Hello all,
This patch series targets a long-standing BPF usability issue - the lack
of general cross-compilation support - by enabling cross-endian usage of
libbpf and bpftool, as well as supporting cross-endian build targets for
selftests/bpf.
Benefits include improved BPF development and testing for embedded systems
based on e.g. big-endian MIPS, more build options e.g for s390x systems,
and better accessibility to the very latest test tools e.g. 'test_progs'.
The series touches many functional areas: BTF.ext handling; object access,
introspection, and linking; generation of normal and "light" skeletons.
Initial development and testing used mips64, since this arch makes
switching the build byte-order trivial and is thus very handy for A/B
testing. However, it lacks some key features (bpf2bpf call, kfuncs, etc)
making for poor selftests/bpf coverage.
Final testing takes the kernel and selftests/bpf cross-built from x86_64
to s390x, and runs the result under QEMU/s390x. That same configuration
could also be used on kernel-patches/bpf CI for regression testing endian
support or perhaps load-sharing s390x builds across x86_64 systems.
This thread includes some background regarding testing on QEMU/s390x and
the generally favourable results:
https://lore.kernel.org/bpf/ZsEcsaa3juxxQBUf@kodidev-ubuntu/
Earlier versions and related discussion of the series are here:
v1: https://lore.kernel.org/bpf/cover.1724216108.git.tony.ambardar@gmail.com/
v2: https://lore.kernel.org/bpf/cover.1724313164.git.tony.ambardar@gmail.com/
v3: https://lore.kernel.org/bpf/cover.1724843049.git.tony.ambardar@gmail.com/
v4: https://lore.kernel.org/bpf/cover.1724976539.git.tony.ambardar@gmail.com/
Feedback and suggestions are welcome!
Best regards,
Tony
Changelog:
---------
v4 -> v5: (feedback from Andrii and Eduard)
- add separate functions to byte-swap info metadata and records, and
ensure ordering so record bswaps occur when metadata is native endian
- use new and existing macros to iterate through info sections/records,
and check embedded record sizes match that of info structs used
- drop use of <cough> evil callbacks
- move setting swapped_endian flag to after byte-swapping functions are
called during initialization, allowing funcs to infer endianness and
drop a 'bool native' call parameter
- simplify byte-swapping macro used to generate light skeleton, and use
internal lib funcs to swap info records instead of assuming all __u32
- change info bswap library funcs to void return
- rework/consolidate new debug statements to reduce their number
- remove some unneeded handling of impossible errors, and drop a safety
check already handled elsewhere
- add and clarify some comments
v3 -> v4:
- fix a use-after-free ELF data-handling error causing rare CI failures
- move bswap functions for func/line/core-relo records to internal header
- use bswap functions also for info blobs in light skeleton
v2 -> v3: (feedback from Andrii)
- improve some log and commit message formatting
- restructure BTF.ext endianness safety checks and byte-swapping
- use BTF.ext info record definitions for swapping, require BTF v1
- follow BTF API implementation more closely for BTF.ext
- explicitly reject loading non-native endianness program into kernel
- simplify linker output byte-order setting
- drop redundant safety checks during linking
- simplify endianness macro and improve blob setup code for light skel
- no unexpected test failures after cross-compiling x86_64 -> s390x
v1 -> v2:
- fixed a light skeleton bug causing test_progs 'map_ptr' failure
- simplified some BTF.ext related endianness logic
- remove an 'inline' usage related to CI checkpatch failure
- improve some formatting noted by checkpatch warnings
- unexpected 'test_progs' failures drop 3 -> 2 (x86_64 to s390x cross)
Tony Ambardar (8):
libbpf: Improve log message formatting
libbpf: Fix header comment typos for BTF.ext
libbpf: Fix output .symtab byte-order during linking
libbpf: Support BTF.ext loading and output in either endianness
libbpf: Support opening bpf objects of either endianness
libbpf: Support linking bpf objects of either endianness
libbpf: Support creating light skeleton of either endianness
selftests/bpf: Support cross-endian building
tools/lib/bpf/bpf_gen_internal.h | 1 +
tools/lib/bpf/btf.c | 242 +++++++++++++++++++++++++--
tools/lib/bpf/btf.h | 3 +
tools/lib/bpf/btf_dump.c | 2 +-
tools/lib/bpf/btf_relocate.c | 2 +-
tools/lib/bpf/gen_loader.c | 191 +++++++++++++++------
tools/lib/bpf/libbpf.c | 57 +++++--
tools/lib/bpf/libbpf.map | 2 +
tools/lib/bpf/libbpf_internal.h | 43 ++++-
tools/lib/bpf/linker.c | 80 +++++++--
tools/lib/bpf/relo_core.c | 2 +-
tools/lib/bpf/skel_internal.h | 3 +-
tools/testing/selftests/bpf/Makefile | 7 +-
13 files changed, 529 insertions(+), 106 deletions(-)
--
2.34.1
Hello all,
This patch series targets a long-standing BPF usability issue - the lack
of general cross-compilation support - by enabling cross-endian usage of
libbpf and bpftool, as well as supporting cross-endian build targets for
selftests/bpf.
Benefits include improved BPF development and testing for embedded systems
based on e.g. big-endian MIPS, more build options e.g for s390x systems,
and better accessibility to the very latest test tools e.g. 'test_progs'.
Initial development and testing used mips64, since this arch makes
switching the build byte-order trivial and is thus very handy for A/B
testing. However, it lacks some key features (bpf2bpf call, kfuncs, etc)
making for poor selftests/bpf coverage.
Final testing takes the kernel and selftests/bpf cross-built from x86_64
to s390x, and runs the result under QEMU/s390x. That same configuration
could also be used on kernel-patches/bpf CI for regression testing endian
support or perhaps load-sharing s390x builds across x86_64 systems.
This thread includes some background regarding testing on QEMU/s390x and
the generally favourable results:
https://lore.kernel.org/bpf/ZsEcsaa3juxxQBUf@kodidev-ubuntu/
Feedback and suggestions are welcome!
Best regards,
Tony
Changelog:
---------
v3 -> v4:
- fix a use-after-free ELF data-handling error causing rare CI failures
- move bswap functions for func/line/core-relo records to internal header
- use bswap functions also for info blobs in light skeleton
v2 -> v3: (feedback from Andrii)
- improve some log and commit message formatting
- restructure BTF.ext endianness safety checks and byte-swapping
- use BTF.ext info record definitions for swapping, require BTF v1
- follow BTF API implementation more closely for BTF.ext
- explicitly reject loading non-native endianness program into kernel
- simplify linker output byte-order setting
- drop redundant safety checks during linking
- simplify endianness macro and improve blob setup code for light skel
- no unexpected test failures after cross-compiling x86_64 -> s390x
v1 -> v2:
- fixed a light skeleton bug causing test_progs 'map_ptr' failure
- simplified some BTF.ext related endianness logic
- remove an 'inline' usage related to CI checkpatch failure
- improve some formatting noted by checkpatch warnings
- unexpected 'test_progs' failures drop 3 -> 2 (x86_64 to s390x cross)
Tony Ambardar (8):
libbpf: Improve log message formatting
libbpf: Fix header comment typos for BTF.ext
libbpf: Fix output .symtab byte-order during linking
libbpf: Support BTF.ext loading and output in either endianness
libbpf: Support opening bpf objects of either endianness
libbpf: Support linking bpf objects of either endianness
libbpf: Support creating light skeleton of either endianness
selftests/bpf: Support cross-endian building
tools/lib/bpf/bpf_gen_internal.h | 1 +
tools/lib/bpf/btf.c | 196 ++++++++++++++++++++++++---
tools/lib/bpf/btf.h | 3 +
tools/lib/bpf/btf_dump.c | 2 +-
tools/lib/bpf/btf_relocate.c | 2 +-
tools/lib/bpf/gen_loader.c | 187 +++++++++++++++++++------
tools/lib/bpf/libbpf.c | 54 ++++++--
tools/lib/bpf/libbpf.map | 2 +
tools/lib/bpf/libbpf_internal.h | 48 ++++++-
tools/lib/bpf/linker.c | 92 ++++++++++---
tools/lib/bpf/relo_core.c | 2 +-
tools/lib/bpf/skel_internal.h | 3 +-
tools/testing/selftests/bpf/Makefile | 7 +-
13 files changed, 502 insertions(+), 97 deletions(-)
--
2.34.1
Hello,
kernel test robot noticed "kunit.list-kunit-test.fail" on:
commit: 17640748eb3875e486805e2d98ca1044a3a69b93 ("list: test: fix tests for list_cut_position()")
https://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git master
[test failed on linux-next/master 5acd9952f95fb4b7da6d09a3be39195a80845eb6]
in testcase: kunit
version:
with following parameters:
group: group-00
compiler: gcc-12
test machine: 8 threads Intel(R) Core(TM) i7-6700 CPU @ 3.40GHz (Skylake) with 16G memory
(please refer to attached dmesg/kmsg for entire log/backtrace)
If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <oliver.sang(a)intel.com>
| Closes: https://lore.kernel.org/oe-lkp/202409161554.6c3e8d5d-oliver.sang@intel.com
below 2 cases can pass on parent but fail on this commit.
13a6473783aaced3 17640748eb3875e486805e2d98c
---------------- ---------------------------
fail:runs %reproduction fail:runs
| | |
:9 67% 6:6 kunit.list-kunit-test.list_test_list_cut_before.fail
:9 67% 6:6 kunit.list-kunit-test.list_test_list_cut_position.fail
[ 143.881460] KTAP version 1
[ 143.884865] 1..3
[ 143.887802] KTAP version 1
[ 143.891541] # Subtest: list-kunit-test
[ 143.896316] # module: list_test
[ 143.896336] 1..39
[ 143.904046] ok 1 list_test_list_init
[ 143.904607] ok 2 list_test_list_add
[ 143.909698] ok 3 list_test_list_add_tail
[ 143.915034] ok 4 list_test_list_del
[ 143.920415] ok 5 list_test_list_replace
[ 143.925428] ok 6 list_test_list_replace_init
[ 143.930900] ok 7 list_test_list_swap
[ 143.936712] ok 8 list_test_list_del_init
[ 143.941952] ok 9 list_test_list_del_init_careful
[ 143.947363] ok 10 list_test_list_move
[ 143.953422] ok 11 list_test_list_move_tail
[ 143.958536] ok 12 list_test_list_bulk_move_tail
[ 143.964192] ok 13 list_test_list_is_head
[ 143.970279] ok 14 list_test_list_is_first
[ 143.975777] ok 15 list_test_list_is_last
[ 143.981871] ok 16 list_test_list_empty
[ 143.987375] ok 17 list_test_list_empty_careful
[ 143.992611] ok 18 list_test_list_rotate_left
[ 143.998511] ok 19 list_test_list_rotate_to_front
[ 144.004367] ok 20 list_test_list_is_singular
[ 144.010375] # list_test_list_cut_position: EXPECTATION FAILED at lib/list-test.c:409
Expected cur == &entries[i], but
cur == ffffc9000126fd70
&entries[i] == ffffc9000126fd50
[ 144.016196] not ok 21 list_test_list_cut_position
[ 144.040950] # list_test_list_cut_before: EXPECTATION FAILED at lib/list-test.c:440
Expected cur == &entries[i], but
cur == ffffc9000129fd60
&entries[i] == ffffc9000129fd50
[ 144.047033] # list_test_list_cut_before: EXPECTATION FAILED at lib/list-test.c:440
Expected cur == &entries[i], but
cur == ffffc9000129fd70
&entries[i] == ffffc9000129fd60
[ 144.071616] not ok 22 list_test_list_cut_before
[ 144.096387] ok 23 list_test_list_splice
[ 144.102425] ok 24 list_test_list_splice_tail
[ 144.107897] ok 25 list_test_list_splice_init
[ 144.113704] ok 26 list_test_list_splice_tail_init
[ 144.119664] ok 27 list_test_list_entry
[ 144.126022] ok 28 list_test_list_entry_is_head
[ 144.131275] ok 29 list_test_list_first_entry
[ 144.137246] ok 30 list_test_list_last_entry
[ 144.143084] ok 31 list_test_list_first_entry_or_null
[ 144.148823] ok 32 list_test_list_next_entry
[ 144.155283] ok 33 list_test_list_prev_entry
[ 144.161048] ok 34 list_test_list_for_each
[ 144.166792] ok 35 list_test_list_for_each_prev
[ 144.172318] ok 36 list_test_list_for_each_safe
[ 144.178278] ok 37 list_test_list_for_each_prev_safe
[ 144.184235] ok 38 list_test_list_for_each_entry
[ 144.190607] ok 39 list_test_list_for_each_entry_reverse
[ 144.196199] # list-kunit-test: pass:37 fail:2 skip:0 total:39
[ 144.202470] # Totals: pass:37 fail:2 skip:0 total:39
[ 144.208908] not ok 1 list-kunit-test
The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20240916/202409161554.6c3e8d5d-oliv…
--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
This patch allows progs to elide a null check on statically known map
lookup keys. In other words, if the verifier can statically prove that
the lookup will be in-bounds, allow the prog to drop the null check.
This is useful for two reasons:
1. Large numbers of nullness checks (especially when they cannot fail)
unnecessarily pushes prog towards BPF_COMPLEXITY_LIMIT_JMP_SEQ.
2. It forms a tighter contract between programmer and verifier.
For (1), bpftrace is starting to make heavier use of percpu scratch
maps. As a result, for user scripts with large number of unrolled loops,
we are starting to hit jump complexity verification errors. These
percpu lookups cannot fail anyways, as we only use static key values.
Eliding nullness probably results in less work for verifier as well.
For (2), percpu scratch maps are often used as a larger stack, as the
currrent stack is limited to 512 bytes. In these situations, it is
desirable for the programmer to express: "this lookup should never fail,
and if it does, it means I messed up the code". By omitting the null
check, the programmer can "ask" the verifier to double check the logic.
Daniel Xu (2):
bpf: verifier: Support eliding map lookup nullness
bpf: selftests: verifier: Add nullness elision tests
kernel/bpf/verifier.c | 56 +++++++
.../bpf/progs/verifier_array_access.c | 143 ++++++++++++++++++
2 files changed, 199 insertions(+)
--
2.46.0
Running this test on a small system produces different failures every
test checking deletions, and some flushes. From different test runs:
TEST: Common host entries configuration tests (L2) [FAIL]
Failed to delete L2 host entry
TEST: Common port group entries configuration tests (IPv4 (S, G)) [FAIL]
IPv4 (S, G) entry with VLAN 10 not deleted when VLAN was not specified
TEST: Common port group entries configuration tests (IPv6 (*, G)) [FAIL]
IPv6 (*, G) entry with VLAN 10 not deleted when VLAN was not specified
TEST: Flush tests [FAIL]
Entry not flushed by specified VLAN ID
TEST: Flush tests [FAIL]
IPv6 host entry not flushed by "nopermanent" state
Add a short sleep after deletion and flush to resolve this.
Tested using 25 test runs in a row, resulting in 100% pass OK.
Create a variable just for this test to allow short sleep, the default
WAIT_TIME of 5 seconds makes the test far longer than necessary.
Fixes: b6d00da08610 ("selftests: forwarding: Add bridge MDB test")
Signed-off-by: Jamie Bainbridge <jamie.bainbridge(a)gmail.com>
---
.../selftests/net/forwarding/bridge_mdb.sh | 32 +++++++++++++++++++
1 file changed, 32 insertions(+)
diff --git a/tools/testing/selftests/net/forwarding/bridge_mdb.sh b/tools/testing/selftests/net/forwarding/bridge_mdb.sh
index d9d587454d207931a539f59be15cbc63d471888f..b3a2a7bc1824f4c394267b283b89e7a3ae19b0fb 100755
--- a/tools/testing/selftests/net/forwarding/bridge_mdb.sh
+++ b/tools/testing/selftests/net/forwarding/bridge_mdb.sh
@@ -30,6 +30,9 @@ ALL_TESTS="
ctrl_test
"
+# time to wait for delete and flush to complete
+: "${SETTLE_DELAY:=0.1}"
+
NUM_NETIFS=4
source lib.sh
source tc_common.sh
@@ -152,6 +155,7 @@ cfg_test_host_common()
check_fail $? "Managed to replace $name host entry"
bridge mdb del dev br0 port br0 grp $grp $state vid 10
+ sleep "$SETTLE_DELAY"
bridge mdb get dev br0 grp $grp vid 10 &> /dev/null
check_fail $? "Failed to delete $name host entry"
@@ -208,6 +212,7 @@ cfg_test_port_common()
check_err $? "Failed to replace $name entry"
bridge mdb del dev br0 port $swp1 $grp_key permanent vid 10
+ sleep "$SETTLE_DELAY"
bridge mdb get dev br0 $grp_key vid 10 &> /dev/null
check_fail $? "Failed to delete $name entry"
@@ -230,6 +235,7 @@ cfg_test_port_common()
check_err $? "$name entry with VLAN 20 not added when VLAN was not specified"
bridge mdb del dev br0 port $swp1 $grp_key permanent
+ sleep "$SETTLE_DELAY"
bridge mdb get dev br0 $grp_key vid 10 &> /dev/null
check_fail $? "$name entry with VLAN 10 not deleted when VLAN was not specified"
bridge mdb get dev br0 $grp_key vid 20 &> /dev/null
@@ -310,6 +316,7 @@ __cfg_test_port_ip_star_g()
bridge -d mdb get dev br0 grp $grp src $src1 vid 10 &> /dev/null
check_err $? "(S, G) entry not created"
bridge mdb del dev br0 port $swp1 grp $grp vid 10
+ sleep "$SETTLE_DELAY"
bridge -d mdb get dev br0 grp $grp vid 10 &> /dev/null
check_fail $? "(*, G) entry not deleted"
bridge -d mdb get dev br0 grp $grp src $src1 vid 10 &> /dev/null
@@ -828,6 +835,7 @@ cfg_test_flush()
bridge mdb add dev br0 port $swp1 grp 239.1.1.8 vid 10 temp
bridge mdb flush dev br0
+ sleep "$SETTLE_DELAY"
num_entries=$(bridge mdb show dev br0 | wc -l)
[[ $num_entries -eq 0 ]]
check_err $? 0 "Not all entries flushed after flush all"
@@ -840,6 +848,7 @@ cfg_test_flush()
bridge mdb add dev br0 port br0 grp 239.1.1.1 vid 10
bridge mdb flush dev br0 port $swp1
+ sleep "$SETTLE_DELAY"
bridge mdb get dev br0 grp 239.1.1.1 vid 10 | grep -q "port $swp1"
check_fail $? "Entry not flushed by specified port"
@@ -849,11 +858,13 @@ cfg_test_flush()
check_err $? "Host entry flushed by wrong port"
bridge mdb flush dev br0 port br0
+ sleep "$SETTLE_DELAY"
bridge mdb get dev br0 grp 239.1.1.1 vid 10 | grep -q "port br0"
check_fail $? "Host entry not flushed by specified port"
bridge mdb flush dev br0
+ sleep "$SETTLE_DELAY"
# Check that when flushing by VLAN ID only entries programmed with the
# specified VLAN ID are flushed and the rest are not.
@@ -864,6 +875,7 @@ cfg_test_flush()
bridge mdb add dev br0 port $swp2 grp 239.1.1.1 vid 20
bridge mdb flush dev br0 vid 10
+ sleep "$SETTLE_DELAY"
bridge mdb get dev br0 grp 239.1.1.1 vid 10 &> /dev/null
check_fail $? "Entry not flushed by specified VLAN ID"
@@ -871,6 +883,7 @@ cfg_test_flush()
check_err $? "Entry flushed by wrong VLAN ID"
bridge mdb flush dev br0
+ sleep "$SETTLE_DELAY"
# Check that all permanent entries are flushed when "permanent" is
# specified and that temporary entries are not.
@@ -879,6 +892,7 @@ cfg_test_flush()
bridge mdb add dev br0 port $swp2 grp 239.1.1.1 temp vid 10
bridge mdb flush dev br0 permanent
+ sleep "$SETTLE_DELAY"
bridge mdb get dev br0 grp 239.1.1.1 vid 10 | grep -q "port $swp1"
check_fail $? "Entry not flushed by \"permanent\" state"
@@ -886,6 +900,7 @@ cfg_test_flush()
check_err $? "Entry flushed by wrong state (\"permanent\")"
bridge mdb flush dev br0
+ sleep "$SETTLE_DELAY"
# Check that all temporary entries are flushed when "nopermanent" is
# specified and that permanent entries are not.
@@ -894,6 +909,7 @@ cfg_test_flush()
bridge mdb add dev br0 port $swp2 grp 239.1.1.1 temp vid 10
bridge mdb flush dev br0 nopermanent
+ sleep "$SETTLE_DELAY"
bridge mdb get dev br0 grp 239.1.1.1 vid 10 | grep -q "port $swp1"
check_err $? "Entry flushed by wrong state (\"nopermanent\")"
@@ -901,6 +917,7 @@ cfg_test_flush()
check_fail $? "Entry not flushed by \"nopermanent\" state"
bridge mdb flush dev br0
+ sleep "$SETTLE_DELAY"
# Check that L2 host entries are not flushed when "nopermanent" is
# specified, but flushed when "permanent" is specified.
@@ -908,16 +925,19 @@ cfg_test_flush()
bridge mdb add dev br0 port br0 grp 01:02:03:04:05:06 permanent vid 10
bridge mdb flush dev br0 nopermanent
+ sleep "$SETTLE_DELAY"
bridge mdb get dev br0 grp 01:02:03:04:05:06 vid 10 &> /dev/null
check_err $? "L2 host entry flushed by wrong state (\"nopermanent\")"
bridge mdb flush dev br0 permanent
+ sleep "$SETTLE_DELAY"
bridge mdb get dev br0 grp 01:02:03:04:05:06 vid 10 &> /dev/null
check_fail $? "L2 host entry not flushed by \"permanent\" state"
bridge mdb flush dev br0
+ sleep "$SETTLE_DELAY"
# Check that IPv4 host entries are not flushed when "permanent" is
# specified, but flushed when "nopermanent" is specified.
@@ -925,16 +945,19 @@ cfg_test_flush()
bridge mdb add dev br0 port br0 grp 239.1.1.1 temp vid 10
bridge mdb flush dev br0 permanent
+ sleep "$SETTLE_DELAY"
bridge mdb get dev br0 grp 239.1.1.1 vid 10 &> /dev/null
check_err $? "IPv4 host entry flushed by wrong state (\"permanent\")"
bridge mdb flush dev br0 nopermanent
+ sleep "$SETTLE_DELAY"
bridge mdb get dev br0 grp 239.1.1.1 vid 10 &> /dev/null
check_fail $? "IPv4 host entry not flushed by \"nopermanent\" state"
bridge mdb flush dev br0
+ sleep "$SETTLE_DELAY"
# Check that IPv6 host entries are not flushed when "permanent" is
# specified, but flushed when "nopermanent" is specified.
@@ -942,16 +965,19 @@ cfg_test_flush()
bridge mdb add dev br0 port br0 grp ff0e::1 temp vid 10
bridge mdb flush dev br0 permanent
+ sleep "$SETTLE_DELAY"
bridge mdb get dev br0 grp ff0e::1 vid 10 &> /dev/null
check_err $? "IPv6 host entry flushed by wrong state (\"permanent\")"
bridge mdb flush dev br0 nopermanent
+ sleep "$SETTLE_DELAY"
bridge mdb get dev br0 grp ff0e::1 vid 10 &> /dev/null
check_fail $? "IPv6 host entry not flushed by \"nopermanent\" state"
bridge mdb flush dev br0
+ sleep "$SETTLE_DELAY"
# Check that when flushing by routing protocol only entries programmed
# with the specified routing protocol are flushed and the rest are not.
@@ -961,6 +987,7 @@ cfg_test_flush()
bridge mdb add dev br0 port br0 grp 239.1.1.1 vid 10
bridge mdb flush dev br0 proto bgp
+ sleep "$SETTLE_DELAY"
bridge mdb get dev br0 grp 239.1.1.1 vid 10 | grep -q "port $swp1"
check_fail $? "Entry not flushed by specified routing protocol"
@@ -970,20 +997,25 @@ cfg_test_flush()
check_err $? "Host entry flushed by wrong routing protocol"
bridge mdb flush dev br0
+ sleep "$SETTLE_DELAY"
# Test that an error is returned when trying to flush using unsupported
# parameters.
bridge mdb flush dev br0 src_vni 10 &> /dev/null
+ sleep "$SETTLE_DELAY"
check_fail $? "Managed to flush by source VNI"
bridge mdb flush dev br0 dst 198.51.100.1 &> /dev/null
+ sleep "$SETTLE_DELAY"
check_fail $? "Managed to flush by destination IP"
bridge mdb flush dev br0 dst_port 4789 &> /dev/null
+ sleep "$SETTLE_DELAY"
check_fail $? "Managed to flush by UDP destination port"
bridge mdb flush dev br0 vni 10 &> /dev/null
+ sleep "$SETTLE_DELAY"
check_fail $? "Managed to flush by destination VNI"
log_test "Flush tests"
--
2.39.2
This is a slight change from the fundamentals of HID-BPF.
In theory, HID-BPF is abstract to the kernel itself, and makes
only changes at the HID level (through report descriptors or
events emitted to/from the device).
However, we have seen a few use cases where HID-BPF might interact with
the running kernel when the target device is already handled by a
specific device.
For example, the XP-Pen/Huion/UC-Logic tablets are handled by
hid-uclogic but this driver is also doing a report descriptor fixup
without checking if the device has already been fixed by HID-BPF.
In the same way, another recent example[0] was when a cheap foot pedal is
used and tricks iPhones and Windows machines by presenting itself as a
known Apple wireless keyboard. The problem is that this fake keyboard is
not presenting a compatible report descriptor and hid-core merges all
device nodes together making libinput ignore the keyboard part for
historical reasons.
Last, there has been a long standing request to allow to disable the
input part of a given gamepad while SDL or Steam opens the device
through hidraw.
This series aims at tackling both of these problems:
- first we had a new hook `hid_bpf_driver_probe` which allows the BPF
program to decide if the curently probed driver should be used or not
- then this same hook can also change the ->driver_data of the struct
hid_device_id argument, and we teach hid-generic to use that field as
the connect mask.
Basically, it means that when we insert a BPF program to fix a device,
we can force hid-generic to handle the device, and thus preventing
any other kernel driver to tamper with our device. We can also
selectively decide to export the hidraw or input nodes when using
hid-generic.
In the SDL/Steam use case, this would means that the gaming application
will load one BPF program per input device it wants to open through
hidraw, that BPF program reassigns the input device to hid-generic and
disables hid-input, then it can open the new hidraw node.
Once that program terminates, the BPF program is removed (either
automatically because no-one has the fd of the links open, or manually
by SDL/Steam), and the normal driver rebinds to the HID device,
restoring full input functionality.
This branch is on top of the for-6.12/hidraw and for-6.12/constify-rdesc
branches of hid.git, mainly because those branch would conflict otherwise.
[0] https://gitlab.freedesktop.org/libinput/libinput/-/issues/1014
Signed-off-by: Benjamin Tissoires <bentiss(a)kernel.org>
---
Changes in v2:
- Refactored the API to not use a new hook but hid_bpf_rdesc_fixup
instead
- Some cleanups in hid-core.c probe() device to not kmemdup multiple
time the report descriptor when it's not required
- I'm still not 100% sure the HID_QUIRK_IGNORE_HIDINPUT is that
required, but I can not think of anything else at the moment to
temporary disable any driver input device.
- Link to v1: https://lore.kernel.org/r/20240903-hid-bpf-hid-generic-v1-0-9511a565b2da@ke…
---
Benjamin Tissoires (11):
HID: bpf: move HID-BPF report descriptor fixup earlier
HID: core: save one kmemdup during .probe()
HID: core: remove one more kmemdup on .probe()
HID: bpf: allow write access to quirks field in struct hid_device
selftests/hid: add dependency on hid_common.h
selftests/hid: cleanup C tests by adding a common struct uhid_device
selftests/hid: allow to parametrize bus/vid/pid/rdesc on the test device
HID: add per device quirk to force bind to hid-generic
selftests/hid: add test for assigning a given device to hid-generic
HID: add quirk to prevent hid-input to be used
selftests/hid: add test to disable hid-input
drivers/hid/bpf/hid_bpf_dispatch.c | 8 +-
drivers/hid/bpf/hid_bpf_struct_ops.c | 1 +
drivers/hid/hid-core.c | 72 ++++++--
drivers/hid/hid-generic.c | 3 +
include/linux/hid.h | 22 ++-
include/linux/hid_bpf.h | 9 +-
tools/testing/selftests/hid/Makefile | 2 +-
tools/testing/selftests/hid/hid_bpf.c | 205 ++++++++++++++++-----
tools/testing/selftests/hid/hid_common.h | 112 +++++++----
tools/testing/selftests/hid/hidraw.c | 36 +---
tools/testing/selftests/hid/progs/hid.c | 13 ++
.../testing/selftests/hid/progs/hid_bpf_helpers.h | 7 +-
12 files changed, 343 insertions(+), 147 deletions(-)
---
base-commit: e1370d5de7b755600df050979e19fbcd625fb4c6
change-id: 20240829-hid-bpf-hid-generic-61579f5b5945
Best regards,
--
Benjamin Tissoires <bentiss(a)kernel.org>
In this series from Geliang, modifying MPTCP BPF selftests, we have:
- A new MPTCP subflow BPF program setting socket options per subflow: it
looks better to have this old test program in the BPF selftests to
track regressions and to serve as example.
Note: Nicolas is no longer working at Tessares, but he did this work
while working for them, and his email address is no longer available.
- A new hook in the same BPF program to do the verification step.
- A new MPTCP BPF subtest validating the new BPF program added in the
first patch, with the help of the new hook added in the second patch.
Signed-off-by: Matthieu Baerts (NGI0) <matttbe(a)kernel.org>
---
Changes in v6:
- Patch 3/3: use usleep() instead of sleep()
- Series: rebased on top of bpf-next/net
- Link to v5: https://lore.kernel.org/r/20240910-upstream-bpf-next-20240506-mptcp-subflow…
Changes in v5:
- See the individual changelog for more details about them
- Patch 1/3: set TCP on the 2nd subflow
- Patch 2/3: new
- Patch 3/3: use the BPF program from patch 2/3 to do the validation
instead of using ss.
- Series: rebased on top of bpf-next/net
- Link to v4: https://lore.kernel.org/r/20240805-upstream-bpf-next-20240506-mptcp-subflow…
Changes in v4:
- Drop former patch 2/3: MPTCP's pm_nl_ctl requires a new header file:
- I will check later if it is possible to avoid having duplicated
header files in tools/include/uapi, but no need to block this series
for that. Patch 2/3 can be added later if needed.
- Patch 2/2: skip the test if 'ip mptcp' is not available.
- Link to v3: https://lore.kernel.org/r/20240703-upstream-bpf-next-20240506-mptcp-subflow…
Changes in v3:
- Sorry for the delay between v2 and v3, this series was conflicting
with the "add netns helpers", but it looks like it is on hold:
https://lore.kernel.org/cover.1715821541.git.tanggeliang@kylinos.cn
- Patch 1/3 includes "bpf_tracing_net.h", introduced in between.
- New patch 2/3: "selftests/bpf: Add mptcp pm_nl_ctl link".
- Patch 3/3: use the tool introduced in patch 2/3 + SYS_NOFAIL() helper.
- Link to v2: https://lore.kernel.org/r/20240509-upstream-bpf-next-20240506-mptcp-subflow…
Changes in v2:
- Previous patches 1/4 and 2/4 have been dropped from this series:
- 1/4: "selftests/bpf: Handle SIGINT when creating netns":
- A new version, more generic and no longer specific to MPTCP BPF
selftest will be sent later, as part of a new series. (Alexei)
- 2/4: "selftests/bpf: Add RUN_MPTCP_TEST macro":
- Removed, not to hide helper functions in macros. (Alexei)
- The commit message of patch 1/2 has been clarified to avoid some
possible confusions spot by Alexei.
- Link to v1: https://lore.kernel.org/r/20240507-upstream-bpf-next-20240506-mptcp-subflow…
---
Geliang Tang (2):
selftests/bpf: Add getsockopt to inspect mptcp subflow
selftests/bpf: Add mptcp subflow subtest
Nicolas Rybowski (1):
selftests/bpf: Add mptcp subflow example
MAINTAINERS | 2 +-
tools/testing/selftests/bpf/prog_tests/mptcp.c | 127 +++++++++++++++++++++
tools/testing/selftests/bpf/progs/mptcp_bpf.h | 42 +++++++
tools/testing/selftests/bpf/progs/mptcp_subflow.c | 128 ++++++++++++++++++++++
4 files changed, 298 insertions(+), 1 deletion(-)
---
base-commit: 23dc9867329c72b48e5039ac93fbf50d9099cdb3
change-id: 20240506-upstream-bpf-next-20240506-mptcp-subflow-test-faef6654bfa3
Best regards,
--
Matthieu Baerts (NGI0) <matttbe(a)kernel.org>