From mboxrd@z Thu Jan 1 00:00:00 1970 From: Paul Mundt Date: Wed, 13 Jun 2012 07:31:14 +0000 Subject: Re: [PATCH 2/2] irqdomain: Support for static IRQ mapping and association. Message-Id: <20120613073114.GM10170@linux-sh.org> List-Id: References: <1337576792-5347-1-git-send-email-lethal@linux-sh.org> <1337576792-5347-2-git-send-email-lethal@linux-sh.org> <20120526015049.1616D3E2336@localhost> <20120611032547.GB10170@linux-sh.org> In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Grant Likely Cc: linux-sh@vger.kernel.org, linux-kernel@vger.kernel.org On Sun, Jun 10, 2012 at 11:02:16PM -0600, Grant Likely wrote: > On Sun, Jun 10, 2012 at 9:25 PM, Paul Mundt wrote: > > On Fri, May 25, 2012 at 07:50:49PM -0600, Grant Likely wrote: > >> It would be really good to make sure the hwirqs aren't already > >> associated before trying to associate them again. ?Unfortunately that > >> can't be done (nicely) until I get rid of the slow path lookup. ?I've > >> got a patch for that which I'll rebase on top of this one and post soon. > >> > > Any updates on this? I have quite a few more changes I intend on making, > > but there's no point in starting in on that until these existing patches > > are sorted out. > > Yeah, I made some progress and have a whole bunch of patches to post > which I had hoped to do last week. Unfortunately the move to the UK > has meant that the last 7 days have been non-stop cleaning, packing > and throwing out stuff. I've not been able to do a final cleanup. > The house gets cleaned on Tuesday so it can go on the market on > Wednesday which means all the house stuff must be done by Monday > evening. Hopefully I'll get some time on Tuesday to do some hacking. > No worries, life happens. Just wanted to make sure you hadn't forgotten about them! > If you want to look at them I've pushed them out to the irqdomain/test > branch on git.secretlab.ca. > Thanks, I've fetched them and been hacking on them a bit. Fortunately you already did some the work I had intended on doing, particularly in relation to slow vs fast-path lookups, so there hasn't been much more for me to do. That being said, you forgot about legacy domains with non-zero hwirq bases in your legacy -> linear revmap conversion, so some (or all) of those IRQs will now find themselves stuffed in the radix tree instead.. I already got bitten by this trying to figure out why my batched disposal helper wasn't doing its job and my linear domain was tripping over the non-zero radix root height warn at domain removal time, so that's fixed now. Patches to follow (also in the common/irqdomain topic branch in my github tree).