This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: LTO streaming of TARGET_OPTIMIZE_NODE
- From: Richard Biener <rguenther at suse dot de>
- To: Bernd Schmidt <bernds_cb1 at t-online dot de>
- Cc: Jan Hubicka <hubicka at ucw dot cz>, gcc-patches at gcc dot gnu dot org, "Joseph S. Myers" <joseph at codesourcery dot com>, Thomas Schwinge <thomas_schwinge at mentor dot com>, Julian Brown <julian at codesourcery dot com>, Cesar Philippidis <cesar_philippidis at mentor dot com>
- Date: Thu, 20 Nov 2014 14:20:45 +0100 (CET)
- Subject: Re: LTO streaming of TARGET_OPTIMIZE_NODE
- Authentication-results: sourceware.org; auth=none
- References: <20141113040652 dot GB1984 at kam dot mff dot cuni dot cz> <546DDE1C dot 6060203 at t-online dot de>
On Thu, 20 Nov 2014, Bernd Schmidt wrote:
> On 11/13/2014 05:06 AM, Jan Hubicka wrote:
> > this patch adds infrastructure for proper streaming and merging of
> > TREE_TARGET_OPTION.
>
> This breaks the offloading path via LTO since it introduces an incompatibility
> in LTO format between host and offload machine.
>
> A very quick patch to fix it is below - the OpenACC testcase I was using seems
> to be working again with this. Thoughts, suggestions?
The offload target needs to have the same target options as the host?
Are the offload functions marked somehow? That is, can we avoid
setting TREE_TARGET_OPTION on them? Or rather we need to have a
default TREE_TARGET_OPTION node for the offload target which we'd
need to set - how would you otherwise transfer different offload
target options to the offload compile? How do you transfer
offload target options to the offload compile at all?
I think this just shows conceptual issues with the LTO approach...
Thanks,
Richard.