Re: final i386.floating.record.patch

[paawan oza <paawan1982@yahoo.com>]

Hi,

As I understand you have tried to run the example test code which I had sent earlier.

please let me know at which c statement (or insn) it is causing this ?
So I can dig into it more.

Thanks & Regards,
Oza.


--- On Tue, 8/4/09, paawan oza <paawan1982@yahoo.com> wrote:

> From: paawan oza <paawan1982@yahoo.com>
> Subject: Re: final i386.floating.record.patch
> To: "Hui Zhu" <teawater@gmail.com>
> Cc: "Michael Snyder" <msnyder@vmware.com>, "gdb-patches@sourceware.org" <gdb-patches@sourceware.org>
> Date: Tuesday, August 4, 2009, 6:32 PM
> Hi Hui,
> I am not sure which test case (insn)  caused this.
> would you please send me the example (.c file) which you
> have tried to run ?
> please also try to let me know at which insn (at which
> point it fails).
> Regards,
> Oza.
> 
> --- On Tue, 8/4/09, Hui Zhu <teawater@gmail.com>
> wrote:
> 
> > From: Hui Zhu <teawater@gmail.com>
> > Subject: Re: final i386.floating.record.patch
> > To: "paawan oza" <paawan1982@yahoo.com>
> > Cc: "Michael Snyder" <msnyder@vmware.com>,
> "gdb-patches@sourceware.org"
> <gdb-patches@sourceware.org>
> > Date: Tuesday, August 4, 2009, 8:49 AM
> > Hi Paawan,
> > 
> > Thanks for your work.
> > 
> > I do a some test works with the test code that you
> sent in
> > before.
> > I get the values of fp reg with command "info
> > all-registers".
> > 
> > It looks like some fp reg's values not right in
> replay
> > mode.
> > 
> > In record mode they are:
> > fctrl         
> > 0x37f    895
> > fstat         
> > 0x0    0
> > ftag       
> >    0xffff    65535
> > fiseg         
> > 0x0    0
> > fioff         
> > 0x0    0
> > ---Type <return> to continue, or q
> <return> to
> > quit---
> > foseg         
> > 0x0    0
> > fooff         
> > 0x0    0
> > fop           
> > 0x0    0
> > 
> > In replay mode they are:
> > fctrl         
> > 0x37f    895
> > fstat         
> > 0x6069    24681
> > ftag       
> >    0x557f    21887
> > fiseg         
> > 0x73    115
> > fioff         
> > 0x8048bfd    134515709
> > ---Type <return> to continue, or q
> <return> to
> > quit---
> > foseg         
> > 0x7b    123
> > fooff         
> > 0x0    0
> > fop           
> > 0x2e9    745
> > 
> > 
> > 
> > Thanks,
> > Hui
> > 
> > 
> > #include <stdio.h>
> > #include <math.h>
> > #include <stdlib.h>
> > 
> > /* the test intends to test following insns.
> > flds faddp fstps fstpl fldl fxch fabs fdivrp fmulp
> fsubrp
> > fucomp fnstsw fsqrt
> > fchs f2xm1 fyl2x fxtract fprem1 fld fdecstp fld1
> fldl2t
> > fldl2e FLDPI
> > FLDLG2 FLDLN2
> > FLDZ fincstp ffree fptan fpatan fincstp fsincos
> frndint
> > fscale fsin fcos fcmovb
> > fcmovbe fcmove fcmovu fcmovnb fcmovnbe fsave frstor
> fstsw
> > */
> > 
> > float no1,no2,no3,no4,no5,no6,no7;
> > double x = 100.345, y = 25.7789;
> > long double ldx = 88888888888888888888.88, ldy =
> > 9999999999999999999.99;
> > float result,resultd,resultld;
> > float *float_memory;
> > 
> > /* initialization of floats */
> > void init_floats()
> > {
> >  no1 = 10.45;
> >  no2 = 20.77;
> >  no3 = 156.89874646;
> >  no4 = 14.56;
> >  no5 = 11.11;
> >  no6 = 66.77;
> >  no7 = 88.88;
> >  float_memory = malloc(sizeof(float) * 4);
> >  *float_memory = 256.256;
> >  *(float_memory + 1) = 356.356;
> >  *(float_memory + 2) = 456.456;
> >  *(float_memory + 3) = 556.556;
> > }
> > 
> > /* marks FPU stack as empty */
> > void empty_fpu_stack()
> > {
> >  asm ("ffree %st(1) \n\t"
> >       "ffree %st(2) \n\t"
> >       "ffree %st(3) \n\t"
> >       "ffree %st(4) \n\t"
> >       "ffree %st(5) \n\t"
> >       "ffree %st(6) \n\t"
> >       "ffree %st(7)");
> > }
> > 
> > /* tests floating point arithmatic */
> > void test_arith_floats()
> > {
> >  result = no1 + no2 + no3 + no4 + no5 + no6 +
> no7;
> >  printf("result is %f\n",result);
> > 
> >  result = fmodf(no2,no1);
> >  printf("result is %f\n",result);
> > 
> >  resultd = fmod(x,y);
> >  printf("result is %f\n",resultd);
> > 
> >  resultld = fmodl(ldy,ldy);
> >  printf("result is %f\n",resultld);
> > 
> >  result = fabsf(no1);
> >  printf("result is %f\n",result);
> > 
> >  result = no3 / no4;
> >  printf("result is %f\n",result);
> > 
> >  result = no1 * no2 * no3 * no4;
> >  printf("result is %f\n",result);
> > 
> >  result = no1 - no2 - no3 - no4;
> >  printf("result is %f\n",result);
> > 
> > 
> >  asm ("fld %0" : :"m"(*float_memory));
> >  asm ("fchs");
> > 
> >  /* test for f2xm1 */
> >  asm ("fld %0" : :"m"(*float_memory));
> >  asm ("f2xm1");
> > 
> >  asm ("fyl2x");
> > 
> >  asm ("fld %0" : :"m"(*float_memory));
> >  asm ("fxtract");
> > 
> >  asm ("fld %0" : :"m"(*float_memory));
> >  asm ("fprem1");
> > 
> >  /* decrement fpu stack pointer only status
> register should
> > get affected */
> >  asm ("fld %0" : :"m"(*float_memory));
> > 
> >  empty_fpu_stack();
> > 
> >  asm ("fld1");
> >  asm ("fldl2t");
> >  asm ("fldl2e");
> >  asm ("fldpi");
> >  asm ("fldlg2");
> >  asm ("fldln2");
> >  asm ("fldz");
> > 
> >  empty_fpu_stack();
> >  /* finishing emptying the stack */
> > 
> >  result = sqrt(no3);
> >  printf("result is %f\n",result);
> > }
> > 
> > void test_log_exp_floats()
> > {
> >  result = log10(no3);
> >  printf("result is %f\n",result);
> > 
> >  result = log(no3);
> >  printf("result is %f\n",result);
> > 
> >  result = exp10(no3);
> >  printf("result is %f\n",result);
> > 
> >  result = exp(no3);
> >  printf("result is %f\n",result);
> > }
> > 
> > void test_trigo_floats()
> > {
> >  result = sin(30);
> >  printf("result is %f\n",result);
> > 
> >  result = cos(30);
> >  printf("result is %f\n",result);
> > 
> >  result = tan(30);
> >  printf("result is %f\n",result);
> > 
> >  result = atan(30);
> >  printf("result is %f\n",result);
> > 
> >  asm ("fld %0" : :"m"(*float_memory));
> >  asm ("fptan");
> > 
> >  /* changes st1 and popping register stack */
> >  asm ("fpatan");
> > 
> >  asm("fincstp");
> >  asm ("fld %0" : :"m"(float_memory));
> >  asm ("fsincos");
> > 
> >  asm ("fld %0" : :"m"(*float_memory));
> >  asm ("frndint");
> > 
> >  asm ("fld %0" : :"m"(*float_memory));
> >  asm ("fld %0" : :"m"(*(float_memory+1)));
> >  asm ("fscale");
> > 
> >  empty_fpu_stack();
> > 
> >  asm ("fld %0" : :"m"(*float_memory));
> >  asm ("fsin");
> >  asm ("fcos");
> > 
> >  /* currently we assume condition likely and
> always record
> > the registers
> >  code could be optimized only if the flag is set
> then
> > record */
> >  asm ("fld %0" : :"m"(*float_memory));
> >  asm ("fld %0" : :"m"(*(float_memory+1)));
> >  asm ("fcmovb %st(1), %st");
> >  asm ("fcmovbe %st(1), %st");
> >  asm ("fcmove %st(1), %st");
> >  asm ("fcmovu %st(1), %st");
> >  asm ("fcmovnb %st(1), %st");
> >  asm ("fcmovnbe %st(1), %st");
> > 
> >  empty_fpu_stack();
> >  /* finished emtyping the stack */
> > }
> > 
> > void test_compare_floats()
> > {
> >    ldy = 88888888888888888888.88;
> >    if (ldx == ldy)
> >      ldy =
> > 7777777777777777777777777777.777;
> >    else
> >      ldy =
> > 666666666666666666666666666.666;
> > }
> > 
> > /* test loading and saving of FPU environment */
> > void test_fpu_env()
> > {
> >  asm ("fsave %0" : "=m"(*float_memory) : );
> >  asm ("frstor %0" : : "m"(*float_memory));
> >  asm ("fstsw %ax");
> > }
> > 
> > int main()
> > {
> >    init_floats();
> >    test_arith_floats();
> >    test_log_exp_floats();
> >    test_trigo_floats();
> >    test_compare_floats();
> >    test_fpu_env();
> > }
> > 
> > 
> > 
> > 
> > On Mon, Aug 3, 2009 at 22:56, paawan oza<paawan1982@yahoo.com>
> > wrote:
> > > Hi,
> > >
> > > please find the patch attached. I have attached
> as per
> > your suggestion.
> > > I am attaching it from opera browser.
> > >
> > > Regards,
> > > Oza.
> > >
> > > --- On Thu, 7/30/09, Michael Snyder <msnyder@vmware.com>
> > wrote:
> > >
> > >> From: Michael Snyder <msnyder@vmware.com>
> > >> Subject: Re: final  
> i386.floating.record.patch
> > >> To: "paawan oza" <paawan1982@yahoo.com>
> > >> Cc: "gdb-patches@sourceware.org"
> > <gdb-patches@sourceware.org>
> > >> Date: Thursday, July 30, 2009, 11:31 PM
> > >> paawan oza wrote:
> > >> > Hi,
> > >> >
> > >> > please find my answers below.
> > >> >
> > >> >  1) Are you using a Windows machine to
> send
> > your
> > >> >  emails?  If so,
> > >> >  is it possible that there is a
> Unix/Linux
> > >> machine you could
> > >> >  try
> > >> >  sending from?  Your attachments look
> OK
> > for
> > >> me, but
> > >> >  some people
> > >> >  seem to have had problems with them.
> > >> >
> > >> > Oza : I used to send all from windows
> till
> > now.
> > >> > but this patch I sent it from
> Linux...from
> > opera.
> > >> >   2) And are you using cut-and-paste
> to
> > >> insert the patches
> > >> >  into the
> > >> >  body of your email?  That would
> certainly
> > >> cause
> > >> >  problems, because
> > >> >  tabs might be changed into spaces
> (which
> > is
> > >> exactly what
> > >> >  was
> > >> >  causing patch to fail for me today).
> > >> >  Oza: yes I am using copy-paste....I
> dont
> > know
> > >> any-other way.
> > >> > because attachmenet are not welcome, so
> I am
> > not sure
> > >> how I could proceed with this.
> > >>
> > >> It's not that attachments are not welcome.
> > >> There are various binary encodings for
> > attachments, and
> > >> some of those binary encodings are not
> welcome. 
> > I
> > >> think
> > >> because there's no open-source way of
> decoding
> > them.
> > >>
> > >> If you look at the list, you'll see that
> > attachments are
> > >> used a lot.
> > >>
> > >> Copy-and-paste, though, in general will not
> work,
> > because
> > >> it
> > >> usually changes tabs into spaces, which ruins
> a
> > patch.
> > >>
> > >>
> > >> > I will send you updated patch.....may be
> I
> > might have
> > >> mistaken of old gdb file. sorry for
> incovenience.
> > >>
> > >> I think it was just the tabs-to-spaces
> issue.
> > >> Why don't you try sending an attachment from
> > Opera?
> > >>
> > >>
> > >
> > >
> > >
> > 
> 
> 
> 
>