From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Egger, Christoph" Subject: Re: [PATCH] gcov: Support gcc 4.7 Date: Mon, 17 Jun 2013 12:49:26 +0200 Message-ID: <51BEE9B6.7080702@amazon.de> References: <1371457748.9654.3.camel@hamster.uk.xensource.com> <1371465779.23802.31.camel@zakaz.uk.xensource.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <1371465779.23802.31.camel@zakaz.uk.xensource.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Ian Campbell Cc: George Dunlap , Frediano Ziglio , Matthew Daley , Miguel Clara , xen-devel@lists.xen.org List-Id: xen-devel@lists.xenproject.org On 17.06.13 12:42, Ian Campbell wrote: > On Mon, 2013-06-17 at 09:29 +0100, Frediano Ziglio wrote: >> gcc 4.7 changed format used internally for coverage data. >> This patch address these changes. >> The information that gcc generate are mostly the same but to support common >> sections. >> The only difference in the blob exported by Xen is that functions have 2 >> different checksums instead of one. > > Do we really need to reflect that in the Xen public API though? Could we > condense it into the things we want/need and try and keep our API > relatively stable? > > I'm a bit concerned that we are going to end up with dozens of variants > of the API (and the internal code) supporting each new variant of this > internal-to-gcc ABI. > > I think you explained before why this couldn't be exported from the > hypervisor as an opaque set of bytes to be interpreted by a userspace > tool, which is a shame. I think this is possible the same way the hypervisor exports machine check error telemetry. Christoph