Re: [PATCH 11/14] bpf/tests: Add test for 32-bit context pointer argument passing

From: Yonghong Song <yhs@fb.com>
To: Johan Almbladh <johan.almbladh@anyfinetworks.com>,
	<ast@kernel.org>, <daniel@iogearbox.net>, <andrii@kernel.org>
Cc: <kafai@fb.com>, <songliubraving@fb.com>,
	<john.fastabend@gmail.com>, <kpsingh@kernel.org>,
	<Tony.Ambardar@gmail.com>, <netdev@vger.kernel.org>,
	<bpf@vger.kernel.org>
Subject: Re: [PATCH 11/14] bpf/tests: Add test for 32-bit context pointer argument passing
Date: Wed, 28 Jul 2021 17:09:45 -0700	[thread overview]
Message-ID: <deee75a2-a4ce-303c-981a-cd5b6e8cecdf@fb.com> (raw)
In-Reply-To: <20210728170502.351010-12-johan.almbladh@anyfinetworks.com>

On 7/28/21 10:04 AM, Johan Almbladh wrote:
> On a 32-bit architecture, the context pointer should occupy the low
> half of R0, and the other half should be zero.

I think this is probably true. The word choice "should" indicates
this doesn't need to be the case if people choose a different
implementation, right?

> 
> Signed-off-by: Johan Almbladh <johan.almbladh@anyfinetworks.com>
> ---
>   lib/test_bpf.c | 16 ++++++++++++++++
>   1 file changed, 16 insertions(+)
> 
> diff --git a/lib/test_bpf.c b/lib/test_bpf.c
> index 55914b6236aa..314af6eaeb92 100644
> --- a/lib/test_bpf.c
> +++ b/lib/test_bpf.c
> @@ -2084,6 +2084,22 @@ static struct bpf_test tests[] = {
>   #undef NUMER
>   #undef DENOM
>   	},
> +#ifdef CONFIG_32BIT
> +	{
> +		"INT: 32-bit context pointer word order and zero-extension",
> +		.u.insns_int = {
> +			BPF_ALU32_IMM(BPF_MOV, R0, 0),
> +			BPF_JMP32_IMM(BPF_JEQ, R1, 0, 3),
> +			BPF_ALU64_IMM(BPF_RSH, R1, 32),
> +			BPF_JMP32_IMM(BPF_JNE, R1, 0, 1),
> +			BPF_ALU32_IMM(BPF_MOV, R0, 1),
> +			BPF_EXIT_INSN(),
> +		},
> +		INTERNAL,
> +		{ },
> +		{ { 0, 1 } }
> +	},
> +#endif
>   	{
>   		"check: missing ret",
>   		.u.insns = {
> 