This is the mail archive of the
binutils@sourceware.cygnus.com
mailing list for the binutils project.
Re: [PATCH] Re: Link error with today's CVS binutils
- To: Ian Lance Taylor <ian at zembu dot com>
- Subject: Re: [PATCH] Re: Link error with today's CVS binutils
- From: "H . J . Lu" <hjl at valinux dot com>
- Date: Tue, 18 Jan 2000 08:03:12 -0800
- Cc: devolder at evsx dot com, nickc at cygnus dot com,binutils at sourceware dot cygnus dot com
- References: <20000117182517.A24406@valinux.com> <20000118053804.1201.qmail@daffy.airs.com>
On Tue, Jan 18, 2000 at 12:38:04AM -0500, Ian Lance Taylor wrote:
>
> Have you followed this thread? Your change:
>
> 1999-01-15 Nick Clifton <nickc@cygnus.com>
>
> * ldlang.c (lang_size_sections): Only update the current
> address of a region if the section just placed into it is an
> allocated section.
>
> seems to cause the regression. This patch seems to fix it. Any
> comments?
>
> Thanks for tracking this down.
>
> I think that checking the size of the section is not the right way to
> go here. Instead, check for SEC_NEVER_LOAD. Look at how
> SEC_NEVER_LOAD is handled in ldwrite.c.
>
> Ian
Thanks. How about this change? Since it looks like SEC_NEVER_LOAD is
only used by COFF, I left in the size check.
--
H.J. Lu (hjl@gnu.org)
---
Tue Jan 18 07:57:08 2000 H.J. Lu <hjl@gnu.org>
* ldlang.c (lang_size_sections): Also update the current
address of a region if the section raw size is > 0 and the
SEC_NEVER_LOAD bit is not set.
Index: ldlang.c
===================================================================
RCS file: /work/cvs/gnu/binutils/ld/ldlang.c,v
retrieving revision 1.1.1.13
diff -u -p -r1.1.1.13 ldlang.c
--- ldlang.c 2000/01/13 19:36:06 1.1.1.13
+++ ldlang.c 2000/01/18 15:56:02
@@ -2823,10 +2823,17 @@ lang_size_sections (s, output_section_st
/* Update dot in the region ?
We only do this if the section is going to be allocated,
since unallocated sections do not contribute to the region's
- overall size in memory. */
+ overall size in memory.
+
+ If the section raw size is > 0, it will affect the
+ addresses of sections after it. We have to update
+ dot. */
if (os->region != (lang_memory_region_type *) NULL
- && (bfd_get_section_flags (output_bfd, os->bfd_section)
- & (SEC_ALLOC | SEC_LOAD)))
+ && ((os->bfd_section->_raw_size > 0
+ && (bfd_get_section_flags (output_bfd, os->bfd_section)
+ & SEC_NEVER_LOAD) == 0)
+ || (bfd_get_section_flags (output_bfd, os->bfd_section)
+ & (SEC_ALLOC | SEC_LOAD))))
{
os->region->current = dot;