public inbox for gdb-patches@sourceware.org
 help / color / mirror / Atom feed
From: Andrew Burgess <aburgess@redhat.com>
To: Bruno Larsen <blarsen@redhat.com>, gdb-patches@sourceware.org
Subject: Re: [PATCH 08/11] gdb/testsuite: disable gdb.cp/typeid.exp when using clang
Date: Wed, 26 Oct 2022 16:37:37 +0100	[thread overview]
Message-ID: <87pmee61fi.fsf@redhat.com> (raw)
In-Reply-To: <20221004170747.154307-10-blarsen@redhat.com>

Bruno Larsen via Gdb-patches <gdb-patches@sourceware.org> writes:

> Since clang chooses to not add any debug information from standard
> libraries, gdb.cp/typeid.exp will always fail, so this test is skipped
> when it clang used.

OK, so I see what you're saying.  But I don't think it follows that we
should skip the whole test.  Also, without digging into the test
further, and looking at the generated symbols, I think the above
explanation makes things sound worse than they are.

So, here's some of the output for 'nm -C' for the test binary (built
with Clang):

  0000000000003b70 V typeinfo for VB2
  0000000000003b38 V typeinfo for Base
  0000000000003b98 V typeinfo for Derived
  0000000000003b28 V typeinfo for char [5]
                   U typeinfo for int@@CXXABI_1.3
  0000000000003bd0 V typeinfo for Base*
                   U typeinfo for char*@@CXXABI_1.3
                   U typeinfo for char const*@@CXXABI_1.3
  0000000000002012 V typeinfo name for VB1
  000000000000201d V typeinfo name for VB2
  0000000000002017 V typeinfo name for Base
  0000000000002009 V typeinfo name for Derived
  0000000000002004 V typeinfo name for char [5]
  0000000000002022 V typeinfo name for Base*

So, you're correct, Clang hasn't included the typeinfo for the base
types.  Except, it sort of has included it, it's just not compiled into
the binary.

So, I think, where the test does this:

  foreach simple_var {i cp ccp ca b} {

We should actually do something like:

  set var {ca b}
  if {$started || ![test_compiler_info clang-*-* c++]} {
      # Clang doesn't place type information for the base types in
      # the executable, and relies on this being linked in from the
      # standard library.  As a result, type information for these
      # variables is only available once the inferior is started.
      lappend var i cp ccp
  }
  foreach simple_var $var {

What do you think?

Thanks,
Andrew


> ---
>  gdb/testsuite/gdb.cp/typeid.exp | 5 +++++
>  1 file changed, 5 insertions(+)
>
> diff --git a/gdb/testsuite/gdb.cp/typeid.exp b/gdb/testsuite/gdb.cp/typeid.exp
> index 6411217191a..d110b3e7695 100644
> --- a/gdb/testsuite/gdb.cp/typeid.exp
> +++ b/gdb/testsuite/gdb.cp/typeid.exp
> @@ -23,6 +23,11 @@ if {[prepare_for_testing "failed to prepare" $testfile $srcfile {debug c++}]} {
>      return -1
>  }
>  
> +if {[test_compiler_info clang-*-*]} {
> +    untested "clang does not add debuginfo for standard libraries"
> +    return
> +}
> +
>  proc do_typeid_tests {started} {
>      global hex
>  
> -- 
> 2.37.3


  reply	other threads:[~2022-10-26 15:37 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-10-04 17:07 [PATCH 00/11] Cleanup gdb.cp tests when running with clang Bruno Larsen
2022-10-04 17:07 ` [PATCH 01/11] gdb/testsuite: ignore Non-C-typedefs for gdb.cp/class2.exp Bruno Larsen
2022-10-26 12:02   ` Andrew Burgess
2022-10-27  8:37     ` Bruno Larsen
2022-10-28 11:37       ` Andrew Burgess
2022-10-04 17:07 ` [PATCH 02/11] gdb/testsuite: enable running gdb.cp/classes.exp with clang Bruno Larsen
2022-10-26 12:19   ` Andrew Burgess
2022-10-04 17:07 ` [PATCH 03/11] gdb/testsuite: account for clang's recursive destructor calls on gdb.cp/mb-ctor.exp Bruno Larsen
2022-10-26 12:35   ` Andrew Burgess
2022-10-04 17:07 ` [PATCH 4/9] gdb/testsuite: add XFAIL to gdb.cp/derivation.exp when using clang Bruno Larsen
2022-10-04 17:09   ` Bruno Larsen
2022-10-04 17:07 ` [PATCH 04/11] " Bruno Larsen
2022-10-26 12:37   ` Andrew Burgess
2022-10-26 14:50   ` Andrew Burgess
2022-10-04 17:07 ` [PATCH 05/11] gdb/testsuite: allow for clang style destructors on gdb.cp/m-static.exp Bruno Larsen
2022-10-26 13:04   ` Andrew Burgess
2022-10-26 14:51   ` Andrew Burgess
2022-10-04 17:07 ` [PATCH 06/11] gdb/testsuite: add XFAIL to gdb.cp/ptype-flags.exp when using clang Bruno Larsen
2022-10-26 14:08   ` Andrew Burgess
2022-10-27 13:17     ` Bruno Larsen
2022-10-28 11:38       ` Andrew Burgess
2022-10-31 12:45         ` Bruno Larsen
2022-10-04 17:07 ` [PATCH 07/11] gdb/testsuite: skip gdb.cp/anon-struct.exp " Bruno Larsen
2022-10-26 14:49   ` Andrew Burgess
2022-10-27 13:46     ` Bruno Larsen
2022-10-04 17:07 ` [PATCH 08/11] gdb/testsuite: disable gdb.cp/typeid.exp " Bruno Larsen
2022-10-26 15:37   ` Andrew Burgess [this message]
2022-10-04 17:07 ` [PATCH 09/11] gdb/testsuite: fix gdb.cp/converts.exp to run with clang Bruno Larsen
2022-10-26 15:54   ` Andrew Burgess
2022-10-31 12:46     ` Bruno Larsen
2022-10-04 17:07 ` [PATCH 10/11] gdb/testsuite: remove XFAIL on gdb.cp/temargs.exp Bruno Larsen
2022-10-26 15:57   ` Andrew Burgess
2022-10-04 17:07 ` [PATCH 11/11] gdb/testsuite: disable gdb.cp/call-method-register.exp with clang Bruno Larsen
2022-10-27  9:49   ` Andrew Burgess
2022-10-31 10:51     ` Bruno Larsen
2022-10-18  8:10 ` [PING][PATCH 00/11] Cleanup gdb.cp tests when running " Bruno Larsen

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=87pmee61fi.fsf@redhat.com \
    --to=aburgess@redhat.com \
    --cc=blarsen@redhat.com \
    --cc=gdb-patches@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).