From: Aaron Merey <amerey@redhat.com>
To: systemtap@sourceware.org
Subject: Re: [PATCH] loc2stap.cxx: Add partial support for DW_OP_bra in DWARF location lists
Date: Tue, 23 Apr 2024 14:12:34 -0400 [thread overview]
Message-ID: <CAJDtP-Te-jABbf6kJk6FkmrksCyRVCdEUOvSGtrkuvL53f+aeA@mail.gmail.com> (raw)
In-Reply-To: <20240419005035.960853-1-amerey@redhat.com>
On Thu, Apr 18, 2024 at 8:50 PM Aaron Merey <amerey@redhat.com> wrote:
>
> Add support for DW_OP_bra when operand is non-negative. Previously
> systemtap would quit probe translation if DW_OP_bra was seen in a
> DWARF location list.
>
> Tested manually on RHEL 8.9 with kernel 4.18.0-513.24.1.el8_9.x86_64.
> Scripts containing a vfs.read probe require DW_OP_bra support when
> run with this kernel.
>
> Support for DW_OP_bra negative operands continues to be deferred due
> to lack of use as well as being more complex to implement.
>
> https://sourceware.org/bugzilla/show_bug.cgi?id=31496
> ---
> loc2stap.cxx | 26 ++++++++++++++++++--------
> 1 file changed, 18 insertions(+), 8 deletions(-)
>
> diff --git a/loc2stap.cxx b/loc2stap.cxx
> index 53316a480..883993040 100644
> --- a/loc2stap.cxx
> +++ b/loc2stap.cxx
> @@ -435,9 +435,26 @@ location_context::translate (const Dwarf_Op *expr, const size_t len,
> }
> break;
>
> - /* Control flow operations. */
> + /* Control flow operations. */
> + case DW_OP_bra:
> + {
> + POP(taken)
> + if (taken == 0)
> + break;
> + }
> + /* FALLTHROUGH */
> +
> case DW_OP_skip:
> {
> + /* A backward branch could lead to an infinite loop. So
> + punt it until we find we actually need it.
> +
> + To support backward branches, we could copy the stack,
> + recurse, and place both traces in arms of a
> + ternary_expression. */
> + if ((Dwarf_Sword) expr[i].number < 0)
> + DIE("backward branch in DW_OP operation not supported");
> +
> Dwarf_Off target = expr[i].offset + 3 + expr[i].number;
> while (i + 1 < len && expr[i + 1].offset < target)
> ++i;
> @@ -446,13 +463,6 @@ location_context::translate (const Dwarf_Op *expr, const size_t len,
> break;
> }
>
> - case DW_OP_bra:
> - // ??? Could copy the stack, recurse, and place both
> - // traces in arms of a ternary_expression. No point
> - // in doing that if this is never used, however.
> - DIE ("conditional branches not supported");
> - break;
> -
> /* Memory access. */
> case DW_OP_deref:
> // ??? Target sizeof, not host sizeof.
> --
> 2.43.0
Pushed as commit 22e146b57ec
Aaron
prev parent reply other threads:[~2024-04-23 18:12 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-04-19 0:50 Aaron Merey
2024-04-23 18:12 ` Aaron Merey [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=CAJDtP-Te-jABbf6kJk6FkmrksCyRVCdEUOvSGtrkuvL53f+aeA@mail.gmail.com \
--to=amerey@redhat.com \
--cc=systemtap@sourceware.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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).