====================
BPF token support in libbpf's BPF object
Add fuller support for BPF token in high-level BPF object APIs. This is the
most frequently used way to work with BPF using libbpf, so supporting BPF
token there is critical.
Patch #1 is improving kernel-side BPF_TOKEN_CREATE behavior by rejecting to
create "empty" BPF token with no delegation. This seems like saner behavior
which also makes libbpf's caching better overall. If we ever want to create
BPF token with no delegate_xxx options set on BPF FS, we can use a new flag to
enable that.
Patches #2-#5 refactor libbpf internals, mostly feature detection code, to
prepare it from BPF token FD.
Patch #6 adds options to pass BPF token into BPF object open options. It also
adds implicit BPF token creation logic to BPF object load step, even without
any explicit involvement of the user. If the environment is setup properly,
BPF token will be created transparently and used implicitly. This allows for
all existing application to gain BPF token support by just linking with
latest version of libbpf library. No source code modifications are required.
All that under assumption that privileged container management agent properly
set up default BPF FS instance at /sys/bpf/fs to allow BPF token creation.
Patches #7-#8 adds more selftests, validating BPF object APIs work as expected
under unprivileged user namespaced conditions in the presence of BPF token.
Patch #9 extends libbpf with LIBBPF_BPF_TOKEN_PATH envvar knowledge, which can
be used to override custom BPF FS location used for implicit BPF token
creation logic without needing to adjust application code. This allows admins
or container managers to mount BPF token-enabled BPF FS at non-standard
location without the need to coordinate with applications.
LIBBPF_BPF_TOKEN_PATH can also be used to disable BPF token implicit creation
by setting it to an empty value. Patch #10 tests this new envvar functionality.
v2->v3:
- move some stray feature cache refactorings into patch #4 (Alexei);
- add LIBBPF_BPF_TOKEN_PATH envvar support (Alexei);
v1->v2:
- remove minor code redundancies (Eduard, John);
- add acks and rebase.
====================
Andrii Nakryiko [Wed, 13 Dec 2023 19:08:42 +0000 (11:08 -0800)]
selftests/bpf: add tests for LIBBPF_BPF_TOKEN_PATH envvar
Add new subtest validating LIBBPF_BPF_TOKEN_PATH envvar semantics.
Extend existing test to validate that LIBBPF_BPF_TOKEN_PATH allows to
disable implicit BPF token creation by setting envvar to empty string.
Andrii Nakryiko [Wed, 13 Dec 2023 19:08:41 +0000 (11:08 -0800)]
libbpf: support BPF token path setting through LIBBPF_BPF_TOKEN_PATH envvar
To allow external admin authority to override default BPF FS location
(/sys/fs/bpf) for implicit BPF token creation, teach libbpf to recognize
LIBBPF_BPF_TOKEN_PATH envvar. If it is specified and user application
didn't explicitly specify neither bpf_token_path nor bpf_token_fd
option, it will be treated exactly like bpf_token_path option,
overriding default /sys/fs/bpf location and making BPF token mandatory.
Andrii Nakryiko [Wed, 13 Dec 2023 19:08:40 +0000 (11:08 -0800)]
selftests/bpf: add tests for BPF object load with implicit token
Add a test to validate libbpf's implicit BPF token creation from default
BPF FS location (/sys/fs/bpf). Also validate that disabling this
implicit BPF token creation works.
Andrii Nakryiko [Wed, 13 Dec 2023 19:08:39 +0000 (11:08 -0800)]
selftests/bpf: add BPF object loading tests with explicit token passing
Add a few tests that attempt to load BPF object containing privileged
map, program, and the one requiring mandatory BTF uploading into the
kernel (to validate token FD propagation to BPF_BTF_LOAD command).
Andrii Nakryiko [Wed, 13 Dec 2023 19:08:38 +0000 (11:08 -0800)]
libbpf: wire up BPF token support at BPF object level
Add BPF token support to BPF object-level functionality.
BPF token is supported by BPF object logic either as an explicitly
provided BPF token from outside (through BPF FS path or explicit BPF
token FD), or implicitly (unless prevented through
bpf_object_open_opts).
Implicit mode is assumed to be the most common one for user namespaced
unprivileged workloads. The assumption is that privileged container
manager sets up default BPF FS mount point at /sys/fs/bpf with BPF token
delegation options (delegate_{cmds,maps,progs,attachs} mount options).
BPF object during loading will attempt to create BPF token from
/sys/fs/bpf location, and pass it for all relevant operations
(currently, map creation, BTF load, and program load).
In this implicit mode, if BPF token creation fails due to whatever
reason (BPF FS is not mounted, or kernel doesn't support BPF token,
etc), this is not considered an error. BPF object loading sequence will
proceed with no BPF token.
In explicit BPF token mode, user provides explicitly either custom BPF
FS mount point path or creates BPF token on their own and just passes
token FD directly. In such case, BPF object will either dup() token FD
(to not require caller to hold onto it for entire duration of BPF object
lifetime) or will attempt to create BPF token from provided BPF FS
location. If BPF token creation fails, that is considered a critical
error and BPF object load fails with an error.
Libbpf provides a way to disable implicit BPF token creation, if it
causes any troubles (BPF token is designed to be completely optional and
shouldn't cause any problems even if provided, but in the world of BPF
LSM, custom security logic can be installed that might change outcome
dependin on the presence of BPF token). To disable libbpf's default BPF
token creation behavior user should provide either invalid BPF token FD
(negative), or empty bpf_token_path option.
BPF token presence can influence libbpf's feature probing, so if BPF
object has associated BPF token, feature probing is instructed to use
BPF object-specific feature detection cache and token FD.
Andrii Nakryiko [Wed, 13 Dec 2023 19:08:37 +0000 (11:08 -0800)]
libbpf: wire up token_fd into feature probing logic
Adjust feature probing callbacks to take into account optional token_fd.
In unprivileged contexts, some feature detectors would fail to detect
kernel support just because BPF program, BPF map, or BTF object can't be
loaded due to privileged nature of those operations. So when BPF object
is loaded with BPF token, this token should be used for feature probing.
This patch is setting support for this scenario, but we don't yet pass
non-zero token FD. This will be added in the next patch.
We also switched BPF cookie detector from using kprobe program to
tracepoint one, as tracepoint is somewhat less dangerous BPF program
type and has higher likelihood of being allowed through BPF token in the
future. This change has no effect on detection behavior.
Andrii Nakryiko [Wed, 13 Dec 2023 19:08:35 +0000 (11:08 -0800)]
libbpf: further decouple feature checking logic from bpf_object
Add feat_supported() helper that accepts feature cache instead of
bpf_object. This allows low-level code in bpf.c to not know or care
about higher-level concept of bpf_object, yet it will be able to utilize
custom feature checking in cases where BPF token might influence the
outcome.
Andrii Nakryiko [Wed, 13 Dec 2023 19:08:34 +0000 (11:08 -0800)]
libbpf: split feature detectors definitions from cached results
Split a list of supported feature detectors with their corresponding
callbacks from actual cached supported/missing values. This will allow
to have more flexible per-token or per-object feature detectors in
subsequent refactorings.
Andrii Nakryiko [Wed, 13 Dec 2023 19:08:33 +0000 (11:08 -0800)]
bpf: fail BPF_TOKEN_CREATE if no delegation option was set on BPF FS
It's quite confusing in practice when it's possible to successfully
create a BPF token from BPF FS that didn't have any of delegate_xxx
mount options set up. While it's not wrong, it's actually more
meaningful to reject BPF_TOKEN_CREATE with specific error code (-ENOENT)
to let user-space know that no token delegation is setup up.
So, instead of creating empty BPF token that will be always ignored
because it doesn't have any of the allow_xxx bits set, reject it with
-ENOENT. If we ever need empty BPF token to be possible, we can support
that with extra flag passed into BPF_TOKEN_CREATE.
Daniel Xu [Mon, 11 Dec 2023 20:20:09 +0000 (13:20 -0700)]
bpf: selftests: Add verifier tests for CO-RE bitfield writes
Add some tests that exercise BPF_CORE_WRITE_BITFIELD() macro. Since some
non-trivial bit fiddling is going on, make sure various edge cases (such
as adjacent bitfields and bitfields at the edge of structs) are
exercised.
Daniel Xu [Mon, 11 Dec 2023 20:20:08 +0000 (13:20 -0700)]
bpf: selftests: test_loader: Support __btf_path() annotation
This commit adds support for per-prog btf_custom_path. This is necessary
for testing CO-RE relocations on non-vmlinux types using test_loader
infrastructure.
Daniel Xu [Mon, 11 Dec 2023 20:20:07 +0000 (13:20 -0700)]
libbpf: Add BPF_CORE_WRITE_BITFIELD() macro
=== Motivation ===
Similar to reading from CO-RE bitfields, we need a CO-RE aware bitfield
writing wrapper to make the verifier happy.
Two alternatives to this approach are:
1. Use the upcoming `preserve_static_offset` [0] attribute to disable
CO-RE on specific structs.
2. Use broader byte-sized writes to write to bitfields.
(1) is a bit hard to use. It requires specific and not-very-obvious
annotations to bpftool generated vmlinux.h. It's also not generally
available in released LLVM versions yet.
(2) makes the code quite hard to read and write. And especially if
BPF_CORE_READ_BITFIELD() is already being used, it makes more sense to
to have an inverse helper for writing.
=== Implementation details ===
Since the logic is a bit non-obvious, I thought it would be helpful
to explain exactly what's going on.
To start, it helps by explaining what LSHIFT_U64 (lshift) and RSHIFT_U64
(rshift) is designed to mean. Consider the core of the
BPF_CORE_READ_BITFIELD() algorithm:
val <<= __CORE_RELO(s, field, LSHIFT_U64);
val = val >> __CORE_RELO(s, field, RSHIFT_U64);
Basically what happens is we lshift to clear the non-relevant (blank)
higher order bits. Then we rshift to bring the relevant bits (bitfield)
down to LSB position (while also clearing blank lower order bits). To
illustrate:
1. Compute a mask where the set bits are the bitfield bits. The first
left shift zeros out exactly the number of blank bits, leaving a
bitfield sized set of 1s. The subsequent right shift inserts the
correct amount of higher order blank bits.
2. On the left of the `|`, mask out the bitfield bits. This creates
0s where the new bitfield bits will go. On the right of the `|`,
bring nval into the correct bit position and mask out any bits
that fall outside of the bitfield. Finally, by bor'ing the two
halves, we get the final set of bits to write back.
Jie Jiang [Tue, 12 Dec 2023 09:39:23 +0000 (09:39 +0000)]
bpf: Support uid and gid when mounting bpffs
Parse uid and gid in bpf_parse_param() so that they can be passed in as
the `data` parameter when mount() bpffs. This will be useful when we
want to control which user/group has the control to the mounted bpffs,
otherwise a separate chown() call will be needed.
Andrei Matei [Sun, 10 Dec 2023 22:51:50 +0000 (17:51 -0500)]
bpf: Comment on check_mem_size_reg
This patch adds a comment to check_mem_size_reg -- a function whose
meaning is not very transparent. The function implicitly deals with two
registers connected by convention, which is not obvious.
The function are defined in the verifier.c file, but not called
elsewhere, so delete the unused function.
kernel/bpf/verifier.c:3448:20: warning: unused function 'bt_set_slot'
kernel/bpf/verifier.c:3453:20: warning: unused function 'bt_clear_slot'
kernel/bpf/verifier.c:3488:20: warning: unused function 'bt_is_slot_set'
Manu Bretelle [Mon, 11 Dec 2023 18:07:33 +0000 (10:07 -0800)]
selftests/bpf: Fixes tests for filesystem kfuncs
`fs_kfuncs.c`'s `test_xattr` would fail the test even when the
filesystem did not support xattr, for instance when /tmp is mounted as
tmpfs.
This change checks errno when setxattr fail. If the failure is due to
the operation being unsupported, we will skip the test (just like we
would if verity was not enabled on the FS.
Before the change, fs_kfuncs test would fail in test_axattr:
Andrii Nakryiko [Mon, 4 Dec 2023 23:39:21 +0000 (15:39 -0800)]
bpf: tidy up exception callback management a bit
Use the fact that we are passing subprog index around and have
a corresponding struct bpf_subprog_info in bpf_verifier_env for each
subprogram. We don't need to separately pass around a flag whether
subprog is exception callback or not, each relevant verifier function
can determine this using provided subprog index if we maintain
bpf_subprog_info properly.
Also move out exception callback-specific logic from
btf_prepare_func_args(), keeping it generic. We can enforce all these
restriction right before exception callback verification pass. We add
out parameter, arg_cnt, for now, but this will be unnecessary with
subsequent refactoring and will be removed.
Andrii Nakryiko [Mon, 11 Dec 2023 17:41:31 +0000 (09:41 -0800)]
selftests/bpf: validate eliminated global subprog is not freplaceable
Add selftest that establishes dead code-eliminated valid global subprog
(global_dead) and makes sure that it's not possible to freplace it, as
it's effectively not there. This test will fail with unexpected success
before 2afae08c9dcb ("bpf: Validate global subprogs lazily").
v2->v3:
- add missing err assignment (Alan);
- undo unnecessary signature changes in verifier_global_subprogs.c (Eduard);
v1->v2:
- don't rely on assembly output in verifier log, which changes between
compiler versions (CI).
32 bytes may be not enough for some custom metadata. Relax the restriction,
allow metadata larger than 32 bytes and make __skb_metadata_differs() work
with bigger lengths.
Now size of metadata is only limited by the fact it is stored as u8 in
skb_shared_info, so maximum possible value is 255. Size still has to be
aligned to 4, so the actual upper limit becomes 252. Most driver
implementations will offer less, none can offer more.
Other important conditions, such as having enough space for xdp_frame
building, are already checked in bpf_xdp_adjust_meta().
Larysa Zaremba [Wed, 6 Dec 2023 20:59:18 +0000 (21:59 +0100)]
selftests/bpf: Increase invalid metadata size
Changed check expects passed data meta to be deemed invalid. After loosening
the requirement, the size of 36 bytes becomes valid. Therefore, increase
tested meta size to 256, so we do not get an unexpected success.
====================
Add new bpf_cpumask_weight() kfunc
It can be useful to query how many bits are set in a cpumask. For
example, if you want to perform special logic for the last remaining
core that's set in a mask. This logic is already exposed through the
main kernel's cpumask header as cpumask_weight(), so it would be useful
to add a new bpf_cpumask_weight() kfunc which wraps it and does the
same.
This patch series was built and tested on top of commit 2146f7fe6e02
("Merge branch 'allocate-bpf-trampoline-on-bpf_prog_pack'").
====================
David Vernet [Thu, 7 Dec 2023 21:08:43 +0000 (15:08 -0600)]
selftests/bpf: Add test for bpf_cpumask_weight() kfunc
The new bpf_cpumask_weight() kfunc can be used to count the number of
bits that are set in a struct cpumask* kptr. Let's add a selftest to
verify its behavior.
David Vernet [Thu, 7 Dec 2023 21:08:42 +0000 (15:08 -0600)]
bpf: Add bpf_cpumask_weight() kfunc
It can be useful to query how many bits are set in a cpumask. For
example, if you want to perform special logic for the last remaining
core that's set in a mask. Let's therefore add a new
bpf_cpumask_weight() kfunc which checks how many bits are set in a mask.
Tiezhu Yang [Thu, 7 Dec 2023 04:08:51 +0000 (12:08 +0800)]
test_bpf: Rename second ALU64_SMOD_X to ALU64_SMOD_K
Currently, there are two test cases with same name
"ALU64_SMOD_X: -7 % 2 = -1", the first one is right,
the second one should be ALU64_SMOD_K because its
code is BPF_ALU64 | BPF_MOD | BPF_K.
Add two tests validating that verifier's precision backtracking logic
handles BPF_ST_MEM instructions that produce fake register spill into
register slot. This is happening when non-zero constant is written
directly to a slot, e.g., *(u64 *)(r10 -8) = 123.
Add both full 64-bit register spill, as well as 32-bit "sub-spill".
Andrii Nakryiko [Sat, 9 Dec 2023 01:09:57 +0000 (17:09 -0800)]
bpf: handle fake register spill to stack with BPF_ST_MEM instruction
When verifier validates BPF_ST_MEM instruction that stores known
constant to stack (e.g., *(u64 *)(r10 - 8) = 123), it effectively spills
a fake register with a constant (but initially imprecise) value to
a stack slot. Because read-side logic treats it as a proper register
fill from stack slot, we need to mark such stack slot initialization as
INSN_F_STACK_ACCESS instruction to stop precision backtracking from
missing it.
====================
The patch set aims to fix the problems found when inspecting the code
related with maybe_wait_bpf_programs().
Patch #1 removes unnecessary invocation of maybe_wait_bpf_programs().
Patch #2 calls maybe_wait_bpf_programs() only once for batched update.
Patch #3 adds the missed waiting when doing batched lookup_deletion on
htab of maps. Patch #4 does wait only if the update or deletion
operation succeeds. Patch #5 fixes the value of batch.count when memory
allocation fails.
====================
Hou Tao [Fri, 8 Dec 2023 10:23:53 +0000 (18:23 +0800)]
bpf: Set uattr->batch.count as zero before batched update or deletion
generic_map_{delete,update}_batch() doesn't set uattr->batch.count as
zero before it tries to allocate memory for key. If the memory
allocation fails, the value of uattr->batch.count will be incorrect.
Fix it by setting uattr->batch.count as zero beore batched update or
deletion.
Hou Tao [Fri, 8 Dec 2023 10:23:52 +0000 (18:23 +0800)]
bpf: Only call maybe_wait_bpf_programs() when map operation succeeds
There is no need to call maybe_wait_bpf_programs() if update or deletion
operation fails. So only call maybe_wait_bpf_programs() if update or
deletion operation succeeds.
Hou Tao [Fri, 8 Dec 2023 10:23:51 +0000 (18:23 +0800)]
bpf: Add missed maybe_wait_bpf_programs() for htab of maps
When doing batched lookup and deletion operations on htab of maps,
maybe_wait_bpf_programs() is needed to ensure all programs don't use the
inner map after the bpf syscall returns.
Instead of adding the wait in __htab_map_lookup_and_delete_batch(),
adding the wait in bpf_map_do_batch() and also removing the calling of
maybe_wait_bpf_programs() from generic_map_{delete,update}_batch().
Hou Tao [Fri, 8 Dec 2023 10:23:50 +0000 (18:23 +0800)]
bpf: Call maybe_wait_bpf_programs() only once for generic_map_update_batch()
Just like commit 9087c6ff8dfe ("bpf: Call maybe_wait_bpf_programs() only
once from generic_map_delete_batch()"), there is also no need to call
maybe_wait_bpf_programs() for each update in batched update, so only
call it once in generic_map_update_batch().
Hou Tao [Fri, 8 Dec 2023 10:23:49 +0000 (18:23 +0800)]
bpf: Remove unnecessary wait from bpf_map_copy_value()
Both map_lookup_elem() and generic_map_lookup_batch() use
bpf_map_copy_value() to lookup and copy the value, and there is no
update operation in bpf_map_copy_value(), so just remove the invocation
of maybe_wait_bpf_programs() from it.
libbpf: Add pr_warn() for EINVAL cases in linker_sanity_check_elf
Before the change on `i686-linux` `systemd` build failed as:
$ bpftool gen object src/core/bpf/socket_bind/socket-bind.bpf.o src/core/bpf/socket_bind/socket-bind.bpf.unstripped.o
Error: failed to link 'src/core/bpf/socket_bind/socket-bind.bpf.unstripped.o': Invalid argument (22)
After the change it fails as:
$ bpftool gen object src/core/bpf/socket_bind/socket-bind.bpf.o src/core/bpf/socket_bind/socket-bind.bpf.unstripped.o
libbpf: ELF section #9 has inconsistent alignment addr=8 != d=4 in src/core/bpf/socket_bind/socket-bind.bpf.unstripped.o
Error: failed to link 'src/core/bpf/socket_bind/socket-bind.bpf.unstripped.o': Invalid argument (22)
Now it's slightly easier to figure out what is wrong with an ELF file.
Merge branch 'bpf: Expand bpf_cgrp_storage to support cgroup1 non-attach case'
Yafang Shao says:
====================
In the current cgroup1 environment, associating operations between a cgroup
and applications in a BPF program requires storing a mapping of cgroup_id
to application either in a hash map or maintaining it in userspace.
However, by enabling bpf_cgrp_storage for cgroup1, it becomes possible to
conveniently store application-specific information in cgroup-local storage
and utilize it within BPF programs. Furthermore, enabling this feature for
cgroup1 involves minor modifications for the non-attach case, streamlining
the process.
However, when it comes to enabling this functionality for the cgroup1
attach case, it presents challenges. Therefore, the decision is to focus on
enabling it solely for the cgroup1 non-attach case at present. If
attempting to attach to a cgroup1 fd, the operation will simply fail with
the error code -EBADF.
Yafang Shao [Wed, 6 Dec 2023 11:53:24 +0000 (11:53 +0000)]
bpf: Enable bpf_cgrp_storage for cgroup1 non-attach case
In the current cgroup1 environment, associating operations between cgroups
and applications in a BPF program requires storing a mapping of cgroup_id
to application either in a hash map or maintaining it in userspace.
However, by enabling bpf_cgrp_storage for cgroup1, it becomes possible to
conveniently store application-specific information in cgroup-local storage
and utilize it within BPF programs. Furthermore, enabling this feature for
cgroup1 involves minor modifications for the non-attach case, streamlining
the process.
However, when it comes to enabling this functionality for the cgroup1
attach case, it presents challenges. Therefore, the decision is to focus on
enabling it solely for the cgroup1 non-attach case at present. If
attempting to attach to a cgroup1 fd, the operation will simply fail with
the error code -EBADF.
Andrii Nakryiko [Fri, 8 Dec 2023 23:30:28 +0000 (15:30 -0800)]
selftests/bpf: fix timer/test_bad_ret subtest on test_progs-cpuv4 flavor
Because test_bad_ret main program is not written in assembly, we don't
control instruction indices in timer_cb_ret_bad() subprog. This bites us
in timer/test_bad_ret subtest, where we see difference between cpuv4 and
other flavors.
For now, make __msg() expectations not rely on instruction indices by
anchoring them around bpf_get_prandom_u32 call. Once we have regex/glob
support for __msg(), this can be expressed a bit more nicely, but for
now just mitigating the problem with available means.
====================
bpf: fix accesses to uninit stack slots
Fix two related issues issues around verifying stack accesses:
1. accesses to uninitialized stack memory was allowed inconsistently
2. the maximum stack depth needed for a program was not always
maintained correctly
The two issues are fixed together in one commit because the code for one
affects the other.
V4 to V5:
- target bpf-next (Alexei)
V3 to V4:
- minor fixup to comment in patch 1 (Eduard)
- C89-style in patch 3 (Andrii)
V2 to V3:
- address review comments from Andrii and Eduard
- drop new verifier tests in favor of editing existing tests to check
for stack depth
- append a patch with a bit of cleanup coming out of the previous review
====================
Andrei Matei [Fri, 8 Dec 2023 03:25:19 +0000 (22:25 -0500)]
bpf: Minor cleanup around stack bounds
Push the rounding up of stack offsets into the function responsible for
growing the stack, rather than relying on all the callers to do it.
Uncertainty about whether the callers did it or not tripped up people in
a previous review.
Andrei Matei [Fri, 8 Dec 2023 03:25:18 +0000 (22:25 -0500)]
bpf: Fix accesses to uninit stack slots
Privileged programs are supposed to be able to read uninitialized stack
memory (ever since 6715df8d5) but, before this patch, these accesses
were permitted inconsistently. In particular, accesses were permitted
above state->allocated_stack, but not below it. In other words, if the
stack was already "large enough", the access was permitted, but
otherwise the access was rejected instead of being allowed to "grow the
stack". This undesired rejection was happening in two places:
- in check_stack_slot_within_bounds()
- in check_stack_range_initialized()
This patch arranges for these accesses to be permitted. A bunch of tests
that were relying on the old rejection had to change; all of them were
changed to add also run unprivileged, in which case the old behavior
persists. One tests couldn't be updated - global_func16 - because it
can't run unprivileged for other reasons.
This patch also fixes the tracking of the stack size for variable-offset
reads. This second fix is bundled in the same commit as the first one
because they're inter-related. Before this patch, writes to the stack
using registers containing a variable offset (as opposed to registers
with fixed, known values) were not properly contributing to the
function's needed stack size. As a result, it was possible for a program
to verify, but then to attempt to read out-of-bounds data at runtime
because a too small stack had been allocated for it.
Each function tracks the size of the stack it needs in
bpf_subprog_info.stack_depth, which is maintained by
update_stack_depth(). For regular memory accesses, check_mem_access()
was calling update_state_depth() but it was passing in only the fixed
part of the offset register, ignoring the variable offset. This was
incorrect; the minimum possible value of that register should be used
instead.
This tracking is now fixed by centralizing the tracking of stack size in
grow_stack_state(), and by lifting the calls to grow_stack_state() to
check_stack_access_within_bounds() as suggested by Andrii. The code is
now simpler and more convincingly tracks the correct maximum stack size.
check_stack_range_initialized() can now rely on enough stack having been
allocated for the access; this helps with the fix for the first issue.
A few tests were changed to also check the stack depth computation. The
one that fails without this patch is verifier_var_off:stack_write_priv_vs_unpriv.
Andrei Matei [Fri, 8 Dec 2023 03:25:17 +0000 (22:25 -0500)]
bpf: Add some comments to stack representation
Add comments to the datastructure tracking the stack state, as the
mapping between each stack slot and where its state is stored is not
entirely obvious.
David Vernet [Fri, 8 Dec 2023 06:17:03 +0000 (00:17 -0600)]
bpf: Load vmlinux btf for any struct_ops map
In libbpf, when determining whether we need to load vmlinux btf, we're
currently (among other things) checking whether there is any struct_ops
program present in the object. This works for most realistic struct_ops
maps, as a struct_ops map is of course typically composed of one or more
struct_ops programs. However, that technically need not be the case. A
struct_ops interface could be defined which allows a map to be specified
which one or more non-prog fields, and which provides default behavior
if no struct_ops progs is actually provided otherwise. For sched_ext,
for example, you technically only need to specify the name of the
scheduler in the struct_ops map, with the core scheduler logic providing
default behavior if no prog is actually specified.
If we were to define and try to load such a struct_ops map, we would
crash in libbpf when initializing it as obj->btf_vmlinux will be NULL:
Reading symbols from minimal...
(gdb) r
Starting program: minimal_example
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/usr/lib/libthread_db.so.1".
Program received signal SIGSEGV, Segmentation fault.
0x000055555558308c in btf__type_cnt (btf=0x0) at btf.c:612
612 return btf->start_id + btf->nr_types;
(gdb) bt
type_name=0x5555555d99e3 "sched_ext_ops", kind=4) at btf.c:914
kind=4) at btf.c:942
type=0x7fffffffe558, type_id=0x7fffffffe548, ...
data_member=0x7fffffffe568) at libbpf.c:948
kern_btf=0x0) at libbpf.c:1017
at libbpf.c:8059
So as to account for such bare-bones struct_ops maps, let's update
obj_needs_vmlinux_btf() to also iterate over an obj's maps and check
whether any of them are struct_ops maps.
====================
bpf: fix verification of indirect var-off stack access
V4 to V5:
- split the test into a separate patch
V3 to V4:
- include a test per Eduard's request
- target bpf-next per Alexei's request (patches didn't change)
V2 to V3:
- simplify checks for max_off (don't call
check_stack_slot_within_bounds for it)
- append a commit to protect against overflow in the addition of the
register and the offset
V1 to V2:
- fix max_off calculation for access size = 0
====================
Andrei Matei [Thu, 7 Dec 2023 04:11:50 +0000 (23:11 -0500)]
bpf: Guard stack limits against 32bit overflow
This patch promotes the arithmetic around checking stack bounds to be
done in the 64-bit domain, instead of the current 32bit. The arithmetic
implies adding together a 64-bit register with a int offset. The
register was checked to be below 1<<29 when it was variable, but not
when it was fixed. The offset either comes from an instruction (in which
case it is 16 bit), from another register (in which case the caller
checked it to be below 1<<29 [1]), or from the size of an argument to a
kfunc (in which case it can be a u32 [2]). Between the register being
inconsistently checked to be below 1<<29, and the offset being up to an
u32, it appears that we were open to overflowing the `int`s which were
currently used for arithmetic.
Andrei Matei [Thu, 7 Dec 2023 04:11:48 +0000 (23:11 -0500)]
bpf: Fix verification of indirect var-off stack access
This patch fixes a bug around the verification of possibly-zero-sized
stack accesses. When the access was done through a var-offset stack
pointer, check_stack_access_within_bounds was incorrectly computing the
maximum-offset of a zero-sized read to be the same as the register's min
offset. Instead, we have to take in account the register's maximum
possible value. The patch also simplifies how the max offset is checked;
the check is now simpler than for min offset.
The bug was allowing accesses to erroneously pass the
check_stack_access_within_bounds() checks, only to later crash in
check_stack_range_initialized() when all the possibly-affected stack
slots are iterated (this time with a correct max offset).
check_stack_range_initialized() is relying on
check_stack_access_within_bounds() for its accesses to the
stack-tracking vector to be within bounds; in the case of zero-sized
accesses, we were essentially only verifying that the lowest possible
slot was within bounds. We would crash when the max-offset of the stack
pointer was >= 0 (which shouldn't pass verification, and hopefully is
not something anyone's code attempts to do in practice).
====================
Allocate bpf trampoline on bpf_prog_pack
This set enables allocating bpf trampoline from bpf_prog_pack on x86. The
majority of this work, however, is the refactoring of trampoline code.
This is needed because we need to handle 4 archs and 2 users (trampoline
and struct_ops).
1/7 through 6/7 refactors trampoline code. A few helpers are added.
7/7 finally let bpf trampoline on x86 use bpf_prog_pack.
Changes in v7:
1. Use kvmalloc for rw_image in x86/arch_prepare_bpf_trampoline. (Alexei)
2. Add comment to explain why we cannot use kvmalloc in
x86/arch_bpf_trampoline_size. (Alexei)
Changes in v6:
1. Rebase.
2. Add Acked-by and Tested-by from Jiri Olsa and Björn Töpel.
Changes in v5:
1. Adjust size of trampoline ksym. (Jiri)
2. Use "unsigned int size" arg in image management helpers.(Daniel)
Changes in v4:
1. Dropped 1/8 in v3, which is already merged in bpf-next.
2. Add Reviewed-by from Björn Töpel.
Changes in v3:
1. Fix bug in s390. (Thanks to Ilya Leoshkevich).
2. Fix build error in riscv. (kernel test robot).
Changes in v2:
1. Add missing changes in net/bpf/bpf_dummy_struct_ops.c.
2. Reduce one dry run in arch_prepare_bpf_trampoline. (Xu Kuohai)
3. Other small fixes.
====================
Song Liu [Wed, 6 Dec 2023 22:40:54 +0000 (14:40 -0800)]
x86, bpf: Use bpf_prog_pack for bpf trampoline
There are three major changes here:
1. Add arch_[alloc|free]_bpf_trampoline based on bpf_prog_pack;
2. Let arch_prepare_bpf_trampoline handle ROX input image, this requires
arch_prepare_bpf_trampoline allocating a temporary RW buffer;
3. Update __arch_prepare_bpf_trampoline() to handle a RW buffer (rw_image)
and a ROX buffer (image). This part is similar to the image/rw_image
logic in bpf_int_jit_compile().
Song Liu [Wed, 6 Dec 2023 22:40:53 +0000 (14:40 -0800)]
bpf: Use arch_bpf_trampoline_size
Instead of blindly allocating PAGE_SIZE for each trampoline, check the size
of the trampoline with arch_bpf_trampoline_size(). This size is saved in
bpf_tramp_image->size, and used for modmem charge/uncharge. The fallback
arch_alloc_bpf_trampoline() still allocates a whole page because we need to
use set_memory_* to protect the memory.
struct_ops trampoline still uses a whole page for multiple trampolines.
With this size check at caller (regular trampoline and struct_ops
trampoline), remove arch_bpf_trampoline_size() from
arch_prepare_bpf_trampoline() in archs.
Also, update bpf_image_ksym_add() to handle symbol of different sizes.
Song Liu [Wed, 6 Dec 2023 22:40:51 +0000 (14:40 -0800)]
bpf, x86: Adjust arch_prepare_bpf_trampoline return value
x86's implementation of arch_prepare_bpf_trampoline() requires
BPF_INSN_SAFETY buffer space between end of program and image_end. OTOH,
the return value does not include BPF_INSN_SAFETY. This doesn't cause any
real issue at the moment. However, "image" of size retval is not enough for
arch_prepare_bpf_trampoline(). This will cause confusion when we introduce
a new helper arch_bpf_trampoline_size(). To avoid future confusion, adjust
the return value to include BPF_INSN_SAFETY.
Song Liu [Wed, 6 Dec 2023 22:40:50 +0000 (14:40 -0800)]
bpf: Add helpers for trampoline image management
As BPF trampoline of different archs moves from bpf_jit_[alloc|free]_exec()
to bpf_prog_pack_[alloc|free](), we need to use different _alloc, _free for
different archs during the transition. Add the following helpers for this
transition:
void *arch_alloc_bpf_trampoline(unsigned int size);
void arch_free_bpf_trampoline(void *image, unsigned int size);
void arch_protect_bpf_trampoline(void *image, unsigned int size);
void arch_unprotect_bpf_trampoline(void *image, unsigned int size);
The fallback version of these helpers require size <= PAGE_SIZE, but they
are only called with size == PAGE_SIZE. They will be called with size <
PAGE_SIZE when arch_bpf_trampoline_size() helper is introduced later.
Song Liu [Wed, 6 Dec 2023 22:40:49 +0000 (14:40 -0800)]
bpf: Adjust argument names of arch_prepare_bpf_trampoline()
We are using "im" for "struct bpf_tramp_image" and "tr" for "struct
bpf_trampoline" in most of the code base. The only exception is the
prototype and fallback version of arch_prepare_bpf_trampoline(). Update
them to match the rest of the code base.
We mix "orig_call" and "func_addr" for the argument in different versions
of arch_prepare_bpf_trampoline(). s/orig_call/func_addr/g so they match.
Song Liu [Wed, 6 Dec 2023 22:40:48 +0000 (14:40 -0800)]
bpf: Let bpf_prog_pack_free handle any pointer
Currently, bpf_prog_pack_free only can only free pointer to struct
bpf_binary_header, which is not flexible. Add a size argument to
bpf_prog_pack_free so that it can handle any pointer.
Andrii Nakryiko [Wed, 6 Dec 2023 19:09:20 +0000 (11:09 -0800)]
bpf: rename MAX_BPF_LINK_TYPE into __MAX_BPF_LINK_TYPE for consistency
To stay consistent with the naming pattern used for similar cases in BPF
UAPI (__MAX_BPF_ATTACH_TYPE, etc), rename MAX_BPF_LINK_TYPE into
__MAX_BPF_LINK_TYPE.
Also similar to MAX_BPF_ATTACH_TYPE and MAX_BPF_REG, add:
#define MAX_BPF_LINK_TYPE __MAX_BPF_LINK_TYPE
Not all __MAX_xxx enums have such #define, so I'm not sure if we should
add it or not, but I figured I'll start with a completely backwards
compatible way, and we can drop that, if necessary.
Also adjust a selftest that used MAX_BPF_LINK_TYPE enum.
====================
BPF token and BPF FS-based delegation
This patch set introduces an ability to delegate a subset of BPF subsystem
functionality from privileged system-wide daemon (e.g., systemd or any other
container manager) through special mount options for userns-bound BPF FS to
a *trusted* unprivileged application. Trust is the key here. This
functionality is not about allowing unconditional unprivileged BPF usage.
Establishing trust, though, is completely up to the discretion of respective
privileged application that would create and mount a BPF FS instance with
delegation enabled, as different production setups can and do achieve it
through a combination of different means (signing, LSM, code reviews, etc),
and it's undesirable and infeasible for kernel to enforce any particular way
of validating trustworthiness of particular process.
The main motivation for this work is a desire to enable containerized BPF
applications to be used together with user namespaces. This is currently
impossible, as CAP_BPF, required for BPF subsystem usage, cannot be namespaced
or sandboxed, as a general rule. E.g., tracing BPF programs, thanks to BPF
helpers like bpf_probe_read_kernel() and bpf_probe_read_user() can safely read
arbitrary memory, and it's impossible to ensure that they only read memory of
processes belonging to any given namespace. This means that it's impossible to
have a mechanically verifiable namespace-aware CAP_BPF capability, and as such
another mechanism to allow safe usage of BPF functionality is necessary.BPF FS
delegation mount options and BPF token derived from such BPF FS instance is
such a mechanism. Kernel makes no assumption about what "trusted" constitutes
in any particular case, and it's up to specific privileged applications and
their surrounding infrastructure to decide that. What kernel provides is a set
of APIs to setup and mount special BPF FS instanecs and derive BPF tokens from
it. BPF FS and BPF token are both bound to its owning userns and in such a way
are constrained inside intended container. Users can then pass BPF token FD to
privileged bpf() syscall commands, like BPF map creation and BPF program
loading, to perform such operations without having init userns privileged.
This version incorporates feedback and suggestions ([3]) received on v3 of
this patch set, and instead of allowing to create BPF tokens directly assuming
capable(CAP_SYS_ADMIN), we instead enhance BPF FS to accept a few new
delegation mount options. If these options are used and BPF FS itself is
properly created, set up, and mounted inside the user namespaced container,
user application is able to derive a BPF token object from BPF FS instance,
and pass that token to bpf() syscall. As explained in patch #3, BPF token
itself doesn't grant access to BPF functionality, but instead allows kernel to
do namespaced capabilities checks (ns_capable() vs capable()) for CAP_BPF,
CAP_PERFMON, CAP_NET_ADMIN, and CAP_SYS_ADMIN, as applicable. So it forms one
half of a puzzle and allows container managers and sys admins to have safe and
flexible configuration options: determining which containers get delegation of
BPF functionality through BPF FS, and then which applications within such
containers are allowed to perform bpf() commands, based on namespaces
capabilities.
Previous attempt at addressing this very same problem ([0]) attempted to
utilize authoritative LSM approach, but was conclusively rejected by upstream
LSM maintainers. BPF token concept is not changing anything about LSM
approach, but can be combined with LSM hooks for very fine-grained security
policy. Some ideas about making BPF token more convenient to use with LSM (in
particular custom BPF LSM programs) was briefly described in recent LSF/MM/BPF
2023 presentation ([1]). E.g., an ability to specify user-provided data
(context), which in combination with BPF LSM would allow implementing a very
dynamic and fine-granular custom security policies on top of BPF token. In the
interest of minimizing API surface area and discussions this was relegated to
follow up patches, as it's not essential to the fundamental concept of
delegatable BPF token.
It should be noted that BPF token is conceptually quite similar to the idea of
/dev/bpf device file, proposed by Song a while ago ([2]). The biggest
difference is the idea of using virtual anon_inode file to hold BPF token and
allowing multiple independent instances of them, each (potentially) with its
own set of restrictions. And also, crucially, BPF token approach is not using
any special stateful task-scoped flags. Instead, bpf() syscall accepts
token_fd parameters explicitly for each relevant BPF command. This addresses
main concerns brought up during the /dev/bpf discussion, and fits better with
overall BPF subsystem design.
This patch set adds a basic minimum of functionality to make BPF token idea
useful and to discuss API and functionality. Currently only low-level libbpf
APIs support creating and passing BPF token around, allowing to test kernel
functionality, but for the most part is not sufficient for real-world
applications, which typically use high-level libbpf APIs based on `struct
bpf_object` type. This was done with the intent to limit the size of patch set
and concentrate on mostly kernel-side changes. All the necessary plumbing for
libbpf will be sent as a separate follow up patch set kernel support makes it
upstream.
Another part that should happen once kernel-side BPF token is established, is
a set of conventions between applications (e.g., systemd), tools (e.g.,
bpftool), and libraries (e.g., libbpf) on exposing delegatable BPF FS
instance(s) at well-defined locations to allow applications take advantage of
this in automatic fashion without explicit code changes on BPF application's
side. But I'd like to postpone this discussion to after BPF token concept
lands.
v11->v12:
- enforce exact userns match in bpf_token_capable() and
bpf_token_allow_cmd() checks, for added strictness (Christian);
v10->v11:
- fix BPF FS root check to disallow using bind-mounted subdirectory of BPF
FS instance (Christian);
- further restrict BPF_TOKEN_CREATE command to be executed from inside
exactly the same user namespace as the one used to create BPF FS instance
(Christian);
v9->v10:
- slight adjustments in LSM parts (Paul);
- setting delegate_xxx options require capable(CAP_SYS_ADMIN) (Christian);
- simplify BPF_TOKEN_CREATE UAPI by accepting BPF FS FD directly (Christian);
v8->v9:
- fix issue in selftests due to sys/mount.h header (Jiri);
- fix warning in doc comments in LSM hooks (kernel test robot);
v7->v8:
- add bpf_token_allow_cmd and bpf_token_capable hooks (Paul);
- inline bpf_token_alloc() into bpf_token_create() to prevent accidental
divergence with security_bpf_token_create() hook (Paul);
v6->v7:
- separate patches to refactor bpf_prog_alloc/bpf_map_alloc LSM hooks, as
discussed with Paul, and now they also accept struct bpf_token;
- added bpf_token_create/bpf_token_free to allow LSMs (SELinux,
specifically) to set up security LSM blob (Paul);
- last patch also wires bpf_security_struct setup by SELinux, similar to how
it's done for BPF map/prog, though I'm not sure if that's enough, so worst
case it's easy to drop this patch if more full fledged SELinux
implementation will be done separately;
- small fixes for issues caught by code reviews (Jiri, Hou);
- fix for test_maps test that doesn't use LIBBPF_OPTS() macro (CI);
v5->v6:
- fix possible use of uninitialized variable in selftests (CI);
- don't use anon_inode, instead create one from BPF FS instance (Christian);
- don't store bpf_token inside struct bpf_map, instead pass it explicitly to
map_check_btf(). We do store bpf_token inside prog->aux, because it's used
during verification and even can be checked during attach time for some
program types;
- LSM hooks are left intact pending the conclusion of discussion with Paul
Moore; I'd prefer to do LSM-related changes as a follow up patch set
anyways;
v4->v5:
- add pre-patch unifying CAP_NET_ADMIN handling inside kernel/bpf/syscall.c
(Paul Moore);
- fix build warnings and errors in selftests and kernel, detected by CI and
kernel test robot;
v3->v4:
- add delegation mount options to BPF FS;
- BPF token is derived from the instance of BPF FS and associates itself
with BPF FS' owning userns;
- BPF token doesn't grant BPF functionality directly, it just turns
capable() checks into ns_capable() checks within BPF FS' owning user;
- BPF token cannot be pinned;
v2->v3:
- make BPF_TOKEN_CREATE pin created BPF token in BPF FS, and disallow
BPF_OBJ_PIN for BPF token;
v1->v2:
- fix build failures on Kconfig with CONFIG_BPF_SYSCALL unset;
- drop BPF_F_TOKEN_UNKNOWN_* flags and simplify UAPI (Stanislav).
====================
Andrii Nakryiko [Thu, 30 Nov 2023 18:52:29 +0000 (10:52 -0800)]
bpf,selinux: allocate bpf_security_struct per BPF token
Utilize newly added bpf_token_create/bpf_token_free LSM hooks to
allocate struct bpf_security_struct for each BPF token object in
SELinux. This just follows similar pattern for BPF prog and map.
Andrii Nakryiko [Thu, 30 Nov 2023 18:52:28 +0000 (10:52 -0800)]
selftests/bpf: add BPF token-enabled tests
Add a selftest that attempts to conceptually replicate intended BPF
token use cases inside user namespaced container.
Child process is forked. It is then put into its own userns and mountns.
Child creates BPF FS context object. This ensures child userns is
captured as the owning userns for this instance of BPF FS. Given setting
delegation mount options is privileged operation, we ensure that child
cannot set them.
This context is passed back to privileged parent process through Unix
socket, where parent sets up delegation options, creates, and mounts it
as a detached mount. This mount FD is passed back to the child to be
used for BPF token creation, which allows otherwise privileged BPF
operations to succeed inside userns.
We validate that all of token-enabled privileged commands (BPF_BTF_LOAD,
BPF_MAP_CREATE, and BPF_PROG_LOAD) work as intended. They should only
succeed inside the userns if a) BPF token is provided with proper
allowed sets of commands and types; and b) namespaces CAP_BPF and other
privileges are set. Lacking a) or b) should lead to -EPERM failures.
Based on suggested workflow by Christian Brauner ([0]).
Andrii Nakryiko [Thu, 30 Nov 2023 18:52:26 +0000 (10:52 -0800)]
libbpf: add BPF token support to bpf_btf_load() API
Allow user to specify token_fd for bpf_btf_load() API that wraps
kernel's BPF_BTF_LOAD command. This allows loading BTF from unprivileged
process as long as it has BPF token allowing BPF_BTF_LOAD command, which
can be created and delegated by privileged process.
Andrii Nakryiko [Thu, 30 Nov 2023 18:52:23 +0000 (10:52 -0800)]
bpf,lsm: add BPF token LSM hooks
Wire up bpf_token_create and bpf_token_free LSM hooks, which allow to
allocate LSM security blob (we add `void *security` field to struct
bpf_token for that), but also control who can instantiate BPF token.
This follows existing pattern for BPF map and BPF prog.
Also add security_bpf_token_allow_cmd() and security_bpf_token_capable()
LSM hooks that allow LSM implementation to control and negate (if
necessary) BPF token's delegation of a specific bpf_cmd and capability,
respectively.
Similarly to bpf_prog_alloc LSM hook, rename and extend bpf_map_alloc
hook into bpf_map_create, taking not just struct bpf_map, but also
bpf_attr and bpf_token, to give a fuller context to LSMs.
Unlike bpf_prog_alloc, there is no need to move the hook around, as it
currently is firing right before allocating BPF map ID and FD, which
seems to be a sweet spot.
But like bpf_prog_alloc/bpf_prog_free combo, make sure that bpf_map_free
LSM hook is called even if bpf_map_create hook returned error, as if few
LSMs are combined together it could be that one LSM successfully
allocated security blob for its needs, while subsequent LSM rejected BPF
map creation. The former LSM would still need to free up LSM blob, so we
need to ensure security_bpf_map_free() is called regardless of the
outcome.
Based on upstream discussion ([0]), rework existing
bpf_prog_alloc_security LSM hook. Rename it to bpf_prog_load and instead
of passing bpf_prog_aux, pass proper bpf_prog pointer for a full BPF
program struct. Also, we pass bpf_attr union with all the user-provided
arguments for BPF_PROG_LOAD command. This will give LSMs as much
information as we can basically provide.
The hook is also BPF token-aware now, and optional bpf_token struct is
passed as a third argument. bpf_prog_load LSM hook is called after
a bunch of sanity checks were performed, bpf_prog and bpf_prog_aux were
allocated and filled out, but right before performing full-fledged BPF
verification step.
bpf_prog_free LSM hook is now accepting struct bpf_prog argument, for
consistency. SELinux code is adjusted to all new names, types, and
signatures.
Note, given that bpf_prog_load (previously bpf_prog_alloc) hook can be
used by some LSMs to allocate extra security blob, but also by other
LSMs to reject BPF program loading, we need to make sure that
bpf_prog_free LSM hook is called after bpf_prog_load/bpf_prog_alloc one
*even* if the hook itself returned error. If we don't do that, we run
the risk of leaking memory. This seems to be possible today when
combining SELinux and BPF LSM, as one example, depending on their
relative ordering.
Also, for BPF LSM setup, add bpf_prog_load and bpf_prog_free to
sleepable LSM hooks list, as they are both executed in sleepable
context. Also drop bpf_prog_load hook from untrusted, as there is no
issue with refcount or anything else anymore, that originally forced us
to add it to untrusted list in c0c852dd1876 ("bpf: Do not mark certain LSM
hook arguments as trusted"). We now trigger this hook much later and it
should not be an issue anymore.
Andrii Nakryiko [Thu, 30 Nov 2023 18:52:20 +0000 (10:52 -0800)]
bpf: consistently use BPF token throughout BPF verifier logic
Remove remaining direct queries to perfmon_capable() and bpf_capable()
in BPF verifier logic and instead use BPF token (if available) to make
decisions about privileges.
Andrii Nakryiko [Thu, 30 Nov 2023 18:52:19 +0000 (10:52 -0800)]
bpf: take into account BPF token when fetching helper protos
Instead of performing unconditional system-wide bpf_capable() and
perfmon_capable() calls inside bpf_base_func_proto() function (and other
similar ones) to determine eligibility of a given BPF helper for a given
program, use previously recorded BPF token during BPF_PROG_LOAD command
handling to inform the decision.
Andrii Nakryiko [Thu, 30 Nov 2023 18:52:18 +0000 (10:52 -0800)]
bpf: add BPF token support to BPF_PROG_LOAD command
Add basic support of BPF token to BPF_PROG_LOAD. Wire through a set of
allowed BPF program types and attach types, derived from BPF FS at BPF
token creation time. Then make sure we perform bpf_token_capable()
checks everywhere where it's relevant.
Andrii Nakryiko [Thu, 30 Nov 2023 18:52:17 +0000 (10:52 -0800)]
bpf: add BPF token support to BPF_BTF_LOAD command
Accept BPF token FD in BPF_BTF_LOAD command to allow BTF data loading
through delegated BPF token. BTF loading is a pretty straightforward
operation, so as long as BPF token is created with allow_cmds granting
BPF_BTF_LOAD command, kernel proceeds to parsing BTF data and creating
BTF object.
Andrii Nakryiko [Thu, 30 Nov 2023 18:52:16 +0000 (10:52 -0800)]
bpf: add BPF token support to BPF_MAP_CREATE command
Allow providing token_fd for BPF_MAP_CREATE command to allow controlled
BPF map creation from unprivileged process through delegated BPF token.
Wire through a set of allowed BPF map types to BPF token, derived from
BPF FS at BPF token creation time. This, in combination with allowed_cmds
allows to create a narrowly-focused BPF token (controlled by privileged
agent) with a restrictive set of BPF maps that application can attempt
to create.
Andrii Nakryiko [Thu, 30 Nov 2023 18:52:15 +0000 (10:52 -0800)]
bpf: introduce BPF token object
Add new kind of BPF kernel object, BPF token. BPF token is meant to
allow delegating privileged BPF functionality, like loading a BPF
program or creating a BPF map, from privileged process to a *trusted*
unprivileged process, all while having a good amount of control over which
privileged operations could be performed using provided BPF token.
This is achieved through mounting BPF FS instance with extra delegation
mount options, which determine what operations are delegatable, and also
constraining it to the owning user namespace (as mentioned in the
previous patch).
BPF token itself is just a derivative from BPF FS and can be created
through a new bpf() syscall command, BPF_TOKEN_CREATE, which accepts BPF
FS FD, which can be attained through open() API by opening BPF FS mount
point. Currently, BPF token "inherits" delegated command, map types,
prog type, and attach type bit sets from BPF FS as is. In the future,
having an BPF token as a separate object with its own FD, we can allow
to further restrict BPF token's allowable set of things either at the
creation time or after the fact, allowing the process to guard itself
further from unintentionally trying to load undesired kind of BPF
programs. But for now we keep things simple and just copy bit sets as is.
When BPF token is created from BPF FS mount, we take reference to the
BPF super block's owning user namespace, and then use that namespace for
checking all the {CAP_BPF, CAP_PERFMON, CAP_NET_ADMIN, CAP_SYS_ADMIN}
capabilities that are normally only checked against init userns (using
capable()), but now we check them using ns_capable() instead (if BPF
token is provided). See bpf_token_capable() for details.
Such setup means that BPF token in itself is not sufficient to grant BPF
functionality. User namespaced process has to *also* have necessary
combination of capabilities inside that user namespace. So while
previously CAP_BPF was useless when granted within user namespace, now
it gains a meaning and allows container managers and sys admins to have
a flexible control over which processes can and need to use BPF
functionality within the user namespace (i.e., container in practice).
And BPF FS delegation mount options and derived BPF tokens serve as
a per-container "flag" to grant overall ability to use bpf() (plus further
restrict on which parts of bpf() syscalls are treated as namespaced).
Note also, BPF_TOKEN_CREATE command itself requires ns_capable(CAP_BPF)
within the BPF FS owning user namespace, rounding up the ns_capable()
story of BPF token.
Andrii Nakryiko [Thu, 30 Nov 2023 18:52:14 +0000 (10:52 -0800)]
bpf: add BPF token delegation mount options to BPF FS
Add few new mount options to BPF FS that allow to specify that a given
BPF FS instance allows creation of BPF token (added in the next patch),
and what sort of operations are allowed under BPF token. As such, we get
4 new mount options, each is a bit mask
- `delegate_cmds` allow to specify which bpf() syscall commands are
allowed with BPF token derived from this BPF FS instance;
- if BPF_MAP_CREATE command is allowed, `delegate_maps` specifies
a set of allowable BPF map types that could be created with BPF token;
- if BPF_PROG_LOAD command is allowed, `delegate_progs` specifies
a set of allowable BPF program types that could be loaded with BPF token;
- if BPF_PROG_LOAD command is allowed, `delegate_attachs` specifies
a set of allowable BPF program attach types that could be loaded with
BPF token; delegate_progs and delegate_attachs are meant to be used
together, as full BPF program type is, in general, determined
through both program type and program attach type.
Currently, these mount options accept the following forms of values:
- a special value "any", that enables all possible values of a given
bit set;
- numeric value (decimal or hexadecimal, determined by kernel
automatically) that specifies a bit mask value directly;
- all the values for a given mount option are combined, if specified
multiple times. E.g., `mount -t bpf nodev /path/to/mount -o
delegate_maps=0x1 -o delegate_maps=0x2` will result in a combined 0x3
mask.
Ideally, more convenient (for humans) symbolic form derived from
corresponding UAPI enums would be accepted (e.g., `-o
delegate_progs=kprobe|tracepoint`) and I intend to implement this, but
it requires a bunch of UAPI header churn, so I postponed it until this
feature lands upstream or at least there is a definite consensus that
this feature is acceptable and is going to make it, just to minimize
amount of wasted effort and not increase amount of non-essential code to
be reviewed.
Attentive reader will notice that BPF FS is now marked as
FS_USERNS_MOUNT, which theoretically makes it mountable inside non-init
user namespace as long as the process has sufficient *namespaced*
capabilities within that user namespace. But in reality we still
restrict BPF FS to be mountable only by processes with CAP_SYS_ADMIN *in
init userns* (extra check in bpf_fill_super()). FS_USERNS_MOUNT is added
to allow creating BPF FS context object (i.e., fsopen("bpf")) from
inside unprivileged process inside non-init userns, to capture that
userns as the owning userns. It will still be required to pass this
context object back to privileged process to instantiate and mount it.
This manipulation is important, because capturing non-init userns as the
owning userns of BPF FS instance (super block) allows to use that userns
to constraint BPF token to that userns later on (see next patch). So
creating BPF FS with delegation inside unprivileged userns will restrict
derived BPF token objects to only "work" inside that intended userns,
making it scoped to a intended "container". Also, setting these
delegation options requires capable(CAP_SYS_ADMIN), so unprivileged
process cannot set this up without involvement of a privileged process.
There is a set of selftests at the end of the patch set that simulates
this sequence of steps and validates that everything works as intended.
But careful review is requested to make sure there are no missed gaps in
the implementation and testing.
This somewhat subtle set of aspects is the result of previous
discussions ([0]) about various user namespace implications and
interactions with BPF token functionality and is necessary to contain
BPF token inside intended user namespace.
Andrii Nakryiko [Thu, 30 Nov 2023 18:52:13 +0000 (10:52 -0800)]
bpf: align CAP_NET_ADMIN checks with bpf_capable() approach
Within BPF syscall handling code CAP_NET_ADMIN checks stand out a bit
compared to CAP_BPF and CAP_PERFMON checks. For the latter, CAP_BPF or
CAP_PERFMON are checked first, but if they are not set, CAP_SYS_ADMIN
takes over and grants whatever part of BPF syscall is required.
Similar kind of checks that involve CAP_NET_ADMIN are not so consistent.
One out of four uses does follow CAP_BPF/CAP_PERFMON model: during
BPF_PROG_LOAD, if the type of BPF program is "network-related" either
CAP_NET_ADMIN or CAP_SYS_ADMIN is required to proceed.
But in three other cases CAP_NET_ADMIN is required even if CAP_SYS_ADMIN
is set:
- when creating DEVMAP/XDKMAP/CPU_MAP maps;
- when attaching CGROUP_SKB programs;
- when handling BPF_PROG_QUERY command.
This patch is changing the latter three cases to follow BPF_PROG_LOAD
model, that is allowing to proceed under either CAP_NET_ADMIN or
CAP_SYS_ADMIN.
This also makes it cleaner in subsequent BPF token patches to switch
wholesomely to a generic bpf_token_capable(int cap) check, that always
falls back to CAP_SYS_ADMIN if requested capability is missing.
====================
Complete BPF verifier precision tracking support for register spills
Add support to BPF verifier to track and support register spill/fill to/from
stack regardless if it was done through read-only R10 register (which is the
only form supported today), or through a general register after copying R10
into it, while also potentially modifying offset.
Once we add register this generic spill/fill support to precision
backtracking, we can take advantage of it to stop doing eager STACK_ZERO
conversion on register spill. Instead we can rely on (im)precision of spilled
const zero register to improve verifier state pruning efficiency. This
situation of using const zero register to initialize stack slots is very
common with __builtin_memset() usage or just zero-initializing variables on
the stack, and it causes unnecessary state duplication, as that STACK_ZERO
knowledge is often not necessary for correctness, as those zero values are
never used in precise context. Thus, relying on register imprecision helps
tremendously, especially in real-world BPF programs.
To make spilled const zero register behave completely equivalently to
STACK_ZERO, we need to improve few other small pieces, which is done in the
second part of the patch set. See individual patches for details. There are
also two small bug fixes spotted during STACK_ZERO debugging.
The patch set consists of logically three changes:
- patch #1 (and corresponding tests in patch #2) is fixing/impoving precision
propagation for stack spills/fills. This can be landed as a stand-alone
improvement;
- patches #3 through #9 is improving verification scalability by utilizing
register (im)precision instead of eager STACK_ZERO. These changes depend
on patch #1.
- patch #10 is a memory efficiency improvement to how instruction/jump
history is tracked and maintained. It depends on patch #1, but is not
strictly speaking required, even though I believe it's a good long-term
solution to have a path-dependent per-instruction information. Kind
of like a path-dependent counterpart to path-agnostic insn_aux array.
v3->v3:
- fixed up Fixes tag (Alexei);
- fixed few more selftests to not use BPF_ST instruction in inline asm
directly, checked with CI, it was happy (CI);
v2->v3:
- BPF_ST instruction workaround (Eduard);
- force dereference in added tests to catch problems (Eduard);
- some commit message massaging (Alexei);
v1->v2:
- clean ups, WARN_ONCE(), insn_flags helpers added (Eduard);
- added more selftests for STACK_ZERO/STACK_MISC cases (Eduard);
- a bit more detailed explanation of effect of avoiding STACK_ZERO in favor
of register spill in patch #8 commit (Alexei);
- global shared instruction history refactoring moved to be the last patch
in the series to make it easier to revert it, if applied (Alexei).
====================
Andrii Nakryiko [Tue, 5 Dec 2023 18:42:47 +0000 (10:42 -0800)]
selftests/bpf: validate precision logic in partial_stack_load_preserves_zeros
Enhance partial_stack_load_preserves_zeros subtest with detailed
precision propagation log checks. We know expect fp-16 to be spilled,
initially imprecise, zero const register, which is later marked as
precise even when partial stack slot load is performed, even if it's not
a register fill (!).
Andrii Nakryiko [Tue, 5 Dec 2023 18:42:46 +0000 (10:42 -0800)]
bpf: track aligned STACK_ZERO cases as imprecise spilled registers
Now that precision backtracing is supporting register spill/fill to/from
stack, there is another oportunity to be exploited here: minimizing
precise STACK_ZERO cases. With a simple code change we can rely on
initially imprecise register spill tracking for cases when register
spilled to stack was a known zero.
This is a very common case for initializing on the stack variables,
including rather large structures. Often times zero has no special
meaning for the subsequent BPF program logic and is often overwritten
with non-zero values soon afterwards. But due to STACK_ZERO vs
STACK_MISC tracking, such initial zero initialization actually causes
duplication of verifier states as STACK_ZERO is clearly different than
STACK_MISC or spilled SCALAR_VALUE register.
The effect of this (now) trivial change is huge, as can be seen below.
These are differences between BPF selftests, Cilium, and Meta-internal
BPF object files relative to previous patch in this series. You can see
improvements ranging from single-digit percentage improvement for
instructions and states, all the way to 50-60% reduction for some of
Meta-internal host agent programs, and even some Cilium programs.
For Meta-internal ones I left only the differences for largest BPF
object files by states/instructions, as there were too many differences
in the overall output. All the differences were improvements, reducting
number of states and thus instructions validated.
Note, Meta-internal BPF object file names are not printed below.
Many copies of balancer_ingress are actually many different
configurations of Katran, so they are different BPF programs, which
explains state reduction going from -16% all the way to 31%, depending
on BPF program logic complexity.
I also tooked a closer look at a few small-ish BPF programs to validate
the behavior. Let's take bpf_iter_netrlink.bpf.o (first row below).
While it's just 8 vs 5 states, verifier log is still pretty long to
include it here. But the reduction in states is due to the following
piece of C code:
You can see that in some situations `ino` is zero-initialized, while in
others it's unknown value filled out by bpf_probe_read_kernel(). Before
this change code after if/else branches have to be validated twice. Once
with (precise) ino == 0, due to eager STACK_ZERO logic, and then again
for when ino is just STACK_MISC. But BPF_SEQ_PRINTF() doesn't care about
precise value of ino, so with the change in this patch verifier is able
to prune states from after one of the branches, reducing number of total
states (and instructions) required for successful validation.
Similar principle applies to bigger real-world applications, just at
a much larger scale.
Andrii Nakryiko [Tue, 5 Dec 2023 18:42:45 +0000 (10:42 -0800)]
selftests/bpf: validate zero preservation for sub-slot loads
Validate that 1-, 2-, and 4-byte loads from stack slots not aligned on
8-byte boundary still preserve zero, when loading from all-STACK_ZERO
sub-slots, or when stack sub-slots are covered by spilled register with
known constant zero value.
Andrii Nakryiko [Tue, 5 Dec 2023 18:42:44 +0000 (10:42 -0800)]
bpf: preserve constant zero when doing partial register restore
Similar to special handling of STACK_ZERO, when reading 1/2/4 bytes from
stack from slot that has register spilled into it and that register has
a constant value zero, preserve that zero and mark spilled register as
precise for that. This makes spilled const zero register and STACK_ZERO
cases equivalent in their behavior.
Andrii Nakryiko [Tue, 5 Dec 2023 18:42:42 +0000 (10:42 -0800)]
bpf: preserve STACK_ZERO slots on partial reg spills
Instead of always forcing STACK_ZERO slots to STACK_MISC, preserve it in
situations where this is possible. E.g., when spilling register as
1/2/4-byte subslots on the stack, all the remaining bytes in the stack
slot do not automatically become unknown. If we knew they contained
zeroes, we can preserve those STACK_ZERO markers.
Add a helper mark_stack_slot_misc(), similar to scrub_spilled_slot(),
but that doesn't overwrite either STACK_INVALID nor STACK_ZERO. Note
that we need to take into account possibility of being in unprivileged
mode, in which case STACK_INVALID is forced to STACK_MISC for correctness,
as treating STACK_INVALID as equivalent STACK_MISC is only enabled in
privileged mode.
Andrii Nakryiko [Tue, 5 Dec 2023 18:42:41 +0000 (10:42 -0800)]
bpf: fix check for attempt to corrupt spilled pointer
When register is spilled onto a stack as a 1/2/4-byte register, we set
slot_type[BPF_REG_SIZE - 1] (plus potentially few more below it,
depending on actual spill size). So to check if some stack slot has
spilled register we need to consult slot_type[7], not slot_type[0].
To avoid the need to remember and double-check this in the future, just
use is_spilled_reg() helper.
Andrii Nakryiko [Tue, 5 Dec 2023 18:42:40 +0000 (10:42 -0800)]
selftests/bpf: add stack access precision test
Add a new selftests that validates precision tracking for stack access
instruction, using both r10-based and non-r10-based accesses. For
non-r10 ones we also make sure to have non-zero var_off to validate that
final stack offset is tracked properly in instruction history
information inside verifier.
Andrii Nakryiko [Tue, 5 Dec 2023 18:42:39 +0000 (10:42 -0800)]
bpf: support non-r10 register spill/fill to/from stack in precision tracking
Use instruction (jump) history to record instructions that performed
register spill/fill to/from stack, regardless if this was done through
read-only r10 register, or any other register after copying r10 into it
*and* potentially adjusting offset.
To make this work reliably, we push extra per-instruction flags into
instruction history, encoding stack slot index (spi) and stack frame
number in extra 10 bit flags we take away from prev_idx in instruction
history. We don't touch idx field for maximum performance, as it's
checked most frequently during backtracking.
This change removes basically the last remaining practical limitation of
precision backtracking logic in BPF verifier. It fixes known
deficiencies, but also opens up new opportunities to reduce number of
verified states, explored in the subsequent patches.
There are only three differences in selftests' BPF object files
according to veristat, all in the positive direction (less states).
Note, I avoided renaming jmp_history to more generic insn_hist to
minimize number of lines changed and potential merge conflicts between
bpf and bpf-next trees.
Notice also cur_hist_entry pointer reset to NULL at the beginning of
instruction verification loop. This pointer avoids the problem of
relying on last jump history entry's insn_idx to determine whether we
already have entry for current instruction or not. It can happen that we
added jump history entry because current instruction is_jmp_point(), but
also we need to add instruction flags for stack access. In this case, we
don't want to entries, so we need to reuse last added entry, if it is
present.
Relying on insn_idx comparison has the same ambiguity problem as the one
that was fixed recently in [0], so we avoid that.
selftests/bpf: Make sure we trigger metadata kfuncs for dst 8080
xdp_metadata test is flaky sometimes:
verify_xsk_metadata:FAIL:rx_hash_type unexpected rx_hash_type: actual 8 != expected 0
Where 8 means XDP_RSS_TYPE_L4_ANY and is exported from veth driver only when
'skb->l4_hash' condition is met. This makes me think that the program is
triggering again for some other packet.
Let's have a filter, similar to xdp_hw_metadata, where we trigger XDP kfuncs
only for UDP packets destined to port 8080.
Dave Marchevsky [Mon, 4 Dec 2023 21:17:22 +0000 (13:17 -0800)]
selftests/bpf: Test bpf_kptr_xchg stashing of bpf_rb_root
There was some confusion amongst Meta sched_ext folks regarding whether
stashing bpf_rb_root - the tree itself, rather than a single node - was
supported. This patch adds a small test which demonstrates this
functionality: a local kptr with rb_root is created, a node is created
and added to the tree, then the tree is kptr_xchg'd into a mapval.
The patchset aims to fix the release of inner map in map array or map
htab. The release of inner map is different with normal map. For normal
map, the map is released after the bpf program which uses the map is
destroyed, because the bpf program tracks the used maps. However bpf
program can not track the used inner map because these inner map may be
updated or deleted dynamically, and for now the ref-counter of inner map
is decreased after the inner map is remove from outer map, so the inner
map may be freed before the bpf program, which is accessing the inner
map, exits and there will be use-after-free problem as demonstrated by
patch #6.
The patchset fixes the problem by deferring the release of inner map.
The freeing of inner map is deferred according to the sleepable
attributes of the bpf programs which own the outer map. Patch #1 fixes
the warning when running the newly-added selftest under interpreter
mode. Patch #2 adds more parameters to .map_fd_put_ptr() to prepare for
the fix. Patch #3 fixes the incorrect value of need_defer when freeing
the fd array. Patch #4 fixes the potential use-after-free problem by
using call_rcu_tasks_trace() and call_rcu() to wait for one tasks trace
RCU GP and one RCU GP unconditionally. Patch #5 optimizes the free of
inner map by removing the unnecessary RCU GP waiting. Patch #6 adds a
selftest to demonstrate the potential use-after-free problem. Patch #7
updates a selftest to update outer map in syscall bpf program.
Please see individual patches for more details. And comments are always
welcome.
Change Log:
v5:
* patch #3: rename fd_array_map_delete_elem_with_deferred_free() to
__fd_array_map_delete_elem() (Alexei)
* patch #5: use atomic64_t instead of atomic_t to prevent potential
overflow (Alexei)
* patch #7: use ptr_to_u64() helper instead of force casting to initialize
pointers in bpf_attr (Alexei)
v4: https://lore.kernel.org/bpf/20231130140120.1736235[email protected]
* patch #2: don't use "deferred", use "need_defer" uniformly
* patch #3: newly-added, fix the incorrect value of need_defer during
fd array free.
* patch #4: doesn't consider the case in which bpf map is not used by
any bpf program and only use sleepable_refcnt to remove
unnecessary tasks trace RCU GP (Alexei)
* patch #4: remove memory barriers added due to cautiousness (Alexei)
v3: https://lore.kernel.org/bpf/20231124113033[email protected]
* multiple variable renamings (Martin)
* define BPF_MAP_RCU_GP/BPF_MAP_RCU_TT_GP as bit (Martin)
* use call_rcu() and its variants instead of synchronize_rcu() (Martin)
* remove unnecessary mask in bpf_map_free_deferred() (Martin)
* place atomic_or() and the related smp_mb() together (Martin)
* add patch #6 to demonstrate that updating outer map in syscall
program is dead-lock free (Alexei)
* update comments about the memory barrier in bpf_map_fd_put_ptr()
* update commit message for patch #3 and #4 to describe more details
v2: https://lore.kernel.org/bpf/20231113123324.3914612[email protected]
* defer the invocation of ops->map_free() instead of bpf_map_put() (Martin)
* update selftest to make it being reproducible under JIT mode (Martin)
* remove unnecessary preparatory patches
Hou Tao [Mon, 4 Dec 2023 14:04:25 +0000 (22:04 +0800)]
selftests/bpf: Test outer map update operations in syscall program
Syscall program is running with rcu_read_lock_trace being held, so if
bpf_map_update_elem() or bpf_map_delete_elem() invokes
synchronize_rcu_tasks_trace() when operating on an outer map, there will
be dead-lock, so add a test to guarantee that it is dead-lock free.
Hou Tao [Mon, 4 Dec 2023 14:04:24 +0000 (22:04 +0800)]
selftests/bpf: Add test cases for inner map
Add test cases to test the race between the destroy of inner map due to
map-in-map update and the access of inner map in bpf program. The
following 4 combinations are added:
(1) array map in map array + bpf program
(2) array map in map array + sleepable bpf program
(3) array map in map htab + bpf program
(4) array map in map htab + sleepable bpf program
Before applying the fixes, when running `./test_prog -a map_in_map`, the
following error was reported:
==================================================================
BUG: KASAN: slab-use-after-free in array_map_update_elem+0x48/0x3e0
Read of size 4 at addr ffff888114f33824 by task test_progs/1858
CPU: 1 PID: 1858 Comm: test_progs Tainted: G O 6.6.0+ #7
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996) ......
Call Trace:
<TASK>
dump_stack_lvl+0x4a/0x90
print_report+0xd2/0x620
kasan_report+0xd1/0x110
__asan_load4+0x81/0xa0
array_map_update_elem+0x48/0x3e0
bpf_prog_be94a9f26772f5b7_access_map_in_array+0xe6/0xf6
trace_call_bpf+0x1aa/0x580
kprobe_perf_func+0xdd/0x430
kprobe_dispatcher+0xa0/0xb0
kprobe_ftrace_handler+0x18b/0x2e0
0xffffffffc02280f7
RIP: 0010:__x64_sys_getpgid+0x1/0x30
......
</TASK>
Hou Tao [Mon, 4 Dec 2023 14:04:23 +0000 (22:04 +0800)]
bpf: Optimize the free of inner map
When removing the inner map from the outer map, the inner map will be
freed after one RCU grace period and one RCU tasks trace grace
period, so it is certain that the bpf program, which may access the
inner map, has exited before the inner map is freed.
However there is no need to wait for one RCU tasks trace grace period if
the outer map is only accessed by non-sleepable program. So adding
sleepable_refcnt in bpf_map and increasing sleepable_refcnt when adding
the outer map into env->used_maps for sleepable program. Although the
max number of bpf program is INT_MAX - 1, the number of bpf programs
which are being loaded may be greater than INT_MAX, so using atomic64_t
instead of atomic_t for sleepable_refcnt. When removing the inner map
from the outer map, using sleepable_refcnt to decide whether or not a
RCU tasks trace grace period is needed before freeing the inner map.
Hou Tao [Mon, 4 Dec 2023 14:04:22 +0000 (22:04 +0800)]
bpf: Defer the free of inner map when necessary
When updating or deleting an inner map in map array or map htab, the map
may still be accessed by non-sleepable program or sleepable program.
However bpf_map_fd_put_ptr() decreases the ref-counter of the inner map
directly through bpf_map_put(), if the ref-counter is the last one
(which is true for most cases), the inner map will be freed by
ops->map_free() in a kworker. But for now, most .map_free() callbacks
don't use synchronize_rcu() or its variants to wait for the elapse of a
RCU grace period, so after the invocation of ops->map_free completes,
the bpf program which is accessing the inner map may incur
use-after-free problem.
Fix the free of inner map by invoking bpf_map_free_deferred() after both
one RCU grace period and one tasks trace RCU grace period if the inner
map has been removed from the outer map before. The deferment is
accomplished by using call_rcu() or call_rcu_tasks_trace() when
releasing the last ref-counter of bpf map. The newly-added rcu_head
field in bpf_map shares the same storage space with work field to
reduce the size of bpf_map.
Hou Tao [Mon, 4 Dec 2023 14:04:21 +0000 (22:04 +0800)]
bpf: Set need_defer as false when clearing fd array during map free
Both map deletion operation, map release and map free operation use
fd_array_map_delete_elem() to remove the element from fd array and
need_defer is always true in fd_array_map_delete_elem(). For the map
deletion operation and map release operation, need_defer=true is
necessary, because the bpf program, which accesses the element in fd
array, may still alive. However for map free operation, it is certain
that the bpf program which owns the fd array has already been exited, so
setting need_defer as false is appropriate for map free operation.
So fix it by adding need_defer parameter to bpf_fd_array_map_clear() and
adding a new helper __fd_array_map_delete_elem() to handle the map
deletion, map release and map free operations correspondingly.
projects / linux.git / log