Cap the maximum (total) value size and bail out if larger than KMALLOC_MAX_SIZE
as otherwise it doesn't make any sense to proceed further, since we're
guaranteed to fail to allocate elements anyway in lpm_trie_node_alloc();
likleyhood of failure is still high for large values, though, similarly
as with htab case in non-prealloc.
Next, make sure that cost vars are really u64 instead of size_t, so that we
don't overflow on 32 bit and charge only tiny map.pages against memlock while
allowing huge max_entries; cap also the max cost like we do with other map
types.
Fixes: b95a5c4db0 ("bpf: add a longest prefix match trie map implementation")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
The patch fixes the case when adding a zero value to the packet
pointer. The zero value could come from src_reg equals type
BPF_K or CONST_IMM. The patch fixes both, otherwise the verifer
reports the following error:
[...]
R0=imm0,min_value=0,max_value=0
R1=pkt(id=0,off=0,r=4)
R2=pkt_end R3=fp-12
R4=imm4,min_value=4,max_value=4
R5=pkt(id=0,off=4,r=4)
269: (bf) r2 = r0 // r2 becomes imm0
270: (77) r2 >>= 3
271: (bf) r4 = r1 // r4 becomes pkt ptr
272: (0f) r4 += r2 // r4 += 0
addition of negative constant to packet pointer is not allowed
Signed-off-by: William Tu <u9012063@gmail.com>
Signed-off-by: Mihai Budiu <mbudiu@vmware.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Alexei Starovoitov <ast@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This work adds a number of tracepoints to paths that are either
considered slow-path or exception-like states, where monitoring or
inspecting them would be desirable.
For bpf(2) syscall, tracepoints have been placed for main commands
when they succeed. In XDP case, tracepoint is for exceptions, that
is, f.e. on abnormal BPF program exit such as unknown or XDP_ABORTED
return code, or when error occurs during XDP_TX action and the packet
could not be forwarded.
Both have been split into separate event headers, and can be further
extended. Worst case, if they unexpectedly should get into our way in
future, they can also removed [1]. Of course, these tracepoints (like
any other) can be analyzed by eBPF itself, etc. Example output:
# ./perf record -a -e bpf:* sleep 10
# ./perf script
sock_example 6197 [005] 283.980322: bpf:bpf_map_create: map type=ARRAY ufd=4 key=4 val=8 max=256 flags=0
sock_example 6197 [005] 283.980721: bpf:bpf_prog_load: prog=a5ea8fa30ea6849c type=SOCKET_FILTER ufd=5
sock_example 6197 [005] 283.988423: bpf:bpf_prog_get_type: prog=a5ea8fa30ea6849c type=SOCKET_FILTER
sock_example 6197 [005] 283.988443: bpf:bpf_map_lookup_elem: map type=ARRAY ufd=4 key=[06 00 00 00] val=[00 00 00 00 00 00 00 00]
[...]
sock_example 6197 [005] 288.990868: bpf:bpf_map_lookup_elem: map type=ARRAY ufd=4 key=[01 00 00 00] val=[14 00 00 00 00 00 00 00]
swapper 0 [005] 289.338243: bpf:bpf_prog_put_rcu: prog=a5ea8fa30ea6849c type=SOCKET_FILTER
[1] https://lwn.net/Articles/705270/
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
William reported couple of issues in relation to direct packet
access. Typical scheme is to check for data + [off] <= data_end,
where [off] can be either immediate or coming from a tracked
register that contains an immediate, depending on the branch, we
can then access the data. However, in case of calculating [off]
for either the mentioned test itself or for access after the test
in a more "complex" way, then the verifier will stop tracking the
CONST_IMM marked register and will mark it as UNKNOWN_VALUE one.
Adding that UNKNOWN_VALUE typed register to a pkt() marked
register, the verifier then bails out in check_packet_ptr_add()
as it finds the registers imm value below 48. In the first below
example, that is due to evaluate_reg_imm_alu() not handling right
shifts and thus marking the register as UNKNOWN_VALUE via helper
__mark_reg_unknown_value() that resets imm to 0.
In the second case the same happens at the time when r4 is set
to r4 &= r5, where it transitions to UNKNOWN_VALUE from
evaluate_reg_imm_alu(). Later on r4 we shift right by 3 inside
evaluate_reg_alu(), where the register's imm turns into 3. That
is, for registers with type UNKNOWN_VALUE, imm of 0 means that
we don't know what value the register has, and for imm > 0 it
means that the value has [imm] upper zero bits. F.e. when shifting
an UNKNOWN_VALUE register by 3 to the right, no matter what value
it had, we know that the 3 upper most bits must be zero now.
This is to make sure that ALU operations with unknown registers
don't overflow. Meaning, once we know that we have more than 48
upper zero bits, or, in other words cannot go beyond 0xffff offset
with ALU ops, such an addition will track the target register
as a new pkt() register with a new id, but 0 offset and 0 range,
so for that a new data/data_end test will be required. Is the source
register a CONST_IMM one that is to be added to the pkt() register,
or the source instruction is an add instruction with immediate
value, then it will get added if it stays within max 0xffff bounds.
>From there, pkt() type, can be accessed should reg->off + imm be
within the access range of pkt().
[...]
from 28 to 30: R0=imm1,min_value=1,max_value=1
R1=pkt(id=0,off=0,r=22) R2=pkt_end
R3=imm144,min_value=144,max_value=144
R4=imm0,min_value=0,max_value=0
R5=inv48,min_value=2054,max_value=2054 R10=fp
30: (bf) r5 = r3
31: (07) r5 += 23
32: (77) r5 >>= 3
33: (bf) r6 = r1
34: (0f) r6 += r5
cannot add integer value with 0 upper zero bits to ptr_to_packet
[...]
from 52 to 80: R0=imm1,min_value=1,max_value=1
R1=pkt(id=0,off=0,r=34) R2=pkt_end R3=inv
R4=imm272 R5=inv56,min_value=17,max_value=17
R6=pkt(id=0,off=26,r=34) R10=fp
80: (07) r4 += 71
81: (18) r5 = 0xfffffff8
83: (5f) r4 &= r5
84: (77) r4 >>= 3
85: (0f) r1 += r4
cannot add integer value with 3 upper zero bits to ptr_to_packet
Thus to get above use-cases working, evaluate_reg_imm_alu() has
been extended for further ALU ops. This is fine, because we only
operate strictly within realm of CONST_IMM types, so here we don't
care about overflows as they will happen in the simulated but also
real execution and interaction with pkt() in check_packet_ptr_add()
will check actual imm value once added to pkt(), but it's irrelevant
before.
With regards to 06c1c04972 ("bpf: allow helpers access to variable
memory") that works on UNKNOWN_VALUE registers, the verifier becomes
now a bit smarter as it can better resolve ALU ops, so we need to
adapt two test cases there, as min/max bound tracking only becomes
necessary when registers were spilled to stack. So while mask was
set before to track upper bound for UNKNOWN_VALUE case, it's now
resolved directly as CONST_IMM, and such contructs are only necessary
when f.e. registers are spilled.
For commit 6b17387307 ("bpf: recognize 64bit immediate loads as
consts") that initially enabled dw load tracking only for nfp jit/
analyzer, I did couple of tests on large, complex programs and we
don't increase complexity badly (my tests were in ~3% range on avg).
I've added a couple of tests similar to affected code above, and
it works fine with verifier now.
Reported-by: William Tu <u9012063@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Cc: Gianluca Borello <g.borello@gmail.com>
Cc: William Tu <u9012063@gmail.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
We need to initialize im_node to NULL, otherwise in case of error path
it gets passed to kfree() as uninitialized pointer.
Fixes: b95a5c4db0 ("bpf: add a longest prefix match trie map implementation")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This trie implements a longest prefix match algorithm that can be used
to match IP addresses to a stored set of ranges.
Internally, data is stored in an unbalanced trie of nodes that has a
maximum height of n, where n is the prefixlen the trie was created
with.
Tries may be created with prefix lengths that are multiples of 8, in
the range from 8 to 2048. The key used for lookup and update operations
is a struct bpf_lpm_trie_key, and the value is a uint64_t.
The code carries more information about the internal implementation.
Signed-off-by: Daniel Mack <daniel@zonque.org>
Reviewed-by: David Herrmann <dh.herrmann@gmail.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds two helpers, bpf_map_area_alloc() and bpf_map_area_free(),
that are to be used for map allocations. Using kmalloc() for very large
allocations can cause excessive work within the page allocator, so i) fall
back earlier to vmalloc() when the attempt is considered costly anyway,
and even more importantly ii) don't trigger OOM killer with any of the
allocators.
Since this is based on a user space request, for example, when creating
maps with element pre-allocation, we really want such requests to fail
instead of killing other user space processes.
Also, don't spam the kernel log with warnings should any of the allocations
fail under pressure. Given that, we can make backend selection in
bpf_map_area_alloc() generic, and convert all maps over to use this API
for spots with potentially large allocation requests.
Note, replacing the one kmalloc_array() is fine as overflow checks happen
earlier in htab_map_alloc(), since it must also protect the multiplication
for vmalloc() should kmalloc_array() fail.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 7bd509e311 ("bpf: add prog_digest and expose it via
fdinfo/netlink") was recently discussed, partially due to
admittedly suboptimal name of "prog_digest" in combination
with sha1 hash usage, thus inevitably and rightfully concerns
about its security in terms of collision resistance were
raised with regards to use-cases.
The intended use cases are for debugging resp. introspection
only for providing a stable "tag" over the instruction sequence
that both kernel and user space can calculate independently.
It's not usable at all for making a security relevant decision.
So collisions where two different instruction sequences generate
the same tag can happen, but ideally at a rather low rate. The
"tag" will be dumped in hex and is short enough to introspect
in tracepoints or kallsyms output along with other data such
as stack trace, etc. Thus, this patch performs a rename into
prog_tag and truncates the tag to a short output (64 bits) to
make it obvious it's not collision-free.
Should in future a hash or facility be needed with a security
relevant focus, then we can think about requirements, constraints,
etc that would fit to that situation. For now, rework the exposed
parts for the current use cases as long as nothing has been
released yet. Tested on x86_64 and s390x.
Fixes: 7bd509e311 ("bpf: add prog_digest and expose it via fdinfo/netlink")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
When structs are used to store temporary state in cb[] buffer that is
used with programs and among tail calls, then the generated code will
not always access the buffer in bpf_w chunks. We can ease programming
of it and let this act more natural by allowing for aligned b/h/w/dw
sized access for cb[] ctx member. Various test cases are attached as
well for the selftest suite. Potentially, this can also be reused for
other program types to pass data around.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, when calling convert_ctx_access() callback for the various
program types, we pass in insn->dst_reg, insn->src_reg, insn->off from
the original instruction. This information is needed to rewrite the
instruction that is based on the user ctx structure into a kernel
representation for the ctx. As we'd like to allow access size beyond
just BPF_W, we'd need also insn->code for that in order to decode the
original access size. Given that, lets just pass insn directly to the
convert_ctx_access() callback and work on that to not clutter the
callback with even more arguments we need to pass when everything is
already contained in insn. So lets go through that once, no functional
change.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 01b3f52157 ("bpf: fix allocation warnings in bpf maps and
integer overflow") has added checks for the maximum allocateable size.
It (ab)used KMALLOC_SHIFT_MAX for that purpose.
While this is not incorrect it is not very clean because we already have
KMALLOC_MAX_SIZE for this very reason so let's change both checks to use
KMALLOC_MAX_SIZE instead.
The original motivation for using KMALLOC_SHIFT_MAX was to work around
an incorrect KMALLOC_MAX_SIZE which could lead to allocation warnings
but it is no longer needed since "slab: make sure that KMALLOC_MAX_SIZE
will fit into MAX_ORDER".
Link: http://lkml.kernel.org/r/20161220130659.16461-3-mhocko@kernel.org
Signed-off-by: Michal Hocko <mhocko@suse.com>
Acked-by: Christoph Lameter <cl@linux.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Andrey Konovalov <andreyknvl@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Make the functions __local_list_pop_free(), __local_list_pop_pending(),
bpf_common_lru_populate() and bpf_percpu_lru_populate() static as they
are not used outide of bpf_lru_list.c
This fixes the following GCC warnings when building with 'W=1':
kernel/bpf/bpf_lru_list.c:363:22: warning: no previous prototype for ‘__local_list_pop_free’ [-Wmissing-prototypes]
kernel/bpf/bpf_lru_list.c:376:22: warning: no previous prototype for ‘__local_list_pop_pending’ [-Wmissing-prototypes]
kernel/bpf/bpf_lru_list.c:560:6: warning: no previous prototype for ‘bpf_common_lru_populate’ [-Wmissing-prototypes]
kernel/bpf/bpf_lru_list.c:577:6: warning: no previous prototype for ‘bpf_percpu_lru_populate’ [-Wmissing-prototypes]
Cc: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Remove the unused but set variable 'first_node' in
__bpf_lru_list_shrink_inactive() to fix the following GCC warning when
building with 'W=1':
kernel/bpf/bpf_lru_list.c:216:41: warning: variable ‘first_node’ set but not used [-Wunused-but-set-variable]
Cc: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
since ARG_PTR_TO_STACK is no longer just pointer to stack
rename it to ARG_PTR_TO_MEM and adjust comment.
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, helpers that read and write from/to the stack can do so using
a pair of arguments of type ARG_PTR_TO_STACK and ARG_CONST_STACK_SIZE.
ARG_CONST_STACK_SIZE accepts a constant register of type CONST_IMM, so
that the verifier can safely check the memory access. However, requiring
the argument to be a constant can be limiting in some circumstances.
Since the current logic keeps track of the minimum and maximum value of
a register throughout the simulated execution, ARG_CONST_STACK_SIZE can
be changed to also accept an UNKNOWN_VALUE register in case its
boundaries have been set and the range doesn't cause invalid memory
accesses.
One common situation when this is useful:
int len;
char buf[BUFSIZE]; /* BUFSIZE is 128 */
if (some_condition)
len = 42;
else
len = 84;
some_helper(..., buf, len & (BUFSIZE - 1));
The compiler can often decide to assign the constant values 42 or 48
into a variable on the stack, instead of keeping it in a register. When
the variable is then read back from stack into the register in order to
be passed to the helper, the verifier will not be able to recognize the
register as constant (the verifier is not currently tracking all
constant writes into memory), and the program won't be valid.
However, by allowing the helper to accept an UNKNOWN_VALUE register,
this program will work because the bitwise AND operation will set the
range of possible values for the UNKNOWN_VALUE register to [0, BUFSIZE),
so the verifier can guarantee the helper call will be safe (assuming the
argument is of type ARG_CONST_STACK_SIZE_OR_ZERO, otherwise one more
check against 0 would be needed). Custom ranges can be set not only with
ALU operations, but also by explicitly comparing the UNKNOWN_VALUE
register with constants.
Another very common example happens when intercepting system call
arguments and accessing user-provided data of variable size using
bpf_probe_read(). One can load at runtime the user-provided length in an
UNKNOWN_VALUE register, and then read that exact amount of data up to a
compile-time determined limit in order to fit into the proper local
storage allocated on the stack, without having to guess a suboptimal
access size at compile time.
Also, in case the helpers accepting the UNKNOWN_VALUE register operate
in raw mode, disable the raw mode so that the program is required to
initialize all memory, since there is no guarantee the helper will fill
it completely, leaving possibilities for data leak (just relevant when
the memory used by the helper is the stack, not when using a pointer to
map element value or packet). In other words, ARG_PTR_TO_RAW_STACK will
be treated as ARG_PTR_TO_STACK.
Signed-off-by: Gianluca Borello <g.borello@gmail.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
commit 484611357c ("bpf: allow access into map value arrays")
introduces the ability to do pointer math inside a map element value via
the PTR_TO_MAP_VALUE_ADJ register type.
The current support doesn't handle the case where a PTR_TO_MAP_VALUE_ADJ
is spilled into the stack, limiting several use cases, especially when
generating bpf code from a compiler.
Handle this case by explicitly enabling the register type
PTR_TO_MAP_VALUE_ADJ to be spilled. Also, make sure that min_value and
max_value are reset just for BPF_LDX operations that don't result in a
restore of a spilled register from stack.
Signed-off-by: Gianluca Borello <g.borello@gmail.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Enable helpers to directly access a map element value by passing a
register type PTR_TO_MAP_VALUE (or PTR_TO_MAP_VALUE_ADJ) to helper
arguments ARG_PTR_TO_STACK or ARG_PTR_TO_RAW_STACK.
This enables several use cases. For example, a typical tracing program
might want to capture pathnames passed to sys_open() with:
struct trace_data {
char pathname[PATHLEN];
};
SEC("kprobe/sys_open")
void bpf_sys_open(struct pt_regs *ctx)
{
struct trace_data data;
bpf_probe_read(data.pathname, sizeof(data.pathname), ctx->di);
/* consume data.pathname, for example via
* bpf_trace_printk() or bpf_perf_event_output()
*/
}
Such a program could easily hit the stack limit in case PATHLEN needs to
be large or more local variables need to exist, both of which are quite
common scenarios. Allowing direct helper access to map element values,
one could do:
struct bpf_map_def SEC("maps") scratch_map = {
.type = BPF_MAP_TYPE_PERCPU_ARRAY,
.key_size = sizeof(u32),
.value_size = sizeof(struct trace_data),
.max_entries = 1,
};
SEC("kprobe/sys_open")
int bpf_sys_open(struct pt_regs *ctx)
{
int id = 0;
struct trace_data *p = bpf_map_lookup_elem(&scratch_map, &id);
if (!p)
return;
bpf_probe_read(p->pathname, sizeof(p->pathname), ctx->di);
/* consume p->pathname, for example via
* bpf_trace_printk() or bpf_perf_event_output()
*/
}
And wouldn't risk exhausting the stack.
Code changes are loosely modeled after commit 6841de8b0d ("bpf: allow
helpers access the packet directly"). Unlike with PTR_TO_PACKET, these
changes just work with ARG_PTR_TO_STACK and ARG_PTR_TO_RAW_STACK (not
ARG_PTR_TO_MAP_KEY, ARG_PTR_TO_MAP_VALUE, ...): adding those would be
trivial, but since there is not currently a use case for that, it's
reasonable to limit the set of changes.
Also, add new tests to make sure accesses to map element values from
helpers never go out of boundary, even when adjusted.
Signed-off-by: Gianluca Borello <g.borello@gmail.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Move the logic to check memory accesses to a PTR_TO_MAP_VALUE_ADJ from
check_mem_access() to a separate helper check_map_access_adj(). This
enables to use those checks in other parts of the verifier as well,
where boundaries on PTR_TO_MAP_VALUE_ADJ might need to be checked, for
example when checking helper function arguments. The same thing is
already happening for other types such as PTR_TO_PACKET and its
check_packet_access() helper.
The code has been copied verbatim, with the only difference of removing
the "off += reg->max_value" statement and moving the sum into the call
statement to check_map_access(), as that was only needed due to the
earlier common check_map_access() call.
Signed-off-by: Gianluca Borello <g.borello@gmail.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Martin reported a verifier issue that hit the BUG_ON() for his
test case in the mark_reg_unknown_value() function:
[ 202.861380] kernel BUG at kernel/bpf/verifier.c:467!
[...]
[ 203.291109] Call Trace:
[ 203.296501] [<ffffffff811364d5>] mark_map_reg+0x45/0x50
[ 203.308225] [<ffffffff81136558>] mark_map_regs+0x78/0x90
[ 203.320140] [<ffffffff8113938d>] do_check+0x226d/0x2c90
[ 203.331865] [<ffffffff8113a6ab>] bpf_check+0x48b/0x780
[ 203.343403] [<ffffffff81134c8e>] bpf_prog_load+0x27e/0x440
[ 203.355705] [<ffffffff8118a38f>] ? handle_mm_fault+0x11af/0x1230
[ 203.369158] [<ffffffff812d8188>] ? security_capable+0x48/0x60
[ 203.382035] [<ffffffff811351a4>] SyS_bpf+0x124/0x960
[ 203.393185] [<ffffffff810515f6>] ? __do_page_fault+0x276/0x490
[ 203.406258] [<ffffffff816db320>] entry_SYSCALL_64_fastpath+0x13/0x94
This issue got uncovered after the fix in a08dd0da53 ("bpf: fix
regression on verifier pruning wrt map lookups"). The reason why it
wasn't noticed before was, because as mentioned in a08dd0da53,
mark_map_regs() was doing the id matching incorrectly based on the
uncached regs[regno].id. So, in the first loop, we walked all regs
and as soon as we found regno == i, then this reg's id was cleared
when calling mark_reg_unknown_value() thus that every subsequent
register was probed against id of 0 (which, in combination with the
PTR_TO_MAP_VALUE_OR_NULL type is an invalid condition that no other
register state can hold), and therefore wasn't type transitioned such
as in the spilled register case for the second loop.
Now since that got fixed, it turned out that 57a09bf0a4 ("bpf:
Detect identical PTR_TO_MAP_VALUE_OR_NULL registers") used
mark_reg_unknown_value() incorrectly for the spilled regs, and thus
hitting the BUG_ON() in some cases due to regno >= MAX_BPF_REG.
Although spilled regs have the same type as the non-spilled regs
for the verifier state, that is, struct bpf_reg_state, they are
semantically different from the non-spilled regs. In other words,
there can be up to 64 (MAX_BPF_STACK / BPF_REG_SIZE) spilled regs
in the stack, for example, register R<x> could have been spilled by
the program to stack location X, Y, Z, and in mark_map_regs() we
need to scan these stack slots of type STACK_SPILL for potential
registers that we have to transition from PTR_TO_MAP_VALUE_OR_NULL.
Therefore, depending on the location, the spilled_regs regno can
be a lot higher than just MAX_BPF_REG's value since we operate on
stack instead. The reset in mark_reg_unknown_value() itself is
just fine, only that the BUG_ON() was inappropriate for this. Fix
it by making a __mark_reg_unknown_value() version that can be
called from mark_map_reg() generically; we know for the non-spilled
case that the regno is always < MAX_BPF_REG anyway.
Fixes: 57a09bf0a4 ("bpf: Detect identical PTR_TO_MAP_VALUE_OR_NULL registers")
Reported-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit aaac3ba95e ("bpf: charge user for creation of BPF maps and
programs") made a wrong assumption of charging against prog->pages.
Unlike map->pages, prog->pages are still subject to change when we
need to expand the program through bpf_prog_realloc().
This can for example happen during verification stage when we need to
expand and rewrite parts of the program. Should the required space
cross a page boundary, then prog->pages is not the same anymore as
its original value that we used to bpf_prog_charge_memlock() on. Thus,
we'll hit a wrap-around during bpf_prog_uncharge_memlock() when prog
is freed eventually. I noticed this that despite having unlimited
memlock, programs suddenly refused to load with EPERM error due to
insufficient memlock.
There are two ways to fix this issue. One would be to add a cached
variable to struct bpf_prog that takes a snapshot of prog->pages at the
time of charging. The other approach is to also account for resizes. I
chose to go with the latter for a couple of reasons: i) We want accounting
rather to be more accurate instead of further fooling limits, ii) adding
yet another page counter on struct bpf_prog would also be a waste just
for this purpose. We also do want to charge as early as possible to
avoid going into the verifier just to find out later on that we crossed
limits. The only place that needs to be fixed is bpf_prog_realloc(),
since only here we expand the program, so we try to account for the
needed delta and should we fail, call-sites check for outcome anyway.
On cBPF to eBPF migrations, we don't grab a reference to the user as
they are charged differently. With that in place, my test case worked
fine.
Fixes: aaac3ba95e ("bpf: charge user for creation of BPF maps and programs")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Geert rightfully complained that 7bd509e311 ("bpf: add prog_digest
and expose it via fdinfo/netlink") added a too large allocation of
variable 'raw' from bss section, and should instead be done dynamically:
# ./scripts/bloat-o-meter kernel/bpf/core.o.1 kernel/bpf/core.o.2
add/remove: 3/0 grow/shrink: 0/0 up/down: 33291/0 (33291)
function old new delta
raw - 32832 +32832
[...]
Since this is only relevant during program creation path, which can be
considered slow-path anyway, lets allocate that dynamically and be not
implicitly dependent on verifier mutex. Move bpf_prog_calc_digest() at
the beginning of replace_map_fd_with_map_ptr() and also error handling
stays straight forward.
Reported-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 57a09bf0a4 ("bpf: Detect identical PTR_TO_MAP_VALUE_OR_NULL
registers") introduced a regression where existing programs stopped
loading due to reaching the verifier's maximum complexity limit,
whereas prior to this commit they were loading just fine; the affected
program has roughly 2k instructions.
What was found is that state pruning couldn't be performed effectively
anymore due to mismatches of the verifier's register state, in particular
in the id tracking. It doesn't mean that 57a09bf0a4 is incorrect per
se, but rather that verifier needs to perform a lot more work for the
same program with regards to involved map lookups.
Since commit 57a09bf0a4 is only about tracking registers with type
PTR_TO_MAP_VALUE_OR_NULL, the id is only needed to follow registers
until they are promoted through pattern matching with a NULL check to
either PTR_TO_MAP_VALUE or UNKNOWN_VALUE type. After that point, the
id becomes irrelevant for the transitioned types.
For UNKNOWN_VALUE, id is already reset to 0 via mark_reg_unknown_value(),
but not so for PTR_TO_MAP_VALUE where id is becoming stale. It's even
transferred further into other types that don't make use of it. Among
others, one example is where UNKNOWN_VALUE is set on function call
return with RET_INTEGER return type.
states_equal() will then fall through the memcmp() on register state;
note that the second memcmp() uses offsetofend(), so the id is part of
that since d2a4dd37f6 ("bpf: fix state equivalence"). But the bisect
pointed already to 57a09bf0a4, where we really reach beyond complexity
limit. What I found was that states_equal() often failed in this
case due to id mismatches in spilled regs with registers in type
PTR_TO_MAP_VALUE. Unlike non-spilled regs, spilled regs just perform
a memcmp() on their reg state and don't have any other optimizations
in place, therefore also id was relevant in this case for making a
pruning decision.
We can safely reset id to 0 as well when converting to PTR_TO_MAP_VALUE.
For the affected program, it resulted in a ~17 fold reduction of
complexity and let the program load fine again. Selftest suite also
runs fine. The only other place where env->id_gen is used currently is
through direct packet access, but for these cases id is long living, thus
a different scenario.
Also, the current logic in mark_map_regs() is not fully correct when
marking NULL branch with UNKNOWN_VALUE. We need to cache the destination
reg's id in any case. Otherwise, once we marked that reg as UNKNOWN_VALUE,
it's id is reset and any subsequent registers that hold the original id
and are of type PTR_TO_MAP_VALUE_OR_NULL won't be marked UNKNOWN_VALUE
anymore, since mark_map_reg() reuses the uncached regs[regno].id that
was just overridden. Note, we don't need to cache it outside of
mark_map_regs(), since it's called once on this_branch and the other
time on other_branch, which are both two independent verifier states.
A test case for this is added here, too.
Fixes: 57a09bf0a4 ("bpf: Detect identical PTR_TO_MAP_VALUE_OR_NULL registers")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Thomas Graf <tgraf@suug.ch>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch allows XDP prog to extend/remove the packet
data at the head (like adding or removing header). It is
done by adding a new XDP helper bpf_xdp_adjust_head().
It also renames bpf_helper_changes_skb_data() to
bpf_helper_changes_pkt_data() to better reflect
that XDP prog does not work on skb.
This patch adds one "xdp_adjust_head" bit to bpf_prog for the
XDP-capable driver to check if the XDP prog requires
bpf_xdp_adjust_head() support. The driver can then decide
to error out during XDP_SETUP_PROG.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commmits 57a09bf0a4 ("bpf: Detect identical PTR_TO_MAP_VALUE_OR_NULL registers")
and 484611357c ("bpf: allow access into map value arrays") by themselves
are correct, but in combination they make state equivalence ignore 'id' field
of the register state which can lead to accepting invalid program.
Fixes: 57a09bf0a4 ("bpf: Detect identical PTR_TO_MAP_VALUE_OR_NULL registers")
Fixes: 484611357c ("bpf: allow access into map value arrays")
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
General assumption is that single program can hold up to BPF_MAXINSNS,
that is, 4096 number of instructions. It is the case with cBPF and
that limit was carried over to eBPF. When recently testing digest, I
noticed that it's actually not possible to feed 4096 instructions
via bpf(2).
The check for > BPF_MAXINSNS was added back then to bpf_check() in
cbd3570086 ("bpf: verifier (add ability to receive verification log)").
However, 09756af468 ("bpf: expand BPF syscall with program load/unload")
added yet another check that comes before that into bpf_prog_load(),
but this time bails out already in case of >= BPF_MAXINSNS.
Fix it up and perform the check early in bpf_prog_load(), so we can drop
the second one in bpf_check(). It makes sense, because also a 0 insn
program is useless and we don't want to waste any resources doing work
up to bpf_check() point. The existing bpf(2) man page documents E2BIG
as the official error for such cases, so just stick with it as well.
Fixes: 09756af468 ("bpf: expand BPF syscall with program load/unload")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
When loading a BPF program via bpf(2), calculate the digest over
the program's instruction stream and store it in struct bpf_prog's
digest member. This is done at a point in time before any instructions
are rewritten by the verifier. Any unstable map file descriptor
number part of the imm field will be zeroed for the hash.
fdinfo example output for progs:
# cat /proc/1590/fdinfo/5
pos: 0
flags: 02000002
mnt_id: 11
prog_type: 1
prog_jited: 1
prog_digest: b27e8b06da22707513aa97363dfb11c7c3675d28
memlock: 4096
When programs are pinned and retrieved by an ELF loader, the loader
can check the program's digest through fdinfo and compare it against
one that was generated over the ELF file's program section to see
if the program needs to be reloaded. Furthermore, this can also be
exposed through other means such as netlink in case of a tc cls/act
dump (or xdp in future), but also through tracepoints or other
facilities to identify the program. Other than that, the digest can
also serve as a base name for the work in progress kallsyms support
of programs. The digest doesn't depend/select the crypto layer, since
we need to keep dependencies to a minimum. iproute2 will get support
for this facility.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Occasionally, clang (e.g. version 3.8.1) translates a sum between two
constant operands using a BPF_OR instead of a BPF_ADD. The verifier is
currently not handling this scenario, and the destination register type
becomes UNKNOWN_VALUE even if it's still storing a constant. As a result,
the destination register cannot be used as argument to a helper function
expecting a ARG_CONST_STACK_*, limiting some use cases.
Modify the verifier to handle this case, and add a few tests to make sure
all combinations are supported, and stack boundaries are still verified
even with BPF_OR.
Signed-off-by: Gianluca Borello <g.borello@gmail.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Couple conflicts resolved here:
1) In the MACB driver, a bug fix to properly initialize the
RX tail pointer properly overlapped with some changes
to support variable sized rings.
2) In XGBE we had a "CONFIG_PM" --> "CONFIG_PM_SLEEP" fix
overlapping with a reorganization of the driver to support
ACPI, OF, as well as PCI variants of the chip.
3) In 'net' we had several probe error path bug fixes to the
stmmac driver, meanwhile a lot of this code was cleaned up
and reorganized in 'net-next'.
4) The cls_flower classifier obtained a helper function in
'net-next' called __fl_delete() and this overlapped with
Daniel Borkamann's bug fix to use RCU for object destruction
in 'net'. It also overlapped with Jiri's change to guard
the rhashtable_remove_fast() call with a check against
tc_skip_sw().
5) In mlx4, a revert bug fix in 'net' overlapped with some
unrelated changes in 'net-next'.
6) In geneve, a stale header pointer after pskb_expand_head()
bug fix in 'net' overlapped with a large reorganization of
the same code in 'net-next'. Since the 'net-next' code no
longer had the bug in question, there was nothing to do
other than to simply take the 'net-next' hunks.
Signed-off-by: David S. Miller <davem@davemloft.net>
Add new cgroup based program type, BPF_PROG_TYPE_CGROUP_SOCK. Similar to
BPF_PROG_TYPE_CGROUP_SKB programs can be attached to a cgroup and run
any time a process in the cgroup opens an AF_INET or AF_INET6 socket.
Currently only sk_bound_dev_if is exported to userspace for modification
by a bpf program.
This allows a cgroup to be configured such that AF_INET{6} sockets opened
by processes are automatically bound to a specific device. In turn, this
enables the running of programs that do not support SO_BINDTODEVICE in a
specific VRF context / L3 domain.
Signed-off-by: David Ahern <dsa@cumulusnetworks.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Code move and rename only; no functional change intended.
Signed-off-by: David Ahern <dsa@cumulusnetworks.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Registers new BPF program types which correspond to the LWT hooks:
- BPF_PROG_TYPE_LWT_IN => dst_input()
- BPF_PROG_TYPE_LWT_OUT => dst_output()
- BPF_PROG_TYPE_LWT_XMIT => lwtunnel_xmit()
The separate program types are required to differentiate between the
capabilities each LWT hook allows:
* Programs attached to dst_input() or dst_output() are restricted and
may only read the data of an skb. This prevent modification and
possible invalidation of already validated packet headers on receive
and the construction of illegal headers while the IP headers are
still being assembled.
* Programs attached to lwtunnel_xmit() are allowed to modify packet
content as well as prepending an L2 header via a newly introduced
helper bpf_skb_change_head(). This is safe as lwtunnel_xmit() is
invoked after the IP header has been assembled completely.
All BPF programs receive an skb with L3 headers attached and may return
one of the following error codes:
BPF_OK - Continue routing as per nexthop
BPF_DROP - Drop skb and return EPERM
BPF_REDIRECT - Redirect skb to device as per redirect() helper.
(Only valid in lwtunnel_xmit() context)
The return codes are binary compatible with their TC_ACT_
relatives to ease compatibility.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
If we have a branch that looks something like this
int foo = map->value;
if (condition) {
foo += blah;
} else {
foo = bar;
}
map->array[foo] = baz;
We will incorrectly assume that the !condition branch is equal to the condition
branch as the register for foo will be UNKNOWN_VALUE in both cases. We need to
adjust this logic to only do this if we didn't do a varlen access after we
processed the !condition branch, otherwise we have different ranges and need to
check the other branch as well.
Fixes: 484611357c ("bpf: allow access into map value arrays")
Reported-by: Jann Horn <jannh@google.com>
Signed-off-by: Josef Bacik <jbacik@fb.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
There's a 'not' missing in one paragraph. Add it.
Fixes: 3007098494 ("cgroup: add support for eBPF programs")
Signed-off-by: Daniel Mack <daniel@zonque.org>
Reported-by: Rami Rosen <roszenrami@gmail.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since we recently converted the BPF filesystem over to use mount_nodev(),
we now have the possibility to also hold mount options in sb's s_fs_info.
This work implements mount options support for specifying permissions on
the sb's inode, which will be used by tc when it manually needs to mount
the fs.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Allow for checking the owner_prog_type of a program array map. In some
cases bpf(2) can return -EINVAL /after/ the verifier passed and did all
the rewrites of the bpf program.
The reason that lets us fail at this late stage is that program array
maps are incompatible. Allow users to inspect this earlier after they
got the map fd through BPF_OBJ_GET command. tc will get support for this.
Also, display how much we charged the map with regards to RLIMIT_MEMLOCK.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Extend the bpf(2) syscall by two new commands, BPF_PROG_ATTACH and
BPF_PROG_DETACH which allow attaching and detaching eBPF programs
to a target.
On the API level, the target could be anything that has an fd in
userspace, hence the name of the field in union bpf_attr is called
'target_fd'.
When called with BPF_ATTACH_TYPE_CGROUP_INET_{E,IN}GRESS, the target is
expected to be a valid file descriptor of a cgroup v2 directory which
has the bpf controller enabled. These are the only use-cases
implemented by this patch at this point, but more can be added.
If a program of the given type already exists in the given cgroup,
the program is swapped automically, so userspace does not have to drop
an existing program first before installing a new one, which would
otherwise leave a gap in which no program is attached.
For more information on the propagation logic to subcgroups, please
refer to the bpf cgroup controller implementation.
The API is guarded by CAP_NET_ADMIN.
Signed-off-by: Daniel Mack <daniel@zonque.org>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds two sets of eBPF program pointers to struct cgroup.
One for such that are directly pinned to a cgroup, and one for such
that are effective for it.
To illustrate the logic behind that, assume the following example
cgroup hierarchy.
A - B - C
\ D - E
If only B has a program attached, it will be effective for B, C, D
and E. If D then attaches a program itself, that will be effective for
both D and E, and the program in B will only affect B and C. Only one
program of a given type is effective for a cgroup.
Attaching and detaching programs will be done through the bpf(2)
syscall. For now, ingress and egress inet socket filtering are the
only supported use-cases.
Signed-off-by: Daniel Mack <daniel@zonque.org>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
All conflicts were simple overlapping changes except perhaps
for the Thunder driver.
That driver has a change_mtu method explicitly for sending
a message to the hardware. If that fails it returns an
error.
Normally a driver doesn't need an ndo_change_mtu method becuase those
are usually just range changes, which are now handled generically.
But since this extra operation is needed in the Thunder driver, it has
to stay.
However, if the message send fails we have to restore the original
MTU before the change because the entire call chain expects that if
an error is thrown by ndo_change_mtu then the MTU did not change.
Therefore code is added to nicvf_change_mtu to remember the original
MTU, and to restore it upon nicvf_update_hw_max_frs() failue.
Signed-off-by: David S. Miller <davem@davemloft.net>
In mlx5e_create_rq(), when creating a new queue, we call bpf_prog_add() but
without checking the return value. bpf_prog_add() can fail since 92117d8443
("bpf: fix refcnt overflow"), so we really must check it. Take the reference
right when we assign it to the rq from priv->xdp_prog, and just drop the
reference on error path. Destruction in mlx5e_destroy_rq() looks good, though.
Fixes: 86994156c7 ("net/mlx5e: XDP fast RX drop bpf programs support")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
I made some invalid assumptions with BPF_AND and BPF_MOD that could result in
invalid accesses to bpf map entries. Fix this up by doing a few things
1) Kill BPF_MOD support. This doesn't actually get used by the compiler in real
life and just adds extra complexity.
2) Fix the logic for BPF_AND, don't allow AND of negative numbers and set the
minimum value to 0 for positive AND's.
3) Don't do operations on the ranges if they are set to the limits, as they are
by definition undefined, and allowing arithmetic operations on those values
could make them appear valid when they really aren't.
This fixes the testcase provided by Jann as well as a few other theoretical
problems.
Reported-by: Jann Horn <jannh@google.com>
Signed-off-by: Josef Bacik <jbacik@fb.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
gcc-6.2.1 gives the following warning:
kernel/bpf/bpf_lru_list.c: In function ‘__bpf_lru_list_rotate_inactive.isra.3’:
kernel/bpf/bpf_lru_list.c:201:28: warning: ‘next’ may be used uninitialized in this function [-Wmaybe-uninitialized]
The "next" is currently initialized in the while() loop which must have >=1
iterations.
This patch initializes next to get rid of the compiler warning.
Fixes: 3a08c2fd76 ("bpf: LRU List")
Reported-by: David Miller <davem@davemloft.net>
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Provide a LRU version of the existing BPF_MAP_TYPE_PERCPU_HASH
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Provide a LRU version of the existing BPF_MAP_TYPE_HASH.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Refactor the codes that populate the value
of a htab_elem in a BPF_MAP_TYPE_PERCPU_HASH
typed bpf_map.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Instead of having a common LRU list, this patch allows a
percpu LRU list which can be selected by specifying a map
attribute. The map attribute will be added in the later
patch.
While the common use case for LRU is #reads >> #updates,
percpu LRU list allows bpf prog to absorb unusual #updates
under pathological case (e.g. external traffic facing machine which
could be under attack).
Each percpu LRU is isolated from each other. The LRU nodes (including
free nodes) cannot be moved across different LRU Lists.
Here are the update performance comparison between
common LRU list and percpu LRU list (the test code is
at the last patch):
[root@kerneltest003.31.prn1 ~]# for i in 1 4 8; do echo -n "$i cpus: "; \
./map_perf_test 16 $i | awk '{r += $3}END{print r " updates"}'; done
1 cpus: 2934082 updates
4 cpus: 7391434 updates
8 cpus: 6500576 updates
[root@kerneltest003.31.prn1 ~]# for i in 1 4 8; do echo -n "$i cpus: "; \
./map_perf_test 32 $i | awk '{r += $3}END{printr " updates"}'; done
1 cpus: 2896553 updates
4 cpus: 9766395 updates
8 cpus: 17460553 updates
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Introduce bpf_lru_list which will provide LRU capability to
the bpf_htab in the later patch.
* General Thoughts:
1. Target use case. Read is more often than update.
(i.e. bpf_lookup_elem() is more often than bpf_update_elem()).
If bpf_prog does a bpf_lookup_elem() first and then an in-place
update, it still counts as a read operation to the LRU list concern.
2. It may be useful to think of it as a LRU cache
3. Optimize the read case
3.1 No lock in read case
3.2 The LRU maintenance is only done during bpf_update_elem()
4. If there is a percpu LRU list, it will lose the system-wise LRU
property. A completely isolated percpu LRU list has the best
performance but the memory utilization is not ideal considering
the work load may be imbalance.
5. Hence, this patch starts the LRU implementation with a global LRU
list with batched operations before accessing the global LRU list.
As a LRU cache, #read >> #update/#insert operations, it will work well.
6. There is a local list (for each cpu) which is named
'struct bpf_lru_locallist'. This local list is not used to sort
the LRU property. Instead, the local list is to batch enough
operations before acquiring the lock of the global LRU list. More
details on this later.
7. In the later patch, it allows a percpu LRU list by specifying a
map-attribute for scalability reason and for use cases that need to
prepare for the worst (and pathological) case like DoS attack.
The percpu LRU list is completely isolated from each other and the
LRU nodes (including free nodes) cannot be moved across the list. The
following description is for the global LRU list but mostly applicable
to the percpu LRU list also.
* Global LRU List:
1. It has three sub-lists: active-list, inactive-list and free-list.
2. The two list idea, active and inactive, is borrowed from the
page cache.
3. All nodes are pre-allocated and all sit at the free-list (of the
global LRU list) at the beginning. The pre-allocation reasoning
is similar to the existing BPF_MAP_TYPE_HASH. However,
opting-out prealloc (BPF_F_NO_PREALLOC) is not supported in
the LRU map.
* Active/Inactive List (of the global LRU list):
1. The active list, as its name says it, maintains the active set of
the nodes. We can think of it as the working set or more frequently
accessed nodes. The access frequency is approximated by a ref-bit.
The ref-bit is set during the bpf_lookup_elem().
2. The inactive list, as its name also says it, maintains a less
active set of nodes. They are the candidates to be removed
from the bpf_htab when we are running out of free nodes.
3. The ordering of these two lists is acting as a rough clock.
The tail of the inactive list is the older nodes and
should be released first if the bpf_htab needs free element.
* Rotating the Active/Inactive List (of the global LRU list):
1. It is the basic operation to maintain the LRU property of
the global list.
2. The active list is only rotated when the inactive list is running
low. This idea is similar to the current page cache.
Inactive running low is currently defined as
"# of inactive < # of active".
3. The active list rotation always starts from the tail. It moves
node without ref-bit set to the head of the inactive list.
It moves node with ref-bit set back to the head of the active
list and then clears its ref-bit.
4. The inactive rotation is pretty simply.
It walks the inactive list and moves the nodes back to the head of
active list if its ref-bit is set. The ref-bit is cleared after moving
to the active list.
If the node does not have ref-bit set, it just leave it as it is
because it is already in the inactive list.
* Shrinking the Inactive List (of the global LRU list):
1. Shrinking is the operation to get free nodes when the bpf_htab is
full.
2. It usually only shrinks the inactive list to get free nodes.
3. During shrinking, it will walk the inactive list from the tail,
delete the nodes without ref-bit set from bpf_htab.
4. If no free node found after step (3), it will forcefully get
one node from the tail of inactive or active list. Forcefully is
in the sense that it ignores the ref-bit.
* Local List:
1. Each CPU has a 'struct bpf_lru_locallist'. The purpose is to
batch enough operations before acquiring the lock of the
global LRU.
2. A local list has two sub-lists, free-list and pending-list.
3. During bpf_update_elem(), it will try to get from the free-list
of (the current CPU local list).
4. If the local free-list is empty, it will acquire from the
global LRU list. The global LRU list can either satisfy it
by its global free-list or by shrinking the global inactive
list. Since we have acquired the global LRU list lock,
it will try to get at most LOCAL_FREE_TARGET elements
to the local free list.
5. When a new element is added to the bpf_htab, it will
first sit at the pending-list (of the local list) first.
The pending-list will be flushed to the global LRU list
when it needs to acquire free nodes from the global list
next time.
* Lock Consideration:
The LRU list has a lock (lru_lock). Each bucket of htab has a
lock (buck_lock). If both locks need to be acquired together,
the lock order is always lru_lock -> buck_lock and this only
happens in the bpf_lru_list.c logic.
In hashtab.c, both locks are not acquired together (i.e. one
lock is always released first before acquiring another lock).
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
