selftests/bpf: Fix chained_global_func_calls_success() for cpu v4
Subsystems (3)
Yonghong Song
reviewerPATCH 1/1 selftests/bpf: Fix chained_global_func_calls_success() for cpu v4
The chained_global_func_calls_success() test hardcodes the instruction counts reported by the verifier's per-subprog stats:
subprog {{[0-9]+}} (global_good) global insns_self 5 insns_total 5 stack processed 14 insns
global_good() does 'return arr[0]', where arr[] is an int array and the return type is long. Without cpu v4 this is a zero-extending load followed by a <<32/s>>32 sign-extension pair. With -mcpu=v4 llvm emits a single sign-extending load instead:
18: (18) r1 = 0xffa00000008eb000
20: (81) r0 = *(s32 *)(r1 +0)
21: (95) exit
so the subprog is 3 insns rather than 5, and the whole program is 12 processed insns rather than 14. test_progs-cpuv4 fails with:
EXPECTED REGEX: 'subprog {{[0-9]+}} (global_good) global insns_self 5 insns_total 5 stack'
#606/1 verifier_global_subprogs/chained_global_func_calls_success:FAIL
Select the expected counts based on __BPF_CPU_VERSION__.
--- tools/testing/selftests/bpf/progs/verifier_global_subprogs.c | 5 +++++ 1 file changed, 5 insertions(+)
tools/testing/selftests/bpf/progs/verifier_global_subprogs.c +5 -0
@@ -54,8 +54,13 @@ __msg("Validating global_good() func") __msg("('global_good') is safe for any args that match its prototype") __msg("subprog 0 (chained_global_func_calls_success) main insns_self 7 insns_total 7 stack") __msg("subprog {{[0-9]+}} (global_calls_good_only) global insns_self 2 insns_total 2 stack") +#if defined(__BPF_CPU_VERSION__) && __BPF_CPU_VERSION__ >=4 +__msg("subprog {{[0-9]+}} (global_good) global insns_self 3 insns_total 3 stack") +__msg("processed 12 insns") +#else __msg("subprog {{[0-9]+}} (global_good) global insns_self 5 insns_total 5 stack") __msg("processed 14 insns") +#endif int chained_global_func_calls_success (void ) {int sum =0 ;
Signature · 1 line
2.53.0-Meta
This patch was applied to bpf/bpf-next.git (master) by Daniel Borkmann <[email protected]>:
Hello:
This patch was applied to bpf/bpf-next.git (master) by Daniel Borkmann <[email protected]>:
Quoted 6 lines · On Thu, 13 Aug 2026 08:06:41 -0700 you wrote:
The
chained_global_func_calls_success()test hardcodes the instruction counts reported by the verifier's per-subprog stats:subprog {{[0-9]+}} (
global_good) globalinsns_self5insns_total5 stack processed 14 insns
Quoted 6 lines
global_good()does 'return arr[0]', where arr[] is an int array and the return type is long. Without cpu v4 this is a zero-extending load followed by a <<32/s>>32 sign-extension pair. With -mcpu=v4 llvm emits a single sign-extending load instead:[...]
Here is the summary with links:
- selftests/bpf: Fix chained_global_func_calls_success() for cpu v4
https://git.kernel.org/bpf/bpf-next/c/9786d424a36b
You are awesome, thank you!
Signature · 2 lines
Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html