(Maybe a good first-timer bug if anyone wants to try contributing during the holiday seasons)
The stable v6.6 kernel currently runs into kernel panic when running the test_progs tests from BPF selftests. Judging by the log it is failing in one of the dummy_st_ops tests (which comes after deny_namespace tests if you look at the output of `test_progs -l`). My guess is that it is related to "check bpf_dummy_struct_ops program params for test runs"[1], perhaps we're missing a commit or two.
Some notes for anyone tackling this for the first time: 1. You'll need to use the stable/linux-6.6.y branch from https://github.com/shunghsiyu/bpf. The current v6.6.66 one fails at compiling of BPF selftests[2] 2. The easiest way to run BPF selftests is to got relevant dependencies[3] installed, and run tools/testing/selftests/bpf/vmtest.sh (need to give it `-i` to download the root image first, and also might need to specify clang and llvm-strip by setting environmental variable CLANG=clang-17 and LLVM_STRIP=llvm-strip-17, respectively). For a more solid setup, see materials[4][5] from Manu Bretelle 3. Patch(es) should be send to stable@vger.kernel.org, following the stable process[6], see [2] as an example
Below is the output from vmtest.sh:
#68/1 deny_namespace/unpriv_userns_create_no_bpf:OK #68/2 deny_namespace/userns_create_bpf:OK #68 deny_namespace:OK [ 26.829153] BUG: kernel NULL pointer dereference, address: 0000000000000000 [ 26.831136] #PF: supervisor read access in kernel mode [ 26.832635] #PF: error_code(0x0000) - not-present page [ 26.833999] PGD 0 P4D 0 [ 26.834771] Oops: 0000 [#1] PREEMPT SMP PTI [ 26.835997] CPU: 2 PID: 119 Comm: test_progs Tainted: G OE 6.6.66-00003-gd80551078e71 #3 [ 26.838774] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.14.0-2 04/01/2014 [ 26.841152] RIP: 0010:bpf_prog_8ee9cbe7c9b5a50f_test_1+0x17/0x24 [ 26.842877] Code: 00 00 00 cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc f3 0f 1e fa 0f 1f 44 00 00 66 90 55 48 89 e5 f3 0f 1e fa 48 8b 7f 00 <8b> 47 00 be 5a 00 00 00 89 77 00 c9 c3 cc cc cc cc cc cc cc cc c0 [ 26.847953] RSP: 0018:ffff9e6b803b7d88 EFLAGS: 00010202 [ 26.849425] RAX: 0000000000000001 RBX: 0000000000000001 RCX: 2845e103d7dffb60 [ 26.851483] RDX: 0000000000000000 RSI: 0000000084d09025 RDI: 0000000000000000 [ 26.853508] RBP: ffff9e6b803b7d88 R08: 0000000000000001 R09: 0000000000000000 [ 26.855670] R10: 0000000000000000 R11: 0000000000000000 R12: ffff9754c0b5f700 [ 26.857824] R13: ffff9754c09cc800 R14: ffff9754c0b5f680 R15: ffff9754c0b5f760 [ 26.859741] FS: 00007f77dee12740(0000) GS:ffff9754fbc80000(0000) knlGS:0000000000000000 [ 26.862087] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 26.863705] CR2: 0000000000000000 CR3: 00000001020e6003 CR4: 0000000000170ee0 [ 26.865689] Call Trace: [ 26.866407] <TASK> [ 26.866982] ? __die+0x24/0x70 [ 26.867774] ? page_fault_oops+0x15b/0x450 [ 26.868882] ? search_bpf_extables+0xb0/0x160 [ 26.870076] ? fixup_exception+0x26/0x330 [ 26.871214] ? exc_page_fault+0x64/0x190 [ 26.872293] ? asm_exc_page_fault+0x26/0x30 [ 26.873352] ? bpf_prog_8ee9cbe7c9b5a50f_test_1+0x17/0x24 [ 26.874705] ? __bpf_prog_enter+0x3f/0xc0 [ 26.875718] ? bpf_struct_ops_test_run+0x1b8/0x2c0 [ 26.876942] ? __sys_bpf+0xc4e/0x2c30 [ 26.877898] ? __x64_sys_bpf+0x20/0x30 [ 26.878812] ? do_syscall_64+0x37/0x90 [ 26.879704] ? entry_SYSCALL_64_after_hwframe+0x78/0xe2 [ 26.880918] </TASK> [ 26.881409] Modules linked in: bpf_testmod(OE) [last unloaded: bpf_testmod(OE)] [ 26.883095] CR2: 0000000000000000 [ 26.883934] ---[ end trace 0000000000000000 ]--- [ 26.885099] RIP: 0010:bpf_prog_8ee9cbe7c9b5a50f_test_1+0x17/0x24 [ 26.886452] Code: 00 00 00 cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc f3 0f 1e fa 0f 1f 44 00 00 66 90 55 48 89 e5 f3 0f 1e fa 48 8b 7f 00 <8b> 47 00 be 5a 00 00 00 89 77 00 c9 c3 cc cc cc cc cc cc cc cc c0 [ 26.890379] RSP: 0018:ffff9e6b803b7d88 EFLAGS: 00010202 [ 26.891450] RAX: 0000000000000001 RBX: 0000000000000001 RCX: 2845e103d7dffb60 [ 26.892779] RDX: 0000000000000000 RSI: 0000000084d09025 RDI: 0000000000000000 [ 26.894254] RBP: ffff9e6b803b7d88 R08: 0000000000000001 R09: 0000000000000000 [ 26.895630] R10: 0000000000000000 R11: 0000000000000000 R12: ffff9754c0b5f700 [ 26.897008] R13: ffff9754c09cc800 R14: ffff9754c0b5f680 R15: ffff9754c0b5f760 [ 26.898337] FS: 00007f77dee12740(0000) GS:ffff9754fbc80000(0000) knlGS:0000000000000000 [ 26.899972] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 26.901076] CR2: 0000000000000000 CR3: 00000001020e6003 CR4: 0000000000170ee0 [ 26.902336] Kernel panic - not syncing: Fatal exception [ 26.903639] Kernel Offset: 0x36000000 from 0xffffffff81000000 (relocation range: 0xffffffff80000000-0xffffffffbfffffff) [ 26.905693] ---[ end Kernel panic - not syncing: Fatal exception ]---
1: https://lore.kernel.org/all/20240424012821.595216-1-eddyz87@gmail.com/t/#u 2: https://lore.kernel.org/all/20241217080240.46699-1-shung-hsi.yu@suse.com/t/#... 3: https://gist.github.com/shunghsiyu/1bd4189654cce5b3e55c2ab8da7dd33d#file-vmt... 4: https://chantra.github.io/bpfcitools/bpf-local-development.html 5: http://oldvger.kernel.org/bpfconf2024_material/BPF-dev-hacks.pdf 6: https://www.kernel.org/doc/html/latest/process/stable-kernel-rules.html
On Tue, Dec 17, 2024 at 11:02 PM Shung-Hsi Yu shung-hsi.yu@suse.com wrote:
(Maybe a good first-timer bug if anyone wants to try contributing during the holiday seasons)
The stable v6.6 kernel currently runs into kernel panic when running the
Does it repro with the latest kernel? If not, please make an effort and figure out which patch is missing in 6.6.
On Tue, Dec 17, 2024 at 11:27:33PM -0800, Alexei Starovoitov wrote:
On Tue, Dec 17, 2024 at 11:02 PM Shung-Hsi Yu shung-hsi.yu@suse.com wrote:
(Maybe a good first-timer bug if anyone wants to try contributing during the holiday seasons)
The stable v6.6 kernel currently runs into kernel panic when running the
Does it repro with the latest kernel?
No, doesn't happen on the latest kernel.
If not, please make an effort and figure out which patch is missing in 6.6.
I'm optimistically hoping this report would persuade someone reading the mailing list into to do the work :)
But if that doesn't happen I'll look into it and have it fixed.
Shung-Hsi
linux-stable-mirror@lists.linaro.org