kernel.git/kernel/bpf, branch linux-6.15.y

bpf: Make reg_not_null() true for CONST_PTR_TO_MAP

2025-08-20T16:36:14Z

[ Upstream commit 5534e58f2e9bd72b253d033ee0af6e68eb8ac96b ] When reg->type is CONST_PTR_TO_MAP, it can not be null. However the verifier explores the branches under rX == 0 in check_cond_jmp_op() even if reg->type is CONST_PTR_TO_MAP, because it was not checked for in reg_not_null(). Fix this by adding CONST_PTR_TO_MAP to the set of types that are considered non nullable in reg_not_null(). An old "unpriv: cmp map pointer with zero" selftest fails with this change, because now early out correctly triggers in check_cond_jmp_op(), making the verification to pass. In practice verifier may allow pointer to null comparison in unpriv, since in many cases the relevant branch and comparison op are removed as dead code. So change the expected test result to __success_unpriv. Signed-off-by: Ihor Solodrai Signed-off-by: Andrii Nakryiko Acked-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20250609183024.359974-2-isolodrai@meta.com Signed-off-by: Sasha Levin

bpf: Forget ranges when refining tnum after JSET

2025-08-20T16:36:05Z

[ Upstream commit 6279846b9b2532e1b04559ef8bd0dec049f29383 ] Syzbot reported a kernel warning due to a range invariant violation on the following BPF program. 0: call bpf_get_netns_cookie 1: if r0 == 0 goto 2: if r0 & Oxffffffff goto The issue is on the path where we fall through both jumps. That path is unreachable at runtime: after insn 1, we know r0 != 0, but with the sign extension on the jset, we would only fallthrough insn 2 if r0 == 0. Unfortunately, is_branch_taken() isn't currently able to figure this out, so the verifier walks all branches. The verifier then refines the register bounds using the second condition and we end up with inconsistent bounds on this unreachable path: 1: if r0 == 0 goto r0: u64=[0x1, 0xffffffffffffffff] var_off=(0, 0xffffffffffffffff) 2: if r0 & 0xffffffff goto r0 before reg_bounds_sync: u64=[0x1, 0xffffffffffffffff] var_off=(0, 0) r0 after reg_bounds_sync: u64=[0x1, 0] var_off=(0, 0) Improving the range refinement for JSET to cover all cases is tricky. We also don't expect many users to rely on JSET given LLVM doesn't generate those instructions. So instead of improving the range refinement for JSETs, Eduard suggested we forget the ranges whenever we're narrowing tnums after a JSET. This patch implements that approach. Reported-by: syzbot+c711ce17dd78e5d4fdcf@syzkaller.appspotmail.com Suggested-by: Eduard Zingerman Acked-by: Yonghong Song Acked-by: Eduard Zingerman Signed-off-by: Paul Chaignon Link: https://lore.kernel.org/r/9d4fd6432a095d281f815770608fdcd16028ce0b.1752171365.git.paul.chaignon@gmail.com Signed-off-by: Alexei Starovoitov Signed-off-by: Sasha Levin

bpf/preload: Don't select USERMODE_DRIVER

2025-08-15T10:16:37Z

[ Upstream commit 2b03164eee20eac7ce0fe3aa4fbda7efc1e5427a ] The usermode driver framework is not used anymore by the BPF preload code. Fixes: cb80ddc67152 ("bpf: Convert bpf_preload.ko to use light skeleton.") Signed-off-by: Thomas Weißschuh Signed-off-by: Daniel Borkmann Reviewed-by: Christoph Hellwig Link: https://lore.kernel.org/bpf/20250721-remove-usermode-driver-v1-1-0d0083334382@linutronix.de Signed-off-by: Sasha Levin

bpf: Ensure RCU lock is held around bpf_prog_ksym_find

2025-08-15T10:16:27Z

[ Upstream commit d090326860096df9dac6f27cff76d3f8df44d4f1 ] Add a warning to ensure RCU lock is held around tree lookup, and then fix one of the invocations in bpf_stack_walker. The program has an active stack frame and won't disappear. Use the opportunity to remove unneeded invocation of is_bpf_text_address. Fixes: f18b03fabaa9 ("bpf: Implement BPF exceptions") Reviewed-by: Emil Tsalapatis Signed-off-by: Kumar Kartikeya Dwivedi Link: https://lore.kernel.org/r/20250703204818.925464-5-memxor@gmail.com Signed-off-by: Alexei Starovoitov Signed-off-by: Sasha Levin

bpf: handle jset (if a & b ...) as a jump in CFG computation

2025-08-15T10:16:24Z

[ Upstream commit 3157f7e2999616ac91f4d559a8566214f74000a5 ] BPF_JSET is a conditional jump and currently verifier.c:can_jump() does not know about that. This can lead to incorrect live registers and SCC computation. E.g. in the following example: 1: r0 = 1; 2: r2 = 2; 3: if r1 & 0x7 goto +1; 4: exit; 5: r0 = r2; 6: exit; W/o this fix insn_successors(3) will return only (4), a jump to (5) would be missed and r2 won't be marked as alive at (3). Fixes: 14c8552db644 ("bpf: simple DFA-based live registers analysis") Reported-by: syzbot+a36aac327960ff474804@syzkaller.appspotmail.com Suggested-by: Alexei Starovoitov Signed-off-by: Eduard Zingerman Link: https://lore.kernel.org/r/20250613175331.3238739-1-eddyz87@gmail.com Signed-off-by: Alexei Starovoitov Signed-off-by: Sasha Levin

selftests/bpf: Add tests with stack ptr register in conditional jmp

2025-08-01T08:51:28Z

commit 5ffb537e416ee22dbfb3d552102e50da33fec7f6 upstream. Add two tests: - one test has 'rX r10' where rX is not r10, and - another test has 'rX rY' where rX and rY are not r10 but there is an early insn 'rX = r10'. Without previous verifier change, both tests will fail. Signed-off-by: Yonghong Song Signed-off-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20250524041340.4046304-1-yonghong.song@linux.dev [ shung-hsi.yu: contains additional hunks for kernel/bpf/verifier.c that should be part of the previous patch in the series, commit e2d2115e56c4 "bpf: Do not include stack ptr register in precision backtracking bookkeeping", which already incorporated. ] Link: https://lore.kernel.org/all/9b41f9f5-396f-47e0-9a12-46c52087df6c@linux.dev/ Signed-off-by: Shung-Hsi Yu Signed-off-by: Greg Kroah-Hartman

bpf: Reject %p% format string in bprintf-like helpers

2025-07-24T06:58:33Z

[ Upstream commit f8242745871f81a3ac37f9f51853d12854fd0b58 ] static const char fmt[] = "%p%"; bpf_trace_printk(fmt, sizeof(fmt)); The above BPF program isn't rejected and causes a kernel warning at runtime: Please remove unsupported %\x00 in format string WARNING: CPU: 1 PID: 7244 at lib/vsprintf.c:2680 format_decode+0x49c/0x5d0 This happens because bpf_bprintf_prepare skips over the second %, detected as punctuation, while processing %p. This patch fixes it by not skipping over punctuation. %\x00 is then processed in the next iteration and rejected. Reported-by: syzbot+e2c932aec5c8a6e1d31c@syzkaller.appspotmail.com Fixes: 48cac3f4a96d ("bpf: Implement formatted output helpers with bstr_printf") Acked-by: Yonghong Song Signed-off-by: Paul Chaignon Link: https://lore.kernel.org/r/a0e06cc479faec9e802ae51ba5d66420523251ee.1751395489.git.paul.chaignon@gmail.com Signed-off-by: Alexei Starovoitov Signed-off-by: Sasha Levin

bpf: Adjust free target to avoid global starvation of LRU map

2025-07-17T16:44:03Z

[ Upstream commit d4adf1c9ee7722545450608bcb095fb31512f0c6 ] BPF_MAP_TYPE_LRU_HASH can recycle most recent elements well before the map is full, due to percpu reservations and force shrink before neighbor stealing. Once a CPU is unable to borrow from the global map, it will once steal one elem from a neighbor and after that each time flush this one element to the global list and immediately recycle it. Batch value LOCAL_FREE_TARGET (128) will exhaust a 10K element map with 79 CPUs. CPU 79 will observe this behavior even while its neighbors hold 78 * 127 + 1 * 15 == 9921 free elements (99%). CPUs need not be active concurrently. The issue can appear with affinity migration, e.g., irqbalance. Each CPU can reserve and then hold onto its 128 elements indefinitely. Avoid global list exhaustion by limiting aggregate percpu caches to half of map size, by adjusting LOCAL_FREE_TARGET based on cpu count. This change has no effect on sufficiently large tables. Similar to LOCAL_NR_SCANS and lru->nr_scans, introduce a map variable lru->free_target. The extra field fits in a hole in struct bpf_lru. The cacheline is already warm where read in the hot path. The field is only accessed with the lru lock held. Tested-by: Anton Protopopov Signed-off-by: Willem de Bruijn Acked-by: Stanislav Fomichev Link: https://lore.kernel.org/r/20250618215803.3587312-1-willemdebruijn.kernel@gmail.com Signed-off-by: Alexei Starovoitov Signed-off-by: Sasha Levin

bpf: Use proper type to calculate bpf_raw_tp_null_args.mask index

2025-06-27T10:13:26Z

[ Upstream commit 53ebef53a657d7957d35dc2b953db64f1bb28065 ] The calculation of the index used to access the mask field in 'struct bpf_raw_tp_null_args' is done with 'int' type, which could overflow when the tracepoint being attached has more than 8 arguments. While none of the tracepoints mentioned in raw_tp_null_args[] currently have more than 8 arguments, there do exist tracepoints that had more than 8 arguments (e.g. iocost_iocg_forgive_debt), so use the correct type for calculation and avoid Smatch static checker warning. Reported-by: Dan Carpenter Signed-off-by: Shung-Hsi Yu Signed-off-by: Andrii Nakryiko Acked-by: Kumar Kartikeya Dwivedi Link: https://lore.kernel.org/bpf/20250418074946.35569-1-shung-hsi.yu@suse.com Closes: https://lore.kernel.org/r/843a3b94-d53d-42db-93d4-be10a4090146@stanley.mountain/ Signed-off-by: Sasha Levin

bpf: Pass the same orig_call value to trampoline functions

2025-06-27T10:13:23Z

[ Upstream commit 94bde253d3ae5d8a01cb958663b12daef1d06574 ] There is currently some confusion in the s390x JIT regarding whether orig_call can be NULL and what that means. Originally the NULL value was used to distinguish the struct_ops case, but this was superseded by BPF_TRAMP_F_INDIRECT (see commit 0c970ed2f87c ("s390/bpf: Fix indirect trampoline generation"). The remaining reason to have this check is that NULL can actually be passed to the arch_bpf_trampoline_size() call - but not to the respective arch_prepare_bpf_trampoline()! call - by bpf_struct_ops_prepare_trampoline(). Remove this asymmetry by passing stub_func to both functions, so that JITs may rely on orig_call never being NULL. Signed-off-by: Ilya Leoshkevich Acked-by: Martin KaFai Lau Link: https://lore.kernel.org/r/20250512221911.61314-2-iii@linux.ibm.com Signed-off-by: Alexei Starovoitov Signed-off-by: Sasha Levin