From: Eduard Zingerman <eddyz87@gmail.com>
To: Alexei Starovoitov <alexei.starovoitov@gmail.com>,
bpf@vger.kernel.org, andrii@kernel.org
Cc: daniel@iogearbox.net, martin.lau@linux.dev, kernel-team@fb.com,
yonghong.song@linux.dev, memxor@gmail.com
Subject: Re: [PATCH bpf-next 01/36] bpf: track may_write flags in liveness
Date: Sun, 27 Sep 2026 01:43:58 -0700 [thread overview]
Message-ID: <0b62d64c669ef14c527758291ad2bdff43b20196.camel@gmail.com> (raw)
In-Reply-To: <DLPD9DRXXOGK.1QIUC9VKTZU08@gmail.com>
On Sat, 2026-09-26 at 15:51 +0000, Alexei Starovoitov wrote:
> On Sat, Sep 26, 2026 at 07:19 AM Eduard Zingerman <eddyz87@gmail.com> wrote:
> > Compared to must_write, may_write includes any possibly modified slot,
> > including partial writes, and does not kill liveness.
>
> It doesn't include the writes done by helpers and kfuncs.
> bpf_helper_stack_access_bytes() returns -size only for MEM_UNINIT args.
> bpf_fib_lookup() has ARG_PTR_TO_MEM | MEM_WRITE for params, so
> record_arg_access() gets +size and the call has 'use' and no 'may_def',
> though the helper writes into params.
> Same for raw uninit mem when !allow_uninit_stack and for every
> kfunc mem arg that is not __uninit.
>
> > -__msg(" 6: (85) call bpf_iter_num_new{{.*}} ; def: fp0-24{{$}}")
> > +__msg(" 6: (85) call bpf_iter_num_new{{.*}} ; def: fp0-24 may_def: fp0-24{{$}}")
> > __msg(" 9: (85) call bpf_iter_num_next{{.*}} ; use: fp0-24{{$}}")
>
> bpf_iter_num_next() is 'use' only too.
>
> Returning a write as a read was the safe direction for liveness.
> For may_write it's the opposite.
> What does SCEV do with a slot that has no may_def inside the loop?
Missing may_write means that SCEV for the slot won't be invalidated on
indirect write that touches it. Which would lead to incorrect SCEV
expression.
This is a good point, I need to fix this bug and collect updated stats.
I think it's the main take-away from the reviews so far.
next prev parent reply other threads:[~2026-09-27 8:44 UTC|newest]
Thread overview: 80+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-09-26 14:19 [PATCH bpf-next 00/36] bpf: use scalar evolution to widen bounded loops Eduard Zingerman
2026-09-26 14:19 ` [PATCH bpf-next 01/36] bpf: track may_write flags in liveness Eduard Zingerman
2026-09-26 15:51 ` Alexei Starovoitov
2026-09-27 8:43 ` Eduard Zingerman [this message]
2026-09-27 20:42 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 02/36] bpf: summarize may write stack slots in insn_aux_data Eduard Zingerman
2026-09-26 15:51 ` Alexei Starovoitov
2026-09-29 20:12 ` Eduard Zingerman
2026-09-27 20:42 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 03/36] bpf: summarize live " Eduard Zingerman
2026-09-26 14:33 ` sashiko-bot
2026-09-27 20:26 ` bot+bpf-ci
2026-09-29 18:16 ` Eduard Zingerman
2026-09-26 14:20 ` [PATCH bpf-next 04/36] bpf: summarize regs that may hold a frame pointer " Eduard Zingerman
2026-09-27 20:27 ` bot+bpf-ci
2026-09-29 20:21 ` Eduard Zingerman
2026-09-26 14:20 ` [PATCH bpf-next 05/36] bpf: record write effects for atomic operations in liveness.c Eduard Zingerman
2026-09-27 20:27 ` bot+bpf-ci
2026-09-29 20:26 ` Eduard Zingerman
2026-09-26 14:20 ` [PATCH bpf-next 06/36] bpf: add tnum_alignment() Eduard Zingerman
2026-09-26 14:20 ` [PATCH bpf-next 07/36] bpf: add cnum{32,64}_union() Eduard Zingerman
2026-09-26 14:20 ` [PATCH bpf-next 08/36] bpf: add cnum64_intersect_linear() Eduard Zingerman
2026-09-26 14:34 ` sashiko-bot
2026-09-29 21:46 ` Eduard Zingerman
2026-09-26 14:20 ` [PATCH bpf-next 09/36] bpf: add bpf_set_reg_range() Eduard Zingerman
2026-09-26 14:36 ` sashiko-bot
2026-09-26 14:20 ` [PATCH bpf-next 10/36] bpf: add bpf_mark_reg_known_scalar() Eduard Zingerman
2026-09-26 14:20 ` [PATCH bpf-next 11/36] bpf: add bpf_reg_union() Eduard Zingerman
2026-09-27 20:42 ` bot+bpf-ci
2026-09-30 0:09 ` Eduard Zingerman
2026-09-26 14:20 ` [PATCH bpf-next 12/36] bpf: expose comparison opcode transformations Eduard Zingerman
2026-09-27 20:26 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 13/36] bpf: allow subrange relations for PTR_TO_STACK in regsafe() Eduard Zingerman
2026-09-27 20:42 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 14/36] bpf: representation for intervals with steps Eduard Zingerman
2026-09-26 14:35 ` sashiko-bot
2026-09-26 14:20 ` [PATCH bpf-next 15/36] bpf: varying offset access support for PTR_TO_BTF_ID pointers Eduard Zingerman
2026-09-26 14:37 ` sashiko-bot
2026-09-27 20:42 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 16/36] bpf: save DFS postorder numbers for program instructions Eduard Zingerman
2026-09-26 14:31 ` sashiko-bot
2026-09-26 14:20 ` [PATCH bpf-next 17/36] bpf: move the live-register and SCC printout to a standalone function Eduard Zingerman
2026-09-27 20:26 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 18/36] bpf: compute immediate dominators Eduard Zingerman
2026-09-26 15:54 ` Alexei Starovoitov
2026-09-27 20:42 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 19/36] bpf: compute loop hierarchy Eduard Zingerman
2026-09-27 20:43 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 20/36] bpf: add a min-heap for ordered analysis worklists Eduard Zingerman
2026-09-27 20:26 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 21/36] bpf: record basic-block ends in insn_aux_data Eduard Zingerman
2026-09-27 20:26 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 22/36] bpf: add bpf_split_cur_state() Eduard Zingerman
2026-09-26 14:20 ` [PATCH bpf-next 23/36] bpf: allow precision backtracking between overlapping checkpoints Eduard Zingerman
2026-09-27 20:27 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 24/36] bpf: compute scalar evolution expressions for loops Eduard Zingerman
2026-09-26 14:38 ` sashiko-bot
2026-09-27 20:43 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 25/36] bpf: use SCEV to widen bounded loops Eduard Zingerman
2026-09-26 14:42 ` sashiko-bot
2026-09-27 20:43 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 26/36] bpf: avoid widening registers that hinder exact stack-slot tracking Eduard Zingerman
2026-09-26 14:46 ` sashiko-bot
2026-09-27 20:43 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 27/36] selftests/bpf: __msg_next tag for matching messages on consecutive lines Eduard Zingerman
2026-09-26 14:20 ` [PATCH bpf-next 28/36] selftests/bpf: test for stack-pointer subrange pruning Eduard Zingerman
2026-09-26 14:32 ` sashiko-bot
2026-09-27 20:26 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 29/36] selftests/bpf: tests for may_write stack-liveness tracking Eduard Zingerman
2026-09-26 14:20 ` [PATCH bpf-next 30/36] selftests/bpf: tests for may_def marks of atomic RMW operations Eduard Zingerman
2026-09-26 14:20 ` [PATCH bpf-next 31/36] selftests/bpf: tests for register base/step arithmetic Eduard Zingerman
2026-09-26 14:20 ` [PATCH bpf-next 32/36] selftests/bpf: tests for register base/step state pruning Eduard Zingerman
2026-09-27 20:27 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 33/36] selftests/bpf: tests for varying offset access to PTR_TO_BTF_ID Eduard Zingerman
2026-09-27 20:42 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 34/36] selftests/bpf: tests for loop hierarchy computation Eduard Zingerman
2026-09-26 14:20 ` [PATCH bpf-next 35/36] selftests/bpf: tests for immediate dominator computation Eduard Zingerman
2026-09-27 20:27 ` bot+bpf-ci
2026-09-26 14:20 ` [PATCH bpf-next 36/36] selftests/bpf: cover SCEV analysis and loop widening Eduard Zingerman
2026-09-27 20:42 ` bot+bpf-ci
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=0b62d64c669ef14c527758291ad2bdff43b20196.camel@gmail.com \
--to=eddyz87@gmail.com \
--cc=alexei.starovoitov@gmail.com \
--cc=andrii@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=daniel@iogearbox.net \
--cc=kernel-team@fb.com \
--cc=martin.lau@linux.dev \
--cc=memxor@gmail.com \
--cc=yonghong.song@linux.dev \
/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