BPF List
 help / color / mirror / Atom feed
From: Stanislav Fomichev <sdf@fomichev.me>
To: Zac Ecob <zacecob@protonmail.com>
Cc: "bpf@vger.kernel.org" <bpf@vger.kernel.org>
Subject: Re: Fixing coerce_subreg_to_size_sx invalidly setting reg->umax_value
Date: Thu, 11 Jul 2024 18:16:05 -0700	[thread overview]
Message-ID: <ZpCD1fDXkr7wLGhZ@mini-arch> (raw)
In-Reply-To: <h3qKLDEO6m9nhif0eAQX4fVrqdO0D_OPb0y5HfMK9jBePEKK33wQ3K-bqSVnr0hiZdFZtSJOsbNkcEQGpv_yJk61PAAiO8fUkgMRSO-lB50=@protonmail.com>

On 07/11, Zac Ecob wrote:
> Hi,
> 
> My fuzzer recently found another bug, in which `reg->umax_value` is being invalidly set in regards to sign extensions.
> 
> The lines below contain the bug:
> ```
> reg->umin_value = reg->u32_min_value = s64_min;                                                             
> reg->umax_value = reg->u32_max_value = s64_max;
> ```
> 
> If `s64_min` / `s64_max` are negative values here, they correctly cast when assigning to the u32 values. However, when assigned to `umin_value` / `umax_value`, it seems there is an implicit (u32) cast applied, causing the top 32 bits to not be set.
> 
> 
> I've attached the files to reproduce, as well as the patch file, based off of 6.10-rc4 - albeit this is my first patch so I'd appreciate someone checking it's formatted fine.
> 
> Thanks.


> From da5ef523f7cd018f3f0991454a18bc961ea1abba Mon Sep 17 00:00:00 2001
> From: Zac Ecob <zacecob@protonmail.com>
> Date: Thu, 11 Jul 2024 17:41:55 +1000
> Subject: [PATCH] Fixed sign-extension issue in coerce_subreg_to_size_sx
> 
> ---
>  kernel/bpf/verifier.c | 10 ++++++++--
>  1 file changed, 8 insertions(+), 2 deletions(-)
> 
> diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
> index 010a6eb864dc..eccf3ac8996a 100644
> --- a/kernel/bpf/verifier.c
> +++ b/kernel/bpf/verifier.c
> @@ -6213,8 +6213,14 @@ static void coerce_reg_to_size_sx(struct bpf_reg_state *reg, int size)
>  	if ((s64_max >= 0) == (s64_min >= 0)) {
>  		reg->smin_value = reg->s32_min_value = s64_min;
>  		reg->smax_value = reg->s32_max_value = s64_max;
> -		reg->umin_value = reg->u32_min_value = s64_min;
> -		reg->umax_value = reg->u32_max_value = s64_max;
> +
> +		// Cannot chain assignments, like reg->umax_val = reg->u32_max_val = (signed input)
> +		// Because of the implicit cast leading to reg->umax_val not being properly set for negative numbers

Pls use /* */ comments instead, use [PATCH bpf] subject in a followup and
try to find a commit that introduced the problem to mention it in the
'Fixes:' tag.

Also, instead of your custom reproducer, can you add a small reproducer
to the test_verifier.c (tools/testing/selftests/bpf/verifier/**) to
demonstrate the issue and avoid similar regressions in the future?

> +		reg->u32_min_value = s64_min;
> +		reg->u32_max_value = s64_max;
> +		reg->umin_value    = s64_min;
> +		reg->umax_value    = s64_max;
> +
>  		reg->var_off = tnum_range(s64_min, s64_max);
>  		return;
>  	}
> -- 
> 2.30.2
> 


      reply	other threads:[~2024-07-12  1:16 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-07-11  8:07 Fixing coerce_subreg_to_size_sx invalidly setting reg->umax_value Zac Ecob
2024-07-12  1:16 ` Stanislav Fomichev [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=ZpCD1fDXkr7wLGhZ@mini-arch \
    --to=sdf@fomichev.me \
    --cc=bpf@vger.kernel.org \
    --cc=zacecob@protonmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox