iv/linux
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
51ae7fa62d
linux/tools/bpf/bpf_exp.y

669 lines
16 KiB
Plaintext
Raw Normal View History

filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
/*
* BPF asm code parser
*
* This program is free software; you can distribute it and/or modify
* it under the terms of the GNU General Public License as published
* by the Free Software Foundation; either version 2 of the License,
* or (at your option) any later version.
*
* Syntax kept close to:
*
* Steven McCanne and Van Jacobson. 1993. The BSD packet filter: a new
* architecture for user-level packet capture. In Proceedings of the
* USENIX Winter 1993 Conference Proceedings on USENIX Winter 1993
* Conference Proceedings (USENIX'93). USENIX Association, Berkeley,
* CA, USA, 2-2.
*
* Copyright 2013 Daniel Borkmann <borkmann@redhat.com>
* Licensed under the GNU General Public License, version 2.0 (GPLv2)
*/
%{
#include <stdio.h>
#include <string.h>
#include <stdint.h>
#include <stdlib.h>
#include <stdbool.h>
#include <unistd.h>
#include <errno.h>
#include <assert.h>
#include <linux/filter.h>
#include "bpf_exp.yacc.h"
enum jmp_type { JTL, JFL, JKL };
extern FILE *yyin;
tools, bpf_asm: simplify parser rule for BPF extensions We can already use yylval in the lexer for encoding the BPF extension number, so that the parser rules can be further reduced to a single one for each B/H/W case. Signed-off-by: Ray Bellis <ray@isc.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-02-22 13:02:40 +03:00
extern int yylineno;
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
extern int yylex(void);
extern void yyerror(const char *str);
extern void bpf_asm_compile(FILE *fp, bool cstyle);
static void bpf_set_curr_instr(uint16_t op, uint8_t jt, uint8_t jf, uint32_t k);
bpf_exp: free duplicated labels at exit time Valgrind found that extracted labels that are passed from the lexer weren't freed upon exit. Therefore, add a small helper function that walks label tables and frees them. Since also NULL can be passed to free(3), we do not need to take care of that here. While at it, fix up a spacing error in bpf_set_curr_label(). Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-16 14:45:01 +04:00
static void bpf_set_curr_label(char *label);
static void bpf_set_jmp_label(char *label, enum jmp_type type);
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
%}
%union {
char *label;
uint32_t number;
}
%token OP_LDB OP_LDH OP_LD OP_LDX OP_ST OP_STX OP_JMP OP_JEQ OP_JGT OP_JGE
%token OP_JSET OP_ADD OP_SUB OP_MUL OP_DIV OP_AND OP_OR OP_XOR OP_LSH OP_RSH
%token OP_RET OP_TAX OP_TXA OP_LDXB OP_MOD OP_NEG OP_JNEQ OP_JLT OP_JLE OP_LDI
%token OP_LDXI
tools, bpf_asm: simplify parser rule for BPF extensions We can already use yylval in the lexer for encoding the BPF extension number, so that the parser rules can be further reduced to a single one for each B/H/W case. Signed-off-by: Ray Bellis <ray@isc.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-02-22 13:02:40 +03:00
%token K_PKT_LEN
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
%token ':' ',' '[' ']' '(' ')' 'x' 'a' '+' 'M' '*' '&' '#' '%'
tools, bpf_asm: simplify parser rule for BPF extensions We can already use yylval in the lexer for encoding the BPF extension number, so that the parser rules can be further reduced to a single one for each B/H/W case. Signed-off-by: Ray Bellis <ray@isc.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-02-22 13:02:40 +03:00
%token extension number label
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
%type <label> label
tools, bpf_asm: simplify parser rule for BPF extensions We can already use yylval in the lexer for encoding the BPF extension number, so that the parser rules can be further reduced to a single one for each B/H/W case. Signed-off-by: Ray Bellis <ray@isc.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-02-22 13:02:40 +03:00
%type <number> extension
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
%type <number> number
%%
prog
: line
| prog line
;
line
: instr
| labelled_instr
;
labelled_instr
: labelled instr
;
instr
: ldb
| ldh
| ld
| ldi
| ldx
| ldxi
| st
| stx
| jmp
| jeq
| jneq
| jlt
| jle
| jgt
| jge
| jset
| add
| sub
| mul
| div
| mod
| neg
| and
| or
| xor
| lsh
| rsh
| ret
| tax
| txa
;
labelled
: label ':' { bpf_set_curr_label($1); }
;
ldb
: OP_LDB '[' 'x' '+' number ']' {
bpf_set_curr_instr(BPF_LD | BPF_B | BPF_IND, 0, 0, $5); }
| OP_LDB '[' '%' 'x' '+' number ']' {
bpf_set_curr_instr(BPF_LD | BPF_B | BPF_IND, 0, 0, $6); }
| OP_LDB '[' number ']' {
bpf_set_curr_instr(BPF_LD | BPF_B | BPF_ABS, 0, 0, $3); }
tools, bpf_asm: simplify parser rule for BPF extensions We can already use yylval in the lexer for encoding the BPF extension number, so that the parser rules can be further reduced to a single one for each B/H/W case. Signed-off-by: Ray Bellis <ray@isc.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-02-22 13:02:40 +03:00
| OP_LDB extension {
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
bpf_set_curr_instr(BPF_LD | BPF_B | BPF_ABS, 0, 0,
tools, bpf_asm: simplify parser rule for BPF extensions We can already use yylval in the lexer for encoding the BPF extension number, so that the parser rules can be further reduced to a single one for each B/H/W case. Signed-off-by: Ray Bellis <ray@isc.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-02-22 13:02:40 +03:00
SKF_AD_OFF + $2); }
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
;
ldh
: OP_LDH '[' 'x' '+' number ']' {
bpf_set_curr_instr(BPF_LD | BPF_H | BPF_IND, 0, 0, $5); }
| OP_LDH '[' '%' 'x' '+' number ']' {
bpf_set_curr_instr(BPF_LD | BPF_H | BPF_IND, 0, 0, $6); }
| OP_LDH '[' number ']' {
bpf_set_curr_instr(BPF_LD | BPF_H | BPF_ABS, 0, 0, $3); }
tools, bpf_asm: simplify parser rule for BPF extensions We can already use yylval in the lexer for encoding the BPF extension number, so that the parser rules can be further reduced to a single one for each B/H/W case. Signed-off-by: Ray Bellis <ray@isc.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-02-22 13:02:40 +03:00
| OP_LDH extension {
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
bpf_set_curr_instr(BPF_LD | BPF_H | BPF_ABS, 0, 0,
tools, bpf_asm: simplify parser rule for BPF extensions We can already use yylval in the lexer for encoding the BPF extension number, so that the parser rules can be further reduced to a single one for each B/H/W case. Signed-off-by: Ray Bellis <ray@isc.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-02-22 13:02:40 +03:00
SKF_AD_OFF + $2); }
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
;
ldi
: OP_LDI '#' number {
bpf_set_curr_instr(BPF_LD | BPF_IMM, 0, 0, $3); }
| OP_LDI number {
bpf_set_curr_instr(BPF_LD | BPF_IMM, 0, 0, $2); }
;
ld
: OP_LD '#' number {
bpf_set_curr_instr(BPF_LD | BPF_IMM, 0, 0, $3); }
| OP_LD K_PKT_LEN {
bpf_set_curr_instr(BPF_LD | BPF_W | BPF_LEN, 0, 0, 0); }
tools, bpf_asm: simplify parser rule for BPF extensions We can already use yylval in the lexer for encoding the BPF extension number, so that the parser rules can be further reduced to a single one for each B/H/W case. Signed-off-by: Ray Bellis <ray@isc.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-02-22 13:02:40 +03:00
| OP_LD extension {
filter: introduce SKF_AD_VLAN_TPID BPF extension If vlan offloading takes place then vlan header is removed from frame and its contents, both vlan_tci and vlan_proto, is available to user space via TPACKET interface. However, only vlan_tci can be used in BPF filters. This commit introduces a new BPF extension. It makes possible to load the value of vlan_proto (vlan TPID) to register A. Support for classic BPF and eBPF is being added, analogous to skb->protocol. Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Jiri Pirko <jpirko@redhat.com> Signed-off-by: Michal Sekletar <msekleta@redhat.com> Acked-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Alexei Starovoitov <ast@plumgrid.com> Reviewed-by: Jiri Pirko <jiri@resnulli.us> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-03-24 16:48:41 +03:00
bpf_set_curr_instr(BPF_LD | BPF_W | BPF_ABS, 0, 0,
tools, bpf_asm: simplify parser rule for BPF extensions We can already use yylval in the lexer for encoding the BPF extension number, so that the parser rules can be further reduced to a single one for each B/H/W case. Signed-off-by: Ray Bellis <ray@isc.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-02-22 13:02:40 +03:00
SKF_AD_OFF + $2); }
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
| OP_LD 'M' '[' number ']' {
bpf_set_curr_instr(BPF_LD | BPF_MEM, 0, 0, $4); }
| OP_LD '[' 'x' '+' number ']' {
bpf_set_curr_instr(BPF_LD | BPF_W | BPF_IND, 0, 0, $5); }
| OP_LD '[' '%' 'x' '+' number ']' {
bpf_set_curr_instr(BPF_LD | BPF_W | BPF_IND, 0, 0, $6); }
| OP_LD '[' number ']' {
bpf_set_curr_instr(BPF_LD | BPF_W | BPF_ABS, 0, 0, $3); }
;
ldxi
: OP_LDXI '#' number {
bpf_set_curr_instr(BPF_LDX | BPF_IMM, 0, 0, $3); }
| OP_LDXI number {
bpf_set_curr_instr(BPF_LDX | BPF_IMM, 0, 0, $2); }
;
ldx
: OP_LDX '#' number {
bpf_set_curr_instr(BPF_LDX | BPF_IMM, 0, 0, $3); }
| OP_LDX K_PKT_LEN {
bpf_set_curr_instr(BPF_LDX | BPF_W | BPF_LEN, 0, 0, 0); }
| OP_LDX 'M' '[' number ']' {
bpf_set_curr_instr(BPF_LDX | BPF_MEM, 0, 0, $4); }
| OP_LDXB number '*' '(' '[' number ']' '&' number ')' {
if ($2 != 4 || $9 != 0xf) {
fprintf(stderr, "ldxb offset not supported!\n");
tools, bpf_asm: Exit non-zero on errors ... so callers can correctly detect failure. Signed-off-by: Ian Denhardt <ian@zenhack.net> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/b0bea780bc292f29e7b389dd062f20adc2a2d634.1614201868.git.ian@zenhack.net
2021-02-24 05:24:00 +03:00
exit(1);
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
} else {
bpf_set_curr_instr(BPF_LDX | BPF_MSH | BPF_B, 0, 0, $6); } }
| OP_LDX number '*' '(' '[' number ']' '&' number ')' {
if ($2 != 4 || $9 != 0xf) {
fprintf(stderr, "ldxb offset not supported!\n");
tools, bpf_asm: Exit non-zero on errors ... so callers can correctly detect failure. Signed-off-by: Ian Denhardt <ian@zenhack.net> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/b0bea780bc292f29e7b389dd062f20adc2a2d634.1614201868.git.ian@zenhack.net
2021-02-24 05:24:00 +03:00
exit(1);
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
} else {
bpf_set_curr_instr(BPF_LDX | BPF_MSH | BPF_B, 0, 0, $6); } }
;
st
: OP_ST 'M' '[' number ']' {
bpf_set_curr_instr(BPF_ST, 0, 0, $4); }
;
stx
: OP_STX 'M' '[' number ']' {
bpf_set_curr_instr(BPF_STX, 0, 0, $4); }
;
jmp
: OP_JMP label {
bpf_set_jmp_label($2, JKL);
bpf_set_curr_instr(BPF_JMP | BPF_JA, 0, 0, 0); }
;
jeq
: OP_JEQ '#' number ',' label ',' label {
bpf_set_jmp_label($5, JTL);
bpf_set_jmp_label($7, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JEQ | BPF_K, 0, 0, $3); }
| OP_JEQ 'x' ',' label ',' label {
bpf_set_jmp_label($4, JTL);
bpf_set_jmp_label($6, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JEQ | BPF_X, 0, 0, 0); }
| OP_JEQ '%' 'x' ',' label ',' label {
bpf_set_jmp_label($5, JTL);
bpf_set_jmp_label($7, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JEQ | BPF_X, 0, 0, 0); }
| OP_JEQ '#' number ',' label {
bpf_set_jmp_label($5, JTL);
bpf_set_curr_instr(BPF_JMP | BPF_JEQ | BPF_K, 0, 0, $3); }
| OP_JEQ 'x' ',' label {
bpf_set_jmp_label($4, JTL);
bpf_set_curr_instr(BPF_JMP | BPF_JEQ | BPF_X, 0, 0, 0); }
| OP_JEQ '%' 'x' ',' label {
bpf_set_jmp_label($5, JTL);
bpf_set_curr_instr(BPF_JMP | BPF_JEQ | BPF_X, 0, 0, 0); }
;
jneq
: OP_JNEQ '#' number ',' label {
bpf_set_jmp_label($5, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JEQ | BPF_K, 0, 0, $3); }
| OP_JNEQ 'x' ',' label {
bpf_set_jmp_label($4, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JEQ | BPF_X, 0, 0, 0); }
| OP_JNEQ '%' 'x' ',' label {
bpf_set_jmp_label($5, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JEQ | BPF_X, 0, 0, 0); }
;
jlt
: OP_JLT '#' number ',' label {
bpf_set_jmp_label($5, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JGE | BPF_K, 0, 0, $3); }
| OP_JLT 'x' ',' label {
bpf_set_jmp_label($4, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JGE | BPF_X, 0, 0, 0); }
| OP_JLT '%' 'x' ',' label {
bpf_set_jmp_label($5, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JGE | BPF_X, 0, 0, 0); }
;
jle
: OP_JLE '#' number ',' label {
bpf_set_jmp_label($5, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JGT | BPF_K, 0, 0, $3); }
| OP_JLE 'x' ',' label {
bpf_set_jmp_label($4, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JGT | BPF_X, 0, 0, 0); }
| OP_JLE '%' 'x' ',' label {
bpf_set_jmp_label($5, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JGT | BPF_X, 0, 0, 0); }
;
jgt
: OP_JGT '#' number ',' label ',' label {
bpf_set_jmp_label($5, JTL);
bpf_set_jmp_label($7, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JGT | BPF_K, 0, 0, $3); }
| OP_JGT 'x' ',' label ',' label {
bpf_set_jmp_label($4, JTL);
bpf_set_jmp_label($6, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JGT | BPF_X, 0, 0, 0); }
| OP_JGT '%' 'x' ',' label ',' label {
bpf_set_jmp_label($5, JTL);
bpf_set_jmp_label($7, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JGT | BPF_X, 0, 0, 0); }
| OP_JGT '#' number ',' label {
bpf_set_jmp_label($5, JTL);
bpf_set_curr_instr(BPF_JMP | BPF_JGT | BPF_K, 0, 0, $3); }
| OP_JGT 'x' ',' label {
bpf_set_jmp_label($4, JTL);
bpf_set_curr_instr(BPF_JMP | BPF_JGT | BPF_X, 0, 0, 0); }
| OP_JGT '%' 'x' ',' label {
bpf_set_jmp_label($5, JTL);
bpf_set_curr_instr(BPF_JMP | BPF_JGT | BPF_X, 0, 0, 0); }
;
jge
: OP_JGE '#' number ',' label ',' label {
bpf_set_jmp_label($5, JTL);
bpf_set_jmp_label($7, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JGE | BPF_K, 0, 0, $3); }
| OP_JGE 'x' ',' label ',' label {
bpf_set_jmp_label($4, JTL);
bpf_set_jmp_label($6, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JGE | BPF_X, 0, 0, 0); }
| OP_JGE '%' 'x' ',' label ',' label {
bpf_set_jmp_label($5, JTL);
bpf_set_jmp_label($7, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JGE | BPF_X, 0, 0, 0); }
| OP_JGE '#' number ',' label {
bpf_set_jmp_label($5, JTL);
bpf_set_curr_instr(BPF_JMP | BPF_JGE | BPF_K, 0, 0, $3); }
| OP_JGE 'x' ',' label {
bpf_set_jmp_label($4, JTL);
bpf_set_curr_instr(BPF_JMP | BPF_JGE | BPF_X, 0, 0, 0); }
| OP_JGE '%' 'x' ',' label {
bpf_set_jmp_label($5, JTL);
bpf_set_curr_instr(BPF_JMP | BPF_JGE | BPF_X, 0, 0, 0); }
;
jset
: OP_JSET '#' number ',' label ',' label {
bpf_set_jmp_label($5, JTL);
bpf_set_jmp_label($7, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JSET | BPF_K, 0, 0, $3); }
| OP_JSET 'x' ',' label ',' label {
bpf_set_jmp_label($4, JTL);
bpf_set_jmp_label($6, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JSET | BPF_X, 0, 0, 0); }
| OP_JSET '%' 'x' ',' label ',' label {
bpf_set_jmp_label($5, JTL);
bpf_set_jmp_label($7, JFL);
bpf_set_curr_instr(BPF_JMP | BPF_JSET | BPF_X, 0, 0, 0); }
| OP_JSET '#' number ',' label {
bpf_set_jmp_label($5, JTL);
bpf_set_curr_instr(BPF_JMP | BPF_JSET | BPF_K, 0, 0, $3); }
| OP_JSET 'x' ',' label {
bpf_set_jmp_label($4, JTL);
bpf_set_curr_instr(BPF_JMP | BPF_JSET | BPF_X, 0, 0, 0); }
| OP_JSET '%' 'x' ',' label {
bpf_set_jmp_label($5, JTL);
bpf_set_curr_instr(BPF_JMP | BPF_JSET | BPF_X, 0, 0, 0); }
;
add
: OP_ADD '#' number {
bpf_set_curr_instr(BPF_ALU | BPF_ADD | BPF_K, 0, 0, $3); }
| OP_ADD 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_ADD | BPF_X, 0, 0, 0); }
| OP_ADD '%' 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_ADD | BPF_X, 0, 0, 0); }
;
sub
: OP_SUB '#' number {
bpf_set_curr_instr(BPF_ALU | BPF_SUB | BPF_K, 0, 0, $3); }
| OP_SUB 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_SUB | BPF_X, 0, 0, 0); }
| OP_SUB '%' 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_SUB | BPF_X, 0, 0, 0); }
;
mul
: OP_MUL '#' number {
bpf_set_curr_instr(BPF_ALU | BPF_MUL | BPF_K, 0, 0, $3); }
| OP_MUL 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_MUL | BPF_X, 0, 0, 0); }
| OP_MUL '%' 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_MUL | BPF_X, 0, 0, 0); }
;
div
: OP_DIV '#' number {
bpf_set_curr_instr(BPF_ALU | BPF_DIV | BPF_K, 0, 0, $3); }
| OP_DIV 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_DIV | BPF_X, 0, 0, 0); }
| OP_DIV '%' 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_DIV | BPF_X, 0, 0, 0); }
;
mod
: OP_MOD '#' number {
bpf_set_curr_instr(BPF_ALU | BPF_MOD | BPF_K, 0, 0, $3); }
| OP_MOD 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_MOD | BPF_X, 0, 0, 0); }
| OP_MOD '%' 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_MOD | BPF_X, 0, 0, 0); }
;
neg
: OP_NEG {
bpf_set_curr_instr(BPF_ALU | BPF_NEG, 0, 0, 0); }
;
and
: OP_AND '#' number {
bpf_set_curr_instr(BPF_ALU | BPF_AND | BPF_K, 0, 0, $3); }
| OP_AND 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_AND | BPF_X, 0, 0, 0); }
| OP_AND '%' 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_AND | BPF_X, 0, 0, 0); }
;
or
: OP_OR '#' number {
bpf_set_curr_instr(BPF_ALU | BPF_OR | BPF_K, 0, 0, $3); }
| OP_OR 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_OR | BPF_X, 0, 0, 0); }
| OP_OR '%' 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_OR | BPF_X, 0, 0, 0); }
;
xor
: OP_XOR '#' number {
bpf_set_curr_instr(BPF_ALU | BPF_XOR | BPF_K, 0, 0, $3); }
| OP_XOR 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_XOR | BPF_X, 0, 0, 0); }
| OP_XOR '%' 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_XOR | BPF_X, 0, 0, 0); }
;
lsh
: OP_LSH '#' number {
bpf_set_curr_instr(BPF_ALU | BPF_LSH | BPF_K, 0, 0, $3); }
| OP_LSH 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_LSH | BPF_X, 0, 0, 0); }
| OP_LSH '%' 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_LSH | BPF_X, 0, 0, 0); }
;
rsh
: OP_RSH '#' number {
bpf_set_curr_instr(BPF_ALU | BPF_RSH | BPF_K, 0, 0, $3); }
| OP_RSH 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_RSH | BPF_X, 0, 0, 0); }
| OP_RSH '%' 'x' {
bpf_set_curr_instr(BPF_ALU | BPF_RSH | BPF_X, 0, 0, 0); }
;
ret
: OP_RET 'a' {
bpf_set_curr_instr(BPF_RET | BPF_A, 0, 0, 0); }
| OP_RET '%' 'a' {
bpf_set_curr_instr(BPF_RET | BPF_A, 0, 0, 0); }
| OP_RET 'x' {
bpf_set_curr_instr(BPF_RET | BPF_X, 0, 0, 0); }
| OP_RET '%' 'x' {
bpf_set_curr_instr(BPF_RET | BPF_X, 0, 0, 0); }
| OP_RET '#' number {
bpf_set_curr_instr(BPF_RET | BPF_K, 0, 0, $3); }
;
tax
: OP_TAX {
bpf_set_curr_instr(BPF_MISC | BPF_TAX, 0, 0, 0); }
;
txa
: OP_TXA {
bpf_set_curr_instr(BPF_MISC | BPF_TXA, 0, 0, 0); }
;
%%
static int curr_instr = 0;
static struct sock_filter out[BPF_MAXINSNS];
bpf_exp: free duplicated labels at exit time Valgrind found that extracted labels that are passed from the lexer weren't freed upon exit. Therefore, add a small helper function that walks label tables and frees them. Since also NULL can be passed to free(3), we do not need to take care of that here. While at it, fix up a spacing error in bpf_set_curr_label(). Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-16 14:45:01 +04:00
static char **labels, **labels_jt, **labels_jf, **labels_k;
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
static void bpf_assert_max(void)
{
if (curr_instr >= BPF_MAXINSNS) {
fprintf(stderr, "only max %u insns allowed!\n", BPF_MAXINSNS);
tools, bpf_asm: Exit non-zero on errors ... so callers can correctly detect failure. Signed-off-by: Ian Denhardt <ian@zenhack.net> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/b0bea780bc292f29e7b389dd062f20adc2a2d634.1614201868.git.ian@zenhack.net
2021-02-24 05:24:00 +03:00
exit(1);
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
}
}
static void bpf_set_curr_instr(uint16_t code, uint8_t jt, uint8_t jf,
uint32_t k)
{
bpf_assert_max();
out[curr_instr].code = code;
out[curr_instr].jt = jt;
out[curr_instr].jf = jf;
out[curr_instr].k = k;
curr_instr++;
}
bpf_exp: free duplicated labels at exit time Valgrind found that extracted labels that are passed from the lexer weren't freed upon exit. Therefore, add a small helper function that walks label tables and frees them. Since also NULL can be passed to free(3), we do not need to take care of that here. While at it, fix up a spacing error in bpf_set_curr_label(). Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-16 14:45:01 +04:00
static void bpf_set_curr_label(char *label)
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
{
bpf_assert_max();
bpf_exp: free duplicated labels at exit time Valgrind found that extracted labels that are passed from the lexer weren't freed upon exit. Therefore, add a small helper function that walks label tables and frees them. Since also NULL can be passed to free(3), we do not need to take care of that here. While at it, fix up a spacing error in bpf_set_curr_label(). Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-16 14:45:01 +04:00
labels[curr_instr] = label;
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
}
bpf_exp: free duplicated labels at exit time Valgrind found that extracted labels that are passed from the lexer weren't freed upon exit. Therefore, add a small helper function that walks label tables and frees them. Since also NULL can be passed to free(3), we do not need to take care of that here. While at it, fix up a spacing error in bpf_set_curr_label(). Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-16 14:45:01 +04:00
static void bpf_set_jmp_label(char *label, enum jmp_type type)
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
{
bpf_assert_max();
switch (type) {
case JTL:
labels_jt[curr_instr] = label;
break;
case JFL:
labels_jf[curr_instr] = label;
break;
case JKL:
labels_k[curr_instr] = label;
break;
}
}
static int bpf_find_insns_offset(const char *label)
{
int i, max = curr_instr, ret = -ENOENT;
for (i = 0; i < max; i++) {
if (labels[i] && !strcmp(label, labels[i])) {
ret = i;
break;
}
}
if (ret == -ENOENT) {
fprintf(stderr, "no such label \'%s\'!\n", label);
tools, bpf_asm: Exit non-zero on errors ... so callers can correctly detect failure. Signed-off-by: Ian Denhardt <ian@zenhack.net> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/b0bea780bc292f29e7b389dd062f20adc2a2d634.1614201868.git.ian@zenhack.net
2021-02-24 05:24:00 +03:00
exit(1);
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
}
return ret;
}
static void bpf_stage_1_insert_insns(void)
{
yyparse();
}
static void bpf_reduce_k_jumps(void)
{
int i;
for (i = 0; i < curr_instr; i++) {
if (labels_k[i]) {
int off = bpf_find_insns_offset(labels_k[i]);
out[i].k = (uint32_t) (off - i - 1);
}
}
}
tools, bpf_asm: Warn when jumps are out of range When compiling larger programs with bpf_asm, it's possible to accidentally exceed jt/jf range, in which case it won't complain, but rather silently emit a truncated offset, leading to a "happy debugging" situation. Add a warning to help detecting such issues. It could be made an error instead, but this might break compilation of existing code (which might be working by accident). Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/20191107100349.88976-1-iii@linux.ibm.com
2019-11-07 13:03:49 +03:00
static uint8_t bpf_encode_jt_jf_offset(int off, int i)
{
int delta = off - i - 1;
tools, bpf_asm: Hard error on out of range jumps Per discussion at [0] this was originally introduced as a warning due to concerns about breaking existing code, but a hard error probably makes more sense, especially given that concerns about breakage were only speculation. [0] https://lore.kernel.org/bpf/c964892195a6b91d20a67691448567ef528ffa6d.camel@linux.ibm.com/T/#t Signed-off-by: Ian Denhardt <ian@zenhack.net> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Link: https://lore.kernel.org/bpf/a6b6c7516f5d559049d669968e953b4a8d7adea3.1614201868.git.ian@zenhack.net
2021-02-24 05:15:31 +03:00
if (delta < 0 || delta > 255) {
fprintf(stderr, "error: insn #%d jumps to insn #%d, "
tools, bpf_asm: Warn when jumps are out of range When compiling larger programs with bpf_asm, it's possible to accidentally exceed jt/jf range, in which case it won't complain, but rather silently emit a truncated offset, leading to a "happy debugging" situation. Add a warning to help detecting such issues. It could be made an error instead, but this might break compilation of existing code (which might be working by accident). Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/20191107100349.88976-1-iii@linux.ibm.com
2019-11-07 13:03:49 +03:00
"which is out of range\n", i, off);
tools, bpf_asm: Hard error on out of range jumps Per discussion at [0] this was originally introduced as a warning due to concerns about breaking existing code, but a hard error probably makes more sense, especially given that concerns about breakage were only speculation. [0] https://lore.kernel.org/bpf/c964892195a6b91d20a67691448567ef528ffa6d.camel@linux.ibm.com/T/#t Signed-off-by: Ian Denhardt <ian@zenhack.net> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Link: https://lore.kernel.org/bpf/a6b6c7516f5d559049d669968e953b4a8d7adea3.1614201868.git.ian@zenhack.net
2021-02-24 05:15:31 +03:00
exit(1);
}
tools, bpf_asm: Warn when jumps are out of range When compiling larger programs with bpf_asm, it's possible to accidentally exceed jt/jf range, in which case it won't complain, but rather silently emit a truncated offset, leading to a "happy debugging" situation. Add a warning to help detecting such issues. It could be made an error instead, but this might break compilation of existing code (which might be working by accident). Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/20191107100349.88976-1-iii@linux.ibm.com
2019-11-07 13:03:49 +03:00
return (uint8_t) delta;
}
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
static void bpf_reduce_jt_jumps(void)
{
int i;
for (i = 0; i < curr_instr; i++) {
if (labels_jt[i]) {
int off = bpf_find_insns_offset(labels_jt[i]);
tools, bpf_asm: Warn when jumps are out of range When compiling larger programs with bpf_asm, it's possible to accidentally exceed jt/jf range, in which case it won't complain, but rather silently emit a truncated offset, leading to a "happy debugging" situation. Add a warning to help detecting such issues. It could be made an error instead, but this might break compilation of existing code (which might be working by accident). Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/20191107100349.88976-1-iii@linux.ibm.com
2019-11-07 13:03:49 +03:00
out[i].jt = bpf_encode_jt_jf_offset(off, i);
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
}
}
}
static void bpf_reduce_jf_jumps(void)
{
int i;
for (i = 0; i < curr_instr; i++) {
if (labels_jf[i]) {
int off = bpf_find_insns_offset(labels_jf[i]);
tools, bpf_asm: Warn when jumps are out of range When compiling larger programs with bpf_asm, it's possible to accidentally exceed jt/jf range, in which case it won't complain, but rather silently emit a truncated offset, leading to a "happy debugging" situation. Add a warning to help detecting such issues. It could be made an error instead, but this might break compilation of existing code (which might be working by accident). Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/20191107100349.88976-1-iii@linux.ibm.com
2019-11-07 13:03:49 +03:00
out[i].jf = bpf_encode_jt_jf_offset(off, i);
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
}
}
}
static void bpf_stage_2_reduce_labels(void)
{
bpf_reduce_k_jumps();
bpf_reduce_jt_jumps();
bpf_reduce_jf_jumps();
}
static void bpf_pretty_print_c(void)
{
int i;
for (i = 0; i < curr_instr; i++)
printf("{ %#04x, %2u, %2u, %#010x },\n", out[i].code,
out[i].jt, out[i].jf, out[i].k);
}
static void bpf_pretty_print(void)
{
int i;
printf("%u,", curr_instr);
for (i = 0; i < curr_instr; i++)
printf("%u %u %u %u,", out[i].code,
out[i].jt, out[i].jf, out[i].k);
printf("\n");
}
static void bpf_init(void)
{
memset(out, 0, sizeof(out));
labels = calloc(BPF_MAXINSNS, sizeof(*labels));
assert(labels);
labels_jt = calloc(BPF_MAXINSNS, sizeof(*labels_jt));
assert(labels_jt);
labels_jf = calloc(BPF_MAXINSNS, sizeof(*labels_jf));
assert(labels_jf);
labels_k = calloc(BPF_MAXINSNS, sizeof(*labels_k));
assert(labels_k);
}
bpf_exp: free duplicated labels at exit time Valgrind found that extracted labels that are passed from the lexer weren't freed upon exit. Therefore, add a small helper function that walks label tables and frees them. Since also NULL can be passed to free(3), we do not need to take care of that here. While at it, fix up a spacing error in bpf_set_curr_label(). Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-16 14:45:01 +04:00
static void bpf_destroy_labels(void)
{
int i;
for (i = 0; i < curr_instr; i++) {
free(labels_jf[i]);
free(labels_jt[i]);
free(labels_k[i]);
free(labels[i]);
}
}
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
static void bpf_destroy(void)
{
bpf_exp: free duplicated labels at exit time Valgrind found that extracted labels that are passed from the lexer weren't freed upon exit. Therefore, add a small helper function that walks label tables and frees them. Since also NULL can be passed to free(3), we do not need to take care of that here. While at it, fix up a spacing error in bpf_set_curr_label(). Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-16 14:45:01 +04:00
bpf_destroy_labels();
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
free(labels_jt);
free(labels_jf);
free(labels_k);
bpf_exp: free duplicated labels at exit time Valgrind found that extracted labels that are passed from the lexer weren't freed upon exit. Therefore, add a small helper function that walks label tables and frees them. Since also NULL can be passed to free(3), we do not need to take care of that here. While at it, fix up a spacing error in bpf_set_curr_label(). Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-16 14:45:01 +04:00
free(labels);
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
}
void bpf_asm_compile(FILE *fp, bool cstyle)
{
yyin = fp;
bpf_init();
bpf_stage_1_insert_insns();
bpf_stage_2_reduce_labels();
bpf_destroy();
if (cstyle)
bpf_pretty_print_c();
else
bpf_pretty_print();
if (fp != stdin)
fclose(yyin);
}
void yyerror(const char *str)
{
tools, bpf_asm: simplify parser rule for BPF extensions We can already use yylval in the lexer for encoding the BPF extension number, so that the parser rules can be further reduced to a single one for each B/H/W case. Signed-off-by: Ray Bellis <ray@isc.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-02-22 13:02:40 +03:00
fprintf(stderr, "error: %s at line %d\n", str, yylineno);
filter: bpf_asm: add minimal bpf asm tool There are a couple of valid use cases for a minimal low-level bpf asm like tool, for example, using/linking to libpcap is not an option, the required BPF filters use Linux extensions that are not supported by libpcap's compiler, a filter might be more complex and not cleanly implementable with libpcap's compiler, particular filter codes should be optimized differently than libpcap's internal BPF compiler does, or for security audits of emitted BPF JIT code for prepared set of BPF instructions resp. BPF JIT compiler development in general. Then, in such cases writing such a filter in low-level syntax can be an good alternative, for example, xt_bpf and cls_bpf users might have requirements that could result in more complex filter code, or one that cannot be expressed with libpcap (e.g. different return codes in cls_bpf for flowids on various BPF code paths). Moreover, BPF JIT implementors may wish to manually write test cases in order to verify the resulting JIT image, and thus need low-level access to BPF code generation as well. Therefore, complete the available toolchain for BPF with this small bpf_asm helper tool for the tools/net/ directory. These 3 complementary minimal helper tools round up and facilitate BPF development. Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-12 02:43:44 +04:00
exit(1);
}
Reference in New Issue Copy Permalink