| ^~~~~~~~~~~~~~~~~~
include/asm-generic/rwonce.h:36:21: note: in expansion of macro '__native_word'
36 | compiletime_assert(__native_word(t) || sizeof(t) == sizeof(long long), \
| ^~~~~~~~~~~~~
include/asm-generic/rwonce.h:49:2: note: in expansion of macro 'compiletime_assert_rwonce_type'
49 | compiletime_assert_rwonce_type(x); \
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
net/core/filter.c:4165:7: note: in expansion of macro 'READ_ONCE'
4165 | xs = READ_ONCE(dev->_rx[queue_id].xsk);
| ^~~~~~~~~
quoted
net/core/filter.c:4165:35: error: 'struct netdev_rx_queue' has no member named 'xsk'
4165 | xs = READ_ONCE(dev->_rx[queue_id].xsk);
| ^
include/linux/compiler_types.h:300:9: note: in definition of macro '__compiletime_assert'
300 | if (!(condition)) \
| ^~~~~~~~~
include/linux/compiler_types.h:320:2: note: in expansion of macro '_compiletime_assert'
320 | _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
| ^~~~~~~~~~~~~~~~~~~
include/asm-generic/rwonce.h:36:2: note: in expansion of macro 'compiletime_assert'
36 | compiletime_assert(__native_word(t) || sizeof(t) == sizeof(long long), \
| ^~~~~~~~~~~~~~~~~~
include/asm-generic/rwonce.h:36:21: note: in expansion of macro '__native_word'
36 | compiletime_assert(__native_word(t) || sizeof(t) == sizeof(long long), \
| ^~~~~~~~~~~~~
include/asm-generic/rwonce.h:49:2: note: in expansion of macro 'compiletime_assert_rwonce_type'
49 | compiletime_assert_rwonce_type(x); \
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
net/core/filter.c:4165:7: note: in expansion of macro 'READ_ONCE'
4165 | xs = READ_ONCE(dev->_rx[queue_id].xsk);
| ^~~~~~~~~
quoted
net/core/filter.c:4165:35: error: 'struct netdev_rx_queue' has no member named 'xsk'
4165 | xs = READ_ONCE(dev->_rx[queue_id].xsk);
| ^
include/linux/compiler_types.h:300:9: note: in definition of macro '__compiletime_assert'
300 | if (!(condition)) \
| ^~~~~~~~~
include/linux/compiler_types.h:320:2: note: in expansion of macro '_compiletime_assert'
320 | _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
| ^~~~~~~~~~~~~~~~~~~
include/asm-generic/rwonce.h:36:2: note: in expansion of macro 'compiletime_assert'
36 | compiletime_assert(__native_word(t) || sizeof(t) == sizeof(long long), \
| ^~~~~~~~~~~~~~~~~~
include/asm-generic/rwonce.h:36:21: note: in expansion of macro '__native_word'
36 | compiletime_assert(__native_word(t) || sizeof(t) == sizeof(long long), \
| ^~~~~~~~~~~~~
include/asm-generic/rwonce.h:49:2: note: in expansion of macro 'compiletime_assert_rwonce_type'
49 | compiletime_assert_rwonce_type(x); \
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
net/core/filter.c:4165:7: note: in expansion of macro 'READ_ONCE'
4165 | xs = READ_ONCE(dev->_rx[queue_id].xsk);
| ^~~~~~~~~
quoted
net/core/filter.c:4165:35: error: 'struct netdev_rx_queue' has no member named 'xsk'
4165 | xs = READ_ONCE(dev->_rx[queue_id].xsk);
| ^
include/linux/compiler_types.h:300:9: note: in definition of macro '__compiletime_assert'
300 | if (!(condition)) \
| ^~~~~~~~~
include/linux/compiler_types.h:320:2: note: in expansion of macro '_compiletime_assert'
320 | _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
| ^~~~~~~~~~~~~~~~~~~
include/asm-generic/rwonce.h:36:2: note: in expansion of macro 'compiletime_assert'
36 | compiletime_assert(__native_word(t) || sizeof(t) == sizeof(long long), \
| ^~~~~~~~~~~~~~~~~~
include/asm-generic/rwonce.h:36:21: note: in expansion of macro '__native_word'
36 | compiletime_assert(__native_word(t) || sizeof(t) == sizeof(long long), \
| ^~~~~~~~~~~~~
include/asm-generic/rwonce.h:49:2: note: in expansion of macro 'compiletime_assert_rwonce_type'
49 | compiletime_assert_rwonce_type(x); \
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
net/core/filter.c:4165:7: note: in expansion of macro 'READ_ONCE'
4165 | xs = READ_ONCE(dev->_rx[queue_id].xsk);
| ^~~~~~~~~
quoted
net/core/filter.c:4165:35: error: 'struct netdev_rx_queue' has no member named 'xsk'
4165 | xs = READ_ONCE(dev->_rx[queue_id].xsk);
| ^
include/linux/compiler_types.h:300:9: note: in definition of macro '__compiletime_assert'
300 | if (!(condition)) \
| ^~~~~~~~~
include/linux/compiler_types.h:320:2: note: in expansion of macro '_compiletime_assert'
320 | _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
| ^~~~~~~~~~~~~~~~~~~
include/asm-generic/rwonce.h:36:2: note: in expansion of macro 'compiletime_assert'
36 | compiletime_assert(__native_word(t) || sizeof(t) == sizeof(long long), \
| ^~~~~~~~~~~~~~~~~~
include/asm-generic/rwonce.h:49:2: note: in expansion of macro 'compiletime_assert_rwonce_type'
49 | compiletime_assert_rwonce_type(x); \
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
net/core/filter.c:4165:7: note: in expansion of macro 'READ_ONCE'
4165 | xs = READ_ONCE(dev->_rx[queue_id].xsk);
| ^~~~~~~~~
quoted
net/core/filter.c:4165:35: error: 'struct netdev_rx_queue' has no member named 'xsk'
4165 | xs = READ_ONCE(dev->_rx[queue_id].xsk);
| ^
include/linux/compiler_types.h:271:13: note: in definition of macro '__unqual_scalar_typeof'
271 | _Generic((x), \
| ^
include/asm-generic/rwonce.h:50:2: note: in expansion of macro '__READ_ONCE'
50 | __READ_ONCE(x); \
| ^~~~~~~~~~~
net/core/filter.c:4165:7: note: in expansion of macro 'READ_ONCE'
4165 | xs = READ_ONCE(dev->_rx[queue_id].xsk);
| ^~~~~~~~~
In file included from ./arch/x86/include/generated/asm/rwonce.h:1,
from include/linux/compiler.h:246,
from include/linux/export.h:43,
from include/linux/linkage.h:7,
from include/linux/kernel.h:7,
from include/linux/list.h:9,
from include/linux/module.h:12,
from net/core/filter.c:20:
quoted
net/core/filter.c:4165:35: error: 'struct netdev_rx_queue' has no member named 'xsk'
From: Björn Töpel <hidden> Date: 2021-01-20 08:42:40
On 2021-01-20 09:25, kernel test robot wrote:
Hi "Björn,
I love your patch! Yet something to improve:
[auto build test ERROR on 95204c9bfa48d2f4d3bab7df55c1cc823957ff81]
url: https://github.com/0day-ci/linux/commits/Bj-rn-T-pel/Introduce-bpf_redirect_xsk-helper/20210120-150357
base: 95204c9bfa48d2f4d3bab7df55c1cc823957ff81
config: x86_64-randconfig-m031-20210120 (attached as .config)
compiler: gcc-9 (Debian 9.3.0-15) 9.3.0
reproduce (this is a W=1 build):
# https://github.com/0day-ci/linux/commit/419b1341d7980ee57fb10f4306719eef3c1a15f8
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Bj-rn-T-pel/Introduce-bpf_redirect_xsk-helper/20210120-150357
git checkout 419b1341d7980ee57fb10f4306719eef3c1a15f8
# save the attached .config to linux build tree
make W=1 ARCH=x86_64
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <redacted>
All errors (new ones prefixed by >>):
In file included from <command-line>:
net/core/filter.c: In function '____bpf_xdp_redirect_xsk':
quoted
quoted
net/core/filter.c:4165:35: error: 'struct netdev_rx_queue' has no member named 'xsk'
This rev is missing proper CONFIG_XDP_SOCKET ifdefs. I'll fix this in
the next spin, but I'll wait a bit for more comments.
Björn
[...]
This series extends bind() for XDP sockets, so that the bound socket
is added to the netdev_rx_queue _rx array in the netdevice. We call
this to register the socket. To redirect packets to the registered
socket, a new BPF helper is used: bpf_redirect_xsk().
For shared XDP sockets, only the first bound socket is
registered. Users that need more complex setup has to use XSKMAP and
bpf_redirect_map().
Now, why would one use bpf_redirect_xsk() over the regular
bpf_redirect_map() helper?
* Better performance!
* Convenience; Most user use one socket per queue. This scenario is
what registered sockets support. There is no need to create an
XSKMAP. This can also reduce complexity from containerized setups,
where users might what to use XDP sockets without CAP_SYS_ADMIN
capabilities.
The first patch restructures xdp_do_redirect() a bit, to make it
easier to add the new helper. This restructure also give us a slight
performance benefit. The following three patches extends bind() and
adds the new helper. After that, two libbpf patches that selects XDP
program based on what kernel is running. Finally, selftests for the new
functionality is added.
Note that the libbpf "auto-selection" is based on kernel version, so
it is hard coded to the "-next" version (5.12). If you would like to
try this is out, you will need to change the libbpf patch locally!
Thanks to Maciej and Magnus for the internal review/comments!
Performance (rxdrop, zero-copy)
Baseline
Two cores: 21.3 Mpps
One core: 24.5 Mpps
Two cores is slower? It used to be faster all the time, didn't it?
Patched
Two cores, bpf_redirect_map: 21.7 Mpps + 2%
One core, bpf_redirect_map: 24.9 Mpps + 2%
Two cores, bpf_redirect_xsk: 24.0 Mpps +13%
Nice, impressive improvement!
One core, bpf_redirect_xsk: 25.5 Mpps + 4%
Thanks!
Björn
v1->v2:
* Added missing XDP programs to selftests.
* Fixed checkpatch warning in selftests.
Björn Töpel (8):
xdp: restructure redirect actions
xsk: remove explicit_free parameter from __xsk_rcv()
xsk: fold xp_assign_dev and __xp_assign_dev
xsk: register XDP sockets at bind(), and add new AF_XDP BPF helper
libbpf, xsk: select AF_XDP BPF program based on kernel version
libbpf, xsk: select bpf_redirect_xsk(), if supported
selftest/bpf: add XDP socket tests for bpf_redirect_{xsk, map}()
selftest/bpf: remove a lot of ifobject casting in xdpxceiver
include/linux/filter.h | 10 +
include/linux/netdevice.h | 1 +
include/net/xdp_sock.h | 12 +
include/net/xsk_buff_pool.h | 2 +-
include/trace/events/xdp.h | 46 ++--
include/uapi/linux/bpf.h | 7 +
net/core/filter.c | 205 ++++++++++--------
net/xdp/xsk.c | 112 ++++++++--
net/xdp/xsk_buff_pool.c | 12 +-
tools/include/uapi/linux/bpf.h | 7 +
tools/lib/bpf/libbpf.c | 2 +-
tools/lib/bpf/libbpf_internal.h | 2 +
tools/lib/bpf/libbpf_probes.c | 16 --
tools/lib/bpf/xsk.c | 83 ++++++-
.../selftests/bpf/progs/xdpxceiver_ext1.c | 15 ++
.../selftests/bpf/progs/xdpxceiver_ext2.c | 9 +
tools/testing/selftests/bpf/test_xsk.sh | 48 ++++
tools/testing/selftests/bpf/xdpxceiver.c | 164 +++++++++-----
tools/testing/selftests/bpf/xdpxceiver.h | 2 +
19 files changed, 554 insertions(+), 201 deletions(-)
create mode 100644 tools/testing/selftests/bpf/progs/xdpxceiver_ext1.c
create mode 100644 tools/testing/selftests/bpf/progs/xdpxceiver_ext2.c
base-commit: 95204c9bfa48d2f4d3bab7df55c1cc823957ff81
@@ -3836,6 +3836,12 @@ union bpf_attr {*Return*ApointertoastructsocketonsuccessorNULLifthefileis*notasocket.+*+*longbpf_redirect_xsk(structxdp_buff*xdp_md,u64action)+*Description+*RedirecttotheregisteredAF_XDPsocket.+*Return+***XDP_REDIRECT**onsuccess,otherwisetheactionparameterisreturned.*/
I think it would be better to make the second argument a 'flags'
argument and make values > XDP_TX invalid (like we do in
bpf_xdp_redirect_map() now). By allowing any value as return you lose
the ability to turn it into a flags argument later...
-Toke
From: Björn Töpel <hidden> Date: 2021-01-20 14:07:05
On 2021-01-20 13:44, Toke Høiland-Jørgensen wrote:
Björn Töpel [off-list ref] writes:
quoted
From: Björn Töpel <redacted>
The XDP_REDIRECT implementations for maps and non-maps are fairly
similar, but obviously need to take different code paths depending on
if the target is using a map or not. Today, the redirect targets for
XDP either uses a map, or is based on ifindex.
Future commits will introduce yet another redirect target via the a
new helper, bpf_redirect_xsk(). To pave the way for that, we introduce
an explicit redirect type to bpf_redirect_info. This makes the code
easier to follow, and makes it easier to add new redirect targets.
Further, using an explicit type in bpf_redirect_info has a slight
positive performance impact by avoiding a pointer indirection for the
map type lookup, and instead use the hot cacheline for
bpf_redirect_info.
The bpf_redirect_info flags member is not used by XDP, and not
read/written any more. The map member is only written to when
required/used, and not unconditionally.
I like the simplification. However, the handling of map clearing becomes
a bit murky with this change:
You're not changing anything in bpf_clear_redirect_map(), and you're
removing most of the reads and writes of ri->map. Instead,
bpf_xdp_redirect_map() will store the bpf_dtab_netdev pointer in
ri->tgt_value, which xdp_do_redirect() will just read and use without
checking. But if the map element (or the entire map) has been freed in
the meantime that will be a dangling pointer. I *think* the RCU callback
in dev_map_delete_elem() and the rcu_barrier() in dev_map_free()
protects against this, but that is by no means obvious. So confirming
this, and explaining it in a comment would be good.
Yes, *most* of the READ_ONCE(ri->map) are removed, it's pretty much only
the bpf_redirect_map(), and as you write, the tracepoints.
The content/element of the map is RCU protected, and actually even the
map will be around until the XDP processing is complete. Note the
synchronize_rcu() followed after all bpf_clear_redirect_map() calls.
I'll try to make it clearer in the commit message! Thanks for pointing
that out!
Also, as far as I can tell after this, ri->map is only used for the
tracepoint. So how about just storing the map ID and getting rid of the
READ/WRITE_ONCE() entirely?
...and the bpf_redirect_map() helper. Don't you think the current
READ_ONCE(ri->map) scheme is more obvious/clear?
(Oh, and related to this I think this patch set will conflict with
Hangbin's multi-redirect series, so maybe you two ought to coordinate? :))
Yeah, good idea! I would guess Hangbin's would go in before this, so I
would need to adapt.
Thanks for taking of look at the series, Toke! Much appreciated!
Björn
On 2021-01-20 13:52, Toke Høiland-Jørgensen wrote:
quoted
Björn Töpel [off-list ref] writes:
quoted
From: Björn Töpel <redacted>
Add detection for kernel version, and adapt the BPF program based on
kernel support. This way, users will get the best possible performance
from the BPF program.
Please do explicit feature detection instead of relying on the kernel
version number; some distro kernels are known to have a creative notion
of their own version, which is not really related to the features they
actually support (I'm sure you know which one I'm referring to ;)).
Right. For a *new* helper, like bpf_redirect_xsk, we rely on rejection
from the verifier to detect support. What about "bpf_redirect_map() now
supports passing return value as flags"? Any ideas how to do that in a
robust, non-version number-based scheme?
Well, having a BPF program pass in a flag of '1' with an invalid lookup
and checking if it returns 1 or 0. But how to do that from libbpf, hmm,
good question. BPF_PROG_RUN()?
An alternative could be to default to a program that will handle both
cases in the BPF code, and make it opt-in to use the optimised versions
if the user knows their kernel supports it?
-Toke
On 2021-01-20 13:52, Toke Høiland-Jørgensen wrote:
quoted
Björn Töpel [off-list ref] writes:
quoted
From: Björn Töpel <redacted>
Add detection for kernel version, and adapt the BPF program based on
kernel support. This way, users will get the best possible performance
from the BPF program.
Please do explicit feature detection instead of relying on the kernel
version number; some distro kernels are known to have a creative notion
of their own version, which is not really related to the features they
actually support (I'm sure you know which one I'm referring to ;)).
Right. For a *new* helper, like bpf_redirect_xsk, we rely on rejection
from the verifier to detect support. What about "bpf_redirect_map() now
supports passing return value as flags"? Any ideas how to do that in a
robust, non-version number-based scheme?
Just so that I understand this correctly. Red^WSome distro vendors
backport the world, and call that franken kernel, say, 3.10. Is that
interpretation correct? My hope was that wasn't the case. :-(
Yup, indeed. All kernels shipped for the entire lifetime of RHEL8 think
they are v4.18.0... :/
I don't think we're the only ones doing it (there are examples in the
embedded world as well, for instance, and not sure about the other
enterprise distros), but RHEL is probably the most extreme example.
We could patch the version check in the distro-supplied version of
libbpf, of course, but that doesn't help anyone using upstream versions,
and given the prevalence of vendoring libbpf, I fear that going with the
version check will just result in a bad user experience...
Would it make sense with some kind of BPF-specific "supported
features" mechanism? Something else with a bigger scope (whole
kernel)?
Heh, in my opinion, yeah. Seems like we'll finally get it for XDP, but
for BPF in general the approach has always been probing AFAICT.
For the particular case of arguments to helpers, I suppose the verifier
could technically validate value ranges for flags arguments, say. That
would be nice as an early reject anyway, but I'm not sure if it is
possible to add after-the-fact without breaking existing programs
because the verifier can't prove the argument is within the valid range.
And of course it doesn't help you with compatibility with
already-released kernels.
-Toke
@@ -3836,6 +3836,12 @@ union bpf_attr {*Return*ApointertoastructsocketonsuccessorNULLifthefileis*notasocket.+*+*longbpf_redirect_xsk(structxdp_buff*xdp_md,u64action)+*Description+*RedirecttotheregisteredAF_XDPsocket.+*Return+***XDP_REDIRECT**onsuccess,otherwisetheactionparameterisreturned.*/
I think it would be better to make the second argument a 'flags'
argument and make values > XDP_TX invalid (like we do in
bpf_xdp_redirect_map() now). By allowing any value as return you lose
the ability to turn it into a flags argument later...
Yes, but that adds a run-time check. I prefer this non-checked version,
even though it is a bit less futureproof.
That...seems a bit short-sighted? :)
Can you actually see a difference in your performance numbers?
I would rather add an additional helper *if* we see the need for flags,
instead of paying for that upfront. For me, BPF is about being able to
specialize, and not having one call with tons of checks.
(Related; Going forward, the growing switch() for redirect targets in
xdp_do_redirect() is a concern for me...)
And yes, even with all those fancy branch predictors, less instructions
is still less. :-) (It shows in my ubenchs.)
Björn
From: Björn Töpel <hidden> Date: 2021-01-20 15:32:23
On 2021-01-20 16:11, Toke Høiland-Jørgensen wrote:
Björn Töpel [off-list ref] writes:
quoted
On 2021-01-20 14:25, Björn Töpel wrote:
quoted
On 2021-01-20 13:52, Toke Høiland-Jørgensen wrote:
quoted
Björn Töpel [off-list ref] writes:
quoted
From: Björn Töpel <redacted>
Add detection for kernel version, and adapt the BPF program based on
kernel support. This way, users will get the best possible performance
from the BPF program.
Please do explicit feature detection instead of relying on the kernel
version number; some distro kernels are known to have a creative notion
of their own version, which is not really related to the features they
actually support (I'm sure you know which one I'm referring to ;)).
Right. For a *new* helper, like bpf_redirect_xsk, we rely on rejection
from the verifier to detect support. What about "bpf_redirect_map() now
supports passing return value as flags"? Any ideas how to do that in a
robust, non-version number-based scheme?
Just so that I understand this correctly. Red^WSome distro vendors
backport the world, and call that franken kernel, say, 3.10. Is that
interpretation correct? My hope was that wasn't the case. :-(
Yup, indeed. All kernels shipped for the entire lifetime of RHEL8 think
they are v4.18.0... :/
I don't think we're the only ones doing it (there are examples in the
embedded world as well, for instance, and not sure about the other
enterprise distros), but RHEL is probably the most extreme example.
We could patch the version check in the distro-supplied version of
libbpf, of course, but that doesn't help anyone using upstream versions,
and given the prevalence of vendoring libbpf, I fear that going with the
version check will just result in a bad user experience...
Ok! Thanks for clearing that out!
quoted
Would it make sense with some kind of BPF-specific "supported
features" mechanism? Something else with a bigger scope (whole
kernel)?
Heh, in my opinion, yeah. Seems like we'll finally get it for XDP, but
for BPF in general the approach has always been probing AFAICT.
For the particular case of arguments to helpers, I suppose the verifier
could technically validate value ranges for flags arguments, say. That
would be nice as an early reject anyway, but I'm not sure if it is
possible to add after-the-fact without breaking existing programs
because the verifier can't prove the argument is within the valid range.
And of course it doesn't help you with compatibility with
already-released kernels.
Hmm, think I have a way forward. I'll use BPF_PROG_TEST_RUN.
If the load fail for the new helper, fallback to bpf_redirect_map(). Use
BPF_PROG_TEST_RUN to make sure that "action via flags" passes.
That should work for you guys as well, right? I'll take a stab at it.
Björn
@@ -3836,6 +3836,12 @@ union bpf_attr {*Return*ApointertoastructsocketonsuccessorNULLifthefileis*notasocket.+*+*longbpf_redirect_xsk(structxdp_buff*xdp_md,u64action)+*Description+*RedirecttotheregisteredAF_XDPsocket.+*Return+***XDP_REDIRECT**onsuccess,otherwisetheactionparameterisreturned.*/
I think it would be better to make the second argument a 'flags'
argument and make values > XDP_TX invalid (like we do in
bpf_xdp_redirect_map() now). By allowing any value as return you lose
the ability to turn it into a flags argument later...
Yes, but that adds a run-time check. I prefer this non-checked version,
even though it is a bit less futureproof.
That...seems a bit short-sighted? :)
Can you actually see a difference in your performance numbers?
-Toke
From: Björn Töpel <hidden> Date: 2021-01-20 15:37:32
On 2021-01-20 14:25, Björn Töpel wrote:
On 2021-01-20 13:52, Toke Høiland-Jørgensen wrote:
quoted
Björn Töpel [off-list ref] writes:
quoted
From: Björn Töpel <redacted>
Add detection for kernel version, and adapt the BPF program based on
kernel support. This way, users will get the best possible performance
from the BPF program.
Please do explicit feature detection instead of relying on the kernel
version number; some distro kernels are known to have a creative notion
of their own version, which is not really related to the features they
actually support (I'm sure you know which one I'm referring to ;)).
Right. For a *new* helper, like bpf_redirect_xsk, we rely on rejection
from the verifier to detect support. What about "bpf_redirect_map() now
supports passing return value as flags"? Any ideas how to do that in a
robust, non-version number-based scheme?
Just so that I understand this correctly. Red^WSome distro vendors
backport the world, and call that franken kernel, say, 3.10. Is that
interpretation correct? My hope was that wasn't the case. :-(
Would it make sense with some kind of BPF-specific "supported features"
mechanism? Something else with a bigger scope (whole kernel)?
Cheers,
Björn
On 2021-01-20 13:44, Toke Høiland-Jørgensen wrote:
quoted
Björn Töpel [off-list ref] writes:
quoted
From: Björn Töpel <redacted>
The XDP_REDIRECT implementations for maps and non-maps are fairly
similar, but obviously need to take different code paths depending on
if the target is using a map or not. Today, the redirect targets for
XDP either uses a map, or is based on ifindex.
Future commits will introduce yet another redirect target via the a
new helper, bpf_redirect_xsk(). To pave the way for that, we introduce
an explicit redirect type to bpf_redirect_info. This makes the code
easier to follow, and makes it easier to add new redirect targets.
Further, using an explicit type in bpf_redirect_info has a slight
positive performance impact by avoiding a pointer indirection for the
map type lookup, and instead use the hot cacheline for
bpf_redirect_info.
The bpf_redirect_info flags member is not used by XDP, and not
read/written any more. The map member is only written to when
required/used, and not unconditionally.
I like the simplification. However, the handling of map clearing becomes
a bit murky with this change:
You're not changing anything in bpf_clear_redirect_map(), and you're
removing most of the reads and writes of ri->map. Instead,
bpf_xdp_redirect_map() will store the bpf_dtab_netdev pointer in
ri->tgt_value, which xdp_do_redirect() will just read and use without
checking. But if the map element (or the entire map) has been freed in
the meantime that will be a dangling pointer. I *think* the RCU callback
in dev_map_delete_elem() and the rcu_barrier() in dev_map_free()
protects against this, but that is by no means obvious. So confirming
this, and explaining it in a comment would be good.
Yes, *most* of the READ_ONCE(ri->map) are removed, it's pretty much only
the bpf_redirect_map(), and as you write, the tracepoints.
The content/element of the map is RCU protected, and actually even the
map will be around until the XDP processing is complete. Note the
synchronize_rcu() followed after all bpf_clear_redirect_map() calls.
I'll try to make it clearer in the commit message! Thanks for pointing
that out!
quoted
Also, as far as I can tell after this, ri->map is only used for the
tracepoint. So how about just storing the map ID and getting rid of the
READ/WRITE_ONCE() entirely?
...and the bpf_redirect_map() helper. Don't you think the current
READ_ONCE(ri->map) scheme is more obvious/clear?
Yeah, after your patch we WRITE_ONCE() the pointer in
bpf_redirect_map(), but the only place it is actually *read* is in the
tracepoint. So the only purpose of bpf_clear_redirect_map() is to ensure
that an invalid pointer is not read in the tracepoint function. Which
seems a bit excessive when we could just store the map ID for direct use
in the tracepoint and get rid of bpf_clear_redirect_map() entirely, no?
Besides, from a UX point of view, having the tracepoint display the map
ID even if that map ID is no longer valid seems to me like it makes more
sense than just displaying a map ID of 0 and leaving it up to the user
to figure out that this is because the map was cleared. I mean, at the
time the redirect was made, that *was* the map ID that was used...
Oh, and as you say due to the synchronize_rcu() call in dev_map_free() I
think this whole discussion is superfluous anyway, since it can't
actually happen that the map gets freed between the setting and reading
of ri->map, no?
quoted
(Oh, and related to this I think this patch set will conflict with
Hangbin's multi-redirect series, so maybe you two ought to coordinate? :))
Yeah, good idea! I would guess Hangbin's would go in before this, so I
would need to adapt.
Thanks for taking of look at the series, Toke! Much appreciated!
On Wed, 20 Jan 2021 15:15:22 +0200
Maxim Mikityanskiy [off-list ref] wrote:
On 2021-01-19 17:50, Björn Töpel wrote:
quoted
This series extends bind() for XDP sockets, so that the bound socket
is added to the netdev_rx_queue _rx array in the netdevice. We call
this to register the socket. To redirect packets to the registered
socket, a new BPF helper is used: bpf_redirect_xsk().
For shared XDP sockets, only the first bound socket is
registered. Users that need more complex setup has to use XSKMAP and
bpf_redirect_map().
Now, why would one use bpf_redirect_xsk() over the regular
bpf_redirect_map() helper?
* Better performance!
* Convenience; Most user use one socket per queue. This scenario is
what registered sockets support. There is no need to create an
XSKMAP. This can also reduce complexity from containerized setups,
where users might what to use XDP sockets without CAP_SYS_ADMIN
capabilities.
I'm buying into the convenience and reduce complexity, and XDP sockets
without CAP_SYS_ADMIN into containers.
People might be surprised that I'm actually NOT buying into the better
performance argument here. At these speeds we are basically comparing
how close we are to zero (and have to use nanosec time scale for our
comparisons), more below.
quoted
The first patch restructures xdp_do_redirect() a bit, to make it
easier to add the new helper. This restructure also give us a slight
performance benefit. The following three patches extends bind() and
adds the new helper. After that, two libbpf patches that selects XDP
program based on what kernel is running. Finally, selftests for the new
functionality is added.
Note that the libbpf "auto-selection" is based on kernel version, so
it is hard coded to the "-next" version (5.12). If you would like to
try this is out, you will need to change the libbpf patch locally!
Thanks to Maciej and Magnus for the internal review/comments!
Performance (rxdrop, zero-copy)
Baseline
Two cores: 21.3 Mpps
One core: 24.5 Mpps
Two cores is slower? It used to be faster all the time, didn't it?
quoted
Patched
Two cores, bpf_redirect_map: 21.7 Mpps + 2%
One core, bpf_redirect_map: 24.9 Mpps + 2%
Two cores, bpf_redirect_xsk: 24.0 Mpps +13%
Nice, impressive improvement!
I do appreciate you work and performance optimizations at this level,
because when we are using this few CPU cycles per packet, then it is
really hard to find new ways to reduce cycles further.
Thank for you saying +13% instead of saying +2.7 Mpps.
It *is* impressive to basically reduce cycles with 13%.
21.3 Mpps = 46.94 nanosec per packet
24.0 Mpps = 41.66 nanosec per packet
21.3 Mpps -> 24.0 Mpps = 5.28 nanosec saved
On my 3.60GHz testlab machine that gives me 19 cycles.
From: Maciej Fijalkowski <maciej.fijalkowski@intel.com> Date: 2021-01-20 16:31:44
On Wed, Jan 20, 2021 at 04:57:08PM +0100, Jesper Dangaard Brouer wrote:
On Wed, 20 Jan 2021 15:15:22 +0200
Maxim Mikityanskiy [off-list ref] wrote:
quoted
On 2021-01-19 17:50, Björn Töpel wrote:
quoted
This series extends bind() for XDP sockets, so that the bound socket
is added to the netdev_rx_queue _rx array in the netdevice. We call
this to register the socket. To redirect packets to the registered
socket, a new BPF helper is used: bpf_redirect_xsk().
For shared XDP sockets, only the first bound socket is
registered. Users that need more complex setup has to use XSKMAP and
bpf_redirect_map().
Now, why would one use bpf_redirect_xsk() over the regular
bpf_redirect_map() helper?
* Better performance!
* Convenience; Most user use one socket per queue. This scenario is
what registered sockets support. There is no need to create an
XSKMAP. This can also reduce complexity from containerized setups,
where users might what to use XDP sockets without CAP_SYS_ADMIN
capabilities.
I'm buying into the convenience and reduce complexity, and XDP sockets
without CAP_SYS_ADMIN into containers.
People might be surprised that I'm actually NOT buying into the better
performance argument here. At these speeds we are basically comparing
how close we are to zero (and have to use nanosec time scale for our
comparisons), more below.
quoted
quoted
The first patch restructures xdp_do_redirect() a bit, to make it
easier to add the new helper. This restructure also give us a slight
performance benefit. The following three patches extends bind() and
adds the new helper. After that, two libbpf patches that selects XDP
program based on what kernel is running. Finally, selftests for the new
functionality is added.
Note that the libbpf "auto-selection" is based on kernel version, so
it is hard coded to the "-next" version (5.12). If you would like to
try this is out, you will need to change the libbpf patch locally!
Thanks to Maciej and Magnus for the internal review/comments!
Performance (rxdrop, zero-copy)
Baseline
Two cores: 21.3 Mpps
One core: 24.5 Mpps
Two cores is slower? It used to be faster all the time, didn't it?
quoted
Patched
Two cores, bpf_redirect_map: 21.7 Mpps + 2%
One core, bpf_redirect_map: 24.9 Mpps + 2%
Two cores, bpf_redirect_xsk: 24.0 Mpps +13%
Nice, impressive improvement!
I do appreciate you work and performance optimizations at this level,
because when we are using this few CPU cycles per packet, then it is
really hard to find new ways to reduce cycles further.
Thank for you saying +13% instead of saying +2.7 Mpps.
It *is* impressive to basically reduce cycles with 13%.
21.3 Mpps = 46.94 nanosec per packet
24.0 Mpps = 41.66 nanosec per packet
21.3 Mpps -> 24.0 Mpps = 5.28 nanosec saved
On my 3.60GHz testlab machine that gives me 19 cycles.
quoted
quoted
One core, bpf_redirect_xsk: 25.5 Mpps + 4%
24.5 Mpps -> 25.5 Mpps = 1.6 nanosec saved
At this point with saving 1.6 ns this is around the cost of a function call 1.3 ns.
We still need these optimization in the kernel, but end-users in
userspace are very quickly going to waste the 19 cycles we found.
I still support/believe that the OS need to have a little overhead as
possible, but for me 42 nanosec overhead is close to zero overhead. For
comparison, I just ran a udp_sink[3] test, and it "cost" 625 ns for
delivery of UDP packets into socket (almost 15 times slower).
I guess my point is that with XDP we have already achieved and exceeded
(my original) performance goals, making it even faster is just showing off ;-P
Even though I'll let Bjorn elaborating on this, we're talking here about
AF-XDP which is a bit different pair of shoes to me in terms of
performance. AFAIK we still have a gap when compared to DPDK's numbers. So
I'm really not sure why better performance bothers you? :)
Let's rather be more harsh on changes that actually decrease the
performance, not the other way around. And I suppose you were the one that
always was bringing up the 'death by a 1000 paper cuts' of XDP. So yeah,
I'm a bit confused with your statement.
On 2021-01-20 15:52, Toke Høiland-Jørgensen wrote:
quoted
Björn Töpel [off-list ref] writes:
quoted
On 2021-01-20 13:44, Toke Høiland-Jørgensen wrote:
quoted
Björn Töpel [off-list ref] writes:
quoted
From: Björn Töpel <redacted>
The XDP_REDIRECT implementations for maps and non-maps are fairly
similar, but obviously need to take different code paths depending on
if the target is using a map or not. Today, the redirect targets for
XDP either uses a map, or is based on ifindex.
Future commits will introduce yet another redirect target via the a
new helper, bpf_redirect_xsk(). To pave the way for that, we introduce
an explicit redirect type to bpf_redirect_info. This makes the code
easier to follow, and makes it easier to add new redirect targets.
Further, using an explicit type in bpf_redirect_info has a slight
positive performance impact by avoiding a pointer indirection for the
map type lookup, and instead use the hot cacheline for
bpf_redirect_info.
The bpf_redirect_info flags member is not used by XDP, and not
read/written any more. The map member is only written to when
required/used, and not unconditionally.
I like the simplification. However, the handling of map clearing becomes
a bit murky with this change:
You're not changing anything in bpf_clear_redirect_map(), and you're
removing most of the reads and writes of ri->map. Instead,
bpf_xdp_redirect_map() will store the bpf_dtab_netdev pointer in
ri->tgt_value, which xdp_do_redirect() will just read and use without
checking. But if the map element (or the entire map) has been freed in
the meantime that will be a dangling pointer. I *think* the RCU callback
in dev_map_delete_elem() and the rcu_barrier() in dev_map_free()
protects against this, but that is by no means obvious. So confirming
this, and explaining it in a comment would be good.
Yes, *most* of the READ_ONCE(ri->map) are removed, it's pretty much only
the bpf_redirect_map(), and as you write, the tracepoints.
The content/element of the map is RCU protected, and actually even the
map will be around until the XDP processing is complete. Note the
synchronize_rcu() followed after all bpf_clear_redirect_map() calls.
I'll try to make it clearer in the commit message! Thanks for pointing
that out!
quoted
Also, as far as I can tell after this, ri->map is only used for the
tracepoint. So how about just storing the map ID and getting rid of the
READ/WRITE_ONCE() entirely?
...and the bpf_redirect_map() helper. Don't you think the current
READ_ONCE(ri->map) scheme is more obvious/clear?
Yeah, after your patch we WRITE_ONCE() the pointer in
bpf_redirect_map(), but the only place it is actually *read* is in the
tracepoint. So the only purpose of bpf_clear_redirect_map() is to ensure
that an invalid pointer is not read in the tracepoint function. Which
seems a bit excessive when we could just store the map ID for direct use
in the tracepoint and get rid of bpf_clear_redirect_map() entirely, no?
Besides, from a UX point of view, having the tracepoint display the map
ID even if that map ID is no longer valid seems to me like it makes more
sense than just displaying a map ID of 0 and leaving it up to the user
to figure out that this is because the map was cleared. I mean, at the
time the redirect was made, that *was* the map ID that was used...
Convinced! Getting rid of bpf_clear_redirect_map() would be good! I'll
take a stab at this for v3!
Cool!
quoted
Oh, and as you say due to the synchronize_rcu() call in dev_map_free() I
think this whole discussion is superfluous anyway, since it can't
actually happen that the map gets freed between the setting and reading
of ri->map, no?
It can't be free'd but, ri->map can be cleared via
bpf_clear_redirect_map(). So, between the helper (setting) and the
tracepoint in xdp_do_redirect() it can be cleared (say if the XDP
program is swapped out, prior running xdp_do_redirect()).
But xdp_do_redirect() should be called on driver flush before exiting
the NAPI cycle, so how can the XDP program be swapped out?
Moving to the scheme you suggested, does make the discussion
superfluous. :-)
From: Björn Töpel <hidden> Date: 2021-01-20 17:02:43
On 2021-01-20 15:52, Toke Høiland-Jørgensen wrote:
Björn Töpel [off-list ref] writes:
quoted
On 2021-01-20 13:44, Toke Høiland-Jørgensen wrote:
quoted
Björn Töpel [off-list ref] writes:
quoted
From: Björn Töpel <redacted>
The XDP_REDIRECT implementations for maps and non-maps are fairly
similar, but obviously need to take different code paths depending on
if the target is using a map or not. Today, the redirect targets for
XDP either uses a map, or is based on ifindex.
Future commits will introduce yet another redirect target via the a
new helper, bpf_redirect_xsk(). To pave the way for that, we introduce
an explicit redirect type to bpf_redirect_info. This makes the code
easier to follow, and makes it easier to add new redirect targets.
Further, using an explicit type in bpf_redirect_info has a slight
positive performance impact by avoiding a pointer indirection for the
map type lookup, and instead use the hot cacheline for
bpf_redirect_info.
The bpf_redirect_info flags member is not used by XDP, and not
read/written any more. The map member is only written to when
required/used, and not unconditionally.
I like the simplification. However, the handling of map clearing becomes
a bit murky with this change:
You're not changing anything in bpf_clear_redirect_map(), and you're
removing most of the reads and writes of ri->map. Instead,
bpf_xdp_redirect_map() will store the bpf_dtab_netdev pointer in
ri->tgt_value, which xdp_do_redirect() will just read and use without
checking. But if the map element (or the entire map) has been freed in
the meantime that will be a dangling pointer. I *think* the RCU callback
in dev_map_delete_elem() and the rcu_barrier() in dev_map_free()
protects against this, but that is by no means obvious. So confirming
this, and explaining it in a comment would be good.
Yes, *most* of the READ_ONCE(ri->map) are removed, it's pretty much only
the bpf_redirect_map(), and as you write, the tracepoints.
The content/element of the map is RCU protected, and actually even the
map will be around until the XDP processing is complete. Note the
synchronize_rcu() followed after all bpf_clear_redirect_map() calls.
I'll try to make it clearer in the commit message! Thanks for pointing
that out!
quoted
Also, as far as I can tell after this, ri->map is only used for the
tracepoint. So how about just storing the map ID and getting rid of the
READ/WRITE_ONCE() entirely?
...and the bpf_redirect_map() helper. Don't you think the current
READ_ONCE(ri->map) scheme is more obvious/clear?
Yeah, after your patch we WRITE_ONCE() the pointer in
bpf_redirect_map(), but the only place it is actually *read* is in the
tracepoint. So the only purpose of bpf_clear_redirect_map() is to ensure
that an invalid pointer is not read in the tracepoint function. Which
seems a bit excessive when we could just store the map ID for direct use
in the tracepoint and get rid of bpf_clear_redirect_map() entirely, no?
Besides, from a UX point of view, having the tracepoint display the map
ID even if that map ID is no longer valid seems to me like it makes more
sense than just displaying a map ID of 0 and leaving it up to the user
to figure out that this is because the map was cleared. I mean, at the
time the redirect was made, that *was* the map ID that was used...
Convinced! Getting rid of bpf_clear_redirect_map() would be good! I'll
take a stab at this for v3!
Oh, and as you say due to the synchronize_rcu() call in dev_map_free() I
think this whole discussion is superfluous anyway, since it can't
actually happen that the map gets freed between the setting and reading
of ri->map, no?
It can't be free'd but, ri->map can be cleared via
bpf_clear_redirect_map(). So, between the helper (setting) and the
tracepoint in xdp_do_redirect() it can be cleared (say if the XDP
program is swapped out, prior running xdp_do_redirect()).
Moving to the scheme you suggested, does make the discussion
superfluous. :-)
Thanks for the input!
Björn
From: Björn Töpel <hidden> Date: 2021-01-20 17:36:18
On 2021-01-20 17:30, Toke Høiland-Jørgensen wrote:
Björn Töpel [off-list ref] writes:
[...]
quoted
It can't be free'd but, ri->map can be cleared via
bpf_clear_redirect_map(). So, between the helper (setting) and the
tracepoint in xdp_do_redirect() it can be cleared (say if the XDP
program is swapped out, prior running xdp_do_redirect()).
But xdp_do_redirect() should be called on driver flush before exiting
the NAPI cycle, so how can the XDP program be swapped out?
To clarify; xdp_do_redirect() is called for each packet in the NAPI poll
loop.
Yeah, you're right. The xdp_do_redirect() is within the RCU scope, so
the program wont be destroyed (but can be swapped though!).
Hmm, so IOW the bpf_clear_redirect_map() is not needed anymore...
Björn
[...]
@@ -3836,6 +3836,12 @@ union bpf_attr {*Return*ApointertoastructsocketonsuccessorNULLifthefileis*notasocket.+*+*longbpf_redirect_xsk(structxdp_buff*xdp_md,u64action)+*Description+*RedirecttotheregisteredAF_XDPsocket.+*Return+***XDP_REDIRECT**onsuccess,otherwisetheactionparameterisreturned.*/
I think it would be better to make the second argument a 'flags'
argument and make values > XDP_TX invalid (like we do in
bpf_xdp_redirect_map() now). By allowing any value as return you lose
the ability to turn it into a flags argument later...
Yes, but that adds a run-time check. I prefer this non-checked version,
even though it is a bit less futureproof.
That...seems a bit short-sighted? :)
Can you actually see a difference in your performance numbers?
I would rather add an additional helper *if* we see the need for flags,
instead of paying for that upfront. For me, BPF is about being able to
specialize, and not having one call with tons of checks.
I get that, I'm just pushing back because omitting a 'flags' argument is
literally among the most frequent reasons for having to replace a
syscall (see e.g., [0]) instead of extending it. And yeah, I do realise
that the performance implications are different for XDP than for
syscalls, but maintainability of the API is also important; it's all a
tradeoff. This will be the third redirect helper variant for XDP and I'd
hate for the fourth one to have to be bpf_redirect_xsk_flags() because
it did turn out to be needed...
(One potential concrete reason for this: I believe Magnus was talking
about an API that would allow a BPF program to redirect a packet into
more than one socket (cloning it in the process), or to redirect to a
socket+another target. How would you do that with this new helper?)
[0] https://lwn.net/Articles/585415/
(Related; Going forward, the growing switch() for redirect targets in
xdp_do_redirect() is a concern for me...)
And yes, even with all those fancy branch predictors, less instructions
is still less. :-) (It shows in my ubenchs.)
Right, I do agree that the run-time performance hit of checking the flag
sucks (along with being hard to check for, cf. our parallel discussion
about version checks). So ideally this would be fixed by having the
verifier enforce the argument ranges instead; but if we merge this
without the runtime check now we can't add that later without
potentially breaking programs... :(
-Toke
On 2021-01-20 16:11, Toke Høiland-Jørgensen wrote:
quoted
Björn Töpel [off-list ref] writes:
quoted
On 2021-01-20 14:25, Björn Töpel wrote:
quoted
On 2021-01-20 13:52, Toke Høiland-Jørgensen wrote:
quoted
Björn Töpel [off-list ref] writes:
quoted
From: Björn Töpel <redacted>
Add detection for kernel version, and adapt the BPF program based on
kernel support. This way, users will get the best possible performance
from the BPF program.
Please do explicit feature detection instead of relying on the kernel
version number; some distro kernels are known to have a creative notion
of their own version, which is not really related to the features they
actually support (I'm sure you know which one I'm referring to ;)).
Right. For a *new* helper, like bpf_redirect_xsk, we rely on rejection
from the verifier to detect support. What about "bpf_redirect_map() now
supports passing return value as flags"? Any ideas how to do that in a
robust, non-version number-based scheme?
Just so that I understand this correctly. Red^WSome distro vendors
backport the world, and call that franken kernel, say, 3.10. Is that
interpretation correct? My hope was that wasn't the case. :-(
Yup, indeed. All kernels shipped for the entire lifetime of RHEL8 think
they are v4.18.0... :/
I don't think we're the only ones doing it (there are examples in the
embedded world as well, for instance, and not sure about the other
enterprise distros), but RHEL is probably the most extreme example.
We could patch the version check in the distro-supplied version of
libbpf, of course, but that doesn't help anyone using upstream versions,
and given the prevalence of vendoring libbpf, I fear that going with the
version check will just result in a bad user experience...
Ok! Thanks for clearing that out!
quoted
quoted
Would it make sense with some kind of BPF-specific "supported
features" mechanism? Something else with a bigger scope (whole
kernel)?
Heh, in my opinion, yeah. Seems like we'll finally get it for XDP, but
for BPF in general the approach has always been probing AFAICT.
For the particular case of arguments to helpers, I suppose the verifier
could technically validate value ranges for flags arguments, say. That
would be nice as an early reject anyway, but I'm not sure if it is
possible to add after-the-fact without breaking existing programs
because the verifier can't prove the argument is within the valid range.
And of course it doesn't help you with compatibility with
already-released kernels.
Hmm, think I have a way forward. I'll use BPF_PROG_TEST_RUN.
If the load fail for the new helper, fallback to bpf_redirect_map(). Use
BPF_PROG_TEST_RUN to make sure that "action via flags" passes.
That should work for you guys as well, right? I'll take a stab at it.
@@ -3836,6 +3836,12 @@ union bpf_attr {*Return*ApointertoastructsocketonsuccessorNULLifthefileis*notasocket.+*+*longbpf_redirect_xsk(structxdp_buff*xdp_md,u64action)+*Description+*RedirecttotheregisteredAF_XDPsocket.+*Return+***XDP_REDIRECT**onsuccess,otherwisetheactionparameterisreturned.*/
I think it would be better to make the second argument a 'flags'
argument and make values > XDP_TX invalid (like we do in
bpf_xdp_redirect_map() now). By allowing any value as return you lose
the ability to turn it into a flags argument later...
Yes, but that adds a run-time check. I prefer this non-checked version,
even though it is a bit less futureproof.
That...seems a bit short-sighted? :)
Can you actually see a difference in your performance numbers?
I would rather add an additional helper *if* we see the need for flags,
instead of paying for that upfront. For me, BPF is about being able to
specialize, and not having one call with tons of checks.
I get that, I'm just pushing back because omitting a 'flags' argument is
literally among the most frequent reasons for having to replace a
syscall (see e.g., [0]) instead of extending it. And yeah, I do realise
that the performance implications are different for XDP than for
syscalls, but maintainability of the API is also important; it's all a
tradeoff. This will be the third redirect helper variant for XDP and I'd
hate for the fourth one to have to be bpf_redirect_xsk_flags() because
it did turn out to be needed...
(One potential concrete reason for this: I believe Magnus was talking
about an API that would allow a BPF program to redirect a packet into
more than one socket (cloning it in the process), or to redirect to a
socket+another target. How would you do that with this new helper?)
[0] https://lwn.net/Articles/585415/
I have a bit of different view. One of the really nice parts about BPF
is exactly specialization. A user can tailor the kernel do a specific
thing. I *don't* see an issue with yet another helper, if that is needed
in the future. I think that is better than bloated helpers trying to
cope for all scenarios. I don't mean we should just add helpers all over
the place, but I do see more lightly on adding helpers, than adding
syscalls.
Elaborating a bit on this: many device drivers try to handle all the
things in the fast-path. I see BPF as one way forward to moving away
from that. Setup what you need, and only run what you currently need,
instead of the current "Is bleh on, then baz? Is this on, then that."
So, I would like to avoid "future proofing" the helpers, if that makes
sense. Use what you need. That's why BPF is so good (one of the things)!
As for bpf_redirect_xsk() it's a leaner version of bpf_redirect_map().
You want flags/shared sockets/...? Well go use bpf_redirect_map() and
XSKMAP. bpf_redirect_xsk() is not for you.
A lot of back-and-forth for *one* if-statement, but it's kind of a
design thing for me. ;-)
Björn
quoted
(Related; Going forward, the growing switch() for redirect targets in
xdp_do_redirect() is a concern for me...)
And yes, even with all those fancy branch predictors, less instructions
is still less. :-) (It shows in my ubenchs.)
Right, I do agree that the run-time performance hit of checking the flag
sucks (along with being hard to check for, cf. our parallel discussion
about version checks). So ideally this would be fixed by having the
verifier enforce the argument ranges instead; but if we merge this
without the runtime check now we can't add that later without
potentially breaking programs... :(
-Toke
On Wed, Jan 20, 2021 at 7:27 AM Björn Töpel [off-list ref] wrote:
quoted
quoted
Would it make sense with some kind of BPF-specific "supported
features" mechanism? Something else with a bigger scope (whole
kernel)?
Heh, in my opinion, yeah. Seems like we'll finally get it for XDP, but
for BPF in general the approach has always been probing AFAICT.
For the particular case of arguments to helpers, I suppose the verifier
could technically validate value ranges for flags arguments, say. That
would be nice as an early reject anyway, but I'm not sure if it is
possible to add after-the-fact without breaking existing programs
because the verifier can't prove the argument is within the valid range.
And of course it doesn't help you with compatibility with
already-released kernels.
Hmm, think I have a way forward. I'll use BPF_PROG_TEST_RUN.
If the load fail for the new helper, fallback to bpf_redirect_map(). Use
BPF_PROG_TEST_RUN to make sure that "action via flags" passes.
+1 to Toke's point. No version checks please.
One way to detect is to try prog_load. Search for FEAT_* in libbpf.
Another approach is to scan vmlinux BTF for necessary helpers.
Currently libbpf is relying on the former.
I think going forward would be good to detect features via BTF.
It's going to be much faster and won't create noise for audit that
could be looking at prog_load calls.
From: Björn Töpel <hidden> Date: 2021-01-20 18:34:24
On 2021-01-20 19:25, Alexei Starovoitov wrote:
On Wed, Jan 20, 2021 at 7:27 AM Björn Töpel [off-list ref] wrote:
quoted
quoted
quoted
Would it make sense with some kind of BPF-specific "supported
features" mechanism? Something else with a bigger scope (whole
kernel)?
Heh, in my opinion, yeah. Seems like we'll finally get it for XDP, but
for BPF in general the approach has always been probing AFAICT.
For the particular case of arguments to helpers, I suppose the verifier
could technically validate value ranges for flags arguments, say. That
would be nice as an early reject anyway, but I'm not sure if it is
possible to add after-the-fact without breaking existing programs
because the verifier can't prove the argument is within the valid range.
And of course it doesn't help you with compatibility with
already-released kernels.
Hmm, think I have a way forward. I'll use BPF_PROG_TEST_RUN.
If the load fail for the new helper, fallback to bpf_redirect_map(). Use
BPF_PROG_TEST_RUN to make sure that "action via flags" passes.
+1 to Toke's point. No version checks please.
One way to detect is to try prog_load. Search for FEAT_* in libbpf.
Another approach is to scan vmlinux BTF for necessary helpers.
Currently libbpf is relying on the former.
I think going forward would be good to detect features via BTF.
It's going to be much faster and won't create noise for audit that
could be looking at prog_load calls.
Thanks Alexei. I'll explore both options for the next spin!
Björn
From: Björn Töpel <hidden> Date: 2021-01-20 20:10:57
On 2021-01-20 14:15, Maxim Mikityanskiy wrote:
On 2021-01-19 17:50, Björn Töpel wrote:
quoted
This series extends bind() for XDP sockets, so that the bound socket
is added to the netdev_rx_queue _rx array in the netdevice. We call
this to register the socket. To redirect packets to the registered
socket, a new BPF helper is used: bpf_redirect_xsk().
For shared XDP sockets, only the first bound socket is
registered. Users that need more complex setup has to use XSKMAP and
bpf_redirect_map().
Now, why would one use bpf_redirect_xsk() over the regular
bpf_redirect_map() helper?
* Better performance!
* Convenience; Most user use one socket per queue. This scenario is
what registered sockets support. There is no need to create an
XSKMAP. This can also reduce complexity from containerized setups,
where users might what to use XDP sockets without CAP_SYS_ADMIN
capabilities.
The first patch restructures xdp_do_redirect() a bit, to make it
easier to add the new helper. This restructure also give us a slight
performance benefit. The following three patches extends bind() and
adds the new helper. After that, two libbpf patches that selects XDP
program based on what kernel is running. Finally, selftests for the new
functionality is added.
Note that the libbpf "auto-selection" is based on kernel version, so
it is hard coded to the "-next" version (5.12). If you would like to
try this is out, you will need to change the libbpf patch locally!
Thanks to Maciej and Magnus for the internal review/comments!
Performance (rxdrop, zero-copy)
Baseline
Two cores: 21.3 Mpps
One core: 24.5 Mpps
Two cores is slower? It used to be faster all the time, didn't it?
Up until busy-polling. Note that I'm using a busy-poll napi budget of 512.
quoted
Patched
Two cores, bpf_redirect_map: 21.7 Mpps + 2%
One core, bpf_redirect_map: 24.9 Mpps + 2%
Two cores, bpf_redirect_xsk: 24.0 Mpps +13%
Nice, impressive improvement!
Thanks! Getting rid of the queue/netdev checks really payed off!
Björn
From: Björn Töpel <hidden> Date: 2021-01-20 20:10:58
On 2021-01-20 13:52, Toke Høiland-Jørgensen wrote:
Björn Töpel [off-list ref] writes:
quoted
From: Björn Töpel <redacted>
Add detection for kernel version, and adapt the BPF program based on
kernel support. This way, users will get the best possible performance
from the BPF program.
Please do explicit feature detection instead of relying on the kernel
version number; some distro kernels are known to have a creative notion
of their own version, which is not really related to the features they
actually support (I'm sure you know which one I'm referring to ;)).
Right. For a *new* helper, like bpf_redirect_xsk, we rely on rejection
from the verifier to detect support. What about "bpf_redirect_map() now
supports passing return value as flags"? Any ideas how to do that in a
robust, non-version number-based scheme?
Thanks,
Björn
@@ -3836,6 +3836,12 @@ union bpf_attr {*Return*ApointertoastructsocketonsuccessorNULLifthefileis*notasocket.+*+*longbpf_redirect_xsk(structxdp_buff*xdp_md,u64action)+*Description+*RedirecttotheregisteredAF_XDPsocket.+*Return+***XDP_REDIRECT**onsuccess,otherwisetheactionparameterisreturned.*/
I think it would be better to make the second argument a 'flags'
argument and make values > XDP_TX invalid (like we do in
bpf_xdp_redirect_map() now). By allowing any value as return you lose
the ability to turn it into a flags argument later...
Yes, but that adds a run-time check. I prefer this non-checked version,
even though it is a bit less futureproof.
Björn
From: Björn Töpel <redacted>
Add detection for kernel version, and adapt the BPF program based on
kernel support. This way, users will get the best possible performance
from the BPF program.
Please do explicit feature detection instead of relying on the kernel
version number; some distro kernels are known to have a creative notion
of their own version, which is not really related to the features they
actually support (I'm sure you know which one I'm referring to ;)).
-Toke
From: Björn Töpel <redacted>
The XDP_REDIRECT implementations for maps and non-maps are fairly
similar, but obviously need to take different code paths depending on
if the target is using a map or not. Today, the redirect targets for
XDP either uses a map, or is based on ifindex.
Future commits will introduce yet another redirect target via the a
new helper, bpf_redirect_xsk(). To pave the way for that, we introduce
an explicit redirect type to bpf_redirect_info. This makes the code
easier to follow, and makes it easier to add new redirect targets.
Further, using an explicit type in bpf_redirect_info has a slight
positive performance impact by avoiding a pointer indirection for the
map type lookup, and instead use the hot cacheline for
bpf_redirect_info.
The bpf_redirect_info flags member is not used by XDP, and not
read/written any more. The map member is only written to when
required/used, and not unconditionally.
I like the simplification. However, the handling of map clearing becomes
a bit murky with this change:
You're not changing anything in bpf_clear_redirect_map(), and you're
removing most of the reads and writes of ri->map. Instead,
bpf_xdp_redirect_map() will store the bpf_dtab_netdev pointer in
ri->tgt_value, which xdp_do_redirect() will just read and use without
checking. But if the map element (or the entire map) has been freed in
the meantime that will be a dangling pointer. I *think* the RCU callback
in dev_map_delete_elem() and the rcu_barrier() in dev_map_free()
protects against this, but that is by no means obvious. So confirming
this, and explaining it in a comment would be good.
Also, as far as I can tell after this, ri->map is only used for the
tracepoint. So how about just storing the map ID and getting rid of the
READ/WRITE_ONCE() entirely?
(Oh, and related to this I think this patch set will conflict with
Hangbin's multi-redirect series, so maybe you two ought to coordinate? :))
-Toke
@@ -3836,6 +3836,12 @@ union bpf_attr {*Return*ApointertoastructsocketonsuccessorNULLifthefileis*notasocket.+*+*longbpf_redirect_xsk(structxdp_buff*xdp_md,u64action)+*Description+*RedirecttotheregisteredAF_XDPsocket.+*Return+***XDP_REDIRECT**onsuccess,otherwisetheactionparameterisreturned.*/
I think it would be better to make the second argument a 'flags'
argument and make values > XDP_TX invalid (like we do in
bpf_xdp_redirect_map() now). By allowing any value as return you lose
the ability to turn it into a flags argument later...
Yes, but that adds a run-time check. I prefer this non-checked version,
even though it is a bit less futureproof.
That...seems a bit short-sighted? :)
Can you actually see a difference in your performance numbers?
I would rather add an additional helper *if* we see the need for flags,
instead of paying for that upfront. For me, BPF is about being able to
specialize, and not having one call with tons of checks.
I get that, I'm just pushing back because omitting a 'flags' argument is
literally among the most frequent reasons for having to replace a
syscall (see e.g., [0]) instead of extending it. And yeah, I do realise
that the performance implications are different for XDP than for
syscalls, but maintainability of the API is also important; it's all a
tradeoff. This will be the third redirect helper variant for XDP and I'd
hate for the fourth one to have to be bpf_redirect_xsk_flags() because
it did turn out to be needed...
(One potential concrete reason for this: I believe Magnus was talking
about an API that would allow a BPF program to redirect a packet into
more than one socket (cloning it in the process), or to redirect to a
socket+another target. How would you do that with this new helper?)
[0] https://lwn.net/Articles/585415/
I have a bit of different view. One of the really nice parts about BPF
is exactly specialization. A user can tailor the kernel do a specific
thing. I *don't* see an issue with yet another helper, if that is needed
in the future. I think that is better than bloated helpers trying to
cope for all scenarios. I don't mean we should just add helpers all over
the place, but I do see more lightly on adding helpers, than adding
syscalls.
Elaborating a bit on this: many device drivers try to handle all the
things in the fast-path. I see BPF as one way forward to moving away
from that. Setup what you need, and only run what you currently need,
instead of the current "Is bleh on, then baz? Is this on, then that."
So, I would like to avoid "future proofing" the helpers, if that makes
sense. Use what you need. That's why BPF is so good (one of the
things)!
Well, it's a tradeoff. We're still defining an API that should not be
(too) confusing...
As for bpf_redirect_xsk() it's a leaner version of bpf_redirect_map().
You want flags/shared sockets/...? Well go use bpf_redirect_map() and
XSKMAP. bpf_redirect_xsk() is not for you.
This argument, however, I buy: bpf_redirect() is the single-purpose
helper for redirecting to an ifindex, bpf_redirect_xsk() is the
single-purpose helper for redirecting to an XSK, and bpf_redirect_map()
is the generic one that does both of those and more. Fair enough,
consider me convinced :)
A lot of back-and-forth for *one* if-statement, but it's kind of a
design thing for me. ;-)
Surely you don't mean to imply that you have *better* things to do with
your time than have a 10-emails-long argument over a single if
statement? ;)
-Toke
On Wed, Jan 20, 2021 at 12:26 PM Toke Høiland-Jørgensen [off-list ref] wrote:
This argument, however, I buy: bpf_redirect() is the single-purpose
helper for redirecting to an ifindex, bpf_redirect_xsk() is the
single-purpose helper for redirecting to an XSK, and bpf_redirect_map()
is the generic one that does both of those and more. Fair enough,
consider me convinced :)
quoted
A lot of back-and-forth for *one* if-statement, but it's kind of a
design thing for me. ;-)
Surely you don't mean to imply that you have *better* things to do with
your time than have a 10-emails-long argument over a single if
statement? ;)
After reading this thread I think I have to pour cold water on the design.
The performance blip comes from hard coded assumptions:
+ queue_id = xdp->rxq->queue_index;
+ xs = READ_ONCE(dev->_rx[queue_id].xsk);
bpf can have specialized helpers, but imo this is beyond what's reasonable.
Please move such things into the program and try to make
bpf_redirect_map faster.
Making af_xdp non-root is orthogonal. If there is actual need for that
it has to be designed thoroughly and not presented as "this helper may
help to do that".
I don't think "may" will materialize unless people actually work
toward the goal of non-root.
From: Björn Töpel <hidden> Date: 2021-01-21 08:24:25
On 2021-01-20 22:15, Alexei Starovoitov wrote:
On Wed, Jan 20, 2021 at 12:26 PM Toke Høiland-Jørgensen [off-list ref] wrote:
quoted
This argument, however, I buy: bpf_redirect() is the single-purpose
helper for redirecting to an ifindex, bpf_redirect_xsk() is the
single-purpose helper for redirecting to an XSK, and bpf_redirect_map()
is the generic one that does both of those and more. Fair enough,
consider me convinced :)
quoted
A lot of back-and-forth for *one* if-statement, but it's kind of a
design thing for me. ;-)
Surely you don't mean to imply that you have *better* things to do with
your time than have a 10-emails-long argument over a single if
statement? ;)
After reading this thread I think I have to pour cold water on the design.
The performance blip comes from hard coded assumptions:
+ queue_id = xdp->rxq->queue_index;
+ xs = READ_ONCE(dev->_rx[queue_id].xsk);
Yes, one can see this as a constrained map:
* The map belongs to a certain netdev.
* Each entry corresponds to a certain queue id.
I.e if we do a successful (non-NULL) lookup, we *know* that all sockets
in that map belong to the netdev, and has the correct queue id.
By doing that we can get rid of two run-time checks: "Is the socket
bound to this netdev?" and "Is this the correct queue id?".
bpf can have specialized helpers, but imo this is beyond what's reasonable.
Please move such things into the program and try to make
bpf_redirect_map faster.
I obviously prefer this path, and ideally combined with a way to even
more specialize xdp_do_redirect(). Then again, you are the maintainer! :-)
Maybe an alternative be adding a new type of XSKMAP constrained in the
similar way as above, and continue with bpf_redirect_map(), but with
this new map the index argument would be ignored. Unfortunately the BPF
context (xdp_buff in this case) is not passed to bpf_redirect_map(), so
getting the actual queue_id in the helper is hard. Adding the context as
an additional argument would be a new helper...
I'll need to think a bit more about it. Input/ideas are welcome!
Making af_xdp non-root is orthogonal. If there is actual need for that
it has to be designed thoroughly and not presented as "this helper may
help to do that".
I don't think "may" will materialize unless people actually work
toward the goal of non-root.
Fair enough! Same goal could be reached using the existing map approach.
Björn
On Wed, 20 Jan 2021 17:19:31 +0100
Maciej Fijalkowski [off-list ref] wrote:
On Wed, Jan 20, 2021 at 04:57:08PM +0100, Jesper Dangaard Brouer wrote:
quoted
On Wed, 20 Jan 2021 15:15:22 +0200
Maxim Mikityanskiy [off-list ref] wrote:
quoted
On 2021-01-19 17:50, Björn Töpel wrote:
quoted
This series extends bind() for XDP sockets, so that the bound socket
is added to the netdev_rx_queue _rx array in the netdevice. We call
this to register the socket. To redirect packets to the registered
socket, a new BPF helper is used: bpf_redirect_xsk().
For shared XDP sockets, only the first bound socket is
registered. Users that need more complex setup has to use XSKMAP and
bpf_redirect_map().
Now, why would one use bpf_redirect_xsk() over the regular
bpf_redirect_map() helper?
* Better performance!
* Convenience; Most user use one socket per queue. This scenario is
what registered sockets support. There is no need to create an
XSKMAP. This can also reduce complexity from containerized setups,
where users might what to use XDP sockets without CAP_SYS_ADMIN
capabilities.
I'm buying into the convenience and reduce complexity, and XDP sockets
without CAP_SYS_ADMIN into containers.
People might be surprised that I'm actually NOT buying into the better
performance argument here. At these speeds we are basically comparing
how close we are to zero (and have to use nanosec time scale for our
comparisons), more below.
quoted
quoted
The first patch restructures xdp_do_redirect() a bit, to make it
easier to add the new helper. This restructure also give us a slight
performance benefit. The following three patches extends bind() and
adds the new helper. After that, two libbpf patches that selects XDP
program based on what kernel is running. Finally, selftests for the new
functionality is added.
Note that the libbpf "auto-selection" is based on kernel version, so
it is hard coded to the "-next" version (5.12). If you would like to
try this is out, you will need to change the libbpf patch locally!
Thanks to Maciej and Magnus for the internal review/comments!
Performance (rxdrop, zero-copy)
Baseline
Two cores: 21.3 Mpps
One core: 24.5 Mpps
Two cores is slower? It used to be faster all the time, didn't it?
quoted
Patched
Two cores, bpf_redirect_map: 21.7 Mpps + 2%
One core, bpf_redirect_map: 24.9 Mpps + 2%
Two cores, bpf_redirect_xsk: 24.0 Mpps +13%
Nice, impressive improvement!
I do appreciate you work and performance optimizations at this level,
because when we are using this few CPU cycles per packet, then it is
really hard to find new ways to reduce cycles further.
Thank for you saying +13% instead of saying +2.7 Mpps.
It *is* impressive to basically reduce cycles with 13%.
21.3 Mpps = 46.94 nanosec per packet
24.0 Mpps = 41.66 nanosec per packet
21.3 Mpps -> 24.0 Mpps = 5.28 nanosec saved
On my 3.60GHz testlab machine that gives me 19 cycles.
quoted
quoted
One core, bpf_redirect_xsk: 25.5 Mpps + 4%
24.5 Mpps -> 25.5 Mpps = 1.6 nanosec saved
At this point with saving 1.6 ns this is around the cost of a function call 1.3 ns.
We still need these optimization in the kernel, but end-users in
userspace are very quickly going to waste the 19 cycles we found.
I still support/believe that the OS need to have a little overhead as
possible, but for me 42 nanosec overhead is close to zero overhead. For
comparison, I just ran a udp_sink[3] test, and it "cost" 625 ns for
delivery of UDP packets into socket (almost 15 times slower).
I guess my point is that with XDP we have already achieved and exceeded
(my original) performance goals, making it even faster is just showing off ;-P
Even though I'll let Bjorn elaborating on this, we're talking here about
AF-XDP which is a bit different pair of shoes to me in terms of
performance. AFAIK we still have a gap when compared to DPDK's numbers. So
AFAIK the gap on this i40e hardware is DPDK=36Mpps, and lets say
AF_XDP=25.5 Mpps, that looks large +10.5Mpps, but it is only 11.4
nanosec.
I'm really not sure why better performance bothers you? :)
Finding those 11.4 nanosec is going to be really difficult and take a
huge effort. My fear is also that some of these optimizations will make
the code harder to maintain and understand.
If you are ready to do this huge effort, then I'm actually ready to
provide ideas on what you can optimize.
E.g. you can get 2-4 nanosec if we prefetch to L2 in driver (assuming
data is DDIO in L3 already). (CPU supports L2 prefetch, but kernel
have no users of that feature).
The base design of XDP redirect API, that have a number of function
calls per packet, in itself becomes a limiting factor. Even the cost
of getting the per-CPU pointer to bpf_redirect_info becomes something
to look at.
Let's rather be more harsh on changes that actually decrease the
performance, not the other way around. And I suppose you were the one that
always was bringing up the 'death by a 1000 paper cuts' of XDP. So yeah,
I'm a bit confused with your statement.
Yes, we really need to protect the existing the performance.
I think I'm mostly demotivated by the 'death by a 1000 paper cuts'.
People with good intentions are adding features, and performance slowly
disappears. I've been spending weeks/months finding nanosec level
improvements, which I don't have time to "defend". Recently I realized
that the 2nd XDP prog on egress, even when no prog is attached, is
slowing down the performance of base redirect (I don't fully understand
why, maybe it is simply the I-cache increase, I didn't realize when
reviewing the code).
--
Best regards,
Jesper Dangaard Brouer
MSc.CS, Principal Kernel Engineer at Red Hat
LinkedIn: http://www.linkedin.com/in/brouer
From: Magnus Karlsson <hidden> Date: 2021-01-22 09:01:29
On Thu, Jan 21, 2021 at 6:07 PM Jesper Dangaard Brouer
[off-list ref] wrote:
On Wed, 20 Jan 2021 17:19:31 +0100
Maciej Fijalkowski [off-list ref] wrote:
quoted
On Wed, Jan 20, 2021 at 04:57:08PM +0100, Jesper Dangaard Brouer wrote:
quoted
On Wed, 20 Jan 2021 15:15:22 +0200
Maxim Mikityanskiy [off-list ref] wrote:
quoted
On 2021-01-19 17:50, Björn Töpel wrote:
quoted
This series extends bind() for XDP sockets, so that the bound socket
is added to the netdev_rx_queue _rx array in the netdevice. We call
this to register the socket. To redirect packets to the registered
socket, a new BPF helper is used: bpf_redirect_xsk().
For shared XDP sockets, only the first bound socket is
registered. Users that need more complex setup has to use XSKMAP and
bpf_redirect_map().
Now, why would one use bpf_redirect_xsk() over the regular
bpf_redirect_map() helper?
* Better performance!
* Convenience; Most user use one socket per queue. This scenario is
what registered sockets support. There is no need to create an
XSKMAP. This can also reduce complexity from containerized setups,
where users might what to use XDP sockets without CAP_SYS_ADMIN
capabilities.
I'm buying into the convenience and reduce complexity, and XDP sockets
without CAP_SYS_ADMIN into containers.
People might be surprised that I'm actually NOT buying into the better
performance argument here. At these speeds we are basically comparing
how close we are to zero (and have to use nanosec time scale for our
comparisons), more below.
quoted
quoted
The first patch restructures xdp_do_redirect() a bit, to make it
easier to add the new helper. This restructure also give us a slight
performance benefit. The following three patches extends bind() and
adds the new helper. After that, two libbpf patches that selects XDP
program based on what kernel is running. Finally, selftests for the new
functionality is added.
Note that the libbpf "auto-selection" is based on kernel version, so
it is hard coded to the "-next" version (5.12). If you would like to
try this is out, you will need to change the libbpf patch locally!
Thanks to Maciej and Magnus for the internal review/comments!
Performance (rxdrop, zero-copy)
Baseline
Two cores: 21.3 Mpps
One core: 24.5 Mpps
Two cores is slower? It used to be faster all the time, didn't it?
quoted
Patched
Two cores, bpf_redirect_map: 21.7 Mpps + 2%
One core, bpf_redirect_map: 24.9 Mpps + 2%
Two cores, bpf_redirect_xsk: 24.0 Mpps +13%
Nice, impressive improvement!
I do appreciate you work and performance optimizations at this level,
because when we are using this few CPU cycles per packet, then it is
really hard to find new ways to reduce cycles further.
Thank for you saying +13% instead of saying +2.7 Mpps.
It *is* impressive to basically reduce cycles with 13%.
21.3 Mpps = 46.94 nanosec per packet
24.0 Mpps = 41.66 nanosec per packet
21.3 Mpps -> 24.0 Mpps = 5.28 nanosec saved
On my 3.60GHz testlab machine that gives me 19 cycles.
quoted
quoted
One core, bpf_redirect_xsk: 25.5 Mpps + 4%
24.5 Mpps -> 25.5 Mpps = 1.6 nanosec saved
At this point with saving 1.6 ns this is around the cost of a function call 1.3 ns.
We still need these optimization in the kernel, but end-users in
userspace are very quickly going to waste the 19 cycles we found.
I still support/believe that the OS need to have a little overhead as
possible, but for me 42 nanosec overhead is close to zero overhead. For
comparison, I just ran a udp_sink[3] test, and it "cost" 625 ns for
delivery of UDP packets into socket (almost 15 times slower).
I guess my point is that with XDP we have already achieved and exceeded
(my original) performance goals, making it even faster is just showing off ;-P
Even though I'll let Bjorn elaborating on this, we're talking here about
AF-XDP which is a bit different pair of shoes to me in terms of
performance. AFAIK we still have a gap when compared to DPDK's numbers. So
AFAIK the gap on this i40e hardware is DPDK=36Mpps, and lets say
AF_XDP=25.5 Mpps, that looks large +10.5Mpps, but it is only 11.4
nanosec.
quoted
I'm really not sure why better performance bothers you? :)
Finding those 11.4 nanosec is going to be really difficult and take a
huge effort. My fear is also that some of these optimizations will make
the code harder to maintain and understand.
If you are ready to do this huge effort, then I'm actually ready to
provide ideas on what you can optimize.
E.g. you can get 2-4 nanosec if we prefetch to L2 in driver (assuming
data is DDIO in L3 already). (CPU supports L2 prefetch, but kernel
have no users of that feature).
In my experience, I would caution starting to use L2 prefetch at this
point since it is somewhat finicky. It is possible to get better
performance using it on one platform, but significantly harder to get
it to work also on a three year older one, not to mention other
platforms and architectures.
Instead, I would start with optimizations that always work such as not
having to execute code (or removing code) that is not necessary for
the functionality that is provided. One example of this is having to
execute all the bpf_redirect code when we are using packet steering in
HW to AF_XDP sockets and other potential consumers. In this scenario,
the core bpf_redirect code (xdp_do_redirect + bpf_xdp_redirect_map +
__xsk_map_redirect for the XSKMAP case) provides no value and amounts
to 19% of the total execution time on my machine (application plus
kernel, and the XDP program itself is another 7% on top of the 19%).
By removing this, you could cut down a large chunk of those 11.4 ns we
would need to find in our optimization work. Just to be clear, there
are plenty of other scenarios where the bpf_redirect functionality
provides great value, so it is really good to have. I just think it is
weird that it has to be executed all the time even when it is not
needed. The big question is just how to achieve this in an
architecturally sound way. Any ideas?
Just note that there are plenty that could be done in other areas too.
For example, the Rx part of the driver is around 22% of the execution
time and I am sure we can optimize this part too and increase the
readability and maintainability of the code at the same time. But not
trim it down to 0% as some of this is actually necessary work in all
cases where traffic is received :-). xp_alloc is another big time
consumer of cycles, but for this one I do have a patch set that cuts
it down from 21% to 8% that I plan on submitting in February.
perf top data:
21.58% [ice] [k] ice_clean_rx_irq_zc
21.21% [kernel] [k] xp_alloc
13.93% [kernel] [k] __xsk_rcv_zc
8.53% [kernel] [k] xdp_do_redirect
8.15% [kernel] [k] xsk_rcv
6.66% [kernel] [k] bpf_xdp_redirect_map
5.08% bpf_prog_992d9ddc835e5629 [k] bpf_prog_992d9ddc835e5629
3.73% [ice] [k] ice_alloc_rx_bufs_zc
3.36% [kernel] [k] __xsk_map_redirect
2.62% xdpsock [.] main
0.96% [kernel] [k] rcu_read_unlock_strict
0.90% bpf_dispatcher_xdp [k] bpf_dispatcher_xdp
0.87% [kernel] [k] bpf_dispatcher_xdp_func
0.51% [kernel] [k] xsk_flush
0.38% [kernel] [k] lapic_next_deadline
0.13% [ice] [k] ice_napi_poll
The base design of XDP redirect API, that have a number of function
calls per packet, in itself becomes a limiting factor. Even the cost
of getting the per-CPU pointer to bpf_redirect_info becomes something
to look at.
quoted
Let's rather be more harsh on changes that actually decrease the
performance, not the other way around. And I suppose you were the one that
always was bringing up the 'death by a 1000 paper cuts' of XDP. So yeah,
I'm a bit confused with your statement.
Yes, we really need to protect the existing the performance.
I think I'm mostly demotivated by the 'death by a 1000 paper cuts'.
People with good intentions are adding features, and performance slowly
disappears. I've been spending weeks/months finding nanosec level
improvements, which I don't have time to "defend". Recently I realized
that the 2nd XDP prog on egress, even when no prog is attached, is
slowing down the performance of base redirect (I don't fully understand
why, maybe it is simply the I-cache increase, I didn't realize when
reviewing the code).
--
Best regards,
Jesper Dangaard Brouer
MSc.CS, Principal Kernel Engineer at Red Hat
LinkedIn: http://www.linkedin.com/in/brouer
From: Maciej Fijalkowski <maciej.fijalkowski@intel.com> Date: 2021-01-22 09:56:29
On Fri, Jan 22, 2021 at 09:59:53AM +0100, Magnus Karlsson wrote:
On Thu, Jan 21, 2021 at 6:07 PM Jesper Dangaard Brouer
[off-list ref] wrote:
quoted
On Wed, 20 Jan 2021 17:19:31 +0100
Maciej Fijalkowski [off-list ref] wrote:
quoted
On Wed, Jan 20, 2021 at 04:57:08PM +0100, Jesper Dangaard Brouer wrote:
quoted
On Wed, 20 Jan 2021 15:15:22 +0200
Maxim Mikityanskiy [off-list ref] wrote:
quoted
On 2021-01-19 17:50, Björn Töpel wrote:
quoted
This series extends bind() for XDP sockets, so that the bound socket
is added to the netdev_rx_queue _rx array in the netdevice. We call
this to register the socket. To redirect packets to the registered
socket, a new BPF helper is used: bpf_redirect_xsk().
For shared XDP sockets, only the first bound socket is
registered. Users that need more complex setup has to use XSKMAP and
bpf_redirect_map().
Now, why would one use bpf_redirect_xsk() over the regular
bpf_redirect_map() helper?
* Better performance!
* Convenience; Most user use one socket per queue. This scenario is
what registered sockets support. There is no need to create an
XSKMAP. This can also reduce complexity from containerized setups,
where users might what to use XDP sockets without CAP_SYS_ADMIN
capabilities.
I'm buying into the convenience and reduce complexity, and XDP sockets
without CAP_SYS_ADMIN into containers.
People might be surprised that I'm actually NOT buying into the better
performance argument here. At these speeds we are basically comparing
how close we are to zero (and have to use nanosec time scale for our
comparisons), more below.
quoted
quoted
The first patch restructures xdp_do_redirect() a bit, to make it
easier to add the new helper. This restructure also give us a slight
performance benefit. The following three patches extends bind() and
adds the new helper. After that, two libbpf patches that selects XDP
program based on what kernel is running. Finally, selftests for the new
functionality is added.
Note that the libbpf "auto-selection" is based on kernel version, so
it is hard coded to the "-next" version (5.12). If you would like to
try this is out, you will need to change the libbpf patch locally!
Thanks to Maciej and Magnus for the internal review/comments!
Performance (rxdrop, zero-copy)
Baseline
Two cores: 21.3 Mpps
One core: 24.5 Mpps
Two cores is slower? It used to be faster all the time, didn't it?
quoted
Patched
Two cores, bpf_redirect_map: 21.7 Mpps + 2%
One core, bpf_redirect_map: 24.9 Mpps + 2%
Two cores, bpf_redirect_xsk: 24.0 Mpps +13%
Nice, impressive improvement!
I do appreciate you work and performance optimizations at this level,
because when we are using this few CPU cycles per packet, then it is
really hard to find new ways to reduce cycles further.
Thank for you saying +13% instead of saying +2.7 Mpps.
It *is* impressive to basically reduce cycles with 13%.
21.3 Mpps = 46.94 nanosec per packet
24.0 Mpps = 41.66 nanosec per packet
21.3 Mpps -> 24.0 Mpps = 5.28 nanosec saved
On my 3.60GHz testlab machine that gives me 19 cycles.
quoted
quoted
One core, bpf_redirect_xsk: 25.5 Mpps + 4%
24.5 Mpps -> 25.5 Mpps = 1.6 nanosec saved
At this point with saving 1.6 ns this is around the cost of a function call 1.3 ns.
We still need these optimization in the kernel, but end-users in
userspace are very quickly going to waste the 19 cycles we found.
I still support/believe that the OS need to have a little overhead as
possible, but for me 42 nanosec overhead is close to zero overhead. For
comparison, I just ran a udp_sink[3] test, and it "cost" 625 ns for
delivery of UDP packets into socket (almost 15 times slower).
I guess my point is that with XDP we have already achieved and exceeded
(my original) performance goals, making it even faster is just showing off ;-P
Even though I'll let Bjorn elaborating on this, we're talking here about
AF-XDP which is a bit different pair of shoes to me in terms of
performance. AFAIK we still have a gap when compared to DPDK's numbers. So
AFAIK the gap on this i40e hardware is DPDK=36Mpps, and lets say
AF_XDP=25.5 Mpps, that looks large +10.5Mpps, but it is only 11.4
nanosec.
quoted
I'm really not sure why better performance bothers you? :)
Finding those 11.4 nanosec is going to be really difficult and take a
huge effort. My fear is also that some of these optimizations will make
the code harder to maintain and understand.
If you are ready to do this huge effort, then I'm actually ready to
provide ideas on what you can optimize.
E.g. you can get 2-4 nanosec if we prefetch to L2 in driver (assuming
data is DDIO in L3 already). (CPU supports L2 prefetch, but kernel
have no users of that feature).
Just to add to what Magnus wrote below, when comparing technologies, XDP
is confronted with kernel's netstack and as we know, difference is really
big and there's no secret who's the winner in terms of performance in
specific scenarios. So, I'm not saying I'm super OK with that, but we can
allow ourselves for having a cool feature that boosts our control plane,
but degrades performance a bit, as you found out with egress XDP prog. We
are still going to be fine.
For AF_XDP the case is the opposite, meaning, we are the chaser in this
scenario. Performance work is still the top priority for AF_XDP in my
eyes.
In my experience, I would caution starting to use L2 prefetch at this
point since it is somewhat finicky. It is possible to get better
performance using it on one platform, but significantly harder to get
it to work also on a three year older one, not to mention other
platforms and architectures.
Instead, I would start with optimizations that always work such as not
having to execute code (or removing code) that is not necessary for
the functionality that is provided. One example of this is having to
execute all the bpf_redirect code when we are using packet steering in
HW to AF_XDP sockets and other potential consumers. In this scenario,
the core bpf_redirect code (xdp_do_redirect + bpf_xdp_redirect_map +
__xsk_map_redirect for the XSKMAP case) provides no value and amounts
to 19% of the total execution time on my machine (application plus
kernel, and the XDP program itself is another 7% on top of the 19%).
By removing this, you could cut down a large chunk of those 11.4 ns we
would need to find in our optimization work. Just to be clear, there
are plenty of other scenarios where the bpf_redirect functionality
provides great value, so it is really good to have. I just think it is
weird that it has to be executed all the time even when it is not
needed. The big question is just how to achieve this in an
architecturally sound way. Any ideas?
Just note that there are plenty that could be done in other areas too.
For example, the Rx part of the driver is around 22% of the execution
time and I am sure we can optimize this part too and increase the
readability and maintainability of the code at the same time. But not
trim it down to 0% as some of this is actually necessary work in all
cases where traffic is received :-). xp_alloc is another big time
consumer of cycles, but for this one I do have a patch set that cuts
it down from 21% to 8% that I plan on submitting in February.
perf top data:
21.58% [ice] [k] ice_clean_rx_irq_zc
21.21% [kernel] [k] xp_alloc
13.93% [kernel] [k] __xsk_rcv_zc
8.53% [kernel] [k] xdp_do_redirect
8.15% [kernel] [k] xsk_rcv
6.66% [kernel] [k] bpf_xdp_redirect_map
5.08% bpf_prog_992d9ddc835e5629 [k] bpf_prog_992d9ddc835e5629
3.73% [ice] [k] ice_alloc_rx_bufs_zc
3.36% [kernel] [k] __xsk_map_redirect
2.62% xdpsock [.] main
0.96% [kernel] [k] rcu_read_unlock_strict
0.90% bpf_dispatcher_xdp [k] bpf_dispatcher_xdp
0.87% [kernel] [k] bpf_dispatcher_xdp_func
0.51% [kernel] [k] xsk_flush
0.38% [kernel] [k] lapic_next_deadline
0.13% [ice] [k] ice_napi_poll
quoted
The base design of XDP redirect API, that have a number of function
calls per packet, in itself becomes a limiting factor. Even the cost
of getting the per-CPU pointer to bpf_redirect_info becomes something
to look at.
quoted
Let's rather be more harsh on changes that actually decrease the
performance, not the other way around. And I suppose you were the one that
always was bringing up the 'death by a 1000 paper cuts' of XDP. So yeah,
I'm a bit confused with your statement.
Yes, we really need to protect the existing the performance.
I think I'm mostly demotivated by the 'death by a 1000 paper cuts'.
People with good intentions are adding features, and performance slowly
disappears. I've been spending weeks/months finding nanosec level
improvements, which I don't have time to "defend". Recently I realized
that the 2nd XDP prog on egress, even when no prog is attached, is
slowing down the performance of base redirect (I don't fully understand
why, maybe it is simply the I-cache increase, I didn't realize when
reviewing the code).
--
Best regards,
Jesper Dangaard Brouer
MSc.CS, Principal Kernel Engineer at Red Hat
LinkedIn: http://www.linkedin.com/in/brouer