From mboxrd@z Thu Jan 1 00:00:00 1970 From: Razvan Cojocaru Subject: Re: [PATCH RFC 1/9] xen: Emulate with no writes; compute current instruction length Date: Thu, 03 Jul 2014 10:42:13 +0300 Message-ID: <53B50955.5040104@bitdefender.com> References: <1404308041-15461-1-git-send-email-rcojocaru@bitdefender.com> <53B4232C.1080204@citrix.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <53B4232C.1080204@citrix.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: Andrew Cooper , xen-devel@lists.xen.org Cc: tim@xen.org List-Id: xen-devel@lists.xenproject.org On 07/02/2014 06:20 PM, Andrew Cooper wrote: > On 02/07/14 14:33, Razvan Cojocaru wrote: >> Added support for emulating an instruction with no memory writes and >> for retrieving the length of the next instruction. Additionally, >> introduced hvm_emulate_one_full(bool_t nowrite), which acts upon all >> possible return values from the hvm_emulate_one() functions (RETRY, >> EXCEPTION, UNHANDLEABLE). >> >> Signed-off-by: Razvan Cojocaru > > xen/arch/x86/x86_emulate/ is the core of the emulation in Xen, and this > looks very much as if it should be part that, rather than wedged on the > side of the hypervisor. > > Amongst other things, x86_emulate already contains instruction decode > tables, so it would appear that extending x86_emulate would result in > less code duplication. Actually I think I'll break this one into two patches: the emulate without writes part, which only needs minor modifications according to Jan's review, and the instruction length part with requires more careful consideration. Thanks, Razvan Cojocaru