iproute2/tc/m_bpf.c

214 lines
5.4 KiB
C
Raw Normal View History

/*
* m_bpf.c BPF based action module
*
* This program is free software; you can redistribute 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.
*
* Authors: Jiri Pirko <jiri@resnulli.us>
* Daniel Borkmann <daniel@iogearbox.net>
*/
#include <stdio.h>
#include <stdlib.h>
{f,m}_bpf: allow for sharing maps This larger work addresses one of the bigger remaining issues on tc's eBPF frontend, that is, to allow for persistent file descriptors. Whenever tc parses the ELF object, extracts and loads maps into the kernel, these file descriptors will be out of reach after the tc instance exits. Meaning, for simple (unnested) programs which contain one or multiple maps, the kernel holds a reference, and they will live on inside the kernel until the program holding them is unloaded, but they will be out of reach for user space, even worse with (also multiple nested) tail calls. For this issue, we introduced the concept of an agent that can receive the set of file descriptors from the tc instance creating them, in order to be able to further inspect/update map data for a specific use case. However, while that is more tied towards specific applications, it still doesn't easily allow for sharing maps accross multiple tc instances and would require a daemon to be running in the background. F.e. when a map should be shared by two eBPF programs, one attached to ingress, one to egress, this currently doesn't work with the tc frontend. This work solves exactly that, i.e. if requested, maps can now be _arbitrarily_ shared between object files (PIN_GLOBAL_NS) or within a single object (but various program sections, PIN_OBJECT_NS) without "loosing" the file descriptor set. To make that happen, we use eBPF object pinning introduced in kernel commit b2197755b263 ("bpf: add support for persistent maps/progs") for exactly this purpose. The shipped examples/bpf/bpf_shared.c code from this patch can be easily applied, for instance, as: - classifier-classifier shared: tc filter add dev foo parent 1: bpf obj shared.o sec egress tc filter add dev foo parent ffff: bpf obj shared.o sec ingress - classifier-action shared (here: late binding to a dummy classifier): tc actions add action bpf obj shared.o sec egress pass index 42 tc filter add dev foo parent ffff: bpf obj shared.o sec ingress tc filter add dev foo parent 1: bpf bytecode '1,6 0 0 4294967295,' \ action bpf index 42 The toy example increments a shared counter on egress and dumps its value on ingress (if no sharing (PIN_NONE) would have been chosen, map value is 0, of course, due to the two map instances being created): [...] <idle>-0 [002] ..s. 38264.788234: : map val: 4 <idle>-0 [002] ..s. 38264.788919: : map val: 4 <idle>-0 [002] ..s. 38264.789599: : map val: 5 [...] ... thus if both sections reference the pinned map(s) in question, tc will take care of fetching the appropriate file descriptor. The patch has been tested extensively on both, classifier and action sides. Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
2015-11-13 07:39:29 +08:00
#include <linux/bpf.h>
#include <linux/tc_act/tc_bpf.h>
#include "utils.h"
#include "tc_util.h"
#include "bpf_util.h"
static const enum bpf_prog_type bpf_type = BPF_PROG_TYPE_SCHED_ACT;
static void explain(void)
{
fprintf(stderr, "Usage: ... bpf ... [ index INDEX ]\n");
fprintf(stderr, "\n");
fprintf(stderr, "BPF use case:\n");
fprintf(stderr, " bytecode BPF_BYTECODE\n");
fprintf(stderr, " bytecode-file FILE\n");
fprintf(stderr, "\n");
fprintf(stderr, "eBPF use case:\n");
fprintf(stderr, " object-file FILE [ section ACT_NAME ] [ export UDS_FILE ]");
fprintf(stderr, " [ verbose ]\n");
{f,m}_bpf: allow for sharing maps This larger work addresses one of the bigger remaining issues on tc's eBPF frontend, that is, to allow for persistent file descriptors. Whenever tc parses the ELF object, extracts and loads maps into the kernel, these file descriptors will be out of reach after the tc instance exits. Meaning, for simple (unnested) programs which contain one or multiple maps, the kernel holds a reference, and they will live on inside the kernel until the program holding them is unloaded, but they will be out of reach for user space, even worse with (also multiple nested) tail calls. For this issue, we introduced the concept of an agent that can receive the set of file descriptors from the tc instance creating them, in order to be able to further inspect/update map data for a specific use case. However, while that is more tied towards specific applications, it still doesn't easily allow for sharing maps accross multiple tc instances and would require a daemon to be running in the background. F.e. when a map should be shared by two eBPF programs, one attached to ingress, one to egress, this currently doesn't work with the tc frontend. This work solves exactly that, i.e. if requested, maps can now be _arbitrarily_ shared between object files (PIN_GLOBAL_NS) or within a single object (but various program sections, PIN_OBJECT_NS) without "loosing" the file descriptor set. To make that happen, we use eBPF object pinning introduced in kernel commit b2197755b263 ("bpf: add support for persistent maps/progs") for exactly this purpose. The shipped examples/bpf/bpf_shared.c code from this patch can be easily applied, for instance, as: - classifier-classifier shared: tc filter add dev foo parent 1: bpf obj shared.o sec egress tc filter add dev foo parent ffff: bpf obj shared.o sec ingress - classifier-action shared (here: late binding to a dummy classifier): tc actions add action bpf obj shared.o sec egress pass index 42 tc filter add dev foo parent ffff: bpf obj shared.o sec ingress tc filter add dev foo parent 1: bpf bytecode '1,6 0 0 4294967295,' \ action bpf index 42 The toy example increments a shared counter on egress and dumps its value on ingress (if no sharing (PIN_NONE) would have been chosen, map value is 0, of course, due to the two map instances being created): [...] <idle>-0 [002] ..s. 38264.788234: : map val: 4 <idle>-0 [002] ..s. 38264.788919: : map val: 4 <idle>-0 [002] ..s. 38264.789599: : map val: 5 [...] ... thus if both sections reference the pinned map(s) in question, tc will take care of fetching the appropriate file descriptor. The patch has been tested extensively on both, classifier and action sides. Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
2015-11-13 07:39:29 +08:00
fprintf(stderr, " object-pinned FILE\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, "\n");
fprintf(stderr, "Where FILE points to a file containing the BPF_BYTECODE string,\n");
{f,m}_bpf: allow for sharing maps This larger work addresses one of the bigger remaining issues on tc's eBPF frontend, that is, to allow for persistent file descriptors. Whenever tc parses the ELF object, extracts and loads maps into the kernel, these file descriptors will be out of reach after the tc instance exits. Meaning, for simple (unnested) programs which contain one or multiple maps, the kernel holds a reference, and they will live on inside the kernel until the program holding them is unloaded, but they will be out of reach for user space, even worse with (also multiple nested) tail calls. For this issue, we introduced the concept of an agent that can receive the set of file descriptors from the tc instance creating them, in order to be able to further inspect/update map data for a specific use case. However, while that is more tied towards specific applications, it still doesn't easily allow for sharing maps accross multiple tc instances and would require a daemon to be running in the background. F.e. when a map should be shared by two eBPF programs, one attached to ingress, one to egress, this currently doesn't work with the tc frontend. This work solves exactly that, i.e. if requested, maps can now be _arbitrarily_ shared between object files (PIN_GLOBAL_NS) or within a single object (but various program sections, PIN_OBJECT_NS) without "loosing" the file descriptor set. To make that happen, we use eBPF object pinning introduced in kernel commit b2197755b263 ("bpf: add support for persistent maps/progs") for exactly this purpose. The shipped examples/bpf/bpf_shared.c code from this patch can be easily applied, for instance, as: - classifier-classifier shared: tc filter add dev foo parent 1: bpf obj shared.o sec egress tc filter add dev foo parent ffff: bpf obj shared.o sec ingress - classifier-action shared (here: late binding to a dummy classifier): tc actions add action bpf obj shared.o sec egress pass index 42 tc filter add dev foo parent ffff: bpf obj shared.o sec ingress tc filter add dev foo parent 1: bpf bytecode '1,6 0 0 4294967295,' \ action bpf index 42 The toy example increments a shared counter on egress and dumps its value on ingress (if no sharing (PIN_NONE) would have been chosen, map value is 0, of course, due to the two map instances being created): [...] <idle>-0 [002] ..s. 38264.788234: : map val: 4 <idle>-0 [002] ..s. 38264.788919: : map val: 4 <idle>-0 [002] ..s. 38264.789599: : map val: 5 [...] ... thus if both sections reference the pinned map(s) in question, tc will take care of fetching the appropriate file descriptor. The patch has been tested extensively on both, classifier and action sides. Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
2015-11-13 07:39:29 +08:00
fprintf(stderr, "an ELF file containing eBPF map definitions and bytecode, or a\n");
fprintf(stderr, "pinned eBPF program.\n");
fprintf(stderr, "\n");
fprintf(stderr, "Where ACT_NAME refers to the section name containing the\n");
fprintf(stderr, "action (default \'%s\').\n", bpf_prog_to_default_section(bpf_type));
fprintf(stderr, "\n");
fprintf(stderr, "Where UDS_FILE points to a unix domain socket file in order\n");
fprintf(stderr, "to hand off control of all created eBPF maps to an agent.\n");
fprintf(stderr, "\n");
fprintf(stderr, "Where optionally INDEX points to an existing action, or\n");
fprintf(stderr, "explicitly specifies an action index upon creation.\n");
}
static void bpf_cbpf_cb(void *nl, const struct sock_filter *ops, int ops_len)
{
addattr16(nl, MAX_MSG, TCA_ACT_BPF_OPS_LEN, ops_len);
addattr_l(nl, MAX_MSG, TCA_ACT_BPF_OPS, ops,
ops_len * sizeof(struct sock_filter));
}
static void bpf_ebpf_cb(void *nl, int fd, const char *annotation)
{
addattr32(nl, MAX_MSG, TCA_ACT_BPF_FD, fd);
addattrstrz(nl, MAX_MSG, TCA_ACT_BPF_NAME, annotation);
}
static const struct bpf_cfg_ops bpf_cb_ops = {
.cbpf_cb = bpf_cbpf_cb,
.ebpf_cb = bpf_ebpf_cb,
};
{f,m}_bpf: allow for sharing maps This larger work addresses one of the bigger remaining issues on tc's eBPF frontend, that is, to allow for persistent file descriptors. Whenever tc parses the ELF object, extracts and loads maps into the kernel, these file descriptors will be out of reach after the tc instance exits. Meaning, for simple (unnested) programs which contain one or multiple maps, the kernel holds a reference, and they will live on inside the kernel until the program holding them is unloaded, but they will be out of reach for user space, even worse with (also multiple nested) tail calls. For this issue, we introduced the concept of an agent that can receive the set of file descriptors from the tc instance creating them, in order to be able to further inspect/update map data for a specific use case. However, while that is more tied towards specific applications, it still doesn't easily allow for sharing maps accross multiple tc instances and would require a daemon to be running in the background. F.e. when a map should be shared by two eBPF programs, one attached to ingress, one to egress, this currently doesn't work with the tc frontend. This work solves exactly that, i.e. if requested, maps can now be _arbitrarily_ shared between object files (PIN_GLOBAL_NS) or within a single object (but various program sections, PIN_OBJECT_NS) without "loosing" the file descriptor set. To make that happen, we use eBPF object pinning introduced in kernel commit b2197755b263 ("bpf: add support for persistent maps/progs") for exactly this purpose. The shipped examples/bpf/bpf_shared.c code from this patch can be easily applied, for instance, as: - classifier-classifier shared: tc filter add dev foo parent 1: bpf obj shared.o sec egress tc filter add dev foo parent ffff: bpf obj shared.o sec ingress - classifier-action shared (here: late binding to a dummy classifier): tc actions add action bpf obj shared.o sec egress pass index 42 tc filter add dev foo parent ffff: bpf obj shared.o sec ingress tc filter add dev foo parent 1: bpf bytecode '1,6 0 0 4294967295,' \ action bpf index 42 The toy example increments a shared counter on egress and dumps its value on ingress (if no sharing (PIN_NONE) would have been chosen, map value is 0, of course, due to the two map instances being created): [...] <idle>-0 [002] ..s. 38264.788234: : map val: 4 <idle>-0 [002] ..s. 38264.788919: : map val: 4 <idle>-0 [002] ..s. 38264.789599: : map val: 5 [...] ... thus if both sections reference the pinned map(s) in question, tc will take care of fetching the appropriate file descriptor. The patch has been tested extensively on both, classifier and action sides. Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
2015-11-13 07:39:29 +08:00
static int bpf_parse_opt(struct action_util *a, int *ptr_argc, char ***ptr_argv,
int tca_id, struct nlmsghdr *n)
{
{f,m}_bpf: allow for sharing maps This larger work addresses one of the bigger remaining issues on tc's eBPF frontend, that is, to allow for persistent file descriptors. Whenever tc parses the ELF object, extracts and loads maps into the kernel, these file descriptors will be out of reach after the tc instance exits. Meaning, for simple (unnested) programs which contain one or multiple maps, the kernel holds a reference, and they will live on inside the kernel until the program holding them is unloaded, but they will be out of reach for user space, even worse with (also multiple nested) tail calls. For this issue, we introduced the concept of an agent that can receive the set of file descriptors from the tc instance creating them, in order to be able to further inspect/update map data for a specific use case. However, while that is more tied towards specific applications, it still doesn't easily allow for sharing maps accross multiple tc instances and would require a daemon to be running in the background. F.e. when a map should be shared by two eBPF programs, one attached to ingress, one to egress, this currently doesn't work with the tc frontend. This work solves exactly that, i.e. if requested, maps can now be _arbitrarily_ shared between object files (PIN_GLOBAL_NS) or within a single object (but various program sections, PIN_OBJECT_NS) without "loosing" the file descriptor set. To make that happen, we use eBPF object pinning introduced in kernel commit b2197755b263 ("bpf: add support for persistent maps/progs") for exactly this purpose. The shipped examples/bpf/bpf_shared.c code from this patch can be easily applied, for instance, as: - classifier-classifier shared: tc filter add dev foo parent 1: bpf obj shared.o sec egress tc filter add dev foo parent ffff: bpf obj shared.o sec ingress - classifier-action shared (here: late binding to a dummy classifier): tc actions add action bpf obj shared.o sec egress pass index 42 tc filter add dev foo parent ffff: bpf obj shared.o sec ingress tc filter add dev foo parent 1: bpf bytecode '1,6 0 0 4294967295,' \ action bpf index 42 The toy example increments a shared counter on egress and dumps its value on ingress (if no sharing (PIN_NONE) would have been chosen, map value is 0, of course, due to the two map instances being created): [...] <idle>-0 [002] ..s. 38264.788234: : map val: 4 <idle>-0 [002] ..s. 38264.788919: : map val: 4 <idle>-0 [002] ..s. 38264.789599: : map val: 5 [...] ... thus if both sections reference the pinned map(s) in question, tc will take care of fetching the appropriate file descriptor. The patch has been tested extensively on both, classifier and action sides. Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
2015-11-13 07:39:29 +08:00
const char *bpf_obj = NULL, *bpf_uds_name = NULL;
struct tc_act_bpf parm = {};
struct bpf_cfg_in cfg = {};
{f,m}_bpf: allow for sharing maps This larger work addresses one of the bigger remaining issues on tc's eBPF frontend, that is, to allow for persistent file descriptors. Whenever tc parses the ELF object, extracts and loads maps into the kernel, these file descriptors will be out of reach after the tc instance exits. Meaning, for simple (unnested) programs which contain one or multiple maps, the kernel holds a reference, and they will live on inside the kernel until the program holding them is unloaded, but they will be out of reach for user space, even worse with (also multiple nested) tail calls. For this issue, we introduced the concept of an agent that can receive the set of file descriptors from the tc instance creating them, in order to be able to further inspect/update map data for a specific use case. However, while that is more tied towards specific applications, it still doesn't easily allow for sharing maps accross multiple tc instances and would require a daemon to be running in the background. F.e. when a map should be shared by two eBPF programs, one attached to ingress, one to egress, this currently doesn't work with the tc frontend. This work solves exactly that, i.e. if requested, maps can now be _arbitrarily_ shared between object files (PIN_GLOBAL_NS) or within a single object (but various program sections, PIN_OBJECT_NS) without "loosing" the file descriptor set. To make that happen, we use eBPF object pinning introduced in kernel commit b2197755b263 ("bpf: add support for persistent maps/progs") for exactly this purpose. The shipped examples/bpf/bpf_shared.c code from this patch can be easily applied, for instance, as: - classifier-classifier shared: tc filter add dev foo parent 1: bpf obj shared.o sec egress tc filter add dev foo parent ffff: bpf obj shared.o sec ingress - classifier-action shared (here: late binding to a dummy classifier): tc actions add action bpf obj shared.o sec egress pass index 42 tc filter add dev foo parent ffff: bpf obj shared.o sec ingress tc filter add dev foo parent 1: bpf bytecode '1,6 0 0 4294967295,' \ action bpf index 42 The toy example increments a shared counter on egress and dumps its value on ingress (if no sharing (PIN_NONE) would have been chosen, map value is 0, of course, due to the two map instances being created): [...] <idle>-0 [002] ..s. 38264.788234: : map val: 4 <idle>-0 [002] ..s. 38264.788919: : map val: 4 <idle>-0 [002] ..s. 38264.789599: : map val: 5 [...] ... thus if both sections reference the pinned map(s) in question, tc will take care of fetching the appropriate file descriptor. The patch has been tested extensively on both, classifier and action sides. Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
2015-11-13 07:39:29 +08:00
bool seen_run = false;
struct rtattr *tail;
{f,m}_bpf: allow for sharing maps This larger work addresses one of the bigger remaining issues on tc's eBPF frontend, that is, to allow for persistent file descriptors. Whenever tc parses the ELF object, extracts and loads maps into the kernel, these file descriptors will be out of reach after the tc instance exits. Meaning, for simple (unnested) programs which contain one or multiple maps, the kernel holds a reference, and they will live on inside the kernel until the program holding them is unloaded, but they will be out of reach for user space, even worse with (also multiple nested) tail calls. For this issue, we introduced the concept of an agent that can receive the set of file descriptors from the tc instance creating them, in order to be able to further inspect/update map data for a specific use case. However, while that is more tied towards specific applications, it still doesn't easily allow for sharing maps accross multiple tc instances and would require a daemon to be running in the background. F.e. when a map should be shared by two eBPF programs, one attached to ingress, one to egress, this currently doesn't work with the tc frontend. This work solves exactly that, i.e. if requested, maps can now be _arbitrarily_ shared between object files (PIN_GLOBAL_NS) or within a single object (but various program sections, PIN_OBJECT_NS) without "loosing" the file descriptor set. To make that happen, we use eBPF object pinning introduced in kernel commit b2197755b263 ("bpf: add support for persistent maps/progs") for exactly this purpose. The shipped examples/bpf/bpf_shared.c code from this patch can be easily applied, for instance, as: - classifier-classifier shared: tc filter add dev foo parent 1: bpf obj shared.o sec egress tc filter add dev foo parent ffff: bpf obj shared.o sec ingress - classifier-action shared (here: late binding to a dummy classifier): tc actions add action bpf obj shared.o sec egress pass index 42 tc filter add dev foo parent ffff: bpf obj shared.o sec ingress tc filter add dev foo parent 1: bpf bytecode '1,6 0 0 4294967295,' \ action bpf index 42 The toy example increments a shared counter on egress and dumps its value on ingress (if no sharing (PIN_NONE) would have been chosen, map value is 0, of course, due to the two map instances being created): [...] <idle>-0 [002] ..s. 38264.788234: : map val: 4 <idle>-0 [002] ..s. 38264.788919: : map val: 4 <idle>-0 [002] ..s. 38264.789599: : map val: 5 [...] ... thus if both sections reference the pinned map(s) in question, tc will take care of fetching the appropriate file descriptor. The patch has been tested extensively on both, classifier and action sides. Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
2015-11-13 07:39:29 +08:00
int argc, ret = 0;
char **argv;
argv = *ptr_argv;
argc = *ptr_argc;
if (matches(*argv, "bpf") != 0)
return -1;
NEXT_ARG();
{f,m}_bpf: allow for sharing maps This larger work addresses one of the bigger remaining issues on tc's eBPF frontend, that is, to allow for persistent file descriptors. Whenever tc parses the ELF object, extracts and loads maps into the kernel, these file descriptors will be out of reach after the tc instance exits. Meaning, for simple (unnested) programs which contain one or multiple maps, the kernel holds a reference, and they will live on inside the kernel until the program holding them is unloaded, but they will be out of reach for user space, even worse with (also multiple nested) tail calls. For this issue, we introduced the concept of an agent that can receive the set of file descriptors from the tc instance creating them, in order to be able to further inspect/update map data for a specific use case. However, while that is more tied towards specific applications, it still doesn't easily allow for sharing maps accross multiple tc instances and would require a daemon to be running in the background. F.e. when a map should be shared by two eBPF programs, one attached to ingress, one to egress, this currently doesn't work with the tc frontend. This work solves exactly that, i.e. if requested, maps can now be _arbitrarily_ shared between object files (PIN_GLOBAL_NS) or within a single object (but various program sections, PIN_OBJECT_NS) without "loosing" the file descriptor set. To make that happen, we use eBPF object pinning introduced in kernel commit b2197755b263 ("bpf: add support for persistent maps/progs") for exactly this purpose. The shipped examples/bpf/bpf_shared.c code from this patch can be easily applied, for instance, as: - classifier-classifier shared: tc filter add dev foo parent 1: bpf obj shared.o sec egress tc filter add dev foo parent ffff: bpf obj shared.o sec ingress - classifier-action shared (here: late binding to a dummy classifier): tc actions add action bpf obj shared.o sec egress pass index 42 tc filter add dev foo parent ffff: bpf obj shared.o sec ingress tc filter add dev foo parent 1: bpf bytecode '1,6 0 0 4294967295,' \ action bpf index 42 The toy example increments a shared counter on egress and dumps its value on ingress (if no sharing (PIN_NONE) would have been chosen, map value is 0, of course, due to the two map instances being created): [...] <idle>-0 [002] ..s. 38264.788234: : map val: 4 <idle>-0 [002] ..s. 38264.788919: : map val: 4 <idle>-0 [002] ..s. 38264.789599: : map val: 5 [...] ... thus if both sections reference the pinned map(s) in question, tc will take care of fetching the appropriate file descriptor. The patch has been tested extensively on both, classifier and action sides. Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
2015-11-13 07:39:29 +08:00
tail = NLMSG_TAIL(n);
addattr_l(n, MAX_MSG, tca_id, NULL, 0);
while (argc > 0) {
if (matches(*argv, "run") == 0) {
NEXT_ARG();
opt_bpf:
seen_run = true;
cfg.type = bpf_type;
cfg.argc = argc;
cfg.argv = argv;
if (bpf_parse_common(&cfg, &bpf_cb_ops, n))
return -1;
argc = cfg.argc;
argv = cfg.argv;
bpf_obj = cfg.object;
bpf_uds_name = cfg.uds;
} else if (matches(*argv, "help") == 0) {
{f,m}_bpf: allow for sharing maps This larger work addresses one of the bigger remaining issues on tc's eBPF frontend, that is, to allow for persistent file descriptors. Whenever tc parses the ELF object, extracts and loads maps into the kernel, these file descriptors will be out of reach after the tc instance exits. Meaning, for simple (unnested) programs which contain one or multiple maps, the kernel holds a reference, and they will live on inside the kernel until the program holding them is unloaded, but they will be out of reach for user space, even worse with (also multiple nested) tail calls. For this issue, we introduced the concept of an agent that can receive the set of file descriptors from the tc instance creating them, in order to be able to further inspect/update map data for a specific use case. However, while that is more tied towards specific applications, it still doesn't easily allow for sharing maps accross multiple tc instances and would require a daemon to be running in the background. F.e. when a map should be shared by two eBPF programs, one attached to ingress, one to egress, this currently doesn't work with the tc frontend. This work solves exactly that, i.e. if requested, maps can now be _arbitrarily_ shared between object files (PIN_GLOBAL_NS) or within a single object (but various program sections, PIN_OBJECT_NS) without "loosing" the file descriptor set. To make that happen, we use eBPF object pinning introduced in kernel commit b2197755b263 ("bpf: add support for persistent maps/progs") for exactly this purpose. The shipped examples/bpf/bpf_shared.c code from this patch can be easily applied, for instance, as: - classifier-classifier shared: tc filter add dev foo parent 1: bpf obj shared.o sec egress tc filter add dev foo parent ffff: bpf obj shared.o sec ingress - classifier-action shared (here: late binding to a dummy classifier): tc actions add action bpf obj shared.o sec egress pass index 42 tc filter add dev foo parent ffff: bpf obj shared.o sec ingress tc filter add dev foo parent 1: bpf bytecode '1,6 0 0 4294967295,' \ action bpf index 42 The toy example increments a shared counter on egress and dumps its value on ingress (if no sharing (PIN_NONE) would have been chosen, map value is 0, of course, due to the two map instances being created): [...] <idle>-0 [002] ..s. 38264.788234: : map val: 4 <idle>-0 [002] ..s. 38264.788919: : map val: 4 <idle>-0 [002] ..s. 38264.789599: : map val: 5 [...] ... thus if both sections reference the pinned map(s) in question, tc will take care of fetching the appropriate file descriptor. The patch has been tested extensively on both, classifier and action sides. Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
2015-11-13 07:39:29 +08:00
explain();
return -1;
} else if (matches(*argv, "index") == 0) {
break;
} else {
if (!seen_run)
goto opt_bpf;
break;
}
NEXT_ARG_FWD();
}
parse_action_control_dflt(&argc, &argv, &parm.action,
false, TC_ACT_PIPE);
if (argc) {
if (matches(*argv, "index") == 0) {
NEXT_ARG();
if (get_u32(&parm.index, *argv, 10)) {
fprintf(stderr, "bpf: Illegal \"index\"\n");
return -1;
}
NEXT_ARG_FWD();
}
}
addattr_l(n, MAX_MSG, TCA_ACT_BPF_PARMS, &parm, sizeof(parm));
tail->rta_len = (char *)NLMSG_TAIL(n) - (char *)tail;
if (bpf_uds_name)
tc: built-in eBPF exec proxy This work follows upon commit 6256f8c9e45f ("tc, bpf: finalize eBPF support for cls and act front-end") and takes up the idea proposed by Hannes Frederic Sowa to spawn a shell (or any other command) that holds generated eBPF map file descriptors. File descriptors, based on their id, are being fetched from the same unix domain socket as demonstrated in the bpf_agent, the shell spawned via execvpe(2) and the map fds passed over the environment, and thus are made available to applications in the fashion of std{in,out,err} for read/write access, for example in case of iproute2's examples/bpf/: # env | grep BPF BPF_NUM_MAPS=3 BPF_MAP1=6 <- BPF_MAP_ID_QUEUE (id 1) BPF_MAP0=5 <- BPF_MAP_ID_PROTO (id 0) BPF_MAP2=7 <- BPF_MAP_ID_DROPS (id 2) # ls -la /proc/self/fd [...] lrwx------. 1 root root 64 Apr 14 16:46 0 -> /dev/pts/4 lrwx------. 1 root root 64 Apr 14 16:46 1 -> /dev/pts/4 lrwx------. 1 root root 64 Apr 14 16:46 2 -> /dev/pts/4 [...] lrwx------. 1 root root 64 Apr 14 16:46 5 -> anon_inode:bpf-map lrwx------. 1 root root 64 Apr 14 16:46 6 -> anon_inode:bpf-map lrwx------. 1 root root 64 Apr 14 16:46 7 -> anon_inode:bpf-map The advantage (as opposed to the direct/native usage) is that now the shell is map fd owner and applications can terminate and easily reattach to descriptors w/o any kernel changes. Moreover, multiple applications can easily read/write eBPF maps simultaneously. To further allow users for experimenting with that, next step is to add a small helper that can get along with simple data types, so that also shell scripts can make use of bpf syscall, f.e to read/write into maps. Generally, this allows for prepopulating maps, or any runtime altering which could influence eBPF program behaviour (f.e. different run-time classifications, skb modifications, ...), dumping of statistics, etc. Reference: http://thread.gmane.org/gmane.linux.network/357471/focus=357860 Suggested-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Reviewed-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Acked-by: Alexei Starovoitov <ast@plumgrid.com>
2015-04-17 03:20:06 +08:00
ret = bpf_send_map_fds(bpf_uds_name, bpf_obj);
{f,m}_bpf: allow for sharing maps This larger work addresses one of the bigger remaining issues on tc's eBPF frontend, that is, to allow for persistent file descriptors. Whenever tc parses the ELF object, extracts and loads maps into the kernel, these file descriptors will be out of reach after the tc instance exits. Meaning, for simple (unnested) programs which contain one or multiple maps, the kernel holds a reference, and they will live on inside the kernel until the program holding them is unloaded, but they will be out of reach for user space, even worse with (also multiple nested) tail calls. For this issue, we introduced the concept of an agent that can receive the set of file descriptors from the tc instance creating them, in order to be able to further inspect/update map data for a specific use case. However, while that is more tied towards specific applications, it still doesn't easily allow for sharing maps accross multiple tc instances and would require a daemon to be running in the background. F.e. when a map should be shared by two eBPF programs, one attached to ingress, one to egress, this currently doesn't work with the tc frontend. This work solves exactly that, i.e. if requested, maps can now be _arbitrarily_ shared between object files (PIN_GLOBAL_NS) or within a single object (but various program sections, PIN_OBJECT_NS) without "loosing" the file descriptor set. To make that happen, we use eBPF object pinning introduced in kernel commit b2197755b263 ("bpf: add support for persistent maps/progs") for exactly this purpose. The shipped examples/bpf/bpf_shared.c code from this patch can be easily applied, for instance, as: - classifier-classifier shared: tc filter add dev foo parent 1: bpf obj shared.o sec egress tc filter add dev foo parent ffff: bpf obj shared.o sec ingress - classifier-action shared (here: late binding to a dummy classifier): tc actions add action bpf obj shared.o sec egress pass index 42 tc filter add dev foo parent ffff: bpf obj shared.o sec ingress tc filter add dev foo parent 1: bpf bytecode '1,6 0 0 4294967295,' \ action bpf index 42 The toy example increments a shared counter on egress and dumps its value on ingress (if no sharing (PIN_NONE) would have been chosen, map value is 0, of course, due to the two map instances being created): [...] <idle>-0 [002] ..s. 38264.788234: : map val: 4 <idle>-0 [002] ..s. 38264.788919: : map val: 4 <idle>-0 [002] ..s. 38264.789599: : map val: 5 [...] ... thus if both sections reference the pinned map(s) in question, tc will take care of fetching the appropriate file descriptor. The patch has been tested extensively on both, classifier and action sides. Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
2015-11-13 07:39:29 +08:00
*ptr_argc = argc;
*ptr_argv = argv;
return ret;
}
{f,m}_bpf: allow for sharing maps This larger work addresses one of the bigger remaining issues on tc's eBPF frontend, that is, to allow for persistent file descriptors. Whenever tc parses the ELF object, extracts and loads maps into the kernel, these file descriptors will be out of reach after the tc instance exits. Meaning, for simple (unnested) programs which contain one or multiple maps, the kernel holds a reference, and they will live on inside the kernel until the program holding them is unloaded, but they will be out of reach for user space, even worse with (also multiple nested) tail calls. For this issue, we introduced the concept of an agent that can receive the set of file descriptors from the tc instance creating them, in order to be able to further inspect/update map data for a specific use case. However, while that is more tied towards specific applications, it still doesn't easily allow for sharing maps accross multiple tc instances and would require a daemon to be running in the background. F.e. when a map should be shared by two eBPF programs, one attached to ingress, one to egress, this currently doesn't work with the tc frontend. This work solves exactly that, i.e. if requested, maps can now be _arbitrarily_ shared between object files (PIN_GLOBAL_NS) or within a single object (but various program sections, PIN_OBJECT_NS) without "loosing" the file descriptor set. To make that happen, we use eBPF object pinning introduced in kernel commit b2197755b263 ("bpf: add support for persistent maps/progs") for exactly this purpose. The shipped examples/bpf/bpf_shared.c code from this patch can be easily applied, for instance, as: - classifier-classifier shared: tc filter add dev foo parent 1: bpf obj shared.o sec egress tc filter add dev foo parent ffff: bpf obj shared.o sec ingress - classifier-action shared (here: late binding to a dummy classifier): tc actions add action bpf obj shared.o sec egress pass index 42 tc filter add dev foo parent ffff: bpf obj shared.o sec ingress tc filter add dev foo parent 1: bpf bytecode '1,6 0 0 4294967295,' \ action bpf index 42 The toy example increments a shared counter on egress and dumps its value on ingress (if no sharing (PIN_NONE) would have been chosen, map value is 0, of course, due to the two map instances being created): [...] <idle>-0 [002] ..s. 38264.788234: : map val: 4 <idle>-0 [002] ..s. 38264.788919: : map val: 4 <idle>-0 [002] ..s. 38264.789599: : map val: 5 [...] ... thus if both sections reference the pinned map(s) in question, tc will take care of fetching the appropriate file descriptor. The patch has been tested extensively on both, classifier and action sides. Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
2015-11-13 07:39:29 +08:00
static int bpf_print_opt(struct action_util *au, FILE *f, struct rtattr *arg)
{
struct rtattr *tb[TCA_ACT_BPF_MAX + 1];
struct tc_act_bpf *parm;
int dump_ok = 0;
if (arg == NULL)
return -1;
parse_rtattr_nested(tb, TCA_ACT_BPF_MAX, arg);
if (!tb[TCA_ACT_BPF_PARMS]) {
fprintf(f, "[NULL bpf parameters]");
return -1;
}
parm = RTA_DATA(tb[TCA_ACT_BPF_PARMS]);
fprintf(f, "bpf ");
if (tb[TCA_ACT_BPF_NAME])
fprintf(f, "%s ", rta_getattr_str(tb[TCA_ACT_BPF_NAME]));
if (tb[TCA_ACT_BPF_OPS] && tb[TCA_ACT_BPF_OPS_LEN]) {
bpf_print_ops(f, tb[TCA_ACT_BPF_OPS],
rta_getattr_u16(tb[TCA_ACT_BPF_OPS_LEN]));
fprintf(f, " ");
}
if (tb[TCA_ACT_BPF_ID])
dump_ok = bpf_dump_prog_info(f, rta_getattr_u32(tb[TCA_ACT_BPF_ID]));
if (!dump_ok && tb[TCA_ACT_BPF_TAG]) {
SPRINT_BUF(b);
fprintf(f, "tag %s ",
hexstring_n2a(RTA_DATA(tb[TCA_ACT_BPF_TAG]),
RTA_PAYLOAD(tb[TCA_ACT_BPF_TAG]),
b, sizeof(b)));
}
print_action_control(f, "default-action ", parm->action, "\n");
fprintf(f, "\tindex %u ref %d bind %d", parm->index, parm->refcnt,
parm->bindcnt);
if (show_stats) {
if (tb[TCA_ACT_BPF_TM]) {
struct tcf_t *tm = RTA_DATA(tb[TCA_ACT_BPF_TM]);
print_tm(f, tm);
}
}
fprintf(f, "\n ");
return 0;
}
struct action_util bpf_action_util = {
{f,m}_bpf: allow for sharing maps This larger work addresses one of the bigger remaining issues on tc's eBPF frontend, that is, to allow for persistent file descriptors. Whenever tc parses the ELF object, extracts and loads maps into the kernel, these file descriptors will be out of reach after the tc instance exits. Meaning, for simple (unnested) programs which contain one or multiple maps, the kernel holds a reference, and they will live on inside the kernel until the program holding them is unloaded, but they will be out of reach for user space, even worse with (also multiple nested) tail calls. For this issue, we introduced the concept of an agent that can receive the set of file descriptors from the tc instance creating them, in order to be able to further inspect/update map data for a specific use case. However, while that is more tied towards specific applications, it still doesn't easily allow for sharing maps accross multiple tc instances and would require a daemon to be running in the background. F.e. when a map should be shared by two eBPF programs, one attached to ingress, one to egress, this currently doesn't work with the tc frontend. This work solves exactly that, i.e. if requested, maps can now be _arbitrarily_ shared between object files (PIN_GLOBAL_NS) or within a single object (but various program sections, PIN_OBJECT_NS) without "loosing" the file descriptor set. To make that happen, we use eBPF object pinning introduced in kernel commit b2197755b263 ("bpf: add support for persistent maps/progs") for exactly this purpose. The shipped examples/bpf/bpf_shared.c code from this patch can be easily applied, for instance, as: - classifier-classifier shared: tc filter add dev foo parent 1: bpf obj shared.o sec egress tc filter add dev foo parent ffff: bpf obj shared.o sec ingress - classifier-action shared (here: late binding to a dummy classifier): tc actions add action bpf obj shared.o sec egress pass index 42 tc filter add dev foo parent ffff: bpf obj shared.o sec ingress tc filter add dev foo parent 1: bpf bytecode '1,6 0 0 4294967295,' \ action bpf index 42 The toy example increments a shared counter on egress and dumps its value on ingress (if no sharing (PIN_NONE) would have been chosen, map value is 0, of course, due to the two map instances being created): [...] <idle>-0 [002] ..s. 38264.788234: : map val: 4 <idle>-0 [002] ..s. 38264.788919: : map val: 4 <idle>-0 [002] ..s. 38264.789599: : map val: 5 [...] ... thus if both sections reference the pinned map(s) in question, tc will take care of fetching the appropriate file descriptor. The patch has been tested extensively on both, classifier and action sides. Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
2015-11-13 07:39:29 +08:00
.id = "bpf",
.parse_aopt = bpf_parse_opt,
.print_aopt = bpf_print_opt,
};