public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug debug/113000] New: Generate BTF always in early-finish regardless of CO-RE
@ 2023-12-13 18:09 jemarch at gcc dot gnu.org
  2023-12-14 13:48 ` [Bug debug/113000] " rguenth at gcc dot gnu.org
  0 siblings, 1 reply; 2+ messages in thread
From: jemarch at gcc dot gnu.org @ 2023-12-13 18:09 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113000

            Bug ID: 113000
           Summary: Generate BTF always in early-finish regardless of
                    CO-RE
           Product: gcc
           Version: unknown
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: debug
          Assignee: unassigned at gcc dot gnu.org
          Reporter: jemarch at gcc dot gnu.org
  Target Milestone: ---

The BPF CO-RE support requires storing information relevant to the BTF debug
info in the .BTF.ext section.  In particular:

 * The CO-RE relocation strings contain references to BTF types, by type ID.
 * The CO-RE information includes strings that are stored in the BTF string
section.

Because of the above, we are generating BTF (call to ctf_debug_finish) in
dwarf2out's early_finish if no -mco-re is specified, but the call is done in
dwarf2out's finish instead if -mco-re is specified.  This implies that there
are differences in the output (also in CTF) because for example entries for
inlined functions are no longer present at 'finish' time but they are at
'early_finish' time.

We should change GCC so it always calls ctf_debug_finish in early_finish,
regardless of whether CO-RE relocs are being generated or not.

^ permalink raw reply	[flat|nested] 2+ messages in thread

* [Bug debug/113000] Generate BTF always in early-finish regardless of CO-RE
  2023-12-13 18:09 [Bug debug/113000] New: Generate BTF always in early-finish regardless of CO-RE jemarch at gcc dot gnu.org
@ 2023-12-14 13:48 ` rguenth at gcc dot gnu.org
  0 siblings, 0 replies; 2+ messages in thread
From: rguenth at gcc dot gnu.org @ 2023-12-14 13:48 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113000

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Note you then need a way to cope with LTO.

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2023-12-14 13:48 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-12-13 18:09 [Bug debug/113000] New: Generate BTF always in early-finish regardless of CO-RE jemarch at gcc dot gnu.org
2023-12-14 13:48 ` [Bug debug/113000] " rguenth at gcc dot gnu.org

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).