This is the mail archive of the
mailing list for the binutils project.
Re: ARC port broken reloc processing
- From: Alan Modra <amodra at gmail dot com>
- To: Cupertino Miranda <Cupertino dot Miranda at synopsys dot com>
- Cc: Andrew Burgess <andrew dot burgess at embecosm dot com>, "binutils at sourceware dot org" <binutils at sourceware dot org>, Claudiu Zissulescu <Claudiu dot Zissulescu at synopsys dot com>
- Date: Tue, 1 Dec 2015 11:34:45 +1030
- Subject: Re: ARC port broken reloc processing
- Authentication-results: sourceware.org; auth=none
- References: <1448580380-2724-1-git-send-email-andrew dot burgess at embecosm dot com> <20151127024411 dot GA8901 at bubble dot grove dot modra dot org> <20151127033938 dot GB8901 at bubble dot grove dot modra dot org> <20151128211826 dot GA4625 at embecosm dot com> <20151130021307 dot GE8901 at bubble dot grove dot modra dot org> <EB86EB452ADE4B44B294F7149B8A2777021853F2 at DE02WEMBXB dot internal dot synopsys dot com>
On Mon, Nov 30, 2015 at 11:57:32AM +0000, Cupertino Miranda wrote:
> I believe in this way we set the mask to a correct value for all the relocation types, without possibly introducing inconsistencies (repetitions) in the reloc definition.
> Nevertheless, the generic relocation patching cannot be used to fix relocations, as FORMULA would not be applied.
Applied. Your patch will fix the relocs used in debug sections, but
as you imply, won't do the right thing for a reloc using other than
an S+A formula. It also won't work for split field relocs. So it
isn't good at supporting linking directly to non-ELF formats, like
SREC. Which doesn't matter too much as it is far better to link
normally, then use objcopy to convert to another format.
Australia Development Lab, IBM