All map types reimplement the field-by-field copy of union bpf_attr
members into struct bpf_map. Add a helper to perform this operation.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Quentin Monnet <quentin.monnet@netronome.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Use the new callback to perform allocation checks for hash maps.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Quentin Monnet <quentin.monnet@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Number of attribute checks are currently performed after hashtab
is already allocated. Move them to be able to split them out to
the check function later on. Checks have to now be performed on
the attr union directly instead of the members of bpf_map, since
bpf_map will be allocated later. No functional changes.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Quentin Monnet <quentin.monnet@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
While using large percpu maps, htab_map_alloc() can hold
cpu for hundreds of ms.
This patch adds cond_resched() calls to percpu alloc/free
call sites, all running in process context.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
There were quite a few overlapping sets of changes here.
Daniel's bug fix for off-by-ones in the new BPF branch instructions,
along with the added allowances for "data_end > ptr + x" forms
collided with the metadata additions.
Along with those three changes came veritifer test cases, which in
their final form I tried to group together properly. If I had just
trimmed GIT's conflict tags as-is, this would have split up the
meta tests unnecessarily.
In the socketmap code, a set of preemption disabling changes
overlapped with the rename of bpf_compute_data_end() to
bpf_compute_data_pointers().
Changes were made to the mv88e6060.c driver set addr method
which got removed in net-next.
The hyperv transport socket layer had a locking change in 'net'
which overlapped with a change of socket state macro usage
in 'net-next'.
Signed-off-by: David S. Miller <davem@davemloft.net>
Introduce the map read/write flags to the eBPF syscalls that returns the
map fd. The flags is used to set up the file mode when construct a new
file descriptor for bpf maps. To not break the backward capability, the
f_flags is set to O_RDWR if the flag passed by syscall is 0. Otherwise
it should be O_RDONLY or O_WRONLY. When the userspace want to modify or
read the map content, it will check the file mode to see if it is
allowed to make the change.
Signed-off-by: Chenbo Feng <fengc@google.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
PCPU_MIN_UNIT_SIZE is an implementation detail of the percpu
allocator. Given we support __GFP_NOWARN now, lets just let
the allocation request fail naturally instead. The two call
sites from BPF mistakenly assumed __GFP_NOWARN would work, so
no changes needed to their actual __alloc_percpu_gfp() calls
which use the flag already.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch writes 'node->ref = 1' only if node->ref is 0.
The number of lookups/s for a ~1M entries LRU map increased by
~30% (260097 to 343313).
Other writes on 'node->ref = 0' is not changed. In those cases, the
same cache line has to be changed anyway.
First column: Size of the LRU hash
Second column: Number of lookups/s
Before:
> echo "$((2**20+1)): $(./map_perf_test 1024 1 $((2**20+1)) 10000000 | awk '{print $3}')"
1048577: 260097
After:
> echo "$((2**20+1)): $(./map_perf_test 1024 1 $((2**20+1)) 10000000 | awk '{print $3}')"
1048577: 343313
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Inline the lru map lookup to save the cost in making calls to
bpf_map_lookup_elem() and htab_lru_map_lookup_elem().
Different LRU hash size is tested. The benefit diminishes when
the cache miss starts to dominate in the bigger LRU hash.
Considering the change is simple, it is still worth to optimize.
First column: Size of the LRU hash
Second column: Number of lookups/s
Before:
> for i in $(seq 9 20); do echo "$((2**i+1)): $(./map_perf_test 1024 1 $((2**i+1)) 10000000 | awk '{print $3}')"; done
513: 1132020
1025: 1056826
2049: 1007024
4097: 853298
8193: 742723
16385: 712600
32769: 688142
65537: 677028
131073: 619437
262145: 498770
524289: 316695
1048577: 260038
After:
> for i in $(seq 9 20); do echo "$((2**i+1)): $(./map_perf_test 1024 1 $((2**i+1)) 10000000 | awk '{print $3}')"; done
513: 1221851
1025: 1144695
2049: 1049902
4097: 884460
8193: 773731
16385: 729673
32769: 721989
65537: 715530
131073: 671665
262145: 516987
524289: 321125
1048577: 260048
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, iproute2's BPF ELF loader works fine with array of maps
when retrieving the fd from a pinned node and doing a selfcheck
against the provided map attributes from the object file, but we
fail to do the same for hash of maps and thus refuse to get the
map from pinned node.
Reason is that when allocating hash of maps, fd_htab_map_alloc() will
set the value size to sizeof(void *), and any user space map creation
requests are forced to set 4 bytes as value size. Thus, selfcheck
will complain about exposed 8 bytes on 64 bit archs vs. 4 bytes from
object file as value size. Contract is that fdinfo or BPF_MAP_GET_FD_BY_ID
returns the value size used to create the map.
Fix it by handling it the same way as we do for array of maps, which
means that we leave value size at 4 bytes and in the allocation phase
round up value size to 8 bytes. alloc_htab_elem() needs an adjustment
in order to copy rounded up 8 bytes due to bpf_fd_htab_map_update_elem()
calling into htab_map_update_elem() with the pointer of the map
pointer as value. Unlike array of maps where we just xchg(), we're
using the generic htab_map_update_elem() callback also used from helper
calls, which published the key/value already on return, so we need
to ensure to memcpy() the right size.
Fixes: bcc6b1b7eb ("bpf: Add hash of maps support")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Avoid two successive functions calls for the map in map lookup, first
is the bpf_map_lookup_elem() helper call, and second the callback via
map->ops->map_lookup_elem() to get to the map in map implementation.
Implementation inlines array and htab flavor for map in map lookups.
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 current map creation API does not allow to provide the numa-node
preference. The memory usually comes from where the map-creation-process
is running. The performance is not ideal if the bpf_prog is known to
always run in a numa node different from the map-creation-process.
One of the use case is sharding on CPU to different LRU maps (i.e.
an array of LRU maps). Here is the test result of map_perf_test on
the INNER_LRU_HASH_PREALLOC test if we force the lru map used by
CPU0 to be allocated from a remote numa node:
[ The machine has 20 cores. CPU0-9 at node 0. CPU10-19 at node 1 ]
># taskset -c 10 ./map_perf_test 512 8 1260000 8000000
5:inner_lru_hash_map_perf pre-alloc 1628380 events per sec
4:inner_lru_hash_map_perf pre-alloc 1626396 events per sec
3:inner_lru_hash_map_perf pre-alloc 1626144 events per sec
6:inner_lru_hash_map_perf pre-alloc 1621657 events per sec
2:inner_lru_hash_map_perf pre-alloc 1621534 events per sec
1:inner_lru_hash_map_perf pre-alloc 1620292 events per sec
7:inner_lru_hash_map_perf pre-alloc 1613305 events per sec
0:inner_lru_hash_map_perf pre-alloc 1239150 events per sec #<<<
After specifying numa node:
># taskset -c 10 ./map_perf_test 512 8 1260000 8000000
5:inner_lru_hash_map_perf pre-alloc 1629627 events per sec
3:inner_lru_hash_map_perf pre-alloc 1628057 events per sec
1:inner_lru_hash_map_perf pre-alloc 1623054 events per sec
6:inner_lru_hash_map_perf pre-alloc 1616033 events per sec
2:inner_lru_hash_map_perf pre-alloc 1614630 events per sec
4:inner_lru_hash_map_perf pre-alloc 1612651 events per sec
7:inner_lru_hash_map_perf pre-alloc 1609337 events per sec
0:inner_lru_hash_map_perf pre-alloc 1619340 events per sec #<<<
This patch adds one field, numa_node, to the bpf_attr. Since numa node 0
is a valid node, a new flag BPF_F_NUMA_NODE is also added. The numa_node
field is honored if and only if the BPF_F_NUMA_NODE flag is set.
Numa node selection is not supported for percpu map.
This patch does not change all the kmalloc. F.e.
'htab = kzalloc()' is not changed since the object
is small enough to stay in the cache.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch allows userspace to do BPF_MAP_LOOKUP_ELEM on
BPF_MAP_TYPE_PROG_ARRAY,
BPF_MAP_TYPE_ARRAY_OF_MAPS and
BPF_MAP_TYPE_HASH_OF_MAPS.
The lookup returns a prog-id or map-id to the userspace.
The userspace can then use the BPF_PROG_GET_FD_BY_ID
or BPF_MAP_GET_FD_BY_ID to get a fd.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
When iterating through a map, we need to find a key that does not exist
in the map so map_get_next_key will give us the first key of the map.
This often requires a lot of guessing in production systems.
This patch makes map_get_next_key return the first key when the key
pointer in the parameter is NULL.
Signed-off-by: Teng Qin <qinteng@fb.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>
There's no need to have struct bpf_map_type_list since
it just contains a list_head, the type, and the ops
pointer. Since the types are densely packed and not
actually dynamically registered, it's much easier and
smaller to have an array of type->ops pointer. Also
initialize this array statically to remove code needed
to initialize it.
In order to save duplicating the list, move it to the
types header file added by the previous patch and
include it in the same fashion.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Conflicts:
drivers/net/ethernet/broadcom/genet/bcmmii.c
drivers/net/hyperv/netvsc.c
kernel/bpf/hashtab.c
Almost entirely overlapping changes.
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds hash of maps support (hashmap->bpf_map).
BPF_MAP_TYPE_HASH_OF_MAPS is added.
A map-in-map contains a pointer to another map and lets call
this pointer 'inner_map_ptr'.
Notes on deleting inner_map_ptr from a hash map:
1. For BPF_F_NO_PREALLOC map-in-map, when deleting
an inner_map_ptr, the htab_elem itself will go through
a rcu grace period and the inner_map_ptr resides
in the htab_elem.
2. For pre-allocated htab_elem (!BPF_F_NO_PREALLOC),
when deleting an inner_map_ptr, the htab_elem may
get reused immediately. This situation is similar
to the existing prealloc-ated use cases.
However, the bpf_map_fd_put_ptr() calls bpf_map_put() which calls
inner_map->ops->map_free(inner_map) which will go
through a rcu grace period (i.e. all bpf_map's map_free
currently goes through a rcu grace period). Hence,
the inner_map_ptr is still safe for the rcu reader side.
This patch also includes BPF_MAP_TYPE_HASH_OF_MAPS to the
check_map_prealloc() in the verifier. preallocation is a
must for BPF_PROG_TYPE_PERF_EVENT. Hence, even we don't expect
heavy updates to map-in-map, enforcing BPF_F_NO_PREALLOC for map-in-map
is impossible without disallowing BPF_PROG_TYPE_PERF_EVENT from using
map-in-map first.
Signed-off-by: Martin KaFai Lau <kafai@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>
In both kmalloc and prealloc mode the bpf_map_update_elem() is using
per-cpu extra_elems to do atomic update when the map is full.
There are two issues with it. The logic can be misused, since it allows
max_entries+num_cpus elements to be present in the map. And alloc_extra_elems()
at map creation time can fail percpu alloc for large map values with a warn:
WARNING: CPU: 3 PID: 2752 at ../mm/percpu.c:892 pcpu_alloc+0x119/0xa60
illegal size (32824) or align (8) for percpu allocation
The fixes for both of these issues are different for kmalloc and prealloc modes.
For prealloc mode allocate extra num_possible_cpus elements and store
their pointers into extra_elems array instead of actual elements.
Hence we can use these hidden(spare) elements not only when the map is full
but during bpf_map_update_elem() that replaces existing element too.
That also improves performance, since pcpu_freelist_pop/push is avoided.
Unfortunately this approach cannot be used for kmalloc mode which needs
to kfree elements after rcu grace period. Therefore switch it back to normal
kmalloc even when full and old element exists like it was prior to
commit 6c90598174 ("bpf: pre-allocate hash map elements").
Add tests to check for over max_entries and large map values.
Reported-by: Dave Jones <davej@codemonkey.org.uk>
Fixes: 6c90598174 ("bpf: pre-allocate hash map elements")
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Optimize:
bpf_call
bpf_map_lookup_elem
map->ops->map_lookup_elem
htab_map_lookup_elem
__htab_map_lookup_elem
into:
bpf_call
__htab_map_lookup_elem
to improve performance of JITed programs.
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
when all map elements are pre-allocated one cpu can delete and reuse htab_elem
while another cpu is still walking the hlist. In such case the lookup may
miss the element. Convert hlist to hlist_nulls to avoid such scenario.
When bucket lock is taken there is no need to take such precautions,
so only convert map_lookup and map_get_next to nulls.
The race window is extremely small and only reproducible with explicit
udelay() inside lookup_nulls_elem_raw()
Similar to hlist add hlist_nulls_for_each_entry_safe() and
hlist_nulls_entry_safe() helpers.
Fixes: 6c90598174 ("bpf: pre-allocate hash map elements")
Reported-by: Jonathan Perry <jonperry@fb.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>
when htab_elem is removed from the bucket list the htab_elem.hash_node.next
field should not be overridden too early otherwise we have a tiny race window
between lookup and delete.
The bug was discovered by manual code analysis and reproducible
only with explicit udelay() in lookup_elem_raw().
Fixes: 6c90598174 ("bpf: pre-allocate hash map elements")
Reported-by: Jonathan Perry <jonperry@fb.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>
All map types and prog types are registered to the BPF core through
bpf_register_map_type() and bpf_register_prog_type() during init and
remain unchanged thereafter. As by design we don't (and never will)
have any pluggable code that can register to that at any later point
in time, lets mark all the existing bpf_{map,prog}_type_list objects
in the tree as __ro_after_init, so they can be moved to read-only
section from then onwards.
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 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 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>
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>
Commit a6ed3ea65d ("bpf: restore behavior of bpf_map_update_elem")
added an extra per-cpu reserve to the hash table map to restore old
behaviour from pre prealloc times. When non-prealloc is in use for a
map, then problem is that once a hash table extra element has been
linked into the hash-table, and the hash table is destroyed due to
refcount dropping to zero, then htab_map_free() -> delete_all_elements()
will walk the whole hash table and drop all elements via htab_elem_free().
The problem is that the element from the extra reserve is first fed
to the wrong backend allocator and eventually freed twice.
Fixes: a6ed3ea65d ("bpf: restore behavior of bpf_map_update_elem")
Reported-by: Dmitry Vyukov <dvyukov@google.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>
The introduction of pre-allocated hash elements inadvertently broke
the behavior of bpf hash maps where users expected to call
bpf_map_update_elem() without considering that the map can be full.
Some programs do:
old_value = bpf_map_lookup_elem(map, key);
if (old_value) {
... prepare new_value on stack ...
bpf_map_update_elem(map, key, new_value);
}
Before pre-alloc the update() for existing element would work even
in 'map full' condition. Restore this behavior.
The above program could have updated old_value in place instead of
update() which would be faster and most programs use that approach,
but sometimes the values are large and the programs use update()
helper to do atomic replacement of the element.
Note we cannot simply update element's value in-place like percpu
hash map does and have to allocate extra num_possible_cpu elements
and use this extra reserve when the map is full.
Fixes: 6c90598174 ("bpf: pre-allocate hash map elements")
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
If kprobe is placed on spin_unlock then calling kmalloc/kfree from
bpf programs is not safe, since the following dead lock is possible:
kfree->spin_lock(kmem_cache_node->lock)...spin_unlock->kprobe->
bpf_prog->map_update->kmalloc->spin_lock(of the same kmem_cache_node->lock)
and deadlocks.
The following solutions were considered and some implemented, but
eventually discarded
- kmem_cache_create for every map
- add recursion check to slow-path of slub
- use reserved memory in bpf_map_update for in_irq or in preempt_disabled
- kmalloc via irq_work
At the end pre-allocation of all map elements turned out to be the simplest
solution and since the user is charged upfront for all the memory, such
pre-allocation doesn't affect the user space visible behavior.
Since it's impossible to tell whether kprobe is triggered in a safe
location from kmalloc point of view, use pre-allocation by default
and introduce new BPF_F_NO_PREALLOC flag.
While testing of per-cpu hash maps it was discovered
that alloc_percpu(GFP_ATOMIC) has odd corner cases and often
fails to allocate memory even when 90% of it is free.
The pre-allocation of per-cpu hash elements solves this problem as well.
Turned out that bpf_map_update() quickly followed by
bpf_map_lookup()+bpf_map_delete() is very common pattern used
in many of iovisor/bcc/tools, so there is additional benefit of
pre-allocation, since such use cases are must faster.
Since all hash map elements are now pre-allocated we can remove
atomic increment of htab->count and save few more cycles.
Also add bpf_map_precharge_memlock() to check rlimit_memlock early to avoid
large malloc/free done by users who don't have sufficient limits.
Pre-allocation is done with vmalloc and alloc/free is done
via percpu_freelist. Here are performance numbers for different
pre-allocation algorithms that were implemented, but discarded
in favor of percpu_freelist:
1 cpu:
pcpu_ida 2.1M
pcpu_ida nolock 2.3M
bt 2.4M
kmalloc 1.8M
hlist+spinlock 2.3M
pcpu_freelist 2.6M
4 cpu:
pcpu_ida 1.5M
pcpu_ida nolock 1.8M
bt w/smp_align 1.7M
bt no/smp_align 1.1M
kmalloc 0.7M
hlist+spinlock 0.2M
pcpu_freelist 2.0M
8 cpu:
pcpu_ida 0.7M
bt w/smp_align 0.8M
kmalloc 0.4M
pcpu_freelist 1.5M
32 cpu:
kmalloc 0.13M
pcpu_freelist 0.49M
pcpu_ida nolock is a modified percpu_ida algorithm without
percpu_ida_cpu locks and without cross-cpu tag stealing.
It's faster than existing percpu_ida, but not as fast as pcpu_freelist.
bt is a variant of block/blk-mq-tag.c simlified and customized
for bpf use case. bt w/smp_align is using cache line for every 'long'
(similar to blk-mq-tag). bt no/smp_align allocates 'long'
bitmasks continuously to save memory. It's comparable to percpu_ida
and in some cases faster, but slower than percpu_freelist
hlist+spinlock is the simplest free list with single spinlock.
As expeceted it has very bad scaling in SMP.
kmalloc is existing implementation which is still available via
BPF_F_NO_PREALLOC flag. It's significantly slower in single cpu and
in 8 cpu setup it's 3 times slower than pre-allocation with pcpu_freelist,
but saves memory, so in cases where map->max_entries can be large
and number of map update/delete per second is low, it may make
sense to use it.
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
bpf_percpu_hash_update() expects rcu lock to be held and warns if it's not,
which pointed out a missing rcu read lock.
Fixes: 15a07b338 ("bpf: add lookup/update support for per-cpu hash and array maps")
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
The functions bpf_map_lookup_elem(map, key, value) and
bpf_map_update_elem(map, key, value, flags) need to get/set
values from all-cpus for per-cpu hash and array maps,
so that user space can aggregate/update them as necessary.
Example of single counter aggregation in user space:
unsigned int nr_cpus = sysconf(_SC_NPROCESSORS_CONF);
long values[nr_cpus];
long value = 0;
bpf_lookup_elem(fd, key, values);
for (i = 0; i < nr_cpus; i++)
value += values[i];
The user space must provide round_up(value_size, 8) * nr_cpus
array to get/set values, since kernel will use 'long' copy
of per-cpu values to try to copy good counters atomically.
It's a best-effort, since bpf programs and user space are racing
to access the same memory.
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Introduce BPF_MAP_TYPE_PERCPU_HASH map type which is used to do
accurate counters without need to use BPF_XADD instruction which turned
out to be too costly for high-performance network monitoring.
In the typical use case the 'key' is the flow tuple or other long
living object that sees a lot of events per second.
bpf_map_lookup_elem() returns per-cpu area.
Example:
struct {
u32 packets;
u32 bytes;
} * ptr = bpf_map_lookup_elem(&map, &key);
/* ptr points to this_cpu area of the value, so the following
* increments will not collide with other cpus
*/
ptr->packets ++;
ptr->bytes += skb->len;
bpf_update_elem() atomically creates a new element where all per-cpu
values are zero initialized and this_cpu value is populated with
given 'value'.
Note that non-per-cpu hash map always allocates new element
and then deletes old after rcu grace period to maintain atomicity
of update. Per-cpu hash map updates element values in-place.
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Both htab_map_update_elem() and htab_map_delete_elem() can be
called from eBPF program, and they may be in kernel hot path,
so it isn't efficient to use a per-hashtable lock in this two
helpers.
The per-hashtable spinlock is used for protecting bucket's
hlist, and per-bucket lock is just enough. This patch converts
the per-hashtable lock into per-bucket spinlock, so that
contention can be decreased a lot.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
The spinlock is just used for protecting the per-bucket
hlist, so it isn't needed for selecting bucket.
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Preparing for removing global per-hashtable lock, so
the counter need to be defined as aotmic_t first.
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For large map->value_size the user space can trigger memory allocation warnings like:
WARNING: CPU: 2 PID: 11122 at mm/page_alloc.c:2989
__alloc_pages_nodemask+0x695/0x14e0()
Call Trace:
[< inline >] __dump_stack lib/dump_stack.c:15
[<ffffffff82743b56>] dump_stack+0x68/0x92 lib/dump_stack.c:50
[<ffffffff81244ec9>] warn_slowpath_common+0xd9/0x140 kernel/panic.c:460
[<ffffffff812450f9>] warn_slowpath_null+0x29/0x30 kernel/panic.c:493
[< inline >] __alloc_pages_slowpath mm/page_alloc.c:2989
[<ffffffff81554e95>] __alloc_pages_nodemask+0x695/0x14e0 mm/page_alloc.c:3235
[<ffffffff816188fe>] alloc_pages_current+0xee/0x340 mm/mempolicy.c:2055
[< inline >] alloc_pages include/linux/gfp.h:451
[<ffffffff81550706>] alloc_kmem_pages+0x16/0xf0 mm/page_alloc.c:3414
[<ffffffff815a1c89>] kmalloc_order+0x19/0x60 mm/slab_common.c:1007
[<ffffffff815a1cef>] kmalloc_order_trace+0x1f/0xa0 mm/slab_common.c:1018
[< inline >] kmalloc_large include/linux/slab.h:390
[<ffffffff81627784>] __kmalloc+0x234/0x250 mm/slub.c:3525
[< inline >] kmalloc include/linux/slab.h:463
[< inline >] map_update_elem kernel/bpf/syscall.c:288
[< inline >] SYSC_bpf kernel/bpf/syscall.c:744
To avoid never succeeding kmalloc with order >= MAX_ORDER check that
elem->value_size and computed elem_size are within limits for both hash and
array type maps.
Also add __GFP_NOWARN to kmalloc(value_size | elem_size) to avoid OOM warnings.
Note kmalloc(key_size) is highly unlikely to trigger OOM, since key_size <= 512,
so keep those kmalloc-s as-is.
Large value_size can cause integer overflows in elem_size and map.pages
formulas, so check for that as well.
Fixes: aaac3ba95e ("bpf: charge user for creation of BPF maps and programs")
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
since eBPF programs and maps use kernel memory consider it 'locked' memory
from user accounting point of view and charge it against RLIMIT_MEMLOCK limit.
This limit is typically set to 64Kbytes by distros, so almost all
bpf+tracing programs would need to increase it, since they use maps,
but kernel charges maximum map size upfront.
For example the hash map of 1024 elements will be charged as 64Kbyte.
It's inconvenient for current users and changes current behavior for root,
but probably worth doing to be consistent root vs non-root.
Similar accounting logic is done by mmap of perf_event.
Signed-off-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We can move bpf_map_ops and bpf_verifier_ops and other structs into ro
section, bpf_map_type_list and bpf_prog_type_list into read mostly.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
- fix NULL pointer dereference:
kernel/bpf/arraymap.c:41 array_map_alloc() error: potential null dereference 'array'. (kzalloc returns null)
kernel/bpf/arraymap.c:41 array_map_alloc() error: we previously assumed 'array' could be null (see line 40)
- integer overflow check was missing in arraymap
(hashmap checks for overflow via kmalloc_array())
- arraymap can round_up(value_size, 8) to zero. check was missing.
- hashmap was missing zero size check as well, since roundup_pow_of_two() can
truncate into zero
- found a typo in the arraymap comment and unnecessary empty line
Fix all of these issues and make both overflow checks explicit U32 in size.
Reported-by: kbuild test robot <fengguang.wu@intel.com>
Signed-off-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
add new map type BPF_MAP_TYPE_HASH and its implementation
- maps are created/destroyed by userspace. Both userspace and eBPF programs
can lookup/update/delete elements from the map
- eBPF programs can be called in_irq(), so use spin_lock_irqsave() mechanism
for concurrent updates
- key/value are opaque range of bytes (aligned to 8 bytes)
- user space provides 3 configuration attributes via BPF syscall:
key_size, value_size, max_entries
- map takes care of allocating/freeing key/value pairs
- map_update_elem() must fail to insert new element when max_entries
limit is reached to make sure that eBPF programs cannot exhaust memory
- map_update_elem() replaces elements in an atomic way
- optimized for speed of lookup() which can be called multiple times from
eBPF program which itself is triggered by high volume of events
. in the future JIT compiler may recognize lookup() call and optimize it
further, since key_size is constant for life of eBPF program
Signed-off-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jakub Kicinski
Eric Dumazet
David S. Miller
Chenbo Feng
Daniel Borkmann
Martin KaFai Lau
Teng Qin
Johannes Berg
Alexei Starovoitov
Michal Hocko
Sasha Levin
tom.leiming@gmail.com
Alexei Starovoitov
Yang Shi
Alexei Starovoitov