From: Shung-Hsi Yu <shung-hsi.yu@suse.com>
To: Andrii Nakryiko <andrii@kernel.org>
Cc: bpf@vger.kernel.org, ast@kernel.org, daniel@iogearbox.net,
martin.lau@kernel.org, kernel-team@meta.com,
Eduard Zingerman <eddyz87@gmail.com>
Subject: Re: [PATCH v6 bpf-next 09/17] bpf: drop knowledge-losing __reg_combine_{32,64}_into_{64,32} logic
Date: Thu, 2 Nov 2023 23:14:15 +0800 [thread overview]
Message-ID: <ZUO8xwx8o_PO8pkD@u94a> (raw)
In-Reply-To: <20231102033759.2541186-10-andrii@kernel.org>
On Wed, Nov 01, 2023 at 08:37:51PM -0700, Andrii Nakryiko wrote:
> When performing 32-bit conditional operation operating on lower 32 bits
> of a full 64-bit register, register full value isn't changed. We just
> potentially gain new knowledge about that register's lower 32 bits.
>
> Unfortunately, __reg_combine_{32,64}_into_{64,32} logic that
> reg_set_min_max() performs as a last step, can lose information in some
> cases due to __mark_reg64_unbounded() and __reg_assign_32_into_64().
> That's bad and completely unnecessary. Especially __reg_assign_32_into_64()
> looks completely out of place here, because we are not performing
> zero-extending subregister assignment during conditional jump.
>
> So this patch replaced __reg_combine_* with just a normal
> reg_bounds_sync() which will do a proper job of deriving u64/s64 bounds
> from u32/s32, and vice versa (among all other combinations).
>
> __reg_combine_64_into_32() is also used in one more place,
> coerce_reg_to_size(), while handling 1- and 2-byte register loads.
> Looking into this, it seems like besides marking subregister as
> unbounded before performing reg_bounds_sync(), we were also performing
> deduction of smin32/smax32 and umin32/umax32 bounds from respective
> smin/smax and umin/umax bounds. It's now redundant as reg_bounds_sync()
> performs all the same logic more generically (e.g., without unnecessary
> assumption that upper 32 bits of full register should be zero).
>
> Long story short, we remove __reg_combine_64_into_32() completely, and
> coerce_reg_to_size() now only does resetting subreg to unbounded and then
> performing reg_bounds_sync() to recover as much information as possible
> from 64-bit umin/umax and smin/smax bounds, set explicitly in
> coerce_reg_to_size() earlier.
>
> Acked-by: Eduard Zingerman <eddyz87@gmail.com>
> Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
LGTM. Seeing __mark_reg{64,32}_unbounded() removed had me spooked
quite a bit though :)
Acked-by: Shung-Hsi Yu <shung-hsi.yu@suse.com>
next prev parent reply other threads:[~2023-11-02 15:14 UTC|newest]
Thread overview: 25+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-11-02 3:37 [PATCH v6 bpf-next 00/17] BPF register bounds logic and testing improvements Andrii Nakryiko
2023-11-02 3:37 ` [PATCH v6 bpf-next 01/17] selftests/bpf: fix RELEASE=1 build for tc_opts Andrii Nakryiko
2023-11-02 3:37 ` [PATCH v6 bpf-next 02/17] selftests/bpf: satisfy compiler by having explicit return in btf test Andrii Nakryiko
2023-11-02 3:37 ` [PATCH v6 bpf-next 03/17] bpf: derive smin/smax from umin/max bounds Andrii Nakryiko
2023-11-02 3:37 ` [PATCH v6 bpf-next 04/17] bpf: derive smin32/smax32 from umin32/umax32 bounds Andrii Nakryiko
2023-11-02 3:37 ` [PATCH v6 bpf-next 05/17] bpf: derive subreg bounds from full bounds when upper 32 bits are constant Andrii Nakryiko
2023-11-02 3:37 ` [PATCH v6 bpf-next 06/17] bpf: add special smin32/smax32 derivation from 64-bit bounds Andrii Nakryiko
2023-11-02 3:37 ` [PATCH v6 bpf-next 07/17] bpf: improve deduction of 64-bit bounds from 32-bit bounds Andrii Nakryiko
2023-11-02 14:39 ` Shung-Hsi Yu
2023-11-02 16:17 ` Andrii Nakryiko
2023-11-03 3:43 ` Shung-Hsi Yu
2023-11-02 3:37 ` [PATCH v6 bpf-next 08/17] bpf: try harder to deduce register bounds from different numeric domains Andrii Nakryiko
2023-11-02 3:37 ` [PATCH v6 bpf-next 09/17] bpf: drop knowledge-losing __reg_combine_{32,64}_into_{64,32} logic Andrii Nakryiko
2023-11-02 15:14 ` Shung-Hsi Yu [this message]
2023-11-02 3:37 ` [PATCH v6 bpf-next 10/17] selftests/bpf: BPF register range bounds tester Andrii Nakryiko
2023-11-02 3:37 ` [PATCH v6 bpf-next 11/17] bpf: rename is_branch_taken reg arguments to prepare for the second one Andrii Nakryiko
2023-11-02 15:15 ` Shung-Hsi Yu
2023-11-02 3:37 ` [PATCH v6 bpf-next 12/17] bpf: generalize is_branch_taken() to work with two registers Andrii Nakryiko
2023-11-02 15:19 ` Shung-Hsi Yu
2023-11-02 3:37 ` [PATCH v6 bpf-next 13/17] bpf: move is_branch_taken() down Andrii Nakryiko
2023-11-02 3:37 ` [PATCH v6 bpf-next 14/17] bpf: generalize is_branch_taken to handle all conditional jumps in one place Andrii Nakryiko
2023-11-02 3:37 ` [PATCH v6 bpf-next 15/17] bpf: unify 32-bit and 64-bit is_branch_taken logic Andrii Nakryiko
2023-11-02 3:37 ` [PATCH v6 bpf-next 16/17] bpf: prepare reg_set_min_max for second set of registers Andrii Nakryiko
2023-11-02 3:37 ` [PATCH v6 bpf-next 17/17] bpf: generalize reg_set_min_max() to handle two sets of two registers Andrii Nakryiko
2023-11-02 16:10 ` [PATCH v6 bpf-next 00/17] BPF register bounds logic and testing improvements patchwork-bot+netdevbpf
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=ZUO8xwx8o_PO8pkD@u94a \
--to=shung-hsi.yu@suse.com \
--cc=andrii@kernel.org \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=daniel@iogearbox.net \
--cc=eddyz87@gmail.com \
--cc=kernel-team@meta.com \
--cc=martin.lau@kernel.org \
/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