public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
From: "simon.marchi at polymtl dot ca" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug debug/100610] New: DWARF5, wrong include_directories[0] when building in /
Date: Sat, 15 May 2021 11:18:27 +0000	[thread overview]
Message-ID: <bug-100610-4@http.gcc.gnu.org/bugzilla/> (raw)

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100610

            Bug ID: 100610
           Summary: DWARF5, wrong include_directories[0] when building in
                    /
           Product: gcc
           Version: unknown
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: debug
          Assignee: unassigned at gcc dot gnu.org
          Reporter: simon.marchi at polymtl dot ca
  Target Milestone: ---

If I build a trivial test file while in the /tmp directory:

$ pwd
/tmp
$ cat test.c
#include <stdio.h>

#define FOO 2

int main() {
        return FOO;
}
$ /opt/gcc/git/bin/gcc -gdwarf-5 -g3 -O0 -o test test.c

The include_directories table of .debug_line looks fine:

$ readelf --debug-dump=rawline test

 The Directory Table (offset 0x22, lines 7, columns 1):
  Entry Name
  0     (line_strp)     (offset: 0x0): /tmp
  1     (line_strp)     (offset: 0xc): /usr/include
  2     (line_strp)     (offset: 0x19): /usr/include/bits
  3     (line_strp)     (offset: 0x2b): /usr/include/sys
  4     (line_strp)     (offset: 0x3c): /usr/include/gnu
  5     (line_strp)     (offset: 0x4d):
/opt/gcc/git/lib/gcc/x86_64-pc-linux-gnu/12.0.0/include
  6     (line_strp)     (offset: 0x85): /usr/include/bits/types

 The File Name Table (offset 0x44, lines 26, columns 2):
  Entry Dir     Name
  0     (udata) 0       (line_strp)     (offset: 0x5): test.c
  1     (udata) 0       (line_strp)     (offset: 0x5): test.c

Files #0 and #1 point to directory 0, which gives /tmp/test.c.

However, if I move test.c in the root directory, at /test.c and move there:

$ pwd
/
$ /opt/gcc/git/bin/gcc -gdwarf-5 -g3 -O0 -o /tmp/test test.c

 The Directory Table (offset 0x22, lines 7, columns 1):
  Entry Name
  0     (line_strp)     (offset: 0x9): /usr/include
  1     (line_strp)     (offset: 0x9): /usr/include
  2     (line_strp)     (offset: 0x16): /usr/include/bits
  3     (line_strp)     (offset: 0x28): /usr/include/sys
  4     (line_strp)     (offset: 0x39): /usr/include/gnu
  5     (line_strp)     (offset: 0x4a):
/opt/gcc/git/lib/gcc/x86_64-pc-linux-gnu/12.0.0/include
  6     (line_strp)     (offset: 0x82): /usr/include/bits/types

 The File Name Table (offset 0x44, lines 26, columns 2):
  Entry Dir     Name
  0     (udata) 0       (line_strp)     (offset: 0x0): test.c
  1     (udata) 0       (line_strp)     (offset: 0x0): test.c

See the files #0 and #1.  They point to directory #0, which is /usr/include. 
That makes it believe that the source file was at /usr/include/test.c.  That
looks wrong to me.

                 reply	other threads:[~2021-05-15 11:18 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=bug-100610-4@http.gcc.gnu.org/bugzilla/ \
    --to=gcc-bugzilla@gcc.gnu.org \
    --cc=gcc-bugs@gcc.gnu.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).