f_bpf update
Subsystems (4)
Daniel Borkmann
maintainercover f_bpf update
Stephen,
Patch 1 can optionally be dropped when you do the header rebase yourself. I have included it here just in case.
Thanks a lot!
Daniel Borkmann (2):
misc: header rebase, add bpf.h
tc: f_bpf: add eBPF support
configure | 26 ++++ include/linux/bpf.h | 183 ++++++++++++++++++++++++ include/utils.h | 5 + tc/Makefile | 5 + tc/f_bpf.c | 36 +++-- tc/tc_bpf.c | 392 ++++++++++++++++++++++++++++++++++++++++++++++++++++ tc/tc_bpf.h | 49 ++++++- 7 files changed, 685 insertions(+), 11 deletions(-) create mode 100644 include/linux/bpf.h
Signature · 1 line
1.9.3
Daniel Borkmann
maintainerPATCH 1/2 misc: header rebase, add bpf.h
Include the bpf.h uapi header file.
--- include/linux/bpf.h | 183 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 183 insertions(+) create mode 100644 include/linux/bpf.h
include/linux/bpf.h new file +183 -0
@@ -0,0 +1,183 @@ +/* Copyright (c) 2011-2014 PLUMgrid, http://plumgrid.com +* +* This program is free software; you can redistribute it and/or +* modify it under the terms of version 2 of the GNU General Public +* License as published by the Free Software Foundation. +*/ +#ifndef __LINUX_BPF_H__ +#define __LINUX_BPF_H__ + +#include <linux/types.h> +#include <linux/bpf_common.h> + +/* Extended instruction set based on top of classic BPF */ + +/* instruction classes */ +#define BPF_ALU64 0x07/* alu mode in double word width */ + +/* ld/ldx fields */ +#define BPF_DW 0x18/* double word */ +#define BPF_XADD 0xc0/* exclusive add */ + +/* alu/jmp fields */ +#define BPF_MOV 0xb0/* mov reg to reg */ +#define BPF_ARSH 0xc0/* sign extending arithmetic shift right */ + +/* change endianness of a register */ +#define BPF_END 0xd0/* flags for endianness conversion: */ +#define BPF_TO_LE 0x00/* convert to little-endian */ +#define BPF_TO_BE 0x08/* convert to big-endian */ +#define BPF_FROM_LE BPF_TO_LE +#define BPF_FROM_BE BPF_TO_BE + +#define BPF_JNE 0x50/* jump != */ +#define BPF_JSGT 0x60/* SGT is signed '>', GT in x86 */ +#define BPF_JSGE 0x70/* SGE is signed '>=', GE in x86 */ +#define BPF_CALL 0x80/* function call */ +#define BPF_EXIT 0x90/* function return */ + +/* Register numbers */ +enum { +BPF_REG_0 =0 , +BPF_REG_1 , +BPF_REG_2 , +BPF_REG_3 , +BPF_REG_4 , +BPF_REG_5 , +BPF_REG_6 , +BPF_REG_7 , +BPF_REG_8 , +BPF_REG_9 , +BPF_REG_10 , + __MAX_BPF_REG, +}; + +/* BPF has 10 general purpose 64-bit registers and stack frame. */ +#define MAX_BPF_REG __MAX_BPF_REG + +struct bpf_insn { +__u8 code;/* opcode */ +__u8 dst_reg:4 ;/* dest register */ +__u8 src_reg:4 ;/* source register */ +__s16 off;/* signed offset */ +__s32 imm;/* signed immediate constant */ +}; + +/* BPF syscall commands */ +enum bpf_cmd { +/* create a map with given type and attributes +* fd = bpf(BPF_MAP_CREATE, union bpf_attr *, u32 size) +* returns fd or negative error +* map is deleted when fd is closed +*/ +BPF_MAP_CREATE , + +/* lookup key in a given map +* err = bpf(BPF_MAP_LOOKUP_ELEM, union bpf_attr *attr, u32 size) +* Using attr->map_fd, attr->key, attr->value +* returns zero and stores found elem into value +* or negative error +*/ +BPF_MAP_LOOKUP_ELEM , + +/* create or update key/value pair in a given map +* err = bpf(BPF_MAP_UPDATE_ELEM, union bpf_attr *attr, u32 size) +* Using attr->map_fd, attr->key, attr->value, attr->flags +* returns zero or negative error +*/ +BPF_MAP_UPDATE_ELEM , + +/* find and delete elem by key in a given map +* err = bpf(BPF_MAP_DELETE_ELEM, union bpf_attr *attr, u32 size) +* Using attr->map_fd, attr->key +* returns zero or negative error +*/ +BPF_MAP_DELETE_ELEM , + +/* lookup key in a given map and return next key +* err = bpf(BPF_MAP_GET_NEXT_KEY, union bpf_attr *attr, u32 size) +* Using attr->map_fd, attr->key, attr->next_key +* returns zero and stores next key or negative error +*/ +BPF_MAP_GET_NEXT_KEY , + +/* verify and load eBPF program +* prog_fd = bpf(BPF_PROG_LOAD, union bpf_attr *attr, u32 size) +* Using attr->prog_type, attr->insns, attr->license +* returns fd or negative error +*/ +BPF_PROG_LOAD , +}; + +enum bpf_map_type { +BPF_MAP_TYPE_UNSPEC , +BPF_MAP_TYPE_HASH , +BPF_MAP_TYPE_ARRAY , +}; + +enum bpf_prog_type { +BPF_PROG_TYPE_UNSPEC , +BPF_PROG_TYPE_SOCKET_FILTER , +BPF_PROG_TYPE_SCHED_CLS , +}; + +#define BPF_PSEUDO_MAP_FD 1 + +/* flags for BPF_MAP_UPDATE_ELEM command */ +#define BPF_ANY 0/* create new element or update existing */ +#define BPF_NOEXIST 1/* create new element if it didn't exist */ +#define BPF_EXIST 2/* update existing element */ + +union bpf_attr { +struct {/* anonymous struct used by BPF_MAP_CREATE command */ +__u32 map_type;/* one of enum bpf_map_type */ +__u32 key_size;/* size of key in bytes */ +__u32 value_size;/* size of value in bytes */ +__u32 max_entries;/* max number of entries in a map */ + }; + +struct {/* anonymous struct used by BPF_MAP_*_ELEM commands */ +__u32 map_fd; +__aligned_u64 key; +union { +__aligned_u64 value; +__aligned_u64 next_key; + }; +__u64 flags; + }; + +struct {/* anonymous struct used by BPF_PROG_LOAD command */ +__u32 prog_type;/* one of enum bpf_prog_type */ +__u32 insn_cnt; +__aligned_u64 insns; +__aligned_u64 license; +__u32 log_level;/* verbosity level of verifier */ +__u32 log_size;/* size of user buffer */ +__aligned_u64 log_buf;/* user supplied buffer */ + }; +} __attribute__((aligned (8 ))); + +/* integer value in 'imm' field of BPF_CALL instruction selects which helper +* function eBPF program intends to call +*/ +enum bpf_func_id { + BPF_FUNC_unspec, + BPF_FUNC_map_lookup_elem,/* void *map_lookup_elem(&map, &key) */ + BPF_FUNC_map_update_elem,/* int map_update_elem(&map, &key, &value, flags) */ + BPF_FUNC_map_delete_elem,/* int map_delete_elem(&map, &key) */ + BPF_FUNC_get_prandom_u32,/* u32 prandom_u32(void) */ + BPF_FUNC_get_smp_processor_id,/* u32 raw_smp_processor_id(void) */ + __BPF_FUNC_MAX_ID, +}; + +/* user accessible mirror of in-kernel sk_buff. +* new fields can only be added to the end of this structure +*/ +struct __sk_buff { +__u32 len; +__u32 pkt_type; +__u32 mark; +__u32 queue_mapping; +}; + +#endif /* __LINUX_BPF_H__ */
Signature · 1 line
1.9.3
Daniel Borkmann
maintainerPATCH 2/2 tc: add eBPF support to f_bpf
In reply to Daniel BorkmannThis work adds the tc frontend for kernel commit e2e9b6541dd4 ("cls_bpf:
add initial eBPF support for programmable classifiers").
A C-like classifier program (f.e. see e2e9b6541dd4) is being compiled via LLVM's eBPF backend into an ELF file, that is then being passed to tc. tc then loads, if any, eBPF maps and eBPF opcodes (with fixed-up eBPF map file descriptors) out of its dedicated sections, and via bpf(2) into the kernel and then the resulting fd via netlink down to cls_bpf. cls_bpf allows for annotations, currently, I've used the file name for that, so that the user can easily identify his filter when dumping configurations back.
Example usage:
clang -O2 -emit-llvm -c cls.c -o - | llc -march=bpf -filetype=obj -o cls.o tc filter add dev em1 parent 1: bpf run object-file cls.o classid x:y
tc filter show dev em1 [...]
filter parent 1: protocol all pref 49152 bpf handle 0x1 flowid x:y cls.o
I placed the parser bits derived from Alexei's kernel sample, into tc_bpf.c as my next step is to also add the same support for BPF action, so we can have a fully fledged eBPF classifier and action in tc.
--- configure | 26 ++++ include/utils.h | 5 + tc/Makefile | 5 + tc/f_bpf.c | 36 ++++-- tc/tc_bpf.c | 392 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++ tc/tc_bpf.h | 49 ++++++- 6 files changed, 502 insertions(+), 11 deletions(-)
configure +26 -0
@@ -266,6 +266,29 @@ EOF rm -f $TMPDIR/ipsettest.c $TMPDIR/ipsettest } +check_elf() +{ + cat >$TMPDIR/elftest.c <<EOF +#include <libelf.h> +#include <gelf.h> +int main(void) +{ + Elf_Scn *scn; + GElf_Shdr shdr; + return elf_version(EV_CURRENT); +} +EOF + + if $CC -I$INCLUDE -o $TMPDIR/elftest $TMPDIR/elftest.c -lelf >/dev/null 2>&1 + then + echo "TC_CONFIG_ELF:=y" >>Config + echo "yes" + else + echo "no" + fi + rm -f $TMPDIR/elftest.c $TMPDIR/elftest +} + check_selinux() # SELinux is a compile time option in the ss utility {
@@ -306,5 +329,8 @@ check_netnsid echo -n "SELinux support: " check_selinux +echo -n "ELF support: " +check_elf + echo -e "\nDocs" check_docs
include/utils.h +5 -0
@@ -157,6 +157,11 @@ void print_nlmsg_timestamp(FILE *fp, const struct nlmsghdr *n);#define ARRAY_SIZE (x) (sizeof(x) / sizeof((x)[0])) +#ifndef __check_format_string +# define __check_format_string (pos_str, pos_args) \ + __attribute__ ((format (printf, (pos_str), (pos_args)))) +#endif +extern int cmdlineno;extern ssize_t getcmdline (char **line,size_t *len,FILE *in);extern int makeargs (char *line,char *argv[],int maxargs);
tc/Makefile +5 -0
@@ -89,6 +89,11 @@ else endif endif +ifeq ($(TC_CONFIG_ELF),y) + CFLAGS += -DHAVE_ELF + LDLIBS += -lelf +endif + TCOBJ += $(TCMODULES) LDLIBS += -L. -ltc -lm
tc/f_bpf.c +26 -10
@@ -34,13 +34,15 @@ static void explain(void)fprintf (stderr,"\n" );fprintf (stderr," [inline]: run bytecode BPF_BYTECODE\n" );fprintf (stderr," [from file]: run bytecode-file FILE\n" ); +fprintf (stderr," [from file]: run object-file FILE\n" );fprintf (stderr,"\n" );fprintf (stderr," [ action ACTION_SPEC ]\n" );fprintf (stderr," [ classid CLASSID ]\n" );fprintf (stderr,"\n" );fprintf (stderr,"Where BPF_BYTECODE := \'s,c t f k,c t f k,c t f k,...\'\n" );fprintf (stderr," c,t,f,k and s are decimals; s denotes number of 4-tuples\n" ); -fprintf (stderr,"Where FILE points to a file containing the BPF_BYTECODE string\n" ); +fprintf (stderr,"Where FILE points to a file containing the BPF_BYTECODE string,\n" ); +fprintf (stderr,"or an ELF file containing eBPF map definitions and bytecode.\n" );fprintf (stderr,"\nACTION_SPEC := ... look at individual actions\n" );fprintf (stderr,"NOTE: CLASSID is parsed as hexadecimal input.\n" ); }
@@ -71,31 +73,40 @@ static int bpf_parse_opt(struct filter_util *qu, char *handle,while (argc >0 ) {if (matches (*argv,"run" ) ==0 ) { -bool from_file; +bool from_file = true, ebpf;struct sock_filter bpf_ops[BPF_MAXINSNS ]; -__u16 bpf_len;int ret;NEXT_ARG ();if (strcmp (*argv,"bytecode-file" ) ==0 ) { - from_file = true; + ebpf = false; }else if (strcmp (*argv,"bytecode" ) ==0 ) { from_file = false; + ebpf = false; + }else if (strcmp (*argv,"object-file" ) ==0 ) { + ebpf = true; }else {fprintf (stderr,"What is \"%s\"?\n" , *argv);explain ();return -1 ; }NEXT_ARG (); - ret =bpf_parse_ops (argc, argv, bpf_ops, from_file); + ret = ebpf ?bpf_open_object (*argv,BPF_PROG_TYPE_SCHED_CLS ) : +bpf_parse_ops (argc, argv, bpf_ops, from_file);if (ret <0 ) { -fprintf (stderr,"Illegal \"bytecode\"\n" ); +fprintf (stderr,"%s\n" , ebpf ? +"Could not load object" : +"Illegal \"bytecode\"" );return -1 ; } - bpf_len = ret; -addattr16 (n,MAX_MSG ,TCA_BPF_OPS_LEN , bpf_len); -addattr_l (n,MAX_MSG ,TCA_BPF_OPS , &bpf_ops, - bpf_len *sizeof (struct sock_filter )); +if (ebpf) { +addattr32 (n,MAX_MSG ,TCA_BPF_FD , ret); +addattrstrz (n,MAX_MSG ,TCA_BPF_NAME , *argv); + }else { +addattr16 (n,MAX_MSG ,TCA_BPF_OPS_LEN , ret); +addattr_l (n,MAX_MSG ,TCA_BPF_OPS , &bpf_ops, + ret *sizeof (struct sock_filter )); + } }else if (matches (*argv,"classid" ) ==0 ||strcmp (*argv,"flowid" ) ==0 ) {unsigned handle;
@@ -153,6 +164,11 @@ static int bpf_print_opt(struct filter_util *qu, FILE *f, sprint_tc_classid(rta_getattr_u32(tb[TCA_BPF_CLASSID]), b1)); } +if (tb[TCA_BPF_NAME ]) +fprintf (f,"%s " ,rta_getattr_str (tb[TCA_BPF_NAME ])); +else if (tb[TCA_BPF_FD ]) +fprintf (f,"pfd %u " ,rta_getattr_u32 (tb[TCA_BPF_FD ])); +if (tb[TCA_BPF_OPS ] && tb[TCA_BPF_OPS_LEN ])bpf_print_ops (f, tb[TCA_BPF_OPS ],rta_getattr_u16 (tb[TCA_BPF_OPS_LEN ]));
tc/tc_bpf.c +392 -0
411 lines not on this page · raw message
tc/tc_bpf.h +48 -1
62 lines not on this page · raw message
Signature · 1 line
1.9.3
Re: [PATCH iproute2 -next 2/2] tc: add eBPF support to f_bpf
Quoted 7 lines · Daniel Borkmann
tc/tc_bpf.h +7
-0+ +/* ELF section names, etc (ABI) */ +#define ELF_SECTION_LICENSE "license" +#define ELF_SECTION_MAPS "maps" +#define ELF_SECTION_CLASSIFIER "classifier" +#define ELF_SECTION_ACTION "action" +
tc/tc_bpf.h line 42
Quoted 10 lines
tc/tc_bpf.h +10
-0+#define ELF_MAX_MAPS 64 +#define ELF_MAX_LICENSE_LEN 128 + +/* ELF map definition (ABI) */ +struct bpf_elf_map { +__u32 type; +__u32 size_key; +__u32 size_value; +__u32 max_elem; +};
I think people might freak out that the above section names and the struct are a kernel ABI. It's obviously not. Would be good to say that this is a present convention between C program that describes tc classifier and tc elf reader and it can change in the future.
Daniel Borkmann
maintainerRe: [PATCH iproute2 -next 2/2] tc: add eBPF support to f_bpf
Quoted 17 lines · Alexei Starovoitov
Quoted 15 lines · Daniel Borkmann
tc/tc_bpf.h +15
-0+ +/* ELF section names, etc (ABI) */ +#define ELF_SECTION_LICENSE "license" +#define ELF_SECTION_MAPS "maps" +#define ELF_SECTION_CLASSIFIER "classifier" +#define ELF_SECTION_ACTION "action" + +#define ELF_MAX_MAPS 64 +#define ELF_MAX_LICENSE_LEN 128 + +/* ELF map definition (ABI) */ +struct bpf_elf_map { +__u32 type; +__u32 size_key; +__u32 size_value;
Quoted 8 lines
tc/tc_bpf.h line 42
Quoted 2 lines
tc/tc_bpf.h +2
-0+__u32 max_elem; +};I think people might freak out that the above section names and the struct are a kernel ABI. It's obviously not. Would be good to say that this is a present convention between C program that describes tc classifier and tc elf reader and it can change in the future.
Ohh, well, it's not a kernel ABI. I actually don't intend to change the section names either as we don't want to break tc users. I could imagine aliases if truly necessary.
I guess, I'll elaborate on the comment, sure.