[parisc-linux] Why do we relax e_ident[EI_OSABI] in gdb?

[parisc-linux] Why do we relax e_ident[EI_OSABI] in gdb?

[Carlos O'Donell]

parisc-linux,

Anyone have some ideas why we have this patch in gdb, and thus relax the 
EI_OSABI check?

--- gdb-5.2.cvs20020401/bfd/elf64-hppa.c~	Sun Mar 31 19:09:41 2002
+++ gdb-5.2.cvs20020401/bfd/elf64-hppa.c	Sat Apr 20 09:35:54 2002
@@ -372,7 +372,8 @@
   i_ehdrp = elf_elfheader (abfd);
   if (strcmp (bfd_get_target (abfd), "elf64-hppa-linux") == 0)
     {
-      if (i_ehdrp->e_ident[EI_OSABI] != ELFOSABI_LINUX)
+      if (i_ehdrp->e_ident[EI_OSABI] != ELFOSABI_LINUX &&
+          i_ehdrp->e_ident[EI_OSABI] != ELFOSABI_NONE)
 	return false;
     }
   else
--- gdb-5.2.cvs20020401/bfd/elf32-hppa.c~	Sun Mar 31 19:09:41 2002
+++ gdb-5.2.cvs20020401/bfd/elf32-hppa.c	Sat Apr 20 09:35:54 2002
@@ -1038,7 +1038,8 @@
   i_ehdrp = elf_elfheader (abfd);
   if (strcmp (bfd_get_target (abfd), "elf32-hppa-linux") == 0)
     {
-      if (i_ehdrp->e_ident[EI_OSABI] != ELFOSABI_LINUX)
+      if (i_ehdrp->e_ident[EI_OSABI] != ELFOSABI_LINUX &&
+          i_ehdrp->e_ident[EI_OSABI] != ELFOSABI_NONE)
 	return false;
     }
   else
===

c.

Re: [parisc-linux] Why do we relax e_ident[EI_OSABI] in gdb?

[Grant Grundler]

On Thu, Dec 26, 2002 at 09:31:37PM -0500, Carlos O'Donell wrote:
> Anyone have some ideas why we have this patch in gdb, and thus relax the 
> EI_OSABI check?

Maybe this mail explains it?
http://lists.parisc-linux.org/hypermail/parisc-linux/2357.html

search for ELF_OSABI on lists.p-l.o turned up alot of hits.
But looks like just one thread.

grant

Re: [parisc-linux] Why do we relax e_ident[EI_OSABI] in gdb?

[Grant Grundler]

On Thu, Dec 26, 2002 at 10:49:02PM -0700, Grant Grundler wrote:
> Maybe this mail explains it?
> http://lists.parisc-linux.org/hypermail/parisc-linux/2357.html

http://lists.parisc-linux.org/hypermail/parisc-linux/2377.html

This is a bit further down the "next thread" list and is Cary
Coutant trying to clarify the intent of the fields.

grant

Re: [parisc-linux] Why do we relax e_ident[EI_OSABI] in gdb?

[Randolph Chung]

> Anyone have some ideas why we have this patch in gdb, and thus relax the 
> EI_OSABI check?
> 
> --- gdb-5.2.cvs20020401/bfd/elf64-hppa.c~	Sun Mar 31 19:09:41 2002
> +++ gdb-5.2.cvs20020401/bfd/elf64-hppa.c	Sat Apr 20 09:35:54 2002
> @@ -372,7 +372,8 @@
>    i_ehdrp = elf_elfheader (abfd);
>    if (strcmp (bfd_get_target (abfd), "elf64-hppa-linux") == 0)
>      {
> -      if (i_ehdrp->e_ident[EI_OSABI] != ELFOSABI_LINUX)
> +      if (i_ehdrp->e_ident[EI_OSABI] != ELFOSABI_LINUX &&
> +          i_ehdrp->e_ident[EI_OSABI] != ELFOSABI_NONE)

well... take a look at
http://bugs.debian.org/cgi-bin/bugreport.cgi?archive=no&bug=137316

but Grant's references explain it in more detail.

randolph