Re: Fw: [Bug 205469] New: x86_32: bpf: multiple test_bpf failures using eBPF JIT

From: Daniel Borkmann <daniel@iogearbox.net>
To: Wang YanQing <udknight@gmail.com>
Cc: Stephen Hemminger <stephen@networkplumber.org>,
	Alexei Starovoitov <ast@kernel.org>,
	Martin KaFai Lau <kafai@fb.com>, Song Liu <songliubraving@fb.com>,
	Yonghong Song <yhs@fb.com>,
	bpf@vger.kernel.org
Subject: Re: Fw: [Bug 205469] New: x86_32: bpf: multiple test_bpf failures using eBPF JIT
Date: Mon, 11 Nov 2019 13:56:31 +0100	[thread overview]
Message-ID: <69c1fa5e-7385-fe8d-ac17-42d22db84cf4@iogearbox.net> (raw)
In-Reply-To: <20191109183602.GA1033@udknight>

On 11/9/19 7:36 PM, Wang YanQing wrote:
> On Sat, Nov 09, 2019 at 12:37:49AM +0100, Daniel Borkmann wrote:
>> [ Cc Wang (x86_32 BPF JIT maintainer) ]
>>
>> On 11/8/19 4:57 PM, Stephen Hemminger wrote:
>>>
>>> Begin forwarded message:
>>>
>>> Date: Fri, 08 Nov 2019 07:35:59 +0000
>>> From: bugzilla-daemon@bugzilla.kernel.org
>>> To: stephen@networkplumber.org
>>> Subject: [Bug 205469] New: x86_32: bpf: multiple test_bpf failures using eBPF JIT
>>>
>>>
>>> https://bugzilla.kernel.org/show_bug.cgi?id=205469
>>>
>>>               Bug ID: 205469
>>>              Summary: x86_32: bpf: multiple test_bpf failures using eBPF JIT
>>>              Product: Networking
>>>              Version: 2.5
>>>       Kernel Version: 4.19.81 LTS
>>>             Hardware: i386
>>>                   OS: Linux
>>>                 Tree: Mainline
>>>               Status: NEW
>>>             Severity: normal
>>>             Priority: P1
>>>            Component: Other
>>>             Assignee: stephen@networkplumber.org
>>>             Reporter: itugrok@yahoo.com
>>>                   CC: itugrok@yahoo.com
>>>           Regression: No
>>>
>>> Created attachment 285829
>>>     --> https://bugzilla.kernel.org/attachment.cgi?id=285829&action=edit
>>> test_bpf failures: kernel 4.19.81/x86_32 (OpenWrt)
>>>
>>> Summary:
>>> ========
>>>
>>> Running the 4.19.81 LTS kernel on QEMU/x86_32, the standard test_bpf.ko
>>> testsuite generates multiple errors with the eBPF JIT enabled:
>>>
>>>     ...
>>>     test_bpf: #32 JSET jited:1 40 ret 0 != 20 46 FAIL
>>>     test_bpf: #321 LD_IND word positive offset jited:1 ret 0 != -291897430 FAIL
>>>     test_bpf: #322 LD_IND word negative offset jited:1 ret 0 != -1437222042 FAIL
>>>     test_bpf: #323 LD_IND word unaligned (addr & 3 == 2) jited:1 ret 0 !=
>>> -1150890889 FAIL
>>>     test_bpf: #326 LD_IND word positive offset, all ff jited:1 ret 0 != -1 FAIL
>>>     ...
>>>     test_bpf: Summary: 373 PASSED, 5 FAILED, [344/366 JIT'ed]
>>>
>>> However, with eBPF JIT disabled (net.core.bpf_jit_enable=0) all tests pass.
>>>
>>>
>>> Steps to Reproduce:
>>> ===================
>>>
>>>     # sysctl net.core.bpf_jit_enable=1
>>>     # modprobe test_bpf
>>>     <Kernel log with failures and test summary>
>>>
>>>
>>> Affected Systems Tested:
>>> ========================
>>>
>>>     OpenWrt master on QEMU/pc-q35(x86_32) [LTS kernel 4.19.81]
>>>
>>>
>>> Kernel Logs:
>>> ============
>>>
>>> Boot log with test results is attached.
>>>
> 
> Hi Daniel Borkmann!
> 
> I have tested and verified that the report bug has been fixed by commit
> 711aef1bbf88 ("bpf, x32: Fix bug for BPF_JMP | {BPF_JSGT, BPF_JSLE, BPF_JSLT, BPF_JSGE}")
> 
> But that fix hasn't been backport to stable trees, so maybe we should do it:)

Yes, given you have access to a x32 setup and are also able to runtime test the backported
JIT changes, please submit it to stable with us in Cc. Thanks Wang!

Best,
Daniel