Received: from eggs.gnu.org ([2001:4830:134:3::10]:57381) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1ZHZVH-0008Pk-9N for qemu-devel@nongnu.org; Tue, 21 Jul 2015 11:30:51 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1ZHZVA-0004On-3C for qemu-devel@nongnu.org; Tue, 21 Jul 2015 11:30:47 -0400 Date: Tue, 21 Jul 2015 17:30:36 +0200 From: Aurelien Jarno Message-ID: <20150721153036.GB30591@aurel32.net> References: <1437455978.5809.2.camel@kernel.crashing.org> <55ADE64A.7050702@twiddle.net> <1437460396.5809.5.camel@kernel.crashing.org> <55ADE936.8090407@twiddle.net> <1437471990.5809.7.camel@kernel.crashing.org> <55AE3542.5080300@suse.de> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <55AE3542.5080300@suse.de> Subject: Re: [Qemu-devel] [PATCH v3] tcg/ppc: Improve unaligned load/store handling on 64-bit backend List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Alexander Graf Cc: Paolo Bonzini , qemu-ppc@nongnu.org, qemu-devel@nongnu.org, Richard Henderson On 2015-07-21 14:04, Alexander Graf wrote: > On 07/21/15 11:46, Benjamin Herrenschmidt wrote: > >On Tue, 2015-07-21 at 07:39 +0100, Richard Henderson wrote: > >>On 07/21/2015 07:33 AM, Benjamin Herrenschmidt wrote: > >>>On Tue, 2015-07-21 at 07:27 +0100, Richard Henderson wrote: > >>>>On 07/21/2015 06:19 AM, Benjamin Herrenschmidt wrote: > >>>>>+ /* Clear the non-page, non-alignment bits from the address */ > >>>>> if (TCG_TARGET_REG_BITS == 32 || TARGET_LONG_BITS == 32) { > >>>>>+ /* We don't support unaligned accesses on 32-bits, preserve > >>>>>+ * the bottom bits and thus trigger a comparison failure on > >>>>>+ * unaligned accesses > >>>>>+ */ > >>>>> tcg_out_rlw(s, RLWINM, TCG_REG_R0, addrlo, 0, > >>>>> (32 - s_bits) & 31, 31 - TARGET_PAGE_BITS); > >>>>Why don't you support this unaligned acess with 32-bit guests? > >>>No reason, I just didn't get to do it yet. It's possible, I was just > >>>lazy :-) It also adds one instruction. On 64-bit we always have 2 > >>>instructions anyway so it wasn't adding any overhead really, on 32-bit > >>>we get away with a single rlwinm, while adding the unaligned support > >>>would make it an addi + rlwinm. > >>Ah, ok. I wondered if some older 32-bit host ppc didn't allow it, and the > >>32-bit guest paid the price. Anyway, > >No, most implementations support unaligned accesses in common cases. I > >think the worst embedded variants might trap when crossing page > >boundaries but that's about it. I don't think we bother emulating > >this though. > > > >>Reviewed-by: Richard Henderson > >Thanks. > > So who can pick those up? PPC TCG is pretty much unmaintained since Malc > disappeared, no? > Richard has reviewed this patch, and there are already generic TCG patches for 2.5 on the list. I guess he can squeeze a few PPC specific commits like this one in the pull request. -- Aurelien Jarno GPG: 4096R/1DDD8C9B aurelien@aurel32.net http://www.aurel32.net