Re: i386.record.floating.point.patch : with more testing and assurity

[paawan oza <paawan1982@yahoo.com>]

Hi,

In My understanding the point was like below.
in the patch there were following register extended in enumeration in 
i386-tdep.h

I386_FSTAT,
I386_FTAG,       
I386_FISEG,
I386_FIOFF,
I386_FOSEG,
I386_FOOFF,
I386_FOP


According to Hui in some of his previous mails...his idea was
> FCTRL, FOP and so on are the fp reg of amd64.  For now, prec is still
> not support amd64 And amd64's support are in amd64-tedp.... files.  >Change i386_regnum is not a good idea. I suggest you divide fp patch to 2 >parts. One is for i386, the other for amd64. For now, just send i386 patch >for review.  And send amd64 patch when prec support amd64"


while, my idea/understanding is:
FCTRL, FOP registers are not only a part of amd64, but also part of i386 (x87 FPU unit) also.
so according to me these registers are part of i386 also and it needed to be also in i386-tdep.h.

Regards,
Oza.





--- On Wed, 7/1/09, Mark Kettenis <mark.kettenis@xs4all.nl> wrote:

> From: Mark Kettenis <mark.kettenis@xs4all.nl>
> Subject: Re: i386.record.floating.point.patch : with more testing and assurity
> To: paawan1982@yahoo.com
> Cc: pedro@codesourcery.com, msnyder@vmware.com, teawater@gmail.com, gdb-patches@sourceware.org
> Date: Wednesday, July 1, 2009, 9:42 PM
> > Date: Wed, 1 Jul 2009 09:02:17
> -0700 (PDT)
> > From: paawan oza <paawan1982@yahoo.com>
> > 
> > Can somebody please help with this open point ?
> 
> Sorry, I don't understand what point.
> 
> > --- On Wed, 7/1/09, paawan oza <paawan1982@yahoo.com>
> wrote:
> > 
> > > From: paawan oza <paawan1982@yahoo.com>
> > > Subject: Re: i386.record.floating.point.patch :
> with more testing and assurity
> > > To: "Pedro Alves" <pedro@codesourcery.com>,
> "Mark Kettenis" <mark.kettenis@xs4all.nl>,
> "Michael Snyder" <msnyder@vmware.com>,
> "Hui Zhu" <teawater@gmail.com>
> > > Cc: gdb-patches@sourceware.org
> > > Date: Wednesday, July 1, 2009, 9:30 PM
> > > Hi Hui,
> > > 
> > > As I clarified earlier, 
> > > these registers are already supported and
> extended by gdb
> > > much before this patch.
> > > all floating point registers are already
> supported by gdb 
> > > (info floats/info all-registers command gives
> it)
> > > 
> > > My aim is : only to make sure that whenever any
> floating
> > > point insn gets executed, we record the registers
> 
> > > (no matter whether it is %st(n) or FCTRL or FTAG
> or
> > > FSTATUS)
> > > there are insns
> > > e.g. 'ffree' changes FTAG register, so we must
> record it.
> > > 
> > > Do you mean to say that we should remove it from
> just
> > > enumaration ?
> > > but anyway we need to record those registers.
> > > 
> > > Regards,
> > > Oza.
> > > 
> > > --- On Wed, 7/1/09, Hui Zhu <teawater@gmail.com>
> > > wrote:
> > > 
> > > > From: Hui Zhu <teawater@gmail.com>
> > > > Subject: Re:
> i386.record.floating.point.patch : with
> > > more testing and assurity
> > > > To: "paawan oza" <paawan1982@yahoo.com>,
> > > "Pedro Alves" <pedro@codesourcery.com>,
> > > "Mark Kettenis" <mark.kettenis@xs4all.nl>,
> > > "Michael Snyder" <msnyder@vmware.com>
> > > > Cc: gdb-patches@sourceware.org
> > > > Date: Wednesday, July 1, 2009, 11:23 AM
> > > > About this patch, I say my idea
> > > > again, I told in
> > > > http://sourceware.org/ml/gdb-patches/2009-06/msg00014.html
> > > > @@ -145,7 +145,22 @@
> > > >    I386_ES_REGNUM,   
> > > >     /* %es */
> > > >    I386_FS_REGNUM,   
> > > >     /* %fs */
> > > >    I386_GS_REGNUM,   
> > > >     /* %gs */
> > > > -  I386_ST0_REGNUM   
> > > >     /* %st(0) */
> > > > +  I386_ST0_REGNUM,   
> > > >     /* %st(0) */
> > > > +  I386_ST1_REGNUM,   
> > > >     /* %st(1) */
> > > > +  I386_ST2_REGNUM,   
> > > >     /* %st(2) */
> > > > +  I386_ST3_REGNUM,   
> > > >     /* %st(3) */
> > > > +  I386_ST4_REGNUM,   
> > > >     /* %st(4) */
> > > > +  I386_ST5_REGNUM,   
> > > >     /* %st(5) */
> > > > +  I386_ST6_REGNUM,   
> > > >     /* %st(6) */
> > > > +  I386_ST7_REGNUM,   
> > > >     /* %st(7) */
> > > > +  I386_FCTRL,       
> > > >     /* floating point env regs :
> FCTRL-FOP
> > > > */    
> > > > +  I386_FSTAT,
> > > > +  I386_FTAG,       
> > > >     
> > > > +  I386_FISEG,
> > > > +  I386_FIOFF,
> > > > +  I386_FOSEG,
> > > > +  I386_FOOFF,
> > > > +  I386_FOP
> > > >  };
> > > > 
> > > > You are working on make prec x86 support fp
> insn, not
> > > to
> > > > extend the fp
> > > > function of i386 (If you want, you can make
> a special
> > > patch
> > > > for it).
> > > > 
> > > > Hui
> > > > 
> > > > On Tue, Jun 30, 2009 at 23:05, paawan
> oza<paawan1982@yahoo.com>
> > > > wrote:
> > > > >
> > > > > Hi,
> > > > > As I am submitting the patch for the
> first time,
> > > I am
> > > > not much aware of gdb test suite.
> > > > > would you please guide me about how I
> can put
> > > the
> > > > things in the testsuite ?
> > > > > is it the testsuite which comes along
> with the
> > > gdb
> > > > source ?
> > > > > gdb\testsuite\gdb.base ??
> > > > > Regards,
> > > > > Oza.
> > > > >
> > > > > --- On Tue, 6/30/09, Pedro Alves <pedro@codesourcery.com>
> > > > wrote:
> > > > >
> > > > >> From: Pedro Alves <pedro@codesourcery.com>
> > > > >> Subject: Re:
> i386.record.floating.point.patch
> > > :
> > > > with more testing and assurity
> > > > >> To: gdb-patches@sourceware.org
> > > > >> Cc: "paawan oza" <paawan1982@yahoo.com>,
> > > > teawater@gmail.com
> > > > >> Date: Tuesday, June 30, 2009, 7:09
> PM
> > > > >> On Tuesday 30 June 2009 14:23:30,
> > > > >> paawan oza wrote:
> > > > >> >
> > > > >> > > As suggested by Hui,
> > > > >> > > I have come up with more
> detailed
> > > and
> > > > granular
> > > > >> test case
> > > > >> > > for the patch which I had
> submitted
> > > last
> > > > week.
> > > > >>
> > > > >> Could you please consider migrating
> that test
> > > into
> > > > the
> > > > >> testsuite?
> > > > >> You've gone through the trouble of
> writing
> > > tests
> > > > to make
> > > > >> sure
> > > > >> the features work now --- putting
> it in the
> > > > testsuite means
> > > > >> we
> > > > >> have an automatic-ish means to
> check that it
> > > > doesn't get
> > > > >> inadvertently broken in the
> future.  The way
> > > it
> > > > is,
> > > > >> when your
> > > > >> code gets in, the test will
> probably end up
> > > lost
> > > > in the
> > > > >> archives.
> > > > >> We wouldn't want that, would we? 
> :-) 
> > > Having
> > > > >> auto-tests, also helps
> > > > >> the person doing the review in
> confirming
> > > things
> > > > work as
> > > > >> expected (without much effort).
> > > > >>
> > > > >> --
> > > > >> Pedro Alves
> > > > >>
> > > > >
> > > > >
> > > > >
> > > > >
> > > > 
> > > 
> > > 
> > > 
> > > 
> > 
> > 
> >       
> > 
>