[RFA] testsuite/gdb.base/pointers.[c,exp]: don't assume >= 8 byte doubles

[RFA] testsuite/gdb.base/pointers.[c,exp]: don't assume >= 8 byte doubles

[Orjan Friberg]

(This is a resubmission.)  In pointers.c, a double D is assigned
-1.375E-123.  This only fits if a double is (at least) 8 bytes.  For a
4-byte double you get an underflow.  The pattern matching of the
decimals when printing the value of D also assumes it's >= 8 bytes.  K&R
says an acceptable minimum for FLT_MIN is 1E-37, so I decided to go with
that and relaxed the decimal pattern matching to the same level as for
floats.

Ok to commit?

Orjan


2001-02-26  Orjan Friberg  <orjanf@axis.com>

	* gdb.base/pointers.c: Don't assume doubles are >= 8 bytes.
	* gdb.base/pointers.exp: Relax pattern match of decimals.


*** /home/orjanf/gdb-HEAD/src/gdb/testsuite/gdb.base/pointers.c Wed Mar
22 00:01
:49 2000
--- pointers.c  Wed Nov  1 14:28:52 2000
*************** int more_code()
*** 194,200 ****
      L = -234;
      UL = 234;
      F = 1.25E10;
!     D = -1.375E-123;
      pC = &C;
      ppC = &pC;
      pppC = &ppC;
--- 194,200 ----
      L = -234;
      UL = 234;
      F = 1.25E10;
!     D = -1.25E-37;
      pC = &C;
      ppC = &pC;
      pppC = &ppC; 


*** /home/orjanf/gdb-HEAD/src/gdb/testsuite/gdb.base/pointers.exp      
Mon Feb 
26 18:34:53 2001
--- pointers.exp        Mon Feb 26 18:34:36 2001
*************** gdb_expect {
*** 527,533 ****
  
  send_gdb "print *pD\n"
  gdb_expect {
!     -re ".\[0-9\]* = -1.375e-123.*$gdb_prompt $" {
          pass "print value of *pD"
        }
      -re ".*$gdb_prompt $" { fail "print value of *pD" }
--- 527,533 ----
  
  send_gdb "print *pD\n"
  gdb_expect {
!     -re ".\[0-9\]* = -1.2\[0-9\]*e\\-37.*$gdb_prompt $" {
          pass "print value of *pD"
        }
      -re ".*$gdb_prompt $" { fail "print value of *pD" }



-- 
Orjan Friberg              E-mail: orjan.friberg@axis.com
Axis Communications AB     Phone:  +46 46 272 17 68

Re: [RFA] testsuite/gdb.base/pointers.[c,exp]: don't assume >= 8 byte doubles

[Fernando Nasser]

Orjan Friberg wrote:
> 
> (This is a resubmission.)  In pointers.c, a double D is assigned
> -1.375E-123.  This only fits if a double is (at least) 8 bytes.  For a
> 4-byte double you get an underflow.  The pattern matching of the
> decimals when printing the value of D also assumes it's >= 8 bytes.  K&R
> says an acceptable minimum for FLT_MIN is 1E-37, so I decided to go with
> that and relaxed the decimal pattern matching to the same level as for
> floats.
> 
> Ok to commit?
> 

Yes, I believe we discussed this long ago and all agreed that it was a sensible change.

Thanks.
Fernando


> Orjan
> 
> 2001-02-26  Orjan Friberg  <orjanf@axis.com>
> 
>         * gdb.base/pointers.c: Don't assume doubles are >= 8 bytes.
>         * gdb.base/pointers.exp: Relax pattern match of decimals.
> 
> *** /home/orjanf/gdb-HEAD/src/gdb/testsuite/gdb.base/pointers.c Wed Mar
> 22 00:01
> :49 2000
> --- pointers.c  Wed Nov  1 14:28:52 2000
> *************** int more_code()
> *** 194,200 ****
>       L = -234;
>       UL = 234;
>       F = 1.25E10;
> !     D = -1.375E-123;
>       pC = &C;
>       ppC = &pC;
>       pppC = &ppC;
> --- 194,200 ----
>       L = -234;
>       UL = 234;
>       F = 1.25E10;
> !     D = -1.25E-37;
>       pC = &C;
>       ppC = &pC;
>       pppC = &ppC;
> 
> *** /home/orjanf/gdb-HEAD/src/gdb/testsuite/gdb.base/pointers.exp
> Mon Feb
> 26 18:34:53 2001
> --- pointers.exp        Mon Feb 26 18:34:36 2001
> *************** gdb_expect {
> *** 527,533 ****
> 
>   send_gdb "print *pD\n"
>   gdb_expect {
> !     -re ".\[0-9\]* = -1.375e-123.*$gdb_prompt $" {
>           pass "print value of *pD"
>         }
>       -re ".*$gdb_prompt $" { fail "print value of *pD" }
> --- 527,533 ----
> 
>   send_gdb "print *pD\n"
>   gdb_expect {
> !     -re ".\[0-9\]* = -1.2\[0-9\]*e\\-37.*$gdb_prompt $" {
>           pass "print value of *pD"
>         }
>       -re ".*$gdb_prompt $" { fail "print value of *pD" }
> 
> --
> Orjan Friberg              E-mail: orjan.friberg@axis.com
> Axis Communications AB     Phone:  +46 46 272 17 68

-- 
Fernando Nasser
Red Hat - Toronto                       E-Mail:  fnasser@redhat.com
2323 Yonge Street, Suite #300
Toronto, Ontario   M4P 2C9

Re: [RFA] testsuite/gdb.base/pointers.[c,exp]: don't assume >= 8 byte doubles

[Orjan Friberg]

Fernando Nasser wrote:
> 
> Orjan Friberg wrote:
> >
> > (This is a resubmission.)  In pointers.c, a double D is assigned
> > -1.375E-123.  This only fits if a double is (at least) 8 bytes.  For a
> > 4-byte double you get an underflow.  The pattern matching of the
> > decimals when printing the value of D also assumes it's >= 8 bytes.  K&R
> > says an acceptable minimum for FLT_MIN is 1E-37, so I decided to go with
> > that and relaxed the decimal pattern matching to the same level as for
> > floats.
> >
> > Ok to commit?
> >
> 
> Yes, I believe we discussed this long ago and all agreed that it was a sensible change.
> 
> Thanks.
> Fernando
> 

Committed.  Thanks to Michael for reviewing and testing the patch.  

BTW, the cvs commit messages for pointers.c and pointers.exp
accidentally got the 2001-02-26 date (which was when I submitted the
patch).  I updated the date in the ChangeLog entry (and ChangeLog's cvs
commit message) though.  Sorry about that.

-- 
Orjan Friberg              E-mail: orjan.friberg@axis.com
Axis Communications AB     Phone:  +46 46 272 17 68

Re: [RFA] testsuite/gdb.base/pointers.[c,exp]: don't assume >= 8 byte doubles

[Michael Elizabeth Chastain]

Hi Orjan,

I proofread this and tested it on a Red Hat 7 Linux native system.
It works for me.

> K&R says an acceptable minimum for FLT_MIN is 1E-37, so I decided to go
> with that and relaxed the decimal pattern matching to the same level as
> for floats.

ISO C standard section 5.2.4.2, "numerical limits", addresses this.
It says: "minimum normalized positive floating-point number:"

  FLT_MIN  1e-37
  DBL_MIN  1e-37
  LDBL_MIN 1e-37

So I agree; it is legal for an implementation to implement
doubles as 32-bit IEEE 754 numbers.

> Ok to commit?

I recommend this patch for approval but I do not have approval power.
Fernando?  Stan?

Michael