[rfc/testsuite] i386-sse.exp: better output if no SSE support

[rfc/testsuite] i386-sse.exp: better output if no SSE support

[Michael Elizabeth Chastain]

This patch fixes i386-sse.exp for compilers that don't support SSE,
specifically gcc 2.95.3.  Before this patch, the results are:

   3 ERROR
   2 WARNING
  33 FAIL
   1 UNRESOLVED

After this patch, the results are a single UNSUPPORTED.
My view is that people should be reading all their non-PASS results,
including the UNRESOLVED, UNSUPPORTED, and UNTESTED results.
We might want to discuss this.

I tested this on native i686-pc-linux-gnu, gcc 2.95.3 and 3.3.3,
dwarf-2 and stabs+.  My processor has SSE (Intel Celeron).

I'm giving this 48 hours for comment and then committing it.

Michael C

2004-06-17  Michael Chastain  <mec.gnu@mindspring.com>

	Fix PR testsuite/1679.
	* gdb.arch/i386-sse.exp: Do not call gdb_suppress_entire file.
	Issue an UNSUPPORTED result instead.

Index: gdb.arch/i386-sse.exp
===================================================================
RCS file: /cvs/src/src/gdb/testsuite/gdb.arch/i386-sse.exp,v
retrieving revision 1.1
diff -c -3 -p -r1.1 i386-sse.exp
*** gdb.arch/i386-sse.exp	7 Jun 2004 15:38:52 -0000	1.1
--- gdb.arch/i386-sse.exp	17 Jun 2004 18:44:46 -0000
*************** set testfile "i386-sse"
*** 35,41 ****
  set srcfile ${testfile}.c
  set binfile ${objdir}/${subdir}/${testfile}
  if { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable {debug}] != "" } {
!     gdb_suppress_entire_file "Testcase compile failed, so all tests in this file will automatically fail."
  }
  
  gdb_exit
--- 35,42 ----
  set srcfile ${testfile}.c
  set binfile ${objdir}/${subdir}/${testfile}
  if { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable {debug}] != "" } {
!     unsupported "compiler does not support SSE"
!     return
  }
  
  gdb_exit

Re: [rfc/testsuite] i386-sse.exp: better output if no SSE support

[Daniel Jacobowitz]

On Thu, Jun 17, 2004 at 02:56:58PM -0400, Michael Chastain wrote:
> This patch fixes i386-sse.exp for compilers that don't support SSE,
> specifically gcc 2.95.3.  Before this patch, the results are:
> 
>    3 ERROR
>    2 WARNING
>   33 FAIL
>    1 UNRESOLVED
> 
> After this patch, the results are a single UNSUPPORTED.
> My view is that people should be reading all their non-PASS results,
> including the UNRESOLVED, UNSUPPORTED, and UNTESTED results.
> We might want to discuss this.

As far as I'm concerned this is absolutely a good change.  I don't
always carefully read the UNSUPPORTED / UNTESTED results, but I do at
least look at them.  And anything that makes gdb_suppress_entire_file
go away is good in my book.

> 	Fix PR testsuite/1679.
> 	* gdb.arch/i386-sse.exp: Do not call gdb_suppress_entire file.
> 	Issue an UNSUPPORTED result instead.

"entire_file".


-- 
Daniel Jacobowitz

Re: [rfc/testsuite] i386-sse.exp: better output if no SSE support

[Jim Blandy]

mec.gnu@mindspring.com (Michael Elizabeth Chastain) writes:
> This patch fixes i386-sse.exp for compilers that don't support SSE,
> specifically gcc 2.95.3.  Before this patch, the results are:
> 
>    3 ERROR
>    2 WARNING
>   33 FAIL
>    1 UNRESOLVED
> 
> After this patch, the results are a single UNSUPPORTED.
> My view is that people should be reading all their non-PASS results,
> including the UNRESOLVED, UNSUPPORTED, and UNTESTED results.
> We might want to discuss this.
> 
> I tested this on native i686-pc-linux-gnu, gcc 2.95.3 and 3.3.3,
> dwarf-2 and stabs+.  My processor has SSE (Intel Celeron).
> 
> I'm giving this 48 hours for comment and then committing it.

Looks good to me.  Just point me at the Kool-aid and I'll drink away.

Re: [rfc/testsuite] i386-sse.exp: better output if no SSE support

[Michael Elizabeth Chastain]

Committed.

===

2004-06-17  Michael Chastain  <mec.gnu@mindspring.com>

	Fix PR testsuite/1679.
	* gdb.arch/i386-sse.exp: Do not call gdb_suppress_entire file.
	Issue an UNSUPPORTED result instead.