Re: [PATCH]Construct canonical scaled address expression in IVOPT

2013-09-26 Thread Richard Biener
tember 26, 2013 8:05 PM >> To: 'Richard Biener'; Bin.Cheng >> Cc: GCC Patches; Richard Earnshaw >> Subject: RE: [PATCH]Construct canonical scaled address expression in IVOPT >> >> >> >> > -Original Message- >> > From: Ri

RE: [PATCH]Construct canonical scaled address expression in IVOPT

2013-09-26 Thread bin.cheng
CC Patches; Richard Earnshaw > Subject: RE: [PATCH]Construct canonical scaled address expression in IVOPT > > > > > -Original Message- > > From: Richard Biener [mailto:richard.guent...@gmail.com] > > Sent: Tuesday, September 24, 2013 7:58 PM > > To: Bin.Cheng

RE: [PATCH]Construct canonical scaled address expression in IVOPT

2013-09-26 Thread bin.cheng
> -Original Message- > From: Richard Biener [mailto:richard.guent...@gmail.com] > Sent: Tuesday, September 24, 2013 7:58 PM > To: Bin.Cheng > Cc: Bin Cheng; GCC Patches; Richard Earnshaw > Subject: Re: [PATCH]Construct canonical scaled address expression in IVOPT &

Re: [PATCH]Construct canonical scaled address expression in IVOPT

2013-09-24 Thread Richard Biener
On Tue, Sep 24, 2013 at 1:40 PM, Bin.Cheng wrote: > On Tue, Sep 24, 2013 at 6:12 PM, Richard Biener > wrote: >> On Tue, Sep 24, 2013 at 8:20 AM, bin.cheng wrote: >>> >>> -Original Message- >> >> Or even [reg*scale] (not sure about that). But yes, at least reg*scale + >> offset >>

Re: [PATCH]Construct canonical scaled address expression in IVOPT

2013-09-24 Thread Bin.Cheng
On Tue, Sep 24, 2013 at 6:12 PM, Richard Biener wrote: > On Tue, Sep 24, 2013 at 8:20 AM, bin.cheng wrote: >> >> >>> -Original Message- > > Or even [reg*scale] (not sure about that). But yes, at least reg*scale + > offset > and reg*scale + reg. > >> Apparently it's infeasible to check

Re: [PATCH]Construct canonical scaled address expression in IVOPT

2013-09-24 Thread Richard Biener
On Tue, Sep 24, 2013 at 12:36 PM, Richard Earnshaw wrote: > On 24/09/13 11:12, Richard Biener wrote: >> Or even [reg*scale] (not sure about that). But yes, at least reg*scale + >> offset >> and reg*scale + reg. > > I can't conceive of a realistic case where one would want to scale the > base add

Re: [PATCH]Construct canonical scaled address expression in IVOPT

2013-09-24 Thread Richard Earnshaw
On 24/09/13 11:12, Richard Biener wrote: > Or even [reg*scale] (not sure about that). But yes, at least reg*scale + > offset > and reg*scale + reg. I can't conceive of a realistic case where one would want to scale the base address. Scaling the offset is fine, but never the base. So reg*scale+

Re: [PATCH]Construct canonical scaled address expression in IVOPT

2013-09-24 Thread Richard Biener
On Tue, Sep 24, 2013 at 8:20 AM, bin.cheng wrote: > > >> -Original Message- >> From: Richard Biener [mailto:richard.guent...@gmail.com] >> Sent: Monday, September 23, 2013 8:08 PM >> To: Bin Cheng >> Cc: GCC Patches >> Subject: Re: [PATCH]Constr

RE: [PATCH]Construct canonical scaled address expression in IVOPT

2013-09-23 Thread bin.cheng
> -Original Message- > From: Richard Biener [mailto:richard.guent...@gmail.com] > Sent: Monday, September 23, 2013 8:08 PM > To: Bin Cheng > Cc: GCC Patches > Subject: Re: [PATCH]Construct canonical scaled address expression in IVOPT > > On Fri, Sep 20, 201

Re: [PATCH]Construct canonical scaled address expression in IVOPT

2013-09-23 Thread Richard Earnshaw
On 23/09/13 13:07, Richard Biener wrote: > What's the problem > with arm supporting reg1 * scale? Why shouldn't it being able to handle > the implicit zero offset? Something like "we don't have an instruction that can do that"... Valid addresses are of the general form address:= '[' base-r

Re: [PATCH]Construct canonical scaled address expression in IVOPT

2013-09-23 Thread Richard Biener
On Fri, Sep 20, 2013 at 12:00 PM, bin.cheng wrote: > Hi, > For now IVOPT constructs scaled address expression in the form of > "scaled*index" and checks whether backend supports it. The problem is the > address expression is invalid on ARM, causing scaled expression disabled in > IVOPT on ARM. Th