Mirror of the gdb mailing list
 help / color / mirror / Atom feed
From: Daniel Jacobowitz <drow@false.org>
To: Christophe LYON <christophe.lyon@st.com>
Cc: gdb@sourceware.org
Subject: Re: Unused variables in testsuite
Date: Fri, 28 Apr 2006 15:11:00 -0000	[thread overview]
Message-ID: <20060428144158.GA29211@nevyn.them.org> (raw)
In-Reply-To: <44522900.4090103@st.com>

On Fri, Apr 28, 2006 at 04:38:56PM +0200, Christophe LYON wrote:
> Is there anywhere a requirement that the compiler should allocate unused 
> variable under -O0 ?

This sort of requirement is never documented anywhere; it's just a
question of common practice.  Which in practice means GCC, until
someone is interested in making it more portable.

> For instance, consider the test in gdb.cp/m-static.cc , m-static.h, 
> m-static1.cc.
> You can see that in main() one defines "test4" which is not used, thus 
> need not be allocated. Then, using "print test4.elsewhere" under GDB 
> fails. [note that in this case the default (empty) constructor for test4 
> has no side effects, thus test4 can be eliminated safely]
> 
> In the same way, in gdb.mi/var-cmd.c , mi-var-display.exp, you can see 
> that function do_special_tests() defines several variables (eg 'u') and 
> does not use them, while the .exp file tries to "-var-create u * u".
> 
> I modified these tests so that the variables are used, and they now pass 
> with our compiler.
> 
> May I submit this small patch, or is there a requirement on GCC to 
> allocate even unused variables in -O0? (our compiler is not GCC)

I'd recommend submitting it.  We (CodeSourcery) have been slowly
submitting patches to allow the testsuite to run with yet another
compiler (ARM's); we hadn't gotten to this one yet, but I think it's in
there somewhere, because I remember having to modify m-static*.

-- 
Daniel Jacobowitz
CodeSourcery


      reply	other threads:[~2006-04-28 14:42 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-04-28 14:45 Christophe LYON
2006-04-28 15:11 ` Daniel Jacobowitz [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=20060428144158.GA29211@nevyn.them.org \
    --to=drow@false.org \
    --cc=christophe.lyon@st.com \
    --cc=gdb@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