From mboxrd@z Thu Jan 1 00:00:00 1970 From: George Dunlap Subject: Re: [PATCH for xen-4.4] libxl: add option for discard support to xl disk configuration Date: Wed, 5 Feb 2014 11:10:33 +0000 Message-ID: <52F21C29.8090607@eu.citrix.com> References: <1391083364-29483-1-git-send-email-olaf@aepfle.de> <20140130162558.GA9033@aepfle.de> <1391099505.9495.23.camel@kazak.uk.xensource.com> <20140130173058.GA12133@aepfle.de> <1391530957.6497.56.camel@kazak.uk.xensource.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; Format="flowed" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <1391530957.6497.56.camel@kazak.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 , Olaf Hering Cc: anthony.perard@citrix.com, xen-devel@lists.xen.org, Ian.Jackson@eu.citrix.com, stefano.stabellini@eu.citrix.com List-Id: xen-devel@lists.xenproject.org On 02/04/2014 04:22 PM, Ian Campbell wrote: > On Thu, 2014-01-30 at 18:30 +0100, Olaf Hering wrote: > > George, any thoughts on: > >>> TBH -- if you (==suse I guess?) are contemplating carrying this as a >>> backport even before 4.4 is out the door we should probably be at least >>> considering a freeze exception for 4.4. George CCd for input. (I >>> appreciate that "backport=>freeze exception" is a potentially slippery >>> slope/ripe for abuse...) >> It will make less work for SUSE if this change would be incorporated >> into 4.4, and later replaced with the "final" version I sent out today. >> However, its small and will be easy to port forward to 4.4.X. >> >> The risk of including such change is small as it requires a patched qemu >> which actually does discard (1.7?), a patched frontend driver (pvops >> 3.15?) before the codepaths it enables are actually executed. Well it looks like in order to keep ABI compatibility (which I don't think we ever promised), you're introducing this weird hack with overloading a putative boolean value with a magic number? I think the patch is really ugly. I assume the reason you're attemping to avoid breaking ABI compatibility is because we're so close to the release? But if so, adding an ugly hack like this is worse, IMHO. -George