Linux MIPS Architecture development
 help / color / mirror / Atom feed
From: Ralf Baechle <ralf@oss.sgi.com>
To: Jun Sun <jsun@mvista.com>
Cc: Keith Owens <kaos@melbourne.sgi.com>, linux-mips@oss.sgi.com
Subject: Re: problems with insmod ...
Date: Tue, 24 Oct 2000 03:18:28 +0200	[thread overview]
Message-ID: <20001024031828.A2816@bacchus.dhis.org> (raw)
In-Reply-To: <39F4E113.6BC85A5C@mvista.com>; from jsun@mvista.com on Mon, Oct 23, 2000 at 06:08:35PM -0700

On Mon, Oct 23, 2000 at 06:08:35PM -0700, Jun Sun wrote:

> > > Jun Sun <jsun@mvista.com> wrote:
> > > >I tried with 2.3.19, and now I am having problem with out of bound index
> > > >in symbol table.  See the output below.
> > > >
> > > >---------
> > > >sh-2.03# insmod hello.o
> > > >hello.o: local symbol gcc2_compiled. with index 10 exceeds
> > > >local_symtab_size 10
> > > >hello.o: local symbol __gnu_compiled_c with index 11 exceeds
> > > >local_symtab_size 10
> > > >---------
> > >
> > > It is a toolchain bug, I think it is in the assembler.  I have a dim
> > > distant memory from about a month ago that somebody on linux-mips found
> > > the problem.  Ask the toolchain experts.
> > 
> > It's a bug bug in ld, one in BFD and a sillyness in IRIX ELF which the linker
> > uses.  IRIX ELF uses different sorting rules for the symbol table, see
> > mips_elf_sym_is_global in bfd/elf32-mips.c.
> > 
> >  - Bug one: ld generated output should follow the same rules as assembler
> >    generated output.
> >  - Bug two is more a design flaw - why does Linux/MIPS and most other
> >    MIPS ELF configurations use IRIX and not ABI ELF?
> >  - Bug three is that mips_elf_sym_is_global applies these IRIX ELF sorting
> >    rules even to ABI ELF.
> > 
> >   Ralf
> 
> This sounds like a serious problem to me.  So here are the questions : 
> 
> 1) is it fixed in the latest binutils?

No, all binutils ever are affected.

> 2) is it worth fixed for binutil v2.8.1?

Probably not.  I believe modulo some testing current CVS binutils are ready
for more serious use than just binutils fixing.  In any case fixing should
be easy.

  Ralf

WARNING: multiple messages have this Message-ID (diff)
From: Ralf Baechle <ralf@oss.sgi.com>
To: Jun Sun <jsun@mvista.com>
Cc: Keith Owens <kaos@melbourne.sgi.com>, linux-mips@oss.sgi.com
Subject: Re: problems with insmod ...
Date: Tue, 24 Oct 2000 03:18:28 +0200	[thread overview]
Message-ID: <20001024031828.A2816@bacchus.dhis.org> (raw)
Message-ID: <20001024011828.HrDE97GCwJEtYlEXhjlcZVAnlzqQTMsBmfsAYtzeE_k@z> (raw)
In-Reply-To: <39F4E113.6BC85A5C@mvista.com>; from jsun@mvista.com on Mon, Oct 23, 2000 at 06:08:35PM -0700

On Mon, Oct 23, 2000 at 06:08:35PM -0700, Jun Sun wrote:

> > > Jun Sun <jsun@mvista.com> wrote:
> > > >I tried with 2.3.19, and now I am having problem with out of bound index
> > > >in symbol table.  See the output below.
> > > >
> > > >---------
> > > >sh-2.03# insmod hello.o
> > > >hello.o: local symbol gcc2_compiled. with index 10 exceeds
> > > >local_symtab_size 10
> > > >hello.o: local symbol __gnu_compiled_c with index 11 exceeds
> > > >local_symtab_size 10
> > > >---------
> > >
> > > It is a toolchain bug, I think it is in the assembler.  I have a dim
> > > distant memory from about a month ago that somebody on linux-mips found
> > > the problem.  Ask the toolchain experts.
> > 
> > It's a bug bug in ld, one in BFD and a sillyness in IRIX ELF which the linker
> > uses.  IRIX ELF uses different sorting rules for the symbol table, see
> > mips_elf_sym_is_global in bfd/elf32-mips.c.
> > 
> >  - Bug one: ld generated output should follow the same rules as assembler
> >    generated output.
> >  - Bug two is more a design flaw - why does Linux/MIPS and most other
> >    MIPS ELF configurations use IRIX and not ABI ELF?
> >  - Bug three is that mips_elf_sym_is_global applies these IRIX ELF sorting
> >    rules even to ABI ELF.
> > 
> >   Ralf
> 
> This sounds like a serious problem to me.  So here are the questions : 
> 
> 1) is it fixed in the latest binutils?

No, all binutils ever are affected.

> 2) is it worth fixed for binutil v2.8.1?

Probably not.  I believe modulo some testing current CVS binutils are ready
for more serious use than just binutils fixing.  In any case fixing should
be easy.

  Ralf


--OgqxwSJOaUobr8KG--

  reply	other threads:[~2000-10-24  1:19 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2000-10-21  2:50 problems with insmod Jun Sun
2000-10-21  3:11 ` Ralf Baechle
2000-10-21  4:12 ` Keith Owens
2000-10-23 18:45   ` Jun Sun
2000-10-23 22:32     ` Keith Owens
2000-10-24  0:40       ` Ralf Baechle
2000-10-24  1:08         ` Jun Sun
2000-10-24  1:18           ` Ralf Baechle [this message]
2000-10-24  1:18             ` Ralf Baechle
2000-10-23 22:42     ` Brady Brown

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=20001024031828.A2816@bacchus.dhis.org \
    --to=ralf@oss.sgi.com \
    --cc=jsun@mvista.com \
    --cc=kaos@melbourne.sgi.com \
    --cc=linux-mips@oss.sgi.com \
    /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